An Overview: Peak To Average Power Ratio (PAPR) in OFDM System Using Some New PAPR Techniques (With Matlab Code)

An Overview : Peak to Average
Power Ratio (PAPR) in OFDM

system using some new PAPR
techniques (with matlab code)
Zainab S. H. AL-Hashmi
An Overview : Peak to Average

Power Ratio (PAPR) in OFDM
system using some new PAPR
techniques (with matlab code)
By
Zainab Saad Hadi AL-Hashmi
A graduate of University of
Baghdad, College of Engineering
Electronic & Communications
Engineering Department

{ }/

.

.

.

ii
Acknowledgments
praise belongs to God who showed favour to us through His
religion, singled us out for His creed, and directed us onto the
roads of His beneficence, in order that through His kindness we
might travel upon them to His good pleasure, a praise which He
will accept from us and through which He will be pleased with
us. !Allah send peace and blessings upon Mohammed and his
progeny (S.A.W.)
Finally I would like to thank my family,
Especially my grandfather Mr. Hassan Ali Zwain,

my mother, Mr. Qasim Hassan Zwain and Mr. Maythem Hassan
Zainab saad hadi

2015
iii
Abstract
The Orthogonal frequency division multiplexing (OFDM) is multicarrier
modulation scheme which has recently become comparatively popular in
both wireless and wired communication systems for transfer the
multimedia data. OFDM could be used at the core of well-known systems
like Asymmetric digital subscriber line (ADSL) internet, digital
television/radio broadcasting, wireless local area network (LANs), and
Long Term Evolution (LTE).
High PAPR is the major drawback of OFDM, which results in lower
power efficiency hence impedes in implementing OFDM. The PAPR
problem is more significant in the uplink because the efficiency of power
amplifier is critical because a mobile terminal has a limited battery
power.
High peak-to-average power ratio (PAPR) occurs due to large envelope
fluctuations in OFDM signal this requires a highly linear high power
amplifier (HPA). Power amplifiers with large linear range are expensive,
bulky 50% of the size of a transmitter lies and difficult to manufacture.
In order to reduce the PAPR, several techniques have been proposed in
this thesis, primarily the repeated frequency domain filtering and clipping
(RFC) has been proposed and compared with the existing method
repeated clipping and frequency domain filtering (RCF). The RFC is
better than RCF in performance especially when I 2, although they have
the same complexity and cost.
The proposed method is not only improving PAPR but also improving
BER. Best case for the bit error rate (BER) is at I =4 and CR =4, where
) improved by (5.7601 dB)
Signal to Noise Ratio (SNR) at BER (
and Complementary Cumulative Distribution Function (CCDF) of PAPR
was improved by (4.775 dB) and PAPR was improved by (11.4177 dB).
The best one improvement in PAPR and CCDF of PAPR So as not to
BER deteriorate is at I =4 and CR =1.75. The improvement in PAPR by =
(18.2789 dB), CCDF of PAPR = (8.0187 dB), and the SNR at
) by = (0.6101 dB).
BER(
In addition to (RFC) six new types of companding have been proposed
and compared with the -law and A-law compandings. all these proposed
methods have better performance than the -law and A-law compandings,
and the best one is Absolute Exponential (AEXP) companding and the
iv
best one improvement in PAPR and CCDF of PAPR is at d= 1.1. The

improvement in PAPR by = (17.6492 dB), and CCDF of PAPR = (7.2405
) deteriorated by = (-3.4186 dB).
dB), while the SNR at BER(
Five types of pre-coding are used in this work and then compared them
with each other. The best type of precoding in term of reduced PAPR and
BER is the Discrete Fourier Transform (DFT) pre-coder, while the least is
the Walsh Hadamard Transform (WHT) pre-coding.
Also four new types of hybrids PAPR reduction techniques have been
proposed. These methods are:
1. RCF with precodings (WHT, Discrete Cosine Transform (DCT),
Discrete Sine Transform (DST),and Discrete Hartley Transform (DHT)).
2. RCF with compandings (the all proposed compandings, -law and Alaw compandings).
3. RFC with compandings (the all proposed compandings, -law and Alaw compandings).
4.and finally precodings (WHT, DCT, DST,and DHT), with compandings
(the all proposed compandings, -law and A-law compandings).
The best one improvement is at (RFC with AEXP) because the PAPR,
CCDF of PAPR, and BER. This improvement in PAPR and CCDF of
PAPR is at d = 0.6 and CR =4. The improvement in PAPR by
(21.0509dB), CCDF of PAPR = (8.7178 dB), and the SNR at
) by (0.0116 dB).
BER(
The DHT with tangent Rooting (tanhR) have acceptable results where the
PAPR and CCDF of PAPR were improved while BER was degarded.
The best one improvement for this case is at k=15, y=.8 and DHT. The
improvement in PAPR by = (22.7711 dB), and CCDF of PAPR = (8.9691
All methods are simulated using matlab.
Contents
Chapter One: Introduction
1.1 Introduction
1.2 Literature survey
Chapter Two: LTE and OFAM
2.1. Introduction
2.2. LTE Requirements
2.3. LTE Architecture
2.4. Air interface in LTE
2.5 History of OFDM
10
2.6 OFDM
12
2.6.1 Orthogonality of the subcarriers and OFDM
15
2.6.2 Guard Interval
17
2.6.3 One-tap equalizer
18
2.7 OFDM based Multiple Access
19
2.8 Orthogonal Frequency Division Multiple Access
20
2.9 SC-FDMA
24
Chapter Three: Peak To Average Power Ratio Reduction
27
3.1Definitions of PAPR
27
3.2 PAPR of OFDM signal
28
3.3 Oversampling discrete OFDM symbols to find true (continuous) peaks
29
3.4 Distribution of PAPR
29
3.5 Identification of the Problem
32
3.5.1 Nonlinear HPA and DAC
32
3.5.2 Power Saving
35
3.6 Factors influencing the PAPR
35
3.6.1 The number of sub carriers
35
3.6.2 The order of Modulation
35
3.6.3 Constellation shape
36
3.6.4 Pulse Shaping
36
vi
3.7 The gauge for judgment of the PAPR reduction in OFDM systems
36
3.8 Fitness function-based approach for determining an appropriate Algorithm
37
Chapter Four: PAPR Reduction Techniques
39
4.1There are three different way to divide the PAPR
39
4.1.1The first way is
39
4.1.2 The second way
43
4.1.3The third way
45
4.1.4 And finally there is Hybrid techniques
45
4.2 Clipping and Filtering
46
4.3 Peak Windowing Method
47
4.4 Envelope Scaling
48
4.5 Peak Reduction Carrier
48
4.6 Companding Technique
49
4.6.1 Square-Rooting Companding Technique ( SQRT) for PAPR Reduction in
50
OFDM Systems
4.6.2 Exponential Companding Algorithm
51
4.6.3 Trapezoidal power companding
53
4.6.4 Hyperbolic tangent (
53
4.6.5 Error Function (
) companding
54
) Companding
4.6.6 Logarithm Function (log) Companding
54
4.7 Coding techniques
54
4.8 Selective Mapping (SLM)
56
4.9 Partial Transmit Sequence (PTS)
57
4.10 Tone Reservation
59
4.11 Tone Injection
60
4.12 Interleaving
61
4.13 Active Constellation Extension (ACE)
61
4.14 Dummy Sequence Insertion (DSI)
62
Chapter Five: Simulation Results and Analysis
63
5.1 OFDM System model
63
5.2 PAPR techniques used
65
vii
5.2.1 Repeated clipping and frequency domain filtering (RCF)
65
5.2.2 Repeated frequency domain filtering and clipping RFC
72
5.2.3 The OFDM System with discrete time companding
77
5.2.3.1 A companding
77
5.2.3.2
81
Companding
5.2.3.3 Rooting Companding Technique (RCT)
85
5.2.3.4 New error function Companding (NERF)
87
5.2.3.5 Absolute Exponential companding (AEXP)
89
5.2.3.6 Cos companding
91
5.2.3.7 tangent Rooting Companding (tanhR)
95
5.2.3.8 Logarithmic Rooting Companding (logR)
101
5.2.4 OFDM System with pre-coding
104
5.2.4.1 Pulse Shaping or Pre-coding
104
5.2.4.2 Discrete Hartley transform (DHT)
105
5.2.4.3 Walsh-Hadamard Transform (WHT)
105
5.2.4.4 Discrete Cosine Transform (DCT):
106
5.2.4.5 Discrete Sine Transform (DST) Precoding Technique
107
5.2.4.6 The Discrete Fourier Transform (DFT) Precoding
107
5.2.4.7 Simulation results and analysis of OFDM system with pre-coding
108
Chapter six: Simulation Results and Analysis of Hybrid PAPR techniques
110
6.1 Hybrid pre-coding with RCF
110
6.2 Hybrids RCF with companding
119
6.2.1 RCF + A companding
119
6.2.2 RCF +
121
6.2.3 RCF + RCT
123
6.2.4 RCF + AEXP
126
6.2.5 RCF + cos
128
6.2.6 RCF + NERF
130
6.2.7 RCF + tanhR
131
6.2.8 RCF +logR
132
6.3 Hybrid RFC with companding
134
6.3.1 RFC + A companding
134
viii
6.3.2 RFC +
137
companding
6.3.3 RFC + RCT
139
6.3.4 RFC + AEXP
141
6.3.5 RFC + cos
143
6.3.6 RFC + NERF
145
6.3.7 RFC + tanhR
146
6.3.8 RFC +logR
147
6.4 Pre-coding + companding
148
6.4.1 Pre-coding + A companding
149
6.4.2 Pre-coding +
152
6.4.3 Pre-coding + RCT
154
6.4.4 Pre-coding + AEXP
156
6.4.5 Pre-coding + cos
159
6.4.6 Pre-coding + tanhR
161
6.4.7 Pre-coding + logR
162
6.4.8 Pre-coding + NERF
163
Chapter seven : Conclusions and future work
165
7.1Conclusions
165
7.2Future work
167
References
168
Appendices
Appendix A : Table of Results
A.1
Appendix B : MATLAB Code
B.1
ix
Chapter One
Introduction
Chapter One
Introduction
1.1 Introduction:
During the last two decades, the demand for multimedia wireless communication
services have grown tremendously and this trend are expected to continue in the near
future. Orthogonal frequency division multiplexing (OFDM) is one of such multicarrier techniques which have attracted vast research attention from academics,
researchers and industries since last two decades. It has become part of new emerging
standards for broadband wireless access [1].
Energy efficiency, particularly matters in future mobile communications networks. A
key driving factor is the growing energy cost of network operation which can make up
as much as 50% of the total operational cost nowadays [2].
The transmitted signal of OFDM exhibits a high Peak-To-Average Power Ratio
(PAPR). This high PAPR reduces the efficiency of high power amplifier and degrades
the performance of the system [3].
A major source for reducing energy costs is to increase the efficiency of the high
power amplifier (HPA) in the radio frequency (RF) front end of the base stations [4].
However, the efficiency of the HPA is directly related to the PAPR of the input signal.
The problem, especially, becomes serious in OFDM multicarrier transmission, which
is applied in many important wireless standards such as the third Generation
Partnership Project (3GPP) Long Term Evolution Advanced (LTE-A). The PAPR
problem still prevents OFDM from being adopted in the uplink of mobile
communication standards, and, besides from power efficiency, it can also place severe
constraints on output power and therefore coverage in the downlink. In the past, there
have been many efforts to deal with the PAPR problem resulting in numerous papers
and several overview articles, e.g., [5], [6], [7].
PAPR has a deleterious effect on battery lifetime in mobile applications. As handy
devices have a finite battery life, it is significant to find ways of reducing the PAPR
allowing for a smaller, more efficient HPA, which in turn will mean a longer lasting
battery life.
In many low-cost applications, the problem of high PAPR may outweigh all the
potential benefits of multicarrier transmission systems [6]. A number of promising
approaches or techniques have been proposed & implemented to reduce PAPR with
the expense of increase transmit signal Power, bit error rate (BER) & computational
complexity and loss of data rate, etc. So, a system trade-off is required [8].
1.2 Literature survey:

In 1996 Robert [9]. The selected mapping was used for the reduction of PAR. The
selected mapping can be used for arbitrary numbers of carriers and any signal
constellation. The selected mapping provides significant gains at moderate additional
complexity. Actually, it is appropriate for all kinds of multiplex techniques, which
transform data symbols to the transmit signal. Even in single carrier systems where
1
Chapter One
Introduction
PAR grows as the roll of factor of the pulse shaping filter decreases, selected mapping
can be applied advantageously.
The first nonlinear companding transform (NCT) for PAPR reduction was given by
Wang et.al in 1999 [10]. It was based on the speech processing algorithm -law and it
has found better performance than that of clipping technique. The -law companding
transform mainly focuses on enlarging small amplitude signals while keeping peak
signals unchanged, and thus it increases the average power of the transmitted signals
and may lead to overcome the saturation region of the HPA to make the performance
of the system worse. In order to overcome the problem of -law companding
(increasing average power) and to have an efficient PAPR reduction. [10]
In 2000 Myonghee et.al [11] Hadamard transform is an effective technique to reduce
the PAPR of an OFDM system. The PAPR can be reduced in OFDM system without
any power increase and side information. The increase of system complexity is not
much. As further study, the equalization problem combining with Hadamard
transform, which is induced to reduce PAPR, over multipath fading channel, is
considered.
In 2001 J. Armstrong [12] the clipping and frequency domain filtering PAPR
reduction technique has been described in which an interpolated version of the
baseband signal is clipped and then filtered with a new form of filter. The filter
consists of a forward and an inverse fast Fourier transform (IFFT). It is designed to
remove the out-of-band (OOB) noise without distorting the in-band discrete signal. It
is shown that significant PAPR reduction can be achieved without any increase in
OOB power. Some in-band distortion results, but this will have negligible effect on
the overall BER in most systems.
In 2002 J. Armstrong [13] the repeated clipping and frequency domain filtering of an
OFDM signal can significantly reduce the PAPR of the transmitted signal. This
method causes any increase in OOB power. Considerable PAPR reduction can be
obtained with only moderate levels of clipping noise.
In 2004 Ryu, et al. [14] The Dummy Sequence Insertion (DSI) technique reduces
PAPR through increased the average power of the signal. Herein, after switchting the
input data stream into parallel through the serial to parallel converter a, dummy
sequence is inserted in the input signal. Thus, the average value is raised and the
PAPR is reduced later.
In 2005 Tao Jiang et.al [15] exponential companding. It can adjust the amplitudes
of both large and small input signals, while maintaining the average power unchanged
by properly choosing transform parameters, so as to make the output signals have a
uniform distribution (with a specific degree). The exponential companding schemes
can efficiently reduce PAPR for various modulation formats and sub-carrier sizes.
2
Chapter One
Introduction
The exponential companding schemes make less spectrum side-lobes than -law
companding. Simulation results have shown that exponential companding schemes
could provide better system performance in terms of PAPR reduction, power
spectrum, BER, and phase error than the -law companding scheme.
In 2007 Wisam et.al [16] square rooting companding (SQRT) companding a simple
method of reducing the PAPR value of OFDM symbol by changing the statistical
characteristics of the output signals . This was achieved by applying a non-linear
square rooting operation of the OFDM signals. The process changed also the
describing parameters of power signals: average and peak power values, and as a
result the PAPR value is reduced. This companding more suitable for OFDM
applications that do not have sophisticated processor, since it allows significant
reduction in PAPR value with very low cost of computational complexity, and only
slight performance degradation.
In 2008 Pisit et.al [17] the simple PAPR reduction method by using the dummy subcarriers. The features of proposed method is to decide the frequency data for dummy
subcarriers theoretically by using the certain number of larger amplitude levels
detected in the time domain signal and to achieve the better PAPR performance with
less computational complexity.
In 2008 Carole et.al [18] they present an incipient PAPR reduction technique which
exploits the utilization of used carriers in addition to the phase information of pilot
signals in OFDM systems to limit the PAPR while not degrading channel estimation
or frequency offset. Compared to conventional techniques like clipping and
windowing, this technique introduces significantly lower OOB distortions and
provides a lower BER since there is no interference to the original data signals. There
is additionally no requisite for side information to be transmitted to the receiver.
In 2009 Kazuki and Fumiyuki [19] A tone injection (TI) has been suggested which
exploits the property of a nonlinear modulo function. The TI is identically equivalent
to the one that superimposes a quadrature amplitude modulation (QAM) signal on the
data symbol to reduce the PAPR. Without the transmission of the side information,
the TI dramatically reduces the PAPR level. Albeit the TI-OFDM reduces the 1%
PAPR level by about 3~4.5dB, the BER performance remarkably degrades. However,
the utilization of antenna diversity reception can reduce the BER performance
degradation.
In 2010 Zhongpeng et.al [20] a combined companding transform and hadamard
transform technique is suggested to reduce PAPR of OFDM signal .Simulation results
shows that the PAPR reduction performance is improved compared with companding
transform used only. On the other hand, the BER of system using proposed PAPR
reduction scheme is not degraded.
3
Chapter One
Introduction
In 2010 Imran and Varun [21] the PAPR of discrete hartley transform (DHT)Precoded OFDM system for M-ary Quadrature Amplitude Modulation (M-QAM)
(where M=16, 32, 64, 256). The Matlab simulation shows that DHT-Precoded OFDM
System shows better PAPR gain as compared to OFDM-Original system, Walsh
Hadamard transformation (WHT)-Precoder Based OFDM system and selective
mapping (SLM) OFDM (with V=2) system respectively. Thus, it is concluded that
DHT Precoder Based OFDM System shows better PAPR reduction then WHTPrecoder Based OFDM System, SLM-OFDM System and OFDM-Original system for
MQAM. Additionally, the DHT-Precoded OFDM system does not require any power
increase, complex optimization and side information to be sent for the receiver.
In 2011 Zhongpeng [22] a combined reduction in PAPR of the
OFDM signals based on the combination of the discrete cosine transform (DCT) with
companding. While taking both BER performance and PAPR performance into
account, a united DCT and companding scheme to reduce the PAPR of OFDM
signals.
In 2011 Hem [23] a combinational method of pre-coding and clipping is proposed to
reduce the PAPR of an OFDM system. The proposed technique is better than
conventional because it does not require any increase in roll-off factor to reduce
PAPR. Thus, it reduces the overhead in comparison to conventional pre-coding
method. Increasing the roll-off factor degrades the BER as given in [24]. The clipping
after pre-coding reduces PAPR but degrades BER. However, this degradation in not
significant in comparison to degradation obtained by increasing roll off factor.
In 2012 Malhar and Prof.Abhishek [25] tone reservation includes no of set of
reservation of tones. By using this technique reserved tones can be utilized to
minimize the PAPR. This method is used for multicarrier transmission and also
demonstrated the reserving tones to limit the PAPR. Advantage of this tone
reservation is very positive that no process is needed at receiver end. Furthermore
there is no need to transmit the side information combined with the transmitted signal.
In 2012 Eugen [26] The PAPR reduction technique based on combination of a WHT
and a new signal companding function. The numerical results show that the hybrid
scheme realizes an improved PAPR reduction than the component methods. The
computation complexity increases linearly with the number of considered signal
variants because of several signal variants are applied to the precoding block. This
problem can be solved, by using few subcarriers as markers.
In 2012 Chau, and Hsuan [27] presents a combination scheme, which using a
combination of precoding by utilizing least null subcarriers in the frequency domain
and nonlinear companding technique by applying proper -Law characteristic in time
domain, for reducing PAPR. Simulation results indicate that the proposed scheme
4
Chapter One
Introduction
achieves a advantageous trade-off between OOB power emission in OFDM systems

and the reduction of PAPR.
In 2013 Sroy et.al [28] an Iterative Clipping and Filtering (ICF) Technique for PAPR
Reduction of OFDM Signals: System Using DCT/ inverse discrete cosine transform
(IDCT) Transform. The OFDM symbol enters the ICF block with DCT/IDCT
technique, then clipping and filtering is iteratively performed. Although we
demonstrate that significant PAPR reduction is obtained through Iterative clipping
and filtering using fast Fourier transform (FFT)/IFFT transform, but better results are
observed applying DCT/IDCT in the classical ICF technique.
In 2013 Zihao et.al [29] a trapezoidal power companding method which could
significantly reduce the PAPR for a complex OFDM or Filterbank Based Multicarrier
Transmission (FBMC) system. The proposed approach provides a convenient way for
designing a compander where the trade-offs among several system performances
(such as PAPR, power spectral density (PSD) and BER) can be made.
In 2013 Mohit et.al [30] the performance of tanh and erf companding is
approximately. Log companding is better than the hyperbolic tangent and error
function companding . -law and A-law companding give the same performance and
the -law and A-law companding is better than the tanh, log and erf companding.
Some more non-linear transform have been suggested in the paper [31, 32, 33, 34, and
35]
In 2013 Jaspreet et.al [36] the performance analyzed in terms of PAPR in Orthogonal
Frequency Division Multiple Access (OFDMA) by utilizing some pre-coding
techniques, called Zadoff-Chu Transform (ZCT) and WHT with the -law
companding to limit the PAPR of the OFDM signals .These pre-coding techniques
produced the lower PAPR as compared to the conventional OFDM system.
Furthermore ZCT is better than WHT because it produced the lowest PAPR than
WHT. -law companding further reduces PAPR of OFDM signal and as with
increasing the value the PAPR reduces. The obtained results approved that the
proposed method have gotten better results than conventional OFDM.
In 2013 Navneet and Lavish [37] The PAPR reduction method is based on combining
clipping with WHT. Combined technique is simple to implement and has no
limitations on the system parameters such as number of subcarriers modulation order,
and constellation type. This system produces the lowest PAPR and is efficient, signal
independent, distortion less and do not require any complex optimizations
representing better PAPR reduction methods than others existing techniques because
it does not require any power increment, complex optimization and side information
to be sent to the receiver.
Chapter One
Introduction
In 2013 Mohit et.al [38] To reduce the PAPR of OFDM has been proposed Hybrid
Clipping-Companding techniques for PAPR Reduction. the performance of hybrid
PAPR reduction scheme with either tanh or erf as companding function is
approximately same .Hybrid PAPR reduction scheme with log companding function
is better than the last two. Hybrid PAPR reduction scheme with either -law or A-law
companding gives the same performance and the Hybrid PAPR reduction scheme
with either -law or A-law companding is best.
In 2013 K. muralibabu et.al [39] In the proposed scheme, a combined reduction in
PAPR of the OFDM system by grouping the sub carrier on the basis of the
combination of joining the Discrete Cosine Transform (DCT) with companding
technique. The simulation results indicat that the proposed scheme can yield good
tradeoff between computational complexity and PAPR reduction performance
In 2014 Jijina et.al [40] a comparative study is made on the three typical linear
precoding techniques: Hadamard transform precoding, Discrete Sine Transform
(DST) precoding and Square root raised cosine function precoding used in OFDMA
system. The performance of these different schemes in terms of PAPR reduction is
analyzed with the conventional Random Interleaved OFDMA system. Linear
precoding schemes are efficient, signal independent, distortion less and do not require
complex optimization when compared to the other reduction schemes.
Chapter Two
LTE and OFAM

Chapter Two
LTE and OFAM
2.1. Introduction:
The growth in data intensive mobile services and applications like Web browsing,
social networking, video streaming and music has become a driving force for
development of the next generation of wireless standards. Thus, new standards are
being developed to provide the data rates and network capacity needful to support
worldwide delivery of these kinds of rich multimedia application. LTE have been
developed to respond to the requirements of this generation and to achieve the aim of
realizing global broadband mobile communications [41].
2.2. LTE Requirements:

The demand for high speed and widespread network access in mobile
communications increases every day as the number of users increases and
applications are constantly developed with greater demand for network resources. As
a result of this trend, mobile communications has experienced significant
developments within the last two decades, which is the result of tremendous research
that has been carried out. [42]
Requirements and objectives for the LTE Discuss the main requirements for the new
LTE system Resulted in a the creation of a formal
Study Item in 3GPP with the specific aim of evolving the 3GPP radio access
technology to guarantee competitiveness over a ten-year time-frame. Depending on
the study of this Study Item, the requirements for LTE Release 8 were revised and
crystallized. They can be summed up as follows [41,43, and 44]:
High peak data rates and diminished delays, in both connection establishment
and transmission latency. These improvements are to be realized through the
simplification of the overall system, the decrease of complexity and the
automated process of system management (i.e. optimization).
greater flexibility of spectrum usage, in each of the new and pre-existing bands;
Seamless
integration
with
existing
systems
(Universal
Mobile
Telecommunications System (UMTS), Wireless Fidelity (Wi-Fi), etc.).
Infrastructure-building economy. Although the implementation of every new
system brings construction and building costs, LTE should be realized through
minimal investment and use as much of the existing mobile communication
infrastructure as possible.
Multi-antenna support.
Improved system capacity and coverage
Reasonable power consumption for the mobile terminal. The mobile terminal is
being associated with mobile phones and similar devices which have limited
battery capacities. Therefore a flexible bandwidth system (with lower
frequencies used for uplink transmission) and automated signal power-level
optimization have to be included into LTE [45].
Seamless mobility, including between different radio-access technologies;
Simplified network architecture;
Increased cell-edge bit-rate, for unification of service provision;
Increased user data rates;
Reduced cost per bit, implying an enhanced spectral efficiency;
7
Chapter Two
LTE and OFAM
Packet switched domain utilization. To eliminate additional system complexity,

introduced through the support of both the circuit switched and packet switched
domain, the circuit switched domain will not be included into the LTE system.
The traditional voice and text messaging services must be replaced with systemexternal subsystems (e.g. Information Management System (IMS)).
High-level security and mobility. As the mobile communication system is now
similar to a data network (e.g. internet), additional emphasis will be set on new
security measures in combination with IP (Internet Protocol)-security functions.
Mobility efficiency is provided through the use of evolved base stations, i.e.
eNodeBs (E-UTRAN Node-B or Evolved Node-B).
These main targets resulted in the creation of additional requirements and spin-off
functionalities, whose realizations were researched, developed and evolved by 3GPP
and hence introduced in LTEs specifications and standardization upgrades.
These improvements were further evolved and enhanced in Release 9, which
contained additional techniques, functionalities and technology approaches to enable a
quick, efficient and low-cost implementation of the LTE system. The following
techniques are included:
introduction to Self-Organizing Networks (SON),
improved approach to emergency calls, as they oppose the systems security
policy,
multiple-eNodeB broadcast signal combination (LTE MBMS),
further improvement of Frequency Division Duplex (LTE-FDD) and Time
Division Duplex (LTE-TDD),
improvement of SON technologies and mechanisms, and
Minimization of system drive-tests (MDT).
The LTE system and its standardization are 3GPPs most significant milestone
achieved so far, triggering an increase of participation in their further projects and
worldwide acknowledgement of their existing work. Takahiro Nakamura, the 3GPP
RAN Chairman, states: Operators need to work on issues that have been created in
signaling and the volume of data being carried. Therefore, further improvements to
the 3GPP system are being driven by that data explosion. A continued evolution of
the system is given in Releases 10, 11 and 12, introducing an improved mobile
communication standard named LTE-Advanced [45].
2.3. LTE Architecture:

The LTE architecture was highly simplified and flattened, as shown in Figure 2.1. The
system contains only two types of nodes named Mobility Management Entity/System
Architecture Evolution Gateway (MME/SAE GW) and evolved Node-B (eNB) [46,
47].
All LTE network interfaces are based on IP protocols and therefore two major
changes were made compared to previous cellular radio architectures. The first
significant modify is that the Radio Network Controller (RNC) is removed from the
data path and its functions are now situated in eNB [46]. The main benefits of this
type of single node access network are the diminished latency and the distribution of
the RNC processing overhead into multiple eNBs. The second major change is that
there are no nodes for Circuit Switched (CS) domain, such as the Mobile Switching
Chapter Two
LTE and OFAM
Centre (MSC). Therefore speech services are handled as Voice over IP (VoIP) calls in
the LTE network [47, 48].
The eNBs are connected to each other via X2 interface and to Evolved Packet Core
(EPC) through S1 interface, as also shown in Figure 2.1. The S1 interface supports in
addition many-to-many relations between MMEs / SAE Gateways and eNBs [46].
SAE Gateway contains two logical gateway entities named as the Serving Gateway
(SGW) and the Packet Data Network Gateway (P-GW). The S-GW is responsible for
receiving and forwarding IP packets. Therefore, it can be seen as a local mobility
anchor to the eNBs [48]. The P-GW, on the other hand, is responsible for handling the
internet protocol functions, like routing, packet filtering, policy enforcement and
address allocation [47].
The new system architecture was designed so that it will reduce the overhead from
increased traffic. This is achieved because only the MME is responsible for signaling
and therefore the user IP packets do not go over MME. This way the network capacity
stays on a good level as the signaling and the traffic can grow separately [49]. The
main responsibilities of MME are idle-mode User Equipment (UE) reachability
including the control and execution of paging retransmission, different type of
authentication procedures with Non-Access Stratum (NAS) signaling, roaming, PGW/S-GW selection, tracking area list management and bearer management
including dedicated bearer establishment [47,48].
2.4. Air interface in LTE:

The air interface and communication environment used in LTE mobile
communication systems is called the LTE Radio Access Network. [45]
The LTE air interface is based on OFDMA for the downlink. OFDMA is an extension
of OFDM for the implementation of a multi-user communication system. For the
uplink, a single-carrier frequency-division multiple access (SC-FDMA) technique has
been selected. Advantages of this method include the relatively low adjacent channel
power, even if the power amplifier is not 100% linear. With SC-FDMA, no exacting
requirements are imposed on the linearity of the power amplifier in the mobile
handset. As a result, power consumption can be kept within limits. [50]
The utilization of OFDM provides considerable advantages over alternative multipleaccess techniques and signals severe departure from the past. Among the benefits are
adaptability for broadband data transmission and high spectral efficiency, impedance
to Inter Symbol Interference (ISI) resulting from the multipath fading, naturally
provide MIMO (Multiple Input Multiple Output) schemes, and provide frequencydomain techniques like frequency-selective scheduling [51].
The design of the time-frequency representation of OFDM to provide high levels of
flexibility in allocation of each of the time frames for transmission and the spectra.
The spectrum flexibility in LTE supports not only a scalable set of bandwidths, but
also a variety of frequency bands. LTE also supplies a small frame size of 10 ms in
order to reduce latency. By designate short frame sizes, LTE allows better channel
estimation to be carried out the mobile, allowing timely feedbacks needful for link
adaptations to be supplied to the base station.[41]
Chapter Two
LTE and OFAM
Figure 2.1: System architecture for LTE Rel-8 network [47].
2.5 History of OFDM:

The initial development of multi-carrier communication system was basically done by
military systems in the late 1950s and mid-1960s. KINEPLEX, ANDEFT and
KATHRYN etc. are the few OFDM based systems utilized by US military systems for
high frequency applications [10].
In 1966, the concept of multicarrier communication was first introduced by Chang
[60] .He suggested a multicarrier scheme utilizing the parallel data transmission by
means of 10 frequency division multiplexing (FDM) with overlapping subcarriers. It
was found to be an efficient scheme for bandwidth utilization and to mitigate the
effect of multipath propagation. It also eliminated the requirement of high-speed
equalization technique. He gave the basic concept of OFDM and outlined a theoretical
way to transmit simultaneous data stream trough linear band limited channel without
Inter Symbol Interference (ISI) and Inter Carrier Interference (ICI) [61] [62].
10
Chapter Two
LTE and OFAM
These systems are called classical Multicarrier Modulation (MCM) system and
transmitted data on non-overlapped band-limited orthogonal signals. These systems
require analog oscillator and filter of much wider bandwidth and sharp cut-off.
Therefore, the concept of OFDM was not gained so much attention or popularity
because of the difficulty in subcarrier recovery without inter-subcarrier interference
by means of analog filters. Due to this reason only, a number of studies in the 1960s
were dedicated for MCM employing overlapped band-limited orthogonal signals [63,
64, and 65]. In the year 1967, B. R. Saltzberg suggested a MCM system employing
Orthogonal time-staggered Quadrature Amplitude Modulation (O-QAM) on the
carriers [63]. The concept of MCM scheme employing time-limited orthogonal
signals, which is similar to OFDM, was first given by H. F. Marmuth [66] in 1960.
[10]
The KINEPLEX system was developed by Collins Radio Company for data
transmission at high frequency over multipath fading channel, in this system, 20 tones
are modulated by DQPSK without filtering, which resulted in overlapping channels.
Initially the implementation of an OFDM system with large number of subcarriers
was very complex and expensive because it requires the array of sinusoidal generators
and coherent demodulators for parallel operations. In order to avoid the crosstalk
between the subcarriers, the system should be free from frequency and timing offsets
[62].
A major breakthrough in the history of OFDM came in 1971 when Weinstein and
Ebert used Discrete Fourier Transform (DFT) to perform baseband modulation and
demodulation which eliminated the need of bank of subcarrier oscillators thus making
the operation efficient and simpler [1,67].
In 1979, after evolutionary growth and development in signal processing and VLSI
technologies, high speed chips can be built around special-purpose hardware
performing the large size Fast Fourier Transform (FFT) (efficient algorithm for DFT)
at affordable price [68], [69].
All the proposals of OFDM systems used guard spaces in frequency domain and
a raised cosine windowing in time domain to combat ISI and ICI. Another milestone
for OFDM history was when Peled and Ruiz introduced Cyclic Prefix (CP) or cyclic
extension in 1980 [67,70] .This solved the problem of maintaining orthogonal
characteristics of the transmitted signals at severe transmission conditions. The
generic idea that they placed was to use cyclic extension of OFDM symbols instead of
using empty guard spaces in frequency domain. This effectively turns the channel as
performing cyclic convolution, which provides orthogonality over dispersive channels
when CP is longer than the channel impulse response [56,70].
Since 1990s, OFDM has been utilized for many broadband communication systems, it
includes high-bit-rate digital subscriber lines (HDSL) [71], asymmetric digital
subscriber lines (ADSL) [72], very high-speed digital subscriber lines (VHDSL) [72],
high definition television (HDTV) terrestrial broadcasting etc. It has also been utilized
by many wireless standards like Digital Audio Broadcasting (DAB) [73] The DAB
standard was in fact the first OFDM-based standard (project started in 1988 by ETSI
and completed in 1995), Digital Video Broadcasting (DVB) [74].
Many standards have been proposed for wireless local area networks (WLANs)
operating in ISM band, which are based on spread-spectrum technology. A number of
studies regarding the commercial applications of OFDM were made during 1990s like
High Bit rate Digital Subscriber Lines (HDSL; 1.6 Mbps), Asymmetric Digital
Subscriber Lines (ADSL; 6 Mbps), Very High Speed Digital Subscriber Lines
11
Chapter Two
LTE and OFAM
(VDSL; 100 Mbps), DAB and High Definition Television (HDTV) terrestrial
broadcasting [75].
In 1997, first OFDM-based WLAN standard, IEEE 802.11, was released. IEEE
802.11 can support a data rate up to 2 Mbps. Later on, in 1999, IEEE approved an
OFDM based standard 802.11a for supporting a data rate up to 54 Mbps. During this
period ETSI has also standardized the HiperLAN/2 standard, which has adopted
OFDM for their PHY standards [1].
In 2001, the FCC came with some new rules for modulations scheme operating in the
2.4 GHz range, which allow IEEE to extend 802.11b to 802.11g standard. Now days,
it has also been used in WiMAX (IEEE 802.16), and mobile broadband wireless
access (MBWA) IEEE 802.10. It is 11 also utilized by 4G wireless communication
systems, such as IMT-A. It is also been considered for 3GPP Long Term Evolution,
which is under deployment [62].
2.6 OFDM:
With the ever growing require of this generation, the necessity for high speed
communication has become a top priority. Different multicarrier modulation
techniques have developed to meet these demands, a few prominent among them
being OFDM and Code Division Multiple Access (CDMA) [52].
The fundamental principle of OFDM is a division of high data rate streams into a
number of lower data rate streams and then transmitted these streams in parallel using
several orthogonal sub-carriers (parallel transmission). Due to this parallel
transmission, the symbol duration increases, thus decrease the prorated amount of
dispersion in time resulting from the multipath delay spread. OFDM can be seen as
either a modulation technique or a multiplexing technique [10].
OFDM communication systems do not depend on increased symbol rates for
achieving higher data rates. That makes the task of managing ISI much easier.
Because data is transmitted in parallel instead of serially, OFDM symbols are
basically much longer than symbols on single carrier systems of equivalent data rate
[53].
The concept of OFDM is very much similar to the well-known and extensively used
technique of Frequency Division Multiplexing (FDM). OFDM uses the principles of
FDM to allow multiple messages to be sent over a single radio channel. It is however
in a much more controlled manner, allowing an improved spectral efficiency [54].
In conventional broadcast, each radio station transmits on a different frequency,
effectively using FDM to maintain a separation between the stations. Due to nonorthogonal nature of carrier frequencies in FDM, a large band gap is required to avoid
inter-channel interference, which reduces the overall spectral efficiency. The
difference between FDM and OFDM is shown in Figure 2.2 [55].
12
Chapter Two
LTE and OFAM
Figure 2.2: Comparison of FDM and OFDM [55]

The sub-carriers are mutually orthogonal (The principle of orthogonality is discussed
in next sub-section.) in the frequency domain which alleviates the effects of ISI as
indicated in the Figure 2.3. All of these sub-carriers experiences flat fading because
they have a bandwidth less than the Mobile channel coherence bandwidth [56].
Figure 2.4 shows a baseband transceiver structure for OFDM utilizing the Fourier
transform for modulation and demodulation. Here the serial data stream is mapped to
complex data symbols (Phase Shift Keying (PSK), QAM, etc.) with a symbol rate
of . The data is then demultiplexed by a serial to parallel converter resulting in a
block of N complex symbols,
.The parallel samples are then passed
through an N point IFFT (in this case no oversampling is assumed) with a rectangular
window of length N.Ts, resulting in complex samples
.Assuming the incoming complex data is random it follows that the IFFT
is a set of
independent random complex sinusoids summed together. The
samples,
are then converted back into a serial data stream producing a
baseband OFDM transmit symbol of length T=N.Ts [57].
A Cyclic Prefix (CP), which is a copy of the final part of the samples, is appended to
the front of the serial data stream before RF up conversion and transmission. The CP
combats the disrupting effects of the channel which introduce ISI.
In the receiver the whole process is reversed to recover the transmitted data, the CP is
removed prior to the FFT which reverses the effect of the IFFT [58]. The complex
symbols at the output of the FFT,
are then decoded and the original bit
steam recovered.
Thus, the IFFT and FFT blocks at the transmitter and at the receiver, respectively, are
important components in an OFDM system. A lot of work has gone into the
optimization of the FFT implementations and the design community has leveraged this
trend to advantage leading to the popularity of OFDM based systems. The time-
13
Chapter Two
LTE and OFAM
frequency view of an OFDM signal is shown in Figure 2.5, where the important
parameters like subcarrier spacing and OFDM symbol period are shown [59].
Figure 2.3 OFDM subcarrier spacing [56].
Input
Signal
Mapper
IDFT
OR
IFFT
S/P
P/S
Add
CP
D/A
Multipath
Fading Ch.
&
noise
output
Signal
demapper
Equalizer
And
P/S
DFT
OR
FFT
Figure 2.4 a block diagram of a basic OFDM system.
14
S/P
Remove
CP
A/D
Chapter Two
LTE and OFAM
Figure 2.5 Time-Frequency view of OFDM signal [59]
2.6.1. Orthogonality of the subcarriers and OFDM:

Two functions or signals are said to be orthogonal if they are mutually independent of
each other. Orthogonality is a feature that lets multiple information signals to be
transmitted skillfully over a common channel with the successful detection [24 and
76].
The subcarrier spacing is chosen so that the waveforms transmitted on different sub
carriers are orthogonal in time, but overlap in frequency. The orthogonality is
achieved by making the peak of each subcarrier signal coincide with the null of the
other subcarrier signals resulting in a perfectly aligned and spaced subcarrier signal
[77].
The principle of orthogonality state that if the time-averaged integral of the product of
( )
( ) }, over a
any two functions from a set of functions { ( ) ( )
joint existence time interval [
] is equal to zero, irrespective of the limit of
existence of the functions, then the functions are told to be orthogonal to each other
within this time-interval [16]. Mathematically, it can be expressed as
( )
( )
(2.1)
The orthogonality property of OFDM signals can be shown with the help of its
spectrum. In the frequency domain every OFDM subcarrier has a
( )
frequency response, as shown in Figure 2.6 [10].
One of the key advantages of OFDM is its efficient use of the frequency band as the
subcarriers are allowed to overlap each other in the frequency domain. The N equally
spaced subcarriers will be orthogonal if the frequency separation between subcarriers
is f =
, where N.Ts is symbol duration, and rectangular windowing of the
IFFT is performed. Under these conditions the subcarriers will have a
waveform
frequency response [78].
Simple rectangular pulse of the length
is used and rectangular shape in time
domain corresponds to a
-square shaped spectrum in frequency domain as
illustrated in Figure 2.6. The LTE sub-carrier spacing is set to f= 15 KHz [62].
15
Chapter Two
LTE and OFAM
Figure 2.6 Per-subcarrier pulse shape and spectrum of basic OFDM transmission [48]
Figure 2.7 shows the frequency response of a 5 carrier system where it is seen that
because of the orthogonal relationship the maximum of a particular sample
corresponds to a null in all other carriers, therefore eliminating the effects of
interference.
Figure 2.7: Frequency spectrum of 5 orthogonal subcarriers of an OFDM transmit

signal [78].
The orthogonality among sub carriers can be viewed in time domain as shown in
Figure 2.8. Each curve represents the time domain view of the wave for a subcarrier.
As seen from Figure 2.3, in a single OFDM symbol duration, there are integer
numbers of cycles of each of the subcarriers [62]
16
Chapter Two
LTE and OFAM
Figure 2.8: Time domain representation of the signal waveforms to show

orthogonality among the subcarriers [62]
2.6.2. Guard Interval:

Individual sub channels can be perfectly separated by the FFT at the receiver when
there are no ISI and Inter-channel Interference (ICI) introduced by channel distortion.
Practically these conditions cannot be acquired. Since the spectra of an OFDM signal
is not precisely band limited, linear distortion like multipath fading caused sub
channel to spread energy in the adjacent channels [79, 80].
Figure 2.9 illustrates the CP insertion technicality, the Guard Interval or CP is a
periodic addition of the final part of an OFDM symbol that is added to the front of the
symbol in the transmitter, and at the receiver the CP is removed before demodulation
[81].
It serves as a recurrence of the end of the symbol, so allowing the linear convolution
of a frequency selective multipath channel to be modeled as circular convolution
which in turn might be transformed to the frequency domain utilizing a discrete
Fourier transform (DFT). This process allows for simple frequency domain processing
like channel estimation and equalization [82].
CP insertion, therefore, increases the size of the data symbol from
to
,
being the duration of the guard-period containing the CP. The standard length of
the guard-period in LTE is defined to be 4.69 s, allowing the system to tolerate path
variations up to 1.4 km (considering the standard LTE symbol length of 66.7 s).
When a cyclic extension longer than a channel impulse response is added, the
negative effect of the previous symbol can be avoided by simply removing that
extension. CP insertion implies the copying of the last part of the OFDM data symbol
and attaching it to the timing at the beginning of the symbol, creating a break between
signals (hence: guarding-period). The receiver can then sample the incoming
waveform at optimum time, as time-dispersion problems (i.e. delays caused by
reflections of the signal) up to the length of the guarding-period are ignored [45].
17
Chapter Two
LTE and OFAM
Figure 2.9 the CP insertion mechanism [83]
2.6.3 One-tap equalizer [10]:

The tap-delay line model with
path is considered for multipath fading channel.
After Considering the effect of the multipath fading channel, the samples of
The
received signal can be expressed as:
(
( ) (
(2.2)
( ) is the impulse response of multipath fading channel with path gains

where,
{ ( )
}, is the path delay of
path, and ( ) is a zeromean, unit variance complex Gaussian noise.
After discarding first G sample of the received signal and taking Z-point FFT, the
output of FFT block is ( ) given as :
(2.3)
Where, the term
is the channel response to the
subcarrier frequency and
is
the Additive white Gaussian noise (AWGN) term in the frequency domain. To
compensate the fading effect of the channel, one-tap equalizer is used and each
element of the vector
is multiplied by an equalized gain factor
the output of
equalizer may be written as
(2.4)
is defined as
Where,
(|
))
(2.5)
18
Chapter Two
LTE and OFAM
2.7 OFDM based Multiple Access:

Various multiple access schemes can be combined with OFDM transmission and they
include orthogonal frequency division multiplexing-time division multiple access
(OFDM-TDMA), OFDMA, and multicarrier code division multiple access (MCCDMA). In OFDM-TDMA, time-slots in multiples of OFDM symbols are used to
separate the transmissions of multiple users as shown in figure. 2.10. This means that
all the used subcarriers are allocated to one of the users for a finite number of OFDM
symbol periods.
The only difference from OFDM-TDMA is that the users capture the channel and use
it for certain duration, i.e., the time dimension is used to separate the user signals [84]
Figure 2.10: Time Frequency view of an OFDM-TDMA Signal

In OFDMA systems, both time and/or frequency resources are used to separate the
multiple user signals. Groups of OFDM symbols and/or groups of subcarriers are the
units used to separate the transmissions to/from multiple users. In figure 2.11, the
time, frequency view of a typical OFDMA signal is shown in a case where there are 3
users. It can be seen from figure 2.11 that users signals are separated either in the
time-domain by using different OFDM symbols and/or in the subcarrier domain.
Thus, both the time and frequency resources are used to support multiuser
transmissions. We shall discuss this technique in more detail in the subsequent
sections and also compare it with OFDM-TDMA [85].
19
Chapter Two
LTE and OFAM
Figure 2.11: Time Frequency view of an OFDMA Signal [85]
2.8 Orthogonal Frequency Division Multiple Access:

The approach used in LTEs access techniques consists of using OFDMA for the
downlink (DL) and SC-FDMA for the uplink (UL).
The main reason that justifies different access techniques for the UL and DL is the
fact that SC-FDMA optimizes range and power consumption at the UE, while
OFDMA minimizes receiver complexity and enables frequency domain scheduling
with flexibility in resource allocation. OFDMA is a multi-carrier transmission scheme
in opposition to SC-FDMA. Both allow multiple user access, depending on the
available bandwidth, by dynamically allocating each user to a specific time-frequency
resource, depending on which duplexing is deployed. OFDM requires a large dynamic
range due to PAPR [86 and 87].
The main difference between an OFDM system and an OFDMA one is represented in
Figure 2.12. The different colors represent different users using resources. In OFDM,
users are assigned to resources in the time domain only, while in OFDMA, users can
be assigned also in the frequency domain, optimizing resource usage.
In OFDMA systems, the multiple user signals are separated in the time and/or
frequency domains. OFDMA has been developed with multi-user operation as its
purpose, allowing a flexible assignment of bandwidth to users according to their
needs.
Typically, a burst in an OFDMA system will consists of several OFDM symbols. The
subcarriers and the OFDM symbol period are the finest allocation units in the
frequency and time domain, respectively. Hence, multiple users are allocated different
slots in the time and frequency domain, i.e., different groups of subcarriers and/ or
OFDM symbols are used for transmitting the signals to/from multiple users. For
instance, we illustrate an example in figure 2.13 wherein the subcarriers in an OFDM
symbol are represented by arrows and the lines shown at different times represent the
different OFDM symbols. We have considered 3 users and we have shown how
resources can be allocated by using the different subcarriers and OFDM symbols [88
and 89].
20
Chapter Two
LTE and OFAM
Figure 2.12 Difference between OFDM and OFDMA resource by user allocation [86].
Figure 2.13: Example allocation of resources to users in an OFDMA system [85].
Figure 2.14 is a detailed block diagram of OFDMA. The LTE PHY (Physical Layer)
specification has been designed to adapt bandwidths from 1.25 MHz to 20 MHz
OFDM was selected as the main modulation scheme due to its robustness with a
severe multipath fading. Downlink multiplexing is achieved through the OFDMA.
OFDM is the modulation scheme for the DL. The primary subcarrier spacing is 15
kHz, with lower subcarrier spacing of 7.5 kHz available for some MB-SFN
(Multicast-broadcast single-frequency network) scenarios. OFDM modulation
parameters summarizes in Table 2-1 [90]
21
Chapter Two
LTE and OFAM
Table 2-1 Downlink OFDM Modulation Parameters [90]

Transmission
1.25 MHz 2.5
5 MHz
10 MHz 15 MHz
BW
MHz
Sub-frame
duration
Sub-carrier
spacing
Sampling
frequency
FFT size
No.
occupied
subcarrier
20 MHz
0.5 ms
15 kHz
192 MHz
(1/2 x 3.84
MHz)
128
of 76
3.84
MHz
7.68
MHz
(2
x 3.84
MHz)
15.36
23.04
30.72 MHz
MHz
MHz (6 (8 x 3.84
(4
x x 3.84 MHz)
3.84
MHz)
MHz)
256
151
512
301
1024
601
22
1536
901
2048
1201
Chapter Two
LTE and OFAM
Figure 2.14 Complete block diagram of an OFDMA transmitter and receiver [91]
23
Chapter Two
LTE and OFAM
2.9 SC-FDMA:
In cellular systems, the wireless communication service in a certain geographical area
is supplied by multiple base stations. The downlink transmissions in cellular systems
are one-to-many, whilst the uplink transmissions are many-to-one. A one-to-many
service means that a base station transmits concurrent signals to multiple users
equipments in its coverage area. This demands that the base station has very high
transmission power ability; as a result of the transmission power is involved for
transmissions to multiple users equipments [92]. On the other hand, in the uplink, a
single users equipment has all its transmission power available for its uplink
transmissions to the base station. In the uplink, the design of an effective multiple
access and multiplexing scheme is more challenging than on the downlink because of
the many-to-one nature of the uplink transmissions. Another consequential requisite
for uplink transmissions is the low signal peakiness by means of the limited
transmission power at the users equipment [92].
One of the main parameters that affect all mobile UE devices is their battery life. It is
therefore necessary to ensure an economic and efficient power use in the transmission
and reception of signals. With the RF power amplifier (i.e enhancer of mixed signals)
and the transmitter being the parts with the highest energy consumption within the
mobile UE; it is essential to establish a transmission model with near constant
operating power level [45].
The downlink physical layer of LTE is depending on OFDMA. Thus, in spite of its
many advantages, OFDMA has specific drawbacks like high sensitivity to frequency
offset (Doppler spread by cause of mobility and Arising from the instability of
electronics) and PAPR. PAPR occurs due to the random constructive addition of subcarriers and results in spectral spreading of the signal which leads to adjacent channel
interference. It is a problem that could be insurmountable with high compression point
power amplifiers and amplifier linearization techniques. While these approaches may
be utilized on the base station, they become costly on the UE [93 and 94].
In LTE, a new concept is used for the access technique of the uplink, called SCFDMA. Its characteristics combine lower PAPR of single-carrier systems because
there is only a single carrier unlike N carriers. (Which allows maintaining a lower
operating power level than OFDMA) with immunity to multipath interference, as well
as flexible subcarrier frequency allocation (as a crucial part of OFDM) [45]. Figure
2.15 shows the concepts of OFDMA and SC-FDMA.
24
Chapter Two
LTE and OFAM
Figure 2.15 frequency domain description of downlink and uplink LTE access
technologies
SC-FDMA differs from OFDMA in one additional transmission step, caused by the
single-path transmission of single-carrier systems. That transmission step, called
resource element mapping (and its counterpart, resource element selection), shifts all
symbols obtained through the FFT to the desired center frequency and passes them on
to the IFFT for further conversion Figure 2.16.
Since the power of the modulation signals used in this process is constant (QPSK
(Quadrature Phase Shift Keying), 16QAM and 64QAM) and the result of the resource
element mapping step is a waveform similar to the original, on another center
frequency; the required result of a constant-power signal is achieved [45].
For practicality, SC-FDMA is implemented in LTE utilizing a Discrete Fourier
Transform Spread OFDM transmission (DFTS-OFDM) which is repeatedly referred
to as a frequency-domain generalization of SC-FDMA. The DFT is used to multiplex
uplink transmissions in definite frequency allocation blocks within the general system
bandwidth in accordance with eNodeB scheduler instructions. The bandwidth of the
single carrier is specified based on the desired data rate by the user. Data remains
serial and not parallelized as done on the downlink with OFDMA (i.e. one
information bit is being transmitted at a time). This results in similar link performance
parameters for the uplink and downlink. Nevertheless, there would be comparatively
high ISI for the uplink because of the single carrier modulation. Thus, the eNodeB
receiver requires a low-complexity equalizer to rectify for the distorting impacts of
the radio channel. SC-FDMA is not as sensitive to Doppler Effect and frequency
instability the as OFDM by cause of its single carrier nature [93].
25
Chapter Two
LTE and OFAM
Figure 2.16 Block diagram of an SC-FDMA transmitter and receiver [37]
26
Chapter Three
Peak-to-Average Power Ratio Reduction

Chapter Three
Peak-to-Average Power Ratio Reduction:
High PAPR of transmitted signals is one of the major issues of the OFDM system. A
large dynamic range of input data symbols is the main cause of getting high PAPR.
An OFDM signal consists of independent data symbols modulated on N orthogonal
subcarriers, and when these signals are added to the same phase, higher peak
amplitude is observed. The value of this peak may be
times of the average
amplitude [10].
3.1 Definitions of PAPR:

For a continuous time baseband OFDM signal, the PAPR of any signal is defined as
the proportion of the maximum instantaneous power of the signal and its average
power. If x (t) is a transmitted baseband OFDM signal, then PAPR is defined as:
,
, ( )-
( )
(3.1)
Where,
is the average power of x (t) and can be computed in frequency domain
because IFFT is a unitary transformation
is useful duration of an OFDM symbol
[95].
For a discrete OFDM signal, the PAPR is computed from
time oversampled
OFDM signal as:
, ( )-
The
[
[
( )
( )
, ( )- at (dB) =
(3.2)
]
[
[
( )
( )
(3.3)
Where, , - denotes the expectation operator and

is the total number of subcarriers. The PAPR of pass band OFDM signal is approximately twice that of
baseband PAPR [95].
The above power characteristics can also be described in terms of their magnitudes
(not power) by defining the crest factor (CF), which is defined as the ratio between
maximum amplitude of OFDM signal ( ) and root-mean-square (RMS) of the
waveform. The CF is defined as:
| ( )|
,|| ( )| |-
(3.4)
In most cases, the peak value of signal ( ) is equals to a maximum value of its
envelope | ( )| However, it can be seen from Figure 3.1 that the appearance of peak
amplitude is very rare, thus it does not make sense to use max | ( )| to represent the
27
Chapter Three
peak value in real application. Therefore, the PAPR performance of OFDM signals is
commonly measured by certain characterization constants which relate to probability
[96].
Figure 3.1: High PAPR when sub-carriers are modulated by same symbols [96]
3.2 PAPR of OFDM signal [62]:

The discrete time baseband OFDM signal is defined in (6). The PAPR of the discrete
time OFDM signal determines the complexity of the digital circuitry in terms of the
number of bits necessary to achieve the desired signal to quantization noise ratio
during signal digitization and recovery. To better approximate the PAPR of a
continuous time OFDM signal, the discrete time OFDM signal is to be obtained by L
times oversampling. The oversampled discrete time OFDM signal can be obtained by
performing LN point IFFT on the data block with (L-1) N zero padding as follows:
, ( )-
, 0 n NL-1
(3.5)
PAPR of the oversampled OFDM signal of becoming

, ( )-
,
,
( )
( )
(3.6)
where, E[. ] denotes the expectation operator and N is total number of sub-carriers.
The PAPR of passband OFDM signal is approximately twice that of baseband PAPR.
Complementary Cumulative Distribution Function (CCDF) for an OFDM signal can
be written as:
P (PAPR > PAP
)=
(3.7)
Where PAP is the clipping level.

This equation can be read as the probability that the PAPR of a symbol block exceeds
some clip level PAP .
28
Chapter Three
3.3 Oversampling discrete OFDM symbols to find true (continuous)

peaks:
The PAPR for the discrete-time baseband signal x [n] may not be the same as that of
the continuous-time baseband signal ( ) In fact, the PAPR for , - is lower than
that for ( ), simply because , - may not have all the peaks of ( ) In practice, the
PAPR for the continuous-time baseband signal can be measured only after
implementing the actual hardware, including digital-to-analog convertor (DAC). In
other words, measurement of the PAPR of the continuous-time baseband signal is not
straightforward. Therefore, there must be some means of estimating the PAPR from
the discrete-time signal , -. Fortunately, it is known that , - can show almost the
same PAPR as ( ) if it is L-times interpolated (oversampled) as shown in Figure 3.2
where L 4 [97 and 98].
Figure 3.2 Block diagram of L times interpolator [83]
3.4 Distribution of PAPR:

To design and develop an effective PAPR reduction technique, it is very important to
accurately identify the distribution of PAPR in OFDM systems. The distribution of
PAPR plays an important role in the design of the whole OFDM system. The
distribution of PAPR can be used in determining the proper output back-off of the
HPA to minimize the total degradation. It can be used directly to calculate the BER
and to estimate the achievable information rates [10].
For the OFDM system, if we assume that the input data stream is statistically
independent and identically distributed (i.e.) then the real and imaginary parts of x[n]
are uncorrelated and orthogonal. From central limit theorem, it follows that, for large
values of N, the real and imaginary parts of x[n] are independent and identically
distributed (i.e.) Gaussian random variables, each with zero mean and variance
,| , - | -
(3.8)
The probability distribution of complex OFDM signals with large N is a complex

Gaussian distribution given by following relation:
* , -+
, -
(3.9)
Where Pr{.} denotes the probability distribution function. Where,

is the variance
of , -.The amplitude of OFDM signal has a Rayleigh distribution and its
probability density function (PDF) is given by:
* , -+
| , -|
| , -|
(3.10)
29
Chapter Three
The histogram plots for the real part, imaginary part and the absolute value of a time
domain OFDM signal are shown in Figure 3.3(a), (b) and (c) respectively. The plots
shown in Figures 3.3(a) and (b) are obtained after performing the computer
simulations of an OFDM system having N=256 QPSK modulated subcarriers as
shown in Fig. 2.4. The signal obtained from IFFT block of Figure 2.4 is complex
OFDM signal. After that real, imaginary and absolute values of OFDM signal (x[n])
are calculated and their histograms are plotted [62].
The power of OFDM signal has chi-square distribution. The distribution of PAPR is
often expressed on the one hand Complementary Cumulative Distribution Function
(CCDF). In probability theory and statistics, the CCDF describes the probability that a
real-valued random variable X with a given probability distribution will be found at a
value greater than or equal to x [99 and 10].
The Cumulative Distribution Function (CDF) of the PAPR of the amplitude of a
signal sample is given by
( )
( )
(3.11)
The CCDF of the PAPR of the data block is desired in our case is to compare outputs
of different reduction techniques. This is given by:
(
(3.12)
( )
(
Where,
(3.13)
(
(3.14)
is the given reference level.
Figure 3.3 (a)

30
Chapter Three
Figure 3.3 (b)
Figure 3.3 (c)

Figure 3.3: Histogram of (a) Real part of OFDM signal amplitude (b) Imaginary part
of OFDM signal amplitude (c) OFDM signal magnitude [63].
31
Chapter Three
3.5 Identification of the Problem:

Multi-carrier phenomena is considered to be one of the major development in wireless
communication and among them OFDM is becoming the important standard.
However, high PAPR is the major drawback of OFDM, which results in lower power
efficiency hence impedes in implementing OFDM. To overcome the low power
efficiency requires not only large back off and large dynamic range DAC but also
highly efficient HPA and linear converters. These demands result in costly hardware
and complex systems. Therefore to lessen the difficulty of complex hardware design it
has become imperative to employ efficient PAPR reduction techniques [100 and 101].
The drawback of a large dynamic range is that it places pressure on the design of
components such as the word length of the IFFT/FFT pair, mixer stages, and most
importantly the HPA, which must be designed to handle irregularly occurring large
peaks, decreases the SQNR (Signal-to-Quantization Noise Ratio) of ADC (Analog-toDigital Converter) and DAC, The PAPR problem is more important in the uplink
since the efficiency of power amplifier is critical due to the limited battery power in a
mobile terminal. Failure to design components with a sufficiently large linear range
result in saturation of the HPA [98, 78]. Saturation creates both in band distortion,
increasing the BER and out of band distortion, or spectral splatter, which causes
Adjacent Channel Interference (ACI). One obvious solution is to design the
components to operate within large linear regions, however this is impractical as the
components will be operating inefficiently and the cost becomes prohibitively high.
This is especially apparent in the HPA where much of the cost and ~50% of the size
of a transmitter lies which will be explained in next sections [98, 78].
3.5.1 Nonlinear HPA and DAC:

HPA are used in the transmitter of communication systems for sufficient transmission
power. To achieve maximum output power efficiency they have to be operated at or
near the saturation region. [100]
If the data on the subcarriers add up in a constructive manner at the transmitter, the
resulting signal could exhibit large PAPR. As a result, the composite transmit signal
could be severely clipped by the DAC and power amplifiers for their bounded
dynamic range as described in Figure 3.4. In this case, the reconstructed output ( )
can possess a significant amount of distortion. It can be reduce the PAPR of an
OFDM signal by modifying the signal characteristics in time-domain or frequency
domain clipping of the composite OFDM signal causes several undesirable outcomes,
such as signal distortion and spectral regrowth. For instance, clipping causes in band
noise that results in a degradation of the BER performance .Moreover, higher-order
harmonics that spill over into OOB spectrum can also result from signal clipping.
Although filtering after the HPA can be employed to remove this spectral leakage, it
is very power-inefficient, so it is an undesirable solution. Therefore, the dynamic
range of DAC should be large enough to accommodate the largest peaks of signals or
high PAPR values [102].
A high-precision DAC support high PAPR with acceptable amount of quantization
noise, but could be very costly to a certain sampling rate of the system. On the other
hand, a low-precision DAC would be cheaper, but the quantization noise will be
significant, which reduces the signal SNR (Signal to Noise Ratio) when the dynamic
range of DAC is increased to support high PAPR. Otherwise, the DAC will saturate
and clipping will occur [48, and 103].
32
Chapter Three
Figure 3.4 An example illustrating effect of clipping.

The dynamic range of the power amplifiers should also be large enough to
accommodate large PAPR values. Otherwise, the power amplifiers may saturate and
clipping might occur. The component cost of the DAC and power amplifiers increase
with the increase in the dynamic range.
33
Chapter Three
It is worth mentioning that the clipping of high signal peaks rarely happens, resulting
in a comparatively low incidence clipping noise. In this manner, the impact of
clipping at the transmitter on the error performance of the OFDM system liable to be
subjected frequency selective fading is minimal [102].
If an HPA with limited linear range is utilized for amplification, it may operate near
saturation and can cause OOB radiations and in-band distortion. The OOB
distortion/noise is a major concern, especially in wireless communications, where
large differences in signal strength from a mobile transmitter impose stringent
requirements on ACI [104]
Figure 3.5 demonstrates a classic input-output characteristic of a power amplifier. For
prevent or limit signal distortion input signals must be preserved below the Non-linear
area. The result is that the amplifier is not completely used [105]
IBO = 10
OBO = 10
(
(
(3.15)
(3.16)
IBO (Input Back-Off) or OBO (Output Back-Off)

High PAPR results in a wide variety of OFDM signal amplitudes which due to
nonlinear characteristics of HPA findings in inter-modulation among the various sub
carriers and leading to an increment in BER. To realize a low BER and minimal
signal distortion, HPA must be a large dynamic range and work in the linear amplifier
region. But, these types of HPA are expensive and smaller power efficient. The power
efficiency in wireless communication is very important for achieving efficient area
coverage and small size terminals. Thus, the power efficient process of non-linear
HPA is so important. Accordingly, it is best to target the reduction of PAPR the
OFDM signals before transmitting the signal into nonlinear DAC and HPA [100].
Figure 3.5 Typical input-output characteristics of a power amplifier showing the

Relation between Output Back-Off (OBO) and Input Back-Off (IBO) [98].
34
Chapter Three
3.5.2 Power Saving [100]:

A high dynamic range HPA has low power efficiency. The power could save by
reducing PAPR. This power saving that is implemented in such a way to provide a
direct correlation with the desired average output power.
On the assumption a linear model of HPA, the power efficiency is:
(3.17)
(3.18)
The = HPA efficiency .
= the average of the output power.
. = A fixed amount of power regardless of their input power.
For example: an OFDM signal with 256 sub carriers that demand an IBO equal to the
PAPR at the probability level lower than 0.01%, i.e. (25.235).This makes
= 0.5/25.2351.98%
The PAPR of OFDM systems has to reduce for avoiding this level of power
inefficiency.
3.6 Factors influencing the PAPR:

3.6.1 The number of sub carriers:
In Multi-Carrier Systems the complex base band signal for one symbol in an OFDM
system can be expressed as follows:
( )
(3.19)
Where N is the modulating symbol and is the number of sub carriers. For moderately
large numbers of m-PSK (multiple phase-shift keying) sub carriers the quadrature
components of x (t) each tends towards a Gaussian distribution (giving the sum of
their power amplitude a Rayleigh distribution). Consequently, whilst the peak value
possible is N times the individual sub carrier peak, the probability of any value close
to that peak occurring is very low. For example, with only 24 sub carriers, the
probability of the PAPR exceeding 4dB is
and of exceeding 8dB is only
[99].
3.6.2 The order of Modulation:

High data bandwidth efficiency (in terms of b/s/Hz) this can be achieved by utilizing
higher order modulations based, for instance, on QAM. When using a higher-order
modulation such as QAM type, the PAPR of the summed OFDM signal is increased
by the PAPR of the QAM constellation utilized. Nevertheless, the probability of these
higher peaks happening is accordingly less. Furthermore, since among benefits of
OFDM is one that sub carriers could have their modulation independently varied to
adapt to channel conditions, the joined PAPR in any system utilizing this technique
might are hard to predict and control. PAPR for an unfiltered base band signal is listed
in the following Table 3.1. [100].
35
Chapter Three
Table 3.1 PAPR for picked modulation formats

Modulation
PAPR
256-QAM
4.23dB
64-QAM
3.68dB
256-QAM (modified)
2.85dB
16-QAM
2.55dB
m-PSK (reference)
0 dB
3.6.3 Constellation shape:

The last entry in Table 3.1 is for a constellation obtained by modifying 256- QAM to
reduce PAPR. This modified constellation shape is shown in figure 3.6. However,
there is an additional processor load associated with encoding and decoding this
constellation.
Figure 3.6 256-QAM constellations: (a) regular and (b) modified mapping to reduce
PAPR
3.6.4 Pulse Shaping:

In terrestrial communications, it is popular to use pulse shaping to the base band
signal, to decrease the bandwidth of the transmitted spectrum, but this causes
overshoot and can increase the PAPR of the modulating signal by 4-5 dB [100].
3.7 The gauge for judgment of the PAPR reduction in OFDM systems
[106, 107, 108]:
Every method used to reduce the PAPR has some drawbacks and merits. There is
always a trade-off between PAPR reduction and some other factors like bandwidth,
computational complexity, average power etc. An ideal PAPR reduction technique
should have following characteristics:
1) High potential to limit the PAPR: It is a key factor to consider in the selection of
technology to reduce the PAPR with few adverse side effects like in-band distortion
and OOB radiation.
2) Low average power: even though it can reduce PAPR through the average power of
the original signals increase, it needs a bigger linear operation region in HPA and
which led in the deterioration of BER performance.
36
Chapter Three
3) Low implementation complexity: mainly, complexity techniques viewing better

capability of PAPR reduction. Nevertheless, practically, both time and hardware
requisites for the PAPR reduction must be minimal.
4) No bandwidth expansion: The bandwidth is an infrequent resource in systems. The
bandwidth expansion has directly resulted in the data code rate loss because of side
information (like the complementary bits in Complement Block Coding (CBC) and
phase factors in PTS). Furthermore, when the side information is received in error
unless some methods of protection like channel coding employed. For that reason,
when channel coding is utilized, the loss in data rate is incremented further due to side
information. Then, the loss in bandwidth because of side information must be avoided
or at least be preserved minimal.
5) No BER performance degradation: The objective of the PAPR reduction is for the
best system performance, including BER than that of the original OFDM system. For
that reason, all the methods, which have an incrementation in BER at the receiver,
must be paid more attention in practice. Additionally, if the side information is
received in error at the receiver, which may also result in entire wrong data frame and
thus the BER performance is reduced.
6) Without the additional power required: The design of a wireless system must
always take into account the efficiency of power. If an operation of the technique
which reduces the PAPR require more extra power, it deteriorates the BER
performance when the transmitted signals are normalized back to the original power
signal [109].
7) No spectral spillage: Any PAPR reduction techniques cannot devastate OFDM
fascinating technical features like immunity to the multipath fading. Thus, the spectral
spillage must be avoided in the PAPR reduction.
8) Other factors: It must be driven greater concentration on the effect of the nonlinear
devices utilized in signal processing loop in the transmitter like DACs, mixers and
HPAs since the PAPR reduction fundamentally avoid nonlinear distortion as a result
of these memories-less devices introducing into the communication channels. At the
same time, the expense of these nonlinear devices is too the important factor to design
the PAPR reduction scheme.
3.8 Fitness function-based approach for determining an appropriate

Algorithm [110]:
In order to determine an appropriate PAPR reduction algorithm for a given system, it
is desirable to consider all above-listed requirements. The number and nature of these
requirements may vary depending upon the system (or user) under consideration. For
a given scenario and requirements, we propose to use the fitness value or
appropriateness value of the algorithm, which is defined as the weighted sum of the
relative changes in the above-listed factors. The appropriateness value provides a
single metric for determining the appropriateness of a PAPR reduction algorithm.
Suppose X1 be the relative degradation in BER performance at certain SNR level, for
given channel conditions, AWGN or multipath, given by:
X1 = 10
(
)
(3.20)
Let X2 be the relative increase in system complexity given by:
X2 = 10
37
(3.21)
Chapter Three
Let X3 be the relative PAPR reduction given by:

X3 = 10
(3.22)
Let X4 be the relative cost savings given by:

X4 = 10
(3.23)
Let X5 be the relative increase in transmit power given by:

X5 = 10
(3.24)
Let X6 be the relative increase in spectral spillage given by:

X6 = 10
(O
(3.25)
Let X7 be the relative reduction in goodput5 given by:

X7 = 10
(3.26)
The aggregate fitness value of the PAPR reduction algorithm can be computed as the
weighted sum of these factors, where the weights correspond to their relative
importance levels. These weights can be determined as per the system or user
requirements. Therefore, the fitness value of the algorithm is given by:
(3.27)
Where
(3.28)
Based on these fitness values, an appropriate algorithm can be chosen in order to

achieve large reduction in PAPR values as well as satisfy other system requirements.
38
Chapter Four
PAPR Reduction Techniques

Chapter Four
4.1There are three different way to divide the PAPR:

4.1.1The first way is [110]:
PAPR reduction techniques can be categorized into deterministic and probabilistic
approaches, as shown in Figure 4.1. Deterministic approaches guarantee that the
PAPR of an OFDM signal does not exceed a predefined threshold, whereas the
probabilistic approaches minimize the probability that the PAPR of an OFDM signal
exceeds a predefined threshold. These categories will be discussed in the following
sections
1) Deterministic Approach
Deterministic PAPR reduction approaches can be classified into techniques that
perform either time-domain based clipping or frequency-domain based coding. The
simplest approach for PAPR reduction is to deliberately clip the amplitude of the
signal to a predefined value before amplification [111]. However, the technique
suffers from various drawbacks, such as signal distortion and spectral regrowth.
Therefore, clipping alone is not a suitable option for PAPR reduction. Modified
clipping techniques exist, which fall under the probabilistic approach explained in the
next section.
Coding techniques are applied to OFDM signals in order to map symbols to codes
with smaller PAPR values [112] .
Each symbol has a choice of two or more codes, where the code yielding the lowest
PAPR is selected. However, this technique works well only when the number of
subcarriers is small. With the increased number of subcarriers, the search space for
finding codes with minimum PAPR increases exponentially and large lookup tables
are needed for encoding and decoding.
2) Probabilistic Approach
Probabilistic approaches attempt to minimize the number of occurrences of OFDM
symbols with PAPR values exceeding a predefined threshold, while simultaneously
minimizing the signal distortion and spectral growth. Probabilistic approaches can be
classified according to whether time domain processing or frequency domain
processing is involved:
time Domain-Based Processing:
Time domain-based processing approaches focus on manipulating the power of the
signal in the time domain. This approach can be further classified into blind and nonblind techniques. Blind techniques imply that the receiver is oblivious to the changes
made at the transmitter side, whereas non-blind techniques imply that the receiver
requires a priori knowledge about the modifications made at the transmitter side for
correctly demodulating the received signals. Thus, non-blind techniques require
additional side information in order to operate, whereas blind techniques might
degrade the error performance of the system since the receiver is transparent to the
changes made at the transmitter side.
39
Chapter Four
The simplest blind technique for PAPR reduction is to clip the amplitude of the signal
to a predefined value and filter the signal to suppress the out-of-band interference
[113,114, 115 ] . The clipping process might result in spectral regrowth, whereas
filtering the signal might result in some peak regrowth. Therefore, clipping may not
be an effective technique when reducing the PAPR of the OFDM signals as long as
the transmitted OFDM signal is strictly band-limited. Even though numerous
algorithms based on amplitude clipping and filtering have been proposed in the
literature, it has been shown that clipping does not improve the reduction of total
degradation [116]. Instead, an unclipped system outperforms a clipped system
because of the inter-carrier interference (ICI) caused by clipping, and offsets the gain
of the PAPR reduction [116]. Another technique called peak windowing can also
reduce the PAPR, where large signal peaks are multiplied with a certain narrowband
window such as Gaussian, Cosine, Kaiser, and Hamming windows [117].
Among the non-blind techniques, several companding4 techniques for compressing
the large peaks of an OFDM signal in time domain, including -law companding , and
exponential companding , have been proposed in literature. By compressing the large
peaks of an OFDM signal by companding, the dynamic range of the D/A converters
are reduced. However, the receiver needs to expand the compressed signal for correct
demodulation.
Frequency Domain-Based Processing
Frequency domain-based processing approaches focus on minimizing the correlation
of the input signals since it is known that the PAPR of an OFDM signal is high when
the input sequences are highly correlated. It has been shown that by altering the phase
and/or power of the input sequence, it is possible to lower the correlation of the input
sequence, thereby reducing the PAPR of an OFDM signal. However, some
approaches also try to directly manipulate the correlation of the input signals.
Frequency domain-based processing approaches can be further classified into blind
and non-blind techniques. In blind phase adjustment-based techniques, the phase of
the subcarriers are adjusted in order to reduce the coherence between the different
subcarriers such that the PAPR value of the OFDM signal is reduced. The phase
adjustments should be kept relatively small so as to minimize bit-error-rate (BER)
performance degradation. For example, signal set expansion technique maps original
signal set into an expanded signal set with two or more points, such as binary phase
shift keying (BPSK) into quadrature phase shift keying (QPSK), which provides more
freedom for phase selection and yields lower PAPR values for the OFDM signal
[118].
Blind power-based techniques alter the power level of the subcarriers such that the
PAPR of an OFDM signal is reduced. These techniques are suitable only for the
MPSK-based OFDM system since the receiver is unaware of the information about
the transmit power levels. For example, the input sequence envelope scaling technique
adjusts the power of the subcarriers so that the power of the individual subcarriers
becomes unequal yielding a minimized PAPR value [119]. Since the phase
information of the original signal is unchanged, the receiver can decode the received
signal without any side information.
In blind power and phase-based techniques, both the phase and the power of the
subcarriers are altered such that the PAPR of an OFDM signal is reduced. If the total
transmit power needs to be kept constant, these techniques are suitable only for low
order modulation techniques since the error robustness of the higher modulation
techniques degrades rapidly with the blind phase and power alterations at the
40
Chapter Four
transmitter. When the order of the modulation techniques in-creases, the complexity
(and limitations) of the algorithm increases as well as transmit power level increases.
For example, the active constellation extension (ACE) [120,121] and dynamic
constellation shaping techniques allow changing the power and phase of some data
symbols without affecting the error probability of the other data symbols.
Non-blind power-based techniques, as well as power and phase-based techniques,
would be suitable for the higher modulation schemes such as MQAM. Non-blind
phase adjustment-based techniques update phases of the input sequence such that the
PAPR of an OFDM signal is reduced. The information about the phase updates is
transmitted to the receiver for correct demodulation. Several modified algorithms are
proposed in literature, which avoid the requirement of explicit side information. For
example, selective mapping (SLM)[9], partial transmit sequences (PTS) [122],
random phase updating [123] techniques add random phase factors to each subcarriers
in order to reduce PAPR with the information about the phase factors transmitted to
the receiver. The blind techniques reduce the PAPR values at the cost of slight
increase in the bit error rate of the system or increased transmit power level since the
adjustments would result into increased noise level at the receiver, whereas the nonblind techniques reduce the PAPR values at the cost of a reduced information rate
since the information about the adjustments made at the transmitter need to be
transmitted to the receiver for the correct demodulation.
A low autocorrelation coefficient of a signal is a sufficient condition for low PAPR.
However this is not a necessary condition [124][125]. Non-blind autocorrelation
minimization techniques attempt to minimize the autocorrelation of the input
sequence `and the information about the changes is transmitted to the receiver for
correct
demodulation. For example, the selective scrambling [126] and interleaving
techniques [127] attempt to break the long correlation patterns of the input sequences
to reduce the PAPR. However, the techniques perform well only when the OFDM
signal has moderate PAPR values since interleaving alone is not effective to break the
correlation pattern when the input sequence are highly correlated.
Attempts have been made to develop OFDM signals with a constant envelope to yield
unity PAPR values [128] . The constant envelope waveforms have a constant
instantaneous power. Continuous phase modulation (CPM) is a class of signaling that
has very low side lobe power while maintaining the constant envelope property.
However, CPM increases the complexity of the receiver and has a poor performance
over frequency selective channels.
41
Chapter Four
Figure 4.1.the first way taxonomy of PAPR Reduction techniques
42
Chapter Four
4.1.2 The second way is :

a) Distortion Based Techniques [11]-[8]-[4]
b) Scrambling Techniques [17]-[16]-[8]
As shown in figure 4.2
a. DISTORTION BASED TECHNIQUES
The schemes that introduce spectral re-growth belong to this category. Distortion
based techniques are the most straightforward PAPR reduction methods. Furthermore,
these techniques distort the spectrum, this spectrum distortion or spectral re-growth
can be corrected to a certain extent by using filtering operation [62 ,129]. These
methods reduce the PAPR by distorting the OFDM signal non-linearly. The methods
like clipping and filtering, peak windowing, and non-linear companding are the
example of these techniques. These techniques are applied after the generation of
OFDM signal (after the IFFT) [130].
The distortion category attempts to reduce PAPR by manipulation of signal before
amplification. Clipping of signal prior to amplification is a simplest method but it
causes increase in both out-of-band (OOB) as well as in-band interference thus
compromises upon performance of system. Amongst this category better techniques
include companding, peak windowing, peak power suppression, peak cancellation,
weighted multicarrier transmission etc. Any technique which is used to reduce PAPR
should not only have high spectral efficiency but must be compatibility with the
existing modulation schemes and at the same time must not be computational
complex [100].
b. Scrambling techniques :
Signal scrambling techniques are all variations on how to scramble the codes to
decrease the PAPR. Coding techniques can be used for signal scrambling. Golay
complementary sequences, Shapiro-Rudin sequences, M sequences, Barker codes can
be used efficiently to reduce the PAPR. However with the increase in the number of
carriers the overhead associated with exhaustive search of the best code would
increase exponentially. More practical solutions of the signal scrambling techniques
are block coding, Selective Level Mapping (SLM) and Partial Transmit Sequences
(PTS). Signal scrambling techniques with side information reduces the effective
throughput since they introduce redundancy [131] [132].
43
Chapter Four
Figure 4.2.the second way taxonomy of PAPR Reduction techniques
44
Chapter Four
4.1.3 The third way is [98]:

These methods are basically divided in five categories:
(1) The clipping technique
(2) Coding Methods,
(3) Probabilistic (Scrambling) Techniques
(4) Pre-distortion Methods.
1. The clipping technique employs clipping or nonlinear saturation around the peaks
to reduce the PAPR. It is simple to implement, but it may cause in-band and out-ofband interferences while destroying the orthogonality among the subcarriers. This
particular approach includes block-scaling technique, clipping and filtering technique,
peak windowing technique, peak cancellation technique, Fourier projection technique,
and decision-aided reconstruction technique [133] [134].
2. The coding technique is to select such code words that minimize or reduce the
PAPR. It causes no distortion and creates no out-of-band radiation, but it suffers from
bandwidth efficiency as the code rate is reduced. It also suffers from complexity to
find the best codes and to store large lookup tables for encoding and decoding,
especially for a large number of subcarriers. Golay complementary sequence, Reed
Muller code, M-sequence, or Hadamard code can be used in this approach [133][134].
3. The probabilistic (scrambling) technique is to scramble an input data block of the
OFDM symbols and transmit one of them with the minimum PAPR so that the
probability of incurring high PAPR can be reduced. While it does not suffer from the
out-of-band power, the spectral efficiency decreases and the complexity increases as
the number of subcarriers increases. Furthermore, it cannot guarantee the PAPR
belowa specified level. This approach includes SLM (Selective Mapping), PTS
(Partial Transmit Sequence).
4. The pre-distortion methods are based on the re-orientation or spreading the energy
of data symbol before taking IFFT. The pre-distortion schemes include DFT
spreading, pulse shaping or precoding and constellation shaping. The methods like
Tone Reservation (TR) and Tone Injection (TI) are the example of constellation
shaping schemes [10].
The DFT-spreading technique is to spread the input signal with DFT, which can be
subsequently taken into IFFT. This can reduce the PAPR of OFDM signal to the level
of
Single-carrier transmission. This technique is particularly useful for mobile terminals
in uplink transmission. It is known as the Single Carrier-FDMA (SC-FDMA), which
is adopted for uplink transmission in the 3GPP LTE standard [135].
4.1.4 And finally there is Hybrid techniques:
Besides these different PAPR reduction techniques, some hybrid methods are also
available in the literature [136 ,137,138 ] . These methods combine two or more than
two techniques for PAPR reduction like clipping with coding, SLM with coding, precoding with clipping, interleaving and companding , Selective Mapping and Binary
Cyclic Codes, combining Hadamard Transform and Hann peak windowing etc. The
hybrid methods are considered as better choice for PAPR reduction because it possess
the advantages of both techniques used in hybridization with slight increases in
complexity.
45
Chapter Four
4.2 Clipping and Filtering :

The clipping is the simplest method of PAPR reduction. Clipping limits the maximum
amplitude of OFDM signal to a pre-specified level. The implementation of clipping is
relatively easy.
The simplest and most widely used technique of PAPR reduction is to basically clip
the parts of the signals that are outside the allowed region .For example; using HPA
with saturation level below the signal span will automatically cause the signal to be
clipped. For amplitude clipping, that is [109]:
(4.1)
Where A is preset clipping level and it is a positive real number
Generally, clipping is performed at the transmitter. However, the receiver need to
estimate the clipping that has occurred and to compensate the received OFDM symbol
accordingly. Typically, at most one clipping occurs per OFDM symbol, and thus the
receiver has to estimate two parameters: location and size of the clip. However, it is
difficult to get this information. Therefore, clipping method introduces both in band
distortion and out of band radiation into OFDM signals, which degrades the system
performance including BER and spectral efficiency. Filtering can reduce out of band
radiation after clipping although it cannot reduce in-band distortion. However,
clipping may cause some peak regrowth so that the signal after clipping and filtering
will exceed the clipping level at some points [108] [109].
It has following drawbacks [98] [139]:
(a) It causes in-band signal distortion, resulting in BER performance degradation.
(b) It also causes out-of-band radiation, which imposes out-of-band interference
signals to adjacent channels. The out-of-band radiation can be reduced by filtering,
but the filtering may affect high-frequency components of in-band signal (aliasing)
when the clipping is performed with the Nyquist sampling rate.
(c) Filtering after clipping can reduce out-of-band radiation at the cost of peak regrowth. The signal after filtering operation may exceed the clipping level specified for
the clipping operation.
To reduce overall peak re-growth, a repeated clipping and filtering can be used to
obtain a desirable PAPR at the cost of increase computational complexity . To reduce
peak regrowth, a repeated clipping-and-filtering operation can be used to obtain a
desirable PAPR at a cost of computational complexity increase. As improved clipping
methods, peak windowing schemes attempt to minimize the out of band radiation by
using narrowband windows such as Gaussian window to attenuate peak signals [140]
Some of clipping techniques:
1. Repeated Clipping [13]
The clipping technique is the simpler one which is used to cut the signal peak up to
desired threshold level. But repeated clipping and filtering technique proved to be
worthy one as it gives better result compared to earlier one. In this technique the peak
regrowth which is generated in filtering can be minimized. So the repeated clip and
filter process reduces these regrowth's in OFDM system
2. Reconstruction of Lost Clipped Signal
46
Chapter Four
To remove the peak regrowth of signal oversampled sequence clipping is used which
can reconstruct the clipped samples and mitigate the clipping distortion in presence of
channel noise at the cost of bandwidth expansion. It is observed that by increasing
small bandwidth , the performance of OFDM system can be improved . PAPR is the
biggest problem in OFDM system. Many techniques are proposed for it. Clipping and
filtering technique is considered to be the simplest one [114][106].
3. Iterative Clipping & Filtering Technique
This technique is used to eliminate the peak regrowth due to CF technique. In each
iteration peak regrowth decreases significantly. The process of iteration undergoes
FFT/IFFT and one extra IFFT is required for conversion into time domain in OFDM
[115][106].
4. Recursive Clipping and Filtering with Bounded Distortion (rcfbd)
In RCF the signal is clipped by repeating process many times before feeding to power
amplifier. When the process of repetition exhibit on the signal the out of band spectral
density and the probability of the occurance of PAPR decreases but error rate
increases due to increase in number of repetitions. The bit error rate increases due to
increase in inband distortion. So to remove this increased error rate another improved
technique is proposed called recursive clipping and filtering with bounded distortion
(RCFBD) to achieve PAPR reduction. The idea of this technique is same as
oversampled digital clipping in time domain and removing out of band components in
frequency domains are used. But additional barrier on in band distortion of each
subcarrier is applied during the recursive process. In this way PAPR can be reduced
without producing any effect on the error rate [114][ 106].
RCFBD minimize PAPR and keeps the control on the distortion of data carried by
each subcarrier. So by using this technique side information can be eliminated and
receiver part becomes less complex and BER performance can be increased more. It is
also more robust against AWGN noise [113].
4.3 Peak Windowing Method:

It is an improved clipping method. The basic aim of peak windowing is to reduce the
out-of-band radiation by using narrow band windows such as Gaussian window to
attenuate peak signals. As a matter of fact, any window which is narrow in time
domain and having good spectral properties can be used [10]. In 2008, an advance
peak windowing method has been given by S. Cha which overcomes the drawback of
normal peak windowing method. It effectively suppresses the peak signals to the
desired threshold level in case when the successive peaks occur within a half of the
window length [10].
The peak windowing method has been suggested by Van Nee and Wild [117]. This
method, proposes that it is possible to remove large peaks at the cost of a slight
amount of self-interference when large peaks arise infrequently. Peak windowing
reduces PAPRs at the cost of increasing the BER and out-of-band radiation. Clipping
is a one kind of simple introduces PAPR reduction technique which is selfinterference. The technique of peak windowing offers better PAPR reduction with
better spectral properties.
(Peak Windowing technique provides better PAPR reduction with better spectral
properties than clipping) [141][142].
47
Chapter Four
In peak windowing method we multiply large signal peak with a specific window, for
example; Gaussian shaped window, cosine, Kaiser and Hamming window. In view of
the fact that the OFDM signal is multiplied with several of these windows,
consequential spectrum is a convolution of the original OFDM spectrum with the
spectrum of the applied window. Thus, the window should be as narrow band as
possible, conversely the window should not be too long in the time domain because
various signal samples are affected, which results an increase in bit error rate (BER).
Windowing method, PAPRs can be obtained to 4dB which from the number of
independent subcarriers. The loss in signal-to-noise ratio (SNR) due to the signal
distortion is limited to about 0.3dB. A back off relative to maximum output power of
about 5.5dB is needed in spectra distortion at least 30dB below the in-band spectral
density [141][142].
The PAPR reduction performance as well as spectral efficiency of peak windowing
technique is better as compared to clipping. The major advantage of peak windowing
is that PAPR reduction is achieved with minimal complexity for any number of sub
carriers. The disadvantages include an increase in out-of-band interference and BER
[100][143].
4.4 Envelope Scaling:

The Envelope Scaling technique has been proposed by Foomooljareon and Fernando.
They anticipated a new algorithm to reduce PAPR by scaling the input envelope for
some subcarriers before they are sent to IFFT. They used 256 subcarriers with QPSK
modulation technique, so that envelopes of all the subcarrie4rs are equal. The key idea
of this scheme is that the input envelope in some sub carrier is scaled to achieve the
smallest amount of PAPR at the output of the IFFT. Thus, the receiver of the system
doesnt need any side information for decoding the receiver sequence. This scheme is
appropriate for QPSK modulation; the envelopes of all subcarriers are equal. Results
show that PAPR can be reduced significantly at around 4 dB [144].
In Envelope Scaling, the input envelopes of sub carriers are scaled prior to IFFT. The
base for this scheme is the facts that with PSK modulation all the sub carriers input
envelops are equal. Hence input envelop of some sub carriers is scaled in such a way
that minimum PAPR is achieved at IFFT output. The input which yields minimum
PAPR is fed into the system. The phase information of the input sequence is similar to
original however envelops are not the same. Hence decoding of sequence can be done
by receiver without any requirement for side information .The major drawback of this
method is that it can only be used when OFDM is employing PSK modulation. On the
other hand if we use this method when QAM modulation is implemented by OFDM,
then there is severe degradation in BER performance results [100] [145] .
4.5 Peak Reduction Carrier:

Peak Reduction Carrier technique is proposed by Tan and Wassell. The technique is
to use the data bearing peak reduction carriers (PRCs) to reduce the effective PAPR in
the OFDM system. It includes the use of a higher order modulation scheme to
represent a lower order modulation symbol. Hence the phase and amplitude of these
carriers remains inside the constellation area which represents the data symbols being
transmitted. This method is suitable for PSK modulation; where the envelopes of all
subcarriers are the same. When the QAM modulation scheme will be implemented in
the OFDM system, the carrier envelope scaling will result in the serious BER
48
Chapter Four
degradation. To limit the BER degradation, amount of the side information would also
be excessive when the number of subcarriers is large [141]
Amongst drawbacks of PRCs, one is that the overall data transmission efficiency of
the system is compromised if we try to achieve maximum PAPR reduction efficiency.
At the same time the BER performance is also affected adversely because of
employing constellation of higher order for carrying symbols of lower order results in
higher probability of error [100]
4.6 Companding Technique:

Non-linear companding is an especial clipping technique which offers good PAPR
reduction with better BER performance, low implementation complexity, and no
bandwidth expansion [109] [145].
The difference between clipping and companding is that the clipping process
deliberately clips the large amplitude signals; therefore the signal cannot be recovered
exactly. On the other hand, the companding transform compand the original signals
using strict monotone increasing function; therefore the companded signals can be
recovered correctly through the corresponding inversion of companding transform at
the receiver. Clipping does not affect small amplitude signal, whereas companding
enlarge the small signals while compressing the large amplitude signals. A lot of
companding techniques are available. The basic concept of most of the companding
techniques is to transform the Rayleigh distributed OFDM signal into a uniformly
distributed signal [10].
It was based on the speech processing algorithm -law and it has shown better
performance than that of clipping method . The -law companding transform mainly
focuses on enlarging small amplitude signals while keeping peak signals unchanged,
and thus it increase the average power of the transmitted signals and possibly results
in exceeding the saturation region of HPA to make the system performance worse
[140]. In order to overcome the problem of -law companding (increasing average
power) and to have efficient PAPR reduction, some more Companding Transform
have been suggested [146,147,148,149,150, and 151] .
Figure 4.3 Block diagram of Companding of OFDM system
49
Chapter Four
4.6.1 Square-Rooting Companding Technique ( SQRT) for PAPR

Reduction in OFDM Systems:
The block diagram of a typical OFDM system using the original SQRT technique for
PAPR reduction is shown in figure 4.4. By using the SQRT technique, the original
OFDM output signals
is processed by (3.21) before they are converted into analog
waveforms and amplified by the power amplifier
| |
(4.2)
is the new OFDM signal, and

is the phase of
During the entire signal
processing, the phases of the OFDM output signals
are kept unchanged while only
the amplitudes are treated and changed [152].
For the complex Gaussian distributed signals, such as OFDM output signals, SQRT
process changes the Rayleigh distribution of these signals into a Gaussian-like, close
to Gaussian, distribution [16,152]; while the Chi-square distribution is converted,
according to the analysis of these signals given in the previous section, to Rayleigh
distribution. The latter is because the Rayleigh distribution in such signals represents
voltage, while the Chi_square distribution represents the power of the same signals.
However, not only the statistical distribution is changed by the SQRT process, but the
values of the mean and variance of the processed OFDM output signals are also
changed, and subsequently the values of the average power and peak power of these
signals are altered also. To understand the effect of SQRT process on the power
values of OFDM output signals, we assume normalized average power (
)
Figure 4.4 Block diagram of an OFDM system using SQRT technique

When the average power is normalized, the value of the peak power is diminished by
N because
for the same PAPR. This assumption is applicable for all
OFDM symbols as the average power is constant and equal to
(
) Hence, the PAPR can be analyzed according to (3.21) through the peak power
only. The new value of normalized peak power is always greater than one because
is constantly greater than
in all OFDM symbols. Therefore, the SQRT
process always causes a reduction in the value of the peak power of the normalized
OFDM symbols, and as a result the PAPR is reduced in all sizes of OFDM blocks, N.
50
Chapter Four
In [16, 152] the SQRT process is applied on the signals of all OFDM output symbols;
therefore, the PAPR reduced without the need to send side information. The SQRT
process changes the distribution of the power signals to Rayleigh distribution and
reduces the value of average power from N to N1/2. The variance of the Rayleigh
distribution equals (
)
[152] which is approximately equal to half the
value of variance of the Gaussian distributed signals. The SQRT process in the SQRT
OFDM system performs this statistical transformation, and therefore results in a
constant degradation in the BER rate equal to 3 dB because of decreasing of variance
to the half of that of the conventional OFDM system (
)
4.6.2 Exponential Companding Algorithm:

A nonlinear companding algorithm, called exponential companding, developed to
reduce the high (PAPR) of (OFDM) signals. Exponential companding technique
adjusts both large and small signals and can keep the average power at the same level.
By transforming the original OFDM signals into uniformly distributed signals, the
exponential companding schemes can effectively reduce PAPR for different
modulation formats and sub-carrier sizes [15]. Let | | be the
power of the
- .The
amplitude of companded signal, have a uniform distribution in the interval ,
exponent is called the degree of a specific exponential companding scheme the
CDF of | | is simply
( )
The amplitude of the | | of companded signal has the following CDF
|
|(
*| |
*| |
(4.3)
(4.4)
(4.5)
(4.6)
The inverse function of

|
|(
| | (x)
(4.7)
On the other hand, given that ( )is a strictly monotonic\ increasing function, we
have,
|
|(
*| |
(4.8)
* (| |)
|
|(
( )+
(4.9)
( )) ,
( )
(4.10)
Considering the phase of input signals, the companding function ( ) is given by:
( )
( )
| |
(4.11)
51
Chapter Four
( ) 0
)1
(4.12)
Where
( ) is sign function. d is the degree of companding scheme, is
the
variance of input signal applied for companding. The positive constant determines
the average power of output signals. In order to keep the input and output signals at
the same average power level, we let
| ]
[|
(4.13)
|
)]
At the receiver side, the inverse function

operation
( )
( )
( ) of is used in the de-companding
(4.14)
Figure 4.5 shows the exponential companding function
( ) with degree as a
parameter. The companded signals have uniformly distributed amplitudes and powers,
respectively for the cases
and
.
When
, the ( ) can compress large input signals and expand small signals
simultaneously. While the -law companding scheme can only enlarge small signals
and does not change the signal peaks, which leads to a higher average power level of
output signals. As seen, the differences between exponential companding functions
are ignorable when
[15] [153].
Figure 4.5 The exponential companding function h(x).

52
Chapter Four
4.6.3 Trapezoidal power companding:

Is a nonlinear companding technique called trapezoidal power companding to
reduce the high PAPR in a complex OFDM by transforming the original signals into
new signals whose power is trapezoidally distributed. A flexible parameter is used to
determine the shape of the trapezium so that the trapezoidal power companding
scheme is able to meet the requirements for various conditions. Given an expected
PAPR value, the scheme provides a closed-form solution that guarantees the actual
PAPR the same as the expected [29].
A flexible trapezoidal design was introduced in [29], [154], transforming the
amplitude of the signals into a distribution of various trapezoidal shapes.
Since that scheme is based on the assumption that all signals are purely real or
imaginary, consequently, when the design is used in a complex system, the
theoretically estimated PAPR would be different than the actual value.
This companding scheme has three desired properties mentioned above. It converts
the power distribution of the original signals (as opposed to the amplitude used in
[29]) into a trapezoidal distribution while keeping the average output power the same
as the original signals. A parameter is used to determine the slope of the hypotenuse
so that the trapezium could have a different shape
The companding function ( ) is given by:
( )
| |
| |
{|
(4.15)
| |
(
|
Where k is the slope of the trapezium. Is the maximum power

The decompanding function at the receiver can be given as:
( )
| |
{|
| | ( | |
| |
)|
(4.16)
)|
Notice that when

, the power distribution is actually a rectangular distribution,
which is the same as the
case in exponential companding .Since sometimes a
received signal is so distorted that the square root part in (3.35) would be an
imaginary or complex number, we then take the absolute value of the square root
parts to eliminate any further potential phase distortion.
When
the flexible trapezoidal companding curve is then the same as the EC
curve [29].
4.6.4 Hyperbolic tangent (

The hyperbolic tangent (
( )=
) companding [30]:
) companding function is defined by
(4.17)
53
Chapter Four
Where
and
the envelope x.
are positive numbers controlling the companding level applied to
4.6.5 Error Function (

The error function (
( )=
Where
and
) Companding [30]:
) is defined by
(4.18)
are positive numbers controlling the level of companding
4.6.6 Logarithm Function (log) Companding [30]:

The logarithm (
( )=
Where and
[83, 84]
) companding function is defined by

)
(4.19)
are two positive numbers controlling the amount of companding.
4.7 Coding techniques:

Many early papers considered how standard coding techniques could be applied to
OFDM. The basic premise of coding is to insert redundant bits into the data stream
which can be used for error correction at the receiver. Their application to PAPR
reduction is in creating sequences of bits which will exhibit low PAPR after the IFFT.
There are 2 types of error detection and correction codes, block codes and
convolutional codes. Most papers relate to the block coding family for PAPR
reduction. During the encoding process k information bits are encoded into n code d
bits, therefore (n-k) redundant non information bits are added to the k information bits
[78].The block code is referred to as an (n,k) code, and the rate of the code as Rc=k/n.
Figure 4.6 is a block diagram showing where coding for PAPR reduction is located in
an OFDM transmitter.
4.6 Block diagram of OFDM transmitter showing PAPR coding

Different codes exhibit different degrees of error correction ability. Another important
property of codes is the weight of the code, which is the number of non-zero elements
in the codeword. Types of block codes are Hamming, Golay, and Reed- Solomon,
some of which are used for PAPR reduction .
54
Chapter Four
The basic idea of all coding schemes for the reduction of PAPR is to reduce the
occurrence probability of the same phase of N signals. The coding method selects
such code words that minimize or reduce the PAPR. It causes no distortion and
creates no out-of-band radiation, but it suffers from bandwidth efficiency as the code
rate is reduced. It also suffers from complexity to find the best codes and to store
large lookup tables for encoding and decoding, especially for a large number of
subcarriers [10].
A simple block coding scheme was introduced by Jones et al.[155], and its basic idea
is that mapping 3 bits data into 4 bits codeword by adding a Simple Odd Parity Code
(SOBC) at the last bit across the channels. The main disadvantage of SOBC method is
that it can reduce PAPR for a 4-bit codeword [109]. Later, in 1996 Wulich applied the
Cyclic Coding (CC) to reduce the PAPR [156]. In 1998, Fragiacomo proposed an
efficient Simple Block Code (SBC) to reduce the PAPR of OFDM signals [157].
However, it is concluded that SBC is not effective when the frame size is large.
Subsequently, Complement Block Coding (CBC) and Modified Complement Block
Coding (MCBC) schemes were proposed to reduce the PAPR without the restriction
of frame size [158][159]. CBC and MCBC are more attractive due to their flexibility
on choosing the coding rate, frame size and low implementation complexity. CBC and
MCBC utilize the complementary bits that are added to the original information bits
to reduce the probability of the peak signals occurrence. To make comparisons, some
results of the PAPR reduction obtained with different coding schemes have been
shown in Table 4.1, in which the number of subblock is 2 and the coding rate
for MCBC.
Table 4.1 PAPR Reduction comparison with different coding schemes
About 3-dB PAPR reduction can be obtained when coding rate

by using
CBC with long frame size. It is also shown that the PAPR reductions obtained with
(
)
(
)
CBC when coding rate
are almost the same as that when
. In
addition, when coding rate is 3/4, more than 3-dB more PAPR reduction can be
obtained using MCBC than the other schemes with any frame size. The flexibility in
coding rate choice and low complexity makes the proposed CBC and MCBC schemes
attractive for OFDM systems with large frame sizes and high coding rates [109].
The [160][161][162] authors used the Golay complementary sequences to achieve the
PAPR reduction, in which more than 3-dB PAPR reduction had been obtained. Codes
55
Chapter Four
with error correcting capabilities has been proposed in [163] to achieve more lower
PAPR for OFDM signals by determining the relationship of the cosets of Reed-Muller
codes to Golay complementary sequences. While these block codes reduce PAPR,
they also reduce the transmission rate, significantly for OFDM systems with large
number of subcarriers. In fact, let C be a code defined over an equal energy
constellation, R denotes the rate and L denotes the length of the C, respectively, then
C has
possible codewords. Therefore, it is possible to compute the codewords
with large PAPR by trying all the codewords of C and computing the peaks of the
corresponding signals at some selected time points [109].
However, it is little hope for computing the PAPR of an arbitrary code when L is
large. Even if it is possible, the complexity is still too high. Based on this motivates,
authors of [159] proposed a novel method of computation and reduction of the PAPR
and it mainly introduced a specific phase shift to each coordinate of all possible
codewords where phase shifts are independent of the codewords and known both to
transceiver, then it can be freely obtained more 4.5-dB PAPR reduction by using the
optimized phase shifts. From this viewpoint, we also consider the coding scheme of
PAPR reduction as a special phase optimization. In summarization, the inherent error
control capability and simplicity of implementation make coding method more
promising for practical OFDM systems design. However, the main disadvantage of
this method is the good performance of the PAPR reduction at the cost of coding rate
loss.
Coding techniques for PAPR reduction where redundant bits are added to the bit
stream before the IFFT. Properly chosen, these codewords ensure that the PAPR after
the IFFT is kept low. These codes can be combined with existing COFDM to reduce
the redundancy and complexity inherent in coding. A disadvantage of coding is that
the complexity becomes prohibitively high with an increase in the number of
subcarriers (>32). Various codewords were presented such as cyclic codes, ShapiroRudin Sequences, Golay Complementary codes, and Reed-Muller codes. Golay codes
and their subset, second order Reed Muller codes were found to have excellent PAPR
properties restricting the PAPR to 3dB. This reduction could be traded off with
reductions in complexity and the code length. Still complexity remains a restrictive
issue in coding [78].
4.8 Selective Mapping (SLM):

In SLM, the basic idea is to generate a set of OFDM signals, all of them representing
the same data block, and then transmitting the one with the lowest PAPR [9][10]. The
major drawback of SLM method is that it is more computationally complex because
more than one IFFT blocks are used. It also decreases the data rate because the
selected signal index, called side information, must also be transmitted to allow for
the recovery of the original data block at the receiver side. The eventual loss of the
side information during transmission significantly degrades the error performance of
the system since the whole data block is lost in this case. Therefore, a lot of work has
been suggested as a modified SLM to reduce the computational complexity [164] and
to reduce or to remove the side information transmitted [125].
In SLM, the input data sequences are multiplied by each of the phase sequences to
generate alternative input symbol sequences. Each of these alternative input data
sequences are then applied to IFFT operation, and then the one with the lowest PAPR
is selected for transmission [165]. A block diagram of SLM techniques is shown in
56
Chapter Four
Figure 4.7. The input data is partitioned into a data block Y of length N. Then these
data block is multiplied element by element with phase sequence ( )
(
( )
(4.20)
resulting into U modified data blocks
( )
( )
- where
(4.21)
After that, the N-point IFFT of each data block

signal is given as
( )
( )
is taken, the resulting OFDM
(4.22)
Among the OFDM data blocks ( )

, only one with the lowest
PAPR is selected for transmission and the corresponding selected phase
factor
also transmitted to receiver as side information. For implementation
of SLM OFDM systems, the SLM technique needs U- IFFT operation and the number
of required bits as side information is ,
- for each data block. Therefore, the
ability of PAPR reduction in SLM depends on the number of phase factors and the
design of the phase factors. The major drawback of SLM method is that it is
more computationally complex and less bandwidth efficient (side information is
required). Therefore, a lot of work has been suggested as a modified SLM to reduce
the computational complexity and to reduce or to remove the side information
transmitted [10].
Figure 4.7 Block diagram of selective mapping (SLM) technique for PAPR reduction
4.9 Partial Transmit Sequence (PTS) :

In PTS, the original data block is divided into multiple non-overlapping sub-blocks.
Then these sub-blocks are rotated with rotation factors which are statistically
independent. After that, the signal with the lowest PAPR is chosen for transmission.
57
Chapter Four
There are several ways for the partition of the data sequence into multiple sub-blocks,
including adjacent partition, interleaved partition and pseudorandom partition [122].
Among them, pseudo-random partitioning has been found to be the best choice.
Similar to SLM, the major drawback of PTS is also the computational complexity
(search complexity for optimal phase factor, and more than one IFFT blocks) and low
data rate (required side information). Several techniques have been proposed in the
literature to reduce the search complexity and overhead (by reducing/avoiding the
usage of side information) [166]. The complexity of PTS is less than SLM [167].
In PTS method, the original frequency-domain data sequence is divided into multiple
disjoint sub-blocks, which are then weighted by a set of phase sequences to create a
set of candidates Finally, the candidate with the lowest PAPR is chosen for
transmission [122]. A block diagram of PTS techniques is shown in Figure 4.8
The input data block in Y is divided in to M disjoint sub-blocks, which are
represented by the vectors { ( )
+ The input data block Y can be
( )
written in terms of
as
( )
for
(4.23)
Where, ( )
with
=
or 0
( )
After that, the sub-blocks
are transformed into M, time-domain partial transmit
sequences by taking the IFFT of length N. These partial transit sequences can be
written as:
( )
( )
] for
(4.24)
These partial sequences ( ) are then independently rotated by phase factors

*
, for
The rotated partial sequences are then
optimally combined to obtain the OFDM signals with lowest PAPR[10] . The time
domain signal after combining is given by
( )
(4.25)
There are two main issues of any PTS scheme: to reduce the computational
complexity for searching the optimal phase factors and to reduce the overhead by
minimizing the side information. Suppose that there are W phase angles to be
allowed, thus
can has the possibility of W different values. Therefore, there are
alternative representations for an OFDM symbol. The search complexity
increases exponentially with the number of sub-blocks M To reduce the search
complexity and overhead (by reducing/avoiding
the
usage
of
side
information)[166].These methods achieve significant reduction in search
complexity with marginal PAPR performance degradation. In 2007, R. J. Baxley
et.al [167] gave a useful comparison between PTS and SLM techniques. It has been
shown that the PTS outperforms SLM in terms of PAPR reduction at the cost of
increase side.
58
Chapter Four
Figure 4.8 Block diagram of partial transmit sequence (PTS) technique for PAPR
Reduction
4.10 Tone Reservation :

In TR subcarriers, called Peak Reduction Tones (PRTs) [168], are set aside for PAPR
reduction as shown in the transceiver block diagram in Figure 4.9.
Tone reservation implemented a projection onto convex sets (POCS) method. Later,
Tellado and Cioffi [169] discussed the idea of tone reservation as a linear
programming problem that has an exact solution (the POCS method is suboptimal).
The linear programming solution can be reached with complexity O [N log N].
The idea behind tone reservation is to isolate energy used to cancel large peaks to a
predefined set of tones. These tones do not bear any useful information and are
orthogonal to the data bearing tones. This orthogonality makes recovering the data
trivial [100].
The advantages of TR technique include:
1. No need for side information
2. Fewer complex-multiplications as only one time IFFT operation is needed. But
multiple iteration operations are needed after IFFT operation.
3. No special receiver operation is needed
While promising, tone reservation has several shortcomings. First the data rate is
necessarily decreased because some tones are used strictly for PAR reduction. The
second problem is the difficulty of selecting which tones to reserve. A random search
over all the possible sets, B, would greatly increase the complexity of tone
reservation.
Often the tones have to be chosen contiguously because fades often affect contiguous
sets of sub carriers. These contiguous sets of tones are known to have bad PAR
reduction abilities. The third issue is a tradeoff between the quantities of reserved
tones and the rise in average power due to tone reservation. More the tones that are
reserved, lesser the power needs to be allocated for PAPR reduction. On other hand,
more reserved tones mean more unused bandwidth that could be data bearing [100].
59
Chapter Four
Figure 4.9 Block diagram of a Tone Reservation (TR) OFDM transceiver.
4.11 Tone Injection:

Motivated by the data rate loss of tone reservation, Tellado introduced a new
technique named tone injection [170] as shown in figure 4.10. It reduces the PAPR
without compromising the data rate. In this method the size of the basic constellation
is increased. Hence mapping of original constellation points into numerous
corresponding points in the new stretched out constellation becomes possible. The
distance between these duplicate points can be calculated as
d , where M=
constellation size, and
.
There is no effect on BER and all we have to do is add a modulo-D subsequent to FFT
in the receiver side. Since mapping of each information unit into numerous
corresponding constellation points is done, it gives a margin of free will which can be
used reduction of PAPR [100]
Figure 4.10 Block diagram of a Tone Injection (TI) OFDM transceiver
60
Chapter Four
4.12 Interleaving [171][172]:

This method is also termed as Adaptive Symbol Selection Method .Multiple OFDM
symbols are created by bit interleaving of input sequences .The basic Idea is to use W
interleaving ways and selecting one with the lowest PAPR.
Figure 4.11 shows an interleaver, PAPR Reduction capability depends on the number
of interleaver used .To recover the signals the receiver need to know the information
about which interleaver is used.
Figure 4.11: Interleaving
4.13 Active Constellation Extension (ACE) [173][174]:

This technique deals with extending the constellation points outside the signal
constellation which is then used to cancel the time domain peaks .Figure 4.12 shows
the points where these constellation points can be extended. Is technique has several
advantages like no loss of data, no degradation in system performance, lower BER as
compared to other techniques and bears no side information like SLM. Some
variations of this method like clipping-based ACE and Adaptive ACE in which
repeated CAF an in later an adaptive control has been used to optimize the
performance.
The drawback is that the technique is useful for larger constellation size modulations
only.
Figure 4.12 Active Constellation Extension (a) for QPSK (b) for 16 QAM
61
Chapter Four
4.14 Dummy Sequence Insertion (DSI)[100]:

In Dummy sequence insertion (DSI) [175], before IFFT stage in input data a dummy
sequence is added. The sequences which are used may be complementary, correlation
or any other sequence. Since dummy sequence is not used as side information hence
any transmission error does not increase BER. DSI technique is united with PAPR
threshold method. After IFFT, if PAPR is below specific threshold then signal is
transmitted but if it is more than this specific level then insertion of dummy sequence
is done to achieve the required results. The block diagram of DSI system is shown in
figure
Figure 4.13 Block diagram of DSI system

The main advantage of this technique is that BER is not degraded due to transmission
errors in the dummy sequence. So far amongst different sequences, use of
complementary sequence produces better results.
62
Chapter Five
Simulation Results and Analysis

Chapter Five
One of the major drawbacks of OFDM system is high PAPR of transmitting signals,
which causes an earnest degradation in performance when a non-linear HPA is
utilized. Therefore, it is compulsory to utilize a congruous PAPR reduction scheme at
the transmitter. In this chapter, the different methods of PAPR reduction are given
with results and new types of PAPR proposed.
5.1 OFDM System model:

The system model used in the work is shown in figure 5.1. The OFDM parameter
used in the test is the LTE parameters as shown in table 5.1. The system was tested
under Rayleigh selective fading channel with parameter given in table 5.2 [176]
I/P
S
/
P
Signal
Mapper
IDFT
OR
IFFT
P
/
S
Add
CP
D
/
A
Multipath Fading Ch.

& noise
O/P
P
Signal
demapper
Equalizer
And
P/S
DFT
OR
FFT
S
/
P
Remove
CP
A
/
D
Figure 5.1 OFDM system model.

Table 5.1 LTE parameter
FFT size
Spacing frequency
BW
CP
No symbol
Sampling frequency
Modulated type
128
15 KHz
1.25MHz
32
1000
192MHz
QPSK
Table 5.2 Average Power and Relative Delays with 6 delay taps [176]
Tap no.
Relative delay (ns)
Average Power (dB)
1
0
0.189
2
0.2
0379
3
0.5
0.239
4
1.6
.095
5
2.3
.061
6
5
.037
63
Chapter Five
The PAPR was evaluated statistically by using the complementary cumulative

distribution function (CCDF). The CCDF of PAPR, for the proposed PAPR reduction
techniques OFDMA downlink signal, is used to express the probability of exceeding a
given threshold PAPR0 (i.e., CCDF (
)). A simulation result was
compared with each other. PAPR was measured for the transmitted OFDM signal
using the equation:
| |
(5.1)
| |
In each case, the BER was measured.

Initially, it is necessary to know the performance of OFDM system without any PAPR
reduction techniques in order to compare it with the PAPR reduction techniques to
find out the amount of improvement in PAPR in each case of PAPR reduction
techniques and their impact on the BER. Fig (5.2) shows the CCDF of PAPR and
SNR at BER
for OFDM system without any PAPR reduction techniques which
is equal to (10.84 dB) with PAPR equal to (25.6015 dB) while shows the BER for
OFDM system without any PAPR reduction techniques and SNR at BER
is
equal to (11.4314 dB).
0
10
CCDF (Pr[PAPR>PAPR0])
Orignal
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
Figure (5.2.a)
Bit error probability curve for qpsk using OFDM
10
simulated
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
Figure (5.2.b)
a) is CCDF of PAPR for OFDM system without any PAPR reduction techniques b)
is BER for OFDM system without any PAPR reduction techniques
64
Chapter Five
5.2 PAPR techniques used:

5.2.1 Repeated clipping and frequency domain filtering (RCF):
In the clipping technique hard limiting is applied to the amplitude of the complex
values of the IFFT output. The filtering technique is designed to alleviate or cancel
OOB distortion dependent on oversampling value but cannot correct in-band
distortion. [98]
Iterative clipping and filtering fft/iffft
Zeroes
0
0
N*
Nonlinear
N*
Point
inverse
DFT
over
Processing
Point
DFT over
Clipping
sampling
rate
(
Zeroes
rate
Ratio =
sampling
N*
)
0
0
Point
inverse
DFT
over
sampling
Add
cp
rate
CR
Input data
zero padded
Interpolated
baseband signal
Clipped
Interpolated
baseband signal
Frequency domain
filtering
Figure 5.3 shows the block diagram of the new PAPR reduction scheme [177].
The input vector
is first converted from the frequency to the time domain
wing an oversize IFFT. N is the number of subcarriers in each OFDM symbol. For an
oversampling factor of, the input vector is extended by adding (
) zeros; in the
middle of the vector. This results in the trigonometric interpolation of the time domain
signal [178].
Trigonometric interpolation gives perfect interpolation when the original signal
consists of integral frequencies over the FFT window. This is the case for OFDM. The
input of the Nyquist frequency,
has been omitted, as the interpolation technique
does not work for this value [178]. This is not a practical limitation as all applications
of OFDM null this input and most do not use a number of adjacent subcarriers. The
interpolated signal is then clipped.
In this Technique hard-limiting is applied to the amplitude of the complex values of
the IFFT output [12]
After an IFFT, the original signal is clipped in the time domain. The clipping can be
described as shown below:
*
Where
,| | -
| |
| |
| |
(5.2)
represents the output of the time domain signal,

,| | -
(5.3)
| |
,
Is the threshold clipping level,
power.
65
; ,| | - Is the mean
Chapter Five
The clipping ratio is defined as the ratio of the clipping level to the mean power of
the unclipped baseband signal.
As shown in the equation (5.2), the discrete time domain signal is clipped in the
amplitude. At every point where the complex time domain signal exceeded the
clipping level, the amplitude was reduced to the clipping level while the phase of the
complex signal was unchanged [179].
The clipping is followed by frequency domain filtering to reduce OOB power caused
by clipping. The filter consists of two FFT operations [12].
The clipped time domain signal c is then converted back into the discrete frequency
domain using an FFT ,The inband discrete frequency components of the clipped
signal
are passed unchanged to the inputs of the second
IFFT while the OOB components,
are nulled [13 and 180] this
technique is repeated, depending on iteration number, usually choose between one

and four.
In this work has been selected four.
Although frequency domain filtering is a common signal processing technique the
form shown in figure 5.3 is unusual. In most filtering applications the filter is
designed to meet particular specifications in the continuous frequency domain. In this
application, the wanted signal is an OFDM signal, which is the sum of discrete
frequency components in each symbol period. The filter must therefore have as little
effect as possible on the in-band discrete frequency domain while attenuating as much
as possible any OOB components. This is precisely what is achieved by the simple
filter structure in Figure 5.3 because the filter operates on a symbol by symbol basis;
there is no filtering across symbol boundaries and so no resultant ISI. The filtering
does cause some peak regrowth. However, this is much less than for clipping before
interpolation [12, and 18]
The clipping noise is added at the transmitter rather than the receiver. In fading
channels this means that in general the clipping noise will cause less degradation in
bit error rate than noise added in the channel since the clipping noise fades along with
the signal.
However the second oversize IFFT could be replaced by any of the transform, up
sampling and filtering arrangements commonly used in OFDM systems. So the
technique can be implemented by replacing the IFFT block in an existing OFDM
system with the new configuration [12].
The FFT/IFFT transform filter can be replaced by DCT/IDCT transform and this
technique has been described in [28].
In paper [18] present a new PAPR reduction technique which exploits the use of
unused carriers as well as the phase information of pilot signals in OFDM systems to
reduce the PAPR while not degrading channel estimation or frequency offset.
Compared to conventional techniques such as clipping and windowing, this technique
introduces much less OOB distortions and provides a lower BER since there is no
interference to the original data signals. There is also no requirement for side
information to be transmitted to the receiver.
To reduce PAPR at LTE downlink, the RCF is applied to OFDM signal for different
CR and oversampling filter and notes their impact on PAPR and BER.
The reason to choose this method is because the filter improves the BER if the
oversampling is high and clipping improves PAPR (it's possible to improve the BER
& PAPR together and this way we have explained previously).
66
Chapter Five
The OFDM system model with RCF as shown in figure 5.4.

For this simulation I = (1, pilot, 1.125, 1.25, 1.5, 2, 3, 4) and CR = (4, 3, 2, 1.75, 1.5),
in order to see the impact of CR on the (BER) and (PAPR), this technique is repeated,
depending on iteration number ( four is used in this simulation)
The transmitted signals pass through Rayleigh fading channel.
D
/
A
Add CP
P
/
S
RCF
IDFT OR IFFT
S
/
P
+pilot symbol
Signal mapper
I
/
P
Multipath
Fading Ch. &
noise
S
/
P
Remove CP
DFT OR FFT
Remove
+pilot symbol
One Tap
Equalizer
And P/S
Signal
demapper
O
/
P
A
/
D
Figure 5.4 the OFDM system model with RCF.

Figure 5.5 illustrate the effect of repetition clipping and filtering on PAPR where CR
=3, I =2, where CCDF of PAPR for, one RCF = 7.7581, two RCF = 6.5462, three
RCF = 5.8319, and four RCF = 5.401,
Note that there is an improvement in CCDF of PAPR for one RCF (2.8935 dB), two
RCF (4.1054 dB), three RCF (4.8197 dB), and four RCF (5.2506 dB). But the
proportion of improvement, between (N) RCF and (N-1) RCF decrease as N increase.
Whenever a CR reduces the PAPR is improving and contrast SNR at BER

is
increased, The best value of PAPR is for CR =1.5, but the SNR at BER
for this
case is the worst, as shown in table CR have a positive relationship with PAPR and
negative relationship with SNR at BER
Whenever oversampling increased the SNR at BER
is improving and contrast
PAPR is increased and vice versa. The best value of PAPR is for I =1 this mean there
is no filter, but the SNR at BER
for this case is the worst, while The best value
of SNR at BER
is for I =4, but the PAPR for this case is the worst, as shown in
table (A.1) I have a positive relationship with SNR at BER
and negative
relationship with PAPR
67
Chapter Five
10
Orignal
One clip and filter
Two clip and filter
Three clip and filter
Four clip and filter
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
Figure 5.5 CCDF of PAPR for OFDM system with repeated clipping and frequency
domain filtering where CR =3, I =2
Figure 5.6 shows the following:
There is a clear improvement in the CR3 CCDF of PAPR reduction in rate
SNR at BER
is relatively small compared with the CR4. Briefly, thats
mean the percentage of improvement in CCDF of PAPR More than the
degradation in BER
For the CR2 the PAPR improved more than CR3 and CR4 but SNR at BER
gets worse
the CR1.75 had a little improvement in CCDF of PAPR in comparison with
the CR2) but SNR at BER
degradation more than The amount of
improvement
For the CR1.5 the PAPR improved PAPR in comparison with the CR1.75 only
in a small proportion, while SNR at BER
Substantially worse.
Figure 5.7 shows the impact of the oversampling (CCDF of PAPR) and (BER),
is conclusion through drawing and table following:
whenever increase the PAPR will increase too only in small percentages, for
this figure PAPR for I4 Worsened by (1.7978 dB) compared with I1
whenever increase the CCDF of PAPR will increase too only in small
percentages, for this figure CCDF of PAPR for I4 Worsened by (.9939 dB)
compared with I1
whenever increase the SNR at BER
will improved , for this for I4 the
SNR at BER
improved by(5.5382) compared with I1
for I2 the SNR at BER
Improved by (2.8466) , CCDF PAPR Worsened
by (.5831) and PAPR Worsened by (.6559 dB) compared with I1
68
Chapter Five
The Conclusion from table (A.1) Summarizes as follows:

CR 4 have the best SNR at BER
and the worst PAPR compared with the
rest of the CR
PAPR at CR 3 better than CR 4 by (2-3 dB improvement in PAPR) but SNR at
BER
at CR3 worse than CR 4 only by small percentage (less than 1 dB
in all cases) PAPR at CR 2 better than CR3 by (2.5 - 3.4 dB improvement in
PAPR) but SNR at BER
at CR2 worse than CR = 3 by (2- 3.7 dB
degradation in SNR at BER
)
PAPR at CR 1.75 better than CR2 by (Maximum improvement is 1.0059) but
SNR at BER
at CR1.75 worse than CR2by (2.2 3.4 dB degradation
in SNR at BER
)
CR 1.5 have the best PAPR and the worst SNR at BER
compared with
the rest of the CR, the SNR should higher than 30 dB have the desired SNR at
BER
thats mean SNR at BER
is deteriorating by a large margin
The following conclusion when comparing the proposed method with an OFDM
system without PAPR reduction method:
As (CR =4, 3 and I =4, 3,2, pilot, 1.5,1.25) and when (CR =2and I =4) and finally
(CR =4 and I = 1.125), The best one improvement in PAPR and CCDF of PAPR
is at I =3 and CR =2. The improvement in PAPR by = (14.9490 dB), CCDF of
PAPR = (6.2850 dB), and the SNR at BER
by = (1.0134 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = Pilot and CR
=2. The improvement in PAPR by = (16.1583 dB), and CCDF of PAPR = (6.9604
dB), while the SNR at BER
deteriorated by = (-1.5686 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 1 and CR =2.
The improvement in PAPR by = (17.3529 dB), and CCDF of PAPR = (7.7214
dB), while the SNR at BER
deteriorated by = (-2.9442 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 1 and CR
=1.75. The improvement in PAPR by = (18.2213 dB), and CCDF of PAPR =
(8.2460 dB), while the SNR at BER
deteriorated by = ( -5.2886 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 1 and
. The improvement in PAPR by = (19.2177 dB), and CCDF of PAPR =
(7.9400 dB), while the SNR at BER
deteriorated by = ( -18.0686 dB).
69
Chapter Five
10
CR
CR
CR
CR
CR
-1
=4
=3
=2
=1.75
=1.5
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 5.6.a
10
CR =4
CR =3
CR =2
CR =1.75
CR =1.5
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
Figure 5.6.b
(a)CCDF of PAPR for OFDM system with RCF where I =2 (b) BER for OFDM
system with RCF where I =2
70
Chapter Five
10
I =1
I =pilot
1.125
I =1.25
I =1.5
I= 2
I =3
I =4
-1
10
-2
10
-3
10
3
PAPR0 [dB]
Figure 5.7.a
10
I =1
I =pilot
1.125
I =1.25
I =1.5
I= 2
I =3
I =4
-1
BER
10
-2
10
-3
10
-4
10
10
SNR
Figure 5.7.b
Figure 5.7 (a) CCDF of PAPR for OFDM system with RCF where CR =3
(b) BER for OFDM system with RCF where CR =3.
71
15
Chapter Five
5.2.2 Repeated frequency domain filtering and clipping RFC:

The proposed method is the same as previous method RCF, but with almost a simple
change and it changes the location of the filter becomes before the clipping as shown
in figure 5.8, the frequency domain filtering that depends on the interpolation As
noted by previous results that improve the BER As noted by previous results but
increases the PAPR. The basic idea of this method is proposed that this filter will
improve the performance of the OFDM to improve the BER and then the clipping will
improves PAPR method is the almost same as RCF, where have the same receiver and
channel But there is a difference in One block in the transmitter. This block is RFC as
shown in figure 5.9. Interpolated baseband signal followed by frequency domain
filtering, the same filter which are explained in the case of RCF. The filtering signal is
clipped in the time domain. The clipping block is described previously in the case of
RCF.
Iterative filtering fft/iffft and clipping
N*
N*
(
Zeroe
Point
DFT
over
Point
inverse
DFT
over
0
0
Zeroes
sampling
rate
Nonlinear
Point
inverse
DFT
over
Processin
g
Ratio =
rate
Input data
zero padded
Interpolated
baseband signal
Add
cp
Clipping
samplin
g rate
sampling
N*
CR
Clipped
Frequency
domain filtering
the filtering signal
Figure 5.8 shows the block of the OFDM system model for this proposed
P
/
S
Add CP
RFC
IDFT OR IFFT
S
/
P
+pilot symbol
Signal mapper
I
/
P
D
/
A
Multipath
Fading Ch. & noise
Figure 5.9 the OFDM system model with RFC.

72
S
/
P
Remove CP
DFT OR FFT
Remove
+pilot symbol
One Tap
Equalizer
And P/S
Signal
demapper
O
/
P
A
/
D
Chapter Five
system without PAPR reduction method as shown in table A.2:
At (CR =4,3 and I =4,3,2, pilot,1.5,1.25) , (CR =2and I =4,3) , ( CR =1.75 and I =
4) and finally (CR =4 and I = 1.125), The best one improvement in PAPR and
CCDF of PAPR is at I =4 and CR =1.75. The improvement in PAPR by = (18.2789
dB), CCDF of PAPR = (8.0187 dB), and the SNR at BER(
) by = (0.6101 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 3 and CR
(8.0088 dB), while the SNR at BER(
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I =2 and CR
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 1.5 and CR
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 1.125 and CR
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 4 and CR =1.5.
) deteriorated by = ( -16.7886 dB).
The following conclusion when comparing the proposed method with RCF:
1. CCDF of PAPR was improved in all cases except when (I = pilot and CR=4
by (-0.3570)). The improvement ratio was increased with the decrease of CR
and the increase of I. The biggest improvement is in the case (I = 3 and CR =4
(2.9062))
2. PAPR was improved in all cases except when (I = 1.125 and CR=2 by (0.0866)) and (I = 1.5 and CR=4 by (-0.0015)) the improvement ratio was
increased with the decrease of CR and the increase of I. The
biggest improvement is in the case( I = 3 and CR =4 (1.5600))
3. SNR at BER(
)
a) SNR at BER(
) was improved for (I =3 and I =4 in all cases of CR )
b) For I =2 SNR at BER(
) was improved in all cases except when (CR=2)
deteriorated by (-0.1548)
c) For I =1.5 SNR at BER(
)was improved except when (CR=4) deteriorated
by (-0.1236)
d) For
I =1.25 SNR at BER(
) was improved except when (CR=4)
deteriorated by (-0.5700) and (CR=2) deteriorated by (-0.0375)
e) For
I =1.125 SNR at BER(
deteriorated by (-0.1700) and (CR=1.75) deteriorated by (-0.2390)
f) For
I =pilot SNR at BER(
deteriorated by (-0.2400) and (CR=3) deteriorated by (-0.1585)
73
Chapter Five
g) The best value of the improvement is where the (I =4 and CR =1. 75 by

(2.1787))
4. RFC is better than RCF because when I increase the SNR at BER(
)
improve and PAPR almost preserves its value
That was the conclusion from a comparison of figure 5.6 and figure 5.10 the
following:
1. There is an obvious improvement in the CCDF of PAPR of the RFC In
comparison with the RCF
2. There is an improvement in the SNR at BER(
) of the RFC In comparison
with the RCF
3. The CCDF of PAPR of the RFC at CR=2 is better than the CCDF of PAPR of
the RCF at CR=1.5, in addition to that the SNR at BER(
) of the RFC at
CR=2 is better than the SNR at BER(
) of the RCF at CR=1.5 by ((17.71721
dB)
4. The amounts of improvement, are described in the table A.2
Figure 5.11 shows the impact of the oversampling (CCDF of PAPR) and (BER), is the
conclusion through drawing and table following:
The PAPR for I(N) was increased only by A small amount compared with I1,
for this figure PAPR was declined amount (0.0251 - 0.3086 dB).
The CCDF of PAPR for I(N) was increased only Avery small amount could
be neglected in comparison with I1, for this figure PAPR was declined
amount ( 0.0102 - 0.0837 dB).
SNR at BER(
) degraded whenever I increased between (0.8374 6.2451dB)
RFC have the same complexity and cost RCF because RFC has not added a new
function for RCF but the only change filter location.
74
Chapter Five
10
CR =4
CR =3
CR =2
CR =1.75
CR =1.5
-1
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 5.10.a
10
CR =4
CR =3
CR =2
CR =1.75
CR =1.5
-1
BER
10
-2
10
-3
10
-4
10
10
15
snr
20
25
30
Figure 5.10.b
Figure 5.10 (a)CCDF of PAPR for OFDM system with RFC where I =2 (b) BER for
OFDM system with RCF where I =2
75
Chapter Five
10
I =pilot
I =1.125
I =1.25
I =1.5
I= 2
I =3
I =4
-1
10
-2
10
-3
10
0.5
1.5
2.5
3
PAPR0 [dB]
3.5
4.5
Figure 5.11.a
10
I =pilot
1.125
I =1.25
I =1.5
I= 2
I =3
I =4
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
Figure 5.11.b
Figure 5.11 (a)CCDF of PAPR for OFDM system with RCF where CR =3 (b) BER
for OFDM system with RCF where CR =3
76
Chapter Five
5.2.3 The OFDM System with discrete time companding:

Compresses the signal at the input and expands the signal at output in order to keep
the signal level above the noise level during processing. At the output, the original
input signal is then restored by a simple attenuation. Companding increases the SNR
when the input signal is low and therefore reduces the effect of a systems noise
source.
5.2.3.1 A-law companding:

In this companding method, the compressor characteristic is piecewise, made up of a
linear segment for low level inputs and a logarithmic segment for high level inputs.
Figure 5.12 shows the A-law compressor characteristics for different values of A.
Corresponding to A=1, we observe that the characteristic is linear (no compression)
which corresponds to a uniform quantization. A-law has mid riser at the origin. Hence
it contains non-zero value. The practically used value of A is 87.6. The A-law
companding is used for PCM telephone systems. The linear segment of the
characteristic is for low level inputs whereas the logarithmic segment is for high level
inputs. This technique can be used to reduce the PAPR which is the main
disadvantage of OFDM [181, and 182].
| |
( )
0
(
| |
( )
| |
1]
(5.4)
| |
( )
Where
x=input signal.
y=output signal.
( ) =sign of the input (+ or -).
|x|=absolute value (magnitude of x).
A=87.6 (defined by CCITT (Consultative Committee for International
Telephony and Telegraphy) ).
This A-law companding technique is used in Europe, Asia, Russia, Africa, China, etc
[183].
Initially, A companding as discussed used with OFDM.
Figure 5.13 illustrates the effect of A parameter on the PAPR, CCDF of PAPR, and
SNR at BER(
). When increasing the values of A parameter, the CCDF of PAPR
improves. The relationship between A parameter and CCDF of PAPR is the inverse
relationship.
CCDF of PAPR (A =20) - CCDF of PAPR (A =120) = (1.15 dB)
A is not linear companding , A possible divided into three areas. The first area is that
when A increases lead to improvement in the CCDF of PAPR is relatively large
compared with the second and third region (the example for this area is A
(CCDF of PAPR (A =5) - CCDF of PAPR (A =20) = (2.955 dB) amount of
improvement in the CCDF of PAPR )
In the second area, when A was increased the CCDF of PAPR was improved but a
small quantity less than the first region example of this area when A
(CCDF of
PAPR (A =20) - CCDF of PAPR (A =120) = (1.15 dB) as is evident A increased by
77
Chapter Five
(100) and the improvement in CCDF of PAPR is (1.15 dB) while in the first area A
increased by (15) but the improvement in CCDF of PAPR (2.955 dB))
In the third area, when A was increased the CCDF of PAPR was not affected even if
improved but very small.
Normalized
Output
Normalized input
Figure 5.12. A-law Compressor Characteristics [99].
22
20
SNR at (BER =10-4)

CCDF of PAPR
PAPR
18
16
[dB]
14
12
10
8
6
4
2
20
40
60
A
80
100
120
Figure 5.13 the relationship between A parameter and (PAPR, CCDF of PAPR and
BER)
78
Chapter Five
When A parameter was increased the PAPR improved while the SNR at BER(
)
deteriorated. The A parameter has a positive relationship BER with but an inverse
relationship with the PAPR.
PAPR (A =5) PAPR (A = 20) = (4.466 dB)
SNR at BER(
) (A= 5) SNR at BER(
) (A=20) = (-4.4 dB)
PAPR (A =20) PAPR (A = 120) = (3.03 dB)
SNR at BER(
) (A= 20) SNR at BER(
) (A=120) = (-3.386 dB)
In the first area A increased by (15) but the improvement in PAPR (4.466 dB) and the
degradation in SNR at BER(
) (4.4 dB) while in the second area A increased by
(100) and the improvement in PAPR (3.03 dB) and the degradation in SNR at
BER(
) (-3.386 dB) . The first and the second area evident in the Figure 5.14
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at A =5. The
improvement in PAPR by = (6.6954 dB), and CCDF of PAPR = (4.200 dB), while the
SNR at BER(
For SNR at BER(
)
improvement in PAPR by = (10.9098 dB), and CCDF of PAPR = (6.1100 dB),
while the SNR at BER(
For SNR at BER(
)
For SNR at BER(
)
Figure 5.14 shows the CCDF of PAPR and the BER of A companding for various A
parameter. For more details see table A.4
79
Chapter Five
10
original
A =5
A =10
A =30
A =50
A =87.6
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
5.14.a
10
A =5
A =10
A =30
A =50
A =87.6
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
5.14.b
Figure 5.14 (a) CCDF of PAPR OFDM system A companding for various A
parameter. (b) BER for OFDM system A companding for various A parameter.
80
Chapter Five
5.2.3.2 -law companding technique:

In the -law companding, the compressor characteristic is piecewise, made up of a
linear segment for low level inputs and a logarithmic segment for high level inputs.
Figure 5.15 shows the -law compressor characteristics for different values of .
Higher the value of more is the compression.
Corresponding to =0, we observe that the characteristic is linear (no compression)
which corresponds to a uniform quantization. -law has mid tread at the origin. Hence
it contains a zero value. The practically used value of is 255 [183].
The signal by utilized -Law compression characteristic is defined as:
( )
| |
( )
(5.5)
Where V is the peak amplitude of the signal, and x is the instantaneous amplitude of
the input signal. Decompression is simply the inverse of (5.5). Compression improves
the quantization resolution of small amplitude signals at the cost of lowering the
resolution of large signals. This also introduces quantization noise; however, the
effect of the quantization noise due to reduction in resolution of the peaks is relatively
small as the peaks occur less frequently. The compression algorithm as described by
amplifying the signals of lower amplitude with the peaks remaining unchanged. [184],
[185].
Normalized
Output
Normalized input
Figure. 5.15 -law Compressor Characteristics [186]

A-law and law coefficients are responsible for the compression ratio. Compression
increases with increasing value of the coefficients. Originally A-law and -law
companders were used for voice compression, as it can be seen, A-law and -law
companders have logarithmic compressing profile. In fact they work as follows,
instead of compressing the high peaks; companding schemes increase the value of
small signals in a way, to bring them in the same level with the high peaks [186].
81
Chapter Five
Thus original Gaussian distributed OFDM signal will be transformed to a signal with
quasi uniform distribution. However, because of increased level of the small signals,
average power of the signal will be increased. That means noise will be increased as
well. This is disadvantage of A-law and -law companding schemes as compared
with exponential companding, which is claimed to adjust both small and large signals
without changing the average power of the signal [187].
Figure 5.16 illustrates the effect of parameter on the PAPR, CCDF of PAPR and
SNR at BER(
). In General, when parameter was increased, the CCDF of PAPR
was decreased except for some cases are as follows:
At
the CCDF of PAPR Larger than the
by (0.06 dB)
At
by (0.014 dB)
At
by (0.205 dB)
At
by (0.01 dB)
At
by (0.09 dB)
At
by ( 0.11 dB )
Even in exceptional cases, the amount of the decline is a few and not exceed
(0.205 dB)
The max CCDF of PAPR at (
) = 6.416 dB while the min CCDF of PAPR at
(
) =2.17 dB
The parameter has a positive relationship SNR at BER(
) with but an inverse
relationship with the PAPR.
When parameter was increased SNR at BER(
) deteriorated except at
be
better than the
by (0.132 dB).
In General, when parameter was increased, the PAPR was decreased except for
some cases are as follows:
At
the PAPR Larger than the
by (0.7103 dB)
At
by (2.3264 dB)
At
by (4.6422 dB)
At
by (0.2429 dB)
At
by (0.3182 dB)
At
by (0.436 dB)
At
by (2.0735 dB)
The max PAPR at (
) = 17.4332 dB while the min PAPR at (
) =10.8218
dB
The max SNR at BER(
) at (
) = 23.764 dB while the min SNR at
BER(
) at (
) =13.3363 dB
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at MU =10. The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at MU =20 . The
82
Chapter Five
For SNR at BER(

)
For SNR at BER(
)
improvement in PAPR by = ( 14.7797 dB), and CCDF of PAPR = (8.6700 dB),
Fig (5.17) shows the CCDF of PAPR of

companding and the BER of
companding for various parameter and illustrates the former explanation.
PAPR improved by ( 8.1683 -14.7797 dB)
CCDF of PAPR improved by (4.4240 - 8.6700 dB)
The amount of SNR at BER(
) degradation is (1.9049 - 12.3326dB )
22
20
18
16
[dB]
14
12
10
SNR at (BER =10-4)
CCDF of PAPR
PAPR
8
6
4
2
50
100
Figure 5.16 the relationship between

BER)
150
MU
200
250
300
parameter and (PAPR, CCDF of PAPR and
83
Chapter Five
10
orginal
MU =5
MU =50
MU =100
MU =160
MU =200
MU = 255
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
Figure 5.17.a
10
MU =5
MU =50
MU =100
MU =160
MU =200
MU = 255
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
5.17.b
Figure 5.17 (a) CCDF of PAPR OFDM system
companding for various
parameters. (b) The BER of companding for various parameters.
84
Chapter Five
5.2.3.3 Rooting Companding Technique (RCT):

The proposed Rooting companding has the same principle of SQRT.
Rooting companding equation is given by:
( )
| |
( )
(5.6)
Where
( )= sign(x)
sign(x) was used in RCT to maintain the phases of the OFDM signal Where the
phases of the OFDM output signals
are kept unchanged while only the amplitudes
are treated and changed . The amount of change in amplitude depends on the value of
R
Rooting decompanding equation is given by:
( )
| |
( )
(5.7)
The following can be observed from table A.6 and figure 5.18
When y parameter decreases the PAPR and CCDF of PAPR also decrease
while SNR at BER(
) increase
The best value for the PAPR is (2.8726) when R =0.1 while the worst value is
(21.8631) when R=0.9
The best value for the CCDF of PAPR is (1.268) when R =0.1 while the worst
value is (9.55) when R =0.9
The best value for the SNR at BER(
) is (11.6765) when R =0.9 while the
worst value is (28.3) when R=0.1
PAPR improved by (3.7384 - 22.7289 dB )
CCDF of PAPR improved by (1.2900 -9.5720 dB )
) degradation is (0.2451 - 16.8686 dB )
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at R =0.7. The
improvement in PAPR by = ( 7.4724 dB), and CCDF of PAPR = ( 2.7820 dB),
For SNR at BER(
)
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at R =0.4 . The
For SNR at BER(
)
For SNR at BER(
)
85
Chapter Five
The best one improvement in PAPR and CCDF of PAPR is at

improvement in PAPR by = (19.6127 dB), and CCDF of PAPR =
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at
improvement in PAPR by = (22.7289 dB), and CCDF of PAPR =
R =0.2. The
(8.2655 dB),
R =0.1. The
(9.5720 dB),
10
original
R = .9
R = .8
R =.7
R =.6
R = .5
R = .4
R =.3
R =.2
R = .1
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
5.18.a
10
R = .9
R = .8
R =.7
R =.6
R = .5
R = .4
R =.3
R =.2
R = .1
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
5.18.b
Figure 5.18 (a)CCDF of PAPR OFDM system RCT for various
BER of RCT for various parameter
86
25
30
parameter. (b) The
Chapter Five
Figure 5.19 illustrates the effect of R parameter on the PAPR, CCDF of PAPR and
SNR at BER(
). The y parameter has a positive relationship with PAPR and
CCDF of PAPR but an inverse relationship with the SNR at BER(
).
30
SNR at (BER =10-4)
CCDF of PAPR
PAPR
25
[dB]
20
15
10
0
0.1
0.2
0.3
0.4

SNR at BER(
))
0.5
R
0.6
0.7
0.8
0.9
parameter and (PAPR, CCDF of PAPR, and
5.2.3.4 New error function Companding (NERF) :

The new type of companding was proposed depends on erf. The NERF companding
equation is:
( )
| |
( )
(5.8)
NERF De_companding:
( )
| |
/|
( )
(5.9)
When used this type of companding the PAPR was improved by (15.422 dB) and the
CCDF of PAPR also was improved by (6.4045 dB) while the SNR at BER(
)
was deteriorated by (2.2466 dB).
The rate of improvement in the PAPR and CCDF of PAPR is greater than the rate of
the decline in SNR at BER(
) as shown in figure 5.20 and table 4.3.
Table 5.3 NERF performance
NERF
PAPR
10.1795
CCDF of PAPR
4.4355
87
SNR at BER(
13.678
Chapter Five
10
Orignal
NERF
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
5.20.a
10
NERF
-1
BER
10
-2
10
-3
10
-4
10
10
12
14
SNR
5.20.b
Figure 5.20 (a) CCDF of PAPR OFDM system NERF companding (b) the BER
of NERF companding.
88
Chapter Five
5.2.3.5 Absolute Exponential companding (AEXP) :

The proposed AEXP equation is derived based on EXP companding and Trapezoidal
power companding:
( ) 0
( )
| |
/1
(5.10 )
EXP companding since received signal with EXP companding is so distorted that the
square root part in (5.10) would be an imaginary or complex number, we then take the
absolute value of the square root parts to eliminate any further potential phase
distortion.
Where
( ) is sign function? The positive constant determines the average power
output signals. In order to keep the input and output signals at the same average power
level
[| | ]
* [
(5.11 )
| |
)] +

operation
( )
( ) |
| |
( ) of is used in the De_companding
(5.12)
The following can be observed from the table (A.7) and figure 5.21
When d parameter decreases the PAPR and CCDF of PAPR also decrease
while SNR at BER(
) increase
Values less than d =.8 the deterioration in the SNR at BER(
) becomes
large
For
The best value for the CCDF of PAPR is (2.92) when d =0.8 while the worst
value is (5.1533) when d =2
The best value for the PAPR is (6.0806) when d =0.8 while the worst value is
(13.0811) when d =2
The best value for the SNR at BER(
) is (14.73) when d =2 while the
worst value is (24.833) when d =0.8
PAPR improved by (12.5205 - 19.5209 dB )
CCDF of PAPR improved by (5.6867 - 7.9136 dB )
) degradation is (3.2986 - 13.4016 dB )
89
Chapter Five
10
orginal
d=2
d = 1.8
d =1.6
d = 1.4
d =1.2
d=1
d = .8
d =.6
d = .4
d =.2
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
5.21.a
10
-1
BER
10
d=2
d = 1.8
d =1.6
d = 1.4
d =1.2
d=1
d = .8
d =.6
d = .4
d =.2
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
5.21.b
Figure 5.21 (a) CCDF of PAPR OFDM system AEXP companding for various
parameters. (b) The BER of AEXP companding for various parameters.
90
Chapter Five
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at d= 1.1. The
For SNR at BER(
)
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at d =0.8.
Figure 5.22 illustrates the effect of d parameter on the PAPR, CCDF of PAPR and
SNR at BER(
). The relationship between the d parameter and SNR at
BER(
) is a direct correlation, while the relationship between the d parameter and
PAPR, and CCDF of PAPR is the inverse relationship.
30
SNR at (BER =10-4)
CCDF of PAPR
PAPR
25
[dB]
20
15
10
0.2
0.4
0.6
0.8

SNR at BER(
))
1
d
1.2
1.4
1.6
1.8
parameter and (PAPR, CCDF of PAPR, and
5.2.3.6 Cos companding:

The new type of companding was proposed depends on cos . The proposed cos
companding eauation is:
( )
( ) 0
| |
/1
(5.13 )
91
Chapter Five

operation,
( )
( )|
| |
)|
( ) of is used in the de-companding
(5.14)
The positive constant determines the average power of output signals. In order to
keep the input and output signals at the same average power level.
[| | ]
[ 0
| |
(5.15 )
/1 ]
The following can be observed from table A.8 and the following figures
At y =2 the PAPR deteriorate incremented by (.7122 dB) as well as the CCDF
of PAPR, deteriorate incremented by (.515 dB) until the SNR at BER(
)
deteriorates by (3.9886 dB)
For
, in this region whenever d decreased lead to improvement in
(PAPR, CCDF of PAPR and SNR at BER(
)) compared with values when y
= 2.
Figure 5.23 shows the best CCDF of PAPR and PAPR at y =0.1. The amount
of improvement in CCDF of PAPR by (9.9192 dB) and in PAPR (23.6085 dB)
compared with OFDM system without companding . Whereas the BER
deteriorates considerably in this case.
Figure 5.23 show the best value for the SNR at BER(
) in cos companding
is when y = 1. Where it has less value deterioration in the SNR at BER(
)
by (0.2717 dB) while PAPR improved by (9.9547 dB), as well as it CCDF of
PAPR improved by (3.8892 dB),compared with OFDM system without
companding.
For
this area is better than
region in terms of PAPR and
CCDF of PAPR and almost have the same SNR at BER(
) as shown in
figure 5.24. So were selected d values less or equal to one.
The relationship between the y parameter in cos companding and PAPR is a direct
correlation, as shown in figure 5.25. Whenever y increased the PAPR and CCDF of
PAPR also increased. But it's different from SNR at BER(
). y =1 is the point of
separation and switching between two contradictory in relation to the SNR at
BER(
). For
whenever y decreased the SNR at BER(
) degradation
increases simply means the relationship is an inverse relationship between y and SNR
at BER(
) when
. While for
is quite unlike the previous case.
Whenever y decreased the SNR at BER(
) degradation also decreased.
92
Chapter Five
10
original
y=1
y = .9
y = .8
y = .7
y = .6
y = .5
y = .4
y = .3
y = .2
y = .1
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
5.23.a
10
-1
BER
10
y
y
y
y
y
y
y
y
y
y
-2
10
-3
10
-4
10
=
=
=
=
=
=
=
=
=
=
5
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
10
15
SNR
20
25
30
5.23.b
Figure 5.23 (a) CCDF of PAPR OFDM system cos companding for various
parameter. (b) The BER of cos companding for various parameter
93
Chapter Five
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at y =0.8. The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at y=0.6. The
For SNR at BER(
)
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at y =0.4. The
For SNR at BER(
)
30
2.1> y >1
1.1>y>0
25
PAPR
20
15
10
0
10
15
20
25
30
SNR at (BER =10 4)
Figure 5.24 the effect of y parameter of cos companding on PAPR and SNR at (BER
=
)
94
Chapter Five
30
SNR at (BER =10-4)
CCDF of PAPR
PAPR
25
[dB]
20
15
10
0
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
Figure 5.25 illustrates the effect of y parameter in cos companding on the PAPR,
CCDF of PAPR and SNR at BER(
)
5.2.3.7 tangent Rooting Companding (tanhR):

The proposed companding depend on tanh and the companding equation will be as
follows:
( )
((| | ) )
( )
(5.16 )
Where k is positive numbers controlling the companding level applied to the envelope
x, | | and sign(x) was used to maintain the phases of the OFDM signal.
Decompanding equation will be as follows:
( )
|(
| |
( )) |
( )
(5.17)
Figure 5.26 shows the CCDF of PAPR and BER for OFDM system with tanhR
companding at k =10 and y change from 0.1 to 1 by 0.1 every time. The CCDF of
PAPR was improved by (6.6795 - 23.9603 dB) and the PAPR was improved by
(6.6999 - 10.1412 dB) while the SNR at BER(
) was deteriorated by (3.0726 18.5686 dB) compared with an OFDM system without companding.
Whenever y decreased the PAPR and CCDF of PAPR was improved while increasing
the SNR at BER(
) values
95
Chapter Five
10
original
y =1
y =.9
y =.8
y =.7
y =.6
y =.5
y =.4
y =.3
y =.2
y =.1
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
25
30
5.26.a
10
-1
BER
10
y
y
y
y
y
y
y
y
y
y
-2
10
-3
10
=1
=.9
=.8
=.7
=.6
=.5
=.4
=.3
=.2
=.1
-4
10
10
15
SNR
20
5.26.b
Figure 5.26 (a) CCDF of PAPR OFDM system tanhR companding for various
parameter. (b) The BER of tanhR companding for various parameters
96
Chapter Five
Figure 5.27 and table A.10 shows the CCDF of PAPR and BER for OFDM system
with tanhR companding at y=1 and with different k,
k = ( 5,10) The CCDF of PAPR was improved by ( 3.6235 - 6.6310 dB) and the
PAPR was improved by (9.1388 - 16.6703 dB) while the SNR at BER (
) was
deteriorated by (0.4931 - 3.2172 dB) compared with an OFDM system without
companding.
k = ( 15 ,20) . The SNR at BER (
) was deteriorated considerably up. where at
k =15 the SNR at BER(
) reach to 30 dB And more than 30 dB at k =20.
While the CCDF of PAPR was improved by (8.2413 - 9.0172 dB) and the PAPR
was improved by ( 20.3701 - 22.2234 dB) compared with an OFDM system
without companding.
Whenever k was increased the PAPR and CCDF of PAPR was improved while
increasing the BER values
Fig (5.28) and table(A.10) shows the CCDF of PAPR and SNR at BER(
) for
OFDM system with tanh companding at y=.8 and with different k .
k = ( 5,10) The CCDF of PAPR was improved by (1.3349- 0.4511 dB) and
the PAPR was improved by (3.0266 - 0.7832 dB) While the SNR at BER
(
) was deteriorated by (0.8925 - 0.2835 dB) compared with an OFDM
system with tanh companding at y=1
k = ( 5,10) The CCDF of PAPR was improved by (4.9584 - 7.0821 dB) and
the PAPR was improved by (12.1654 - 17.4535 dB) While the SNR at BER
(
) was deteriorated by (1.3856 - 3.5007 dB) compared with an OFDM
system without companding.
k =15. The SNR at BER (
) was improved by (7.5344) and the CCDF of
PAPR was improved by (0.0722 dB) While the PAPR was deteriorated by (0.0115 dB) but the deterioration ratio is less than the proportion of
improvement.
At k =20, The SNR at BER (
) was improved ,the CCDF of PAPR was
improved by (0.0184 dB) and the PAPR was improved by ( 0.2169 dB).
k = ( 15,20). where at k =15 the The SNR at BER (
) was deteriorated
by (11.0342 dB )And more than 30 dB at k =20. While the CCDF of PAPR
was improved by (8.3135 - 8.9988 dB) and the PAPR was improved by
(20.3586 - 22.0065 dB) compared with an OFDM system without
companding.
Figure 5.29 and table A.9 shows the following:

For k =5, 10, whenever y was decreased the PAPR and CCDF of PAPR improved
while deteriorating the SNR at BER(
).
For k =15, whenever y was decreased the PAPR and CCDF of PAPR improved,
but the rate of improvement is less than at k = 5, 10. The best value of SNR at
BER(
) at y =.5 while the worst at y =1, where up to 30 dB.
For k =20, The influence of y on PAPR and CCDF of PAPR very little, except
when y = .2 which improve the SNR at BER(
), and CCDF PAPR of PAPR
with a small amount
97
Chapter Five
10
k=
k=
k=
k=
5
10
15
20
-1
10
-2
10
-3
10
4
PAPR0 [dB]
Figure 5.27.a
10
k=
k=
k=
k=
-1
BER
10
5
10
15
20
-2
10
-3
10
-4
10
10
15
snr
20
25
30
Figure 5.27.b
parameters at y=1. (b) The BER of tanhR companding for various parameters at
y=1.
98
Chapter Five
10
k=
k=
k=
k=
5
10
15
20
-1
10
-2
10
-3
10
3
PAPR0 [dB]
Figure 5.28.a
10
k=
k=
k=
k=
-1
BER
10
5
10
15
20
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
Figure 5.28.b
parameters at y=.8. (b) The BER of tanhR companding for various parameters at
y=.8.
99
Chapter Five
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =5 ,y =0.8 . The
while the BER deteriorated by = (-1.2398 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =10 ,y =0.9 .
The improvement in PAPR by = ( 17.0445 dB), and CCDF of PAPR = (6.8431
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=10 ,y =0.6 . The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =10 , y=0.5. The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k = 5, y=0.2. The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =20 ,y =0.2. The
30
25
[dB]
20
15
10
BER k = 5
BER k = 10
BER k = 15
BER k = 20
PAPR k = 5
PAPR k = 10
PAPR k = 15
PAPR k = 20
0
0.2
0.3
0.4
0.5
0.6
y
0.7
0.8
0.9
Figure 5.29 illustrates the effect of y and k parameter in tanhR companding on the
PAPR and SNR at BER(
)
100
Chapter Five
5.2.3.8 Logarithmic Rooting Companding (logR):

The logarithm (
( )
) Rooting companding equation will be as follows:
((| |
( )
(5.18)
Decompanding equation will be as follows:

( )
|(
| |
. /
) |
( )
(5.19)
Where is positive number controlling the amount of companding.

We used to control the companding level applied to the envelope x, | | and sign(x)
was used to maintain the phases of the OFDM signal.
Figure 5.30 shows the CCDF of PAPR and BER for OFDM system with log
companding at y =1 and k change from 10 to 100 by 10 every time.
The CCDF of PAPR was improved by (15.5911 - 6.6150 dB) and the PAPR was
improved by (8.8595 -15.5911 dB) while the SNR at BER(
) was deteriorated by
(1.0686 -18.5686 dB) compared with an OFDM system without companding.
Whenever k was increased the PAPR and CCDF of PAPR was decreased while
Figure 5.31 shows the CCDF of PAPR and BER for OFDM system with tanh
companding at k =10 and y change from 0.1 to 1 by 0.1 every time. The CCDF of
PAPR was improved by (3.5255 - 9.9080 dB) and the PAPR was improved by
(8.8595 -23.4987 dB) while the SNR at BER(
) was deteriorated by (1.0686 18.1686 dB) compared with an OFDM system without companding.
Whenever y decreased the PAPR and CCDF of PAPR was improved while
The following conclusion when comparing the proposed method (from the table
(A.12)) with an OFDM system without PAPR reduction method:
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =10, y =0.9. The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=10 ,y =0.6. The
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =20 ,y =0.5 .
The improvement in PAPR by = ( 16.6873 dB), and CCDF of PAPR = (6.9120
For SNR at BER(
)
101
Chapter Five
For SNR at BER(

)
The best one improvement
in PAPR and CCDF of PAPR is at . The
For SNR at BER(
)
The best one improvement
in PAPR and CCDF of PAPR is at . The
0
10
-1
10
original
k = 10
k = 20
k = 30
k = 40
k = 50
k =60
k = 70
k = 80
k = 90
k = 100
-2
10
-3
10
6
PAPR0 [dB]
10
12
Figure 5.30.a
10
k
k
k
k
k
k
k
k
k
k
-1
BER
10
-2
10
= 10
= 20
= 30
= 40
= 50
=60
= 70
= 80
= 90
= 100
-3
10
-4
10
10
15
SNR
20
25
30
Figure 5.30.b
Figure 5.30 (a)CCDF of PAPR OFDM system logR companding for various
parameter. (b) the BER of logR companding for various parameter.
102
Chapter Five
10
original
y=1
y = .9
y = .8
y = .7
y = .6
y = .5
y = .4
y = .3
y = .2
y = .1
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
Figure 5.31.a
10
-1
BER
10
y
y
y
y
y
y
y
y
y
y
-2
10
-3
10
-4
10
=
=
=
=
=
=
=
=
=
=
5
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
10
15
SNR
20
25
30
Figure 5.31.b
Figure 5.31 (a)CCDF of PAPR OFDM system logR companding for various
parameter. (b) The BER of logR companding for various parameter.
103
Chapter Five
5.2.4 Pre-distortion methods:

The pre-distortion methods are based on the reorientation or spreading the energy of
data symbol before taking IFFT. The pre-distortion schemes include DFT spreading,
pulse shaping or precoding and constellation shaping. The methods like Tone
Reservation (TR) and Tone Injection (TI) are the example of constellation shaping
schemes [188]
5.2.4.1 Pulse Shaping or Pre-coding:

The pulse shaping or pre-coding technique is an efficient and flexible way for
reducing the PAPR of OFDM signals. In this method, each data block is multiplied by
a pre-coding matrix prior to OFDM modulation and transmission. This method is
data-independent and, thus, avoids block based optimization. It also works with an
arbitrary number of subcarriers and any type of baseband modulation used. In terms
of BER performance, it takes advantage of the frequency variation of the fading
multipath channel and improves the BER of OFDM signals in comparison to
conventional OFDM (no pre-coding). The implementation complexity of the proposed
technique is acceptable, since a predefined pre-coding matrix is used and thus, no
handshake is needed between transmitter and receiver. Having the same pre-coding
matrix for all OFDM blocks avoids all the processing needed in block-based
optimization methods [189].
Precoded OFDMA consists of using a precoding matrix P that spreads the energy of
symbols over the subcarriers allocated to the user. Uniform energy distribution is
favored in practice. [190]
The OFDM system with an orthogonal precoder is considered. In precoded OFDM
system instead of sending uncoded symbols (one per subcarrier), the idea is to send
different linear combinations of the information symbols on the subcarriers. This
corresponds to signal space diversity. [191]
Precoding based techniques are simple linear techniques. These techniques can reduce
the PAPR up to the PAPR of single carrier systems (Slimane, 2007). WHT precoding
based techniques, DCT precoding based techniques, DHT precoding based techniques
are common examples of precoding based PAPR reduction techniques (Slimane,
2007; Min & Jeoti, 2007; Baig & Jeoti, 2010a, 2010b, 2010c) [14]
Figure. 5.32 shows the block diagram of Precoding Based OFDM System. We
implemented the Precoding matrix P of dimension N N before the IFFT to reduce
the PAPR.
The Precoding matrix P can be written as:
(5.20)
)]
)(
Where P is a Precoding Matrix of size N N is shown in equation (5.20). The complex

baseband OFDM signal with N subcarriers can be written as:
( )
t
104
NT
(5.21)
Chapter Five
We can express modulated OFDM vector signal with N subcarriers as:

*
(5.22)
[192], [193]
Add CP
S
/
P
IDFT OR IFFT
Signal mapper
I
/
P
P
/
S
Multipath
Fading Ch. &
noise
S
/
P
Remove CP
DFT OR FFT
One Tap
Equalizer
And P/S
Signal
demapper
O
/
P
Figure 5.32 Block diagram of Precoding based OFDM system
5.2.4.2 Discrete Hartley transform (DHT) :

The DHT is a linear transform. In DHT, N real numbers
are
transformed into N real numbers
a. According to [91], the N-point
DHT can be defined as follows:
Where
( )
(5.23)
and
P is precoding matrix of size NN shown, m and n are integers from

DHT is also invertible transform which allows us to recover the
from
inverse can be obtained by simply multiplying DHT of
by [194].
. The
and
5.2.4.3 Walsh-Hadamard Transform (WHT):

The Hadamard transform (also known as the WalshHadamard transform, Hadamard
RademacherWalsh transform, Walsh transform, or WalshFourier transform) is an
example of a generalized class of Fourier transforms. It performs an orthogonal,
symmetric, involutional, linear operation on
real numbers (or complex numbers,
although the Hadamard matrices themselves are purely real). The Hadamard
transform can be regarded as being built out of size
DFTs, and is
in fact equivalent to a multidimensional DFT of size. It decomposes an arbitrary input
vector into a superposition of Walsh functions [195].
105
Chapter Five
The WHT is a non-sinusoidal and it is an orthogonal technique which decomposes a

signal into set of basic functions. These functions are called Walsh functions, the
hadamard transform scheme reduce the occurrence of the high peaks comparing the
conventional OFDM system. The hadamard transform used because it reduce the
autocorrelation of the input sequence to reduce the PAPR of OFDM signal. It also not
requires to send side information to the receiver [36]
WHT can be implemented by a butterfly structure as in FFT. This means that
applying WHT does not require the extensive increase of system complexity. The
kernel of WHT can be written as follows:, -
(5.24)
(5.25)
(5.26)
Where
denotes the binary complement of
[14]
Note that Hadamard transform is an orthogonal linear transform and can be
implemented by a butterfly structure as in FFT. This means that applying Hadamard
transform does not require the extensive increase of system complexity. The received
vector signal
corrupted by noise vector n can be recovered to
as [11]:
*
*
+
*
(5.27)
(5.28)
The FWHT for a signal x of length N is defined as:
Where i = 0,1,..., N-1 and
(5.29)
(
) are Walsh functions
5.2.4.4 Discrete Cosine Transform (DCT):

DCT is a technique to transform a signal into frequency domain. DCT denotes a row
of data in terms of the sum of cosine functions that oscillate at different frequencies.
DCT is similar to the DFT, but the DCT only uses real number without imaginary
component. The idea of using DCT in this study is to reduce the autocorrelation of the
input row to reduce PAPR and it does not require the information transmitted to the
receiver. The idea of using DCT is for reduce auto-correlation from input data to
reduce PAPR problem [117].
DCT matrix P of size N-by-N can be created by using equation
106
Chapter Five
(5.30)
and DCT can be defined as:
, .
/ - k=0,1 N-1
(5.31)
5.2.4.5 Discrete Sine Transform (DST) Precoding Technique:

For an input signal
, the discrete sine transform

(
)(
can be defined as:
)- k=1, 2 N-1
(5.32)
DST precoding matrix D can be generated as follows:
(5.33)
The DST matrix must satisfy the following criteria:

1. Same magnitude for all the elements of the precoding matrix.
2. The magnitude must be equal to
.
3. The DST precoding matrix must be non-singular matrix.

These criteria ensure that every output symbol has the same amount of information of
every input data; it preserves the power at the output and also ensures the recovery of
the original data at the receiver. When DST precoding is applied to the complex input
vector of size M, this input vector is transformed to a new vector of size L that can be
written as follows:
Y = D. X = ,
(5.34)
Where D is a DST precoder matrix of size N = L L, generated by Eq. (5.33) and
correspond to
(5.35)
row and
column of DST precoder matrix [40].
5.2.4.6 The Discrete Fourier Transform (DFT) Precoding:

The only difference between the DFT-spread OFDM and the conventional OFDM is
the presence of a DFT and an IDFT block in the transmitter and receiver, respectively
107
Chapter Five
In the DFT-spread OFDM, the PAPR of the signal is fairly low as compared with the
conventional OFDM because the DFT operation spreads data into subcarriers [197].
The DFT of a sequence of length N can be defined as
( )
( )
(5.36)
The sinusoids of the DFT (or Inverse Discrete Fourier Transform (IDFT)) form an
orthogonal basis set and a signal in vector space of the DFT (or IDFT) can be
represented as a linear combination of the orthogonal sinusoids. Thus the IDFT at the
transmitter maps an input signal into a set of orthogonal subcarriers. Similarly the
transform DFT is used at the receiver to reverse the mapping of IDFT and signal from
the subcarriers are combined to form an estimate of the source signal from the
transmitter. Since the basis function of DFT is uncorrelated, the correlation performed
in DFT for the given subcarrier only sees energy for that corresponding subcarrier.
The energy from other subcarrier does not contribute because they are uncorrelated.
This separation of the signal energy is the reason that OFDM subcarriers spectrum
can overlap without causing interference. [198]
5.2.4.7 Simulation results and analysis of OFDM system with precoding:

5 types of Pre-coding are used in this section and then compare them with each other.
The best type of reduced PAPR and BER is the DFT pre-coder. The best type of
reduced PAPR and BER is the DFT pre-coder as shown in figure 5.34 and table (A.3)
but suffer from link performance loss in a frequency-selective channel when highorder modulation techniques are used. The presence of carrier frequency offsets
(CFOs) between the transmitter and the receiver results in a loss of orthogonality
among subcarriers and an intercarrier interference (ICI). CFOs also introduce multiple
access interference (MAI) and degrade the bit error rate (BER) performance in the
DFT pre-coder system. [92]
The following is the conclusion from the table (A.3) and figure 5.33
WHT pecoder was improved each of the PAPR by (2.7941 dB), CCDF of
PAPR by (0.8684 dB) and SNR at BER(
) (0.01dB). But the amount of
improvement in WHT pre-coding is the least in comparison with the rest
kinds of pre-coding
DCT pecoder was improved each of the PAPR by (7.5208 dB), CCDF of
) (0.012dB). DCT pre-coding
results better than WHT pre-coding but worse than the rest
DST pecoder was improved each of the PAPR by (8.1669 dB), CCDF of
) (0.012 dB).
DST and DCT have the same SNR at BER(
) but DST better than DCT in
PAPR and CCDF of PAPR
DHT pecoder was improved each of the PAPR by (18.6731 dB), CCDF of
) (0.058dB). DHT pecoder results
are better than other types of pre-coding except DFT pre-coder.
DFT pecoder was improved each of the PAPR by (25.6118 dB), CCDF of
) (0.171dB). DFT pecoder
results are the best compared with other types of pre-coding
108
Chapter Five
10
orignal
WHT
DCT
DST
DHT
DFT
-1
10
-2
10
-3
10
6
PAPR0 [dB]
10
12
5.33.a
10
orignal
WHT
DCT
DST
DHT
DFT
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
5.33.b
Figure 5.33 (a)CCDF of PAPR for OFDM system with different type of pre-coding
(b) BER for OFDM system with different type of pre-coding
109
Simulation Results and Analysis of Hybrid PAPR techniques

Chapter six
6.1 Hybrid pre-coding with RCF:
Proposed a method based on the integration of all of precoding with RCF as shown in
figure 6.1. The results of this method better than the results of the RCF and precoding each alone, except in the case of DHT with RCF (I = 2, pilot) ,where the
results of the DHT itself better than the pre-coding with RCF hybrid (DHT with
RCF).The best result for the PAPR is when RCF (I =1) with (DHT).
WHT, DCT, DST, and DHT pre-coders are used with RCF is used with the following
specifications (I =1, pilot, and 2, CR =4, 3, 2) The OFDM system model with the
proposed technique as shown in figure 6.1.
D
/
A
Add CP
P
/
S
RCF
IDFT OR IFFT
S
/
P
+pilot symbol
Signal mapper
I
/
P
Multipath
Fading Ch. &
noise
S
/
P
Remove CP
DFT OR FFT
Remove
+pilot symbol
One Tap
Equalizer
And P/S
Signal
demapper
O
/
P
A
/
D
Figure 6.1 the OFDM system model with precoding + RCF.

The following conclusion from table A.29 when comparing the proposed method with
an OFDM system without PAPR reduction method:
There are improved in PAPR, CCDF of PAPR and BER in many points that have
been tested, but The best one improvement in PAPR and CCDF of PAPR is at I =
pilot, CR =2,and DHT. The improvement in PAPR by = (17.2780 dB), CCDF of
PAPR = (7.2062 dB), and the SNR at BER(
) by = (0.1105 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at I = 1, CR =1.5,
and DHT. The improvement in PAPR by = (20.4339 dB), and CCDF of PAPR =
For SNR at BER(
)
110

The best one improvement in PAPR and CCDF of PAPR is at I = 1, CR =1.3,
and DHT. The improvement in PAPR by = (21.0373 dB), and CCDF of PAPR =
an OFDM system with RCF method:
The PAPR and the CCDF of PAPR were improved and the least amount of
improvement was when I=1, CR = 4 and WHT .The PAPR improvement is equal
to (0.1148 dB) and the CCDF of PAPR improvement is equal to (0.0031 dB),
while the vast amount of improvement is where I=1, CR = 4 and DHT the PAPR
improvement is equal to (7.1348 dB) and the CCDF of PAPR improvement is
equal to (3.0141 dB )
The SNR at BER(
) at I =2, pilot, there is an improvement in some of the
points and the degradation the other .The largest amount of improvement is when
I=2, CR =1.5 and DHT is equal to (17.1418 dB).The largest amount of
degradation is when I=2, CR =1.5 and WHT is equal to (-0.4 dB).
The SNR at BER(
) was improved at I =1.The least amount of improvement in
SNR at BER(
) when DST and CR = 4 and is equal to (0.063 dB). The largest
amount of improvement is when DHT and CR =1.5 is equal to (16.905 dB).
an OFDM system with pre-coding method:
The PAPR was improved, except when (DHT, I =2, pilot, and CR = 4,3 ,2 ,1.5)
PAPR was degraded and the maximum degradation is (-7.269 dB).the least
amount of improvement was at (DHT, I =1, and CR = 2) and is equal to (0.0008
dB), while the vast amount of improvement is where (WHT, I =1, and CR = 1.1)
and is equal to (18.3732 dB).
The CCDF of PAPR was improved, except when (DHT, I =2, pilot, and CR = 4,3
,2 ,1.5) and (DHT, I =2, and CR = 1.5) CCDF of PAPR were degraded and the
maximum degradation is (-2.9015 dB).the least amount of improvement was at
(DHT, I = Pilot, and CR = 1.5) and is equal to (0.1727 dB), while the vast amount
of improvement is where (WHT, I =1, and CR = 1.1) and is equal to (8.6977 dB).
The SNR at BER(
) was degraded, except when (DHT, I =2, pilot, and CR =
4,3 ,2 ) SNR at BER(
) were degraded and the maximum improvement is
(2.7701 dB).The least amount of degradation in SNR at BER(
) when d=1.1
and DHT and is equal to (dB). The largest amount of degradation is when I =1,
CR =1.1, 1.3 and WHT is equal to (-18.37 dB).
Figures (6.2, 6.3, 6.4, and 6.5) shows the performance of the hybrid pre-coding with
RCF (at I =1 for different CR).
Hybrid pre-coding with RCF is the same as the RCF where whenever CR decreased
the (PAPR and CCDF of PAPR) improved and the SNR at BER(
) degraded but
the proposed better than RCF because the improvement (CCDF and PAPR of PAPR)
is greater than the amount of degradation in the SNR at BER(
) if we compared
both with the original signal and RCF to gather.
The best species is (`DHT + RCF) comes after ((DCT + RCF and RCF + DST)
together) and in the end come (WHT + RCF)
111

0
10
WHT+RCF (CR = 4)
WHT+RCF(CR = 3)
WHT+RCF(CR = 2)
WHT+RCF(CR = 1.5)
WHT+RCF(CR = 1.3)
WHT+RCF(CR = 1.1)
-1
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 6.2.a
10
WHT+RCF (CR = 4)
WHT+RCF(CR = 3)
WHT+RCF(CR = 2)
WHT+RCF(CR = 1.5)
WHT+RCF(CR = 1.3)
WHT+RCF(CR = 1.1)
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
Figure 6.2.b
Figure 6.2 (a)CCDF of PAPR for OFDM system with WHT +RCF where I =1 for
different CR (b) BER for OFDM system with WHT +RCF where I =1 for different
CR
112

0
10
DCT+RCF (CR = 4)
DCT+RCF(CR = 3)
DCT+RCF(CR = 2)
DCT+RCF(CR = 1.5)
DCT+RCF(CR = 1.3)
DCT+RCF(CR = 1.1)
-1
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 6.3.a
10
DCT+RCF (CR = 4)
DCT+RCF(CR = 3)
DCT+RCF(CR = 2)
DCT+RCF(CR = 1.5)
DCT+RCF(CR = 1.3)
DCT+RCF(CR = 1.1)
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
Figure 6.3.b
Figure 6.3 (a)CCDF of PAPR for OFDM system with DCT +RCF where I =1 for
different CR (b) BER for OFDM system with DCT +RCF where I =1 for different CR
113

0
10
DST+RCF (CR = 4)
DST+RCF(CR = 3)
DST+RCF(CR = 2)
DST+RCF(CR = 1.5)
DST+RCF(CR = 1.3)
DST+RCF(CR = 1.1)
-1
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 6.4.a
10
DST+RCF (CR = 4)
DST+RCF(CR = 3)
DST+RCF(CR = 2 )
DST+RCF(CR = 1.5)
DST+RCF(CR = 1.3)
DST+RCF(CR = 1.1)
-1
BER
10
-2
10
-3
10
-4
10
10
15
SNR
20
25
30
Figure 6.4.b
Figure 6.4 (a)CCDF of PAPR for OFDM system with DST +RCF where I =1 for
different CR (b) BER for OFDM system with DST +RCF where I =1 for different CR
114
10
-1
10
-2
10
DHT+RCF(CR =4)
DHT+RCF(CR =3)
DHT+RCF(CR =2)
DHT+RCF(CR =1.5)
DHT+RCF(CR =1.3)
DHT+RCF(CR =1.1)
-3
10
0.5
1.5
2
PAPR0 [dB]
2.5
3.5
Figure 6.5.a
10
DHT+RCF
DHT+RCF
DHT+RCF
DHT+RCF
DHT+RCF
DHT+RCF
-1
BER
10
(CR
(CR
(CR
(CR
(CR
(CR
=
=
=
=
=
=
4)
3)
2)
1.5)
1.3)
1.1)
-2
10
-3
10
-4
10
10
15
SNR
Figure 6.5.b
Figure 6.5 (a)CCDF of PAPR for OFDM system with DHT +RCF where I =1 for
different CR (b) BER for OFDM system with DHT +RCF where I =1 for different CR
Figure 6.6 shows the proposed method with I =2 and CR =4, Can note the following
form the figure 6.6, primarily the proposed method on despite of the different type of
pre-coding used but it has almost the same PAPR (0.1690 dB), CCDF of PAPR
(0.1493 dB) and SNR at BER(
) (0.1111 dB). This means that it is not based on
the type of pre-coding.
115
10
WHT + RCF
DCT + RCF
DST + RCF
DHT+ RCF
-1
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 6.6.a
10
WHT + RCF
DCT + RCF
DST + RCF
DHT + RCF
-1
BER
10
-2
10
-3
10
-4
10
5
SNR
10
Figure 6.6.b
+RCF where I =2, CR =4 (b) BER for OFDM with different type of pre-coding +RCF
where I =2, CR =4
Figure 6.7 shows the proposed method with I =pilot and CR =4 , Can note the
following form the figure 6.7 ,primarily the proposed method on despite of the
different type of pre-coding used but it has almost the same PAPR (0.0378 dB),
116

CCDF of PAPR(0.0219 dB) and SNR at BER(
is not based on the type of pre-coding.
) 0.3759 dB). This means that it
10
DCT+RCF
DST+RCF
DHT +RCF
-1
10
-2
10
-3
10
3
4
PAPR0 [dB]
Figure 6.7.a
10
DCT +RCF
DST +RCF
DHT +RCF
-1
BER
10
-2
10
-3
10
-4
10
5
snr
10
Figure 6.7.b
+RCF where I =pilot, CR =4 (b) BER for OFDM with different type of pre-coding
+RCF where I =pilot, CR =4
117

Figure 6.8 shows the proposed method with I =1 and CR =1.5. Here depends on the
type of the pre-coding, where different values of (PAPR, CCDF of PAPR and BER).
For each type of pre-coding, and best as is evident is the DHT of figure 6.8.
10
WHT + RCF
DCT + RCF
DST + RCF
DHT + RCF
-1
10
-2
10
-3
10
0.5
1.5
2.5
PAPR0 [dB]
Figure 6.8.a
10
WHT + RCF
DCT + RCF
DST + RCF
DHT + RCF
-1
BER
10
-2
10
-3
10
-4
10
10
15
20
25
SNR
Figure 6.8.b
Figure 6.8(a) CCDF of PAPR for OFDM system with different type of pre-coding
+RCF where I =pilot, CR =4 (b) BER for OFDM with different type of pre-coding
+RCF where I =1, CR =1.5
118

6.2 Hybrids RCF with companding:
The clipping is the easiest technique to reduce the power by setting a maximum level
for the transmitted signal. In addition to these benefits in clipping, the use of
frequency domain filtering, this improves the BER. On the other hand, the
companding has also been considered a good technique, because it has the good
PAPR reduction capability with no bandwidth expansion and low computational
complexity. The other advantage of companding is that the signal can be recovered at
the receiver through inverse companding transform [10]
With the understanding on RCF and companding techniques, an idea emerged to
combine the philosophy of companding and RCF.
This hybrid technique shows good results because of first RCF reduce the PAPR and
improves the BER constant and then companding more reduces the amount of the
PAPR.
The OFDM system model with the proposed technique is as shown in figure 6.9. RCF
is used with the following specifications (I =2, CR =4, 3, 2) as for the companding has
been using all kinds of previous companding.
Add CP
P
/
S
Companding
RCF
IDFT OR IFFT
S
/
P
+pilot symbol
Signal mapper
I
/
P
D
/
A
Multipath
Fading Ch. &
noise
Remove CP
S
/
P
De-Companding
DFT OR FFT
Remove
+pilot symbol
One Tap
Equalizer
And P/S
Signal
demapper
O
/
P
A
/
D
Figure 6.9 the OFDM system model with RCF with companding .
6.2.1 RCF + A companding:

The following conclusion from table A.13 and figure 6.10 when
comparing the proposed method with an OFDM system without PAPR reduction
method:
At A =5 and CR =4 , There is an improvement in PAPR by = 17.1305 dB ,CCDF
of PAPR = 6.9730 dB, and the BER by =0.3409 dB. This point was chosen
because all the variables improved.
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at A =10 and CR =4
. The improvement in PAPR by = (18.8751 dB), and CCDF of PAPR = (7.7750
For SNR at BER(
)
119
The best one improvement in PAPR and CCDF of PAPR is at A = 20 and CR =4

For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at A =4 and CR =4.
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at A=200 and CR
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at A =90 and CR =2.
comparing the proposed method with an OFDM system with a companding PAPR
reduction method:
The PAPR was improved and the least amount of improvement was when A = 90
and CR = 4 and is equal to (6.9071 dB), While the vast amount of improvement is
where A = 5 and CR = 1.5 and is equal to (13.4875 dB).
Less the amount of improvement in CCDF of PAPR when A = 100 and CR = 4
and is equal to (0.815 dB), While the vast amount of improvement is where A = 5
and CR = 1.5 and is equal to (4.64 dB).
The SNR at BER(
)was improved when CR =4,3 . The vast amount of
improvement is where A = 10 and CR = 4 and is equal to (3.1882 dB), while Less
the amount of improvement in SNR at BER(
)when A = 30 and CR = 3 and is
equal to (0.88 dB)
The SNR at BER(
) was degraded when CR = 2. The least amount of
) when A =80 and CR = 2 and is equal to (1.2408 dB). The largest amount of degradation is when A= 90 and is equal to (3.0453 dB).
The SNR at BER(
) was degraded when CR = 1.5, when A =5 the amount of
degradation is equal to (-16.4 dB).
The following conclusion from table A.13 and figure 6.10 when comparing the
proposed method with an OFDM system with RCF method:
improvement was when A =5 and CR = 1.5 .The PAPR improvement is equal to
(2.9643 dB) and the CCDF of PAPR improvement is equal to (1.6257 dB), while
the vast amount of improvement is where A = 200 and CR = 4 and the PAPR
equal to (4.64 dB )
The SNR at BER(
) was degraded, except when A =5 and CR =1.5 the BER
maintains its value. The least amount of degradation in SNR at BER(
) when
A =5 and CR = 4 and is equal to (-2.3226 dB). The largest amount of degradation
is when A= 90 and CR =2 is equal to (-12.072 dB).
120

30
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
A
RCF (CR=4) + A
RCF (CR=3) + A
RCF (CR=2) + A
25
PAPR
20
15
10
10
12
14
16
18
20
22
24
26
SNR at (BER =10- 4)
Figure 6.10.a
11
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
A
RCF (CR=4) + A
RCF (CR=3) + A
RCF (CR=2) + A
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
12
14
16
18
20
22
24
26
SNR at (BER =10- 4)
Figure 6.10.b
Figure 6.10(a) Shows the values of the PAPR and SNR at BER =
for each of the
RCF, Acompanding , and Hybird (RCF+A ) b) Shows the values of the CCDF of
PAPR and SNR at BER =
for each of the RCF, Acompanding , and Hybird
(RCF+A )
6.2.2 RCF + :
The following conclusion when from table A.14 and figure 6.11 comparing the
proposed method with an OFDM system without PAPR reduction method:
At =5 and CR = 4, There is an improvement in PAPR by = (16.5081 dB), CCDF
of PAPR = (6.7470 dB), and the SNR at BER(
) by = (0.8014 dB). This point
was chosen because all the variables improved.
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at =10 and CR
121
For SNR at BER(

)
The best one improvement in PAPR and CCDF of PAPR is at =30 and CR =3.
dB), while the BER deteriorated by = (-3.5686 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at =70 and CR =3 .
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at = 220 and CR
comparing the proposed method with an OFDM system with companding PAPR
reduction method:
The PAPR was improved and the least amount of improvement was when = 220
where = 80 and CR = 2 and is equal to (12.6705 dB).
Less the amount of improvement in CCDF of PAPR when = 220 and CR = 4 and
is equal to (0.8325 dB), while the vast amount of improvement is where = 5 and
CR = 1.5 and is equal to (4.1715 dB).
The SNR at BER(
) was improved when CR =4,3 . The vast amount of
improvement is where = 50 and CR = 4 and is equal to (3.0554 dB), while Less
the amount of improvement in SNR at BER(
)when = 255 and CR = 3 and is
equal to (1.468 dB)
The SNR at BER(
) when =240 and is equal to (-0.985 dB). The
largest amount of degradation is when = 20 and is equal to (-2.625 dB).
The SNR at BER(
) was degraded when CR = 1.5, when =5 the amount of
degradation is equal to (-16.6637dB).
comparing the proposed method with an OFDM system with RCF method:
and CR = 1.5 and is equal to (2.8281 dB), while the vast amount of improvement is
Less the amount of improvement in CCDF of PAPR when = 5 and CR = 2 and is
equal to (1.4656 dB), while the vast amount of improvement is where = 255 and
CR = 4 and is equal to (4.6677 dB).
The SNR at BER(
) was degraded, except when =5 and CR =1.5 the SNR at
BER(
) maintains its value. The least amount of degradation in SNR at
BER(
) when MU =5 and CR = 4 and is equal to (-1.8621 dB). The largest
amount of degradation is when MU= 220 and CR =2 is equal to (-10.872 dB).
122

30
original
RCF (CR =4)
RCF (CR =3)
RCF (CR=2)
MU
RCF (CR =4) + MU
RCF (CR =3) + MU
RCF (CR=2) + MU
25
PAPR
20
15
10
10
12
14
16
18
20
22
24
SNR at (BER =10- 4)
Figure 6.11.a
11
original
RCF (CR =4)
RCF (CR =3)
RCF (CR=2)
MU
RCF (CR =4) + MU
RCF (CR =3) + MU
RCF (CR=2) + MU
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
12
14
16
18
20
22
24
SNR at (BER =10- 4)
Figure 6.11.b
Figure 6.11 ( a) Shows the values of the PAPR and SNR at BER =
for each of
the RCF, companding , and Hybird (RCF+ ). b) Shows the values of the CCDF of
for each of the RCF, companding , and Hybird
(RCF+ ).
6.2.3 RCF + RCT:

method:
At these values (R =0.9, 0.8, 0.7, 0.6 and CR = 4, 3) There are improved in PAPR,
CCDF of PAPR and the BER dB). Point was chosen because all the variables
improved. The best one improvement in PAPR and CCDF of PAPR is at R =0.6
123
and CR =3. The improvement in PAPR by = (17.2514 dB), CCDF of PAPR =

(7.0968 dB), and the SNR at BER(
) by = (0.5079 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at R =0.5 and CR =3.
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at R =0.4 and CR =3
For SNR at BER(
)
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at R= 0.2 and CR =3.
For SNR at BER(
)
proposed method with an OFDM system with RCT PAPR reduction method:
improvement was when R =0.1 and CR = 4 .The PAPR improvement is equal to
(0.9227 dB) and the CCDF of PAPR improvement is equal to(0.363 dB), While
the vast amount of improvement is where R =0.9 and CR = 1.5 and the PAPR
equal to (6.201dB )
The SNR at BER(
) was improved when CR =4,3 . The vast amount of
improvement is where R =0.7 and CR = 4 and is equal to (3.0902 dB), while Less
the amount of improvement in CCDF of PAPR when R =0.1 and CR = 3 and is
equal to (1.8866 dB)
The SNR at BER(
) when R =0.5 and is equal to (-0.7285 dB). The
largest amount of degradation is when R =0.3 and is equal to (-1.475 dB).
The SNR at BER(
) was degraded when CR = 1.5, when R =0.9 the amount of
degradation is equal to (-18.3235dB).
The PAPR was improved and the least amount of improvement was when R =0.9
where R =0.1and CR = 4 and is equal to (12.5433 dB).
124
Less the amount of improvement in CCDF of PAPR when R = 0.9 and CR =2 and
is equal to (0.1506 dB), while the vast amount of improvement is where R= 0.1
and CR = 4 and is equal to (5.5127 dB).
The SNR at BER(
) was degraded, except when R =0.1 and CR =1.5 the SNR
at BER(
) maintains its value. The least amount of degradation in CCDF of
PAPR when R =0.9 and CR = 4 and is equal to (-0.1291 dB). The largest amount
of degradation is when R= 0.1 and CR =2 is equal to (-17.181 dB).
30
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
Roots
RCF (CR=4) + Rooting
25
PAPR
20
15
10
10
15
20
25
30
SNR at (BER =10- 4)
figure 6.12.a
12
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
Roots
10
CCDF of PAPR
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.12. b
figure 6.12 (a) Shows the values of the PAPR and SNR at BER =
for each of the
RCF, RCT , and Hybird (RCF+RCT) b) Shows the values of the CCDF of PAPR and
SNR at BER =
for each of the RCF, RCT, and Hybird (RCF+ RCT).
125

6.2.4 RCF + AEXP:
method:
At these values ((d =2-1by .1 every time and CR = 4) and when (d =1.8, 1.5 and
CR =3) There are improved in PAPR, CCDF of PAPR and SNR at BER(
)).
The best one improvement in PAPR and CCDF of PAPR is at d = 1 and CR =4.
The improvement in PAPR by = (18.7316 dB) ,CCDF of PAPR = (7.7135 dB),
and the SNR at BER(
) by = ( 0.2467 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at d =0.9 and CR =3.
The improvement in PAPR by = (19.6985 dB), and CCDF of PAPR = (8.1400 dB),
For SNR at BER(
)
For SNR at BER(
)
For SNR at BER(
)
comparing the proposed method with an OFDM system with AEXP companding
PAPR reduction method:
The PAPR was improved except when d =1.6 and CR =3 the PAPR was degraded
by (-1.6855dB) .the least amount of improvement was when d =0.4 and CR = 4
and is equal to (0.1356 dB), while the vast amount of improvement is where d =2
Less the amount of improvement in CCDF of PAPR when d =0.4 and CR = 4 and
is equal to (0.14 dB), while the vast amount of improvement is where d =2 and CR
= 1.5 and is equal to (1.7799 dB).
The SNR at BER(
) was improved when CR =4,3 except when (d =0.4 and CR
=4,3) and when (d =0.5 and CR =3) the SNR at BER(
) maintains its value.
The vast amount of improvement is where d =0.7 and CR = 4 and is equal to
(15.5893 dB), while Less the amount of improvement in SNR at BER(
) when
d=0.1 and CR = 4 and is equal to (0.79 dB)
The SNR at BER(
) was degraded when CR = 2 except when (d =0.6, 0.5, 0.4)
the SNR at BER(
) maintains its value. The least amount of degradation in
SNR at BER(
) when d =0.7 and is equal to (-0.9 dB). The largest amount of
degradation is when R =1.1 and is equal to (-7.65 dB).
The SNR at BER(
) was degraded when CR = 1.5, when d =2. the amount of
126
30
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
AEXP
RCF (CR=4) + AEXP
RCF(CR=3) +AEXP
RCF(CR=2) +AEXP
25
PAPR
20
15
10
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.13.a
11
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
AEXP
RCF (CR=4) + AEXP
RCF (CR=3) + AEXP
RCF(CR=2) + AEXP
10
9
8
CCDF of PAPR
The PAPR was improved and the least amount of improvement was when d =2 and
CR = 1.5 and is equal to (0.0555 dB), while the vast amount of improvement is
where d =0.4 and CR = 4 and is equal to (11.2249 dB).
Less the amount of improvement in CCDF of PAPR when d =2 and CR =2 and is
equal to (0.3051 dB), while the vast amount of improvement is where d =0.4 and
The SNR at BER(
) was degraded, except when d =2 and CR =1.5 the SNR at
BER(
BER(
) when d =1.3 and CR = 4 and is equal to (-1.6321 dB). The largest amount
of degradation is when d =0.4 and CR =4 is equal to (-21.2321 dB).
7
6
5
4
3
2
1
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.13.b
Figure 6.13 (a) Shows the values of the PAPR and SNR at BER =
RCF, AEXP companding , and Hybird (RCF+AEXP)
b) Shows the values of the CCDF of PAPR and SNR at BER =
RCF, AEXP companding , and Hybird (RCF+ AEXP)
127
for each of the

for each of the

6.2.5 RCF + cos :
method:
At these values ((y =1,0.9, 0.8,0.7 and CR = 4,3) and when ( y =0.6 and CR =4)
There are improved in PAPR, CCDF of PAPR and BER ). The best one
improvement in PAPR and CCDF of PAPR is at y =.7 and CR =3. The
improvement in PAPR by = (17.1463 dB),CCDF of PAPR = (7.0651 dB), and the
BER by = ( 0.5250 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at y=0.5 and CR =4.
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at y = 0.6 and CR
For SNR at BER(
)
For SNR at BER(
)
For SNR at BER(
)
proposed method with an OFDM system with cos companding PAPR reduction
method:
The PAPR was improved and the least amount of improvement was when y =0.3
where y =1 and CR = 1.5 and is equal to (8.2456 dB).
The CCDF of PAPR was improved and the least amount of improvement is where
y =0.3 and CR = 4 and is equal to (0.4597 dB).
, while the vast amount of improvement is where y =1 and CR = 1.5 and is equal
to (3.4937 dB).
The SNR at BER(
) was improved when CR =4, 3. The vast amount of
improvement when y=0.3 and CR = 4 and is equal to (4.4287 dB), while Less the
amount of improvement in BER when y = 1 and CR = 3 and is equal to (2.061dB)
The SNR at BER(
) when y =0.6 and is equal to (-0.3705 dB). The
largest amount of degradation is when y =0.3 and is equal to (-8.3 dB).
The SNR at BER(
) was degraded when CR = 1.5, when y =1 the amount of
128

The PAPR was improved and the least amount of improvement was when y =1 and
CR = 1.5 and is equal to (1.0679 dB), While the vast amount of improvement is
where y =0.1 and CR = 4 and is equal to (12.7916 dB).
Less the amount of improvement in CCDF of PAPR when y=1 and CR = 1.5 and is
equal to (0.2694 dB), while the vast amount of improvement is where y=0.1 and
The SNR at BER(
) was degraded, except when y =1 and CR =1.5 the BER
maintains its value. The least amount of degradation in SNR at BER(
) when y
=1 and CR = 4 and is equal to (-0.3076 dB). The largest amount of degradation is
when y =0.2, 0.1 and CR =4 is equal to (21.2321 dB).
30
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
cos
RCF (CR=4) + cos
RCF (CR=3) +cos
RCF (CR=2) + cos
25
PAPR
20
15
10
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.14.a
11
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
cos
RCF (CR=4) + cos
RCF (CR=3) + cos
RCF (CR=2) + cos
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.14.b
for each of the
RCF, cos companding , and Hybird (RCF+cos) (b)Shows the values of the CCDF of
for each of the RCF, cos companding , and Hybird
(RCF+ cos)
129

6.2.6 RCF + NERF :
method:
At these values (CR = 4, 3) and There are improved in PAPR, CCDF of PAPR and
BER). The best one improvement in PAPR and CCDF of PAPR is at CR =3. The
improvement in PAPR by = (17.0615 dB), CCDF of PAPR = (7.2730 dB), and the
SNR at BER(
) by = (0.5314 dB).
26
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
RCF (CR=1.5)
NERF
RCF (CR=4) +
RCF (CR=3) +
RCF (CR=2) +
RCF (CR=1.5)
24
22
20
PAPR
18
16
14
NERF
NERF
NERF
+ NERF
12
10
8
6
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.15.a
11
original
RCF (CR=4)
RCF (CR=3)
RCF (CR=2)
RCF (CR=1.5)
NERF
RCF (CR=4) + NERF
RCF (CR=3) + NERF
RCF (CR=2) + NERF
RCF (CR=1.5) + NERF
10
9
CCDF of PAPR
8
7
6
5
4
3
2
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.15.b
for each of the
RCF, cos companding , and Hybird (RCF+NERF) b) Shows the values of the CCDF
of PAPR and SNR at BER =
for each of the RCF, cos companding , and Hybird
(RCF+ NERF)
130

6.2.7 RCF + tanhR :
At these values (k =5,10,15 and y =1 ,0.8 for CR = 4,3) and when (k =20 and y
=1 ,0.8 for CR =3) There are improved in PAPR, CCDF of PAPR and BER ). The
best one improvement in PAPR and CCDF of PAPR is at k=20 y = .8 for CR =4 .
The improvement in PAPR by = (18.6958 dB), CCDF of PAPR = (7.5530 dB), and
the SNR at BER(
) by = ( 0.1527dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=10, y=.5 and
CR =4 . The improvement in PAPR by = (19.3352 dB), and CCDF of PAPR =
The best one improvement in CCDF of PAPR is at k=5 , y=.5 and CR =3 . The
) deteriorated by = (--1.4569dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =20, y=0.5 and
CR =3. The improvement in PAPR by = (21.1850 dB), and CCDF of PAPR =
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k =30, y=0.5 and
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=40, y =0.5and
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k= 40, y =0.2 and
CR =3.The improvement in PAPR by = (23.9630 dB), and CCDF of PAPR =
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k = 20, y=.2 and
an OFDM system with tanhR companding PAPR reduction method:
The PAPR was degraded at (k =15, 20, y =1,0.8 and CR = 4, 3, 2), (k= 10, y
=1,0.8 and CR=4,3), (k =15 ,y =0.5 ,and CR =4) , and when (k =20 ,y =0.5 ,and
CR =4,3). The least amount of degradation in PAPR when k=10, y=0.8 and CR = 3
and is equal to (-0.5458 dB). The largest amount of degradation is when k =15, y =
1 and CR =4 is equal to (-4.717 dB).
The CCDF of PAPR was degraded at [for CR = 4, 3, 2 at (k =20, y =1,0.8,0.5 )
and (k =15, y =1,0.8) ], [for CR =4,3 at (k =15 ,y =0.5 ),and((k= 10, y =1,0.8 and
CR=4,3)], and when [k =10 ,y =0.5 ,and CR =4] The least amount of degradation
in CCDF of PAPR when k=20, y=0.5 and CR = 2 and is equal to (-0.0085 dB).
The largest amount of degradation is when k =20, y = 1 and CR =4 is equal to (2.0457 dB).
131
Except the points already mentioned, the PAPR was improved and the least
amount of improvement was when k =20, y =0.5 and CR = 2 and is equal to
(0.0043 dB), while the vast amount of improvement is where k =5, y =1 and CR =
1.5 and is equal to (8.0365dB).
Except the points already mentioned, the CCDF of PAPR was improved and Less
the amount of improvement in CCDF of PAPR when k =10, y =1 and CR = 2 and
is equal to (0.3585 dB), While the vast amount of improvement is where k =5, y =1
The SNR at BER(
) was degraded when CR =2 at (k =5, y =1,0.8,0.5,0.2 ), ( y
=0.2, k =10,15,20), and when (k =10 ,y =0.5). The least amount of degradation in
SNR at BER(
) when k=5, y=0.8 and is equal to (-0.11 dB). The largest
amount of degradation is when k =10, y = 0.2 is equal to (-2.4641 dB).
The SNR at BER(
) was degraded when CR = 1.5, when k =5, y =1 .The
amount of degradation is equal to (-18.0900 dB).
Except the points already mentioned, the SNR at BER(
) was improved and
the vast amount of improvement is where k =15, y =1 and CR = 4 and is equal to
(19.803 dB), while Less the amount of improvement in SNR at BER(
) when k
=5, y=0.8 and CR = 2 and is equal to (1.1217 dB).
The PAPR was improved and the least amount of improvement was when k=5 ,y
=1 and CR = 1.5 and is equal to (0.0685 dB), While the vast amount of
improvement is where k =40, y =0.2 and CR = 4 and is equal to (12.6433 dB).
The CCDF of PAPR was improved, except when k =5 ,y =1 and CR =2 the CCDF
of PAPR was degraded by (-0.0264 dB). Less the amount of improvement in
CCDF of PAPR when k =5, y= 1 and CR = 3 and is equal to (0.027 dB), While the
vast amount of improvement is where k =40, y =0.2 and CR = 4 and is equal to
(5.5377 dB).
The SNR at BER(
) was degraded, except when k =5 ,y =1 and CR =1.5 the
SNR at BER(
BER(
) when k =5, y =1 and CR = 3 and is equal to (-0.1386 dB). The largest
amount of degradation is when k =50, y =1, 0.8, 0.2 and CR =3 is equal to (20.5860 dB).
6.2.8 RCF +logR :

At these values (k =5,10,20 and y =1 ,0.8 for CR = 4,3) and when (k =30,40,50
,70 and y =1 ,0.8 for CR =4) and finally, when (k =30,40 and y =1 for CR =3)
There are improved in PAPR, CCDF of PAPR and BER ). The best one
improvement in PAPR and CCDF of PAPR is at k=70, y = .8 for CR =4 . The
improvement in PAPR by = (18.0511 dB) ,CCDF of PAPR = (7.3815 dB), and the
SNR at BER(
) by = (0.2025 dB).
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=5, y =0.5 and
132
For SNR at BER(

)
The best one improvement in PAPR and CCDF of PAPR is at k=90, y=0.5 and
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=50 ,y =0.5and
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=70 ,y =0.2and
CR =3. The improvement in PAPR by = (23.3205dB), and CCDF of PAPR =
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=.9, y =0.2 and
an OFDM system with logR companding PAPR reduction method:
The PAPR was improved and the least amount of improvement was when k =5, y
=1 and CR = 1.5 and is equal to (0.4088 dB), While the vast amount of
The CCDF of PAPR was improved and the least amount of improvement was
when k =5, y =1 and CR = 2 and is equal to (0.2136 dB), while the vast amount of
The SNR at BER(
) was degraded when k =5 ,y =1 and CR =1.5 the BER
degraded by (-17.9300 dB).
The SNR at BER(
) was improved when CR =4, 3. The least amount of
improvement in BER when k =5, y =1 and CR = 3 and is equal to (2.0783 dB). The
largest amount of improvement is when k =90, y =1 and CR =4 is equal to (18.215
dB).
The SNR at BER(
) was degraded when CR = 2, except when (k =30,90 and y
=1) and CR =1.5 the SNR at BER(
) was improved by (0.5122 - 0.825 dB).
The least amount of degradation in SNR at BER(
) when k=10, y =1 and is
equal to (-0.1639 dB). The largest amount of degradation is when k=90, y =0.8 and
is equal to (-4.0838 dB).
The PAPR was improved and the least amount of improvement was when k=5, y
=1 and CR = 1.5 and is equal to (0.4088 dB), while the vast amount of
Less the amount of improvement in CCDF of PAPR when k =5, y= 1 and CR = 2
and is equal to (0.2136 dB), While the vast amount of improvement is where k
=90, y =0.2 and CR = 4 and is equal to (5.2067 dB).
The SNR at BER(
) was degraded, except when k =5 ,y =1 and CR =1.5 the
SNR at BER(
BER(
) when k =5, y =1 and CR = 3 and is equal to (-0.1171 dB). The largest
amount of degradation is when k =90, y =1 and CR =2 is equal to (-17.047 dB).
133

6.3 Hybrid RFC with companding:
The process of companding enlarges the amplitudes of the small signals, while the
peaks remain unchanged. Therefore, the average power is increased and thus the
Peak-to Average Power Ratio (PAPR) of the OFDM systems can be reduced, which
in turn helps in increasing the efficiency of the power amplifiers and also reduces the
complexity of the Analog-to-Digital Converter (ADC) and Digital-to-Analog
Converter (DAC) [53].
As we demonstrated earlier RFC batter than RCF, because it reduces the impact of the
filter on the PAPR and also when increasing I the BER improves.
The previous method hybrid technique depend on RCF and companding give a good
result and as we demonstrated earlier RFC batter than RCF, because it reduces the
impact of the filter on the PAPR and also the BER improves. When increasing I. So
we proposed a new hybrid method based on RFC and companding. The OFDM
system model with RFC and companding was shown in the Figure 6.16.
This hybrid technique shows good results better than the previous method, because of
first RFC improve the PAPR and the BER more than RCF and then companding more
reduces the amount of the PAPR.
RFC is used with the following specifications (I =4, CR =4, 3, 2) as for the
companding has been using all kinds of previous companding.
Add CP
P/
S
Companding
RFC
IDFT OR IFFT
S/
P
+pilot symbol
Signal mapper
I
/
P
D
/
A
Multipath
Fading Ch. &
noise
Remove CP
S/
P
De-Companding
DFT OR FFT
Remove
+pilot symbol
Signal
demapper
One Tap Equalizer

And P/S
O
/
P
A
/
D
Figure 6.16 the OFDM system model with RFC + companding .
6.3.1 RFC + A companding:

At these values (A =5, 10 for CR = 4, 3) and There are improved in PAPR, CCDF
of PAPR and BER). The best improvement in PAPR and CCDF of PAPR is at
A=10 for CR =3. The improvement in PAPR by = (19.8043 dB), CCDF of PAPR =
(8.4633 dB), and the SNR at BER(
) by = ( 0.5300 dB).
For SNR at BER(
)
134

The best improvement in PAPR and CCDF of PAPR is at A =20 and CR =3. The
The best improvement in CCDF of PAPR is at A =20 and CR =2. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at A = 80 and CR =4 .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at A = 100 and CR =2.
The best improvement in CCDF of PAPR is at A = 120 and CR =3. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at A=80 and CR =2. The
improvement in PAPR by = (22.2111 dB), and CCDF of PAPR = (9.8161dB),
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at A =90 and CR =2. The
comparing the proposed method with an OFDM system with A companding PAPR
reduction method:
The PAPR was improved and the least amount of improvement was when A = 40
where A = 5 and CR = 1.5 and is equal to (14.3373 dB).
The least amount of improvement in CCDF of PAPR when A = 100 and CR = 4
and is equal to (0.8784 dB), while the vast amount of improvement is where A = 5
The SNR at BER(
) was improved when CR =4,3,2 . The vast amount of
improvement is where A = 70 and CR = 4 and is equal to (6.3199 dB), while The
least amount of improvement in SNR at BER(
) when A = 5 and CR = 2 and
is equal to (0.5206 dB)
The SNR at BER(
) was degraded when CR = 1.5, when A =5. The amount of
degradation is equal to (-16 dB).
improvement was when A =5 and CR = 1.5 .The PAPR improvement is equal to
(1.8221 dB) and the CCDF of PAPR improvement is equal to(0.9794 dB), while
the vast amount of improvement is where A = 200 and CR = 4 and the PAPR
135
The SNR at BER(

) was degraded. The least amount of degradation in SNR at
BER(
) when A =5 and CR = 1.5 and is equal to (-1.38 dB). The largest
amount of degradation is when A= 80 and CR =2 is equal to (-11.2484 dB).
30
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
A
RFC (CR=4) + A
RFC (CR=3) + A
RFC (CR=2) + A
25
PAPR
20
15
10
10
12
14
16
18
20
22
SNR at (BER =10-4)
Figure 6.17.a
12
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
A
RFC (CR=4) + A
RFC (CR=3) + A
RFC (CR=2) + A
10
CCDF of PAPR
10
12
14
16
18
20
22
SNR at (BER =10 4)
Figure 6.17.b
for each of the
RFC, companding , and Hybird (RFC+ ). b) Shows the values of the CCDF of
for each of the RCF, companding , and Hybird
(RFC+ ).
136

6.3.2 RFC +
companding:
method:
At these values ( =5,10,20 for CR = 4,3) and when ( =30 ,40 for CR = 4)
There are improved in PAPR, CCDF of PAPR and SNR at BER(
)). The best
improvement in PAPR is at =40 for CR =4. The improvement in PAPR by =
(20.0157 dB) ,CCDF of PAPR = (8.3564 dB), and the SNR at BER(
) by = (
0.0116dB).
The best improvement in CCDF of PAPR is at
=20 for CR =3 . The
improvement in PAPR by = (19.9252 dB) ,CCDF of PAPR = (8.5044 dB), and
the SNR at BER(
) by = ( 0.2052 dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at = 40 and CR =3. The
For SNR at BER(
)
The best improvement in PAPR is at =180 and CR =4. The improvement in
PAPR by = (21.4145 dB) , and CCDF of PAPR = (9.0247 dB), while the SNR at
BER(
=80 and CR =3. The
improvement in PAPR by = (21.1482 dB) , and CCDF of PAPR = (9.0856 dB),
For SNR at BER(
)
The best improvement in PAPR is at =240 and CR =4. The improvement in
PAPR by = (21.605 dB), and CCDF of PAPR = (9.0964 dB), while the SNR at
BER(
=255and CR =4. The
improvement in PAPR by = (21.1370 dB),and CCDF of PAPR = (9.1356 dB),
A.22 and figure 6.18 when comparing the proposed method with an OFDM system
with companding PAPR reduction method:
The least amount of improvement in CCDF of PAPR when = 220 and CR = 4
and is equal to (0.9042 dB), while the vast amount of improvement is where = 5
The SNR at BER(
) was improved when CR =4,3. The vast amount of
improvement is where = 50 and CR = 4 and is equal to (6.4429 dB), while The
) when = 5 and CR = 3 and is
The SNR at BER(
) was improved when CR =2. The vast amount of
improvement is where = 220 and is equal to (2.4273 dB), while The least
amount of improvement in CCDF of PAPR when = 30 and is equal to (1.2065
dB)
The SNR at BER(
) was degraded when CR = 1.5, when =5 .the amount of
137

comparing the proposed method with an OFDM system with RFC method:
and CR = 1.5 and is equal to (1.4566 dB), While the vast amount of improvement
is where = 240 and CR = 4 and is equal to (10.1873 dB).
The least amount of improvement in CCDF of PAPR when = 5 and CR = 1.5 and
is equal to (0.8031 dB), while the vast amount of improvement is where = 255
The SNR at BER(
) was degraded the least amount of degradation in SNR at
BER(
) when =5 and CR = 1.5 and is equal to (-1.78 dB). The largest
amount of degradation is when = 255 and CR =2 is equal to (-10.882 dB).
30
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
MU
RFC (CR=4) + MU
RFC (CR=3) + MU
RFC (CR=2) + MU
25
PAPR
20
15
10
10
12
14
16
18
20
22
SNR at (BER =10-4)
Figure 6.18.a
12
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
MU
RFC (CR=4) + MU
RFC (CR=3) + MU
RFC (CR=2) + MU
10
CCDF of PAPR
10
12
14
16
18
20
22
SNR at (BER =10- 4)
Figure 6.18.b
for each of the
RFC, companding , and Hybird (RFC+ ) b) Shows the values of the CCDF of
138

(RFC+ )
for each of the RFC,
companding , and Hybird
6.3.3 RFC + RCT:

method:
At these values (R =0.9 ,0.8 ,0.7,0.6,0.5 and CR = 4,3,2) and when (R =0.4 and
CR = 4,3) There are improved in PAPR, CCDF of PAPR and the SNR at
BER(
) dB). These points were chosen because all the variables improved. The
best improvement in PAPR and CCDF of PAPR is at R =0.5 and CR =2. The
improvement in PAPR by = (20.5192 dB) ,CCDF of PAPR = (8.7312 dB), and the
SNR at BER(
) by = ( 0.0156 dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R = 0.4 and CR = 2 .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R = 0.2 and CR =3 .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R= 0.2 and CR =2.
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R= 0.1 and CR =3.
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R = 0.1and CR =2 .
comparing the proposed method with an OFDM system with RCT companding PAPR
reduction method:
improvement was when R =0.1 and CR = 4 .The PAPR improvement is equal to
(0.9577 dB) and the CCDF of PAPR improvement is equal to(0.3934 dB), While
the vast amount of improvement is where R =0.9 and CR = 1.5 and the PAPR
The SNR at BER(
) was improved when CR =4,3,2 . The vast amount of
improvement is where R =0.6 and CR = 4 and is equal to (6.298 dB), while The
) when R =0.4 and CR = 2 and
is equal to (2.3582 dB)
139
The SNR at BER(

) was degraded when CR = 1.5, when R =0.9 the amount of
30
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
Roots
RFC (CR=4) + Rooting
25
PAPR
20
15
10
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.19.a
12
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
Roots
10
CCDF of PAPR
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.19.b
figure 6.19 (a) Shows the values of the PAPR and SNR at BER =
for each of the
RFC, RCT, and Hybird (RFC+ RCT) b) Shows the values of the CCDF of PAPR and
SNR at BER =
for each of the RFC, RCT, and Hybird (RFC+ RCT).
The PAPR was improved and the least amount of improvement was when R =0.9
where R =0.1and CR = 4 and is equal to (12.2689 dB).
The least amount of improvement in CCDF of PAPR when R = 0.9 and CR =2
and is equal to (0.1621 dB), while the vast amount of improvement is where R=
0.1 and CR = 4 and is equal to (5.1904 dB).
The SNR at BER(
BER(
) when R =0.9 and CR = 4 and is equal to (-0.2943 dB). The largest
amount of degradation is when R= 0.1 and CR =3 is equal to (-17.4447 dB).
140

6.3.4 RFC + AEXP:
method:
At these values ((d =2-0.7 by .1 every time and CR = 4,3) and when (d =2 and
CR =2) and finally when(d =0.6 and CR =2)There are improved in PAPR, CCDF
of PAPR and SNR at BER(
)). The best improvement in PAPR and CCDF of
PAPR is at d = 0.6 and CR =4. The improvement in PAPR by = (21.0509dB)
,CCDF of PAPR = (8.7178 dB), and the SNR at BER(
) by = ( 0.0116 dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d =0.6 and CR = 3.
The improvement in PAPR by = (21.3545 dB) , and CCDF of PAPR = (8.8589
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d= 0.5 and CR =4 .
) deteriorated by= (-2.9912 dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d = 0.7 and CR =2 .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d =0.6 and CR =2. The
The PAPR was improved the least amount of improvement was when d =0.4 and
CR = 4 and is equal to (0.1528 dB), while the vast amount of improvement is
where d =2 and CR = 1.5 and is equal to (5.8486 dB).
The least amount of improvement in CCDF of PAPR when d =1.2 and CR = 4 and
is equal to (0.0659 dB), while the vast amount of improvement is where d =2 and
CR = 1.5 and is equal to (2.6752 dB).
The SNR at BER(
) was improved when CR =4,3,2 except when (d =0.4 and
CR =4, 3, 2) and when (d =0.5 and CR =2) the SNR at BER(
) maintains its
value. The vast amount of improvement is where d =0.7 and CR = 4 and is equal
to (19.5598 dB), while The least amount of improvement in CCDF of PAPR when
d=1 and CR = 2 and is equal to (0.217 dB)
The SNR at BER(
) was degraded when CR = 1.5, when d =2 The amount of
The PAPR and the CCDF of PAPR were improved except when d =2 and CR
=1.5,2 they were degraded. The least amount of improvement was when d =1.9 and
CR = 2 and the PAPR improvement is equal to (0.2144 dB) and the CCDF of
PAPR improvement is equal to (0.0784 dB), while the vast amount of
141
improvement is where d =0.4 and CR = 4 and the PAPR improvement is equal to

(10.9327 dB) and the CCDF of PAPR improvement is equal to (4.5281 dB).
The SNR at BER(
BER(
) when d =2 and CR = 4 and is equal to (-1.5141 dB). The largest
amount of degradation is when d =0.4 and CR =4 is equal to (-24.3287 dB).
30
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
AEXP
RFC (CR=4) + AEXP
RFC (CR=3) + AEXP
RFC (CR=2) + AEXP
25
PAPR
20
15
10
10
15
20
25
30
SNR at (BER =10 4)
Figure (6.20.a)
11
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
AEXP
RFC (CR=4) + AEXP
RFC (CR=3) + AEXP
RFC (CR=2) + AEXP
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
15
20
25
30
SNR at (BER =10- 4)
Figure (6.20.b)
RFC, AEXP companding , and Hybird (RFC+AEXP)
RFC, AEXP companding , and Hybird (RFC+ AEXP).
142
for each of the

for each of the

6.3.5 RFC + cos :
At these values ((y =1,0.9, 0.8,0.7 and CR = 4,3,2) and when ( y =0.6,0.5 and CR
=4,3 ) and finally at (y =.4 and CR =4) There are improved in PAPR, CCDF of
PAPR and SNR at BER(
)). The best improvement in PAPR and CCDF of
PAPR is at y =0.7 and CR =2. The improvement in PAPR by = (19.9896 dB),
CCDF of PAPR = (8.3417 dB), and the SNR at BER(
) by = (1.1469 dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y = 0.5 and CR =2 .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y =0.4 and CR =2 .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y =0.3 and CR =2.
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y=0.2 and CR =3. The
comparing the proposed method with an OFDM system with cos companding PAPR
reduction method:
improvement was when y =0.3 and CR = 4.The PAPR improvement is equal to
the vast amount of improvement is where y =1 and CR = 1.5 and the PAPR
The SNR at BER(
) was improved when CR =4, 3,2. The vast amount of
improvement when y=0.3 and CR = 4 and is equal to (7.6706 dB), while The least
amount of improvement in SNR at BER(
) when y = 0.8 and CR = 2 and is
The SNR at BER(
) was degraded when CR = 1.5, when y =1 the amount of
proposed method with an OFDM system with RFC method:
improvement was when y =1 and CR = 1.5 .The PAPR improvement is equal to
the vast amount of improvement is where y =0.1 and CR = 4 and the PAPR
143
The SNR at BER(

) was degraded, the least amount of degradation in SNR at
BER(
) when y =1 and CR = 2 and is equal to (-0.0444 dB). The largest
amount of degradation is when (y =0.2, 0.1 and CR =4)(y =0.1 and CR =3) is equal
to (-23.9287 dB).
30
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
cos
RFC (CR=4) + cos
RFC (CR=3) + cos
RFC (CR=2) + cos
25
PAPR
20
15
10
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.21.a
12
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
cos
RFC (CR=4) + cos
RFC (CR=3) + cos
RFC (CR=2) + cos
10
CCDF of PAPR
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.21.b
RCF, cos companding , and Hybird (RCF+cos)
RCF, cos companding , and Hybird (RCF+ cos)
144
for each of the

for each of the

6.3.6 RFC + NERF :
method:
At these values ( CR = 4,3) and There are improved in PAPR, CCDF of PAPR and
SNR at BER(
)). The best improvement in PAPR and CCDF of PAPR is at CR
=3. The improvement in PAPR by = (17.3493 dB) ,CCDF of PAPR = (7.3822 dB),
and the SNR at BER(
) by = ( 3.6500 dB).
30
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
RFC (CR=1.5)
NERF
RFC (CR=4) +
RFC (CR=3) +
RFC (CR=2) +
RFC (CR=1.5)
25
PAPR
20
NERF
NERF
NERF
+ NERF
15
10
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.22.a
11
original
RFC (CR=4)
RFC (CR=3)
RFC (CR=2)
RFC (CR=1.5)
NERF
RFC (CR=4) + NERF
RFC (CR=3) +NERF
RFC (CR=2) + NERF
RFC (CR=1.5) + NERF
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
15
20
25
30
SNR at (BER =10- 4)
Figure 6.22.b
Figure (a) Shows the values of the PAPR and SNR at BER =
for each of the
RCF, cos companding , and Hybird (RCF+NERF) (b)Shows 6.22 the values of the
CCDF of PAPR and SNR at BER =
for each of the RCF, cos companding , and
Hybird (RCF+ NERF)
145

6.3.7 RFC + tanhR :
) in many
points that have been tested, but The best improvement in PAPR and CCDF of
PAPR is at k=40, y = .5 for CR =4 . The improvement in PAPR by = (20.7866
dB),CCDF of PAPR = (8.5636 dB), and the SNR at BER(
) by = ( 0.1129
dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k =20, y =0.5 and CR
=2 . The improvement in PAPR by = (21.4382 dB) , and CCDF of PAPR =
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k =30, y=0.5 and CR
=2 . The improvement in PAPR by = (21.7998 dB), and CCDF of PAPR =
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k =40, y=0.5 and CR
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k=40, y=0.2 and CR
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is k=40, y=0.2 and CR =2.
an OFDM system with tanhR companding PAPR reduction method:
The PAPR was degraded at (k =20, y =1,0.8,0.5 and CR = 4, 3, 2), (k= 10, y
=1,0.8 and CR=4,3), (k =15 ,y =0.5 ,and CR =4,3) , and when (k =10 ,y =0.5 ,and
CR =4). The least amount of degradation in SNR at BER(
) when k=20, y=0.5
and CR = 2 and is equal to (-0.3872 dB). The largest amount of degradation is
when k =20, y = 1 and CR =4 is equal to (-8.4594 dB).
The CCDF of PAPR was degraded at (k =15, 20, y =1,0.8,0.5 and CR = 4, 3), (k=
20, y =1,0.8 and CR=2), (k =10 ,y =1,0.8 ,and CR =4,3) , and when (k =10 ,y =0.5
,and CR =4). The least amount of degradation in SNR at BER(
) when k=15,
y=0.8 and CR = 2 and is equal to (-0.2065 dB). The largest amount of degradation
is when k =20, y = 1 and CR =4 is equal to (-3.426 dB).
Except the points already mentioned, the PAPR was improved and the least
amount of improvement was when k =20, y =0.2 and CR = 4 and is equal to
(0.1722 dB), while the vast amount of improvement is where k =5, y =1 and CR =
1.5 and is equal to (10.4523 dB).
Except the points already mentioned, the CCDF of PAPR was improved and the
least amount of improvement in CCDF of PAPR when k =10, y =1 and CR = 2 and
is equal to (0.0055 dB), While the vast amount of improvement is where k =5, y =1
146
The SNR at BER(

) was degraded when CR = 1.5, when k =5, y =1 the
amount of degradation is equal to (-16.346 dB).
Except the points already mentioned, the SNR at BER(
) was improved and
The vast amount of improvement is where k =20, y =1 and CR = 4 and is equal to
(23.5747 dB), while the least amount of improvement in SNR at BER(
) when
k =5, y=0.2 and CR = 2 and is equal to (3.0185 dB).
an OFDM system with RFC method:
The PAPR was improved except when k =5, y =1 and CR =2 the SNR at
BER(
) degraded by (-0.1918 dB), the least amount of improvement was
when k=5, y =1 and CR = 1.5 and is equal to (0.1303 dB), While the vast amount
of improvement is where k =40, y =0.2 and CR = 4 and is equal to (12.0608 dB).
The CCDF of PAPR was improved , except when k =5,y =1 and CR =1.5,2,3,4 the
CCDF of PAPR was degraded. The least amount of improvement in CCDF of
PAPR when k =10, y= 1 and CR = 2 and is equal to (0.0255 dB), while the vast
amount of improvement is where k =40, y =0.2 and CR = 4 and is equal to (5.0792
dB).
The SNR at BER(
) was degraded, except when k =10, y =1 and CR =3 the
SNR at BER(
) was improved by (0.018 dB). The least amount of degradation
in SNR at BER(
) when k =10, y =1 and CR = 4 and is equal to (-0.1085 dB).
The largest amount of degradation is when k =30, y =0.2 and CR =2 is equal to (16.7147 dB).
6.3.8 RFC +logR :

) in many
points that have been tested, but the best improvement in PAPR and CCDF of
PAPR is at k=90, y = .5 for CR =3 . The improvement in PAPR by = (20.6844 dB)
,CCDF of PAPR = (8.6603 dB), and the SNR at BER(
) by = ( 0.2112 dB).
For SNR at BER(
)
=2 . The improvement in PAPR by = (20.9759 dB) , and CCDF of PAPR =
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k=50 ,y =0.5 and CR
=2 . The improvement in PAPR by = (21.6271 dB), and CCDF of PAPR =
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k =90 ,y= 0.5 and CR
For SNR at BER(
)
=. The improvement in PAPR by = (23.3566 dB), and CCDF of PAPR = (9.8323
For SNR at BER(
)
147

The best improvement in PAPR and CCDF of PAPR is at = and CR =. The
an OFDM system with logR companding PAPR reduction method:
The PAPR was improved and the least amount of improvement was when k =90 ,y
=1 and CR = 4 and is equal to (0.4691 dB), while the vast amount of improvement
is where k =10, y =1 and CR = 2 and is equal to (8.8239 dB).
The CCDF of PAPR was improved and the least amount of improvement was
when k =70 , y =1 and CR = 4 and is equal to (0.2321 dB), while the vast amount
of improvement is where k =10, y =1 and CR = 2 and is equal to (4.1088 dB).
The SNR at BER(
) was improved . The least amount of improvement in
SNR at BER(
) when k =90 ,y =0.2 and CR = 2 and is equal to (1.437 dB).
The largest amount of improvement is when k =90, y =1 and CR =4 is equal to
(22.5168 dB).
an OFDM system with RFC method:
The PAPR was improved and the least amount of improvement was when k=5 ,y
=1 and CR = 1.5 and is equal to (0.0422 dB), while the vast amount of
The CCDF of PAPR was improved, except when k =5 ,y =1 and CR =2 the CCDF
of PAPR was degraded by (-0.0018 dB). The least amount of improvement in
CCDF of PAPR when k =5, y= 1 and CR = 3 and is equal to (0.0727 dB), while the
vast amount of improvement is where k =90, y =0.2 and CR = 4 and is equal to
(4.8674 dB).
The SNR at BER(
) was degraded, except when k =5 ,y =1 and CR =1.5, 2. the
BER was improved by (8.7803 - 0.1415 dB). The least amount of degradation in
SNR at BER(
) when k =5, y =1 and CR = 4 and is equal to (-0.0429 dB). The
largest amount of degradation is when k =90, y =0.2 and CR =2 is equal to (14.496 dB).
6.4 Pre-coding + companding:

The companding scheme can be implemented with low complexity, without any
iterative computations, comparing with coding, partial transmit and selective mapping
schemes, in which either delay due to coding or extra overheads to synchronize
transmitter and receiver are required. On the other hand, the pre-coding has also been
considered as a best among all these techniques, because it improves the PAPR
without increasing much complexity and destroying the orthogonality between
subcarriers. The pre-coding also improves the BER in comparison to the normal
OFDM system because of diversity gain obtained due to the spreading of the data
symbol on more than one subcarrier.
The OFDM system model with the proposed technique is as shown in figure 6.23.
WHT, DCT, DST, and DHT pre-coders are used as for the companding has been
using all kinds of previous companding. The results of the proposed method are good
and the best result for the PAPR is when (DHT + tanhR).
148
Add CP
P
/
S
Companding
S
/
P
IDFT OR IFFT
Signal mapper
I
/
P
D
/
A
Multipath
Fading Ch. &
noise
S
/
P
Remove CP
De-Companding
DFT OR FFT
One Tap
Equalizer
And P/S
Signal
demapper
O
/
P
A
/
D
Figure 6.23 the OFDM system model with precoding + companding .
6.4.1 Pre-coding + A companding:

For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at A =5 and DHT . The
For SNR at BER(
)
The best improvement in PAPR is at A = 20 and DHT. The improvement in
BER(
The best improvement in CCDF of PAPR is at A = 15 and DHT. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at A =120 and DHT. The
For SNR at BER(
)
The best improvement in CCDF of PAPR is at A =120 and DST. The
proposed method with an OFDM system with A companding PAPR reduction
method:
The PAPR was improved except when (WHT and A =30, 87.6 ,100,120) the
PAPR was degraded and maximum degraded at A =30 by (-0.9647). The least
amount of improvement was when A = 40 and WHT and is equal to (0.0499 dB),
149
While the vast amount of improvement is where A = 5 and DHT and is equal to
(14.2226 dB).
The least amount of improvement in CCDF of PAPR when A = 30 and WHT and
is equal to (0.1178 dB), while the vast amount of improvement is where A = 5 and
DHT and is equal to (4.2193 dB).
The SNR at BER(
) was degraded at DST and WHT. The least amount of
degradation in The SNR at BER(
) when A =70 and WHT and is equal to (0.036 dB). The largest amount of degradation is when A =50 and WHT and is
equal to (-1.5167 dB).
The SNR at BER(
) was improved at DHT. The least amount of improvement
in The SNR at BER(
) when A =5 and is equal to (1.1517 dB). The largest
amount of improvement is when A= 120 is equal to (5.377 dB).
The SNR at BER(
) was improved at DCT and A= 30, 40, 70, 87.6, 100,120.
The least amount of improvement in The SNR at BER(
) when A =40 and is
equal to (0.0804 dB). The largest amount of improvement is when A= 70 is equal
to (0.4065 dB).
The SNR at BER(
) was degraded at DCT and A= 5, 10, 15, 20, 35, 50. The
least amount of degradation in the SNR at BER(
) when A =15 and DHT and
is equal to (-0.0769 dB). The largest amount of degradation is when A= 5 and
WHT is equal to (-0.5053 dB).
comparing the proposed method with an OFDM system with pre-coding method:
improvement was when A =5 and DHT.The PAPR improvement is equal to
(2.2752 dB) and the CCDF of PAPR improvement is equal to ( 0.9293 dB), while
the vast amount of improvement is where A = 120 and WHT and the PAPR
The SNR at BER(
) was degraded. The least amount of degradation in the
SNR at BER(
) when A =5 and DHT and is equal to (-0.8663 dB). The largest
amount of degradation is when A= 90 and WHT is equal to (-10.9249 dB).
150

30
original
WHT
DCT
DST
DHT
A
WHT + A
DCT+ A
DST + A
DHT + A
25
PAPR
20
15
10
0
10
12
14
16
18
20
22
24
SNR at (BER =10-4)
Figure 6.24.a
11
original
WHT
DCT
DST
DHT
A
WHT + A
DCT+ A
DST + A
DHT + A
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
12
14
16
18
20
22
24
SNR at (BER =10-4)
Figure 6.24.b
for each of the
precodings, companding , and Hybird (precodings + ). b) Shows the values of the
for each of the precodings, companding ,
and Hybird (precodings + ).
151

6.4.2 Pre-coding + :
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at MU =10 and DHT.
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at MU =50 and DHT. The
For SNR at BER(
)
The best improvement in PAPR is at MU =320 and DHT. The improvement in
BER(
The best improvement in CCDF of PAPR is at MU =320 and DHT. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at MU = 1000 and DHT.
For SNR at BER(
)
The best improvement in CCDF of PAPR is at MU = 1000 and DCT. The
comparing the proposed method with an OFDM system with A companding PAPR
reduction method:
The PAPR was improved except when (WHT and MU =60, 160,180) the PAPR
was degraded and the maximum degraded at MU=160 by (-0.7606 dB). The least
amount of improvement was when MU = 220 and WHT and is equal to (0.0305
dB), while the vast amount of improvement is where MU = 20 and DHT and is
The least amount of improvement in CCDF of PAPR when MU = 160 and WHT
and is equal to (0.0919 dB), while the vast amount of improvement is where MU =
5 and DHT and is equal to (3.8402 dB).
The SNR at BER(
) was degraded at DST and WHT except at (MU =700 and
WHT the SNR at BER(
) was improved by (0.1826 dB) . The least amount of
degradation in The SNR at BER(
) when MU =100 and WHT and is equal to
(-0.071dB). The largest amount of degradation is when MU =30 and DST and is
equal to (-0.8137 dB).
The SNR at BER(
in The SNR at BER(
) when MU =5 and is equal to (1.0094 dB). The largest
amount of improvement is when MU= 700 is equal to (6.2032 dB).
The SNR at BER(
) at DCT there is an improvement in some of the points
and the degradation the other .The largest amount of improvement is when MU
=120 is equal to (0.4501 dB).The largest amount of degradation is when MU =10
152

is equal to (-0.3151 dB). Is clearly the amount of improvement and degradation
less than 0.5 in all cases.
30
original
WHT
DCT
DST
DHT
MU
WHT + MU
DCT+ MU
DST + A
DHT + MU
25
PAPR
20
15
10
0
10
12
14
16
18
20
22
24
SNR at (BER =10- 4)
Figure 6.25.a
11
original
WHT
DCT
DST
DHT
MU
WHT + MU
DCT+ MU
DST + MU
DHT + MU
10
9
CCDF of PAPR
8
7
6
5
4
3
2
1
10
12
14
16
18
20
22
24
SNR at (BER =10-4)
Figure 6.25.b
for each of the
precodings, companding , and Hybird (precodings + ). b) Shows the values of the
for each of the precodings, companding ,
and Hybird (precodings + ).
153

proposed method with an OFDM system with pre-coding method:
improvement was when MU =5 and DHT .The PAPR improvement is equal to
(1.9728dB) and the CCDF of PAPR improvement is equal to (0.7742 dB), While
the vast amount of improvement is where MU = 1000 and WHT and the PAPR
equal to (8.0035 dB ).
The SNR at BER(
BER(
) when MU =5 and DHT and is equal to (-0.7449 dB). The largest
amount of degradation is when MU =1000 and WHT is equal to (-12.5842 dB).
6.4.3 Pre-coding + RCT:

method:
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R =0.6 and DHT . The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R =0.4 and DHT . The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R =0.3 and DHT. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R= 0.2 and DHT. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R = 0.1 and DHT. The
The best improvement in PAPR and CCDF of PAPR is at R = 0.2 and DST. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at R =0.1 and DST. The
154

30
original
WHT
DCT
DST
DHT
Rooting
WHT + Rooting
DCT+ Rooting
DST + Rooting
DHT + Rooting
25
PAPR
20
15
10
0
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.26.a
12
original
WHT
DCT
DST
DHT
Rooting
WHT + Rooting
DCT+ Rooting
DST + Rooting
DHT + Rooting
10
CCDF of PAPR
0
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.26.b
for each of
the precodings, RCT, and Hybird (precodings +RCT). b) Shows the values of the
for each of the precodings, RCT, and Hybird
(precodings +RCT ).
155

The following conclusion from table A.32 and figure 6.26 when comparing
the proposed method with an OFDM system with RCT PAPR reduction method:
The PAPR and the CCDF of PAPR were improved .The least amount of
improvement was when R =0.1 and WHT.The PAPR improvement is equal to
(0.218 dB) and the CCDF of PAPR improvement is equal to (0.0754 dB), while
the vast amount of improvement is where R =0.9 and DHT and the PAPR
The SNR at BER(
in The SNR at BER(
) when R = 0.8 and is equal to (0.1258 dB). The largest
amount of improvement is when R=0.1 is equal to (7.1402 dB).
The SNR at BER(
) at DCT, DST and WHT there is an improvement in some
of the points and the degradation the other .The largest amount of improvement is
when R = 0.2 and DCT is equal to (0.4058 dB).The largest amount of degradation
is when R=0.4 and WHT is equal to (-0.4257 dB). Is clearly the amount of
improvement and degradation less than 0.5 in all cases.
improvement was when R =0.9 and DHT.The PAPR improvement is equal to
(0.356 dB) and the CCDF of PAPR improvement is equal to (0.1052 dB), While
the vast amount of improvement is where R =0.1 and WHT and the PAPR
The SNR at BER(
) was degraded, except when R=0.9 and DHT the SNR at
BER(
) was improved by (0.0718 dB).The least amount of degradation in
SNR at BER(
) when R=0.9 and DCT and is equal to (-0.0147 dB). The
largest amount of degradation is when R =0.1 and WHT is equal to (-17.0412 dB).
6.4.4 Pre-coding + AEXP:

For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d =0.8 and DHT . The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d =0.4 and DHT. The
) deteriorated by (-3.1563 dB).
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at d =0.3 and DHT. The
) deteriorated by (-4.2985dB).
For SNR at BER(
)
156

The best improvement in PAPR and CCDF of PAPR is at d =0.2 and DHT . The
The PAPR was improved, except when (DHT and d =0.1, 0.2, 0.3, 0.4) and (WHT
and d =0.1) PAPR was degraded and the maximum degradation is(-2.1186 dB).the
least amount of improvement was when d = 0.2 and WHT and is equal to (0.0148
dB), while the vast amount of improvement is where d = 1.9 and DHT and is
The CCDF of PAPR was improved, except when (DHT and d =0.1, 0.2, 0.3, 0.4,
0.5) and (WHT and d =0.1, 0.2, 0.3) PAPR were degraded and the maximum
degradation is (-1.5086 dB). The least amount of improvement in CCDF of PAPR
when d =0.7 and WHT and is equal to (0.017 dB), while the vast amount of
improvement is where d =0.9 and DHT and is equal to (2.085 dB).
The SNR at BER(
in The SNR at BER(
) when d =2 and is equal to (1.5749 dB). The largest
amount of improvement is when d =0.3 is equal to (14.2701 dB).
The SNR at BER(
when d = 0.7 and WHT is equal to (11.8003 dB).The largest amount of
degradation is when d = 1 and DST is equal to (-14.3dB).
The PAPR was improved, except when DHT and d =2, 1.9, 1.8, 1.7 PAPR was
degraded and the maximum degradation is(-0.6937 dB).the least amount of
improvement was when d = 1.6 and DHT and is equal to (0.0648 dB), while the
vast amount of improvement is where d = 0.1 and WHT and is equal to (21.8494
dB).
The CCDF of PAPR was improved. The least amount of improvement in CCDF
of PAPR when d =2 and DHT and is equal to (0.1988 dB), while the vast amount
of improvement is where d =0.1 and WHT and is equal to (9.3313 dB).
The SNR at BER(
BER(
) when d=1.1 and DHT and is equal to (-0.3896 dB). The largest
amount of degradation is when d =0.2 and DHT is equal to (-18.418 dB).
157

30
original
WHT
DCT
DST
DHT
AEXP
WHT + AEXP
DCT+ AEXP
DST +AEXP
DHT +AEXP
25
PAPR
20
15
10
0
10
15
20
25
30
SNR at (BER =10 4)
Figure 6.27.a
12
original
WHT
DCT
DST
DHT
AEXP
WHT + AEXP
DCT+ AEXP
DST +AEXP
DHT +AEXP
10
CCDF of PAPR
0
10
15
20
25
30
SNR at (BER =10 4)
figure 6.27.b
for each of the
precodings, AEXP companding , and Hybird (precodings +AEXP ). b) Shows the
values of the CCDF of PAPR and SNR at BER =
for each of the
precodings,AEXP companding , and Hybird (precodings +AEXP ).
158

6.4.5 Pre-coding + cos :
method:
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y=0.6 and DHT . The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y = 0.4 and DHT .
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y = 0.3 and DHT. The
For SNR at BER(
)
The best improvement in CCDF of PAPR is at y = 0.3 and DHT. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at y = 0.2 and DST. The
comparing the proposed method with an OFDM system with cos companding PAPR
reduction method:
The PAPR was improved .The least amount of improvement was when y = 0.3
and WHT and is equal to (0.0641 dB), while the vast amount of improvement is
where y = 1 and DHT and is equal to (7.7327 dB).
The CCDF of PAPR was improved except at y =1 and WHT the CCDF of PAPR
was degraded by (-0.0179 dB). The least amount of improvement in CCDF of
PAPR when y =0.4 and WHT is equal to (0.0715 dB), while the vast amount of
improvement is where y =0.1 and DHT and is equal to (3.7092 dB).
The SNR at BER(
in The SNR at BER(
) when y = 1 and is equal to (0.0513 dB). The largest
amount of improvement is when y =0.3 is equal to (5.7052 dB).
The SNR at BER(
when y = 0.3 and DST is equal to (1.0307 dB).The largest amount of degradation
is when y =0.3 and DCT is equal to (-1.6335 dB).
The PAPR was improved, except when DHT and y =1, 9, the PAPR was degraded
and the maximum degradation is (-0.8692 dB). the least amount of improvement
was when y = 0.8 and DHT and is equal to (0.0275 dB), while the vast amount of
improvement is where y = 0.1 and WHT and is equal to (20.8281 dB).
159
of PAPR when y =1 and DHT is equal to (0.2092 dB), while the vast amount of
improvement is where y =0.1 and WHT and is equal to (8.9201 dB).
The SNR at BER(
BER(
) when y =1 and DCT and is equal to (-0.0147 dB). The largest amount
of degradation is when y=0.2 and DHT is equal to (-18.418 dB).
30
original
WHT
DCT
DST
DHT
cos
WHT + cos
DCT+ cos
DST +cos
DHT +cos
25
PAPR
20
15
10
0
10
12
14
16
18
20
22
24
26
28
30
SNR at (BER =10- 4)
Figure 6.28.a
12
original
WHT
DCT
DST
DHT
cos
WHT + cos
DCT+ cos
DST +cos
DHT +cos
10
CCDF of PAPR
0
10
12
14
16
18
20
22
24
26
28
30
SNR at (BER =10- 4)
Figure 6.28.b
for each of the
precodings, cos companding , and Hybird (precodings +cos ). b) Shows the values of
the CCDF of PAPR and SNR at BER =
for each of the precodings,cos
companding , and Hybird (precodings +cos ).
160

6.4.6 Pre-coding + tanhR :
For SNR at BER(
)
The best improvement in PAPR is at k=15, y=.8 and DHT. The improvement in
BER(
The best improvement in CCDF of PAPR is at k=20, y=1 and DHT. The
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k= 5, y =0.2 and DCT.
For SNR at BER(
)
The best improvement in PAPR and CCDF of PAPR is at k = 30, y=.2 and DCT.
an OFDM system with a tanhR companding method:
The PAPR, there is an improvement in some of the points and the degradation the
other .The largest amount of improvement is when k =5, y = 1 and DST is equal to
(15.7833 dB).The largest amount of degradation is when k =20, y =1 and WHT is
equal to (-19.3807dB).
The CCDF of PAPR, there is an improvement in some of the points and the
degradation the other .The largest amount of improvement is when k =5, y = 1 and
DST is equal to (6.7541dB).The largest amount of degradation is when k =20, y
=1 and WHT is equal to (-7.703 dB).
The SNR at BER(
) at DHT, WHT, and DCT, there is an improvement in
some of the points and the degradation the other .The largest amount of
improvement is when k=15, y = 1 and WHT is equal to (17.8776 dB).The largest
amount of degradation is when k=5, y =0.3 and DHT is equal to (-0.856 dB).
The SNR at BER(
) at DST was degraded more than 30 dB in all cases
The following conclusion from table A.36 when comparing the
proposed method with an OFDM system with pre-coding method:
The PAPR was improved, except when WHT and y =1, k =5 the PAPR was
degraded by (-0.1807 dB).the least amount of improvement was when k=10, y = 1
where k =40, y = 0.1 and WHT and is equal to (19.3685 dB).
of PAPR when k=5, y =1 and WHT is equal to (0.2938 dB), while the vast amount
of improvement is where k =40, y =0.2 and WHT and is equal to (8.3837 dB).
The SNR at BER(
BER(
) when k=30, y =1 and WHT and is equal to (-0.1416 dB). The largest
amount of degradation is when k=30, y =0.2 and DCT is equal to (-17.5814 dB).
161

6.4.7 Pre-coding + logR :
For SNR at BER(
)
The best improvement in PAPR is at k=90, y =0.5 and DHT. The improvement
in PAPR by = (21.9089 dB) , and CCDF of PAPR = (8.6957 dB), while the SNR
at BER(
The best improvement in CCDF of PAPR is at k=70, y =0.5 and DHT. The
For SNR at BER(
)
The best improvement in PAPR is at k=90, y=0.2 and DHT. The improvement
in PAPR by = (22.2456 dB), and CCDF of PAPR = (8.8018 dB), while the SNR at
BER(
The best improvement in CCDF of PAPR is at k=20, y=0.2 and DHT. The
For SNR at BER(
)
The best improvement in PAPR is at k=90 ,y =0.2and DCT. The improvement
in PAPR by = (22.8075 dB), and CCDF of PAPR = (9.5916 dB), while the SNR
at BER(
The best improvement in CCDF of PAPR is at k=10 ,y =0.2 and DST. The
For SNR at BER(
)
The best improvement in PAPR is at k=90, y =0.2 and DST. The improvement in
BER(
The best improvement in CCDF of PAPR is at k=70 , y =0.2 and DST. The
The following conclusion from table A.37 when comparing the
proposed method with an OFDM system with a logR companding method:
The PAPR, there is an improvement in some of the points and the degradation the
other .The largest amount of improvement is when k =10, y = 1 and DHT is equal
to (12.2879 dB).The largest amount of degradation is when k =70, y =1 and WHT
is equal to (-7.3198 dB).
The CCDF of PAPR, there is an improvement in some of the points and the
degradation the other .The largest amount of improvement is when k =10, y = 1
and DHT is equal to (4.9342 dB).The largest amount of degradation is when k
=90, y =1 and WHT is equal to (-2.7575 dB).
The SNR at BER(
) at DHT, WHT, and DCT, there is an improvement in
some of the points and the degradation the other .The largest amount of
improvement is when k=90 , y = 1 and WHT is equal to (17.7596 dB).The largest
amount of degradation is when k=70, y =0.2 and DHT is equal to (-0.7874 dB).
The SNR at BER(
) at DST was degraded, except when k =90 ,y =1 the SNR
at BER(
) maintains its value .The least amount of degradation in SNR at
162

BER(
) when k =50, y =1 and is equal to (-1.012 dB). The largest amount of
degradation is when k=40, y=0.2 is equal to (-14.3135dB).
an OFDM system with pre-coding method:
The PAPR was improved, the least amount of improvement was when k=5, y = 1
where k =40, y = 0.2 and WHT and is equal to (18.9242 dB).
of PAPR when k=5, y =1 and WHT is equal to (0.5366 dB), while the vast amount
of improvement is where k =90, y =0.2 and WHT and is equal to (8.2281 dB).
The SNR at BER(
) was degraded, except at WHT, y =1 and k=5, 10, 40 the
SNR at BER(
) was improved and the maximum improvement is (0.0852
dB). The least amount of degradation in SNR at BER(
) when k=50, y =1 and
WHT and is equal to (-0.1544 dB). The largest amount of degradation is when
k=50, y =0.8 and DST is equal to (-17.2375 dB).
6.4.8 Pre-coding + NERF:

This method did not work with DST and DHT, whereas the BER gives us an error.
method:
When using DCT and WHT with ERF, there are improved in PAPR, CCDF of PAPR.
The best improvement in PAPR and CCDF of PAPR is at DCT. The improvement in
PAPR by = (16.5312 dB),CCDF of PAPR = (6.8931 dB), while the SNR at
BER(
) degraded by = (-1.9789 dB).
26
original
WHT
DCT
DST
DHT
NERF
WHT + NERF
DCT+NERF
24
22
20
PAPR
18
16
14
12
10
8
6
11.5
12
12.5
13
SNR at (BER =10- 4)
Figure 6.29.a
163
13.5
14

11
original
WHT
DCT
DST
DHT
NERF
WHT + NERF
DCT+ NERF
10
CCDF of PAPR
9
8
7
6
5
4
3
11.5
12
12.5
13
13.5
14
SNR at (BER =10- 4)
Figure 6.29.b
for each of the
precodings, cos companding , and Hybird (precodings +cos ). b) Shows the values of
the CCDF of PAPR and SNR at BER =
for each of the precodings,cos
companding , and Hybird (precodings +cos ).
164
Chapter seven
Conclusions and future work

Chapter seven
7.1Conclusions:
1. The RFC and RCF can improve the PAPR and BER at the same time.
2. RFC is better than RCF in performance especially when I 2 while maintaining
the complexity and price of RCF.
3. The performance of all kinds of proposed companding is better than the
performance -law and A-law compandings
4. AEXP can be considered the best types of companding that we used in terms of
BER and performance, followed by tanhR.
5. TanhR has better results when used in the hybrid technique and also the y, k
parameters give it a kind of flexibility
6. TanhR and logR and NERF is better than tanh, log, and erf that Mohit was used in
his paper, because the performance of the proposed techniques better than -law
and A-law compandings and also the -law and A-law compandings better than
Mohit methods.
7. The performance of logR companding asymptotic to tanhR but the tanhR have
better results in most cases.
8. The performance of cos companding asymptotic to AEXP but the AEXP have
better results in most cases.
9. The best type precoding in term of reduced PAPR and BER is the DFT
10. DST and DCT precodings give almost the same performance, the DST improves
the PAPR more than DCT even a few percent.
11. The worst type of precoding in term of reducing the PAPR and BER is the WHT.
12. As it is clear from the results that the hybrid methods have better results but at the
expense of complexity.
13. The results of hybrid pre-coding with RCF is better than the results of the RCF
and pre-coding each alone, except in the case of DHT with RCF (I = 2, pilot)
where the results of the DHT is better
14. For the hybrid pre-coding with RCF, the PAPR value is better when RCF (I = 1)
because in this case the effect of the filter on the PAPR cancels.
15. The hybrid RCF with companding shows good results better than the results of the
RCF and pre-coding each alone, because of RCF reduces the PAPR and improves
the BER constant and then companding more reduces the amount of the PAPR.
16. The hybrid RCF with companding can improve the PAPR and BER at the same
time with amount greater than the RCF and the best one improvement in PAPR is
at (RCF + AEXP).
17. The hybrid RFC with companding shows good results better than the results of
the hybrid RCF with companding, because as we demonstrated earlier RFC batter
than RCF.
18. The hybrid RFC with companding can also improve the PAPR and BER at the
same time, and the best one improvement in PAPR is at (RFC + AEXP).
19. The results of the hybrid precoding with companding are provides good results
and the best result for the PAPR is when (DHT with tanhR) except at (DST with
tanhR, DST with NERF, DHT with NERF). At DST with tanhR the BER
performance significantly degraded
20. The best results are obtained at these techniques:
RFC:
A. 165
Chapter seven
The best one improvement in PAPR and CCDF of PAPR is at I =4 and CR =1.75. The
improvement in PAPR by = (18.2789 dB), CCDF of PAPR = (8.0187 dB), and the
SNR at BER (
) by = (0.6101 dB).
AEXP companding:
For SNR at BER(
)
improvement in PAPR by = (18.8515 dB), and CCDF of PAPR = (7.6480 dB), while
the SNR at BER (
LogR companding:
For SNR at BER(
)
the SNR at BER (
TanhR companding:
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k = 5, y=0.2. The
the SNR at BER (
LogR companding:
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at . The improvement in
PAPR by = (23.5788 dB), and CCDF of PAPR = (9.9600 dB), while the SNR at BER
(
Hybird
RFC + AEXP:
The best one improvement in PAPR and CCDF of PAPR is at d = 0.6 and CR =4. The
improvement in PAPR by = (21.0509dB), CCDF of PAPR = (8.7178 dB), and the
SNR at BER (
) by = (0.0116 dB).
Pre-coding + tanhR :
For SNR at BER(
)
The best one improvement in PAPR is at k=15, y=.8 and DHT. The improvement in
PAPR by = (22.7711 dB), and CCDF of PAPR = (8.9691 dB), while the SNR at BER
(
RFC + tanhR
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at k=40, y=0.2 and CR
dB), while the SNR at BER (
RFC + tanhR
For SNR at BER(
A. 166
Chapter seven
The best one improvement in PAPR and CCDF of PAPR is k=40, y=0.2 and CR =2.
while the SNR at BER (
RFC + RCT
For SNR at BER(
)
The best one improvement in PAPR and CCDF of PAPR is at R = 0.1and CR =2. The
the SNR at BER (
7.2 Future work:

1. Use another type of filter that does not effect on the PAPR or have little
impact with clipping.
2. Find a new type of companding to recduce the PAPR with maintaining the
BER performance.
3. The proposed companding PAPR reduction methods can be combined with
different PAPR reduction techniques such as PTS, SLM, TR and etc.
4. The proposed RFC can be combined with different PAPR reduction
techniques such as coding, interleaving, TI and DSI etc.
5. The RCF, proposed RFC can be combined with different existing companding
techniques such as airy companding, linear companding, Trapezoidal power
companding and etc.
6. The proposed companding PAPR reduction methods can be combined with
Zadoff-Chu matrix Transform precoding.
7. Analysis of the proposed techniques and find out its impact on the PAPR
mathematically.
8. proposed new hybrid techniques by using the proposed method
9. Study the impact of these proposed techniques on bandwidth, noise , distortion
and the ratio of power saving.
10. Study the impact of these proposed techniques on statistical distribution.
11. The proposed PAPR reduction methods can be used with MIMO OFDM
system.
12. The proposed PAPR reduction methods can be used with other multicarrier
system
A. 167
References
References
[1] K. Fazel and S. Kaiser, Multi-Carrier and Spread Spectrum Systems: From OFDM
and MC-CDMA to LTE and WiMAX, 2nd Edition, John Wiley & Sons, 2008.
[2] G. Wunder, R. F. H. Fischer, H. Boche, S. Litsyn, J,-S. No, The PAPR Problem
in OFDM Transmission: New Directions for a Long-Lasting Problem, IEEE Signal
Processing Magazine, vol. 20, no. 6, pp. 130-144, November 2013.
[3] T. Jiang, and Y. Wu An Overview: Peak-to-Average Power Ratio Reduction
Techniques for OFDM Signals IEEE Transactions on Wireless Communications,
vol.54, no2, pp.257 - 268. June 2008.
[4] L.M. Correia, D. Zeller, O. Blume, D. Ferling, Y. Jading , I. Gdor, G. Auer and
L. Van der Perre Challenges and Enabling Technologies for Energy Aware Mobile
Radio Networks, IEEE Communications Magazine, vol. 48, pp. 66-72, November
2010.
[5] S. Litsyn. Peak Power Control in Multicarrier Communications. Cambridge
University Press, January 2007.
[6] S. H. Han and J. H. Lee ,An overview of peak-to-average power ratio reduction
techniques for multicarrier transmission, IEEE Wireless Communications, ,vol.12 ,
no: 2 ,pp 56-65. April 2005.
[7] A. Gangwar, and M. Bhardwaj. An Overview: Peak to Average Power Ratio in
OFDM system & its Effect, International Journal of Communication and Computer
Technologies, vol. 01, pp.22-25, no.2, September 2012.
[8] M. M. Mowla, and S.M. M. Hasan Performance Improvement of PAPR
Reduction for OFDM Signal In LTE System, International Journal of Wireless &
Mobile Networks , vol. 5, no. 4,pp. 35-47,August 2013.
[9] R.W. Bauml, R.F.H. Fischer, and J. B. Huber Reducing the peak-to-average
power ratio of multicarrier modulation by selected mapping, IEEE Electronics
Letters ,vol. 33, pp. 2056 - 2057, Oct 1996.
[10] H. D. Joshi. Performance augmentation of OFDM system, Ph.D. dissertation,
Jaypee Univ. of engineering and Technology, India, May 2012.
[11] M. Park, H. Jun, J. Cho, N. Cho, D. Hong and C. Kang PAPR Reduction in
OFIIM Transmission using Hadamard Transform, IEEE International Conference on
Communications, vol.1, pp. 430-433, Jun 2000.
[12] J. Armstrong. New OFDM Peak-to-Average Power Reduction Scheme, IEEE
VTS 53rd Vehicular Technology Conference, Spring, vol. 01, pp. 756 760, May
2001.
168
References
[13] J. Armstrong. Peak-to-average power reduction for OFDM by repeated clipping
and frequency domain filtering, IEEE Electronics Letters, vol. 38, pp. 246 - 247, Feb
2002.
[14] R. C. Hincapie and J. E. Sierra. Advanced Transmission Techniques in WiMAX.
InTech, January 2012.
[15] T. Jiang, Y. Yang, and Y. Song. Companding Technique for PAPR Reduction
in OFDM Systems Based on An Exponential Function, Global Telecommunications
Conference, GLOBECOM '05. IEEE , vol. 05, pp. 2798-2801, Dec. 2005.
[16] W. F. Al-Azzo, B. M. Ali, S. Khatun and S. Jamalullail Time Domain
aStatistical Control for PAPR Reduction in OFDM System Proceedings of AsiaPacific Conference on Communications 2007, pp. 141-144.
[17] P. BOONSRIMUANG, K. MORI, T. PAUNGMA, and H. KOBAYASHI
Proposal of Simple PAPR Reduction Method for OFDM Signal by Using Dummy
Sub -Carriers The Institute of Electronics, Information and Communication
Engineers, vol. E91B, no. 3, pp. 784-794, march 2008.
[18] C. A. Devlin, A. Zhu, and T.J. Brazil Peak to Average Power Ratio Reduction
Technique for OFDM Using Pilot Tones and Unused Carriers, Radio and Wireless
Symposium, 2008 IEEE, pp. 33 36, Jan. 2008.
[19] K. TAKEDA and F. ADACHI A PAPR Reduction Scheme Without Side
Information For OFDM Signal Transmissions, IEEE Transactions on Signal
Processing, vol. 60, pp. 3657- 3669, July 2012.
[20] Z.Wang, S. zhang, and B. qiu.PAPR Reduction of OFDM Signal by Using
Hadamard Transform in Companding Techniques, IEEE International Conference on
Communication Technology (ICCT), pp. 320 323, Nov. 2010.
[21] I. Baig and V. Jeoti PAPR Analysis of DHT-Precoded OFDM System for MQAM, IEEE 2010 International Conference on Intelligent and Advanced Systems
(ICIAS), pp.1-4, June 2010.
[22] Z. Wang. Combined DCT and Companding for PAPR Reduction in OFDM
Signals, Journal of Signal and Information Processing, vol. 2, pp. 100-104, 2011.
[23] H. D. Joshi and R. Saxena PAPR Reduction in OFDM Systems Using
Precoding With Clipping IEEE Communications, Computing and Control
Applications, 2011, pp.1-5.
[24] B. P. Lathi, Modern Digital and Analog Communication Systems. New York:
Oxford University Press, 1998.
[25] M. Chauhan, and A. Chobe PAPR Reduction in OFDM system Using Tone
Reservation Technique, IJCTEE, vol. 2, pp. 57- 60, August 2012.
169
References
[26] E. V. Cuteanu and D. Isar Hybrid PAPR Reduction Scheme using Walsh
Hadamard Precoding and Signal Companding International Symposium on
Electronics and Telecommunications, 2012, pp. 195-198.
[27] C. Hsu, and H. Liao PAPR Reduction Using the Combination of Precoding and
Mu-Law Companding Techniques for OFDM Systems IEEE 11th International
Conference on Signal Processing, 2012, vol.1, pp.1-4.
[28] S. Abouty, L. Renfa, Z. Fanzi and F. Mangone A Novel Iterative Clipping and
Filtering Technique for PAPR Reduction of OFDM Signals: System Using
DCT/IDCT Transform, International Journal of Future Generation Communication
and Networking, vol. 6, no. 1 , pp. 1-8, February 2013.
[29] Z. You, I. Lu, R. Yang, J. Li Flexible Companding Design for PAPR Reduction
in OFDM and FBMC Systems, IEEE International Conference on Computing,
Networking and Communications, 2013, pp. 408-412.
[30] M. K. Singh, P. Siingh , A. Gupta, S. S. Sharma, and S. K. Singh Performance
Evaluation of Different Companding Techniques for Peak-to-Average Power
Reduction of an OFDM Signal, International Journal of Advances in Electrical and
Electronics Engineering, vol.2, no. 1, pp. 204-209, 2013.
[31] T. Jiang, W. Xiang, P. C. Richardson, D. Qu, and G. Zhu On the Nonlinear
Companding Transform for Reduction in PAPR of MCM Signals, IEEE
Transactions On Wierless Communications, vol. 6, no. 6, June 2007.
[32] C. Wang and S. Ku A Low-Complexity Companding Transform For Peak To
Average Power Ratio Reduction In OFDM Systems, IEEE International Conference
on Acoustics, Speech and Signal Processing, 2006, vol. 4, pp. 329-332.
[33] S. A. Aburakhia, E.F. Badran and D. A. E. Mohamed Linear Companding
Transform for the Reduction of Peak-to Average Power Ratio of OFDM Signals,
IEEE Transactions on Broadcasting, vol. 55, pp. 155 -160, February 2009.
[34] O. Gazi A New Companding Technique f or PAPR Reduction in OFDM
Communication Systems, IEEE International Congress on Ultra-Modern
Telecommunications and Control Systems and Workshops, 2011, pp.1-5.
[35] K.H. Sankar, K.A. Babu Airy Function Based PAPR Reduction Method for
OFDM Systems IJMER, vol.2, pp-4295-4297, Nov-Dec. 2012.
[36] J. Virdi , and S. Kumar PAPR Reduction Based on Precoding Techniques with
Companding in OFDM Systems, International Journal of Scientific & Engineering
Research, vol. 4, pp.1064 -1070, May-2013.
[37] N. Kaur and L. Kansal Peak To Average Power Ratio Reduction Of OFDM
Signal By Companding Clipping With Walsh Hadamard Transform, International
Journal of Wireless & Mobile Networks , vol. 5, no. 1,pp.33-45 , February 2013.
170
References
[38] M. K .Singh , A. Gupta, P. Siingh, S. S.Sharma, and S.t K. Singh Hybrid
Clipping-Companding Schemes for Peak to Average Power Reduction of OFDM
Signal, International Journal of Advances in Engineering Science and Technology,
vol. 2, no. 1, pp. 114-119, 2013.
[39] K. MURALIBABU, and M.SUNDHARARAJAN DCT Based PAPR and ICI
Reduction Using Companding Technique in MIMO-OFDM System , International
Journal of Advanced Research in Computer Science and Software Engineering. vol. 3,
pp. 49-53, August 2013
[40] Jijina N., and S. S. Pillai. Linear Precoding Schemes for PAPR Reduction in
Mobile WiMAX OFDMA System, International Journal of Advanced Research in
Electrical,
Electronics and Instrumentation Engineering, vol. 3,pp. 8873-8884, April 2014.
[41] H. Zarrinkoub, Understanding LTE With MATLAB, Massachusetts, USA:
John Wiley & Sons, Ltd, 2014.
[42] G. M. Kebede Performance Evaluation of LTE Downlink with MIMO
Techniques. M.A. thesis, School of Engineering, Blekinge Institute of Technology,
Sweden, November 2010.
[43] X. Zhang and X. Zhou, LTE-Advanced Air Interface Technology. CRC Press,
2013.
[44] S. Sesia, I. Toufik, and M. Baker, LTE The UMTS Long Term Evolution. 2nd
Edition., John Wiley & Sons, 2011.
[45] M. oa The challenges of LTE technologies M.A. thesis, polytechnic
university of Zagreb, June 2013.
[46] 3GPP TS 36.300, LTE; Evolved Universal Terrestrial Radio Access (E-UTRA)
and Evolved Universal Terrestrial Radio Access Network (E-UTRAN); Overall
Description; Stage 2 (Release 10), version 10.7.0, April 2010.
[47] H. Holma and A.Toskala, LTE for UMTS- OFDMA and SC-FDMA Based
Radio Access. John Wiley & Sons, 2009.
[48] E. Dahlman, S. Parkvall, J. Skld, and P. Beming. 3G Evolution: HSPA and LTE
for Mobile Broadband. Academic Press, July 2007.
[49] F. KHAN. LTE for 4G Mobile Broadband: Air Interface Technologies and
Performance. Cambridge University Press, April 2009.
[50] Fact sheet 3GPP LTE. The "Long Term Evolution" air interface. Federal Office
of Communications OFCOM, January 2011.
[51] T. Bhandare, LTE and WiMAX Comparison, Santa Clara University, 2008.
171
References
[52] A. ARUN DASH OFDM Systems And PAPR Reduction Techniques In OFDM
Systems National Institute Of Technology, Rourkela, 2010.
[53] J. Zyren Overview of the 3GPP Long Term Evolution Physical Layer,
White Paper, freescale semiconductor, 2007.
[54] S. M. Abbas, M. N. Abbas, and S. A. Mohammed Slantlet Transform-Based
OFDM Scheme, Journal of Engineering, vol. 13,no. 3,pp. 1638-1647,September
2006.
[55] V.R.Prakash, and P.Kumaraguru Performance analysis of OFDM with QPSK
using AWGN and Rayleigh Fading Channel, Conference: ICIESP, 2012, pp.10 -15.
[56] R. Prasad l and F. J. Velez. WiMAX Networks: Techno-Economic Vision and
Challenges. Springer, 2010.
[58] M. Sajedin, A. Ghorbani, and H. R. A. Davar Nonlinearity Compensation for
High Power Amplifiers Based on Look-Up Table Method for OFDM
TransmittersInternational Journal of Advanced Computer Science and Information
Technology, vol. 3, no. 4, pp. 354-367, 2014.
[59] F.A. Ibikunle, MNSE, and MIEEE Broadband Wireless Access Solution based
on OFDMA Technique in WiMAX IEEE 802.16, The Pacific Journal of Science and
Technology, vol. 10, no 1, pp. 286-294, May 2009.
[60] Chang R. W., Synthesis of band-limited orthogonal signals for multichannel
data transmission, Bell Systems Technical Journal, vol. 46, pp. 1775-1796, Dec.
1966.
[61] Part 16: Air Interface for Fixed Broadband Wireless Access Systems Amendment
2: Medium Access Control Modifications and Additional Physical Layer
Specifications for 2-11 GHz, IEEE, Standard IEEE Std. 802.16a-2003, 2003.
[62] A. Goel Improved PAPR Reduction in OFDM Systems, Ph.D. dissertation,
JAYPEE Institute of Information Technology, India, April 2013.
[63] B. R. Saltzberg, Performance of an efficient parallel data transmission
system,IEEE Transactions on Communication Technology, vol. COM-15, no.6,
pp.805813, Dec.1967.
[64] R. W. Chang ,Synthesis of band-limited orthogonal signals for multichannel
data transmission, Bell System Technology Journal, vol. 45, pp. 17751796, Dec.
1966.
[65] R. W. Chang and R. A. Gibby ,A theoretical study of performance of an
orthogonal multiplexing data transmission scheme, IEEE Transactions on
Communication Technology, vol. COM-16, no.4, pp. 529540, August 1968.
172
References
[66] H. F. Marmuth, On the transmission of information by orthogonal time
functions, AIEE Transactions of communication and Electronics, vol. 79, pp. 248
255, July 1960.
[67] V.Tarokh, New Directions in Wireless Communications Research. Spring,
September 7, 2009.
[68] Despain A. M., Very Fast Fourier Transform Algorithms Hardware for
Implementation, IEEE Trans. Comp., Vol. C-28, no. 5, pp. 333-341, May 1979.
[69] Bidet E., Castelain D., Joanblanq C. , Senn P., A fast single-chip
implementation of 8192 complex point FFT, IEEE Journal of Solid State Circ., Vol.
30, no. 3, pp. 300-305, Mar. 1995.
[70] R. Peled, A. Ruiz, Frequency domain data transmission using reduced
computational complexity algorithms, in Proceeding of the IEEE International
Conference on Acoustics, Speech, and Signal Processing, ICASSP 80 ,USA, pp.
964967,1980.
[71] Chow P. S., Dhahir N. AL, Cioffi J.,Bingham M., A Multicarrier E1-HDSL
transceiver system with coded modulation, J. Eur. Transactions on Telecom. Rel.
Tech. (ETT) (Special Issue: Applications of Coded Modulation Techniques), vol. 4,
no. 3, pp: 257 66, June 1993.
[72] Chow P.S., Tu J.C. and Cioffi J.M., Performance Evaluation of a Multichannel
Transceiver System for ADSL and VHDSL services", IEEE Journal on Selected
Areas in Communications, vol. 9, no. 6, Aug. 1991.
[73] ETSI, Radio broadcasting systems; Digital Audio Broadcasting (DAB) to
mobile, portable and fixed receivers, European Telecommunication Standard,
Standard EN-300- 401, May 1997.
[74] ETSI, Digital Video Broadcast (DVB); Framing Structure, Channel Coding and
Modulation for Digital Terrestrial Television, ETSI EN 300 744 v1.1.2, 1997.
[75] ETSI, Broadband Radio Access Networks (BRAN): Hiperlan Type 2;
Technical Specification Part 1-Physical Layer, DTS/BRAN 030003-1, Oct. 1999.
[76] B. Sklar, Digital Communications. Fundamentals and Applications. 2nd Edition,
Prentice Hall, 2001.
[77] F. Tariq Impact of PAPR on Link Adaptation Strategies of OFDM Based
Systems, M.A. thesis, Chalmers University of Technology, May 2007.
[78] G. Hill, Peak Power Reduction in Orthogonal Frequency Division Multiplexing
Transmitters. Ph.D. Thesis, Victoria University of Technology, March 2011.
[79] D. Mati OFDM as a possible modulation technique for multimedia
applications in the range of mm waves, CiteSeerX, 1998.
173
References
[80] A. Sahu , and S. Behera PAPR analysis and channel estimation techniques for
3GPP LTE system. National Institute of Technology Rourkela , 2011.
[81] S. Bhavsar, H. Pandey, P. Chouhan Design and Implementation of OFDM
Trans-Receiver for IEEE 802.11(WLAN), International Journal Of Modern
Engineering Research, vol. 4, pp. 55-67, Jan. 2014.
[82] J. Mohanty Analysis And Study Of Multi-Symbol Encapsulated Orthogonal
Frequency Division Multiplexing. National Institute of Technology, Rourkela,
2010.
[83] E.Dahlman. S. Parkvall, and J. Skld ,4G LTE/LTE-Advanced for Mobile
Broadband. Academic Press, 2011.
[84] S. L. Morancho. Access modes based on coordinated multipoint for relay
transmissions in 4G systems. M.A. thesis, University of Catalonia, 2011.
[85] S. K. Rangineni. Multihop Concept in Cellular Systems. M.A. thesis, Czech
Technical University, Sweden, 2008.
[86] J. A. Pereira Small Cell Deployment Evaluation on LTE, M.A. thesis, Tcnico
Lisboa, October 2013.
[87] T. Zemen OFDMA/SC-FDMA Basics for 3GPP LTE (E-UTRA),
Forschungszentrum Telekommunikation Wien, April 24, 2008.
[88] S. Singh, M. Kaur Physical Layer Simulation of WIMAX 802.16, IJECT, vol.
4, pp.84-87, July - September 2013.
[89] S. Singh IMT-Advanced Requirements for 4G Technology and its Components
IJECT, vol. 4, pp. 58-60, Jan - March 2013.
[90] J. Zyren Overview of the 3GPP Long Term Evolution Physical Layer. White
Paper, Freescale Semiconductor, July 2007.
[91] C. Cox. An Introduction To LTE: LTE, LTE-Advsnced, SAE And 4G Mobile
Communications. John Wiley & Sons, April 2012.
[92] F. E. Abd El-Samie, F. S. Al-kamali, A. Y. Al-nahari, and M. I. Dessouky .SCFDMA for Mobile Communications. CRC Press, July 2013
[93] LTE in a Nutshell: The Physical Layer White paper, Telesystem Innovations
Inc, 2010.
[94] R. Siddavaatam Joint Graph Coloring And Ant Colony Optimization Based
Sub-Channel Allocation Algorithms For Downlink LTE-A Hetnets, M.A. thesis,
Toronto, Ontario, Canada, August 2012.
[95] C. Choudhary, and V. Gupta A Study of Performance Enhancement Schemes
for Multicarrier Transmission, International Journal of Computer Applications, vol.
68, no. 5, pp. 50-54, April 2013.
174
References
[96] K. Srinivasarao, B. Prabhakararao, and M. V. S. Sairam Peak-to-Average Power
Reduction In MIMO OFDM Systems using Sub-Optimal Algorithm,International
Journal of Distributed and Parallel Systems, vol.3, no.3, pp. 261-273, May 2012.
[97] N. Kaur, and A. S. Sappal Peak to Average Power Ratio Measurement Using
Oversampling Technique International Journal for Multi Disciplinary Engineering
and Business Management, vol. 2, pp. 125-129, July-September 2014.
[98] Y. S. Cho, J. Kim, W. Y. Yang and C. G. Kahn, MIMO OFDM wireless
communications with Matlab. Asia: John Wiley & Sons, 2010.
[99] R. V. Nee and R. Prasad, OFDM for Wireless Multimedia Communication.
London: Artech house Publisher, 2000.
[100] S. Khalid Peak to Average Power Ratio Reduction in Orthogonal Frequency
Division Multiplexing Systems. Ph.D. dissertation, University of Engineering and
Technology Taxila, Pakistan, Dec. 2009.
[101] R. D. Kanti, and R.V. Ch. Sekhar Rao Systematic Comparison of Different
PAPR Reduction Methods in OFDM Systems International Journal of Electronics
and Communication Engineering, vol. 7, no.1, pp. 21-30, 2014.
[102] R. Rajbanshi OFDM-Based Cognitive Radio for DSA Networks. Ph.D.
dissertation, The University of Kansas, September 2007.
[103] V. Vijayarangan, and R. Sukanesh An Overview Of Techniques For Reducing
Peak To Average Power Ratio And Its Selection Criteria For Orthogonal Frequency
Division Multiplexing Radio Systems, Journal of Theoretical and Applied
Information Technology, vol. 5, no. 1, pp.25-36, January 2009.
[104] Y. G. Li and G.L. Stuber, Orthogonal Frequency Division Multiplexing for
Wireless Communications. New York, NY, USA: Springer, 2006.
[105] P. H. Lehne, and F. Bhagen .OFDM(A) for wireless communication .Telenor
ASA, 2008.
[106] Sanjeev Saini, and O.P. Sahu Peak to Average Power Ratio Reduction in
OFDM System by Clipping and Filtering, International Journal of Electronics
Communication and Computer Technology, vol. 2, pp.105-109, May 2012.
[107] A.Tiwari, and K. Markam An Overview: Peak-to-Average Power Ratio
Reduction Techniques for OFDM Signals, International Journal of Computer &
Communication Engineering Research, vol. 2, pp.23-28, January 2014.
[108] T. Jiang, and Y. Wu, An Overview: Peak-to-Average Power Ratio Reduction
Techniques for OFDM Signals, IEEE Transactions on Broadcasting, vol. 54, no. 2,
pp.257-268, June 2008.
175
References
[109] M. El-Tarhuni, M. Hassan, and A. B. Sediq, A Joint Power Allocation and
Adaptive Channel Coding Scheme for Image Transmission over Wireless Channels,
International Journal of Computer Networks & Communications (IJCNC), Vol.2,
No.3, May 2010.
[110] R. Rajbanshi OFDM-Based Cognitive Radio for DSA Networks. Technical
Report, The University of Kansas, September 2007.
[111] X. Li and L. J. Cimini, Effects of clipping and filtering on the performance of
OFDM, IEEE Commun. Lett., vol. 2, pp. 131133, May 1998.
[112] T. A. Wilkison and A. E. Jones, Minimization of the peak to mean envelope
power ratio of multicarrier transmission schemes by block coding, in Proc. 45th
IEEE Veh. Technol. Conf., vol. 2, Chicago, IL, USA, July 1995, pp. 825 829.
[113] H. Chen and A. M. Haimovich, Iterative estimation and cancellation of
clipping noise for OFDM signals, IEEE Commun. Lett., vol. 7, pp. 305307, July
2003.
[114] L. Wang and C. Tellambura, A simplified clipping and filtering technique for
PAR reduction in OFDM systems, IEEE Signal Processing Lett., vol. 12, pp. 453
456, June 2005.
[115] S.-K. Deng and M.-C. Lin, OFDM PAPR reduction using clipping with
distortion control, in Proc. IEEE Int. Conf. Commun., vol. 4, Seoul, Korea, May
2005, pp. 2563 2567.
[116] S. C. Thompson, J. G. Proakis, and J. R. Zeidler, The effectiveness of signal
clipping for PAPR and total degradation reduction in OFDM systems, in Proc. IEEE
Global Telecommun. Conf., St. Louis, MO, USA, Nov. 2005, pp.28072811.
[117] R. van Nee and A. de Wild, Reducing the peak-to-average power ratio of
OFDM, in Proc. 48th IEEE Veh. Technol. Conf., Ottawa, Ontario, Canada, May
1998, pp. 20722076.
[118] S. H. Han and J. H. Lee, Peak-to-average power ratio reduction of an OFDM
signal by signal set expansion, in Proc. IEEE Int. Conf. Commun.,vol. 2, Paris,
France, June 2004, pp. 867971.
[119] P. Foomooljareon, W. A. C. Fernando, and K. M. Ahmed, PAPR reduction of
OFDM systems using input sequence envelope scaling, in Proc. 57th IEEE Veh.
Technol. Conf. - Spring, vol. 2, Jeju, Korea, Apr. 2003, pp. 1243 1247.
[120] B. S. Krongold and D. L. Jones, PAR reduction in OFDM via active
constellation extension, IEEE Trans. Broadcast., vol. 49, pp. 258 268, Sept. 2003.
[121] Y. Kou, L. Wu-Sheng, and A. Antoniou, New peak-to-average powerratioreduction algorithms for multicarrier communications, IEEE Trans. Circuits
Syst., vol. 51, pp. 1790 1800, Sept. 2004.
176
References
[122] S. H. Muller and J. B. Huber, OFDM with reduced peak to average power ratio
by optimum combination of partial transmit sequences, Electron. Lett., vol. 33, pp.
368 369, Feb. 1997.
[123] H. Nikookar and K. S. Lidsheim, Random phase updating algorithm for
OFDM transmission with low PAPR, IEEE Trans. Broadcast., vol. 48,
no. 6, pp. 123 128, June 2002.
[124] N. Y. Ermolova and P. Vainikainen, On the relationship between peak
factor of a multicarrier signal and aperiodic autocorrelation of the generating
sequence, IEEE Commun. Lett., vol. 7, pp. 107108, Mar. 2003.
[125] S. Y. Le Goff, S.S. Al-Samahi, B.K. Khoo, C.C. Tsimenidis, and B.S. Sharif
(July 2009) Selected mapping without side information for PAPR reduction in
OFDM, IEEE Transactions on Wireless Communication, vol. 8, no. 7, pp. 33203325.
[126] P. van Eetvelt, G. wade, and M. Tomlinson, Peak to average power reduction
for OFDM schemes by selective scrambling, Electron. Lett., vol. 32, pp. 19631964,
Oct. 1996.
[127] A. D. S. Jayalath and C. Tellambura, The use of interleaving to reduce
the peak to average power ratio of an OFDM signal, in Proc. IEEE Global
Telecommun. Conf., vol. 1, San Francisco, CA, USA, Nov. 2000, pp. 82 86.
[128] S. C. Thompson, Constant envelope OFDM phase modulation, Ph.D.
dissertation, University of California, San Diego, CA, USA, 2005.
[129] K. K. Sen, and V. sahu, PAPR Reduction Using PTS with DCT SLM
Technique, irdindia., Vol. 1, no.3, pp. 23-25, 2014.
[130] B. Torun Peak-to-Average Power Ratio Reduction Techniques for Wavelet
Packet Modulation. M.A. thesis, Department of Telecommunications at Delft
University of Technology, Delft, The Netherlands. 27 October 2010.
[131] Md. A. Islam, N. Ahmed, N. U.Ahamed, M. Rahman, S. A. Aljunid, PAPR
Reduction in an OFDM system using Recursive Clipping and Filtering Technique,
WSEAS TRANSACTIONS on COMMUNICATIONS, Vol. 13, pp.291-297, 2014.
[132] M. Bala, M. Kumar, K. Rohilla, PAPR Reduction in OFDM Signal Using
Signal Scrambling Techniques, International Journal of Engineering and Innovative
Technology (IJEIT), Vol. 3, No. 11, pp.140-143, May 2014.
[133] S. Sujitha, and R. Ramachandran Performance Analysis of PAPR Reduction in
MIMO OFDM System Using Modified Constant Modulus Algorithm, International
Journal of Advanced Research in Electrical, Electronics and Instrumentation
Engineering, Vol. 3, No. 2, pp520-526, April 2014.
177
References
[134] M. F. Ghanim, and M.F.L.Abdullah, Reduction Techniques of Peak-toAverage Power Ratio in MCCDMA Systems, International Journal of Informatics
and Communication Technology (IJ-ICT), Vol.1, No.2, pp. 82-90, December 2012.
[135] M. Gouda, K. A. Shehata, M. Hussien, PAPR Reduction Performance for LTE
OFDM Systems with Different Techniques, International Journal of Scientific &
Engineering Research, Vol. 4, No. 5, pp. 2335-2344, May 2013.
[136] H. Sakran, M. Shokair, and A. Abou Elazm Combined Interleaving And
Companding For PAPR Reduction in OFDM Systems, Progress In Electromagnetics
Research C, vol. 6, pp. 6778, 2009.
[137] H. Chen and H. Liang. Combined Selective Mapping and Binary Cyclic Codes
for PAPR Reduction in OFDM Systems. IEEE Transactions on Wireless
Communications, vol.6, pp. 3524 - 3528. Oct.2007.
[138] K. M. K. Pervez, and M. M. Hossain A New Proposed Scheme for PAPR
Reduction of OFDM System Combining Hadamard Transform and Hann Peak
Windowing, International Journal of Computer and Information Technology, vol. 2,
pp. 62-66, 2012.
[139] M. M. Mowla, L. C. Paul and M. R. Hasan COMPARATIVE
PERFORMANCE ANALYSIS OF DIFFERENT MODULATION TECHNIQUES
FOR PAPR REDUCTION OF OFDM SIGNAL,, International Journal of Computer
Networks & Communications (IJCNC) Vol.6, No.3,pp.63-73 , May 2014.
[140] C. R. ANTHIKKAD ,and I. A. BAIG Performance Enhancement Of OFDM
In PAPR Reduction Using New Companding Transform And Adaptive AC
Extension Algorithm For Next Generation Networks. M.A. thesis, School of
Engineering Blekinge, Institute of Technology, Sweden, September 2013.
[141] M. Z. Parvez ,and Md. A. Al Baki Peak To Average Power Ratio (PAPR)
Reduction In OFDM Based Radio Systems, M.A. thesis, School of Engineering,
Blekinge Institute of Technology, Sweden, May 2010.
[142] B. U. Rindhe, J.Digge, and S. K. Narayankhedkar, Performance Analysis of
OFDM Based System for PAPR Reduction Techniques with Optical Fiber Link,
International Journal of Engineering Science and Innovative Technology (IJESIT)
Vol. 3, No. 2, pp.20-32, March 2014.
[143] Van Nee, R., and Wild, A., Reducing the peak to average power ratio of
OFDM, IEEE Vehicular Technology Conference, Vol. 3, May 1998.
[144] Wilkison, T. A. and Jones A. E., "Minimization of the Peak to mean Envelope
Power Ratio of Multicarrier Transmission Schemes by Block Coding," IEEE,
Vehicular Conference, Vol.2, Jul. 1995.
[145] Foomooljareon, P. and Fernando, W.A.C., Input sequence envelope scaling in
PAPR reduction of OFDM, IEEE 5th international symposium on wireless personal
multimedia communications, Vol. 1, Oct 2002.
178
References
[146] P. Dhok, S.V. Rathkantiwar, Performance Improvement in BER and PAPR
Reduction in OFDM System Using Companding Technique, International Journal of
Application or Innovation in Engineering & Management (IJAIEM), Volume 2,
No.1,pp.232-235, January 2013.
[147] S. A. Aburakhia, E. F. Badran, and D. A. E. Mohamed Linear Companding
Transform for the Reduction of Peak-toAverage Power Ratio of OFDM Signals,
IEEE Transactions on Broadcasting, Vol.55 , No.1 , pp.155-160, March 2009.
[148] L. Shu-hua, L. Yong-zhao1, Z. Hai-lin1, and L. Ying-ting , Threshold-based
piecewise companding transform for PAPR reduction in OFDM systems, The
Journal of China Universities of Posts and Telecommunications, No.2, April 2013.
[149] B. Ragini, M. S. Babu ,and K. K. Rao , Companding Technique for Reducing
Peak-to-Average Power Ratio in OFDM Linear Coded Systems, International
Journal of Power System Operation and Energy Management, Vol.1, No. 2,pp. 48-53,
2011.
[150] T. Jiang, W. Yao, P. Guo, Y. Song, and D. Qu , Two Novel Nonlinear
Companding Schemes With Iterative Receiver to Reduce PAPR in Multi-Carrier
Modulation Systems, IEEE TRANSACTIONS ON BROADCASTING, Vol.. 52,
No. 2, June 2006.
[151] T. Moazzeni , H. Selvaraj ,and Y. Jiang, A Novel Multi Exponential Functionbased Companding Technique for Uniform Signal Compression over Channels with
Limited Dynamic Range , International Journal of Electronics and
Telecommunications Vol. 56, pp.125-128, Jun 2010.
[152] W. F. Al-Azzo, and B. M. Ali Adaptive Square-Rooting Companding
Technique for PAPR Reduction in OFDM Systems, World Academy of Science,
Engineering and Technology, vol.5, pp. 507-511, 2011.
[153] T. Jiang, Y. Yang, and Y. Song Exponential Companding Technique for PAPR
Reduction in OFDM Systems IEEE Transactions on Broadcasting, vol. 51,pp.244247 , no. 2, June 2005.
[154] Yogita1 and R. Kumar A Review on PAPR Reduction Techniques
International Journal of Engineering, Applied and Management Sciences Paradigms,
vol. 16, pp.45-48, June 2014.
[155] A. E. Jones, T. A. Wilkinson, and S. K. Barton, Block coding scheme for
reduction of peak-to-average envelope power ratio of multicarrier transmission
systems, IEE Electronics Letters, vol. 30, no. 8, pp. 20982099, Dec. 1994.
[156] D. Wulich, Reduction of peak to mean ratio of multicarrier modulation using
cyclic coding, IEE Electronics Letters, vol. 32, no. 29, pp. 432433, Feb. 1996.
179
References
[157] S. Fragicomo, C. Matrakidis, and J. J. OReilly, Multicarrier transmission peakto-average power reduction using simple block code, IEEElectronics Letters, vol. 34,
no. 14, pp. 953954, May 1998.
[158] T. Jiang and G. X. Zhu, OFDM peak-to-average power ratio reduction by
complement block coding scheme and its modified version, in The 60th IEEE
Vehicular Technology Conference 2004 Fall, Los Angeles, USA, pp. 448451, Sept.
2004.
[159] T. Jiang and G. X. Zhu, Complement block coding for reduction in peak-toaverage power ratio of OFDM signals, IEEE Communications Magazine, vol. 43, no.
9, pp. S17S22, Sept. 2005.
[160] S. Boyd, Multitone signals with low crest factor, IEEE Trans. Circuits
Systems, vol. CAS-33, no. 10, pp. 10181022, Oct. 1986.
[161] B. M. Popovic, Synthesis of power efficient multitone signals with flat
amplitude spectrum, IEEE Trans. Communications, vol. 39, no. 7, pp. 10311033,
Jul. 1991.
[162] T. Jiang, G. X. Zhu, and J. B. Zheng, A block coding scheme for reducing
PAPR in OFDM systems with large number of subcarriers, Journal of Electronics,
vol. 21, no. 6, pp. 482489, Dec. 2004.
[163] J. A. Davis and J. Jedwab, Peak-to-mean power control in OFDM,Golay
complementary sequences and Reed-Muller codes, IEEE Trans. Information
Theory, vol. 45, no. 7, pp. 23972417, Nov. 1997.
[164] C. L. Wang, S. J. Ku, and C-Ju Yang, A low-complexity PAPR estimation
scheme for OFDM signals and its application to SLM-based PAPR reduction, IEEE
Journal of Selected Topics in Signal Processing, vol. 4, no. 3, pp. 637-645, June 2010.
[165] Zhongpeng Wang, Reduction PAPR of OFDM Signals by Combining SLM
with DCT Transform, Int. J. Communications, Network and System Sciences, 2010,
3, pp. 888-892 , November 2010 .
[166] A. Ghassemi and T. A. Gulliver, Partial selective mapping OFDM with low
complexity IFFTs, IEEE Communication Letters, vol. 12, no. 1, pp. 4-6, January
2008.
[167] R. J. Baxley and G. T. Zhou ,Comparing selected mapping and partial transmit
sequence for PAR reduction, IEEE Transactions on Broadcasting, vol. 53, no. 4, pp.
797803, December 2007.
[168] Gatherer, A. and Polley, M., Controlling clipping probability in DMT
transmission, Asilomar Conference record, vol. 1, pp. 578 - 584, Nov. 1997.
[169] Tellado, J. and Cioffi, J. M., Efficient algorithms for reducing PAR in
multicarrier systems, Proc. IEEE International Symposium on Information Theory, p.
191, Aug. 1998.
180
References
[170] J. Tellado, Multicarrier Modulation with Low PAR: Applications to DSL and
Wireless, Kluwer Academic Publishers, 2000.
[171] L. Dewangan, M. Singh, and N. Dewangan A Survey of PAPR Reduction
Techniques in LTE-OFDM System, International Journal of Recent Technology and
Engineering (IJRTE), Vol.1, No.5,pp.10-13 November 2012.
[172] Mukul , and S. singh An Overview and Comparative Study of Different PAPR
Reduction Techniques in OFDM System, National Conference on Innovative Trends
in Computer Science Engineering ,pp. 118-122, April 2015.
[173] D. Narendra, and P.S. Reddy PAPR Reduction Technique in OFDM System
For 4G Wireless Applications Using Partial Transmit Sequence Method, Quest
Journals, Journal of Electronics and Communication Engineering Research, Vol.1,
No.1 ,pp. 38-42,2013.
[174] M. Petermann, D. Wubben and K. Kammeyer Joint Constellation Extension
And Tone Reservation For PAPR Reduction In Adaptive OFDM Systems, IEEE
10th Workshop on Signal Processing Advances in Wireless Communications, 2009.
SPAWC '09, pp. 439 443, 21-24 June 2009.
[175] Ryu, H., Lee, J., and Park, J., Dummy Sequence insertion (DSI) for PAPR
reduction in the OFDM communication system, IEEE Transactions on Consumer
Electronics, Vol. 50, Feb. 2004.
[176] D. Pramudiwati LTE System Performance In Relation To Widband Channel
Properties . M.A. thesis, Delft University of Technology.
[177] ITU-R, Report M.2134, Requirements related to technical performance for
IMT-Advance Radio Interface,2008.
[178] R. Prasad. OFDM for Wireless Communications Systems. Artech House
Publishers, August 2004.
[179] B. Schmidt, C. Ng, P. Yien, C. Harris, U. Saripalle, A. Price, S. Brewer, G.
Scelsi, R. Slaviero and J. Armstrong' Efficient Algorithms for PAPR Reduction in
OFDM Transmitters Implemented using Fixed-Point DSPs, IEEE 63rd Vehicular
Technology Conference, 2006, vol. 4 ,pp. 2023 2027.
[180] J. Armstrong Peak-to-Average Power Reduction in Digital Television
Transmitters, Digital Image Computing Techniques and Applications, pp.1-6,
January 2002.
[181] V. N. Sonawane, and S. V. Khobragade Reduction of PAPR in OFDM System
using A-law Companding Technique, International Journal of Scientific &
Engineering Research, vol. 4, pp.1450-1453 , May-2013.
[182] Y. S. Reddy1, M V. K. Reddy, K Ayyanna, and G. V. Ravikumar The Effect
of NCT Techniques on SC-FDMA System in presence of HPA International Journal
181
References
of Research in Computer and Communication Technology, vol. 3,pp.844-848, August
2014.
[183] V. N. Sonawane1, and S. V. Khobragade Comparative Analysis between Alaw & -law Companding Technique for PAPR Reduction in OFDM International
Journal of Advanced Research in Computer and Communication Engineering, Vol. 2,
PP. 2210-2214, May 2013.
[184] S. S. Bhadoriya, and R. Bhatia Novel Method For PAPR Reduction In
OFDM Using Cascading Of Companding Technique And Selective Mapping,
International Journal of Engineering Sciences and Management, vol. 2, pp. 167-169,
2012.
[185] A. Vallavaraj, B.G. Stewart, D. K. Harrison, and F. G. McIntosh. Reduction
of peak-to-average power ratio of OFDM system using a companding technique,
IEEE Transactions on Broadcasting, vol. 45, pp. 303-307, Sep 1999.
[186] T. Kozhakhmetov Reduction of Peak-to-Average Power Ratio in OFDM
systems using companding schemes Helmut Schmidt University, 2008.
[187] S. L. Kotgire and S. B. Deosarkar Low Noise Technique for Reduction of
Peak Power and BER in High Rate Wireless Communication System, International
Journal of Current Engineering and Technology, vol. 3,pp.1330-1334, no.4, October
2013.
[188] K. K. Sen, and V. sahu PAPR Reduction Using PTS with DCT SLM
Technique International Journal of Computer Network And Security, vol. 1, pp. 2729, 2014.
[189] M. I. Ishaq., Y. A.Khan, and M. T. Gul. Precoding in MIMO, OFDM to reduce
PAPR (Peak to Average Power Ratio), M.A. thesis, Linnaeus University, Sweden ,
2012.
[190] C. Ciochina, D. Mottier, and H. Sari An Analysis of OFDMA, Precoded
Ofdma and SC-FDMA for the Uplink in Cellular Systems, Lecture Notes
Electrical Engineering ,vol.1, pp. 25-36, 2007.
[191] F. S. akar A Study Of Precoding Schemes For OFDM Systems. M.A. thesis,
Middle East Technical University, August 2008.
[192] M. A. Aboul-Dahab, E.A. A. A. Hagras, and A. A. Elhaseeb PAPR Reduction
Based on DFT Precoding for OFDM Signals, International Journal ofFuture
Computer and Communication, vol. 2, no. 4, pp. 325-328 , August 2013.
[193] A. V. Sivaram and R. S. Rao PAPR Reduction of DHT and WHT-Precoded
OFDM System for M-QAM ITSI Transactions on Electrical and Electronics
Engineering, vol. 1, pp.113-117 , 2013.
182
References
[194] G. D. Prakash and M.Sharmila Analysis of PAPR in Precoded OFDM Systems
for M-QAM International Journal of Engineering Trends and Technology, vol. 4,
pp.2318 - 2323, June 2013.
[195] B. Pavan and Kumar.M Comparitive Study Of PAPR Of DHT-Precoded
OFDM System With DFT & WHT, International Journal of Engineering Research
and Applications, vol. 3, pp.527-532, January-February 2013.
[196] D. A. Nugroho, and D. Kim Precoded DCT and Low Complexity SLM for
PAPR Reduction in OFDM Systems International Journal of Engineering and
Industries, vol. 4, no. 4, pp. 43-49, December 2013.
[197] B. S. Kumari.PAPR and Fiber Nonlinearity Mitigation of DFT-Precoded
OFDM System, International Journal of Advanced Research in Electrical,
Electronics and Instrumentation Engineering, vol. 3, pp. 9801- 9808, June 2014.
[198] S. Chaudhari Peak to Average Power Ratio Reduction Schemes and
Synchronization Algorithms for OFDM Based Systems. M.A. thesis, Indian Institute
of Science, India, July 2004.
183
Appendix A:
Tables of Results
Appendix A
Tables of Results
A.1 RCF Results

Table A.1 RCF Results
Oversampling(I)
1
pilot
1.125
1.25
1.5
CR
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
PAPR
14.0866
11.5702
8.2486
7.3802
6.3838
14.1973
11.9169
9.4432
8.5889
8.0108
14.1010
11.6977
8.4887
7.4828
6.6241
14.1392
11.7044
8.5374
7.7741
6.8767
14.1728
11.8777
8.9992
8.0760
7.8858
14.4932
12.2742
9.6073
9.0189
8.3829
15.3850
13.2843
10.6525
9.8847
9.4795
15.5738
13.4298
10.7538
10.0109
9.2193
A. 1
CCDF of
PAPR
6.0244
4.785
3.1186
2.594
2.9
6.248
5.225
3.8796
3.56
3.3568
6.0337
4.8436
3.2413
2.7737
2.3785
6.06
4.8772
3.43
3.023
2.618
6.145
5
3.6
3.315
2.976
6.4177
5.355
4.1216
3.785
3.6257
6.7374
5.79
4.555
4.2435
3.9
6.8674
5.8
4.4432
4.0712
3.725
SNR
(BER=
11.82
12.3711
14.3756
16.72
29.5
9.36
9.8415
13
16.063
29.6(
11.3068
11.65
14.4
16.875
29.6(
10.572
11.4
13.72
16
29.6(
10
10.4765
13.7445
16.4373
29.6(
8.7679
9.414
12.128
15
29.6(
7.1838
7.665
10.418
13.1725
29.6(
5.8315
6.0725
9.6955
13
29.6(
Appendix A:
Tables of Results
A.2 RFC Results

If (A, B, and C) positive values that's mean there is an improvement,
while if the negative values this mean there is a deterioration in values
The (PAPR, CCDF OF PAPR and BER) were calculated with different
value of (CR (4, 3, 2, 1.75, 1.5) and I (1, pilot, 1.125, 1.25, 1.5, 2, 3, 4)
These values have been placed on the table. A, B, and C also added to the
table for comparison with the RCF
Where A = PAPR (RCF) PAPR (RFC)
B =CCDF of PAPR (RCF) CCDF of PAPR (RFC)
C = SNR at BER (
) (RCF) SNR at BER (
) (RFC)
Table A.2 RFC Results
I
Pilot
1.125
1.25
1.5
CR
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
2
1.75
1.5
4
3
A
0.0232
.0906
.6703
.334
1.1464
0.0042
0.0159
-0.0866
0.1119
0.0771
0.0215
0.0583
0.2114
0.2031
0.2754
-0.0015
0.0533
0.4553
0.6558
1.3184
0.3083
0.5039
1.0271
1.4327
1.7478
1.1714
1.5283
PAPR
14.1741
11.8263
8.7729
8.2549
6.8644
14.0968
11.6818
8.5753
7.3709
6.5470
14.1177
11.6461
8.3260
7.5710
6.6013
14.1743
11.8244
8.5439
7.4202
6.5674
14.1849
11.7703
8.5802
7.5862
6.6351
14.2136
11.7560
B
-0.3570
0.3490
0.5706
0.7329
0.9448
0.0069
0.0436
0.0513
0.0837
0.2372
0.0279
0.0732
0.1750
0.3118
0.3556
0.1018
0.1655
0.3289
0.2343
0.5919
0.3577
0.5010
0.7472
0.9370
1.1757
0.6830
0.9373
A. 2
CCDF of
PAPR
6.605
4.876
3.309
2.8271
2.412
6.0268
4.8
3.19
2.69
2.1413
6.0321
4.804
3.255
2.7112
2.2624
6.0432
4.8345
3.2711
3.0807
2.3841
6.06
4.854
3.3744
2.848
2.45
6.0544
4.8527
-0.2400
-0.1585
0.5000
1.0630
0
0.0708
-0.1700
0.7455
-0.2390
0.9000
-0.5700
0.1900
-0.0375
0.3286
0
-0.1236
0.0510
0.5733
0.1286
0
0.1089
0.0440
-0.1548
0.3875
0
0.4163
0.2842
SNR
(BER=
9.6
10
12.5
15
29.6(
)
11.236
11.82
13.6545
17.114
28.7
11.142
11.21
13.7575
15.6714
29.6(
)
10.1236
10.4255
13.1712
16.3087
29.6(
)
8.659
9.37
12.2828
14.6125
29.6(
6.7675
7.3808
Appendix A:
2
1.75
1.5
4
3
2
1.75
1.5
Tables of Results
2.1453
2.2903
2.9062
1.3900
1.6108
2.2738
2.6883
2.8284
8.5072
7.5944
6.5733
14.1838
11.8190
8.4800
7.3226
6.3909
1.2550
1.4123
1.5600
0.8024
0.9360
1.1382
1.2499
1.3092
3.3
2.8312
2.34
6.065
4.864
3.305
2.8213
2.4158
0.4380
1.4725
0
0.1602
0.0340
1.2685
2.1787
1.3800
9.98
11.7
29.6(
)
5.6713
6.0385
8.427
10.8213
28.22
Table A.3 Precoding Results

Precoding
WHT
DCT
DST
DHT
DFT
Table A.4
5
10
15
20
30
35
40
50
70
80
87.6
90
100
120
Table A.5
PAPR
25.6318
22.8377
18.1110
17.4649
6.9587
0.0200
CCDF of PAPR
10.773
9.9046
7.664
7.523
3.35
0
SNR (BER=
13.6
16.12
17.576
18
19.2
19.2
19.757
20.116
21.13
21.2592
21.372
21.1547
21.62
21.863
Companding Results
PAPR
18.9061
14.6917
14.3265
14.4401
11.8545
12.9100
12.6028
12.5906
12.1503
11.8738
11.7819
11.7517
11.3543
11.4101
CCDF of PAPR
6.64
4.73
4.126
3.685
3.32
3.413
3.26
3.1
2.837
2.7542
2.723
2.739
2.58
2.535
Companding Results
PAPR
5
10
20
30
40
50
60
70
SNR (BER=
11.64
11.63
11.628
11.628
11.582
11.469
CCDF of
PAPR
6.416
5.77
5.078
4.4
4
3.8
3.66
3.646
17.4332
16.5470
17.2573
14.6356
13.8453
13.8622
13.7327
14.2237
A. 3
SNR (BER=
13.3363
14.64
16.25
17.165
17.75
18.27
18.777
19
Appendix A:
80
90
100
120
140
160
180
200
220
240
250
255
260
280
300
320
500
700
1000
Tables of Results
16.5501
12.8584
12.4142
12.1142
12.7379
12.2083
12.5858
12.2423
11.2722
15.9144
12.4358
11.7434
11.9863
11.5891
11.9073
12.3703
11.1951
10.8218
12.8953
3.33
3.3
3.12
3
3.06
2.866
2.88
2.76
2.645
2.85
2.723
2.68
2.666
2.6
2.61
2.7
2.3
2.17
2.28
19.474
19.6
20
20.28
20.5
20.475
21.05
21
21.2385
21.6
21.6
21.468
21.7
21.9
22
22.125
22.92
23.5
23.764
Table A.6 RCT Results

R
.9
.8
.7
.6
.5
.4
.3
.2
.1
PAPR
21.8631
21.1311
18.1291
15.3547
13.9264
11.5292
8.5529
5.9888
2.8726
CCDF of PAPR
9.55
8.6815
8.058
6.6825
5.835
4.8215
3.71
2.5745
1.268
SNR (BER=
11.6765
11.987
12.4137
13.4
14.45
16.145
18.525
22.25
28.3
CCDF of PAPR
5.1533
5.14
4.9185
4.77
4.585
4.358
4.1465
3.98
3.806
3.5995
3.374
3.192
2.9264
2.637
SNR (BER=
)
14.73
14.7
14.858
14.45
14.2
14.3138
14.5685
14.3
15.33
14.85
15.3
16.3
24.833
30 (
)
Table A.7 AEXP Companding Results

AEXP d
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
PAPR
13.0811
13.0240
12.1983
11.2173
10.6962
10.0664
9.6145
8.9815
8.4500
7.9523
7.3774
6.7500
6.0806
5.5253
A. 4
Appendix A:
.6
.5
.4
.3
.2
.1
Tables of Results
4.7892
4.1344
3.4039
2.6518
1.8358
0.9690
2.34
2.1075
1.74
1.3272
.9425
.5116
30 (
30 (
30 (
30 (
30 (
30 (
)
)
)
)
)
)
Table A.8 Cos Companding Results

y
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
PAPR
26.3137
25.5112
24.6392
23.3428
22.8762
21.4160
20.1972
19.4063
18.1770
16.8064
15.6468
14.3909
13.1204
11.7106
10.2401
8.7575
7.2067
5.5700
3.8470
1.9930
CCDF of PAPR
11.355
10.858
10.47
10.129
9.682
9.256
8.875
8.5
7.919
7.48
6.9508
6.3817
5.796
5.2196
4.6249
3.9743
3.3453
2.59
1.8074
.9208
SNR (BER=
15.42
15
14.255
13.75
13.675
12.765
12.25
11.95
12
11.832
11.7031
12.1428
12.6966
13.6465
14.2953
15.7648
17.7538
29.2836
>30
>>30
Table A.9 tanhR Companding Results

k
PAPR
5
5
5
5
10
10
10
10
10
1
.8
.5
.2
1
.9
.8
.7
.6
16.7129
13.8472
8.9043
3.5446
8.9016
8.5570
8.2624
7.7099
7.0350
A. 5
CCDF of
PAPR
7.4076
6.0581
3.8861
1.5275
4.1605
3.9969
3.915
3.5129
3.2427
SNR (BER=
12.1294
12.6712
15.5877
24.7231
14.504
14.8376
15.1789
15.8577
16.8
Appendix A:
Tables of Results
10
10
10
10
10
15
15
15
15
20
20
20
20
.5
.4
.3
.2
.1
1
.8
.5
.2
1
.8
.5
.2
6.5160
5.4230
4.4318
3.0296
1.6412
5.2314
5.2429
4.9419
2.7437
3.3781
3.5950
3.7761
2.5710
2.9176
2.4367
1.9475
1.3133
.6988
2.5987
2.5265
2.2321
1.2027
1.8228
1.8412
1.7895
1.1315
17.8871
20.2622
22.3763
25.7854
>30
29.6
22.4656
21.4747
26.5576
30(
)
30
30
28.9392
Table A.10 tanhR Companding Results at y =1

k
PAPR
CCDF of
PAPR
5
10
15
20
1
1
1
1
16.4627
8.9312
5.2314
3.3781
7.2165
4.209
2.5987
1.8228
SNR at BER
11.9245
14.6486
30
30(
)
Table A.11 tanhR Companding Results at y =0.8

k
PAPR
CCDF of
PAPR
5
10
15
20
.8
.8
.8
.8
13.4361
8.1480
5.2429
3.5950
5.8816
3.7579
2.5265
1.8412
SNR (BER=
12.817
14.9321
22.4656
Table A.12 logR Companding Results

y
1
1
1
.9
.8
.7
.6
.5
.4
.3
.2
1
5
10
10
10
10
10
10
10
10
10
PAPR
23.9381
19.4187
16.7420
15.7785
14.7339
12.3933
11.1271
9.7424
8.1924
5.9023
4.0933
CCDF
OF
PAPR
10.14
8.54
7.3145
6.783
6.3775
5.526
4.87
4.24
3.532
2.625
1.8
SNR
(BER=
11.65
12.07
12.5
12.712
13.2
13.8265
14.7521
15.75
17.27
20
23.65
A. 6
PAPR
)
1
5
20
20
20
20
20
20
20
20
20
23.9381
19.4187
14.4171
13.5847
12.8391
11.3170
10.1279
8.9142
7.4148
5.8346
4.1963
CCDF
OF
PAPR
10.14
8.54
6.291
6.078
5.53
5
4.4434
3.928
3.317
2.526
1.892
SNR
(BER=
11.65
12.07
13.616
13.913
13.838
14.54
15.325
16.4332
18.082
20.237
23.9
Appendix A:
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
10
30
30
30
30
30
30
30
30
30
30
50
50
50
50
50
50
50
50
50
50
70
70
70
70
70
70
70
70
70
70
90
90
90
90
90
90
90
90
90
90
2.1028
13.7103
12.5344
11.5140
10.4388
9.6178
8.4340
7.2406
5.7490
3.7918
2.2351
11.6535
11.1455
10.2693
9.9933
9.0562
7.7808
6.5612
5.3972
3.7525
2.0379
10.3556
10.3901
9.7716
8.9933
8.8132
7.6750
6.3892
5.5562
3.9753
2.2114
9.6991
9.6399
9.0726
8.7749
8.2079
7.0926
6.6478
6.0873
3.6976
2.1337
Tables of Results
.932
5.76
5.465
5.02
4.538
4.18
3.6
3.15
2.5
1.686
.882
5.08
4.803
4.492
4.2685
3.94
3.412
2.8765
2.306
1.668
.908
4.592
4.5
4.3
3.935
3.64
3.366
2.83
2.446
1.705
.9335
4.42
4.282
3.95
3.707
3.488
3.124
2.889
2.24
1.6
.89
29.6
14.7383
14.71
14.7
15.37
16
17.0828
18
20.627
23.78
30
16.58
16.82
16.26
16.478
16.93
17.424
18.893
21
24.4
29.6
20
18.3715
17.725
17.65
17.43
18.128
19.15
21.25
23.88
29.6
30
21.9
20
18.85
19.337
18.9814
19.16
21.5175
24.36
29.6
A. 7
20
40
40
40
40
40
40
40
40
40
40
60
60
60
60
60
60
60
60
60
60
80
80
80
80
80
80
80
80
80
80
100
100
100
100
100
100
100
100
100
100
2.0352
12.4590
11.7370
10.8267
10.1572
9.0205
8.3535
7.3063
5.4107
3.8688
2.1230
11.7483
10.3340
10.2586
9.5086
8.4069
7.8869
6.5492
5.2164
4.0939
2.1340
10.4009
10.2191
9.6145
8.7912
8.0294
7.5870
6.6794
5.1697
3.6822
2.0227
10.0104
9.7016
9.2385
8.2944
7.9701
7.1942
6.2888
5.0748
3.6944
2.0774
.892
5.417
5.2
4.7
4.5
4
3.62
3.05
2.4
1.6785
.885
4.6
4.515
4.483
4.0443
3.685
3.5
2.6864
2.2685
1.6474
.8863
4.6
4.2185
4.288
3.84
3.56
3.3025
3
2.28
1.626
.88
4.225
4.238
4.0666
3.65
3.504
3.2
2.755
2.275
1.63
.9
29.4665
15.6865
15.9
16
15.68
16.4
17.612
18.95
20.778
23.8868
30
18.2856
17.52
17.08
17.36
17.085
18
19.03
21.056
23.888
29.6
24.085
20
19.1
17.638
18.18
17.3734
19.4576
21.188
24.158
29.6
30
24.5
20.46
18.745
19.13
19.45
19.816
21.6
24.4
29.6
Appendix A:
Tables of Results
A.5 Hybrid RCF with companding Results:
X = PAPR (Companding ) PAPR (Companding +RCF)

Y =CCDF of PAPR (Companding) - CCDF of PAPR (AV+RCF)
Z= SNR (BER=
) (Companding) SNR (BER=
) (Companding +RCF)
X1 == PAPR (RCF) PAPR (Companding +RCF)
Y1 =CCDF of PAPR (RCF) - CCDF of PAPR (Companding +RCF)
Z1= SNR (BER=
) (RCF) SNR (BER=
) (Companding +RCF)
Table A.13 (RCF+A) Results and compared with the results of each of (RCF) and (A companding)
A
CR
X1
PAPR
Y1
5
10
20
30
40
50
60
70
80
87.6
90
100
120
140
160
2
2
2
2
2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
10.4351
7.9653
8.7683
9.3538
7.0551
8.0339
8.1579
7.8844
7.7334
7.6864
6.9071
7.3797
7.4983
6.0222
7.7668
8.9350
9.4069
9.6938
9.9243
10.0605
10.2273
10.3528
10.3977
9.6486
10.5186
10.5814
10.6902
10.7995
8.4710
6.7264
5.5582
5.0863
4.7994
4.5689
4.4327
4.2659
4.1404
4.0955
4.8446
3.9746
3.9118
3.8030
3.6937
2.7730
1.6650
1.6433
1.4150
1.1950
1.2070
1.1400
0.9340
0.9127
0.8872
0.9250
0.8150
0.8535
2.5507
3.3527
3.9350
4.1477
4.2927
4.3647
4.4577
4.5147
4.5762
4.5819
4.6037
4.6527
4.7362
4.7527
4.7847
A. 8
CCDF of
PAPR
3.867
3.065
2.4827
2.27
2.125
2.053
1.96
1.903
1.8415
1.8358
1.814
1.765
1.6815
1.665
1.633
Z1
2.5095
3.1882
2.5760
1.8420
2.2000
2.3762
2.4280
2.6545
2.8042
2.7380
2.6547
2.9200
3.0930
-2.3226
-4.1639
-6.2321
-7.3901
-8.2321
-8.6129
-8.9201
-9.7076
-9.6871
-9.8661
-9.7321
-9.9321
-10.0021
-10.3171
-10.8121
SNR
(BER=
)
11.0905
12.9318
15
16.158
17
17.3808
17.688
18.4755
18.455
18.634
18.5
18.7
18.77
19.085
19.58
Appendix A:
180
200
5
10
20
30
40
50
60
70
80
87.6
90
5
10
20
30
40
50
60
70
80
87.6
90
5
2
2
2
4
4
3
3
3
2
2
2
1.5
11.6867
8.8266
9.3274
9.8178
7.5657
8.5241
8.6321
8.2973
8.0553
8.0755
8.0445
13.0333
9.7802
10.0069
10.5073
8.1122
8.9745
9.0741
8.6284
8.5552
8.3071
8.4543
13.4875
10.8244
10.8726
5.0548
6.4091
7.2751
7.6519
7.9854
8.1955
8.3157
8.4212
8.4557
8.5678
8.5670
3.7345
4.6958
5.2877
5.6745
5.8650
5.9790
6.0908
6.0854
6.2887
6.1325
6.3099
2.9643
Tables of Results
3.6688
3.6206
7.2194
5.8651
4.9991
4.6223
4.2888
4.0787
3.9585
3.8530
3.8185
3.7064
3.7072
5.8728
4.9115
4.3196
3.9328
3.7423
3.6283
3.5165
3.5219
3.3186
3.4748
3.2974
5.4186
3.4560
2.1970
2.0140
1.7750
1.5435
1.5333
1.4440
1.2262
1.1762
1.1790
1.1990
4.1750
2.7300
2.5260
2.2375
1.9420
1.9300
1.8065
1.6420
1.5542
1.5230
1.5570
4.64
4.8297
4.8657
2.1710
2.8220
3.2430
3.4450
3.5785
3.6283
3.6990
3.7442
3.7770
3.8110
3.8150
1.6566
2.1216
2.5216
2.6741
2.7436
2.7916
2.8281
2.9266
2.9216
2.9216
2.9396
1.6257
1.588
1.552
3.184
2.533
2.112
1.91
1.7765
1.7267
1.656
1.6108
1.578
1.544
1.54
2.465
2
1.6
1.4475
1.378
1.33
1.2935
1.195
1.2
1.2
1.182
2
1.6000
2.1200
1.4040
0.8800
1.4000
1.2927
1.3397
2.0412
2.0229
1.8220
1.9285
-3.0400
-2.6729
-2.8240
-2.7420
-1.8000
-2.7930
-2.1605
-1.7200
-1.2408
-1.4280
-3.0453
-16.4
Table A.14 (RCF+ ) Results and compared with the results of each of (RCF) and ( companding)
A. 9
-10.9921
-11.1128
-2.5860
-4.5860
-6.7580
-7.7060
-8.3860
-9.0503
-9.3623
-9.6748
-9.8223
-10.1360
-9.8122
-4.5120
-6.6649
-8.2720
-8.6140
-8.8720
-10.4220
-10.1485
-10.7220
-10.3720
-10.6720
-12.0720
0
19.76
19.8807
12
14
16.172
17.12
17.8
18.4643
18.7763
19.0888
19.2363
19.55
19.2262
16.64
18.7929
20.4
20.742
21
22.55
22.2765
22.85
22.5
22.8
24.2
30
Appendix A:
5
10
20
30
40
50
60
70
80
90
100
120
140
160
180
200
220
240
255
5
10
20
30
40
50
60
70
Tables of Results
CR
X1
PAPR
Y1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
4
4
4
4
8.3398
8.7462
10.6280
8.5826
8.1186
8.3905
8.5050
9.1783
11.6283
8.0138
7.5968
7.5493
8.2890
7.8826
8.3477
8.0239
7.1092
11.8374
7.7076
9.6333
9.7698
11.4793
9.3130
8.8229
9.0029
9.0355
9.7156
5.3998
6.6924
7.8639
8.4402
8.7665
9.0215
9.2655
9.4478
9.5714
9.6486
9.6758
9.9283
10.0443
10.1675
10.2551
10.2748
10.3302
10.4162
10.4574
4.4743
5.4970
6.4962
6.9516
7.2518
7.4149
7.5770
7.7661
9.0934
7.8008
6.6293
6.0530
5.7267
5.4717
5.2277
5.0454
4.9218
4.8446
4.8174
4.5649
4.4489
4.3257
4.2381
4.2184
4.1630
4.0770
4.0358
7.7999
6.7772
5.7780
5.3226
5.0224
4.8593
4.6972
4.5081
2.3230
2.2700
2.1110
1.6640
1.4470
1.3690
1.3100
1.3860
1.1230
1.1540
1.0110
0.9738
1.0600
0.9080
1.0160
0.9200
0.8325
1.0750
0.9300
2.9530
2.8265
2.5780
2.1220
1.8440
1.7330
1.6600
1.7260
2.3247
2.9177
3.4507
3.6817
3.8647
3.9867
4.0677
4.1577
4.2107
4.2717
4.3087
4.3915
4.4177
4.4597
4.5537
4.5777
4.6052
4.6427
4.6677
1.8920
2.4115
2.8550
3.0770
3.1990
3.2880
3.3550
3.4350
3
3
3
A. 10
CCDF OF
PAPR
4.093
3.5
2.967
2.736
2.553
2.431
2.35
2.26
2.207
2.146
2.109
2.0262
2
1.958
1.864
1.84
1.8125
1.775
1.75
3.463
2.9435
2.5
2.278
2.156
2.067
2
1.92
Z1
2.7063
2.8310
2.7100
2.9360
2.8084
3.0554
2.8770
2.7650
2.9880
2.7524
2.9370
2.7300
2.8215
2.7580
2.8232
2.7820
2.8935
2.9780
2.7810
1.8508
1.7762
2.1180
2.1650
1.7760
1.8700
1.8585
2.1287
-1.8621
-3.0411
-4.7721
-5.4611
-6.1737
-6.4467
-7.1321
-7.4671
-7.7181
-8.0797
-8.2951
-8.7821
-8.9106
-8.9491
-9.4589
-9.4501
-9.5771
-9.8541
-9.9191
-2.0715
-3.4498
-4.7180
-5.5860
-6.5600
-6.9860
-7.5045
-7.4573
SNR
(BER=
)
10.63
11.809
13.54
14.229
14.9416
15.2146
15.9
16.235
16.486
16.8476
17.063
17.55
17.6785
17.717
18.2268
18.218
18.345
18.622
18.687
11.4855
12.8638
14.132
15
15.974
16.4
16.9185
16.8713
Appendix A:
80
90
100
120
140
160
180
200
220
240
255
5
10
20
30
40
50
60
70
80
90
100
120
140
160
180
200
220
240
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
12.1143
8.4708
8.0750
7.9673
8.6065
8.2211
8.7098
8.4603
7.9536
12.2074
8.0323
10.9878
10.9800
12.1035
10.1735
9.4721
9.5534
9.7161
10.2669
12.6705
9.0121
8.6436
8.5096
9.1446
8.7169
8.9937
8.7313
7.9605
12.5271
7.8384
7.8866
7.9350
8.1273
8.1428
8.2870
8.3982
8.4922
8.9556
8.5672
8.5631
3.1619
4.0403
4.4535
5.1452
5.2341
5.2985
5.5907
5.6505
5.7277
5.7610
5.8367
6.0027
6.0140
6.1159
6.0152
6.0963
6.2956
6.2200
Tables of Results
4.4358
4.3876
4.3392
4.1469
4.1314
3.9872
3.8760
3.7820
3.3186
3.7070
3.7111
6.4454
5.5670
5.1538
4.4621
4.3732
4.3088
4.0166
3.9568
3.8796
3.8463
3.7706
3.6046
3.5933
3.4914
3.5921
3.5110
3.3117
3.3873
1.4350
1.4885
1.3600
1.3040
1.3750
1.2240
1.2734
1.2350
1.0450
1.3300
1.1800
3.7600
3.5200
3.1310
2.6080
2.3440
2.1870
2.1380
2.1940
1.8640
1.8850
1.7370
1.6740
1.7565
1.5840
1.6390
1.5600
1.4520
1.6650
A. 11
3.4600
3.5435
3.5950
3.6590
3.6700
3.7130
3.7484
3.8300
3.7550
3.8350
3.8550
1.4656
1.8716
2.1746
2.3296
2.4656
2.5086
2.5996
2.6696
2.6556
2.7066
2.7386
2.7956
2.8181
2.8396
2.8806
2.9216
2.9286
2.9366
1.895
1.8115
1.76
1.696
1.685
1.642
1.6066
1.525
1.6
1.52
1.5
2.656
2.25
1.947
1.792
1.656
1.613
1.522
1.452
1.466
1.415
1.383
1.326
1.3035
1.282
1.241
1.2
1.193
1.185
2.2240
2.1840
2.2374
2.2800
2.0000
1.8250
2.1320
1.6372
2.0058
2.2420
1.4680
-1.6637
-2.6100
-2.6250
-1.6850
-1.8810
-1.6875
-1.2230
-2.1380
-2.2047
-1.7000
-1.7400
-1.7200
-1.9200
-2.1100
-1.1377
-1.2575
-1.7615
-0.9850
-7.8360
-8.0020
-8.3486
- 8.5860
-9.0860
-9.2360
-9.5040
-9.9488
-9.8187
-9.9440
-10.5860
-2.8720
-5.1220
-6.7470
-6.7220
-7.5030
-7.8295
-7.8720
-9.0100
-9.5507
-9.1720
-9.6120
-9.8720
-10.2920
-10.4570
-10.0597
-10.1295
-10.8720
-10.4570
17.25
17.416
17.7626
18
18.5
18.65
18.918
19.3628
19.2327
19.358
20
15
17.25
18.875
18.85
19.631
19.9575
20
21.138
21.6787
21.3
21.74
22
22.42
22.585
22.1877
22.2575
23
22.585
Appendix A:
255
5
2
2
8.3049
11.8784
1.5
6.1688
2.8281
Tables of Results
3.4385
5.5548
1.5205
4.1715
2.9621
1.4805
1.1595
2.2445
-1.4387
-16.6637
-10.7787
0
22.9067
>30
Table A.15 (RCF+ RCT) Results and compared with the results of each of (RCF) and (RCT)
R
.9
.8
.7
.6
.5
.4
.3
.2
.1
.9
.8
.7
.6
.5
.4
.3
.2
.1
.9
.8
.7
I
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
CR
X1
PAPR
Y1
8.4451
8.8719
7.0561
5.5844
5.5235
4.5640
3.1480
2.2574
0.9227
10.5065
10.7468
8.7402
7.0046
6.7408
5.5512
3.9120
2.7595
1.1778
12.9726
12.9771
10.6546
1.0752
2.2340
3.4202
4.7229
6.0903
7.5280
9.0883
10.7618
12.5433
0.9176
1.8899
2.8853
3.9241
5.0886
6.2962
7.6333
9.0449
10.5794
0.7168
1.4533
2.1328
13.4180
12.2592
11.0730
9.7703
8.4029
6.9652
5.4049
3.7314
1.9499
11.3566
10.3843
9.3889
8.3501
7.1856
5.9780
4.6409
3.2293
1.6948
8.8905
8.1540
7.4745
3.4485
3.0435
3.0450
2.2125
2.0020
1.6643
1.2494
0.8589
0.3630
4.4625
3.8165
3.8460
2.9393
2.5280
2.0849
1.5900
1.0745
0.4965
5.5790
5.0235
4.8455
0.3162
0.7797
1.4047
1.9477
2.5847
3.2605
3.9571
4.7021
5.5127
0.2675
0.4900
1.1430
1.6118
2.0480
2.6184
3.2350
3.8550
4.5835
0.1506
0.4636
0.9091
3
3
2
2
2
A. 12
CCDF OF
PAPR
6.1015
5.638
5.013
4.47
3.833
3.1572
2.4606
1.7156
0.905
5.0875
4.865
4.212
3.7432
3.307
2.7366
2.12
1.5
.7715
3.971
3.658
3.2125
Z1
2.7795
2.9000
3.0902
2.8434
2.7734
2.8350
2.7515
3.0366
2.3826
2.0289
2.2570
2.0987
2.4765
2.2000
2.3021
2.0918
2.2500
1.8866
-0.7535
-1.1263
-0.9963
-0.1291
-0.3191
-0.5556
-1.7887
-2.9087
-4.5421
-7.0056
-10.4455
-17.1495
-0.2336
-0.3160
-0.9010
-1.5095
-2.8360
-4.4289
-7.0192
-10.5860
-16.9994
-0.3020
-0.9853
-1.2820
SNR
(BER=
)
8.897
9.087
9.3235
10.5566
11.6766
13.31
15.7735
19.2134
25.9174
9.6476
9.73
10.315
10.9235
12.25
13.8429
16.4332
20
26.4134
12.43
13.1133
13.41
Appendix A:
.6
.5
.4
.3
.2
.1
.9
2
2
2
2
2
2
2
2
1.5
8.7222
8.2882
6.8280
4.8319
3.4152
1.5161
13.9836
2.9748
3.9691
4.9061
5.8863
7.0337
8.2508
.5034
Tables of Results
6.6325
5.6382
4.7012
3.7210
2.5736
1.3565
7.8795
3.7330
3.3380
2.7185
2.0413
1.4145
0.6680
6.201
1.1721
1.6246
2.0186
2.4529
2.9616
3.5216
.2767
2.9495
2.497
2.103
1.6687
1.16
0.6
3.349
-1.3655
-0.7285
-1.2550
-1.4750
-0.9680
-1.0090
-183235
-2.6375
-3.0505
-5.2720
-7.8720
-11.0900
-17.1810
0
14.7655
15.1785
17.4
20
23.218
29.309
> 30
Table A.16 (RCF+AEXP) Results and compared with the results of each of (RCF) and (AEXP companding)
d
CR
X1
PAPR
Y1
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
4.8
.7
.6
.5
.4
4
4
4
4
4
1.6576
2.0456
1.6168
0.9849
0.8895
0.8895
0.7767
0.6955
0.5710
0.5453
0.5075
0.4226
0.3368
0.3553
0.2150
0.2011
0.1356
3.0697
3.5148
3.9117
4.2608
4.6865
5.2963
5.6554
6.2072
6.6142
7.0862
7.6233
8.1658
8.7494
9.3232
9.9190
10.5599
11.2249
11.4235
10.9784
10.5815
10.2324
9.8067
9.1969
8.8378
8.2860
7.8790
7.4070
6.8699
6.3274
5.7438
5.1700
4.5742
3.9333
3.2683
0.3643
0.4800
0.4185
0.4080
0.3930
0.3410
0.2839
0.2534
0.2895
0.2820
0.2475
0.2760
0.2575
0.2310
0.1875
0.2485
0.1400
1.6287
1.7577
1.9177
2.0557
2.2257
2.4007
2.5551
2.6911
2.9012
3.1002
3.2912
3.5017
3.7552
4.0117
4.2652
4.5587
4.8177
A. 13
CCDF OF
PAPR
4.789
4.66
4.5
4.362
4.192
4..017
3.8626
3.7266
3.5165
3.3175
3.1265
2.916
2.6625
2.406
2.1525
1.859
1.6
Z1
4.2300
3.9830
4.0281
3.7540
3.5558
3.4196
3.9685
3.9000
4.6170
4.0570
4.1153
4.0250
6.5750
15.5893
13.8000
0.7900
0
-1.7321
-1.9491
-2.0620
-1.9281
-1.8763
-2.1263
-1.8321
-1.6321
-1.9451
-2.0251
-2.4168
-3.5071
-3.9021
-5.6428
-7.4321
-20.4421
-21.2321
SNR
(BER=
)
10.5
10.717
10.8299
10.696
10.6442
10.8942
10.6
10.4
10.713
10.793
11.1847
12.275
12.67
14.4107
16.2
29.21
> 30
Appendix A:
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
3
3
3
2
2
2
2.6002
2.9078
2.3311
1.9316
-1.6855
1.5155
1.3912
1.2246
1.1408
1.0906
0.9998
0.8470
0.6953
0.6599
0.4746
0.4315
0.3165
3.7697
4.0878
3.2999
3.1314
2.9557
2.7710
2.6346
2.3627
2.2259
2.0747
1.8764
1.7267
1.7933
2.1580
2.4070
2.9885
3.2635
3.7233
4.0509
4.5173
4.9650
5.4125
5.8966
6.3712
6.8889
7.4088
7.9596
8.5713
9.1868
0.2959
0.6711
0.7089
1.5214
1.8668
2.3119
2.6274
2.9885
3.3832
3.7297
4.1063
4.5840
10.4809
10.1162
9.8672
9.2857
9.0107
8.5509
8.2233
7.7569
7.3092
6.8617
6.3776
5.9030
5.3853
4.8654
4.3146
3.7029
3.0874
9.3114
8.9362
8.8984
8.0859
7.7405
7.2954
6.9799
6.6188
6.2241
5.8776
5.5010
5.0233
Tables of Results
0.6733
0.8370
0.6955
0.7300
0.6150
0.5820
0.5390
0.5370
0.5380
0.4970
0.4495
0.4920
0.4200
0.3505
0.3400
0.3125
0.2740
1.3368
1.4775
1.3285
1.3250
1.2940
1.1373
1.0605
1.0460
1.0160
0.9575
0.8755
0.8580
0.8750
1.0520
1.1320
1.3150
1.3850
1.5790
1.7475
1.9120
2.0870
2.2525
2.4305
2.6550
2.8550
3.0685
3.3550
3.5600
3.8890
0.3051
0.4591
0.5316
0.6766
0.8306
0.9009
1.0356
1.1876
1.3316
1.4796
1.6231
1.7876
A. 14
4.48
4.303
4.223
4.04
3.97
3.776
3.6075
3.443
3.268
3.1025
2.9245
2.7
2.5
2.2865
2
1.795
1.466
3.8165
3.6625
3.59
3.445
3.291
3.2207
3.086
2.934
2.79
2.642
2.4985
2.334
3.2800
3.2458
3.6960
2.9420
2.4616
2.9138
3.0907
2.7642
3.5735
2.6963
3.1313
3.3500
5.4050
15.2617
9.1820
0
0
-1.0610
-3.5845
-4.1178
-4.8455
-4.3000
-5.0002
-3.6315
-4.1380
-5.6200
-7.6500
-6.3838
-4.0670
-2.0360
-2.0402
-1.7480
-2.0940
-2.3244
-1.9860
-2.0638
-2.1218
-2.3425
-2.7397
-2.7547
-3.5360
-4.4260
-5.3243
-11.4040
-20.5860
-20.5860
-3.6630
-6.1565
-6.8478
-7.1675
-6.3720
-7.1860
-6.0720
-6.3100
-8.8220
-9.5558
-10.3720
-8.2390
11.45
11.4542
11.162
11.508
11.7384
11.4
11.4778
11.5358
11.7565
12.1537
12.1687
12.95
13.84
14.7383
20.818
> 30
> 30
15.791
18.2845
18.9758
19.2955
18.5
19.314
18.2
18.438
20.95
22.5
21.6838
20.367
Appendix A:
.8
.7
.6
.5
.4
2
1.5
1.4732
1.3760
1.0899
0.9386
0.7417
4.7537
4.9999
5.4580
5.9080
6.4115
6.9451
.0555
Tables of Results
4.6074
4.1493
3.6993
3.1958
2.6622
8.3274
0.7860
0.6705
0.5820
0.6000
0.4680
1.7799
1.9876
2.1551
2.3636
2.6141
2.8496
.4116
2.134
1.9665
1.758
1.5075
1.272
3.3734
-2.5675
-0.9000
0
0
0
-15.27
-9.6845
-16.9720
-17.8720
-17.8720
-17.8720
0
21.8125
29.1
>30
>30
>30
>30
Table A.17 (RCF+ cos) Results and compared with the results of each of (RCF) and (cos companding)
y
CR
X1
PAPR
Y1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
3.0167
2.7783
2.4748
2.0799
1.7684
1.4604
1.1521
0.8658
4.7448
4.3655
3.9220
3.2140
2.7977
2.2782
1.8853
1.4064
12.5439
11.6223
10.6347
9.5893
8.5374
7.3086
6.0541
4.7101
3.2819
1.7016
10.8158
10.0351
9.1875
8.4552
7.5081
6.4908
5.3209
4.1695
1.3219
1.2531
1.1279
0.9978
0.7710
0.6732
0.5450
0.4597
1.9493
2.8709
3.8585
4.9039
5.9558
7.1846
8.4391
9.7831
11.2113
12.7916
1.4584
2.2391
3.0867
3.8190
4.7661
5.7834
6.9533
8.1047
0.8896
1.2548
1.6806
2.1340
2.5857
3.1359
3.6837
4.2774
4.9241
5.6361
0.5467
0.9077
1.2406
1.5801
2.0047
2.4661
2.9620
3.4718
2.0417
1.9687
1.7506
1.5066
1.2527
1.0661
0.8860
0.7168
A. 15
CCDF OF
PAPR
5.5281
5.1629
4.7371
4.2837
3.832
3.2818
2.734
2.1403
1.4936
.7816
4.8083
4.4473
4.1144
3.7749
3.3503
2.8889
2.393
1.8832
Z1
2.7245
2.9981
2.8724
3.2051
3.3736
3.2650
3.3207
4.4287
-0.3076
-0.4590
-0.9397
-1.4110
-2.3970
-3.8271
-5.4714
-8.1034
> -21.2321
> -21.2321
-0.3250
-0.4264
-1.0708
-1.4924
-2.7060
-3.8565
-5.9436
-7.7338
2.0610
2.3846
2.0952
2.4776
2.4185
2.5895
2.2024
4.1522
SNR
(BER=
)
9.0755
9.2269
9.7076
10.1789
11.1649
12.595
14.2393
16.8713
29.6
30
9.739
9.8404
10.4848
10.9064
12.12
13.2705
15.3576
17.1478
Appendix A:
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
7.1336
6.4694
5.7883
5.0089
4.4710
3.6352
2.9108
2.2328
.1
1
1.5
8.2456
Tables of Results
9.3722
10.7498
1.1803
1.6761
2.2861
2.9470
3.7725
4.4735
5.3119
6.2642
7.3113
2.9020
1.5244
8.4270
7.9312
7.3212
6.6603
5.8348
5.1338
4.2954
3.3431
2.2960
8.3921
1.2152
1.0679
7.3150
3.0466
2.8082
2.5530
2.2845
1.9571
1.6068
1.3159
1.0410
3.4937
4.0344
4.6579
0.3182
0.5138
0.8096
1.1246
1.4757
1.7734
2.1585
2.5626
3.0777
1.3206
.6971
3.8034
3.6078
3.312
2.997
2.6459
2.3482
1.9631
1.559
1.0439
3.5707
.5509
0.2694
3.3563
-17.6437
> -20.5860
-0.4563
-0.7894
-1.3324
-2.2901
-2.7810
-6.9048
-4.6413
-17.4720
> -17.8720
-0.7843
-0.6924
-0.8804
-1.0341
-0.3705
-3.1728
-0.7907
-8.3000
> -17.8720
>-18.2000
27.0577
29.6
12.5843
12.9174
13.4604
14.4181
14.909
19.0328
16.7693
29.6
29.6
29.6
29.6
Table A.18 (RCF+NERF) Results and compared with the results of each of (RCF) and (NERF companding)
NERF
CR
X1
PAPR
Y1
4
3
2
1.5
0.9191
1.6395
2.7940
3.0791
5.2328
3.7342
2.2218
1.2825
9.2604
8.5400
7.3855
7.1004
0.4355
0.8685
1.5235
1.8955
2.4177
1.7880
1.2096
1.0857
CCDF OF
PAPR
4
3.567
2.912
2.54
Z1
3.2135
2.7780
-2.3220
-16.3220
-1.6966
-1.4860
-3.8720
-0.4000
SNR
(BER=
)
10.4645
10.9
16
> 30
Table A.19 (RCF+tanhR) Results and compared with the results of each of (RCF) and (tanhR companding)
k
CR
X1
PAPR
A. 16
Y1
CCDF
Z1
SNR
Appendix A:
5
5
5
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
5
5
5
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
4
4
4
4
2.8874
2.4646
1.6787
0.8605
-3.0836
-1.7229
0.1334
0.6159
-4.7170
-3.0465
-0.5541
0.4601
-4.5883
-3.3107
-1.1082
0.4112
3
3
3
4.9487
4.0556
2.6254
0.6677
3.1106
7.2676
11.8091
2.4784
4.6223
8.2269
12.0589
4.5448
6.2038
8.9972
12.2096
6.5268
7.5875
9.6089
12.3334
9.4116
9.6678
10.5681
12.5057
11.1190
11.0445
11.2677
12.6433
0.5100
2.4826
5.9953
Tables of Results
13.8255
11.3826
7.2256
2.6841
12.0148
9.8709
6.2663
2.4343
9.9484
8.2894
5.4960
2.2836
7.9664
6.9057
4.8843
2.1598
5.0816
4.8254
3.9251
1.9875
3.3742
3.4487
3.2255
1.8499
11.7642
9.7916
6.2789
1.1576
0.8777
0.5861
0.2635
-1.3540
-0.7381
-0.1059
0.1921
-1.9818
-1.3095
-0.3264
0.1342
-2.0457
-1.4458
-0.5220
0.1085
2.0796
1.6441
0.9621
A. 17
0.1677
1.2373
3.1177
5.1537
0.8547
1.9217
3.4837
5.2647
1.8372
2.5817
3.8592
5.3492
2.5492
3.1307
4.1062
5.3947
3.8792
4.0477
4.5127
5.4807
4.5622
4.6592
4.8527
5.5377
0.0270
0.9410
2.4310
OF
PAPR
6.25
5.1804
3.3
1.264
5.563
4.496
2.934
1.153
4.5805
3.836
2.5585
1.0685
3.8685
3.287
2.3115
1.023
2.5385
2.37
1.905
.937
1.8555
1.7585
1.565
.88
5.328
4.414
2.924
3.1158
3.3330
3.5127
4.1231
5.6486
5.2944
4.9151
5.0159
19.8030
12.2656
8.1255
4.8634
19.0682
18.3213
15.0955
7.1422
2.5768
2.8712
2.6994
- 0.2457
-0.5703
-3.3071
-11.8321
-0.2321
-0.8698
-3.9459
-12.2321
-1.0291
-1.4321
-4.5813
-12.9263
-1.7639
-2.5108
-5.7366
-13.0291
-4.7878
-4.8484
-6.9271
-14.4491
-19.5481
-10.3821
-9.1149
-14.5471
-0.1386
-0.3860
-3.4743
(BER=
)
9.0136
9.3382
12.075
20.6
9
9.6377
12.7138
21
9.797
10.2
13.3492
21.6942
10.5318
11.2787
14.5045
21.797
13.5557
13.6163
15.695
23.217
28.316
19.15
17.8828
23.315
9.5526
9.8
12.8883
Appendix A:
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
50
5
5
5
5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
2
2
3
3
2
2
2
1.1979
-1.6137
-0.5458
0.8648
0.9020
-3.7961
-2.2420
0.0189
0.7330
-4.1116
-2.8070
-0.6404
0.6658
7.3271
5.9194
3.7924
1.6223
9.9275
1.7293
3.5804
6.7393
10.1260
3.2467
4.7893
7.3512
10.2635
4.7845
5.8722
7.8577
10.3690
7.2887
7.6359
8.6651
10.5174
8.8946
8.8910
9.2710
10.6357
9.8817
9.7344
9.7324
9.8883
0.2215
1.6795
4.4954
7.6850
Tables of Results
2.3467
10.5449
8.6938
5.5349
2.1482
9.0275
7.4849
4.9230
2.0107
7.4897
6.4020
4.4165
1.9052
4.9855
4.6383
3.6091
1.7568
3.3796
3.3832
3.0032
1.6385
2.3925
2.5398
2.5418
2.3859
9.3858
7.9278
5.1119
1.9223
0.4465
-0.6095
-0.1761
0.2921
0.3341
-1.5233
-0.9965
-0.1039
0.2677
-1.6947
-1.1964
-0.2755
0.2270
3.2596
3.0001
1.6301
0.6385
A. 18
4.2740
0.5365
1.4210
2.8190
4.3440
1.2330
1.8320
3.0190
4.4200
1.8375
2.3174
3.2900
4.4505
2.8900
3.0680
3.6140
4.5395
3.5970
3.6190
3.9270
4.5790
4.0150
4.0445
4.0990
4.0850
-0.0264
1.0636
1.8656
3.2326
1.081
4.8185
3.934
2.536
1.011
4.122
3.523
2.336
.935
3.5175
3.0376
2.065
.9045
2.465
2.287
1.741
.8155
1.758
1.736
1.428
.776
1.34
1.3105
1.256
1.27
4.148
3.058
2.256
.889
3.0276
4.6886
3.8892
3.9912
3.4909
19.2188
11.2571
7.4021
4.5003
18.1168
17.4851
14.6743
6.0819
-0.4980
-0.1100
-0.8703
-0.9024
-12.2815
-0.5460
-1.6289
-4.2237
-13.1110
-0.9672
-1.7945
-4.6586
-12.6433
-2.0692
-2.7009
-5.5117
-13.4433
-5.1985
-5.6554
-7.1240
-14.2860
-20.5860
-13.3568
-10.2676
-15.1671
>-20.5860
>-20.5860
-15.2120
>-20.5860
-0.4994
-0.6532
-4.3300
-13.4975
21.6955
9.96
11.0429
13.6377
22.525
10.3812
11.2085
14.0726
22.0573
11.4832
12.1149
14.9257
22.8573
14.6125
15.0694
16.538
23.7
>30
22.7708
19.6816
24.5811
>30
>30
24.626
>30
12.6274
12.7812
16.458
25.6255
Appendix A:
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
5
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
0.1714
0.8666
1.7957
1.2846
-2.6181
-1.1756
0.7929
1.0662
-3.4074
-2.0768
0.0043
0.9859
1.5
8.0365
0.8475
2.3259
5.0033
7.8417
1.7578
3.1888
5.4583
7.9298
2.8218
3.9355
5.8355
8.0222
4.6925
5.2686
6.4575
8.1430
6.1219
6.2434
6.8934
8.2333
0.0685
Tables of Results
8.7598
7.2814
4.6040
1.7656
7.8495
6.4185
4.1490
1.6775
6.7855
5.6718
3.7718
1.5851
4.9148
4.3387
3.1498
1.4643
3.4854
3.3639
2.7139
1.3740
8.3144
0.3585
0.4979
0.7281
0.5421
-0.9813
-0.3955
0.3171
0.4467
-1.2977
-0.7468
-0.0085
0.3955
0.2711
0.8616
2.0216
3.3186
0.5416
1.1996
2.2066
3.3656
1.0011
1.5336
2.3236
3.3856
1.7481
1.9186
2.6191
3.4246
2.2986
2.4566
2.8286
3.4806
0.0729
3.6052
3.8505
3.26
2.1
.803
3.58
2.922
1.915
0.756
3.1205
2.588
1.798
.736
2.3735
2.203
1.5025
.697
1.823
1.665
1.293
.641
3.5528
2.0108
1.2845
-0.7796
-2.4641
15.0296
7.5001
1.1217
-2.1924
13.0242
12.6110
6.3155
-0.6608
>-18.0900
-0.5098
-1.5196
-6.2805
-16.3520
-2.4424
-2.8375
-8.2250
-16.6220
-4.4478
-4.8610
-11.1565
-17.4720
-17.8720
-14.4193
-17.3187
>-17.8720
>-17.8720
>-17.8720
>-17.8720
>-17.8720
0
12.6378
13.6476
18.4085
28.48
14.5704
14.9655
20.353
28.75
16.5758
16.989
23.2845
29.6
>30
26.5473
29.4467
>30
>>30
>>30
>30
>30
>30
Table A.20 (RCF+logR) Results and compared with the results of each of (RCF) and (logR companding)
K
CR
5
5
5
1
.8
.5
4
4
4
X1
PAPR
1.2906
3.5653
7.4063
13.2026
10.9279
7.0869
Y1
0.5137
1.4507
3.1367
A. 19
CCDF OF
PAPR
5.904
4.967
3.281
Z1
-0.2189
-0.5014
-3.8893
SNR
(BER=
)
8.9868
9.2693
12.6572
Appendix A:
5
10
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
50
70
70
70
70
90
90
90
90
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4.4581
4.5315
3.0208
1.2544
3.3770
3.5281
2.5891
1.4327
3.4707
2.7496
2.3515
1.0726
2.8036
2.4673
2.4487
1.1876
2.4698
2.2419
2.0257
1.0953
1.8443
2.2212
2.1335
1.3557
1.6931
1.8910
1.7246
1.1122
11.5744
2.2093
4.2908
7.7716
11.6543
3.4531
5.1822
8.1681
11.7296
4.2536
5.7288
8.4107
11.7740
4.8378
6.1338
8.5884
11.8120
5.3095
6.4658
8.7381
11.8360
5.9819
6.9428
8.9517
11.8736
6.4872
7.3116
9.1252
11.9078
Tables of Results
2.9188
12.2839
10.2024
6.7216
2.8389
11.0401
9.3110
6.3251
2.7636
10.2396
8.7644
6.0825
2.7192
9.6554
8.3594
5.9048
2.6812
9.1837
8.0274
5.7551
2.6572
8.5113
7.5504
5.5415
2.6196
8.0060
7.1816
5.3680
2.5854
1.7353
1.6805
1.1055
0.4720
1.1940
1.2650
1.0200
0.5980
1.0538
0.9815
0.7820
0.4210
1.0170
0.8890
0.8983
0.4235
0.8367
0.7788
0.7408
0.4230
0.7005
0.8415
0.8038
0.4858
0.7215
0.6245
0.6245
0.3890
A. 20
5.0617
0.8385
1.7207
3.2832
5.0897
1.3207
2.1527
3.5097
5.1237
1.7115
2.3792
3.5997
5.1527
2.0177
2.6067
3.6960
5.1627
2.1744
2.7045
3.7465
5.1727
2.5262
2.9592
3.8555
5.1985
2.7192
3.0922
3.9182
5.2067
1.356
5.5792
4.697
3.1345
1.328
5.097
4.265
2.908
1.294
4.7062
4.0385
2.818
1.265
4.4
3.811
2.7217
1.255
4.2433
3.7132
2.6712
1.245
3.8915
3.4585
2.5622
1.2192
3.6985
3.3255
2.4995
1.211
3.3237
3.5893
2.9930
3.2073
3.9373
3.8130
3.5467
3.3966
4.9448
4.3135
3.6750
2.9800
5.4633
5.4110
4.1120
3.2819
6.1574
5.0437
3.7357
3.5624
9.0967
6.4961
4.5018
3.0288
18.2150
7.8815
5.0814
3.7194
-11.8750
-0.4084
-0.8428
-3.9891
-11.6748
-0.9108
-1.2571
-4.1186
-11.7355
-1.0256
-1.6186
-4.6399
-12.0321
-1.4553
-1.8211
-4.7321
-11.8370
-1.6547
-2.4484
-4.9204
-12.0697
-2.1354
-2.4610
-4.8583
-12.0833
-3.0171
-3.3506
-5.1321
-11.8727
20.6429
9.1763
9.6107
12.757
20.4427
9.6787
10.025
12.8865
20.5034
9.7935
10.3865
13.4078
20.8
10.2232
10.589
13.5
20.6049
10.4226
11.2163
13.6883
20.8376
10.9033
11.2289
13.6262
20.8512
11.785
12.1185
13.9
20.6406
Appendix A:
5
5
5
5
10
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
50
70
70
70
70
90
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
6.2371
5.9788
3.9290
1.6290
4.8603
4.7576
3.4143
1.7983
4.8063
3.9069
3.1544
1.4254
4.0178
3.5545
3.2209
1.5353
3.6384
3.2426
2.7517
1.4373
2.9375
3.2059
2.8683
1.6943
2.6972
1.0477
2.9539
6.1589
9.7371
1.7693
3.5191
6.4608
9.8099
2.7174
4.1927
6.7743
9.8762
3.3702
4.6671
6.9946
9.9078
3.8330
5.0020
7.1416
9.9407
4.2591
5.2475
7.2451
9.9590
4.8561
5.7085
7.4675
9.9932
5.2723
Tables of Results
11.2265
9.3203
6.1153
2.5371
10.5049
8.7551
5.8134
2.4643
9.5568
8.0815
5.4999
2.3980
8.9040
7.6071
5.2796
2.3664
8.4412
7.2722
5.1326
2.3335
8.0151
7.0267
5.0291
2.3152
7.4181
6.5657
4.8067
2.2810
7.0019
2.6175
2.4579
1.5920
0.6580
1.9060
1.8177
1.3930
0.7803
1.7293
1.4890
1.1335
0.5853
1.5754
1.4000
1.2790
0.5945
1.4450
1.2997
1.1300
-0.1060
1.1777
1.2800
1.1529
0.6430
1.1900
A. 21
0.2240
1.0718
2.5305
4.1768
0.6580
1.4354
2.7070
4.2130
0.9700
1.6427
2.8200
4.2433
1.3243
1.8240
2.8885
4.2543
1.5134
2.0550
3.0140
4.2710
1.7200
2.1627
3.0730
3.5810
1.9407
2.3350
3.1419
4.2930
2.1250
5.131
4.2832
2.8245
1.1782
4.697
3.9196
2.648
1.142
4.385
3.7123
2.535
1.1117
4.0307
3.531
2.4665
1.1007
3.8416
3.3
2.341
1.084
3.635
3.1923
2.282
1.774
3.4143
3.02
2.2131
1.062
3.23
2.7294
3.0403
2.0783
2.4248
3.1218
2.9968
2.6157
2.6000
4.1696
3.2438
2.9597
2.4858
4.3566
4.4867
3.6694
2.7652
4.5271
3.6600
3.0435
2.9000
7.5555
5.2071
3.7626
2.2988
16.9738
-0.1171
-1.1021
-3.5519
-11.7935
-0.3566
-0.7457
-4.2577
-11.8112
-1.0802
-1.4272
-4.4035
-11.8860
-1.1547
-2.0422
-4.7091
-11.8802
-1.9159
-2.0993
-4.5286
-11.7076
-2.6389
-3.1860
-4.9665
-12.0860
-3.0305
-3.1039
-4.9514
-12.1672
-3.6122
9.5311
10.5161
12.9659
21.2075
9.7706
10.1597
13.6717
21.2252
10.4942
10.8412
13.8175
21.3
10.5687
11.4562
14.1231
21.2942
11.3299
11.5133
13.9426
21.1216
12.0529
12.6
14.3805
21.5
12.4445
12.5179
14.3654
21.5812
13.0262
Appendix A:
90
90
90
5
5
5
5
10
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
50
70
70
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
3
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2.8076
2.4246
1.4463
8.2868
7.6483
5.0356
2.1069
6.6065
6.3082
4.4669
2.2520
6.3973
5.3452
4.2164
1.9076
5.6362
4.8415
4.2284
1.9958
5.0052
4.5085
3.7058
1.8782
4.2424
4.3524
6.0092
7.6062
10.0229
0.6047
2.1294
4.6828
7.5585
1.1521
2.5217
4.9005
7.6209
1.7967
3.0764
5.1600
7.6630
2.2943
3.4385
5.3897
7.7231
2.7845
3.6221
5.4822
7.7343
2.9590
3.8465
5.5323
7.7330
3.4941
4.1881
Tables of Results
6.2650
4.6680
2.2513
9.0026
7.4779
4.9245
2.0488
8.4552
7.0856
4.7068
1.9864
7.8106
6.5309
4.4473
1.9443
7.3130
6.1688
4.2176
1.8842
6.8228
5.9852
4.1251
1.8730
6.6483
5.7608
4.0750
1.8743
6.1132
5.4192
1.0630
0.9726
0.5540
3.4773
3.2775
2.1470
0.9035
2.9073
2.6340
1.9210
1.0097
2.5700
2.2390
1.6813
0.8307
2.3297
2.0360
1.7700
0.8282
2.1132
1.9000
1.5884
0.8302
1.8590
1.8465
A. 22
2.4680
3.2036
4.3090
0.2136
0.7711
1.9060
3.2011
0.2844
1.0216
2.0286
3.2251
0.7379
1.2256
2.1146
3.2393
0.9316
1.3406
2.2029
3.2663
1.0343
1.4576
2.2716
3.2713
1.1548
1.5296
2.2980
3.2838
1.3886
1.6681
2.887
2.1514
1.046
3.908
3.3505
2.2156
0.9205
3.8372
3.1
2.093
.8965
3.3837
2.896
2.007
.8823
3.19
2.781
1.9187
.8553
3.0873
2.664
1.85
.8503
2.9668
2.592
1.8236
.8378
2.733
2.4535
6.3905
4.4034
3.1088
-0.1639
-0.9584
-2.4240
-2.1900
-0.8286
-2.3474
-1.8076
-2.2310
0.5122
-0.7515
-1.0756
-1.9290
-1.4135
-1.9750
-1.3729
-1.6906
-1.6049
-0.9329
-2.2943
-2.1000
-0.4088
-1.8350
-4.1955
-5.1640
-11.8372
-0.6420
-1.7059
-5.2076
-13.3424
-0.5359
-2.0304
-6.0460
-13.7120
-2.3166
-4.0574
-6.1128
-14.0030
-2.0981
-3.3235
-6.0304
-13.5810
-4.9720
-5.8470
-6.8569
-13.4494
-6.0569
-5.0649
-7.5903
-14.3720
-8.2808
-7.4320
13.6095
14.578
21.2512
12.77
13.8339
17.3356
25.4704
12.6639
14.1584
18.174
25.84
14.4446
16.1854
18.2408
26.131
14.2261
15.4515
18.1584
25.709
17.1
17.975
18.9849
25.5774
18.1849
17.1929
19.7183
26.5
20.4088
19.56
Appendix A:
70
70
90
90
90
90
5
.5
.2
1
.8
.5
.2
1
2
2
2
2
2
2
1.5
3.7415
2.1221
3.8737
3.9332
3.2689
1.8688
11.4446
5.6738
7.7541
3.7819
4.4679
5.7836
7.7785
0.4088
Tables of Results
3.9335
1.8532
5.8254
5.1394
3.8237
1.8288
7.9741
1.5870
0.8657
1.7740
1.6365
1.3934
0.7643
5.1875
2.3426
3.2823
1.4756
1.8081
2.3910
3.2859
0.2732
1.779
.8393
2.646
2.3135
1.7306
.8357
3.3525
-2.1860
-3.5928
0.8250
-4.0838
-2.2159
-3.1778
>-17.9300
-8.1860
-15.3448
-17.0470
-11.9558
-9.0693
-15.4098
0
20.314
27.4728
29.175
24.0838
21.1973
27.5378
>30
A.6 Hybrid RFC with companding Results:
X = PAPR (Companding ) PAPR (Companding +RFC)

Y =CCDF of PAPR (Companding) - CCDF of PAPR (Companding + RFC)
Z= SNR (BER=
) (Companding +RFC)
X1 == PAPR (RFC) PAPR (Companding + RFC)
Y1 =CCDF of PAPR (RFC) - CCDF of PAPR (Companding + RFC)
Z1= SNR (BER=
) (RFC) SNR (BER=
) (Companding + RFC)
Table A.21 (RFC+A) Results and compared with the results of each of (RFC) and (A companding)
A
CR
X1
PAPR
Y1
5
10
20
30
40
50
60
70
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
10.6770
8.1290
8.8708
9.4736
7.1833
8.1376
8.2792
7.9605
5.9547
7.6211
8.7281
9.2173
9.5126
9.7186
9.8724
9.9940
8.2291
6.5627
5.4557
4.9665
4.6712
4.4652
4.3114
4.1898
2.9485
1.8080
1.7175
1.5030
1.2752
1.3045
1.2560
1.0095
2.3735
3.1430
3.6565
3.8830
4.0202
4.1095
4.2210
4.2375
A. 23
CCDF OF
PAPR
3.6915
2.922
2.4085
2.182
2.0448
1.9555
1.844
1.8275
Z1
5.5216
6.2537
5.8070
5.2467
5.4416
5.6391
5.5909
6.3199
-2.4071
-4.1950
-6.0977
-7.0820
-8.0871
-8.4466
-8.8538
-9.1388
SNR
(BER=
)
8.0784
9.8663
11.769
12.7533
13.7584
14.1179
14.5251
14.8101
Appendix A:
80
87.6
90
100
120
140
160
180
200
5
10
20
30
40
50
60
70
80
87.6
90
100
120
5
10
20
30
40
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
7.7809
7.7519
7.7326
7.4070
7.5977
11.8201
8.8945
9.4051
9.9401
7.5889
8.5004
8.6499
8.2973
8.1005
8.0606
8.0332
7.5943
7.7448
13.2210
10.0137
9.3891
9.8898
8.2320
10.0909
10.1538
10.1647
10.2365
10.3714
10.4564
10.5338
9.1488
10.6436
4.7330
6.0218
6.8976
7.3190
7.5534
7.7166
7.8783
7.9660
8.0457
8.0977
8.1005
8.0590
8.1537
2.7949
3.8020
3.5426
3.9297
4.8575
Tables of Results
4.0929
4.0300
4.0191
3.9473
3.8124
3.7274
3.6500
5.0350
3.5402
7.0860
5.7972
4.9214
4.5000
4.2656
4.1024
3.9407
3.8530
3.7733
3.7213
3.7185
3.7600
3.6653
5.6851
4.6780
4.9374
4.5503
3.6225
0.9546
0.9755
1.0037
0.8784
0.8928
3.6748
2.3533
2.1496
1.9007
1.6440
1.6596
1.5366
1.3512
1.3053
1.2983
1.3020
1.1672
1.1669
4.5796
3.1030
2.1530
1.8913
2.1484
A. 24
4.2654
4.3175
4.3297
4.3634
4.4228
4.4580
4.4941
4.5362
4.5529
1.8988
2.4873
2.8876
3.0797
3.1880
3.2636
3.3006
3.3782
3.4151
3.4393
3.4270
3.4512
3.4959
1.2446
1.6780
1.3320
1.5113
2.1334
1.7996
1.7475
1.7353
1.7016
1.6422
1.607
1.5709
1.5288
1.5121
2.9652
2.3767
1.9764
1.7843
1.676
1.6004
1.5634
1.4858
1.4489
1.4247
1.437
1.4128
1.3681
2.0604
1.627
1.973
1.7937
1.1716
6.2943
6.2175
5.8577
6.1995
6.0463
4.7945
5.2186
4.5924
4.4688
5.0920
4.5339
4.8399
5.5438
5.2229
5.2748
5.1440
5.1326
5.1104
0.5206
1.2700
4.7008
3.9879
0.8444
-9.2936
-9.4832
-9.6257
-9.7492
-10.1454
-10.5325
-10.8208
-10.8888
-11.0074
-2.7670
-4.8629
-6.9451
-7.4927
-8.0695
-9.1846
-9.2376
-9.5477
-9.9978
-10.0587
-9.9722
-10.4489
-10.7141
-4.6524
-6.4230
-4.4482
-5.5851
-9.9286
14.9649
15.1545
15.297
15.4205
15.8167
16.2038
16.4921
16.5601
16.6787
8.8055
10.9014
12.9836
13.5312
14.108
15.2231
15.2761
15.5862
16. 0363
16.0972
16.0107
16.4874
16.7526
13.0794
14.85
12.8752
14.0121
18.3556
Appendix A:
50
60
70
80
87.6
90
100
120
5
4
4
4
4
4
4
4
4
4
2
2
2
2
2
2
2
2
1.5
9.0584
8.5255
8.1824
8.4834
8.1676
8.0728
7.7330
8.1811
14.3373
4.9356
4.4149
4.5121
5.0896
4.8657
4.8011
4.8587
5.2510
1.8221
Tables of Results
3.5444
4.0651
3.9679
3.3904
3.6143
3.6789
3.6213
3.2290
4.5688
2.1234
1.5636
1.3459
1.7303
1.7146
1.2991
1.1672
1.5670
5.2036
2.1684
1.7686
1.8139
2.2811
2.2966
1.8651
1.8922
2.3370
0.9794
1.1366
1.5364
1.4911
1.0239
1.0084
1.4399
1.4128
.968
1.4364
1.3836
4.9809
5.2696
1.5838
2.0854
5.3091
5.7459
2.3035
-16
-9.9464
-6.7081
-7.4334
-11.2484
-10.8596
-7.4186
-7.4471
-11.1325
-1.3800
18.3734
15.1351
15.8604
19.6754
19.2866
15.8456
15.8741
19.5595
29.6
Z1
5.5832
6.1305
6.2688
6.0706
6.3302
6.4429
6.2734
6.1287
6.2719
6.2989
6.2706
6.2921
6.1176
5.8882
6.2099
-2.0818
-2.8382
-4.3099
-5.4231
-5.7485
-6.1558
-6.8323
-7.2000
-7.5308
-7.6298
-8.0581
-8.3166
-8.7111
-8.9155
-9.1688
SNR
(BER=
)
7.7531
8.5095
9.9812
11.0944
11.4198
11.8271
12.5036
12.8713
13.2021
13.3011
13.7294
13.9879
14.3824
14.5868
14.8401
Table A.22 (RFC+ ) Results and compared with the results of each of (RFC) and ( companding)
MU
CR
X1
PAPR
Y1
5
10
20
30
40
50
60
70
80
90
100
120
140
160
180
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
8.5326
8.9351
10.1373
8.6956
8.2595
8.5378
8.5780
9.2492
11.6899
6.8796
7.7661
6.3834
8.3693
7.9319
8.3988
5.2832
6.5719
7.0638
8.2438
8.5980
8.8594
9.0291
9.2093
9.3236
8.2050
9.5357
8.4530
9.8152
9.9074
9.9968
8.9006
7.6119
7.1200
5.9400
5.5858
5.3244
5.1547
4.9745
4.8602
5.9788
4.6481
5.7308
4.3686
4.2764
4.1870
2.4646
2.3622
2.1758
1.7481
1.5164
1.4583
1.3896
1.4627
1.2103
1.2110
1.0783
1.0264
1.1766
0.9970
1.0647
2.1136
2.6572
3.1628
3.4131
3.5814
3.7233
3.7946
3.8817
3.9453
3.9760
4.0233
4.0914
4.1816
4.1960
4.2497
A. 25
CCDF OF
PAPR
3.9514
3.4078
2.9022
2.6519
2.4836
2.3417
2.2704
2.1833
2.1197
2.089
2.0417
1.9736
1.8834
1.869
1.8153
Appendix A:
200
220
240
255
5
10
20
30
40
50
60
70
80
90
100
120
140
160
180
200
220
240
255
5
10
20
30
40
50
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
2
8.1269
7.2315
11.9179
7.2789
9.7665
9.9560
11.5810
9.3580
8.8961
9.0117
9.0525
9.7253
12.0968
8.5998
8.2039
7.9277
8.7502
8.2815
8.7284
8.4567
7.5390
12.2278
8.0884
11.4573
11.2272
12.5123
10.1818
9.4679
9.6186
10.0684
10.1431
10.1873
9.7193
4.1523
5.2280
6.1427
6.5414
6.8698
6.9685
7.1388
7.3206
7.3657
7.5604
7.6087
7.6325
7.8313
7.8922
7.9616
8.0334
8.0858
8.1324
8.1640
2.5041
3.1602
3.7350
4.0262
4.1026
4.2364
Tables of Results
4.1154
4.0407
3.9965
4.4645
7.6667
6.5910
5.6763
5.2776
4.9492
4.8505
4.6802
4.4984
4.4533
4.2586
4.2103
4.1865
3.9877
3.9268
3.8574
3.7856
3.7332
3.6866
3.6550
5.9759
5.3198
4.7450
4.4538
4.3774
4.2436
0.9662
0.9042
1.1064
0.9756
3.2220
3.0278
2.7424
2.2537
1.9936
1.8652
1.8009
1.8636
1.5756
1.5980
1.4553
1.3889
1.4974
1.3435
1.3896
1.2902
1.1998
1.4267
1.2715
4.2422
3.8944
3.4619
2.9164
2.5916
2.4543
A. 26
4.2712
4.3242
4.3214
4.3606
1.6700
2.1218
2.5284
2.7177
2.8576
2.9292
3.0049
3.0816
3.1096
3.1620
3.1993
3.2529
3.3014
3.3415
3.3736
3.3942
3.4188
3.4407
3.4555
1.1312
1.4294
1.6889
1.8214
1.8966
1.9593
1.7938
1.7408
1.7436
1.7044
3.194
2.7422
2.3356
2.1463
2.0064
1.9348
1.8591
1.7824
1.7544
1.702
1.6647
1.6111
1.5626
1.5225
1.4904
1.4698
1.4452
1.4233
1.4085
2.1738
1.8756
1.6161
1.4836
1.4084
1.3457
5.9880
6.0007
6.2370
6.1604
4.8568
4.9688
5.0238
5.0583
5.1076
5.0674
5.4896
5.2608
5.2740
5.3629
5.8642
5.4133
5.2935
5.3149
5.2071
5.2732
5.2154
5.5058
5.3381
1.6003
1.4241
1.3077
1.2065
1.9988
2.0125
-9.3407
-9.5665
-9.6917
-9.6363
-2.4410
-3.6327
-5.1877
-6.0682
-6.6039
-7.1641
-7.2489
-7.7007
-8.1615
-8.1986
-8.0973
-8.8282
-9.1680
-9.1216
-9.8044
-9.6883
-9.9846
-10.0557
-10.0914
-3.3090
-4.7889
-6.5153
-7.5315
-7.3242
-7.8305
15.012
15.2378
15.363
15.3076
8.4795
9.6712
11.2262
12.1067
12.6424
13.2026
13.2874
13.7392
14.2
14.2371
14.1358
14.8667
15.2065
15.1601
15.8429
15.7268
16.0231
16.0942
16.1299
11.736
13.2159
14.9423
15.9585
15.7512
16.2575
Appendix A:
60
70
80
90
100
120
140
160
180
200
220
240
255
5
4
4
4
4
4
4
4
4
4
4
4
4
4
4
2
2
2
2
2
2
2
2
2
2
2
2
2
1.5
9.5917
10.1649
12.3265
8.6925
8.2980
8.0803
9.2679
8.7930
9.2168
8.9133
7.9705
12.5551
8.3361
12.4989
4.3390
4.4212
4.2564
4.3141
4.3638
4.4461
5.0100
5.0647
5.1110
5.1510
5.1783
5.1207
5.0727
1.4566
Tables of Results
4.1410
4.0588
4.2236
4.1659
4.1162
4.0339
3.4700
3.4153
3.3690
3.3290
3.3017
3.3593
3.4073
4.9343
2.3625
2.3873
2.1150
2.1121
1.9557
1.8746
1.9463
1.7787
1.8151
1.7145
1.6223
1.8317
1.6868
4.8033
2.0075
2.0463
2.0900
2.1171
2.1407
2.1796
2.1913
2.2177
2.2401
2.2595
2.2823
2.2867
2.3118
0.8031
1.2975
1.2587
1.215
1.1879
1.1643
1.1254
1.1137
1.0873
1.0649
1.0455
1.0227
1.0183
.9932
1.6127
2.1959
1.3881
1.8028
1.7990
2.1021
2.2005
2.0155
1.7904
2.1229
1.8744
2.4273
2.4166
2.1590
-16.6637
Table A.23 (RFC+ RCT) Results and compared with the results of each of (RFC) and (RCT).
I
CR
X
X1
PAPR
Y
Y1
CCDF OF
Z
R
.9
.8
.7
.6
.5
.4
.3
.2
.1
.9
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
3
8.7058
8.2277
7.2828
5.7647
5.6633
4.7000
3.2463
2.3157
0.9577
10.8872
1.0265
1.2804
3.3375
4.5938
5.9207
7.3546
8.8772
10.5107
12.2689
0.8431
13.1573
12.9034
10.8463
9.5900
8.2631
6.8292
5.3066
3.6731
1.9149
10.9759
3.7793
3.3454
3.2633
2.4276
2.1581
1.7575
1.3323
0.9217
0.3934
4.9102
A. 27
0.2943
0.7289
1.2703
1.8101
2.3881
3.0010
3.6873
4.4122
5.1904
0.2242
PAPR
5.7707
5.3361
4.7947
4.2549
3.6769
3.064
2.3777
1.6528
.8746
4.6398
5.7109
5.8086
5.9381
6.2980
6.0880
6.0777
5.6848
5.9429
5.5206
5.2180
-8.1541
-9.1849
-9.2442
-9.3740
-9.4709
-9.6525
-10.0575
-10.2576
-10.5001
-10.6986
-10.7564
-10.3842
-10.8820
-1.7800
Z1
-0.2943
-0.5071
-0.8043
-1.4307
-2.6907
-4.3960
-7.1689
-10.6358
-17.1081
-0.4200
16.5811
17.6119
17.6712
17.801
17.8979
18.0795
18.4845
18.6846
18.9271
19.1256
18.8112
19.1834
19.309
>30
SNR
(BER=
)
5.9656
6.1784
6.4756
7.102
8.362
10.0673
12.8402
16.3071
22.7794
6.4585
Appendix A:
.8
.7
.6
.5
.4
.3
.2
.1
.9
.8
.7
.6
.5
.4
.3
.2
.1
.9
3
3
3
3
3
3
3
3
2
2
2
2
2
2
2
2
2
1.5
11.0343
9.0586
7.3092
6.9813
5.7331
4.0685
2.8625
1.2197
13.9172
13.9228
11.5339
9.4644
8.8441
7.2554
5.2124
3.6394
1.6257
15.9263
1.7222
2.7485
3.7735
4.8739
6.0229
7.3346
8.6927
10.1661
0.5341
1.2717
1.8848
2.5897
3.3977
4.2062
5.1395
6.1306
7.2331
0.4541
Tables of Results
10.0968
9.0705
8.0455
6.9451
5.7961
4.4844
3.1263
1.6529
7.9459
7.2083
6.5952
5.8903
5.0823
4.2738
3.3405
2.3494
1.2469
5.9368
4.3707
4.1650
3.2110
2.8005
2.2985
1.7236
1.1770
0.5409
6.4071
5.7633
5.4034
4.2677
3.7262
3.0466
2.2784
1.5677
0.7444
7.3321
0.5532
0.9710
1.3925
1.8295
2.3410
2.8776
3.4665
4.1369
0.1621
0.3868
0.6504
0.8902
1.1962
1.5301
1.8734
2.2982
2.7814
0.1979
4.3108
3.893
3.4715
3.0345
2.523
1.9864
1.3975
.7271
3.1429
2.9182
2.6546
2.4148
2.1088
1.7749
1.4316
1.0068
.5236
2.2179
5.4803
5.2811
5.4153
5.1927
5.1535
5.5226
5.3988
4.8168
2.6339
2.8523
2.9250
3.1766
3.0342
2.3582
2.8951
3.4017
2.6712
-17.0065
-0.4682
-1.0941
-1.9462
-3.2188
-4.9530
-6.9639
-10.8127
-17.4447
-0.6156
-0.7077
-1.0617
-1.7964
-2.9888
-5.3598
-7.2029
-10.4213
-17.2018
-0.4630
6.5067
7.1326
7.9847
9.2573
10.9915
13.0024
16.8512
23.4832
9.0426
9.1347
9.4887
10.2234
11.4158
13.7868
15.6299
18.8483
25.6288
28.683
Table A.24 (RFC+AEXP) Results and compared with the results of each of (RFC) and (AEXP companding)
AEXP
CR
X1
PAPR
Y1
2
1.9
1.8
1.7
1.6
1.5
4
4
4
4
4
4
4
4
4
4
4
4
1.9702
2.2773
1.8198
1.2434
1.1310
0.9692
3.0729
3.4371
3.8053
4.2099
4.6186
5.0866
11.1109
10.7467
10.3785
9.9739
9.5652
9.0972
0.3967
0.5200
0.4753
0.4379
0.4047
0.3714
1.3084
1.4450
1.6218
1.7329
1.8847
2.0784
A. 28
CCDF OF
PAPR
4.7566
4.62
4.4432
4.3321
4.1803
3.9866
Z1
7.5446
6.7516
7.3982
6.4435
6.2825
6.8105
-1.5141
-2.2771
-1.7885
-2.3352
-2.2462
-1.8320
SNR
(BER=
)
7.1854
7.9484
7.4598
8.0065
7.9175
7.5033
Appendix A:
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
0.9623
0.7656
0.7023
0.6813
0.6085
0.4981
0.3743
0.3806
0.2386
0.2162
0.1528
2.8409
3.1584
2.6683
2.0759
1.9290
1.6787
1.6218
1.4161
1.2859
1.2248
1.1195
0.9545
0.7515
0.7338
0.5422
0.4644
0.3578
4.4446
5.5316
5.9679
6.4361
6.9128
7.4149
7.9319
8.4775
9.0391
9.6332
10.2656
10.9327
1.5788
1.9534
2.2890
2.6776
3.0518
3.4313
3.8263
4.2536
4.6549
5.0915
5.5611
6.0235
6.4899
7.0275
7.5720
8.1490
8.7729
-0.1565
Tables of Results
8.6522
8.2159
7.7477
7.2710
6.7689
6.2519
5.7063
5.1447
4.5506
3.9182
3.2511
10.2402
9.8656
9.5300
9.1414
8.7672
8.3877
7.9927
7.5654
7.1641
6.7275
6.2579
5.7955
5.3291
4.7915
4.2470
3.6700
3.0461
8.6365
0.3238
0.3317
0.0659
0.3208
0.2994
0.3327
0.2977
0.2528
0.2178
0.2695
0.2031
0.8554
0.9622
0.8424
0.8486
0.8007
0.6796
0.6111
0.6129
0.5822
0.5470
0.5258
0.5417
0.5069
0.4166
0.3589
0.4074
0.3188
1.8041
2.2423
2.4167
2.3249
2.7863
2.9904
3.2057
3.4427
3.6808
3.9428
4.2270
4.5281
0.5661
0.6862
0.7879
0.9426
1.0797
1.1856
1.3286
1.4969
1.6402
1.8115
2.0158
2.2137
2.4509
2.6436
2.8829
3.1639
3.4428
-0.0442
A. 29
3.8227
3.6483
3.7401
3.2787
3.0746
2.8593
2.6223
2.3842
2.1222
1.838
1.5369
4.2979
4.1778
4.0761
3.9214
3.7843
3.6784
3.5354
3.3671
3.2238
3.0525
2.8482
2.6503
2.4131
2.2204
1.9811
1.7001
1.4212
3.3492
7.1086
7.0868
7.6358
6.8206
7.4259
7.5934
10.0902
19.5598
18.5802
15.5774
0
6.6001
6.4858
6.5050
6.2358
5.9424
5.9209
6.0119
5.7748
6.7358
6.2308
6.3021
7.0545
9.2588
19.1217
17.5183
1.3672
0
4.4356
-1.7886
-1.5419
-2.0229
-2.3581
-2.2028
-3.0353
-3.4835
-4.7689
-5.7485
-8.7513
-24.3287
-2.0914
-2.1757
-2.3145
-2.1757
-2.2191
-2.3544
-2.5181
-2.4867
-2.5557
-2.5807
-2.9594
-3.2070
-3.9477
-4.8398
-6.4432
-22.5943
-23.9615
-1.8674
7.4599
7.2132
7.6942
8.0294
7.8741
8.7066
9.1548
10.4402
11.4198
14.4226
>30
8.1299
8.2142
8.353
8.2142
8.2576
8.3929
8.5566
8.5252
8.5942
8.6192
8.9979
9.2455
9.9862
10.8783
12.4817
28.6328
>30
10.2944
Appendix A:
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1.5
4.7584
4.2805
3.5883
3.4069
3.2267
3.0173
2.7391
2.6453
2.5080
2.2867
2.0616
1.7864
1.6340
1.3458
1.1395
0.9063
5.8486
0.2144
0.5622
0.8510
1.1907
1.6403
1.8828
2.2376
2.6753
3.0357
3.3893
3.7916
4.1858
4.5887
5.0366
5.4851
5.9824
-0.8416
Tables of Results
8.2656
7.9178
7.6290
7.2893
6.8397
6.5972
6.2424
5.8047
5.4443
5.0907
4.6884
4.2942
3.8913
3.4434
2.9949
2.4976
7.2325
1.9134
1.7817
1.6806
1.5898
1.5115
1.3648
1.3171
1.3019
1.2229
1.1397
1.1004
0.9863
0.8742
0.7287
0.7272
0.5759
2.6752
0.0784
0.1682
0.2156
0.3098
0.4585
0.5233
0.6421
0.8009
0.9284
1.0707
1.2134
1.3713
1.5422
1.6937
1.9247
2.1409
-0.0623
3.2266
3.1368
3.0894
2.9952
2.8465
2.7817
2.6629
2.5041
2.3766
2.2343
2.0916
1.9337
1.7628
1.6113
1.3803
1.1641
2.4781
2.6868
2.7087
1.2358
0.5798
0.8877
0.9430
0.6155
1.5764
1.1190
0.2170
2.2184
3.6618
12.9485
1.6400
0
0
-15.2700
-3.5862
-3.7223
-4.7872
-5.1932
-4.9991
-5.1985
-5.2575
-5.3266
-5.3040
-6.6560
-5.6546
-7.1562
-8.6245
-19.9330
-21.5730
-21.5730
-1.7800
12.0132
12.1493
13.2142
13.6202
13.4261
13.6255
13.6845
13.7536
13.731
15.083
14.0816
15.5832
17.0515
28.36
>30
>30
>30
Table A.25 (RFC+ cos) Results and compared with the results of each of (RFC) and (cos companding)
Cos y
CR
X1
PAPR
Y1
1
.9
.8
.7
.6
.5
.4
4
4
4
4
4
4
4
3.3647
2.3785
2.4468
2.3786
2.0945
1.6828
1.2488
1.9879
2.1617
3.5211
4.8932
5.9725
7.0976
8.2264
12.1959
12.0221
10.6627
9.2906
8.2113
7.0862
5.9574
1.4804
1.4216
1.2605
1.1151
0.9266
0.7674
0.6048
0.6954
1.0706
1.4605
1.8986
2.3886
2.8774
3.3908
A. 30
CCDF OF
PAPR
5.3696
4.9944
4.6045
4.1664
3.6764
3.1876
2.6742
Z1
5.9268
6.1069
5.9421
6.1775
6.4601
6.2525
6.2306
-0.2019
-0.4468
-0.9666
-1.5352
-2.4071
-3.9362
-5.6581
SNR
(BER=
)
5.8732
6.1181
6.6379
7.2065
8.0784
9.6075
11.3294
Appendix A:
Tables of Results
.3
.2
.1
4
4
4
1.0157
9.6236
11.0172
12.5275
4.5602
3.1666
1.6563
0.5335
3.9985
4.6068
5.3006
2.0665
1.4582
.7644
7.6706
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
3
3
3
3
3
3
3
3
3
3
5.3429
4.9243
4.4042
3.7962
3.3283
2.7294
2.1840
1.6589
1.6013
2.3427
3.1137
3.9460
4.8415
5.7794
6.7968
7.9020
9.0885
10.3771
10.2177
9.4763
8.7053
7.8730
6.9775
6.0396
5.0222
3.9170
2.7305
1.4419
2.3797
2.2367
2.0116
1.8062
1.4903
1.2001
1.0164
0.8270
0.3937
0.6847
1.0106
1.3887
1.7513
2.1091
2.6014
3.0910
3.6135
4.2062
4.4703
4.1793
3.8534
3.4753
3.1127
2.7549
2.2626
1.773
1.2505
.6578
5.5047
5.4913
5.5914
5.4577
5.9005
5.6383
5.8226
7.1450
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
2
2
2
2
2
2
2
2
2
2
8.4031
7.7215
6.8456
6.0573
5.2889
4.4202
3.5833
2.6995
1.3225
1.8009
2.2161
2.8681
3.4631
4.1312
4.8571
5.6036
6.4762
7.4125
7.1575
6.6791
6.2639
5.6119
5.0169
4.3488
3.6229
2.8764
2.0038
1.0675
3.7063
3.4932
3.1303
2.7832
2.3470
2.0022
1.6358
1.3013
0.1613
0.3822
0.5703
0.8067
1.0490
1.3522
1.6618
2.0063
2.3977
2.8196
3.1437
2.9228
2.7347
2.4983
2.256
1.9528
1.6432
1.2987
.9073
.4854
1.5
10.1993
1.0296
5.3613
4.5714
0.1372
2.2786
A. 31
-7.9581
>-23.9287
>>23.9287
-0.2568
-0.6952
-0.9501
-1.8878
-2.5995
-4.1832
-5.6989
-8.1165
-19.2430
-23.5615
13.6294
>30
>>30
3.3286
3.1239
2.5218
3.0995
2.6977
3.0493
2.9223
4.0627
-0.0444
-0.6741
-1.6312
-1.8575
-3.4138
-4.3837
-6.2107
-8.8103
-21.1730
-21.1730
8.4714
9.1011
10.0582
10.2845
11.8408
12.8107
14.6377
17.2373
29.6
29.6
-17.8000
-1.3800
6.2953
6.7337
6.9886
7.9263
8.638
10.2217
11.7374
14.155
25.2815
29.6
29.6
Appendix A:
Tables of Results
Table A.26 (RFC+NERF) Results and compared with the results of each of (RFC) and (NERF companding)
NERF
CR
X1
PAPR
Y1
4
4
4
4
4
4
3
2
1.75
1.5
0.9191
1.6395
2.7940
3.7575
3.0791
5.2328
3.7342
2.2218
9.0903
8.2522
6.9153
6.4220
5.7098
0.4355
0.8685
1.5235
2.1590
1.8955
2.4177
1.7880
1.2096
1.2825
1.0857
CCDF OF
PAPR
3.8875
3.4578
2.5844
2.2765
1.9541
Z1
3.2135
2.7780
-2.3220
-5.7976
-16.3220
-1.6966
-1.4860
-3.8720
-0.4000
SNR
(BER=
)
7.2096
7.7814
11.883
19.4756
>30
Table A.27 (RFC+tanhR) Results and compared with the results of each of (RFC) and (tanhR companding)
k
CR
X1
PAPR
Y1
5
5
5
5
10
10
10
10
15
15
15
15
20
20
20
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
2.6944
2.1182
1.2755
0.6910
-4.5995
-2.9960
-0.6902
0.4108
-7.5448
-5.2051
-1.6597
0.2390
-8.4594
-6.0915
-2.7713
0.1653
2.4548
6.5550
11.3302
0.6531
3.0398
7.0939
11.5444
1.4076
3.7358
7.5822
11.6791
2.3463
4.4973
7.6364
14.0185
11.7290
7.6288
2.8536
13.5307
11.1440
7.0899
2.6394
12.7762
10.4480
6.6016
2.5047
11.8375
9.6865
6.5474
1.2590
0.8791
0.4358
0.2234
-1.7440
-1.2074
-0.3799
0.1330
-3.0363
-2.1608
-0.7619
0.0530
-3.4260
-2.5352
-1.0317
-0.0836
0.8860
2.6147
4.7609
0.1120
1.0997
2.8570
4.8529
0.4300
1.3777
3.0710
4.9153
0.8162
1.6886
3.2438
4
4
4
4
4
4
4
4
4
A. 32
CCDF
OF
PAPR
6.1486
5.179
3.4503
1.3041
5.953
4.9653
3.208
1.2121
5.635
4.6873
2.994
1.1497
5.2488
4.3764
2.8212
Z1
6.3440
6.2949
6.8333
7.2231
8.8688
8.7788
8.5310
8.5822
23.6963
15.9471
11.7960
8.9374
23.5747
22.8019
19.6576
-0.1141
-0.7050
-3.0831
-11.8287
-0.1085
-0.4820
-3.4266
-11.7624
-0.2324
-0.8472
-4.0074
-11.9489
-0.3540
-1.1268
-4.2711
SNR
(BER=
)
5.7854
6.3763
8.7544
17.5
5.7798
6.1533
9.0979
17.4337
5.9037
6.5185
9.6787
17.6202
6.0253
6.7981
9.9424
Appendix A:
20
30
30
30
30
40
40
40
40
5
5
5
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
4
4
4
4
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
0.1722
5.0723
4.0340
2.4146
1.0832
-2.4226
-1.2989
0.3078
0.7564
-5.6505
-3.7325
-0.7879
0.5623
-6.7832
-4.7701
-1.5784
0.4900
11.7850
4.3979
5.9931
8.7585
11.9430
6.3037
7.3308
9.3689
12.0608
0.1784
2.0058
5.3293
9.3576
0.4652
2.3721
5.7271
9.5252
0.9371
2.8436
6.0892
9.6376
1.6577
3.4539
6.4645
9.7380
3.0272
4.4995
7.0108
9.8546
Tables of Results
2.3988
9.7859
8.1907
5.4253
2.2408
7.8801
6.8530
4.8149
2.1230
11.6406
9.8132
6.4897
2.4614
11.3538
9.4469
6.0919
2.2938
10.8819
8.9754
5.7298
2.1814
10.1613
8.3651
5.3545
2.0810
8.7918
7.3195
4.8082
1.9644
0.0269
2.4801
1.8728
1.0524
0.4255
-0.6289
-0.3228
0.1424
0.3116
-2.0568
-0.3544
-0.2991
0.2234
-2.6056
-1.8404
-0.6063
0.1826
A. 33
4.9604
1.5013
2.2362
3.5268
5.0263
2.2698
2.7876
3.7886
5.0792
-0.0635
0.6787
2.0303
3.7620
0.0261
0.7833
2.1783
3.8305
0.2085
1.9831
2.3328
3.8847
0.4356
1.1824
2.4682
3.9151
0.9312
1.5702
2.6746
3.9733
1.1046
4.5637
3.8288
2.5382
1.0387
3.7952
3.2774
2.2764
.9858
4.9275
4.1853
2.8337
1.102
4.8379
4.0807
2.6857
1.0335
4.6555
2.8809
2.5312
.9793
4.4284
3.6816
2.3958
.9489
3.9328
3.2938
2.1894
.8907
10.7485
5.9113
5.8850
6.4305
6.8634
8.6281
8.4958
7.5664
6.5865
23.2474
15.3959
11.6768
8.0920
22.9245
22.1648
19.1135
9.8503
-12.5194
-1.1109
-1.9585
-4.5051
-12.8835
-1.9663
-2.6475
-5.6472
-13.2070
-0.1796
-0.7477
-3.1187
-11.8212
0.0180
-0.3978
-4.0240
-13.3909
-0.3141
-1.0312
-3.7594
-12.4271
-0.6370
-1.3967
-4.4480
-13.0504
-0.8914
-2.1879
-5.1420
-12.8308
18.1907
6.7822
7.6298
10.1764
18.5548
7.6376
8.3188
11.3185
18.8783
6.2181
6.7862
9.1572
17.8597
6.0205
6.4363
10.0625
19.4294
6.3526
7.0697
9.7979
18.4656
6.6755
7.4352
10.4865
19.0889
6.9299
8.2264
11.1805
18.8693
Appendix A:
40
40
40
40
50
50
50
50
5
5
5
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
3
3
3
3
3
3
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
8.0411
6.4911
3.9618
1.6309
0.6809
1.2029
1.8224
1.2947
-2.8631
-1.5051
0.5603
1.0594
-4.3947
-2.8497
-0.3872
0.9528
4.5177
5.5636
7.5154
9.9583
5.8206
6.4840
7.9373
10.0415
-0.1918
1.1239
3.5375
6.5663
0.2297
1.5349
3.9027
6.7245
0.3855
1.7320
4.0984
6.7957
0.7072
2.0353
4.3167
6.8618
1.4478
2.6476
4.6783
Tables of Results
7.3013
6.2554
4.3036
1.8607
5.9984
5.3350
3.8817
1.7775
8.6718
7.3561
4.9425
1.9137
8.2503
6.9451
4.5773
1.7555
8.0945
6.7480
4.3816
1.6843
7.7728
6.4447
4.1633
1.6182
7.0322
5.8324
3.8017
4.0083
3.1445
1.8462
0.7088
0.9295
0.9545
0.9068
0.5888
-0.6314
-0.2065
0.4052
0.4813
-1.3229
-0.8356
0.0055
0.4239
A. 34
1.5349
2.0098
2.9066
4.0222
2.0640
2.3953
3.0958
4.0640
-0.0943
0.3914
1.2651
2.4863
0.0255
0.5016
1.3837
2.5487
0.0749
0.5720
1.4781
2.5836
0.1593
0.6282
1.5210
2.5974
0.3891
0.8513
1.6656
3.3291
2.8542
1.9574
.8418
2.8
2.4687
1.7682
.8
3.3993
2.9136
2.0399
.8187
3.2795
2.8034
1.9213
.7563
3.2301
2.733
1.8269
.7214
3.1457
2.6768
1.784
.7076
2.9159
2.4537
1.6394
3.4383
3.3481
3.0606
3.0185
6.0106
5.1929
4.9274
4.2034
20.9592
12.6710
8.2439
3.5147
20.4398
19.6499
16.6531
5.9826
-2.1879
-2.9035
-6.1790
-13.4003
-3.7195
-4.0708
-7.0648
-14.0866
-0.2641
-0.8961
-4.1001
-13.2776
-0.2110
-1.3122
-4.2745
-13.3855
-0.2138
-1.3676
-4.8038
-14.6159
-0.7332
-1.5231
-4.5199
-14.5296
-1.0211
-2.8904
-6.2576
8.2264
8.942
12.2175
19.4388
9.758
10.1093
13.1033
20.1251
8.6911
9.3231
12.5271
21.7046
8.638
9.7392
12.7015
21.8125
8.6408
9.7946
13.2308
23.0429
9.1602
9.9501
12.9469
22.9566
9.4481
11.3174
14.6846
Appendix A:
30
40
40
40
40
5
.2
1
.8
.5
.2
1
2
2
2
2
2
1.5
10.4523
Tables of Results
6.9505
2.2898
3.2772
4.9997
7.0196
0.1303
1.5295
6.1902
5.2028
3.4803
1.4604
6.2606
4.9898
2.6311
0.6198
1.0325
1.7771
2.6697
-0.0020
.6739
2.6852
2.2725
1.5279
.6353
2.4178
-16.3460
-16.7147
-2.3539
-3.5287
-7.5339
-16.0484
-0.2554
25.1417
10.7809
11.9557
15.9609
24.4754
28.4754
Table A.28 (RFC+logR) Results and compared with the results of each of (RFC) and (logR companding)
k
CR
5
5
5
5
10
10
10
10
20
20
20
20
30
30
30
30
40
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
X1
PAPR
2.7526
4.0614
2.7840
1.2286
2.3809
2.0822
1.8293
1.2155
2.3488
1.9860
2.0414
1.0410
1.6240
0.7014
3.0106
6.9250
11.2475
0.1944
3.5113
7.2254
11.3191
2.1476
3.4269
7.0989
11.2030
2.8223
4.6558
7.7912
11.4330
3.3488
13.4824
11.1732
7.2588
2.9363
13.9894
10.6725
6.9584
2.8647
12.0362
10.7569
7.0849
2.9808
11.3615
9.5280
6.3926
2.7508
10.8350
Y1
1.5031
1.6195
1.1077
0.4931
0.9625
1.0705
0.9371
0.6147
0.7103
0.7510
0.7017
0.4266
0.5382
0.1402
1.1054
2.7955
4.7255
0.2536
1.3070
2.9327
4.7581
0.7365
1.6055
3.0741
4.7877
1.0153
1.7960
3.1667
4.8056
1.1862
A. 35
CCDF OF
PAPR
5.9248
4.9596
3.2695
1.3395
5.8114
4.758
3.1323
1.3069
5.3285
4.4595
2.9909
1.2773
5.0497
4.269
2.8983
1.2594
4.8788
Z1
6.6311
6.5858
6.4264
6.4737
7.5177
7.1717
6.6767
6.5743
8.5090
7.5945
7.3627
6.3249
8.9930
-0.0429
-0.7108
-3.5314
-11.5564
-0.1976
-0.9429
-3.6523
-11.5050
-0.4270
-0.9950
-4.0852
-11.6544
-0.5580
-1.4342
-4.0488
-11.7838
-1.0222
SNR
(BER=
)
5.7142
6.3821
9.2027
17.2277
5.8689
6.6142
9.3236
17.1763
6.0983
6.6663
9.7565
17.3257
6.2293
7.1055
9.7201
17.4551
6.6935
Appendix A:
40
40
40
50
50
50
50
70
70
70
70
90
90
90
90
5
5
5
5
10
10
10
10
20
20
20
20
30
30
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
1.6777
2.1372
1.1544
1.2444
1.4152
1.6948
1.0611
0.6110
1.3851
1.7977
1.3230
0.4691
1.0355
1.3705
1.0737
5.9143
5.6943
3.8312
1.6255
4.2590
4.4077
3.2711
1.8024
4.0813
3.4342
5.0348
7.9675
11.4694
3.7747
5.3297
8.0978
11.4924
4.4392
5.7973
8.3065
11.5315
4.9538
6.1467
8.4617
11.5599
0.5230
2.4889
5.6872
9.3116
0.9913
2.7794
5.9078
9.3512
1.6609
3.3876
6.1759
9.4251
2.1900
3.7392
Tables of Results
9.1490
6.2163
2.7144
10.4091
8.8541
6.0860
2.6914
9.7446
8.3865
5.8773
2.6523
9.2300
8.0371
5.7221
2.6239
11.2960
9.3301
6.1318
2.5074
10.8277
9.0396
5.9112
2.4678
10.1581
8.4314
5.6431
2.3939
9.6290
8.0798
0.5574
0.7850
0.4373
0.4157
0.5253
0.6536
0.4371
0.2321
0.5107
0.6907
0.4884
0.2755
0.3334
0.5281
0.4024
2.6999
2.4796
1.6309
0.7015
1.9290
1.8493
1.4476
0.8319
1.5873
1.4834
A. 36
1.9224
3.2300
4.8238
1.4007
2.0983
3.3066
4.8341
1.7051
2.2757
3.3897
4.8484
1.9205
2.4484
3.4691
4.8674
0.0727
0.8322
2.1320
3.7351
0.2494
0.9661
2.2549
3.7655
0.5020
1.1833
2.3836
3.8039
0.6913
1.3274
4.1426
2.835
1.2412
4.6643
3.9667
2.7584
1.2309
4.3599
3.7893
2.6753
1.2166
4.1445
3.6166
2.5959
1.1976
4.7913
4.0318
2.732
1.1289
4.6146
3.8979
2.6091
1.0985
4.362
3.6807
2.4804
1.0601
4.1727
3.5366
8.7135
7.8613
6.3458
10.3272
9.0464
6.9655
6.8945
12.8556
10.3224
7.7414
6.1660
22.5168
12.1155
8.8508
6.5459
6.2141
6.3045
5.9720
5.5242
7.1443
6.7307
6.3066
6.2466
7.9106
6.9501
-1.6152
-4.0794
-11.8697
-0.5815
-1.5423
-4.7872
-11.8342
-1.4731
-1.7313
-4.7153
-12.0427
-1.8119
-2.2132
-4.4593
-12.1428
-0.2954
-1.0780
-3.8176
-11.9709
-0.2474
-0.8570
-3.7395
-12.0873
-0.4332
-1.0688
-4.0881
-11.6149
-0.7892
-1.7114
7.2865
9.7507
17.541
6.2528
7.2136
10.4585
17.5055
7.1444
7.4026
10.3866
17.7149
7.4832
7.8845
10.1306
17.8141
6.3339
7.1165
9.8561
18.0094
6.2859
6.8955
9.778
18.1258
6.4717
7.1073
10.1266
17.6534
6.8277
7.7499
Appendix A:
30
30
40
40
40
40
50
50
50
50
70
70
70
70
90
90
90
90
5
5
5
5
10
10
10
10
20
20
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
2
2
2
2
2
2.9839
1.4195
3.1671
3.0396
3.0029
1.5260
2.7298
2.6866
2.5838
1.4481
1.9549
2.6206
2.6296
1.7008
1.7931
2.1595
2.1755
1.4527
8.8239
8.1036
5.3311
2.2409
6.7883
6.4508
6.3689
9.4467
2.5271
4.0319
6.4684
9.4762
2.8953
4.2363
6.6220
9.5146
3.4183
4.6680
6.7736
9.5445
3.9130
4.9059
6.9019
9.5741
0.2534
1.5881
3.8544
6.5880
0.5619
1.8497
4.0687
6.6276
0.8512
2.0917
Tables of Results
5.4501
2.3723
9.2919
7.7871
5.3506
2.3428
8.9237
7.5827
5.1970
2.3044
8.4007
7.1510
5.0454
2.2745
7.9060
6.9131
4.9171
2.2449
8.2266
6.8919
4.6256
1.8920
7.9181
6.6303
4.4113
1.8524
7.6288
6.3883
1.2057
0.6260
1.3860
1.2730
1.2607
0.6341
1.2264
1.1700
1.1024
0.6294
0.9412
1.1170
1.1271
0.6822
0.9073
0.9197
0.9443
0.5923
4.1088
3.6671
2.3911
1.0076
3.2716
2.9631
A. 37
2.4697
3.8040
0.8330
1.4370
2.5047
3.8196
1.0104
1.5420
2.5544
3.8254
1.2132
1.6810
2.6251
3.8412
1.3513
1.8337
2.6843
3.8563
-0.0018
0.5126
1.3946
2.4961
0.0993
0.5946
1.4561
2.5126
0.2856
0.7381
2.3943
1.06
4.031
3.427
2.3593
1.0444
3.8536
3.322
2.3096
1.0386
3.6508
3.183
2.2389
1.0228
3.5127
3.0303
2.1797
1.0077
3.3068
2.7924
1.9104
.8089
3.2057
2.7104
1.8489
.7924
3.0194
2.5669
6.8281
5.7628
8.5830
8.2716
6.8099
5.7427
9.2743
8.1233
6.8099
6.4055
12.6132
9.5620
6.9099
5.7203
22.0765
11.0954
7.7612
6.0773
3.8634
3.3864
2.5828
2.6103
4.5313
3.2202
-4.2162
-11.9787
-1.0650
-1.6899
-4.7636
-12.1056
-1.2672
-2.0982
-4.5756
-11.9560
-1.3483
-2.1245
-5.1796
-12.1212
-1.8850
-2.8661
-5.1817
-12.2442
0.1415
-0.8028
-4.3434
-13.2386
-0.2096
-1.3866
-4.7402
-12.6127
-0.6577
-2.1908
10.2547
18.0172
7.1035
7.7284
10.8021
18.1441
7.3057
8.1367
10.6141
17.9945
7.3868
8.163
11.2181
18.1597
7.9235
8.9046
11.2202
18.2827
8.2855
9.2298
12.7704
21.6656
8.6366
9.8136
13.1672
21.0397
9.0847
10.6178
Appendix A:
20
20
30
30
30
30
40
40
40
40
50
50
50
50
70
70
70
70
90
90
90
90
5
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1.5
4.6377
2.3661
6.3349
5.3149
4.2482
1.9742
5.2026
4.7025
4.3637
2.0398
4.9773
4.5914
3.8064
1.9923
3.9401
4.3097
3.8191
2.1731
3.5714
3.7503
3.3643
1.9762
4.2035
6.6498
1.1046
2.2809
4.2942
6.6624
1.2236
2.3558
4.4902
6.6510
1.8038
2.8021
4.5056
6.7198
2.0645
3.0181
4.6241
6.6778
2.3523
3.1577
4.7517
6.7586
0.0422
Tables of Results
4.2765
1.8302
7.3754
6.1991
4.1858
1.8176
7.2564
6.1242
3.9898
1.8290
6.6762
5.6779
3.9744
1.7602
6.4155
5.4619
3.8559
1.8022
6.1277
5.3223
3.7283
1.7214
6.3487
2.1625
1.1165
2.8383
2.5214
1.8696
0.9202
2.5925
2.2848
1.9292
0.9285
2.3391
2.1381
1.7476
0.9215
1.9476
2.0283
1.7605
0.9599
1.9041
1.7519
1.5721
0.8578
A. 38
1.5395
2.5295
0.3833
0.8064
1.5746
2.5392
0.4805
0.8898
1.6142
2.5550
0.5641
0.9511
1.6406
2.5585
0.6606
1.0333
1.6995
2.5599
0.7891
1.1069
1.7531
2.5628
0.1114
1.7655
.7755
2.9217
2.4986
1.7304
.7658
2.8245
2.4152
1.6908
.75
2.7409
2.3539
1.6644
.7465
2.6444
2.2717
1.6055
.7451
2.5159
2.1981
1.5519
.7422
2.3044
2.9487
2.0485
4.4838
3.8255
3.0074
1.9224
5.9716
4.2817
3.3653
1.5957
6.0264
4.6848
3.3485
2.5252
7.8528
5.0058
3.1049
2.0422
16.2341
7.4641
3.3538
1.4370
-5.0575
-13.4245
-1.8275
-2.4475
-5.6484
-13.4306
-1.2879
-3.2913
-5.8197
-13.8641
-2.1266
-3.1482
-5.6485
-13.4478
-3.7202
-4.2922
-6.5961
-13.4108
-5.3389
-4.1089
-7.2006
-14.4960
8.7803
13.4845
21.8515
10.2545
10.8745
14.0754
21.8576
9.7149
11.7183
14.2467
22.2911
10.5536
11.5752
14.0755
21.8748
12.1472
12.7192
15.0231
21.8378
13.7659
12.5359
15.6276
22.923
19.4397
Appendix A:
Tables of Results
A.7 Hybrids RCF with companding Results:

X = PAPR (Pre-coding ) PAPR (Pre-coding +RCF)
Y =CCDF of PAPR (Pre-coding) - CCDF of PAPR (Pre-coding +RCF)
Z= SNR (BER=
) (Pre-coding) SNR (BER=
) (Pre-coding +RCF)
X1 == PAPR (RCF) PAPR (Pre-coding +RCF)
Y1 =CCDF of PAPR (RCF) - CCDF of PAPR (Pre-coding +RCF)
Z1= SNR (BER=
) (RCF) SNR (BER=
) (Pre-coding +RCF)
Table A.29 (Pre-coding +RCF) Results and compared with the results of each of (Pre-coding) and (RCF).
Precoding
CR
X1
PAPR
Y1
CCDF OF
PAPR
Z1
WHT
WHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
4
3
2
1.5
1.3
1.1
4
3
2
1.5
1.3
1.1
4
3
2
8.8659
11.4801
14.8677
16.6813
17.6205
18.3732
4.2276
6.9879
10.3525
12.1038
12.7949
13.3286
3.5817
6.3420
9.7467
0.1148
0.2126
0.2786
0.2274
13.9718
11.3576
7.9700
6.1564
5.2172
4.4645
13.8834
11.1231
7.7585
6.0072
5.3161
4.7824
13.8832
11.1229
7.7182
3.8833
5.1262
6.8155
7.8561
8.2340
8.6977
1.6434
2.8918
4.6089
5.6285
6.0657
6.4410
1.5024
2.7508
4.4656
0.0031
0.0066
0.0295
0.8515
6.0213
4.7784
3.0891
2.0485
1.6706
1.2069
6.0206
4.7722
3.0551
2.0355
1.5983
1.223
6.0206
4.7722
3.0574
-0.1058
-0.6262
-2.5798
-13.3450
-18.3700
-18.3700
0.0853
0.0641
-1.8862
-5.2472
-10.2117
-18.3720
-0.1290
-0.3210
-1.4552
0.0842
0.1149
0.1658
4.5250
0.2032
0.4471
0.4901
0.3766
0.2034
0.4473
0.5304
A. 39
0.0038
0.0128
0.0635
0.8645
0.0038
0.0128
0.0612
0.2773
0.8072
0.8614
12.6248
0.0630
0.4221
1.2924
SNR
(BER=
)
11.7358
12.2562
14.2098
24.975
>30
>30
11.5427
11.5639
13.5142
16.8752
21.8397
>30
11.757
11.949
13.0832
Appendix A:
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DHT
DHT
DHT
DHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
1
1
1
1
1
1
1
1
1
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
Pilot
2
2
2
2
2
2
2
2
1.5
1.3
1.1
4
3
2
1.5
1.3
1.1
4
3
2
1.5
4
3
2
1.5
4
3
2
1.5
4
3
2
1.5
4
3
2
1.5
11.6978
12.7190
13.0431
0.0069
0.0115
0.0008
1.7911
2.3945
0.0323
4.1673
6.7675
9.3391
10.5003
3.5404
6.1023
8.8583
9.9215
-6.9472
-4.2539
-1.3648
-0.5168
8.4655
10.7834
13.4396
14.6473
3.9078
6.3048
9.1329
10.1414
0.6167
7.1348
4.6230
1.2907
1.2162
0.2536
0.5734
0.6713
0.4001
0.2728
0.5543
0.8366
0.4674
0.2914
0.7043
1.1197
0.5353
0.1210
0.2199
0.2092
0.1925
0.2900
0.4680
0.6292
0.4133
Tables of Results
5.7671
4.7459
4.4218
6.9518
6.9472
6.9579
5.1676
4.5642
6.9264
13.9437
11.3435
8.7719
7.6107
13.9245
11.3626
8.6066
7.5434
13.9059
11.2126
8.3235
7.4755
14.3722
12.0543
9.3981
8.1904
14.2032
11.8062
8.9781
7.9696
5.4920
5.9291
6.3389
0.3397
0.3397
0.3397
1.4546
1.6229
1.4687
1.5259
2.6731
3.9837
4.5440
1.3769
2.4963
3.8617
4.4768
-2.7662
-1.6302
-0.2838
0.1727
3.5038
4.6331
5.9192
6.3965
1.3830
2.4546
3.6787
4.2608
A. 40
0.8690
3.0141
1.7747
0.1083
1.0046
0.1099
0.2341
0.1993
0.2368
0.1019
0.1983
0.2183
0.3106
0.1318
0.2448
0.2458
0.1795
0.0169
0.0835
0.1362
0.1176
0.1367
0.1456
0.1363
0.2225
2.031
1.5939
1.1841
3.0103
3.0103
3.0103
1.8954
1.7271
1.8813
6.1381
4.9909
3.6803
3.12
6.1461
5.0267
3.6613
3.0462
6.1162
4.9802
3.6338
3.1773
6.4008
5.2715
3.9854
3.5081
6.281
5.2094
3.9853
3.4032
-5.1509
-10.4894
-18.3720
0.0759
0.0272
-0.0615
-1.0130
-1.4779
-1.5015
2.4153
2.0345
-1.2236
-10.8087
2.2559
1.7912
-1.1227
-8.0414
1.9934
1.8664
0.2611
-2.5626
2.8937
1.9358
-0.8971
-18.3700
2.7806
2.3883
-0.3089
-7.3951
12.7211
0.3139
0.8163
2.7321
16.9050
0.1473
0.2480
0.1484
7.1633
-0.0121
0.0047
0.2493
9.9306
-0.2286
0.1259
1.6791
15.4554
0.0316
-0.2802
-0.3991
-0.4000
-0.0795
0.1743
0.1911
10.5769
16.7789
22.1174
>30
11.5061
11.5548
11.6435
12.595
13.0599
13.0835
9.2127
9.5935
12.8516
22.4367
9.3721
9. 8368
12.7507
19.6694
9.5886
9.7156
11.3209
14.1446
8.7363
9.6942
12.5271
>30
8.8474
9.2397
11.9369
19.0231
Appendix A:
DST
DST
DST
DST
DHT
DHT
DHT
DHT
2
2
2
2
2
2
2
2
4
3
2
1.5
4
3
2
1.5
3.2174
5.7121
8.4680
9.5290
-7.2690
-4.7567
-1.9188
-0.5578
0.2457
0.5214
0.6104
0.4470
0.2655
0.5588
0.7298
0.8664
Tables of Results
14.2475
11.7528
8.9969
7.9359
14.2277
11.7154
8.8775
7.5165
1.2382
2.3688
3.6055
4.1101
-2.9015
-1.7979
-0.6022
-0.0681
0.1329
0.2008
0.2041
0.2128
0.1662
0.2071
0.1694
0.2076
6.2848
5.1542
3.9175
3.4129
6.2515
5.1479
3.9522
3.4181
2.8864
2.6270
-0.3350
-7.2475
2.7701
2.6878
0.7115
-0.8762
0.0263
0.4130
0.1650
10.7245
-0.0440
0.5198
1.2575
17.1418
8.7416
9.001
11.963
18.8755
8.8119
8.8942
10.8705
12.4582
A.8 Hybrid Pre-coding with Companding Results:

X = PAPR (Pre-coding ) PAPR (Pre-coding + Companding)
Y =CCDF of PAPR (Pre-coding) - CCDF of PAPR(Pre-coding + Companding)
Z= SNR (BER=
) (Pre-coding) SNR (BER=
) (Pre-coding + Companding)
X1 == PAPR (Companding) PAPR (Pre-coding + Companding)
Y1 =CCDF of PAPR (Companding) - CCDF of PAPR (Pre-coding + Companding)
Z1= SNR (BER=
) (Pre-coding + Companding)
Table A.30 (Pre-coding +A) Results and compared with the results of each of (Pre-coding) and (A companding).
Precoding
X1
PAPR
Y1
WHT
WHT
WHT
WHT
WHT
WHT
WHT
5
10
15
20
30
35
40
6.4003
8.2527
9.0269
9.4809
10.0185
10.1382
10.2848
2.4687
0.1067
0.5157
1.0833
-0.9647
0.2105
0.0499
16.4374
14.5850
13.8108
13.3568
12.8192
12.6995
12.5529
4.2596
5.5140
6.0383
6.3435
6.7024
6.9423
7.0335
0.9950
0.3394
0.2597
0.1239
0.1178
0.4507
0.3889
A. 41
CCDF OF
PAPR
5.645
4.3906
3.8663
3.5611
3.2022
2.9623
2.8711
Z1
-2.2393
-4.7003
-6.2322
-6.6714
-8.0292
-8.1607
-8.5875
-0.2693
-0.2103
-0.2862
-0.3014
-0.4592
-0.5907
-0.4605
SNR
(BER=
)
13.8693
16.3303
17.8622
18.3014
19.6592
19.7907
20.2175
Appendix A:
WHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
50
70
87.6
100
120
5
10
15
20
30
35
40
50
70
87.6
100
120
5
10
15
20
30
35
40
50
70
87.6
100
10.5114
10.8155
10.9967
11.0811
11.2105
6.2514
7.7579
8.3981
8.7766
9.5223
9.6724
9.5424
9.9838
9.9556
10.1086
10.6923
10.3032
6.1302
7.6475
8.2865
8.6628
9.1103
8.2173
8.3394
8.5285
8.7829
8.9348
9.1724
0.2643
0.1281
-0.0591
-0.4023
-0.2171
7.0465
4.3386
4.6136
5.1057
3.2658
4.4714
4.0342
4.4634
3.9949
3.7795
3.9356
3.6023
7.5714
4.8743
5.1481
5.6380
3.4999
3.6624
3.4773
3.6542
3.4683
3.2518
3.0618
Tables of Results
12.3263
12.0222
11.8410
11.7566
11.6272
11.8596
10.3531
9.7129
9.3344
8.5887
8.4386
8.5686
8.1272
8.1554
8.0024
7.4187
7.8078
11.3347
9.8174
9.1784
8.8021
8.3546
9.2476
9.1255
8.9364
8.6820
8.5301
8.2925
7.1740
7.3616
7.4729
7.5333
7.6154
3.4636
4.3642
4.7366
4.9509
5.2077
5.2937
5.4200
5.4722
5.5693
5.6578
5.7713
5.7703
3.3458
4.2613
4.6392
4.8599
5.1202
5.1115
5.1855
5.2998
5.4526
5.5434
5.6106
A. 42
0.3694
0.2940
0.2913
0.2087
0.2458
2.4396
1.4302
1.1986
0.9719
0.8637
1.0427
1.0160
0.9082
0.7423
0.7168
0.6873
0.6413
2.4628
1.4683
1.2422
1.0219
0.9172
1.0015
0.9225
0.8768
0.7666
0.7434
0.6676
2.7306
2.543
2.4317
2.3713
2.2892
4.2004
3.2998
2.9274
2.7131
2.4563
2.3703
2.244
2.1918
2.0947
2.0062
1.8927
1.8937
4.1772
3.2617
2.8838
2.6631
2.4028
2.4115
2.3375
2.2232
2.0704
1.9796
1.9124
-10.0027
-9.5360
-10.0027
-10.6669
-10.9249
-2.4773
-4.6072
-6.0249
-6.7160
-7.3912
-7.7982
-8.0486
-8.6099
-9.0955
-9.5582
-9.7262
-10.0745
-2.7785
-4.8245
-6.4536
-7.0850
-8.3413
-8.2486
-8.4777
-9.0263
-9.6570
-10.1584
-10.2688
-1.5167
-0.0360
-0.2607
-0.6769
-0.6919
-0.5053
-0.1152
-0.0769
-0.3440
0.1808
-0.2262
0.0804
-0.1219
0.4065
0.1858
0.2658
0.1605
-0.8065
-0.3325
-0.5056
-0.7130
-0.7693
-0.6766
-0.3487
-0.5383
-0.1550
-0.4144
-0.2768
21.6327
21.166
21.6327
22.2969
22.5549
14.1053
16.2352
17.6529
18.344
19.0192
19.4262
19.6766
20.2379
20.7235
21.1862
21.3542
21.7025
14.4065
16.4525
18.0816
18.713
19.9693
19.8766
20.1057
20.6543
21.285
21.7864
21.8968
Appendix A:
DST
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
120
5
10
15
20
30
35
40
50
70
87.6
100
120
9.2816
2.2752
2.6637
2.8227
2.9158
3.0324
3.0690
3.0988
3.1450
3.2239
3.2613
3.2819
3.3088
3.2268
14.2226
10.3967
10.1905
10.3972
7.9282
9.0203
8.7429
8.7769
8.4155
8.0845
7.6775
7.7602
Tables of Results
8.1833
4.6835
4.2950
4.1360
4.0429
3.9263
3.8897
3.8599
3.8137
3.7348
3.6974
3.6768
3.6499
5.6740
0.9293
1.0712
1.1343
1.1712
1.1238
1.1358
1.1456
1.1608
1.1915
1.2050
1.2125
1.2224
0.6860
4.2193
2.4512
1.9103
1.5062
1.0938
1.1988
1.0556
0.9108
0.6785
0.5780
0.4425
0.4074
1.849
2.4207
2.2788
2.2157
2.1788
2.2262
2.2142
2.2044
2.1892
2.1585
2.145
2.1375
2.1276
-10.5360
-0.8663
-1.5606
-2.2378
-2.7303
-3.4581
-3.6582
-3.7728
-4.0772
-4.4893
-4.6584
-4.7606
-4.9040
-0.3010
1.1517
2.9774
3.7562
3.6877
4.1599
3.9598
4.4022
4.4568
5.0587
5.1316
5.2774
5.3770
22.164
12.4483
13.1426
13.8198
14.3123
15.0401
15.2402
15.3548
15.6592
16.0713
16.2404
16.3426
16.486
Table A.31(Pre-coding + ) Results and compared with the results of each of (Pre-coding) and ( companding).
Precoding
WHT
WHT
WHT
WHT
WHT
WHT
WHT
5
10
20
30
40
50
60
X1
PAPR
Y1
6.3192
8.7224
9.9641
8.8772
9.2710
10.2718
8.9088
0.9147
2.4317
4.3837
0.6751
0.2786
1.2963
-0.1962
16.5185
14.1153
12.8736
13.9605
13.5667
12.5659
13.9289
4.1987
5.1385
5.8731
6.1212
6.3771
6.6101
6.4984
0.7101
1.0039
1.0465
0.6166
0.4725
0.5055
0.2538
A. 43
CCDF OF
PAPR
5.7059
4.7661
4.0315
3.7834
3.5275
3.2945
3.4062
Z1
-1.8052
-3.1978
-4.9404
-5.8312
-6.7931
-7.0780
-7.2855
-0.0989
-0.1878
-0.3204
-0.2962
-0.6731
-0.4380
-0.1385
SNR
(BER=
)
13.4352
14.8278
16.5704
17.4612
18.4231
18.708
18.9155
Appendix A:
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
70
80
90
100
120
140
160
180
200
220
240
250
255
260
280
300
320
500
700
1000
5
10
20
30
40
50
60
70
9.0814
11.7882
11.9079
10.9096
11.5372
11.6736
9.8688
9.9644
11.3965
11.5960
11.6595
12.1288
12.1429
11.3810
11.8364
11.8831
11.8574
12.1312
12.4720
13.5600
5.2168
6.4678
6.0880
9.0769
8.2610
8.7019
8.8635
8.3644
0.4674
5.5006
1.9286
0.4861
0.8137
1.5738
-0.7606
-0.2875
0.8011
0.0305
4.7362
1.7269
1.0486
0.5296
0.5878
0.9527
1.3900
0.4886
0.4561
3.6176
4.5390
4.9038
5.2343
5.6015
3.9953
4.4531
4.4852
4.4771
Tables of Results
13.7563
11.0495
10.9298
11.9281
11.3005
11.1641
12.9689
12.8733
11.4412
11.2417
11.1782
10.7089
10.6948
11.4567
11.0013
10.9546
10.9803
10.7065
10.3657
9.2777
12.8942
11.6432
12.0230
9.0341
9.8500
9.4091
9.2475
9.7466
6.6125
7.0090
7.0824
6.9954
7.1759
7.2552
7.1305
7.1930
7.3336
7.3780
7.4171
7.5244
7.5329
7.5448
7.4979
7.5278
7.5393
7.7131
7.8745
8.0035
3.2318
3.8305
4.1559
4.7366
4.8099
5.0522
5.1270
5.1100
A. 44
0.3539
0.4344
0.4778
0.2108
0.2713
0.4106
0.0919
0.1684
0.1890
0.1184
0.3625
0.3428
0.3083
0.3062
0.1933
0.2332
0.3347
0.1085
0.1399
0.3789
1.9838
1.9365
1.5699
1.4726
1.1459
1.1882
1.1230
1.0920
3.2921
2.8956
2.8222
2.9092
2.7287
2.6494
2.7741
2.7116
2.571
2.5266
2.4875
2.3802
2.3717
2.3598
2.4067
2.3768
2.3653
2.1915
2.0301
1.9011
4.4322
3.8335
3.5081
2.9274
2.8541
2.6118
2.537
2.554
-7.4804
-8.0466
-8.1737
-8.4410
-9.1505
-9.4242
-9.2081
-9.8181
-10.0901
-10.3356
-10.5607
-10.3413
-10.4272
-10.3295
-10.6842
-10.8583
-11.0382
-11.5297
-11.9464
-12.5842
-1.8527
-3.3271
-4.7831
-5.5879
-6.2800
-6.7103
-7.2375
-7.4138
-0.1104
-0.2026
-0.2037
-0.0710
-0.5005
-0.5542
-0.3631
-0.3981
-0.7201
-0.7271
-0.5907
-0.3713
-0.5892
-0.2595
-0.4142
-0.4883
-0.5432
-0.2397
-0.0764
-0.4502
-0.1444
-0.3151
-0.1611
-0.0509
-0.1580
-0.0683
-0.0885
-0.0418
19.1104
19.6766
19.8037
20.071
20.7805
21.0542
20.8381
21.4481
21.7201
21.9656
22.1907
21.9713
22.0572
21.9595
22.3142
22.4883
22.6682
23.1597
23.5764
24.2142
13.4807
14.9551
16.4111
17.2159
17.908
18.3383
18.8655
19.0418
Appendix A:
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DST
DST
DST
DST
DST
80
90
100
120
140
160
180
200
220
240
250
255
260
280
300
320
500
700
1000
5
10
20
30
40
50
60
70
80
9.4263
9.7652
9.4938
8.9284
10.3110
9.7695
10.1595
10.2138
9.3948
10.3095
9.0514
10.4596
9.5897
10.6507
9.8587
10.7427
10.1111
10.8348
11.1737
5.4486
6.8133
6.7980
8.3677
7.6474
9.2676
9.0030
9.2551
9.2642
7.8654
4.5126
3.7970
2.9316
4.9379
3.8668
4.6343
4.3451
2.5560
8.1129
3.3762
4.0920
3.4650
4.1288
3.6550
5.0020
3.1952
3.5456
5.9580
5.4169
5.8954
6.5904
5.5384
4.0278
5.6649
5.2708
6.0139
8.3494
Tables of Results
8.6847
8.3458
8.6172
9.1826
7.8000
8.3415
7.9515
7.8972
8.7162
7.8015
9.0596
7.6514
8.5213
7.4603
8.2523
7.3683
7.9999
7.2762
6.9373
12.0163
10.6516
10.6669
9.0972
9.8175
8.1973
8.4619
8.2098
8.2007
5.3400
5.3276
5.3421
5.3968
5.5709
5.5726
5.6011
5.5775
5.6864
5.6913
5.7476
5.7487
5.6988
5.8074
5.7534
5.8207
5.9902
6.0670
6.1554
3.0512
3.7000
4.1850
4.6087
4.7031
5.0013
5.0522
5.1230
5.2023
A. 45
1.0060
0.9636
0.7981
0.7328
0.9669
0.7746
0.8171
0.6735
0.6674
0.8773
0.8066
0.7647
0.7008
0.7434
0.6994
0.8567
0.6262
0.5730
0.7714
1.9442
1.9470
1.7400
1.4857
1.1801
1.2783
1.1892
1.2460
1.0093
2.324
2.3364
2.3219
2.2672
2.0931
2.0914
2.0629
2.0865
1.9776
1.9727
1.9164
1.9153
1.9652
1.8566
1.9106
1.8433
1.6738
1.597
1.5086
4.4718
3.823
3.338
2.9143
2.8199
2.5217
2.4708
2.4
2.3207
-7.4558
-8.0063
-8.2689
-8.6476
-8.9126
-8.9073
-9.0835
-9.6093
-9.1604
-9.7126
-9.8144
-9.8207
-10.0232
-10.0021
-10.2307
-10.4537
-10.8432
-11.7915
-12.0486
-2.3432
-3.5551
-5.0565
-6.3507
-6.8122
-6.9744
-7.7744
-8.1161
-8.4165
0.3902
-0.0343
0.1031
0.0044
-0.0406
-0.0603
0.3385
-0.2373
0.4501
0.2594
0.1576
0.0193
0.0488
0.2699
0.1413
0.0433
0.4488
0.0805
0.0874
-0.6349
-0.5431
-0.4345
-0.8137
-0.6902
-0.3324
-0.6254
-0.7441
-0.5705
19.0838
19.6343
19.8969
20.2756
20.5406
20.5353
20.7115
21.2373
20.7884
21.3406
21.4424
21.4487
21.6512
21.6301
21.8587
22.0817
22.4712
23.4195
23.6766
13.9712
15.1831
16.6845
17.9787
18.4402
18.6024
19.4024
19.7441
20.0445
Appendix A:
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
90
100
120
140
160
180
200
220
240
250
255
260
280
300
320
500
700
1000
5
10
20
30
40
50
60
70
80
90
9.4738
9.3820
9.7971
9.7060
10.0042
9.8794
10.1197
9.5615
10.2651
9.6395
9.5708
8.1344
9.7061
8.2172
9.1446
10.3526
8.6398
10.6545
1.9728
2.3552
2.6599
2.8047
2.8756
2.9382
2.9855
3.0230
3.0564
3.0824
4.8673
4.3313
4.4464
4.9790
4.7476
5.0003
4.8971
3.3688
8.7146
4.6104
3.8493
2.6558
3.8303
2.6596
4.0500
4.0828
1.9967
6.0849
12.4473
11.9435
12.9585
10.4816
9.7622
9.8417
9.7595
10.2880
12.6478
8.9821
Tables of Results
7.9911
8.0829
7.6678
7.7589
7.4607
7.5855
7.3452
7.9034
7.1998
7.8254
7.8941
9.3305
7.7588
9.2477
8.3203
7.1123
8.8251
6.8104
4.9859
4.6035
4.2988
4.1540
4.0831
4.0205
3.9732
3.9357
3.9023
3.8763
5.2485
5.3318
5.3680
5.4550
5.4857
5.5538
5.5763
5.6042
5.6338
5.6467
5.6370
5.5533
5.6830
5.6017
5.6361
5.8364
5.8487
6.0095
0.7742
0.9217
1.0222
1.0703
1.0526
1.0762
1.0940
1.1081
1.0898
1.0982
A. 46
1.0255
0.9288
0.8450
0.9920
0.8287
0.9108
0.8133
0.7262
0.9608
0.8467
0.7940
0.6963
0.7600
0.6887
0.8131
0.6134
0.4957
0.7665
3.8402
3.3417
2.7502
2.1203
1.7026
1.5262
1.4040
1.4041
1.0698
1.0482
2.2745
2.1912
2.155
2.068
2.0373
1.9692
1.9467
1.9188
1.8892
1.8763
1.886
1.9697
1.84
1.9213
1.8869
1.6866
1.6743
1.5135
2.5758
2.4283
2.3278
2.2797
2.2974
2.2738
2.256
2.2419
2.2602
2.2518
-8.5426
-8.6108
-9.0312
-9.3167
-9.6425
-9.7710
-10.0101
-10.1189
-10.4585
-10.3842
-10.4538
-10.7142
-10.7494
-10.8662
-10.8145
-11.6766
-11.6894
-12.5300
-0.7449
-1.2059
-2.1132
-2.4565
-2.7943
-3.3194
-3.5662
-3.7728
-3.7046
-3.7978
-0.5706
-0.2388
-0.3792
-0.4447
-0.7955
-0.3490
-0.6381
-0.5084
-0.4865
-0.4122
-0.6138
-0.6422
-0.4774
-0.4942
-0.3175
-0.3846
0.1826
-0.3940
1.0094
1.8521
2.5548
3.1265
3.3737
3.3686
3.6288
3.6452
4.1874
4.2202
20.1706
20.2388
20.6592
20.9447
21.2705
21.399
21.6381
21.7469
22.0865
22.0122
22.0818
22.3422
22.3774
22.4942
22.4425
23.3046
23.3174
24.158
12.3269
12.7879
13.6952
14.0385
14.3763
14.9014
15.1482
15.3548
15.2866
15.3798
Appendix A:
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
100
120
140
160
180
200
220
240
250
255
260
280
300
320
500
700
1000
3.0929
3.1412
3.1703
3.1929
3.2132
3.2306
3.2458
3.2399
3.2541
3.2490
3.2519
3.2708
3.2715
3.2898
3.3475
3.3767
3.4307
8.5484
8.2967
8.9495
8.4425
8.8403
8.5142
7.5593
12.1956
8.7312
8.0337
8.2795
7.9012
8.2201
8.7014
7.5839
7.2398
9.3673
Tables of Results
3.8658
3.8175
3.7884
3.7658
3.7455
3.7281
3.7129
3.7188
3.7046
3.7097
3.7068
3.6879
3.6872
3.6689
3.6112
3.5820
3.5280
1.1892
1.1173
1.1267
1.2065
1.2134
1.2193
1.2244
1.1561
1.2272
1.1591
1.1600
1.2337
1.2504
1.2412
1.2639
1.2848
1.3010
0.9592
0.7673
0.8367
0.7225
0.7434
0.6293
0.5194
0.6561
0.6002
0.4891
0.4760
0.4837
0.5104
0.5912
0.2139
0.1048
0.2310
2.1608
2.2327
2.2233
2.1435
2.1366
2.1307
2.1256
2.1939
2.1228
2.1909
2.19
2.1163
2.0996
2.1088
2.0861
2.0652
2.049
-3.7443
-4.0950
-4.2397
-4.2496
-4.3751
-4.4275
-4.5322
-4.8847
-4.9731
-5.2350
-5.2350
-5.0874
-5.0789
-5.3548
-5.7347
-5.7148
-6.0038
4.6737
4.6030
4.6783
4.6434
5.0929
4.9905
5.1243
5.1333
5.0449
4.6510
4.8830
5.2306
5.3391
5.1882
5.6033
6.2032
6.1782
15.3263
15.677
15.8217
15.8316
15.9571
16.0095
16.1142
16.4667
16.5551
16.817
16.817
16.6694
16.6609
16.9368
17.3167
17.2968
17.5858
Table A.32 (Pre-coding + RCT) Results and compared with the results of each of (Pre-coding) and (RCT)
Pre-coding
X1
PAPR
Y1
WHT
WHT
WHT
WHT
WHT
WHT
WHT
.9
.8
.7
.6
.5
.4
.3
3.3459
5.1820
6.1784
8.2743
10.1461
13.2863
15.4834
2.3713
3.4754
1.4698
0.7913
1.2348
1.9778
1.1986
19.4918
17.6557
16.6593
14.5634
12.6916
9.5514
7.3543
1.3932
2.1560
2.6885
3.5811
4.3915
5.6906
6.6295
1.0386
0.9329
0.8419
0.3590
0.3219
0.6075
0.4349
A. 47
CCDF OF
PAPR
8.5114
7.7486
7.2161
6.3235
5.5131
4.214
3.2751
Z1
-0.1006
-0.2737
-1.0924
-1.9051
-3.0635
-4.9407
-7.1992
-0.0541
0.0833
-0.3087
-0.1351
-0.2435
-0.4257
-0.3042
SNR
(BER=
)
11.7306
11. 9037
12.7224
13.5351
14.6935
16.5707
18.8292
Appendix A:
WHT
WHT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DST
DST
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
.2
.1
.9
.8
.7
.6
.5
.4
.3
.2
.1
.9
.8
.7
.6
.5
.4
.3
.2
.1
.9
.8
.7
.6
.5
.4
.3
.2
.1
17.7489
20.1831
2.6159
4.0365
5.5106
6.0660
7.8308
9.6726
11.9638
13.6369
15.8025
1.7433
3.1975
4.7059
6.2744
7.9097
9.1755
11.0822
13.0856
15.2166
0.3560
0.7383
1.1257
1.5174
1.9183
2.3380
2.7565
3.1938
3.6504
0.9000
0.2180
6.3680
7.0566
5.5287
3.3097
3.6462
3.0908
2.4057
1.5147
0.5641
6.1415
6.8637
5.3701
4.1642
4.3712
3.2398
2.1702
1.6095
0.6243
15.2604
14.9107
12.2961
9.9134
8.8860
6.9085
4.3507
2.2239
-0.4357
Tables of Results
5.0888
2.6546
15.4951
14.0745
12.6004
12.0450
10.2802
8.4384
6.1472
4.4741
2.3085
15.7216
14.2674
12.7590
11.1905
9.5552
8.2894
6.3827
4.3793
2.2483
6.6027
6.2204
5.8330
5.4413
5.0404
4.6207
4.2022
3.7649
3.3083
7.6444
8.7120
0.8902
1.5129
2.1557
2.5025
3.2498
4.0337
4.9843
5.7340
6.6729
0.7555
1.3692
2.0010
2.6500
3.3369
3.9292
4.7562
5.6273
6.5535
0.1052
0.2788
0.5084
0.6703
0.7518
0.8374
1.0510
1.0568
1.2249
A. 48
0.3143
0.0754
2.7762
2.5304
2.5497
1.5210
1.4208
1.1912
1.0303
0.6445
0.2769
2.7825
2.5277
2.5360
1.8095
1.6489
1.2277
0.9432
0.6788
0.2985
6.3052
5.6103
5.2164
4.0028
3.2368
2.3089
1.4110
0.2813
-0.8571
2.2602
1.1926
6.7738
6.1511
5.5083
5.1615
4.4142
3.6303
2.6797
1.93
.9911
6.7675
6.1538
5.522
4.873
4.1861
3.5938
2.7668
1.8957
.9695
3.2448
3.0712
2.8416
2.6797
2.5982
2.5126
2.299
2.2932
2.1251
-10.7564
-17.0412
-0.0147
-0.3815
-1.0401
-1.8911
-3.0857
-4.7014
-6.6401
-10.2162
-16.5112
-0.0393
-0.3827
-0.9126
-1.8332
-2.9499
-4.6498
-7.0794
-10.5597
-16.8057
0.0718
-0.2792
-0.6722
-1.1254
-1.5191
-2.5360
-3.4451
-5.8003
-9.5778
-0.1364
-0.3712
0.0338
-0.0225
-0.2544
-0.1191
-0.2637
-0.1844
0.2569
0.4058
0.1608
0.0092
-0.0237
-0.1269
-0.0612
-0.1279
-0.1328
-0.1824
0.0623
-0.1337
0.1663
0.1258
0.1595
0.6926
1.3489
2.0270
3.4979
4.8677
7.1402
22.3864
28.6712
11.6427
12.0095
12.6681
13.5191
14.7137
16.3294
18.2681
21.8442
28.1392
11.6673
12.0107
12.5406
13.4612
14.5779
16.2778
18.7074
22.1877
28.4337
11.5102
11.8612
12.2542
12.7074
13.1011
14.118
15.0271
17.3823
21.1598
Appendix A:
Tables of Results
Table A.33(Pre-coding +AEXP) Results and compared with the results of each of (Pre-coding) and (AEXP companding).
Precoding
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
AEXP d
X1
PAPR
Y1
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
2
1.9
1.8
1.7
10.7804
11.3253
11.7057
12.2595
12.3909
12.9152
13.4360
14.2743
14.8045
15.2924
15.8627
16.4192
17.0353
17.4907
18.2253
18.8350
19.5373
20.2553
21.0167
21.8494
6.7007
7.1704
7.6401
8.2400
1.0238
1.5116
1.0663
0.6391
0.2494
0.1439
0.2128
0.4181
0.4168
0.4070
0.4024
0.3315
0.2782
0.1783
0.1768
0.1317
0.1035
0.0694
0.0148
-0.0193
1.6708
2.0834
1.7274
1.3463
12.0573
11.5124
11.1320
10.5782
10.4468
9.9225
9.4017
8.5634
8.0332
7.5453
6.9750
6.4185
5.8024
5.3470
4.6124
4.0027
3.3004
2.5824
1.8210
0.9883
11.4103
10.9406
10.4709
9.8710
4.9111
5.0575
5.2111
5.4254
5.5148
5.6605
5.8436
6.0031
6.2983
6.4313
6.6393
6.8279
7.1031
7.2846
7.6039
7.9506
8.2612
8.5546
8.9118
9.3313
2.9261
3.0746
3.2279
3.4745
0.1598
0.2929
0.2250
0.2908
0.1952
0.1139
0.0855
0.0785
0.1997
0.1262
0.1087
0.1153
0.1185
0.0170
0.0393
0.1535
0.0966
-0.0228
-0.0503
-0.0617
0.4154
0.5506
0.4824
0.5805
A. 49
CCDF OF
PAPR
4.9935
4.8471
4.6935
4.4792
4.3898
4.2441
4.061
3.9015
3.6063
3.4733
3.2653
3.0767
2.8015
2.62
2.3007
1.954
1.6434
1.35
.9928
.5733
4.7379
4.5894
4.4361
4.1895
Z1
-4.3415
-4.5962
-5.3536
-4.7950
-4.3508
-5.2656
-4.0073
-4.0536
-4.1271
-4.4413
-4.2999
-4.2078
-5.5993
-6.5697
-9.5950
-18.3700
-18.3700
-18.3700
-18.3700
-18.3700
-3.9507
-3.9175
-3.8548
-3.1582
-1.2415
-1.5262
-2.1256
-1.9750
-1.7808
-2.5818
-1.0688
-1.3836
-0.4271
-1.2213
-0.6299
0.4622
2.0157
11.8003
8.7750
0
0
0
0
0
-0.8487
-0.8455
-0.6248
-0.3362
SNR
(BER=
)
15.9715
16.2262
16.9836
16.425
15.9808
16.8956
15.6373
15.6836
15.7571
16.0713
15.9299
15.8378
17.2293
18.1997
21.225
>30
>30
>>30
>>30
>>>30
15.5787
15.5455
15.4828
14.7862
Appendix A:
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
8.5743
9.0476
9.5246
10.0065
10.4319
10.9577
11.4605
11.9813
12.5214
13.1121
13.6710
14.2768
14.9229
15.6162
16.3628
17.1819
5.7603
6.2487
6.7376
7.4201
7.9659
8.2102
8.6838
9.2006
9.6836
10.2053
10.6868
11.3528
11.9005
1.1595
1.0030
1.0281
0.8770
0.7709
0.7990
0.7269
0.6203
0.4910
0.5264
0.3492
0.3002
0.2158
0.1570
0.0876
0.0399
1.3765
1.8078
1.4710
1.1725
1.1972
0.8117
0.8334
0.7172
0.6687
0.6927
0.5993
0.6379
0.5162
Tables of Results
9.5367
9.0634
8.5864
8.1045
7.6791
7.1533
6.6505
6.1297
5.5896
4.9989
4.4400
3.8342
3.1881
2.4948
1.7482
0.9291
11.7046
11.2162
10.7273
10.0448
9.4990
9.2547
8.7811
8.2643
7.7813
7.2596
6.7781
6.1121
5.5644
3.6095
3.7703
3.9270
4.0957
4.5142
4.3928
4.5868
4.8018
5.0146
5.3103
5.5512
5.8225
6.1132
6.4346
6.8036
7.1850
2.8291
2.9692
3.1148
3.3569
3.4962
3.5812
3.7896
3.9663
4.1170
4.3442
4.5007
4.7532
4.9746
A. 50
0.5305
0.4643
0.4095
0.4117
0.6562
0.3283
0.2968
0.3298
0.2706
0.2833
0.2272
0.2660
0.1892
0.0978
0.0821
0.0326
0.4594
0.5862
0.5103
0.6039
0.5582
0.4162
0.4131
0.4233
0.4000
0.4207
0.3517
0.4222
0.3716
4.0545
3.8937
3.737
3.5683
3.1498
3.2712
3.0772
2.8622
2.6494
2.3537
2.1128
1.8415
1.5508
1.2294
.8604
.479
4.6939
4.5538
4.4082
4.1661
4.0268
3.9418
3.7334
3.5567
3.406
3.1788
3.0223
2.7698
2.5484
-3.2536
-2.9112
-2.7500
-2.6924
-2.4536
-2.4922
-3.0850
-3.7744
-5.2869
-6.8777
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-3.4924
-2.9424
-1.7942
-2.1349
-1.5678
-1.8026
-2.0358
-2.4312
-3.5155
-4.6118
-17.9720
-18.3720
-18.3720
-0.6816
-0.2254
0.1905
-0.0204
1.2484
0.7298
0.5870
0.8976
2.3301
11.4943
0
0
0
0
0
0
-0.3904
0.1296
1.4358
0.6871
1.0042
0.8832
0.9047
0.2408
0.1865
-1.3898
-14.3000
-13.7000
-10.7550
14.8816
14.5392
14.378
14.3204
14.0816
14.1202
14.713
15.4024
16.9149
18.5057
>30
>30
>30
>>30
>>30
>>>30
15.1204
14.5704
13.4222
13.7629
13.1958
13.4306
13.6638
14.0592
15.1435
16.2398
29.6
>30
>30
Appendix A:
DST
DST
DST
DST
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
.7
.6
.5
.4
.3
.2
.1
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
12.4544
13.0419
13.6703
14.3023
15.0261
15.7624
16.5640
-0.6937
-0.4865
-0.3933
-0.2155
0.0648
0.3050
0.2838
0.5355
0.8006
1.1528
1.5750
1.8437
2.1033
2.3210
2.5943
2.8462
3.1118
3.3695
3.6525
3.8711
0.5148
0.3662
0.3398
0.2413
0.2130
0.1333
0.0681
5.4287
5.5788
4.8463
4.0431
3.8023
3.4127
2.9396
2.5583
2.2919
2.1464
1.9937
1.6350
1.2252
0.8876
0.4248
0.0219
-0.4430
-0.9374
-1.4704
-2.1186
Tables of Results
5.0105
4.4230
3.7946
3.1626
2.4388
1.7025
0.9009
7.6524
7.4452
7.3520
7.1742
6.8939
6.6537
6.6749
6.4232
6.1581
5.8059
5.3837
5.1150
4.8554
4.6377
4.3644
4.1125
3.8469
3.5892
3.3062
3.0876
5.2316
5.4339
5.6985
5.9822
6.3203
6.6784
7.0644
0.1988
0.2950
0.3265
0.3825
0.5006
0.5780
0.6027
0.6680
0.7343
0.7608
0.9006
0.9706
1.0415
1.0391
1.1280
1.1485
1.2126
1.2777
1.3818
1.3298
A. 51
0.3456
0.2509
0.2830
0.1992
0.1245
0.0979
0.0530
2.0021
2.0850
1.8950
1.8025
1.7356
1.5860
1.3992
1.2980
1.1903
1.0103
0.9246
0.8126
0.6115
0.3261
0.1180
-0.0940
-0.3974
-0.7451
-1.0257
-1.5086
2.2914
2.0891
1.8245
1.5408
1.2027
.8446
.4586
3.1512
3.055
3.0235
2.9675
2.8494
2.772
2.7473
2.682
2.6157
2.5892
2.4494
2.3794
2.3085
2.3109
2.222
2.2015
2.1374
2.0723
1.9682
2.0202
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-1.9173
-1.5431
-1.5752
-1.0598
-0.8207
-0.8038
-1.0845
-1.0560
-0.8581
-0.3896
-0.5745
-0.6731
-0.8873
-1.4496
-1.7265
-2.3714
-3.0057
-4.1479
-18.4180
-18.4180
0
0
0
0
0
0
0
1.2307
1.5749
1.7008
1.8082
1.7973
1.9280
1.9020
1.6620
2.8899
2.8784
3.1435
4.0449
6.7757
6.9684
16.6915
16.0466
15.4123
14.2701
0
0
>30
>>30
>>30
>>30
>>30
>>>30
>>>30
13.4993
13.1251
13.1572
12.6418
12.4027
12.3858
12.6665
12.638
12.4401
11.9716
12.1565
12.2551
12.4693
13.0316
13.3085
13.9534
14.5877
15.7299
17.8183 -30
>>30
Appendix A:
Tables of Results
Table A.34 (Pre-coding +cos) Results and compared with the results of each of (Pre-coding) and (cos companding).
Precoding
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DST
Cos
X1
PAPR
Y1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
7.3889
8.6546
9.9471
11.3571
12.8150
14.2197
15.7226
17.3259
19.0411
20.8281
4.5103
5.5939
6.7188
7.8898
9.1125
10.0773
11.4935
12.9874
14.5824
16.2663
3.4964
5.0333
6.0390
7.0375
8.3999
0.1118
0.2175
0.2189
0.1886
0.2831
0.1510
0.0911
0.0641
15.4488
14.1831
12.8906
11.4806
10.0227
8.6180
7.1151
5.5118
3.7966
2.0096
13.6007
12.5171
11.3922
10.2212
8.9985
8.0337
6.6175
5.1236
3.5286
1.8447
13.9685
12.4316
11.4259
10.4274
9.0650
3.0367
3.5637
4.2290
4.7941
5.4976
6.0483
6.6971
7.4495
8.1231
8.9201
1.7336
2.2109
2.6944
3.1956
3.7198
4.1538
4.7682
5.4192
6.0790
6.8477
1.4699
2.0815
2.5870
3.0940
3.5805
-0.0179
0.0751
0.1894
0.1710
0.1960
0.0987
0.0715
0.1449
1.9599
1.8835
1.7173
1.4480
1.3073
0.7353
0.5887
0.4523
1.5921
1.9690
1.6836
1.2418
1.2408
A. 52
0.9196
0.9629
0.8954
0.8131
0.6588
0.4448
0.3832
0.3552
0.7969
0.9745
0.9290
0.8525
0.6605
CCDF of
PAPR
6.8679
6.3409
5.6756
5.1105
4.407
3.8563
3.2075
2.4551
1.7815
.9845
5.9304
5.4531
4.9696
4.4684
3.9442
3.5102
2.8958
2.2448
1.585
.8163
6.0531
5.4415
4.936
4.429
3.9425
Z1
-0.0936
-0.3785
-0.8393
-1.5116
-2.6935
-3.9494
-6.0413
-9.0243
-18.3700
-18.3700
-0.0147
-0.3795
-1.0026
-1.5364
-2.5236
-3.9940
-5.9277
-11.3055
-18.3720
-18.3720
-0.5719
-0.6383
-1.3140
-2.1014
-3.0753
0.0764
0.2165
0.1107
0.2424
0.2150
0.2806
-0.1113
0.6457
0.1573
0.2175
-0.0506
0.2196
0.3869
0.2380
0.0043
-1.6335
-0.3999
-0.0413
-0.3620
-0.3454
-0.1648
SNR
(BER=
)
11.7236
12.0085
12.4693
13.1416
14.3235
15.5794
17.6713
20.6543
>30
>>30
11.6427
12.0075
12.6306
13.1644
14.1516
15.622
17.5557
22.9335
>30
>>30
12.1999
12.2663
12.942
13.7294
14.7033
Appendix A:
DST
DST
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
.5
.4
.3
.2
.1
1
.9
.8
.7
.6
.5
.4
.3
.2
.1
9.6431
11.1108
12.4423
13.9884
15.6603
-0.8692
-0.3893
0.0275
0.5943
1.2213
1.6803
2.1705
2.6262
3.1364
3.6225
Tables of Results
0.9472
0.8521
0.5533
7.8218
6.3541
5.0226
3.4765
1.8046
7.8279
7.3480
6.9312
6.3644
5.7374
5.2784
4.7882
4.3325
3.8223
3.3362
7.7327
7.0526
6.1783
5.3048
4.5684
3.4906
2.4180
1.2434
4.1717
4.6548
5.3188
5.9826
6.7346
0.2092
0.3310
0.4508
0.5863
0.6807
0.7570
0.8625
1.0832
1.2095
1.3423
0.6037
0.4108
0.3958
3.7092
3.3970
2.9658
2.5178
1.9337
1.3620
0.7915
0.3332
3.3513
2.8682
2.2042
1.5404
.7884
3.1408
3.019
2.8992
2.7637
2.6693
2.593
2.4875
2.2668
2.1405
2.0077
-4.0155
-6.3518
-8.6413
-17.8447
-18.3720
-0.1667
-0.3262
-0.5046
-0.9311
-1.1133
-2.2444
-2.6397
-4.0128
-18.4180
-18.4180
0.2165
-0.4198
1.0307
0.0513
0.3168
0.4934
0.8709
1.8432
2.0336
3.3383
5.7052
15.6435
17.9798
20.2693
29.4727
>>>30
11.7487
11.9082
12.0866
12.5131
12.6953
13.8264
14.2217
15.5948
30
>>30
Table A.35 (Pre-coding +NERF) Results and compared with the results of each of (Pre-coding) and (NERF companding).
Precoding
NERF
WHT
DCT
DST
DHT
X1
PAPR
Y1
12.7937
9.0407
8.4569
1.1145
0.3968
1.3705
1.4328
4.5966
10.0440
9.0703
9.0080
5.8442
5.6684
3.7171
3.6090
0.6177
0.2838
0.5731
0.6060
1.7877
CCDF of
PAPR
4.2362
3.9469
3.914
2.7323
Z1
-2.1831
-1.7823
-0.4681
-0.0653
SNR
(BER=
)
13.8131
13.4103
non-NaN.
non-NaN
Table A.36 (Pre-coding + tanhR) Results and compared with the results of each of (Pre-coding) and (tanhR companding).
Precoding
X1
PAPR
A. 53
Y1
CCDF of
Z1
SNR
Appendix A:
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
DCT
DCT
DCT
DCT
5
5
5
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
5
5
5
5
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
-0.1807
3.7573
10.6855
18.3499
0.0656
4.2484
11.1658
18.6064
0.3235
4.9015
10.8964
18.7511
0.0789
5.7588
12.0332
19.1134
2.9821
6.1658
12.9867
18.9871
4.2670
7.5802
13.8960
19.3685
4.9948
7.4490
11.3493
15.4567
-6.3055
-5.2332
-3.2479
-0.9432
-13.8409
-10.4413
-5.2722
-1.1811
-17.2828
-12.6933
-6.9994
-1.3429
-19.3807
-13.4839
-7.0284
-1.1533
3.5967
3.1852
2.1426
0.8903
Tables of Results
23.0184
19.0804
12.1522
4.4878
22.7721
18.5893
11.6719
4.2313
22.5142
17.9362
11.9413
4.0866
22.7588
17.0789
10.8045
3.7243
19.8556
16.6719
9.8510
3.8506
18.5707
15.2575
8.9417
3.4692
13.1162
10.6620
6.7617
2.6543
0.2938
2.2354
4.7480
7.9541
0.3874
2.1231
4.9280
8.0609
0.4487
2.3646
5.1420
8.1767
0.3788
2.6213
5.1936
8.2956
1.4998
2.7222
5.6462
8.2285
1.8924
3.4669
5.9673
8.3837
1.8882
2.9477
4.6478
6.4640
A. 54
-2.2032
-1.6111
-1.2705
-0.4230
-5.3082
-4.0236
-2.1485
-0.4986
-6.8572
-5.0135
-2.5305
-0.5252
-7.7030
-5.4421
-2.9215
-0.4775
1.6318
1.3418
0.8699
0.3275
PAPR
9.6108
7.6692
5.1566
1.9505
9.5172
7.7815
4.9766
1.8437
9.4559
7.54
4.7626
1.7279
9.5258
7.2833
4.711
1.609
8.4048
7.1824
4.2584
1.6761
8.0122
6.4377
3.9373
1.5209
5.7758
4.7163
3.0162
1.2
0.1112
-0.6251
-3.0382
-10.9634
0.1112
-0.4102
-3.1630
-11.2273
-0.0924
-0.6197
-3.2753
-11.4066
-0.2736
-0.7083
-3.4580
-11.7669
0.1416
-0.6555
-3.4154
-12.3505
0.0095
-0.6978
-3.9522
-12.5536
-0.2843
-1.2305
-4.0362
-12.4393
0.6106
0.4161
0.9195
2.1297
3.1298
2.8919
2.8359
3.1586
17.8776
10.2159
6.5694
3.5210
17.6964
17.2617
14.5120
5.5423
0.2171
-0.1873
-0.0765
0.6558
(BER=
)
11.5188
12.2551
14.6682
22.5934
11.5188
12.0402
14.793
22.8573
11.7224
12.2497
14.9053
23.0366
11.9036
12.3383
15.088
23.3969
11.4884
12.2855
15.0454
23.9805
11.6205
12.3278
15.5822
24.1836
11.9123
12.8585
15.6642
24.0673
Appendix A:
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DST
DST
DST
DST
DST
DST
DST
DST
DST
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
.5
.5
.5
.5
1
1
1
1
5
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
10.0154
10.8270
12.7260
15.6243
13.2826
13.3063
13.9101
15.8469
14.9735
14.7766
14.7322
16.0073
16.4535
16.2397
15.7594
16.2183
17.0593
16.9063
16.3777
16.3828
2.7791
5.6384
10.1178
14.6923
7.7531
9.1948
11.8357
15.0636
16.5353
0.8356
0.8640
1.0147
0.5635
0.4030
0.4382
0.7410
0.4796
0.2406
0.2606
0.3973
0.4673
15.7833
Tables of Results
8.0956
7.2840
5.3850
2.4867
4.8284
4.8047
4.2009
2.2641
3.1375
3.3344
3.3788
2.1037
1.6575
1.8713
2.3516
1.8927
1.0517
1.2047
1.7333
1.7282
14.6858
11.8265
7.3471
2.7726
9.7118
8.2701
5.6292
2.4013
0.9296
4.0445
4.4516
5.3042
6.5815
5.2743
5.3592
5.7489
6.6630
5.9965
5.9706
6.1017
6.7400
6.6685
6.5711
6.5149
6.8333
6.9437
6.9363
6.7772
6.8898
1.2087
2.4243
4.3530
6.3053
3.1540
3.7915
4.9741
6.4409
6.8695
A. 55
0.5895
0.5455
0.4683
0.2626
0.2090
0.2217
0.3170
0.2017
0.1553
0.1478
0.2272
0.2075
6.7541
3.6195
3.2124
2.3598
1.0825
2.3897
2.3048
1.9151
1.001
1.6675
1.6934
1.5623
.924
.9955
1.0929
1.1491
.8307
.7203
.7277
.8868
.7742
6.3143
5.0987
3.17
1.2177
4.369
3.7315
2.5489
1.0821
.6535
-2.1407
-2.6784
-6.2448
-13.8424
-7.3080
-7.1678
-8.8231
-14.9256
-18.3720
-18.3720
-13.4514
-16.1143
-18.3720
-18.3720
-18.3720
-17.5814
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
0.8799
0.6257
-0.2439
0.5455
10.6640
3.6698
1.0236
0.0040
-0.4000
-0.4000
4.5206
1.1969
-17.8706
13.7687
14.3064
17.8728
25.4704
18.936
18.7958
20.4511
26.5536
>30
>30
25.0794
27.7423
>>30
>>30
>30
29.2094
>>30
>>30
>>30
>30
>30
>>30
>>>30
>>>30
>30
>30
>>>30
>>>30
>>>30
Appendix A:
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
5
5
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
5
5
5
5
10
10
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
16.3710
15.8264
15.8076
17.1973
17.1359
16.8045
16.1937
17.3352
17.3044
17.1086
16.3513
17.3829
17.3648
17.2385
16.4904
17.4153
17.4074
17.3451
16.6910
17.4283
17.4237
17.3863
16.7987
4.0819
4.0969
3.9269
3.8097
4.1283
4.1282
12.7533
7.2658
1.8873
8.6636
7.8190
5.7393
1.7790
5.1017
5.0824
4.5856
1.6301
3.2961
3.4949
3.5497
1.5965
13.8361
10.9854
5.8725
0.3956
6.1008
5.3175
Tables of Results
1.0939
1.6385
1.6573
0.2676
0.3290
0.6604
1.2712
0.1297
0.1605
0.3563
1.1136
0.0820
0.1001
0.2264
0.9745
0.0496
0.0575
0.1198
0.7739
0.0366
0.0412
0.0786
0.6662
2.8768
2.8618
3.0318
3.1490
2.8304
2.8305
6.8535
6.6730
6.7862
7.2727
7.2537
7.1194
6.9533
7.3498
7.3447
7.2769
7.0354
7.4045
7.3987
7.3456
7.0845
7.4273
7.4277
7.3967
7.1582
7.4550
7.4562
7.4420
7.2140
1.4368
1.4315
1.4552
1.3754
1.4791
1.4791
A. 56
5.3886
3.0361
0.7907
3.9587
3.4886
2.4245
0.7754
2.4255
2.3482
1.9860
0.7151
1.7043
1.7169
1.6121
0.6930
5.4944
4.1396
1.9913
-0.4471
2.3381
1.8870
.6695
.85
.7368
.2503
.2693
.4036
.5697
.1732
.1783
.2461
.4876
.1185
.1243
.1774
.4385
.0957
.0953
.1263
.3648
.068
.0668
.081
.309
1.9132
1.9185
1.8948
1.9746
1.8709
1.8709
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-18.3720
-1.4034
-0.9347
-1.3122
-5.7323
-1.5015
-1.0322
-17.3288
-14.4123
-5.2769
-15.3514
-15.0679
-12.3711
-3.9841
-0.4000
-7.5344
-8.5253
-3.4424
-0.4000
-0.4000
-0.4000
-1.0608
-0.8560
0.1545
2.6935
7.4088
1.5651
2.3179
>>>30
>>30
>30
>>>30
>>>30
>>>30
>>30
>>>30
>>>30
>>>30
>>30
>>>30
>>>30
>>>30
>>30
>>>30
>>>30
>>>30
>>>30
>>>30
>>>30
>>>30
>>>30
12.9854
12.5167
12.8942
17.3143
13.0835
12.6142
Appendix A:
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
4.0944
3.8978
4.1283
4.1283
4.1193
3.9387
4.0983
4.0983
4.0954
3.9368
4.1032
4.1264
4.1028
3.9758
4.1264
4.1024
4.1263
4.0197
Tables of Results
3.5354
-0.0107
2.4010
2.4125
2.1025
-0.2763
0.5177
0.7346
0.9128
-0.4509
2.8643
3.0609
2.8304
2.8304
2.8394
3.0200
2.8604
2.8604
2.8633
3.0219
2.8555
2.8323
2.8559
2.9829
2.8323
2.8563
2.8324
2.9390
1.4683
1.4054
1.4791
1.4791
1.4762
1.4185
1.5718
1.5619
1.5609
1.5058
1.5600
1.5207
1.4790
1.4365
1.5207
1.5619
1.5207
1.4816
0.9464
-0.5995
0.7278
0.6556
0.3583
-0.7288
0.0446
0.0531
0.0004
-0.7127
1.8817
1.9446
1.8709
1.8709
1.8738
1.9315
1.7782
1.7881
1.7891
1.8442
1.79
1.8293
1.871
1.9135
1.8293
1.7881
1.8293
1.8684
-1.1244
-5.7624
-1.5015
-1.0322
-1.0754
-5.7564
-1.3866
-0.6494
-1.2818
-5.5015
-1.3244
-0.8384
-1.0972
-5.7159
-1.4180
-0.8458
-1.2255
-5.5194
4.9225
8.6715
16.5165
9.8514
8.8173
9.2192
16.6314
17.3686
16.7362
11.8557
12.7064
17.3444
13.0835
12.6142
12.6574
17.3384
12.9686
12.2314
12.8638
17.0835
12.9064
12.4204
12.6792
17.2979
13
12.4278
12.8075
17.1014
Table A.37 (Pre-coding +logR) Results and compared with the results of each of (Pre-coding) and (logR companding).
Precoding
WHT
WHT
WHT
WHT
WHT
5
5
5
5
10
1
.8
.5
.2
1
0.4263
4.6287
11.9168
18.3833
1.0356
X1
PAPR
-5.0601
22.4114
18.2090
10.9209
4.4544
21.8021
0.5366
2.2852
5.1927
7.9769
0.7807
A. 57
Y1
-1.8094
CCDF of
PAPR
9.368
7.6194
4.7119
1.9277
9.1239
Z1
0.0852
-0.4767
-3.2966
-11.4059
0.0587
0.9287
SNR
(BER=
)
11.5448
12.1067
14.9266
23.0359
11.5713
Appendix A:
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
WHT
DCT
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
50
70
70
70
70
90
90
90
90
5
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
5.2030
11.7711
18.4830
1.9256
5.9356
12.1732
18.5564
2.8015
6.5904
12.4999
18.6257
4.7516
8.0905
13.3591
18.9242
4.2553
7.6258
12.9890
18.7335
5.1623
8.2282
13.2299
18.7585
5.9939
8.7922
13.4752
18.8058
4.0783
-2.9008
-1.3242
-0.2614
-6.4950
-4.0630
-1.7503
-0.0850
-6.3259
-4.7333
-1.9038
-0.4202
-5.6271
-3.9205
-1.1251
-0.0447
-6.9289
-4.9426
-2.0679
-0.3517
-7.3198
-4.8379
-1.9328
-0.1039
-7.1447
-4.9729
-2.2699
-0.3343
Tables of Results
17.6347
11.0666
4.3547
20.9121
16.9021
10.6645
4.2813
20.0362
16.2473
10.3378
4.2120
18.0861
14.7472
9.4786
3.9135
18.5824
15.2119
9.8487
4.1042
17.6754
14.6095
9.6078
4.0792
16.8438
14.0455
9.3625
4.0319
14.0327
2.5204
5.1833
8.0193
1.4846
3.0211
5.4977
8.1109
1.4262
3.0154
5.5439
8.1529
2.1352
3.5446
5.7738
8.2183
2.0956
3.4779
5.6726
8.1261
2.6547
3.8796
5.9123
8.1934
2.7271
3.9383
5.9566
8.2281
1.5751
A. 58
-1.0067
-0.4813
-0.0853
-2.1290
-1.3535
-0.4789
0.0983
-2.7184
-1.8692
-0.7607
-0.0657
-2.3524
-1.6600
-0.5108
-0.0078
-2.7290
-1.9347
-0.8200
-0.1105
-2.6579
-1.7250
-0.6263
-0.0062
-2.7575
-2.0163
-0.8240
-0.0765
7.3842
4.7213
1.8853
8.42
6.8835
4.4069
1.7937
8.4784
6.8892
4.3607
1.7517
7.7694
6.36
4.1308
1.6863
7.809
6.4267
4.232
1.7785
7.2499
6.025
3.9923
1.7112
7.1775
5.9663
3.948
1.6765
6.0889
-0.5679
-3.3075
-11.4272
-0.2059
-0.7820
-3.4292
-11.4118
-0.2365
-0.8880
-3.4371
-11.8373
0.0341
-0.8033
-3.5819
-11.7032
-0.1544
-0.9347
-3.9124
-11.5775
-0.4487
-1.1765
-4.1543
-11.6655
-0.6104
-1.3420
-4.0355
-12.0414
-0.4586
1.0021
0.8125
0.5928
1.7801
1.4260
1.3740
0.8582
2.8718
2.1820
2.0157
0.3127
4.0906
3.5667
2.4001
0.5536
4.7956
3.6953
1.8816
1.1925
7.9213
4.9185
2.3437
0.5845
17.7596
7.0280
3.3159
0.6886
12.1979
14.9375
23.0572
11.8359
12.412
15.0592
23.0418
11.8665
12.518
15.0671
23.4673
11.5959
12.4333
15.2119
23.3332
11.7844
12.5647
15.5424
23.2075
12.0787
12.8065
15.7843
23.2955
12.2404
12.972
15.6655
23.6714
12.0866
Appendix A:
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
DCT
5
5
5
10
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
50
70
70
70
70
90
90
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
5.8776
10.1251
14.7651
5.2700
7.5462
10.9668
14.9984
6.4419
8.8030
11.5302
15.1147
8.4747
9.6024
11.9149
14.9918
8.4754
9.4797
11.6821
15.0467
9.6001
9.8944
11.8863
15.0848
9.7691
10.4531
12.1597
15.2596
10.5879
11.1396
3.9010
4.1691
2.5982
0.9807
2.7480
3.5311
2.3334
1.2000
4.0740
3.0054
2.2379
0.6726
2.8234
2.1954
1.9246
0.8045
3.1426
2.0527
1.5561
0.7263
2.0137
2.1137
1.7237
1.1239
2.1760
2.1012
Tables of Results
12.2334
7.9859
3.3459
12.8410
10.5648
7.1442
3.1126
11.6691
9.3080
6.5808
2.9963
9.6363
8.5086
6.1961
3.1192
9.6356
8.6313
6.4289
3.0643
8.5109
8.2166
6.2247
3.0262
8.3419
7.6579
5.9513
2.8514
7.5231
6.9714
2.4715
4.2709
6.2418
2.0919
3.0042
4.5075
6.2912
2.7293
3.5713
4.7449
6.3263
3.3960
3.8900
4.9083
6.3167
3.4950
3.9296
4.8823
6.3449
3.8828
4.0978
4.9704
6.3745
4.0838
4.3864
5.1309
6.4068
4.3312
4.5770
A. 59
1.7424
1.7177
1.0835
0.4272
1.3563
1.4373
1.0089
0.5543
1.4920
1.2460
0.8443
0.3387
1.2480
0.9656
0.8383
0.3594
1.2988
0.9258
0.7184
0.3785
1.0118
1.0224
0.8329
0.4478
1.0872
0.8630
5.1925
3.3931
1.4222
5.5721
4.6598
3.1565
1.3728
4.9347
4.0927
2.9191
1.3377
4.268
3.774
2.7557
1.3473
4.169
3.7344
2.7817
1.3191
3.7812
3.5662
2.6936
1.2895
3.5802
3.2776
2.5331
1.2572
3.3328
3.087
-0.8859
-4.2414
-11.6890
-0.7392
-1.4409
-4.2249
-11.9552
-1.6394
-2.3940
-5.0332
-12.1126
-2.4175
-2.9251
-5.0536
-12.0250
-3.5951
-3.7052
-5.4937
-12.5112
-3.6787
-3.9254
-6.2297
-12.2911
-5.0376
-4.8662
-6.3668
-13.0394
-6.3649
-5.9901
0.1328
0.1311
-0.1029
0.0668
0.3486
-0.1840
-0.2280
0.1594
0.6928
0.1469
0.4012
0.1270
0.4634
0.6668
0.4903
-0.2524
1.2733
0.7066
-0.4337
0.4809
3.3344
1.2308
0.1332
-0.7874
12.0071
2.3819
12.5139
15.8694
23.317
12.3672
13.0689
15.8529
23.5832
13.2674
14.022
16.6612
23.7406
14.0455
14.5531
16.6816
23.653
15.2231
15.3332
17.1217
24.1392
15.3067
15.5534
17.8577
23.9191
16.6656
16.4942
17.9948
24.6674
17.9929
17.6181
Appendix A:
DCT
DCT
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
DST
90
90
5
5
5
5
10
10
10
10
15
15
15
15
20
20
20
20
30
30
30
30
40
40
40
40
50
50
50
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
12.5958
15.3170
8.5202
9.3524
11.4249
14.5077
9.8718
10.0125
11.7969
14.6244
10.4876
10.8296
12.3484
14.6892
10.9548
11.2190
12.3938
14.6429
11.6232
11.7005
12.5336
14.7947
11.9575
12.1610
12.9059
14.8824
11.9685
12.1433
12.8274
1.5774
0.9036
9.1489
7.2815
4.0744
1.2528
7.9070
6.5932
3.8431
1.3743
7.8686
5.7496
3.5027
1.1216
6.9516
5.5228
3.7945
1.2863
6.1571
4.9477
3.1433
Tables of Results
5.5152
2.7940
8.9447
8.1125
6.0400
2.9572
7.5931
7.4524
5.6680
2.8405
6.9773
6.6353
5.1165
2.7757
6.5101
6.2459
5.0711
2.8220
5.8417
5.7644
4.9313
2.6702
5.5074
5.3039
4.5590
2.5825
5.4964
5.3216
4.6375
5.2242
6.4156
3.6111
4.0014
4.8296
6.2632
4.1997
4.3860
5.0592
6.3016
4.5158
4.6424
5.2379
6.2822
4.6516
4.8452
5.2302
6.3268
4.9818
5.0121
5.3564
6.3350
5.0939
5.1844
5.5148
6.3855
5.1368
5.2135
5.5130
A. 60
0.6842
0.3516
3.9912
3.2405
1.7762
0.5786
3.4196
2.8522
1.6352
0.6958
3.2188
2.5091
1.4334
0.4980
2.9879
2.3614
1.6118
0.5410
2.6938
2.1825
1.4020
2.4398
1.2484
3.9119
3.5216
2.6934
1.2598
3.3233
3.137
2.4638
1.2214
3.0072
2.8806
2.2851
1.2408
2.8714
2.6778
2.2928
1.1962
2.5412
2.5109
2.1666
1.188
2.4291
2.3386
2.0082
1.1375
2.3862
2.3095
2.01
-6.7500
-12.6807
-2.6156
-3.4388
-5.9240
-13.2083
-5.1839
-5.3699
-7.6083
-13.0385
-7.2293
-6.8672
-8.0515
-13.6972
-8.9870
-8.1981
-8.7103
-13.8951
-15.0476
-11.0759
-9.2805
-13.9518
-18.3720
-13.6810
-9.9412
-13.8786
-18.3720
-17.2375
-10.5717
0.6034
0.0513
-4.3119
-3.7979
-3.4863
-1.0165
-6.9990
-5.9881
-3.9051
-1.6231
-11.9373
-8.0039
-3.8257
-1.7998
-14.3135
-9.3090
-3.9572
-1.6198
-13.4200
-12.6055
-4.7757
18.378
24.3087
14.2436
15.0668
17.552
24.8363
16.8119
16.9979
19.2363
24.6665
18.8573
18.4952
19.6795
25.3252
20.615
19.8261
20.3383
25.5231
26.6756
22.7039
20.9085
25.5798
>30
25.309
21.5692
25.5066
>30
28.8655
22.1997
Appendix A:
DST
DST
DST
DST
DST
DST
DST
DST
DST
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
50
70
70
70
70
90
90
90
90
5
5
5
5
10
10
10
10
20
20
20
20
30
30
30
30
40
40
40
40
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
14.8421
12.3682
12.5000
13.0636
14.7955
12.8291
12.9326
13.4032
14.9895
2.1510
2.3800
2.8097
3.4943
2.5046
2.6241
2.9457
3.5365
2.7535
2.8192
3.0364
3.5319
2.8996
2.9317
3.1032
3.5497
2.9635
3.0145
3.1604
3.5777
1.1297
5.2589
4.8067
3.2737
1.3059
5.0633
4.5403
3.0309
1.2222
12.2879
10.3993
5.7294
0.6711
10.2119
8.6996
4.9919
0.7695
9.6512
7.4870
4.5785
0.3828
8.4638
6.8825
4.5552
0.4878
Tables of Results
2.6228
5.0967
4.9649
4.4013
2.6694
4.6358
4.5323
4.0617
2.4754
4.8077
4.5787
4.1490
3.4644
4.4541
4.3346
4.0130
3.4222
4.2052
4.1395
3.9223
3.4268
4.0591
4.0270
3.8555
3.4090
3.9952
3.9442
3.7983
3.3810
6.3588
5.2927
5.3530
5.6055
6.4405
5.4648
5.5119
5.7247
6.4377
0.8595
0.9299
1.0939
1.3200
0.9697
1.0064
1.1058
1.2908
1.1068
1.1329
1.2124
1.3662
1.0915
1.0909
1.1471
1.2944
1.1013
1.1159
1.1686
1.3201
A. 61
0.5038
2.3617
2.1300
1.4485
0.6225
2.3618
1.9389
1.3257
0.5147
4.9342
4.0339
1.9958
-0.2592
4.0478
3.3129
1.7904
-0.0918
3.5015
2.7609
1.3971
-0.3696
3.1683
2.4659
1.4386
-0.3514
1.1642
2.2303
2.17
1.9175
1.0825
2.0582
2.0111
1.7983
1.0853
2.4905
2.4201
2.2561
2.03
2.3803
2.3436
2.2442
2.0592
2.2432
2.2171
2.1376
1.9838
2.2585
2.2591
2.2029
2.0556
2.2487
2.2341
2.1814
2.0299
-13.7840
-18.3720
-18.3720
-11.9018
-14.2162
-18.3720
-18.3720
-11.9831
-14.6467
-0.3995
-0.4631
-1.8985
-6.1537
-0.8641
-0.5591
-1.5443
-6.0508
-0.5381
-0.4235
-1.4934
-5.6755
-1.0229
-0.5152
-1.3132
-5.8967
-1.0149
-0.4055
-1.4070
-5.5958
-1.0120
-10.0000
-12.2750
-5.4018
-1.9642
0
-10.0000
-4.6297
-1.9147
0.0539
1.0589
2.6237
6.0172
1.4959
1.8325
3.3578
6.6425
2.1334
2.6028
4.1876
6.3013
3.0896
4.0125
4.6230
6.7090
25.412
>30
>30
23.5298
25.8442
>30
>30
23.6111
26.2747
11.9815
12.0451
13.4805
17.7357
12.4461
12.1411
13.1263
17.6328
12.1201
12.0055
13.0754
17.2575
12.6049
12.0972
12.8952
17.4787
12.5969
11.9875
12.989
17.1778
Appendix A:
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
DHT
50
50
50
50
70
70
70
70
90
90
90
90
1
.8
.5
.2
1
.8
.5
.2
1
.8
.5
.2
3.0338
3.0455
3.1724
3.5636
3.0908
3.1424
3.2468
3.6069
3.1589
3.1763
3.2661
3.6028
7.7286
6.3561
3.9945
0.3574
6.4877
5.9553
3.9631
0.6235
5.8993
5.2902
3.4000
0.3417
Tables of Results
3.9249
3.9132
3.7863
3.3951
3.8679
3.8163
3.7119
3.3518
3.7998
3.7824
3.6926
3.3559
1.1228
1.1738
1.2150
1.3427
1.1885
1.1830
1.2180
1.3394
1.1722
1.1776
1.2057
1.3118
A. 62
2.8528
2.3158
1.2770
-0.3393
2.4305
2.1330
1.2340
-0.3056
2.2422
1.7776
0.9797
-0.4382
2.2272
2.1762
2.135
2.0073
2.1615
2.167
2.132
2.0106
2.1778
2.1724
2.1443
2.0382
-0.9711
-0.5152
-1.3591
-6.0558
-0.8581
-0.6713
-1.2287
-5.5747
-1.1746
-0.6321
-1.3063
-5.9216
4.0269
4.1628
4.4829
6.7622
7.5599
5.4717
5.3173
6.7233
17.2434
7.7859
6.0931
6.8564
12.5531
12.0972
12.9411
17.6378
12.4401
12.2533
12.8107
17.1567
12.7566
12.2141
12.8883
17.5036
Appendix B
MATLAB Code
Appendix B
MATLAB Code
&& OFDM CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
% QPSK signal constellatio

% have 128 data point
% Sampling period of channel

% Max Doppler frequency shift
% -----------------% TRANSMITTER
% -----------------% Generate 1 x 128 vector of random data points
source = randsrc(1, 128*(nsym-2), 0:M-1); %the data source
cp_length = .25*fft_size; % length of cyclic prefix
sp = reshape(source , 128 , nsym-2);% s/p
s = size (sp);
tx = [];
for i=2:nsym-1
% QPSK modulation (mapping)
qpsk_mod = pskmod(sp(:,i-1), M);
% making s/p
serial_to_paralle = reshape(qpsk_mod, fft_size, 1); % s/p
% to know the start and end of cp
cp_start = fft_size-cp_length;
cp_end = fft_size;
% make ifft to each block and add CP
ifft_data_matrix = ifft(serial_to_paralle,fft_size);
% Compute and append Cyclic Prefix
for j=1:cp_length,
B. 1
Appendix B
MATLAB Code
cp(j,1) = ifft_data_matrix(j+cp_start,1);
end
% Append the CP to the existing block to create the actual OFDM block
ifft_data = vertcat(cp,ifft_data_matrix);
% s/p for transmission
[rows_ifft_data, cols_ifft_data]=size(ifft_data);
length_ofdm_data = rows_ifft_data*cols_ifft_data;
pilot = zeros(1,length_ofdm_data);
ofdm_signal = reshape(ifft_data, 1, length(ifft_data)); %p/s
Signal_Power = abs(ofdm_signal.^2);
Peak_Power = max(Signal_Power,[],2);
Mean_Power = mean(Signal_Power,2);
PAPR_Orignal(i) = 10*log10(Peak_Power./Mean_Power);
tx = [tx ofdm_signal];
end
[cdf0, PAPR0] = ecdf(PAPR_Orignal);
%-------------------------------------------------------------------------semilogy(PAPR0,1-cdf0,'-b')
legend('Orignal')
xlabel('PAPR0 [dB]');
ylabel('CCDF (Pr[PAPR>PAPR0])');
tt =[ pilot tx pilot];
x_abs=abs(tt);
papr=10*log(max(x_abs.^2)/mean(x_abs.^2))
% -----------% CHANNEL
% -----------% Create Rayleigh fading channel object.
% Frequency selective channel with 4 taps
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h = rayleighchan(Ts, Fd,tau,pdb);
h.StoreHistory = 0;
h.StorePathGains = 1;
h.ResetBeforeFiltering = 1;
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
B. 2
% Path delays
% Avg path
Appendix B
MATLAB Code
for ii=1:length(snr)
rx= [];
rx_signal_matrix = reshape(tt,length(ifft_data),nsym);
[~, c] = size(rx_signal_matrix);
for j = 2: nsym-1
hx = filter(h,rx_signal_matrix(:,j).');
% Pass through Rayleigh channel
a = h.PathGains;
AM = h.channelFilter.alphaMatrix;
g = a*AM;
% Channel coefficients
G(j,:) = fft(g,fft_size);
% DFT of channel coefficients
% Add AWGN no
rx_signal = awgn(hx,snr(ii),'measured');
d=size(rx_signal);
% Convert Data back to "parallel" form to perform FFT
con=reshape( rx_signal , length(ifft_data),1);
% Remove CP
con(1:cp_length,:)=[];
% Perform FFT
% FFT
fft_data_matrix = fft(con,fft_size);
% p/s
rx_serial_data = reshape(fft_data_matrix, 1,fft_size);
fftrx = rx_serial_data./G(j,:);
% Demodulate the data
qpsk_dem_data = pskdemod(fftrx,M);
rx =[rx qpsk_dem_data];
end
figure(2)
[no_of_error(ii),ratio(ii)]=biterr(source , rx); % error rate calculation
end
semilogy(snr,ratio,'--*r','linewidth',1);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
title('Bit error probability curve for qpsk using OFDM');
B. 3
Appendix B
MATLAB Code
&& RCF CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
CR = 3;
ITERATE_NUM = 4;
K = 128;
% SIZE OF OFDM Symbol
IF = 2;
% Interpolation factor
(Oversamplingfactor)
fft_size = K*IF;
% SIZE OF FFT
mm = 193
%when IF =1.125 =81 ;when IF =1.25 =97 ;when IF
=1.5 =129; when IF = 2 =193; when IF =3 =321; when IF = 4=449
% -----------------% TRANSMITTER
source = randsrc(1, K*(nsym-2), 0:M-1); %the data source
cp_length = .25*K; % length of cyclic prefix
sp = reshape(source , K , nsym-2);% s/p
s = size (sp);
tx = [];
PAPR_Orignal = zeros(1,nsym);
PAPR_RCF = zeros(ITERATE_NUM,nsym);
for i=2:nsym-1
% making s/p
serial_to_paralle = reshape(qpsk_mod, K , 1);% s/p
xy = [serial_to_paralle(1:K/2) ; zeros(fft_size-K,1); serial_to_paralle(K/2+1:K)];
ifft_data_matrix = ifft(xy,fft_size);
pilot = zeros(1,length(ifft_data_matrix));
B. 4
Appendix B
MATLAB Code
ofdm_signal = reshape(ifft_data_matrix, 1, length(ifft_data_matrix));%p/s

for nIter=1:ITERATE_NUM
% Clipping
x_tmp = ofdm_signal(Signal_Power>CR*Mean_Power);
x_tmp = sqrt(CR*Mean_Power)*x_tmp./abs(x_tmp);
ofdm_signal(Signal_Power>CR*Mean_Power) = x_tmp;
% Filtering
XX = fft(ofdm_signal,[],2);
XX(K/2+(1:fft_size-K)) = zeros(1,fft_size-K);
ofdm_signal = ifft(XX,[],2);
% PAPR Compute
PAPR_RCF(nIter,i) = 10*log10(Peak_Power./Mean_Power);
end
serial_to_paralle2 = reshape(ofdm_signal,fft_size , 1);% s/p
cp_end = fft_size;
for j=1:cp_length,
cp(j,1) =serial_to_paralle2(j+cp_start,1);
end
ifft_data_cp = vertcat(cp,serial_to_paralle2);
ofdm = reshape(ifft_data_cp, 1, length(ifft_data_cp));%p/s
pilot = zeros(1,length(ifft_data_cp));
tx = [tx ofdm];
end
[cdf1, PAPR1] = ecdf(PAPR_RCF(1,:));
%-------------------------------------------------------------------------B. 5
Appendix B
MATLAB Code
semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r',PAPR2,1-cdf2,'-g',PAPR3,1-cdf3,'c',PAPR4,1-cdf4,'-m')
legend('Orignal','One clip and filter','Two clip and filter','Three clip and filter','Four
clip and filter')
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
gains
h.StoreHistory = 0;
% Path delays
% Avg path power
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
rx_signal_matrix =reshape(tt,length( ifft_data_cp),nsym);
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
rx_signal = awgn(hx ,snr(ii),'measured');
con=reshape( rx_signal, length( ifft_data_cp),1);
%
Remove CP
B. 6
Appendix B
MATLAB Code
% Perform FFT
% FFT
du = [fft_data_matrix(1:K/2) ; [];fft_data_matrix(mm:fft_size)];
% p/s
rx_serial_data = reshape(du, 1,K);
end
figure(2)
end
ofdm_signal = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('SNR');
ylabel('BER')
&& RCF I=1 CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
CR = 2;
ITERATE_NUM = 4;
% -----------------B. 7
Appendix B
MATLAB Code
% TRANSMITTER
source = randsrc(1, fft_size*(nsym-2), 0:M-1); %the data source
sp = reshape(source , fft_size , nsym-2);% s/p
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
serial_to_paralle = reshape(qpsk_mod, fft_size , 1);% s/p

cp_end = fft_size;
ifft_data_matrix = ifft(serial_to_paralle ,fft_size);
% Clipping
% PAPR Compute
end
B. 8
Appendix B
MATLAB Code

cp_end = fft_size;
for j=1:cp_length,
end
tx = [tx ofdm];
end
figure(1)
%-------------------------------------------------------------------------semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r',PAPR2,1-cdf2,'-g',PAPR3,1-cdf3,'c',PAPR4,1-cdf4,'-m')
clip and filter')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
B. 9
% Path delays
% Avg path
Appendix B
MATLAB Code
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
% Remove CP
% Perform FFT
% FFT
% p/s
end
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
B. 10
Appendix B
MATLAB Code
grid on
xlabel('snr');
ylabel('BER')
&& RCF( I =pilot =76 in this case )CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
CR = 4;
ITERATE_NUM = 4;
K = 76;
% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
B. 11
Appendix B
MATLAB Code
% Clipping
% Filtering
% PAPR Compute
end
cp_end = fft_size;
for j=1:cp_length,
end
tx = [tx ofdm];
end
figure(1)
%-------------------------------------------------------------------------B. 12
Appendix B
MATLAB Code
clip and filter')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
B. 13
Appendix B
MATLAB Code
% Remove CP
% Perform FFT
% FFT
du = [fft_data_matrix(1:K/2) ; [];fft_data_matrix(91:fft_size)];
% p/s
end
% r = rx(1,(K+1:length(rx)-K));
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& RFC CODE :

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
CR = 3;


B. 14
Appendix B
MATLAB Code
ITERATE_NUM = 4;
K = 76;
% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
% Filtering
% Clipping
% PAPR Compute
B. 15
Appendix B
MATLAB Code
end
cp_end = fft_size;
for j=1:cp_length,
end
tx = [tx ofdm];
end
figure(1)
legend('I =pilot ','1.125','I =1.25 ','I =1.5 ','I= 2','I =3','I =4')
t = size (tx);
Q = size(tt);
figure(2)
plot(real(tt)); xlabel('Time'); ylabel('Amplitude');
title('OFDM Signal');grid on;
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
% Path delays
B. 16
Appendix B
MATLAB Code
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];

power gains
h.StoreHistory = 0;
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
% Remove CP
% Perform FFT
% FFT
% p/s
end
figure(2)
end
B. 17
Appendix B
MATLAB Code
v = size(rx);
semilogy(snr,ratio,'--*b','linewidth',1);
hold on;
axis([0 30 10^-4 1])
legend('I =pilot ','1.125','I =1.25 ','I =1.5 ','I= 2','I =3','I =4')
grid on
xlabel('snr');
ylabel('BER')
&& A_ law CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;


% -----------------% TRANSMITTER
%figure(1)
cp_length = .25*128 ; % length of cyclic prefix
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
B. 18
Appendix B
MATLAB Code

cp_end = fft_size;
for j=1:cp_length,
end
ofdm_signal = reshape(ifft_data, 1, length(ifft_data));%p/s
A = 90; % Parameter for A-law compander
V = max(abs(ofdm_signal));
compsig = compand(ofdm_signal,A,V,'A/compressor');
Signal_Power = abs(compsig.^2);
PAPR_Orignal1(i) = 10*log10(Peak_Power./Mean_Power);
tx = [tx compsig];
end
figure(1)
[cdf1, PAPR1] = ecdf(PAPR_Orignal1);
%-------------------------------------------------------------------------semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
B. 19
Appendix B
MATLAB Code
papr=10*log(max(x_abs.^2)/mean(x_abs.^2));
disp('PAPR of original signal in dB');
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
V = max(abs(rx_signal));
compsig = compand(rx_signal,A, V,'A/expander');
con=reshape( compsig, length(ifft_data),1);
% Remove CP
% Perform FFT
% FFT
% p/s
B. 20
Appendix B
MATLAB Code

end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&&
CODE
clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;


% -----------------% TRANSMITTER
s = size (sp);
tx = [];
B. 21
Appendix B
MATLAB Code
for i=2:nsym-1
%scatterplot(qpsk_mod);
%title('MODULATED TRANSMITTED DATA');
% making s/p
cp_end = fft_size;
for j=1:cp_length,
end
ofdm_signal = reshape(ifft_data, 1, length(ifft_data));%p/
Mu = 700; % Parameter for mu-law compander
x = compand(ofdm_signal,Mu,V,'mu/compressor');
Signal_Power = abs(x.^2);
tx = [tx x];
end
figure(1)
%-------------------------------------------------------------------------B. 22
Appendix B
MATLAB Code
semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r')
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
xx = compand(rx_signal,Mu,V,'mu/expander');
con=reshape(xx , length(ifft_data),1);
%
Remove CP
B. 23
Appendix B
MATLAB Code
% Perform FFT
% FFT
% p/s
end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& Rooting CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
y = .5;
% -----------------% TRANSMITTER
B. 24
Appendix B
MATLAB Code

s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
cp_end = fft_size;
for j=1:cp_length,
end
a =abs (ofdm_signal );
b =a.^y;
x= complex(b.*cos(angle(ofdm_signal )),b.*sin(angle(ofdm_signal )));
tx = [tx x];
end
%-------------------------------------------------------------------------B. 25
Appendix B
MATLAB Code
semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
aa =abs (rx_signal);
bb =aa.^(1/y);
xx= complex(bb.*cos(angle(rx_signal)),bb.*sin(angle(rx_signal)));
B. 26
Appendix B
MATLAB Code
% Remove CP
% Perform FFT
% FFT
% p/s
end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& NERF CODE :

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;


% -----------------B. 27
Appendix B
MATLAB Code
% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
cp_end = fft_size;
for j=1:cp_length,
end
b =erf(((abs(ofdm_signal)))./(sqrt(2).*std(ofdm_signal)));
f= (((2).*std(ofdm_signal).*b));
h= sign(ofdm_signal).*f;
Signal_Power = abs(h.^2);
B. 28
Appendix B
MATLAB Code
tx = [tx h];
end
t = size (tx);
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
B. 29
Appendix B
MATLAB Code
a =abs(rx_signal);
dd=erfinv((a)./(2).*std(rx_signal));
v =sqrt(2).*std(rx_signal);
s =(v.*dd);
ff=abs(s);
rr =sign(rx_signal).*ff;
con=reshape( rr , length(ifft_data),1);
% Remove CP
% Perform FFT
% FFT
% p/s
end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& AEXP CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;

B. 30
Appendix B
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
MATLAB Code

% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
cp_end = fft_size;
for j=1:cp_length,
end
B. 31
Appendix B
MATLAB Code
d =1.4;
a =var(abs(ofdm_signal))+ mean(abs(ofdm_signal));
b =exp(-((abs(ofdm_signal)).^2)./var(ofdm_signal));
c =(1-b).^2;
e =(c).^(d/2);
E =( a./mean(e)).^(d/2);
f= (E.*(1-b)).^(d/2);
h= sign(ofdm_signal).*f;
tx = [tx h];
end
t = size (tx);
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% --------------% RECEIVER
% -----------no_of_error=[];
B. 32
% Path delays
% Avg path
Appendix B
MATLAB Code
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
a =abs(rx_signal);
c =a.^(2/d);
aa =var(abs(rx_signal))+ mean(abs(rx_signal));
b =exp(-((abs(rx_signal)).^2)./var(rx_signal));
cc =(1-b).^2;
e =(cc).^(d/2);
E =( aa./mean(e)).^(d/2);
dd=log(1-(c./E));
v =var(rx_signal);
s =sqrt(-v.*dd);
ff=abs(s);

con=reshape( rr , length(ifft_data),1);
% Remove CP
% Perform FFT
% FFT
% p/s
end
figure(2)
end
x = size (source);
B. 33
Appendix B
MATLAB Code
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& COS CODE :

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;


% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
cp_end = fft_size;
B. 34
Appendix B
MATLAB Code
for j=1:cp_length,
end
a =abs (ofdm_signal );
b =sqrt(a);
x= complex(b.*cos(angle(ofdm_signal )),b.*sin(angle(ofdm_signal )));
tx = [tx x];
end
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
B. 35
Appendix B
MATLAB Code
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];

pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
aa =abs (rx_signal);
bb =aa.^2;
xx= complex(bb.*cos(angle(rx_signal)),bb.*sin(angle(rx_signal)));
% Remove CP
% Perform FFT
% FFT
% p/s
end
figure(2)
B. 36
Appendix B
MATLAB Code

end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
$$ tanhR CODE:
clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 10;
snr = 0:0.8:30;
k=5;
k1 =1;
y = 1;


% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
serial_to_paralle = reshape(qpsk_mod, fft_size, 1);% s/p
B. 37
Appendix B
MATLAB Code

cp_end = fft_size;
for j=1:cp_length,
end
x = k1.*tanh((((abs(ofdm_signal).*k).^(y)))).* sign(ofdm_signal);
tx = [tx x];
end
figure(1)
x_abs=abs(tt);
disp(papr);
% -----------B. 38
Appendix B
MATLAB Code
% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
xx = abs((atanh(abs(rx_signal)./(k))).^(1/y))./((k1).^(1/y)) .* sign(rx_signal);
% Remove CP
% Perform FFT
% FFT
% p/s
rx_serial_data = reshape( fft_data_matrix, 1,fft_size);
B. 39
Appendix B
MATLAB Code
end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& logR CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;


% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
%scatterplot(qpsk_mod);
%title('MODULATED TRANSMITTED DATA');
% making s/p
B. 40
Appendix B
MATLAB Code

cp_end = fft_size;
for j=1:cp_length,
end
k=10;
k1 =1;
y = .6;
x = k1.*log((((abs(ofdm_signal).*k).^(y))+1)).* sign(ofdm_signal);
tx = [tx x];
end
figure(1)
t = size (tx);
Q = size(tt);
B. 41
Appendix B
MATLAB Code
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
rx_signal = awgn(hx ,snr(ii),'measured');h;
xx = abs((exp(abs(rx_signal)./k)-1).^(1/y))./((k1).^(1/y)) .* sign(rx_signal);
% Remove CP
% Perform FFT
% FFT
% p/s
B. 42
Appendix B
MATLAB Code

end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& dht function CODE:

function X=dht(x)
N=length(x);
X=zeros(size(x));
i=sqrt(-1);
for k=0:N-1
for n=0:N-1
X(k+1)=X(k+1)+x(n+1)*(cos(2*pi*k*n/N)+sin(2*pi*k*n/N));
end
end
&&idht function CODE:
function x=idht(X)
N=length(X);
x=zeros(size(X));
i=sqrt(-1);
for k=0:N-1
for n=0:N-1,
x(k+1)=x(k+1)+X(n+1)*(cos(2*pi*k*n/N)+sin(2*pi*k*n/N));
end
end
x=x/N;
B. 43
Appendix B
MATLAB Code
&& precoding CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;


% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
serial_to_paralle = reshape(qpsk_mod, fft_size, 1);% s/p
ddg = dht(serial_to_paralle);
% ddg = fft(serial_to_paralle); % for DFT precoding
% ddg = dct(serial_to_paralle); % for DCT precoding
% ddg = dst(serial_to_paralle); % for DST precoding
% ddg = fwht(serial_to_paralle); % for WHT precoding
cp_end = fft_size;
ifft_data_matrix = ifft(ddg,fft_size);
for j=1:cp_length,
end
B. 44
Appendix B
MATLAB Code
tx = [tx ofdm_signal];
end
legend('Orignal')
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
B. 45
% Path delays
% Avg path
Appendix B
MATLAB Code
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
d=size(rx_signal);
con=reshape( rx_signal , length(ifft_data),1);
% Remove CP
% Perform FFT
% FFT
dg = idht(fft_data_matrix);
% dg = ifft(fft_data_matrix); % for DFT precoding
% ddg = idct(serial_to_paralle); % for DCT precoding
% ddg = idst(serial_to_paralle); % for DST precoding
% ddg = ifwht(serial_to_paralle); % for WHT precoding
% p/s
rx_serial_data = reshape(dg, 1,fft_size);
end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
B. 46
Appendix B
MATLAB Code
hybrid
&& precoding + RCF CODE:
clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
CR =4;
ITERATE_NUM = 4;
K = 128;
IF = 2;
% Interpolation factor (Oversampling factor)
fft_size = K*IF;
% SIZE OF FFT
mm=193;
%when IF =1.125 =81 ;when IF =1.25 =97 ;when IF =1.5 =129;
when IF = 2 =193; when IF =3 =321; when IF = 4=449
% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
B. 47
You can use

another type of
precoding
compnding
Appendix B
MATLAB Code
xy = [ddg(1:K/2) ; zeros(fft_size-K,1); ddg(K/2+1:K)];

% Clipping
% Filtering
% PAPR Compute
end
cp_end = fft_size;
for j=1:cp_length,
end
tx = [tx ofdm];
end
figure(1)
B. 48
Appendix B
MATLAB Code

clip and filter')
t = size (tx);
Q = size(tt);
figure(2)
plot(real(tt)); xlabel('Time'); ylabel('Amplitude');
title('OFDM Signal');grid on;
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;

B. 49
Appendix B
MATLAB Code
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
% Remove CP
% Perform FFT
% FFT
du = [fft_data_matrix(1:K/2) ; [];fft_data_matrix(mm:fft_size)];
dg = idht(du);
You can use
% ddg = ifft(du);
% for DFT precoding
another type of
% ddg = idct(du); % for DCT precoding
precoding
% ddg = idst(du); % for DST precoding
% ddg = ifwht(du); % for WHT precoding
compnding
% p/s
rx_serial_data = reshape(dg, 1,K);
end
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
B. 50
Appendix B
MATLAB Code
&& precoding +RCF I =1 CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 10;
snr = 0:0.8:30;
CR = 2;
ITERATE_NUM = 4;
% -----------------% TRANSMITTER
source = randsrc(1, fft_size*(nsym-2), 0:M-1); %the data source
sp = reshape(source , fft_size , nsym-2);% s/p
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
serial_to_paralle = reshape(qpsk_mod, fft_size , 1);% s/p
cp_end = fft_size;
B. 51
Appendix B
MATLAB Code

% Clipping
% PAPR Compute
end
cp_end = fft_size;
for j=1:cp_length,
end
tx = [tx ofdm];
end
figure(1)
%-------------------------------------------------------------------------B. 52
Appendix B
MATLAB Code
clip and filter')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
B. 53
Appendix B
MATLAB Code
% Remove CP
% Perform FFT
% FFT
% dg = idct(fft_data_matrix); % for DCT precoding
% dg = idst(fft_data_matrix); % for DST precoding
% dg = ifwht(fft_data_matrix); % for WHT precoding
% p/s
end
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& precoding + RCF( I =pilot =76 in this case )CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
B. 54
Appendix B
MATLAB Code
nsym = 1000;
snr = 0:0.8:30;
CR = 1.5;
ITERATE_NUM = 4;
K = 76;
% -----------------% TRANSMITTER
%figure(1)
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
xy = [ddg(1:K/2) ; zeros(fft_size-K,1); ddg(K/2+1:K)];
% Clipping
% Filtering
B. 55
Appendix B
MATLAB Code
% PAPR Compute
end
cp_end = fft_size;
for j=1:cp_length,
end
tx = [tx ofdm];
end
figure(7)
clip and filter')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
B. 56
Appendix B
MATLAB Code
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
% Remove CP
% Perform FFT
% FFT
dg = idht(du);
% ddg = ifft(du);
% for DFT precoding
% ddg = idct(du); % for DCT precoding
% ddg = idst(du); % for DST precoding
B. 57
Appendix B
MATLAB Code
% ddg = ifwht(du); % for WHT precoding
% p/s
rx_serial_data = reshape(dg, 1,K);
end
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& precoding + companding code:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 100;
snr = 0:0.8:30;
% -----------------% TRANSMITTER
cp_length = .25*128 ; % length of cyclic prefix
s = size (sp);
B. 58
Appendix B
MATLAB Code
tx = [];
for i=2:nsym-1
% making s/p
You can use
another type of
precoding
compnding
cp_end = fft_size;
for j=1:cp_length,
end
ofdm_signal = reshape(ifft_data, 1, length(ifft_data));%p/s
A = 10; % Parameter for A-law compander
compsig = compand(ofdm_signal,A,V,'A/compressor');
Signal_Power = abs(compsig.^2);
tx = [tx compsig];
B. 59
You can use

another type of
compnding
compnding
Appendix B
MATLAB Code
end
figure(1)
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
disp(papr);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
h.StoreHistory = 0;
% Path delays
% Avg path power gains
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
% Add AWGN no
B. 60
Appendix B
MATLAB Code

compsig = compand(rx_signal,A, V,'A/expander');
con=reshape( compsig, length(ifft_data),1);
You can use

another type of
compnding
compnding
% Remove CP
% Perform FFT
% FFT
% dg = idct(fft_data_matrix); % for DCT precoding
% dg = idst(fft_data_matrix); % for DST precoding
% dg = ifwht(fft_data_matrix); % for WHT precoding
You can use

another type of
precoding
compnding
% p/s
end
figure(2)
end
x = size (source);
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& RCF + companding CODE:

clear all
clc
close
% --------------% Parameters
B. 61
Appendix B
MATLAB Code
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
nsym = 1000;
snr = 0:0.8:30;
CR = 4;
ITERATE_NUM = 4;
K = 128;
IF = 2;
% Interpolation factor
(Oversampling factor)
fft_size = K*IF; % SIZE OF FFT
% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
% Clipping
B. 62
Appendix B
MATLAB Code
% Filtering
% PAPR Compute
end
cp_end = fft_size;
for j=1:cp_length,
end
k= 5;
k1 =1;
y = 1;
x = k1.*tanh((((abs(ofdm).*k).^(y)))).* sign(ofdm);
tx = [tx x];
end
figure(1)
B. 63
You can use

another type of
compnding
compnding
Appendix B
MATLAB Code

%-------------------------------------------------------------------------semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r',PAPR2,1-cdf2,'-g',PAPR3,1-cdf3,'c',PAPR4,1-cdf4,'-m',PAPR5,1-cdf5,'-k')
clip and filter','tanh')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
B. 64
Appendix B
MATLAB Code
xx = abs((atanh(abs(rx_signal)./(k))).^(1/y))./((k1).^(1/y)) .* sign(rx_signal);
con=reshape(xx , length(ifft_data_cp),1);
% Remove CP
% Perform FFT
You can use

another type of
compnding
compnding
% FFT
du = [fft_data_matrix(1:K/2) ; [];fft_data_matrix((fft_size/2+K/2)+1:fft_size)];
% p/s
end
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
&& RFC + companding CODE:

clear all
clc
close
% --------------% Parameters
% --------------M = 4;
fft_size = 128;
fspacing=15000;
fs=15000*128;
Ts = 1/fs;
Fd = 0;
B. 65
Appendix B
nsym = 1000;
snr = 0:0.8:30;
CR = 3;
ITERATE_NUM = 4;
K = 128;
IF = 2;
factor)
fft_size = K*IF; % SIZE OF FFT
d =.8;
MATLAB Code

% Interpolation factor (Oversampling
% -----------------% TRANSMITTER
s = size (sp);
tx = [];
for i=2:nsym-1
% making s/p
% Filtering
% Clipping
B. 66
Appendix B
MATLAB Code
% PAPR Compute
end
cp_end = fft_size;
for j=1:cp_length,
end
a =var(abs(ofdm))+ mean(abs(ofdm));
bb =cos(-((abs(ofdm)))./std(ofdm));
b =exp(-((abs(ofdm)))./std(ofdm));
c =(1-b).^2;
e =(c).^(d/2);
E1 =( a./mean(e)).^(d/2);
f= (E1.*(1-bb)).^(d/2);
h= sign(ofdm).*f;
You can use

another type of
compnding
compnding
tx = [tx h];
end
figure(1)
B. 67
Appendix B
MATLAB Code
%-------------------------------------------------------------------------semilogy(PAPR0,1-cdf0,'-b',PAPR1,1-cdf1,'-r',PAPR2,1-cdf2,'-g',PAPR3,1-cdf3,'c',PAPR4,1-cdf4,'-m',PAPR5,1-cdf5,'-k')
clip and filter','COS')
t = size (tx);
Q = size(tt);
x_abs=abs(tt);
% -----------% CHANNEL
tau = [0 .2e-9 .5e-9 1.6e-9 2.3e-9 5e-9];
pdb = [0.189 0.379 0.239 0.095 0.061 0.037];
power gains
h.StoreHistory = 0;
% Path delays
% Avg path
% --------------% RECEIVER
% -----------no_of_error=[];
ratio=[];
rx= [];
for j = 2: nsym-1
a = h.PathGains;
g = a*AM;
G(j,:) = fft(g,K);
% Add AWGN no
B. 68
Appendix B
MATLAB Code
a =abs(rx_signal);
c =a.^(2/d);
aa =var(abs(rx_signal))+ mean(abs(rx_signal));
b =exp(-((abs(rx_signal)).^2)./var(rx_signal));
cc =(1-b).^2;
e =(cc).^(d/2);
E =( aa./mean(e)).^(d/2);
dd=acos(1-(c)./E);
v =std(rx_signal);
s =(-v.*dd);
ff=abs(s);
You can use

another type of
compnding
compnding

con=reshape( rr , length(ifft_data_cp),1);
% Remove CP
% Perform FFT
% FFT
du = [fft_data_matrix(1:K/2) ; [];fft_data_matrix((fft_size/2+K/2)+1:fft_size)];
% p/s
end
figure(2)
end
v = size(rx);
hold on;
axis([0 30 10^-4 1])
legend('simulated')
grid on
xlabel('snr');
ylabel('BER')
B. 69

( )OFDM ,

OFDM .
( )ADSL / ()LANs
(.)LTE
) (PAPR OFDM
.OFDM PAPR

.
PAPR OFDM
( .)HPA

.
PAPR
( )RFC
( RFC .)RCF RCF I
.
PAPR
) .(BER BER CR = 4 I = 4
) (SNR )
( BER (.6
) ( )CCDF PAPR (.66
) PAPR ( .66).
CCDF of PAPR PAPR BER I = 4
.CR = 1.75 8.681( = PAPR)(=CCDF of PAPR
8.0187) )
( SNR at BER = ( .).
( )RFC companding
- .- compandings A
compandings A companding ))AEXP
CCDF of PAPR PAPR .. =d
6.1( = PAPR) 7.2405( = CCDF of PAPR )
)
( SNR at BER= ( .8-).
( )precoding
. precoding BER PAPR
) (DFT (.(WHT
B. 70
.PAPR :
RCF . WHT) precodings ) (DCT
( )DST (.))DHT
RCF . . ( compandings compandings -
.)compandings A-
RFC . . ( compandings compandings -
)compandings A-
. .)DHT DST DCT WHT) precodings ( compandings
compandings -)compandings A-
( RFC )AEXP CCDF of PAPR
.BER PAPR CCDF of PAPR, PAPR . = d
.CR = 4 .1( = PAPR)8.7178(= CCDF of PAPR,
) )
( .( = SNR at BER).
DHT ( )tanhR PAPR
CCDF of PAPR BER .
CCDF of PAPR, PAPR .DHT .8 = y = k
.66( = PAPR) 8.9691(=CCDF of PAPR )
)
( SNR at BER = ( .88-).
.
B. 71
:

( )
1436
2015
:

( )

An Overview: Peak To Average Power Ratio (PAPR) in OFDM System Using Some New PAPR Techniques (With Matlab Code)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Overview: Peak To Average Power Ratio (PAPR) in OFDM System Using Some New PAPR Techniques (With Matlab Code)

Uploaded by

Copyright:

Available Formats

An Overview : Peak to Average

Power Ratio (PAPR) in OFDM

An Overview : Peak to Average

Especially my grandfather Mr. Hassan Ali Zwain,

Zainab saad hadi

best one improvement in PAPR and CCDF of PAPR is at d= 1.1. The

1.2 Literature survey

Chapter Two: LTE and OFAM

2.2. LTE Requirements

2.3. LTE Architecture

2.4. Air interface in LTE

2.5 History of OFDM

2.6.1 Orthogonality of the subcarriers and OFDM

2.6.2 Guard Interval

2.6.3 One-tap equalizer

2.7 OFDM based Multiple Access

2.8 Orthogonal Frequency Division Multiple Access

Chapter Three: Peak To Average Power Ratio Reduction

3.2 PAPR of OFDM signal

3.3 Oversampling discrete OFDM symbols to find true (continuous) peaks

3.4 Distribution of PAPR

3.5 Identification of the Problem

3.5.1 Nonlinear HPA and DAC

3.5.2 Power Saving

3.6 Factors influencing the PAPR

3.6.1 The number of sub carriers

3.6.2 The order of Modulation

3.6.3 Constellation shape

3.6.4 Pulse Shaping

3.8 Fitness function-based approach for determining an appropriate Algorithm

Chapter Four: PAPR Reduction Techniques

4.1There are three different way to divide the PAPR

4.1.1The first way is

4.1.2 The second way

4.1.3The third way

4.1.4 And finally there is Hybrid techniques

4.2 Clipping and Filtering

4.3 Peak Windowing Method

4.4 Envelope Scaling

4.5 Peak Reduction Carrier

4.6 Companding Technique

4.6.1 Square-Rooting Companding Technique ( SQRT) for PAPR Reduction in

4.6.3 Trapezoidal power companding

4.6.4 Hyperbolic tangent (

4.6.5 Error Function (

4.6.6 Logarithm Function (log) Companding

4.7 Coding techniques

4.8 Selective Mapping (SLM)

4.9 Partial Transmit Sequence (PTS)

4.10 Tone Reservation

4.11 Tone Injection

4.13 Active Constellation Extension (ACE)

4.14 Dummy Sequence Insertion (DSI)

Chapter Five: Simulation Results and Analysis

5.1 OFDM System model

5.2 PAPR techniques used

5.2.1 Repeated clipping and frequency domain filtering (RCF)

5.2.2 Repeated frequency domain filtering and clipping RFC

5.2.3 The OFDM System with discrete time companding

5.2.3.3 Rooting Companding Technique (RCT)

5.2.3.4 New error function Companding (NERF)

5.2.3.5 Absolute Exponential companding (AEXP)

5.2.3.6 Cos companding