JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO.
4, FEBRUARY 15, 2014
Real-time Digital Signal Processing for Optical
OFDM-Based Future Optical Access Networks
Roger Giddings
(Invited Tutorial)
Abstract—In this paper, key aspects associated with realizing
real-time digital signal processing (DSP)-based optical transceivers
for use in future optical access networks are considered. The fun-
damental principles of optical orthogonal frequency division mul-
tiplexing (OOFDM)-based transceivers and their associated DSP
implementation are presented. This is followed by an extensive re-
view of the real-time DSP implemented in the end-to-end OOFDM
transceivers experimentally demonstrated over the past several
years.
Index Terms—Digital signal processing (DSP), optical access,
optical communication, orthogonal frequency division multiple
access (OFDMA), orthogonal frequency division multiplexing
(OFDM), passive optical network (PON).
I. INTRODUCTION
IGITAL signal processing (DSP) is widely exploited in the
D modern world to enable a vast array of high performance
services and devices that were unimaginable several years ago.
As a highly pervasive technology, DSP considerably enhances
everyday life by enabling applications ranging from antilock
breaking systems to satellite navigation and sophisticated med-
ical imaging. DSP has also been an enabler for many of the
highly successful communications technologies over the last
20 years [1]–[3].
It is only in recent years that advanced DSP has been utilized
in optical communications to realize commercial long-haul op-
tical systems in the form of DSP-enable coherent optical re-
ceivers [4], which not only offer high transmission capacities
of the order of 100 Gb/s per wavelength, but achieve ultrasen-
sitive receivers for radically increasing unrepeated transmission
distances.
To enable the wide use of DSP in other areas of optical com-
munications, there is growing interest in the exploitation of
DSP to solve the challenges facing the future optical access
networks. The passive optical network (PON) [5], [6] has been
widely adopted as one of the main fiber-to-the-home (FTTH)
solutions capable of meeting the low-cost demands of the ac-
cess networks. PON technologies are expected to deliver an
Manuscript received June 23, 2013; revised August 4, 2013 and August 31,
2013; accepted September 3, 2013. Date of publication September 16, 2013;
date of current version January 10, 2014.
The author is with the School of Electrical Engineering, Bangor University,
Bangor LL57 1UT, U.K. (e-mail: r.p.giddings@bangor.ac.uk).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/JLT.2013.2281628
0733-8724 © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
553
aggregate capacity of 40 Gb/s in the near future and the NG-
PON2 standardization work has addressed this by the decision
to adopt a time-division multiplexing/wavelength-division mul-
tiplexing (TDM/WDM) approach [7], this maintains the use of
conventional on-off keying (OOK) modulation with transmis-
sion speeds preserved at 10 Gb/s per wavelength. It is widely
accepted that in the long term the future generation PON tech-
nologies must exceed the 10 Gb/s per wavelength threshold to
further increase network capacity throughput. It is technically
highly challenging to achieve this with the conventional binary
OOK modulation. DSP-enabled PON technologies on the other
hand offer far greater flexibility in signal generation and decod-
ing allowing compensation of signal distortions and/or utiliza-
tion of advanced modulation formats which are inherently more
tolerant to the fiber distortion effects. DSP can also allow the
use of spectrally efficient modulation techniques, which means
increased network capacity can be achieved through efficient
exploitation of component bandwidths, thus great commercial
benefits may be attained if the established, mature optical com-
ponent sources for NG-PON1 and NG-PON2 can be exploited.
DSP can also enable adaptive modulation techniques which can
adapt to the varying spectral characteristics of the network due
to the natural variations in optical fiber, optical component, and
radio frequency (RF) component characteristics. DSP can also
enable important features such as dynamic bandwidth allocation
(DBA) to improve capacity utilization efficiency. DSP-enabled
optical access networks can thus potentially provide network ad-
ministrators with on-demand adaptability down to the physical
layer making the networks highly adaptable to the fluctuating
end-user service demands.
This paper presents a tutorial on real-time DSP for optical
access networks, first exploring general implementation as-
pects of any DSP-based optical transceiver and then provid-
ing an in-depth examination of real-time DSP for OOFDM-
based transceivers based on the world first real-time end-to-end
OOFDM transmission systems developed at Bangor Univer-
sity. The rest of this paper is organized as follows; Section II
discusses the basic DSP-based transceiver structure and key ele-
ments with a detailed discussion of digital-to-analog converters
(DAC) and analog-to-digital converters (ADC). In Section III,
the implementation of real-time DSP with field programmable
gate arrays (FPGAs) is presented covering issues such as
DAC/ADC interfacing and parallel processing and pipelining
techniques. Example state of the art FPGAs are also described.
In Section IV, the use of DSP in OOFDM-based access net-
works is discussed. The basic principles of the OFDM modu-
lation format are also covered, with the fundamental concepts
554
Fig. 1. System elements of DSP-based optical transceivers.
of adaptively modulated OOFDM explained. The section also
describes in detail the architecture and fundamental functions
of a DSP-based OOFDM transceiver. In Section V, the real-
time OOFDM transceiver implemented at Bangor University
is described with sections describing in detail the IFFT/FFT
function, channel estimation and equalization functions, and
symbol synchronization. Section VI presents the advantages of
the multiband OOFDM technique over the single-band OOFDM
technique when considering PON applications. Section VII con-
cludes the paper.
II. REAL-TIME DSP-BASED OPTICAL TRANSCEIVERS
A. Transceiver Structure and Key Elements
The basic structure of a DSP-based optical transceiver is
shown in Fig. 1. The key elements in the transmitter are: high-
speed digital logic for hardware-based DSP, high speed DAC for
conversion of the digital signal samples to an analog electrical
signal, a wideband RF section to amplify, filter and possibly up-
convert the signal onto an RF carrier, and finally an electrical-to-
optical (EO) converter that converts the analog electrical signal
into an optical signal for fiber launching. The key elements in the
receiver are: an optical-to-electrical (OE) converter to detect the
optical signal and convert to an electrical signal, a wideband RF
section to filter, amplify and possibly down-convert the signal,
and a high speed ADC to convert the analog electrical signal to
digital samples for processing by the high-speed digital logic.
The DSP functions must be implemented in digital hardware
due to the ultrahigh processing speeds necessary to support
the multi-Gb/s optical signals. It may be feasible to implement
some transceiver DSP functions in software which can oper-
ate by subsampling the received signal, (e.g., synchronization
functions). However, for the majority of DSP functions, it is
essential to employ digital hardware operating at clock speeds
of several 100 MHz to achieve sufficient processing through-
put. For prototyping real-time DSP hardware, FPGAs offer the
ideal solution due to their reprogrammability. This enables rapid
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014
evaluation, exploration, and optimization of the hardware-based
algorithms. The high cost and power consumption of FPGAs,
however, makes them inappropriate for the cost and power sensi-
tive PON applications. It is therefore necessary to employ cus-
tom designed application specific integrated circuits (ASICs)
for real-time DSP in commercial products. ASICs obviously
require significant capital investment for development but reap
the benefits of low costs associated with high volume mass pro-
duction of integrated circuits. ASICs also offer the advantage of
significant power reduction compared to FPGAs. The DAC and
ADC are highly critical components in the transceiver and are
discussed in more detail in Section II-B.
The transceiver structure in Fig. 1 is for operation with
baseband electrical signals. It is also possible to construct
a transceiver which employs modulation of a single RF up-
converted signal or multiple RF signals in a single or multi-
band architecture. Moreover, transceiver architectures can com-
bine baseband signals with RF up-converted signals. Generally
speaking, in a multiband transceiver, multiple DACs and ADCs
are required according to the number of subbands and also
depending on whether in-phase/quadrature (IQ) modulation is
used. As DACs/ADCs are critical components, it is greatly ben-
eficial if advanced DSP algorithms can be used to relax both
the requirements and the number of DACs/ADCs required by
a specific transmission system. A transceiver employing RF
signals obviously requires more complex RF sections, and is-
sues such as RF carrier phase and frequency offsets must be
addressed. Unless specifically stated, the baseband transceiver
will be considered throughout this paper. However, the multi-
band transceiver architecture is discussed in more detail in Sec-
tion VI where its advantages over the single-band baseband
transceiver are analyzed. It should be noted, however, that data
conveyed by all signals are generated and recovered at baseband
regardless of the transmission frequency band, such that the im-
plemented DSP functionality is similar for all sub-bands. It is
of course possible to use ultrawideband DACs and ADCs for
direct digital-to-RF conversion [8] thus eliminating the analog
RF front-ends, but this approach is, at least for the present time,
most likely too costly for application in cost-sensitive PONs.
Due to the cost sensitivity of the optical access network it is
necessary to employ low-cost optical front ends. For low cost
optics, the intensity-modulation direct-detection (IMDD) tech-
nique [9] is unrivalled. IMDD operates by either direct modula-
tion or external modulation of a laser source. Directly modulated
lasers (DML) offer the lowest cost solution. However, DMLs
suffer from the phenomenon of frequency chirp [10] which
can degrade transceiver performance compared with the almost
chirp-free external modulation scheme. For direct detection, a
photodiode or avalanche photodiode is employed which is a so-
called square-law detector as the electrical current generated is
proportional to the square of the optical field and therefore the
optical signal intensity. The photodiode is followed by a tran-
simpedance amplifier to convert the detected current to a voltage
for the following RF section. For ultralow cost IMDD optics,
a highly promising laser source is the vertical cavity surface
emitting laser (VCSEL) [11] as these lasers can be produced
at extremely low cost mainly due to the reduced manufacturing
processes involved.
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS 555

A key point to emphasize is that the utilization of DSP enables

the use of low-cost optical components in high performance
optical transceivers, as their associated characteristics of limited
bandwidth, higher signal distortions, and wider tolerances can
be compensated for by the DSP algorithms either directly or
indirectly through advanced modulation formats that are more
tolerant to the component deficiencies. For example, the high
spectral efficiency and variable signal spectrum of adaptively
modulated OFDM, is able to fully utilize the lower bandwidth,
and adapt to the varying frequency response, of low cost optical
components.
As with conventional optical transceivers, the downstream
Fig. 2. DAC/ADC sample rate versus line rate for different spectral
and upstream optical interfaces can either operate at the same efficiencies.
wavelength for a dual feeder fiber-based PON, or as is more
typical, at different wavelengths for operation with a single
feeder fiber-based PON. Other advanced optical transmission
schemes, such as wavelength remodulation [12] and a lightwave-
centralized architecture [13] can also be employed with DSP-
based optical transceivers.

B. DACs and ADCs

The DAC and ADC are highly critical components in DSP-
based optical transceivers. The required DAC/ADC basic char-
acteristics are: high sample rates of the order of several GS/s,
bit resolutions in the region of 8 bits (modulation format depen-
dent), high linearity and low noise. DAC/ADC aspects that can
have impact on transceiver performance include: quantization
noise due to the discrete signal levels, non-ideal linear behav-
ior which causes the effective number of bits (ENOB) to be
lower than the physical resolution, and the ENOB decreasing
with signal frequency. The full-scale of the DAC/ADC should
be utilized to minimize the effect of quantization noise, which
can necessitate automatic gain control (AGC) before the ADC.
DACs also typically have a characteristic roll-off in frequency
response due to the inherent sin(x)/x shaping due to the zero-
order-hold output format, as well as low pass filtering effects of
the on-chip analog front end. The sampling clock quality can
moreover affect performance due to clock jitter and frequency
offset. It should be emphasized here that DSP algorithms can be
exploited to mitigate some of the nonideal DAC/ADC properties
and/or relax the required DAC/ADC performance requirements.
The required DAC/ADC sampling rate for a given line rate
of R (bits/s) is dependent on the electrical spectral efficiency
E (b/s/Hz) of the adopted modulation format. The required sig-
nal bandwidth is B = R/E (Hz). Therefore assuming operation
over the entire Nyquist band and single-band transmission, the
required sampling rate is S = 2·B = 2·(R/E) (samples/s). Fig. 2
shows graphically the variation of sample rate against line rate
for different spectral efficiencies. It can be seen, for example,
that if the sampling rate is limited to 20 GS/s, a 40 Gb/s line rate
would require a modulation format with at least 4 b/s/Hz spec-
tral efficiency. Modulation formats with high spectral efficiency
are thus important to minimize DAC/ADC sample rates.
Fig. 3 shows the bit resolution and sample rates of some com-
mercial high speed DACs and ADCs currently available. The
trend in DAC/ADC sampling rates has shown a steady growth
over the last 5 years [14] and developments are generally led
Fig. 3. Bit resolutions and sample rates of commercially available
DAC/ADCs.
by the progress in high-end test equipment such as digital sam-
pling oscilloscopes (ADCs) and arbitrary waveform generators
(DACs). The DAC/ADC can contribute a significant portion of
the total power consumption in an optical transceiver [15], so
this is obviously a key area to be addressed in the development
of future DAC/ADC targeted at access network applications.
III. REAL-TIME DSP IMPLEMENTATION WITH FPGAS
A. FPGA Technology
State-of-the-art FPGAs are unrivalled as a development plat-
form for high-speed real-time signal processing. Modern FPGAs
support features such as:
r Vast array of logic elements;
r Ultra-high speed transceivers (10 s Gb/s);
r Huge resource of high speed IO (Gb/s);
r Embedded memory;
r Dedicated multiplier units;
r High-performance embedded DSP blocks;
r Embedded hard functions such as phase-locked loops
(PLLs);
r Soft microprocessor support.
To illustrate the performance available from high-end FPGAs,
Table 1 summarizes the features of Altera’s Stratix V family
of FPGAs [16] implemented in 28 nm complementary metal–
oxide–semiconductor (CMOS) technology. Not only are some
devices offering almost 1 million logic elements and hundreds
556
TABLE I
FEATURES OF ALTERA’S STRATIX V FPGA FAMILY [16]
Fig. 4. Parallel and pipelined processing.
or thousands of dedicated DSP blocks they also support an
immense digital interface bandwidth which is essential for in-
terfacing to the multi-GS/s DACs and ADCs. The huge digital
interface bandwidth is provided by the multi-Gb/s embedded
transceivers, which can offer bidirectional peak bandwidths of
over 1 Tb/s.
B. Parallelism and Pipelining for High Speed DSP
For the DSP-based optical transceiver, analog signal sample
rates are of the order of several GS/s, whereas the digital logic
can be clocked at speeds on the order of several 100 MHz.
To overcome this speed disparity parallelism and pipelining
techniques must be fully exploited to achieve the required pro-
cessing throughput. Fig. 4 shows the principle of the technique.
Incoming digital samples at multi-GS/s from the ADC are first
passed through a serial-to-parallel (S/P) converter which gener-
ates parallel samples at a reduced sample rate compatible with
the FPGA logic speed. In order to maintain the necessary sample
throughput, the parallel samples are processed simultaneously.
Furthermore, to maximize the clock speed of a digital logic func-
tion, it is partitioned into a series of sequential functions with
the intermediate samples stored by registers, this is known as a
pipelining. Each sequential function is then clocked simultane-
ously with its input samples taken from the registered outputs of
the previous function. The maximum achievable clock speed is
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014
Fig. 5. FPGA SERDES.
thus determined by the function with the longest propagation de-
lay, which is significantly shorter than that of the corresponding
nonpipelined function. Skillful partitioning of the higher level
function can thus enable maximization of a DSP function’s clock
frequency.
The sample throughput of a function (samples/s) is the prod-
uct of the number of parallel samples and the clock frequency.
Thus, if more (less) parallel samples are employed for a given
throughput, the necessary clock speed is reduced (increased).
The required logic resources are proportional to the number of
parallel samples. There can thus be a tradeoff between logic
resources and clock speed. As power consumption is a function
of clock speed, it is possible to tradeoff die area and power
consumption [17].
The pipelining approach may lead to an increase in total func-
tion propagation delay. However, the high clock speeds involved
mean that this is unlikely to cause unacceptable processing la-
tency. Pipelining a function increases clocking frequency, and
as new output samples are generated on every clock cycle, the
overall function throughput is significantly increased.
C. FPGA Interfacing to DACs and ADCs
The multi-GS/s DACs and ADCs employed mean that ultra-
high bandwidth digital interfaces between the FPGA and the
DAC/ADC are necessary. A 56 GS/s, 8-bit converter requires a
digital bandwidth of 448 Gb/s for example. As previously indi-
cated, FPGAs offer large resources of high speed input/output
(I/O) supporting speeds in the order of 1 Gb/s, as well as high
speed digital transceivers supporting speeds of several 10 s Gb/s,
as illustrated in Table I. As the FPGA logic cannot operate at
these speeds, high speed serializer and deserializer (SERDES)
circuits implemented in dedicated circuitry are employed in the
FPGA, as illustrated in Fig. 5. The SERDES circuits typically
have a programmable range of parallelization ratios thus per-
mitting the logic array to operate with parallel data at a suitable
clock frequency, which is a sub-multiple of the interface clock
frequency. Due to the high frequencies involved, the logic sig-
nals at the digital interface are typically differential logic carried
by controlled impedance signal pairs, thus requiring impedance
matched interconnections. An example common interface logic
standard is low voltage differential signaling (LVDS), which has
an impedance of 100 Ω and logic levels of ±350 mV. As the
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS
Fig. 6. FPGA to ADC digital interface example.
digital transceivers operate at higher frequencies than the I/O,
they also incorporate embedded circuits to ensure high signal
integrity. These include clock data recovery (CDR) and pro-
grammable equalization in the individual deserializer inputs,
and programmable preemphasis at the individual serializer out-
puts. To programme the equalization and preemphasis features,
characterization of the interconnection is of course necessary.
The SERDES can also have arbitrary phase offsets at power up,
so it can be necessary to synchronize all SERDES when initial-
izing the system. Test pattern generation by the ADC may thus
be necessary in order to correctly synchronize the deserializers.
Fig. 6 shows an example interface between an FPGA and a
10 GS/s, 8-bit, 4 port ADC. The interface consists of 32 signals
operating at 2.5 GHz. 32 × 10:1deserializers are used in the
FPGA to give 320 parallel signals at 250 MHz.
IV. DSP IN OOFDM-BASED OPTICAL ACCESS NETWORKS
A. OFDM Modulation
OFDM is a multicarrier modulation (MCM) technique first
proposed in the 1960s [18]–[20] but at that time its implementa-
tion was impractical. Salz and Weinstein [21] first proposed the
use of the discreet Fourier transform (DFT) [22] for the gener-
ation of OFDM signals in 1969. It was not until semiconductor
electronics achieved sufficient processing power, however, that
implementation of OFDM with the DFT was feasible. Today the
OFDM modulation technique is widely adopted in numerous
communication standards such as digital subscriber line (DSL)
and its many variants, wireless local area networks (WLAN)
and digital audio and video broadcast (DAB, DVB). OFDM is
now widely recognized as a potential modulation technique for
application in future optical access networks [14], [23]–[28].
In [14], [23] authors provide an extensive coverage of OFDM in
optical communications. The following sections, of this paper,
provide an overview of the key principles of OFDM and adap-
tive OFDM, pertinent to the real-time OOFDM implementation
presented in this paper.
To illustrate the principle of the MCM method, Fig. 7 shows a
block diagram of a generic multicarrier transmission system. In
the transmitter, the incoming binary data stream is converted to
a number of lower bit rate binary data streams, with the separate
557
Fig. 7. Block diagram of a generic multicarrier transmission system.
data streams encoded onto multiple carrier frequencies, also
known as subcarriers, and subsequently summed together for
transmission over the channel.
At the receiver, the same carrier frequencies are used to ex-
tract the separate signals from the received signal containing all
subcarriers. The low pass filters remove the unwanted products
generated from other subcarriers before the signal is decoded to
recover the binary data stream. The multiple binary data streams
are then combined to form a single high speed bit stream. It is
extremely difficult to implement the subcarrier generation and
detection functions with discrete RF components even with a
limited number of subcarriers. However, DSP is highly suited
to performing these functions, even when the number of sub-
carriers is of the order of several hundreds or even approaches
1000s [29].
OFDM is a special case of multicarrier modulation where
there is a harmonic relationship between subcarrier frequencies,
the frequency spacing Δf is defined as
1
Δf = (1)
Tb
where Tb is the symbol period. Symbols are defined as the fixed
length time intervals into which the OFDM signal is divided for
transmission. Data are encoded onto the subcarriers in each sym-
bol using both amplitude and phase modulation such as M-ary
QAM modulation [30], where M is the number of distinct com-
binations of subcarrier phase and amplitude. An M -ary QAM
signal thus encodes b bits were b = log2 (M ). For example, if
there are 64 subcarriers each encoded with 16-QAM the number
of bits carried by each symbol is 64 × (log2 16) = 256. As an
OFDM symbol encodes multiple bits its period is significantly
longer than a single bit period in a conventional OOK modulated
signal for the same bit rate. This makes OFDM more tolerant to
dispersive channels such as wireless channels and optical fibers,
as the extended symbol period greatly reduces the impact of the
dispersive channel’s pulse spreading effect.
The frequency, fk , of the kth subcarrier is thus
k
fk = f0 + k = 1, 2, 3, . . . , N s (2)
Tb
558
Fig. 8. Principle of OFDM signal generation in the time domain.
where f0 is a fixed frequency offset and NS is the total number
of subcarriers. f0 is 0 for a baseband system and represents a
frequency offset due to the frequency up-shifting in a carrier-
based system. Fig. 8 illustrates the OFDM signal generation
principle in the time domain for the case of 4 subcarriers over
one symbol period.
The spreading effect associated with a dispersive channel
causes adjacent received symbols to overlap, a phenomenon
known as intersymbol-interference (ISI). To further improve the
dispersion tolerance of OFDM, an intersymbol gap can be in-
serted between two adjacent symbols to avoid the ISI occurring
in the wanted signal region. To ensure the temporal spreading
of the signal in the intersymbol gap does not distort the wanted
signal region, each subcarrier is simply extended into the inter-
symbol gap [23]. As the subcarriers are all cyclic the simplest
way to achieve this is to take an appropriate portion from the
end of the symbol and prefix it to the front of the symbol, this
is thus known as a cyclic prefix (CP). The CP causes a trans-
mission overhead and so reduces the net bit rate, the length of
the CP should thus be only as long as necessary to eliminate ISI
from the wanted signal region of the symbol and to also provide
sufficient system operation robustness.
In addition to dispersion tolerance, OFDM has the extremely
beneficial characteristic of high spectral efficiency due to the
subcarrier orthogonality property allowing the subcarriers to
overlap in the frequency domain without interference. To un-
derstand the spectrum of an OFDM signal the general form of
the signal s(t) should be considered
 
+∞ N SC
 
s (t) = Rk ,m ej θ k , m Υ (t) ej 2π f k (t−m T b )
m =−∞ k =1
where k and m denote the mth symbol and the kth subcarrier
and Rk ,m and θk ,m are the amplitude and phase respectively of
the encoded data. Υ (t)is a rectangular pulse shaping waveform
which ensures the mth encoded data is zero outside the mth
symbol, Υ (t)is given by
 a particular subcarrier can be varied to change the minimum
1, (0 < t ≤ Tb )
Υ (t) = (4)
0, (t ≤ 0, t > Tb ) .
A subcarrier within each symbol can thus be considered as the
product of a modulated yet continuous sine wave and the rectan-
gular waveform Υ (t) . The spectrum of Υ (t) is the sinc function
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014
Fig. 9. OFDM signal spectrum showing four orthogonal subcarriers.
which has the form of sin(x)/x, multiplication by the continu-
ous sine wave causes an up-shifting of the sinc function to be
symmetrical about the sine wave frequency. This is illustrated
by the four subcarrier spectra shown in Fig. 9, each is a sinc
function centered at the corresponding subcarrier frequency.
An important property of the sinc function is the zero-crossing
points which are spaced at the subcarrier spacing frequency Δf ,
this means that although the subcarrier spectra overlap at any
subcarrier frequency all other subcarriers are zero valued. As
the OFDM signal is processed digitally using the DFT only the
discrete subcarrier frequencies are considered, at these discrete
frequencies only one subcarrier is nonzero thus the individual
subcarriers can spectrally overlap without intersubcarrier inter-
ference. This spectral overlapping thus makes the OFDM signal
very spectrally efficient.
B. Adaptively Modulated OFDM
An important characteristic of OFDM is the ability to modu-
late each subcarrier independently [31], [32] which allows the
signal to adapt to the spectral characteristics of the complete
transmission channel which includes the fiber and transceiver
components. As for any modulated signal its bit error rate (BER)
performance is dependent on the received signal-to-noise ratio
(SNR) thus for a desired BER there is a corresponding mini-
mum SNR requirement. The minimum SNR will be modulation
format dependent, as the number of encoded bits increases the
signal’s tolerance to noise and distortion decreases thus the min-
imum required SNR will increase. Each OFDM subcarrier can
experience different noise and distortion due to their frequency
(3)
dependent nature thus SNR is subcarrier frequency dependent.
There are two basic methods to ensure the minimum SNR is
achieved for a specific subcarrier. First, for a fixed modulation
format an individual subcarrier’s transmitted power level can be
adjusted to achieve the minimum SNR at the receiver. Second, if
the transmitted subcarrier power is fixed the SNR at the receiver
cannot be adjusted, however the modulation format adopted on
required SNR to be below but as near as possible to the actual
SNR.
To illustrate the adaptive loading principle we consider a
channel where noise and distortion are similar at all subcar-
rier frequencies and the received subcarrier SNRs are only
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS 559

Fig. 11. Functional DSP architecture of an OOFDM transceiver employing

baseband transmission of real-valued signals.

C. DSP-Based Optical OFDM

Fig. 10. (a) Power loading and (b) bit loading of OFDM subcarriers.
dependent on the channel loss which is frequency dependent.
Fig. 10(a) shows how subcarrier power loading is employed
Fig. 11 illustrates the basic functional architecture of the DSP
employed within an optical OFDM transceiver which operates
with real-valued baseband signals. The DSP architecture con-
sists of various functional blocks. However, the core functions
at the heart of the transceiver are the inverse DFT (IDFT) and
DFT. The IDFT in the transmitter converts a series of N complex
coefficients, Xk , representing the amplitude and phase of the
discrete subcarrier frequencies, fk , (both positive and negative)
to a series of N coefficients xn representing the corresponding
discrete complex time domain samples. The IDFT is defined as
follows:
2 −1
N

at the transmitter to compensate for the channel frequency re- 1  2π N N

xn = Xk ej N k n n=− , . . . , 0, . . . , − 1.
sponse and ensure a constant SNR across all subcarriers at the N −N
2 2
k= 2
receiver. Fig. 10(b) illustrates how subcarrier bit loading can be
(5)
employed for the same case. The transmitted modulation format
is selected to ensure the minimum required SNR is aligned to the
The IDFT thus represents a bank of IQ modulators each mod-
actual subcarrier SNR resulting from the frequency dependent
ulating one of the Δf spaced subcarriers. Where Δf is as de-
channel loss. It is important to note that power loading alone
fined in (1).
allows the optimization of BER across subcarriers at a fixed bit
The DFT in the receiver converts a series of complex time do-
rate whereas bit loading provides the added flexibility of con-
main coefficients, xn , representing the OFDM symbol samples
trolling aggregated bit rate and can thus maximize the bit rate
into the corresponding complex frequency domain coefficients,
for a given channel while meeting the required BER target. Bit
Xk , and is defined as follows
loading also allows dynamic bandwidth redistribution between
N
−1
subcarriers enabling increased control of bandwidth allocation 2 −2 π N N
to end users. It is of course possible to employ a combination of Xk = xn ej N kn
k=− , . . . , 0, . . . , − 1.
2 2
both power and bit loading [33] for optimum flexibility. n = −N
2
It is important to consider the impact on DSP complexity of (6)
the three possible adaptive loading techniques of power loading The DFT thus represents a bank of IQ demodulators each
(PL), bit loading (BL), and combined bit and power loading demodulating one of the Δf spaced subcarriers.
(BPL). Implementing BL or BPL requires significantly more The signal flow through the transmitter DSP blocks is as fol-
complex DSP compared to PL due to the additional logic re- lows: Incoming serial binary data are serial-to-parallel converted
source required to implement the encoders and decoders sup- to a sequence of parallel data words, each word is combined
porting reconfigurable modulation formats [34]. with pilot data for use in channel estimation and subdivided
560
into (N /2)-1 data words representing the data to be encoded
on each of the (N/2)-1 subcarriers. A bank of encoders then
convert each data word to a complex number according to the
modulation format used by the encoder, where the complex
number represents the subcarrier’s frequency domain amplitude
and phase. It should be highlighted that the encoders can be
of a fixed type or adaptive encoders can be employed to al-
low the selection of different modulation formats (bit loading)
and/or power levels (power loading). To ensure the generated
time domain signal is real valued only for use with the IMDD-
based optics, the frequency coefficients at the IDFT inputs are
arranged with Hermitian symmetry [22] such that the encoded
complex data are applied to the positive (N /2)-1 frequency bins
and the complex-conjugates are applied to the corresponding
(N /2)-1 negative frequency bins, such that
 
∗ N
X−k = Xk k = 1, 2, ..., −1 (7)
2
and X0 = X−N /2 = 0 as the subcarrier at zero frequency can-
not carry data. Also, due to the need for symmetry, there is no
corresponding positive frequency for X−N /2 . The IDFT thus
converts the N frequency-domain coefficients to N time do-
main samples which represent one OFDM symbol. Due to the
large peak-to-average-power-ratio (PAPR) of an OFDM signal,
the time domain signal is clipped to reduce the PAPR and is also
quantized to the number of bits of resolution required by the
DAC. Although clipping distorts the signal, it increases the av-
erage signal power, this reduces the effect of both quantization
noise and independent analog noise sources, such as thermal
noise, due to the resultant increase in SNR at the receiver. There
is however an optimum clipping level [35] which maximizes
system performance. A cyclic prefix is then added by prefixing
η · N samples at the start of the symbol, where η is the cyclic
prefix parameter to create a symbol of (η+1)·N samples, where
η is typically <1. The parallel samples of the whole symbol
are then passed to the DAC typically via a number of serializer
circuits as discussed in Section III-C.
The signal flow through the receiver DSP blocks is essen-
tially the reverse of that in the transmitter, with the addition
of symbol alignment and channel estimation and equalization
functions. The data from the ADC passes through a number of
deserializer circuits to provide parallel signal samples typically
corresponding in length to one symbol period, but arbitrarily
located and so not aligned to the symbol boundaries. A symbol
alignment block is employed to detect the symbol offset so that
it can be compensated for in the symbol offset adjustment block
and achieve symbol realignment. The cyclic prefix can then be
removed as it has no useful data. The remaining N time domain
samples are then processed by the DFT block to recover the N
frequency domain coefficients from which (N /2)-1coefficients
are selected from the positive frequency bins. At the output of the
DFT, a channel estimation function must detect the location of
the subcarriers carrying pilot data. From these pilot-subcarriers,
the channel can be estimated based on the known transmitted
pilot-subcarriers. Once the channel transfer function (CTF) is
known for each subcarrier frequency, it can be used to equalize
the corresponding received subcarriers by a simple multiplica-
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014
Fig. 12. Basic DSP architecture for an OOFDM transceiver employing an IQ
modulated RF carrier.
tion by the inverse CTF at the given frequency. This is commonly
referred to as single-tap equalization. The equalized subcarrier
frequency domain coefficients are then fed to a bank of decoders
corresponding to the encoders used in the transmitter to recover
the transmitted data words. The data words are recombined and
the pilot data removed to form a single parallel data word con-
taining all data bits transmitted in one symbol. The parallel word
is then passed to a parallel-to-serial converter for output at the
transceiver’s digital data output. It should be noted that the dig-
ital data interfaces of the transceiver would in reality typically
consist of multiple serial interfaces, as the transceiver would
effectively multiplex several lower bit rate signals into one high
speed optical signal. For example, the transceiver may have 4
× 10 Gb/s Ethernet ports which are multiplexed into a 40 Gb/s
optical signal.
As a comparison, the DSP functional architecture for an
OOFDM transceiver employing an IQ modulated signal is
shown in Fig. 12. It can be seen that the DSP is essentially
the same as that employed in the transceiver employing base-
band transmission. The main difference is that N -1 subcarriers
can be used to carry data (zero frequency cannot be used) and
the time domain signal is now a complex value. The phase offset
between the received carrier and the receiver’s local oscillator
(LO) results in a phase rotation of the I and Q components
which must be detected and corrected. Any frequency offset
between the transmitter’s carrier frequency and the receiver’s
LO must also be accommodated, although this is not explicitly
shown in Fig. 12.
V. REAL-TIME DSP FOR OPTICAL OFDM
Since 2009, there have been a number of real-time OOFDM
demonstrations world-wide. Nearly all demonstrations are tar-
geted at coherent long-haul systems [29], [36]–[44]. A real-time
OOFDM receiver targeted at PON applications was also demon-
strated by NEC [45].
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS
Fig. 13. DSP architecture of the adaptively modulated OOFDM transceiver.
TABLE II
KEY SYSTEM REAL-TIME OOFDM TRANSCEIVER PARAMETERS
Bangor University’s Optical Communications Research
Group were the first to demonstrate real-time end-to-end IMDD-
based OOFDM transmission over SMF and MMF fibers in a se-
ries of world first real-time OOFDM transmission experiments
[35], [46]–[53]. More recently a group at ETRI also demon-
strated a real-time end-to-end OOFDM system [54] targeted at
PON applications. This section gives a detailed overview of the
real-time DSP employed in Bangor’s OOFDM transceivers.
A. Implementation
The DSP architecture of Bangor’s most recent real-time
OOFDM transceiver design, based on Altera’s Stratix II GX
FPGAs, is shown in Fig. 13 and key transceiver parameters are
presented in Table II.
The architecture is similar to the functional DSP architecture
shown in Fig. 11. The key differences are discussed here. In the
transmitter, a data generator block is implemented to provide
parallel pseudorandom binary data for transmission, 15 adap-
tive modulators are employed each of which can perform either
16, 32, 64, or 128-QAM encoding by selecting one of four
distinct encoders [34], and a power loading block allows live
adjustment of individual subcarrier power. The 32 point IDFT
is implemented with an inverse fast Fourier transform (IFFT),
561
as this is a resource efficient form of the IDFT as discussed
in Section V-B. The clipping and quantization block have an
online adjustable clipping level for live optimization. After the
cyclic prefix is added the signed digital samples are converted to
unsigned samples required by the DAC. This block also inserts
a low power synchronization signal for symbol alignment as de-
scribed in Section V-D. The samples and bits are then correctly
organized for interfacing to 32×10:1 serializers feeding 32 I/O
operating at 1 GHz. The interface to the DAC thus consists of
4 × 8 bit ports such that 4 digital samples are transferred in
parallel to the 4 GS/s DAC. All online controlled parameters
are controlled via embedded memory accessed via the FPGA’s
Joint Test Action Group (JTAG) [55] interface.
In the receiver, the ADC interface is the reverse of the DAC
interface such that 32 1 GHz I/O feed 32×10:1 deserializers.
A reorganization block restructures the parallel samples cor-
rectly. These samples are used by the symbol offset detection
block to determine the arbitrary sample offset from the symbol
boundaries (described in Section V-D). The incoming samples
are thus realigned to the symbol boundary and the cyclic prefix
removed to provide 32 real-valued time domain samples cor-
responding to one OFDM symbol. The 32 real-valued samples
are converted to 32 signed complex samples by removing the
DC offset added by the ADC and setting all imaginary compo-
nents to zero. The 32 complex time domain samples are then
fed to a 32 point FFT. The FFT is used as it provides a highly
efficient implementation of the DFT. From the 32 complex fre-
quency domain coefficients output from the FFT, only the 15
positive frequencies are selected. These correspond to the 15
data-encoded frequency domain subcarriers. The pilot detec-
tion block operates on the 15 frequency domain subcarriers to
detect the pilot subcarriers which are used by the channel es-
timation block to determine the CTF as described in Section
V-C. The 15 subcarriers will subsequently be equalized using
the estimated CTF. The encoded binary data are then decoded
using adaptive decoders with modulation formats selected to
correspond to those used in the transmitter. Finally, the total
channel BER and individual subcarrier BERs are calculated
in a BER analyzer block in combination with a parallel pseu-
dorandom binary data generator identical to that used in the
transmitter.
The real-time adaptively modulated OOFDM transceiver has
been used to achieve the following world first demonstrations
involving real-time end-to-end OOFDM transmission in both
SMF and MMF links by fully exploiting its ability to dynami-
cally adapt to the system spectral characteristics.
r Aggregated upstream baseband OOFDM transmission at
11.25 Gb/s in an OFDMA PON [56].
r Colourless OLT-seeded REAM-based baseband OOFDM
upstream transmission at 10 Gb/s over bidirectional 25 km
SSMF [57].
r 19.125 Gb/s dual-band OOFDM transmission over 25 km
SSMF employing an EML as intensity modulator [51].
r 19.25 Gb/s dual-band OOFDM colourless transmission
over 25 km SSMF employing a REAM as intensity modu-
lator [52].
r 20 Gb/s dual-band OOFDM transmission over 500 m OM2
MMF [58].
562 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014

r 25.25 Gb/s triple-sub-band OOFDM transmission over

300 m OM2 MMF.
r Triple subband OOFDM transceivers operation at 30 Gb/s
[53].
Fig. 14. Radix-2 decimation in time butterfly element.

B. IFFT and FFT
To explicitly compute xn and Xk from the definitions of an N
point IDFT and DFT, as given in (5) and (6) respectively, would
require N 2 complex multiplications and N 2 –N complex addi-
tions. For a hardware-based implementation of the transforms,
it is highly advantageous to minimize computational complex-
ity in order to minimize design complexity. Furthermore, the
extremely high IDFT/DFT real-time computational throughput
inherent to OOFDM implies that a highly parallel and pipelined
architecture is necessary. This makes it difficult to reuse complex
functions for more than one calculation during one transform
radix values that can be employed are therefore dependent of
the required value of N . To allow more flexibility in the value
of N , it is also possible to create mixed radix FFTs [60].
A detailed examination of the conversion of one N -point DFT
into two N /2-point DFTs will be presented as this also explains
the origin of the fundamental building block of the FFT: the
butterfly operator. If the original time domain sequence xn is
split into its even and odd sequences of yn = x2n and zn =
x2n +1 , respectively, for n = 0,1,..,(N /2)–1 and substituted into
the DFT as defined in (6), this becomes
−1
N

cycle. Therefore, minimizing the number of discrete instances 

2
 (2n +1)k
2n k
Xk = yn ωN + zn ω N (8)
of complex functions in the algorithm is vitally important if
n =0
chip cost and power consumption targets are to be met. The
fast Fourier transform (FFT) and inverse FFT (IFFT) are highly for k = 0,1,..(N –1) and ωN = e−j 2π /N . (8) can be rewritten
computationally efficient algorithms for computing the DFT nk
using the relation ωN 2n k = ωN /2 as
and IDFT, respectively. The FFT was first introduced by Tukey
−1
N
−1
N
and Cooley in 1965 in their seminal paper “An algorithm for 
2 
2
nk k nk
the Machine Calculation of Complex Fourier Series” [59]. The Xk = yn ωN /2 + ωN zn ω N /2 . (9)
n =0 n =0
drastic reduction in computational complexity offered by the
FFT and IFFT makes them highly appropriate for implemen- The original DFT has now been expressed as a simple com-
tation in physical hardware and thus ideal for use in real-time bination of two DFTs each of length N /2. The DFT on the right
k
OOFDM transceivers. The IFFT can be created from an FFT by of (9) is multiplied by the factor ωN which accounts for the
simple modification. Therefore, the following discussions will relative time shift between the sub-sequences and is known as
concentrate on the FFT although they are equally applicable to the twiddle factor. If these N /2-point DFTs are denoted as Yk
the IFFT. and Zk respectively we can write
The fundamental principle of the FFT is to take the input k
Xk = Yk + ωN Zk (10)
sequence xn of N coefficients and split it into shorter subse-
quences. For this example, we split the original N point se- and
quence into two sequences of length N /2. Two DFTs of length (k + N2 )
N/2 can now be performed and the resulting two N /2 output Xk + N = Yk + N + ωN Zk + N (11)
2 2 2

sequences can be recombined to form the N point output se-

for k = 0,1,..,(N /2)–1. (11) can be further simplified as ωN N /2
quence of the original DFT. The recombination process takes
= −1 and Yk and Zk have a period of N /2, which gives the
care of the associated time shift between the sequences. This
following pair of equations known as butterfly operators:
splitting approach can be continued on the subsequences by fur-
k
ther division of the subsequences for M = log2 N total steps Xk = Yk + ωN Zk (12)
until the N -point DFT is replaced by N DFTs each of length
Xk + N = Yk − ωN
k
Zk . (13)
1. The DFT operation is now trivial as the DFT of a 1- point 2

sequence is itself, thus requiring no computational stage. The
task of the FFT is then to correctly recombine the 1-point se-
quences according to the splitting method, thus recombining the
N 1-point sequences into N /2 2-point sequences, and then these
into N /4 4-point sequences, and so on, until the final single N -
point transform sequence Xk is formed. The N -point FFT thus
consists of a total of log2 N recombination stages.
When the original sequence and all subsequences are equally
divided at each step, this is a radix-2 FFT with N = 2M , where
M is the number of recombination steps. Other radices are cre-
ated when the sequence is split into more than two subsequences
at each stage. For example, if the original sequence is first split
into four subsequences, this is repeated for M = log4 N steps.
In this case N = 4M and so it is a radix-4 FFT. The possible
These butterfly operators are the fundamental FFT building
blocks used at the recombination stages of the radix-2 FFT, and
are depicted by the example symbol shown in Fig. 14. To convert
two subsequences Yk and Zk to the single N -point sequence Xk
will thus require N /2 discreet butterfly operators.
Different radix values and sequence splitting methods will
have their own corresponding butterfly elements. The radix-4
butterfly, for example, has 4 coefficient inputs, three twiddle
factor inputs, and 4 coefficient outputs.
The selected splitting method can affect the order of the se-
quence reordering and the sequence recombination. The afore-
mentioned example splits the original sequence and the subse-
quences into even and odd sequences. This requires the input
coefficients xn to first be reordered and then the sequences to be
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS 563

recombined. As xn is reordered, this type of FFT architecture is

known as decimation-in-time. If, however, the splitting method
divides according to the first-half and last-half subsequences, xn
and xn +N /2 , where n = 0,1,..,(N /2)–1, the FFT first performs
the recombination of the naturally ordered input sequence and
then reorders the Xk coefficients. This type of FFT architecture
is therefore known as decimation-in-frequency.
The selected architecture for the FFT and IFFT implemented
in the real-time OOFDM transceiver is based on the fact that a
32-point DFT is required. A radix-2 FFT can be used as N =
32 = 25 . There is no difference in the complexity of decimation-
in-time and decimation-in-frequency, therefore decimation-in-
time is selected. The implemented FFT architecture is therefore
the Cooley–Tukey radix-2 decimation-in-time.
To implement the IFFT it is only necessary to modify the
twiddle factors, which is apparent from the opposite sign of the
exponential power in the IDFT in (5) compared to the DFT in
k
(6). Thus, for the IFFT, the twiddle factors are now ωN . To
convert the FFT function to an IFFT function, the twiddle factor
values are thus simply replaced with their complex conjugates.
It is important to consider the savings in computational com-
plexity achieved by the implemented FFT architecture compared
to the explicit computation according to the DFT definition in
(6). Explicit computation requires N 2 complex multiplications
and N 2 − N complex additions, whereas a radix-2 N -point FFT
will have (N /2)log2 N butterfly operators, each consisting of one
complex multiplier and two complex additions. Thus, in total
there are (N /2)log2 N complex multipliers and N log2 N com-
plex additions. For the case of the implemented 32-point FFT,
the computational saving is ∼92% for the complex multiplica-
tions and ∼84% for the complex additions. The actual saving is
higher when taking into account the instances where the twid-
dle factors are unity. The immense computational efficiency of
the IFFT is clear from the savings achieved. Furthermore, the
computational saving increases further with higher values of N .
For a hardware-based implementation of the DFT, the FFT is
therefore indispensable due to the advantages associated with
the vast reduction in the required logic resources. The FFT and
IFFT will still constitute one of the largest logic functions, if
not the largest logic function in the OOFDM transceiver, and so
the optimization of the FFT logic is an important issue.
As an example, Fig. 15 shows the structure of a 16 point radix-
2 decimation-in-time FFT. The repeated divide by 2 structure
is apparent in the right-to-left direction and the decimation-in-
time architecture results in the reordering of the xn values. An
important issue with hardware implementation is bit resolution
control of the intermediate stage sample values. As the butterfly
elements within each stage contain multiplication and addition
functions, the bit resolutions of the output samples will increase
at each stage. If this is not restricted, the final stage will have an
excessive bit resolution and so large logic resources will be con-
sumed. It is thus critical that the intermediate sample resolutions
are truncated to limit the excessive escalation of sample bit res-
olutions while maintaining sufficient calculation precision. As
shown in Fig. 15, sample resolution reduction must be built
into the FFT structure between stages. Another important factor
which can affect logic resource usage and calculation precision
is the bit resolution of the twiddle factor values. This must be
Fig. 15. Structure of a 16-point radix-2 decimation-in-time FFT.
carefully selected as overly high resolution can cause excessive
use of logic, whereas overly low resolution can cause insuffi-
cient calculation precision. The implemented twiddle factor is a
6 bit signed complex value.
For the implemented real-time transceiver, it is worth noting
the processing power of the FFT and IFFT. There are log2 (32) =
5 recombination stages each consisting of 32/2 = 16 radix-2 but-
terfly elements, giving a total of 80 butterflies. There are thus
80 complex multipliers and 160 complex adders giving a to-
tal of 240 complex operations. As the FFT/IFFT is clocked at
100 MHz, this results in a processing throughput of 24×109
complex operations per second. In comparison, direct computa-
tion with the DFT would require 1024 complex multipliers and
961 complex adders, resulting in close to 200×109 complex
operations per second.
C. Pilot Detection, Channel Estimation, and Equalization
The frequency response of the transmission channel intro-
duces subcarrier amplitude and phase changes during transmis-
sion. The received signal is therefore no longer a direct repre-
sentation of the transmitted signal. To compensate the effect of
the channel response, the inverse channel response is applied in
the receiver, which is termed channel equalization. In order to
perform equalization, the CTF must be estimated. An advantage
of OFDM is that channel equalization can be extremely simple.
As the amplitude and phase of each subcarrier are determined
at a discrete frequency, the CTF only needs to be known at the
corresponding frequency to allow the subcarrier to be equal-
ized. Equalization can then be achieved by a single complex
multiplication in the frequency domain.
To determine the CTF at the subcarrier frequencies, pilot
symbols Pk are periodically transmitted on each subcarrier with
known amplitude, Ak and phase θk , defined as
k
Pk = Ak ej (2π N n +θ k )
. (14)
564 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014

can be expressed as

pk ,m (m − k) = qNs
Xk ,m =
dk ,m (m − k) = qNs (19)
q = 0, 1, 2, 3, . . .
where pk ,m and dk ,m are the encoded complex pilot and data
values, respectively, and Ns is the total number of data bearing
Fig. 16. Subcarrier equalization to compensate for channel induced amplitude
and phase changes.
subcarriers (Ns is 15 in this case). The diagonal pilot mapping
approach was adopted as it has the advantage that no buffering
of the incoming data is required when all subcarriers carry the
same number of bits. However, it is still necessary to direct the
Ns –1 incoming data streams and the single pilot data value to
the appropriate subcarriers on a per OFDM symbol basis.
In the receiver, the 15 data-bearing OFDM subcarriers in the
positive frequency bins are selected for channel estimation and
subsequent data recovery at the FFT output. When transmission
is first established, the pilot detection block must locate the
symbols such that the first subcarrier is a pilot subcarrier. These
symbols are regarded as pilot subcarrier reference points relative
Fig. 17. Pilot and data bearing subcarrier mapping in the OOFDM time-
frequency symbol space. to which all other pilot subcarriers can be located. At the output
of the FFT, the identification of the received pilot subcarriers is
first made by performing operations (20) and (21) to subcarrier
The corresponding received pilot symbol Rk is 1 of consecutive symbols, where
k
Rk = Bk ej (2π N n +φ k )
+ Wk (15) ∗
Dm ,1 = Xm ,1 · X(m (20)
+N s ),1

C −1
2
1

where the received subcarrier amplitude and phase are Bk , and

φk respectively and Wk is the noise component of the kth sub- Qm ,1 =

D(m +iN s ),1
(21)
carrier after the receiver FFT. The CTF in the frequency domain, C
i=0

Hk , is then determined as
R k − Wk Bk j (φ k −θ k )
Hk = = e
Pk Ak
the CTF is thus estimated as
Rk Wk
Ĥk = = Hk + . (17)
Pk Pk
Large pilot symbol amplitude therefore reduces error due
to noise. To further reduce the effects of channel noise, the
estimated CTF can be averaged over many pilot symbols as
long as the channel can be considered to be static over the
averaging period. To equalize the received frequency domain
complex data, dk ,m , encoded onto the kth subcarrier, a single
multiplication by the inverse CTF estimate, Ĥk−1 , is applied.
The equalized encoded complex data value dk ,m is therefore
defined as
dk ,m = Ĥk−1 dk ,m .
The subcarrier equalization principle is illustrated in Fig. 16.
The real-time OOFDM transceiver implements pilot
subcarrier-based channel estimation in the following way. In
the transmitter, the pilot insertion function follows the parallel
data generator, such that one extra parallel bit sequence of a
fixed pattern, representing known pilot subcarrier data, is diag-
onally mapped into the OOFDM time-frequency symbol space
as shown in Fig. 17. Mathematically, the pilot and data-bearing
subcarrier mapping onto the frequency domain subcarriers Xk ,m
∗
such that Xm ,1 (X(m +N s),1 ) is the received complex (complex
conjugate) value of subcarrier 1 of the mth [(m+NS ) th] symbol.
(16)
C is a preset integer number determining the total number of
Ns symbol-spaced D values used for averaging. The magnitude
squared function is used in (21) as this gives a real-valued Q
value to simplify peak detection, and is also easier to compute
than the absolute magnitude which would require a square-root
operation. As the pilot mapping sequence repeats every NS
symbols, Qm ,1 must be determined for NS adjacent symbol lo-
cations. For the implemented design with NS = 15, 15 values
of Qm ,1 for consecutive symbols positions must be determined.
The data-bearing subcarriers are modulated with complex val-
ues encoded using a random data sequence. This results in mini-
mized Q values due to the averaging process. On the other hand,
each of the pilot subcarriers is modulated with a fixed complex
number of maximal amplitude, causing the occurrence of a Q
peak corresponding to the symbol locations where subcarrier
(18) 1 is the pilot subcarrier, as illustrated in Fig. 18. A large C
will make the Q peak more distinguishable, but this requires a
longer time and more logic resources to conduct the averaging
operation, such that C should be optimized. Experimental mea-
surements show that C = 16 is adequate for reliable detection
of pilot subcarriers. By locating the peak in the 15 detected and
stored Q values the symbols are identified where subcarrier 1 is
the pilot. Based on this reference pilot, all other pilot subcarriers
in subsequent symbols can be easily identified due to their fixed
relative positions. In the implemented design, the pilot detection
function operates continuously. However, after identifying the
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS
Fig. 18. Pilot subcarrier identification using Q peaks after FFT in the receiver.
reference pilot subcarriers, the pilot detection process could be
terminated and only needs to be reactivated following a break
in the transmission.
Making use of the known transmitted pilot subcarriers and the
received pilot subcarriers, the channel estimation block deter-
mines the complex CTF, Hk (k = 1, 2, . . .,Ns ), by performing
the operation defined below in (22) as:
M −1 −1
1  R(k +iN s ),k 
M
1 perimentally demonstrated that is highly suitable for applica-
Hk = = R(k +iN s ),k (22)
M i=0 P(k +iN s ),k M · PC i=0
where R(k +iN s),k (P(k +iN s),k ) is the received (assigned) com-
plex value of the kth pilot subcarrier in the (k + iNs ) th symbol.
As a constant power, PC , is assigned to the pilot subcarriers the
simplified expression on the right of (22) is used. To reduce the
noise effect associated with the transmission system, frequency
response averaging is performed over M pilot subcarriers at
each frequency. Here, M is taken to be 32, which is an op-
timum value identified experimentally [61]. Thus, to compute
Hk , parallel summation functions with suitable scaling are im-
plemented over 32 pilots for each subcarrier. The 15 computed
complex values forming the CTF are stored and fed to the chan-
nel equalization block with new values continuously computed
every 32 symbols.
The CTF obtained in the channel estimation function is then
used by the channel equalization block to equalize each individ-
ual subcarrier using the following operation:
 ST X = SOOFDM + SALIGN .
Xm ,k
Xm ,k =
Hk
 The corresponding received signal SR X can be written as
where Xm ,k is the received complex value of the kth unequal-
ized subcarrier in the mth symbol. The channel equalization
function thus consists of 15 parallel complex dividers. The
equalized subcarriers, Xm ,k , provide the inputs to the 15 parallel
adaptive demodulators.
The real and imaginary parts of the 15 complex CFT parame-
ters Hk determined by the channel estimation block are probed
by the Signal Tap II embedded logic analyzer. H1 to H15 are
extracted by the Signal Tap IIapplication in order to view the
live system frequency response from IFFT input to FFT output.
This feature is utilized in combination with individual subcarrier
BER measurements to manually determine suitable levels for the
variable power loading profile employed in the transmitter.
565
It is important to note that, due to the quasi-static nature of the
optical channel, a low CTF estimate update rate can be employed
without degrading system performance. The channel estimation
technique can therefore only insert pilot data in periodic bursts
of pilot subcarriers. This allows all 15 subcarriers to be used for
data transmission between pilot bursts. The insertion rate of the
pilot bursts can be as low as 10 Hz [61], corresponding to an
extremely low overhead of 0.001% for the channel estimation
function.
D. Symbol Synchronization
Symbol timing offset (STO) is the difference between the
correct symbol start position and the estimated symbol start po-
sition. Symbol synchronization is necessary to minimize STO,
which is ideally zero, as nonzero STO leads to degraded BER
performance if the processed samples do not all originate from
the same symbol. It should be noted, however, that if the CP
length exceeds the ISI length by y samples, an STO of up to
y samples can be tolerated without performance degradation.
STO tolerance can thus be improved by increasing CP length.
A DSP-based symbol synchronization method has been ex-
tion in OOFDM multiple access-based passive optical networks
(OOFDMA-PONs). This is because the technique can achieve
symbol, timeslot, and frame alignment of an optical network
unit’s (ONU’s) upstream and downstream signals without the
need to interrupt existing ONU traffic.
The symbol synchronization technique [62], [63] is based on
the principle that the dc level of each OOFDM symbol has no
influence on the nonzero frequency subcarriers at the output of
the FFT. A different dc offset can therefore be applied to each
individual OOFDM symbol. This operation thus produces an
encoded synchronization signal, SALIGN , which is added to the
OOFDM signal, SOOFDM , at the transmitter. For simplicity but
without losing generality, two different dc offsets with identical
amplitude but opposite polarities, +P and –P, are alternately
added to successive symbols. SALIGN is thus a square wave
with a period of 2 symbol periods (2·TS ), and a peak-to-peak
amplitude of 2P . The transmitted signal ST X as illustrated in
Fig. 19 is as follows:
(24)
(23)

SR X = SOOFDM 
+ SALIGN + SN (25)
where SN represents system noise.
In the receiver, a cross-correlation method is used to detect the

position of SALIGN . A signal SCORR is generated which has an
identically shaped waveform to SALIGN and amplitude of ±1
to simplify computation. By computing the cross-correlation
between SR X and SCORR , symbol alignment offset can be de-
termined based on the location of the correlation peaks. This
is because there is no correlation between SCORR and either

SOOFDM or SN due to their Gaussian random characteris-
tics. The cross-correlation is therefore entirely dependent on
566
Fig. 19. OOFDM signal combination with dc offset symbol alignment signal
(alignment signal dc levels are exaggerated for clarity).

SALIGN . An arbitrarily positioned sequence of 2·M · Z sam-
ples is processed, where Z is the total number of samples in an
OOFDM symbol, and M is a sufficiently large integer selected
to give clear correlation peaks. As SALIGN is cyclic, a sym-
bol summation, or accumulation, can be performed before the
cross-correlation. A signal SSUM is calculated using (26) where
SSUM (n) is the nth sample within SSUM and n = 1 to Z.

M −1 
M −1
 variation of |COR(ν)| against offset ν for an arbitrary symbol
SSUM (n) = SR X (n + 2kZ) = SOOFDM (n + 2kZ)
k =0 k =0
 
+ SALIGN (n + 2kZ) + SN (n + 2kZ)
where SSUM is thus the sum of M sequences of Z consecu-
tive samples spaced at intervals of 2·Z samples. If M is large
enough, the waveform of SSUM will take on the shape of SALIGN
as the Gaussian random characteristics of S OOFDM and SN re-
sult in their summations both tending to zero. The exact shape
of SSUM will depend on the symbol alignment offset relative to
the arbitrarily selected samples. Signal transitions from positive
to negative, and vice-versa, will thus coincide with the OOFDM
symbol boundaries. The cross-correlation is then performed be-
tween SSUM and SCORR with the relative offset, v, of SCORR
varied from 0 to (2·Z)–1 and the correlation value COR(ν)
for each offset calculated using (27). The sequence of values
COR(0) to COR(2Z–1) provides a correlation profile, CPROF ,
where the position of the peaks indicates the offset where the
highest correlation between S ALIGN and SCORR occurs, thus
identifying the position of the OOFDM symbol.

(2Z )−1
COR (v) = SSUM (k) · SCORR (k + v) .
k =0
A positive (negative) peak will occur in CPROF when SCORR
and S ALIGN are in phase (in opposite phase) both of which
indicate symbol alignment as SCORR and SALIGN have a period
of 2·TS . By taking |COR(ν)|, only positive peaks then occur in
CPROF and it is only necessary to select Z samples in every 2·Z
samples to ensure a peak is detected. Fig. 20 shows the ideal
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014
Fig. 20. Ideal correlation profile for symbol offset w 0 .
Fig. 21. Symbol offset detection block diagram.
alignment offset of w0 .
The addition of the dc offset level is performed in the signed
(26) to unsigned block in Fig. 13. The added dc offset is online
adjustable to allow optimization. In the experimental demon-
stration [62], it was shown that a dc offset as small as ±1
quantization level was sufficient and resulted in no reduction in
system BER performance.
A block diagram of the implemented symbol offset detec-
tion function is shown in Fig. 21. The sum and accumulate
block consists of 40 parallel accumulators, corresponding to the
40 samples per symbol period, to generate a new SSUM every
10 000 symbols as M = 5000. As each accumulator sums a total
of 5000 8 bit samples, between resets, scaling is used to limit the
accumulator outputs to 12 bits. 40 parallel cross-correlators are
employed to generate the correlation profile. A peak detector
detects the position of the correlation profile peak to determine
the symbol offset value.
It should be noted that as the correlation signal SCORR
has values of ±1, the cross-correlation function consists of 40
add/subtract operators each with 40 inputs. The offset of the cor-
responding SCORR value determining if a sample is added or
subtracted. The use of multipliers is thus avoided to reduce de-
(27)
sign complexity. Also, it should be noted that although multiple
parallel cross-correlators were employed, it would be possible to
significantly reduce logic resources by implementing the func-
tion with a single cross-correlator, and sequentially increment
the offset of the correlation signal to build up the correlation
profile one value at a time.
As previously discussed, the symbol synchronization tech-
nique is designed for application in OOFDMA-PONs to achieve
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS
downstream and upstream alignment of symbols, timeslots and
frames. Here, the mechanism for achieving synchronization of
an ONU in a live PON is described. For downstream symbol
alignment, the optical line terminal (OLT) continuously trans-
mits a synchronization signal which all ONUs use for symbol
alignment. In the upstream direction, the OLT controls the syn-
chronization process, allowing only one ONU to transmit a
synchronization signal at any one time. For each ONU, the OLT
detects its upstream symbol offset and then notifies the ONU
via a control channel, so that it can correctly realign its sym-
bol positions. The ONU thus contains a symbol offset detection
function in its receiver and a symbol offset adjustment function
in both its transmitter and receiver. The OLT only requires the
symbol offset detection function in its receiver.
By constructing the synchronization signal from a coded
sequence of dc offsets, enhancements can be made that of-
fer a number of key features. For upstream and downstream
frame/timeslot alignment, a suitably coded synchronization sig-
nal with the same length as one or more OOFDMA frames
allows the OLT to detect an ONU’s frame alignment offset by
performing a cross-correlation over a period equivalent to one
or more coded sequence lengths. ONU frame alignment offset is
detected and then corrected in the ONU, such that each ONU is
timeslot aligned to the network before initiating OOFDMA sig-
nal transmission and hence avoiding any upstream ONU signal
collisions in the operational network. As symbol offset can drift
slowly over time, the OLT must periodically track and correct
any symbol offset drift for each ONU in turn.
Furthermore, coding the downstream synchronization signal
with a sufficiently long encrypted key code will make it virtually
impossible for an unauthorized user to achieve synchronization,
thus achieving network security at the physical layer.
An alternative symbol synchronization technique employing
a subtraction-based correlation method for cyclic prefix location
has also been implemented and fully verified in the real-time
transceiver [64]–[66].
E. Clock Synchronization
Accurate synchronization of the OOFDM receiver and trans-
mitter sampling clocks is essential to minimize their sam-
pling frequency offset (SFO) [67]. SFO induces interchannel-
interference (ICI) which produces increasing received signal
distortion with increasing SFO. ICI results from the loss of
subcarrier orthogonality due to the mismatch between the dis-
crete subcarrier frequencies in the receiver compared to those
in the transmitter. SFO also induces a drift in symbol alignment
necessitating periodic symbol realignment.
Due to the noise-like nature of the OOFDM signal clock re-
covery is not straightforward. However, asynchronous (nonzero
SFO) and synchronous (zero SFO) clocking techniques in real-
time OOFDM transmissions have been demonstrated. The CP
detection-based symbol alignment method [66] supports asyn-
chronous clocking. The technique is able to compensate for SFO
as it continuously readjusts the symbol alignment and so pre-
vents the accumulation of excessive STO. Suitably fast symbol
offset tracking and symbol realignment was demonstrated to
compensate for a SFO as high as 4000 ppm. This method [66]
567
Fig. 22. Multiband OFDM signal generation principle.
can also estimate SFO to an accuracy of <1 ppm which could
allow highly accurate synchronization of the receiver clock via
a feedback loop from a voltage controlled oscillator (VCO).
For synchronous clocking, a DSP-free technique employing
an out-of-band auxiliary clock signal was employed [68]–[70].
At the transmitter, a 4 GHz sine wave is combined with the 0–
2 GHz baseband OFDM electrical signal, before driving the
optical modulator. At the receiver the clock signal and the
OFDM signal are separated using electrical filters. The received
clock signal is prescaled and feeds a PLL generating a low jit-
tered clock at the required frequency for clocking the receiver.
The synchronous clocking technique has particular significance
for implementing accurate synchronization of OOFDMA-based
PONs.
VI. MULTIBAND OOFDM-BASED PONS FOR IMPROVED
COST EFFECTIVENESS
As the cost of the ONU is critical in an OOFDMA PON, it
is important to avoid unnecessary overengineering of the ONU.
Employing single-band OOFDM transceivers in a PON leads to
the undesirable scenario, where all ONUs support the full peak
PON capacity. In practice, however, an ONU will only ever need
to operate at a reduced peak capacity. If a multiband OOFDM
approach [51], [52] is adopted this can overcome the drawbacks
of the single-band approach. Fig. 22 illustrates the multiband
OFDM signal generation principle. Each OFDM transceiver
generates a baseband signal, which is then up-converted using
a unique RF carrier frequency such that the generated subbands
do not overlap in the frequency domain. A frequency division
multiplexing (FDM) method is thus adopted to combine the
OOFDM subbands together. In the downstream direction, the
OOFDM subbands are electrically summed in the OLT before
EO conversion, whereas for the upstream direction, the summa-
tion occurs in the optical domain in the optical coupler in the
PON’s remote node.
There are many advantages associated with the multiband
OOFDM technique particularly when considering the ONU im-
plementation. It offers the key advantage of flexibility in adopted
DAC/ADC bandwidth as this is no longer dictated by the total
PON capacity. Moreover, ONU signal processing complexity is
568 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014

reduced, as only one PON subband is processed. If the subband TABLE III
RELATIVE SUBBAND PARAMETERS FOR MULTIBAND OOFDM
transceiver is designed to support subband tunability, it pro-
vides increased network operation efficiency in terms of both
dynamic bandwidth provisioning and equipment logistics. Fur-
thermore, the reduced complexity leads to a reduction in ONU
power consumption. Although the OLT must support all sub-
bands, the same tunable subband transceiver electronics, as used
in the ONUs, can be employed. This provides the benefits of
economies of scale in transceiver manufacturing, and also al-
lows a scalable OLT architecture, where capacity can expand
in sub-band capacity increments in line with service take up.
It would also be possible to implement dynamic traffic redis-
tribution across subbands, so that when PON traffic levels per-
mit, transceivers can be powered down to reduce OLT energy
consumption. As higher cost and complexity can be tolerated
at the OLT side of a multiband OOFDMA-PON, an alterna-
tive OLT architecture employing wideband DACs and ADCs,
for direct digital-to-RF conversion of all subbands [71], is also
conceivable.
An OOFDM transceiver, designed to support a single subband
in a multi band system, will additionally require, two RF mixers,
a single tunable LO and RF filters to support the up-conversion
and down-conversion of the OFDM signal. The IMDD optical
components will be similar for the single-band and multiband
transceivers; although the multiband approach will require wider
optical component bandwidths. This is because the optical band Fig. 23. Relative IFFT/FFT complexity and DAC/ADC bandwidth per ONU
must encompass all subbands to support dynamic subband tun- for varying number of side-bands in a multiband OOFDMA-PON.
ability. Furthermore, double side-band subbands and the inter-
subband spacing will increase the required optical bandwidth. If the relative IFFT/FFT processing complexity and
It is important to compare the difference between subband DAC/ADC bandwidths, for a single subband ONU, are cal-
generation using a single carrier, and using two orthogonal culated for the case of IQ modulated subbands. The relative
carriers for IQ modulation. IQ modulated subbands have the parameter values in Table III and Fig. 23 show almost ex-
advantage of increasing spectral efficiency. This theoretically actly the same trends when considering a PON supporting ∼500
halves the subband bandwidth for a given data capacity. As a subcarriers.
consequence, optical component bandwidth requirements are Using the adaptively modulated 4 GS/s real-time OOFDM
significantly reduced. For IQ modulation, DAC and ADC pairs transceivers, the feasibility of multiband OOFDM has been
are needed, as illustrated in Fig. 12, though the bandwidths are demonstrated by real-time dual-band OOFDM experimental
now halved compared to a single carrier generated subband of demonstrations employing baseband (0–2 GHz) and passband
the same data capacity. (∼4–8 GHz) OFDM signals. OOFDM signal transmission at
The savings in IFFT/FFT processing complexity and >19 Gb/s over 25 km of SSMF has been successfully demon-
DAC/ADC bandwidths will now be considered in detail for strated [51], [52]. Furthermore, by using this approach, real-time
the case of an ONU that supports one subband using a single OFDM signal generation and detection at 30 Gb/s has been
RF carrier. For an IMDD system employing real-valued time demonstrated [53].
domain signals, for each subband, the relationship between the
number of data-carrying subcarriers NS and the IFFT/FFT size
N , is NS = (N /2)–1. For an N point radix-2 decimation-in- VII. CONCLUSION
time IFFT/FFT architecture, the number of complex operations This paper has provided an overview of the implementation
is N log2 N complex additions and (N/2)log2 N complex multi- aspects associated with DSP-based optical transceivers for fu-
plications, as described in Section V-B. If the total capacity and ture access networks by examining the optical transceiver struc-
thus the number of subcarriers in a PON is fixed, then the effect ture and the key transceiver constituent elements. This paper fo-
of employing multiple bands is to reduce the number of sub- cuses on DSP functionality and architecture of OOFDM-based
carriers per subband and also reduce the required bandwidth of optical transceivers. Real-time OOFDM transceivers experi-
each subband. If each ONU supports one subband, the required mentally demonstrated by Bangor University are also analyzed
IFFT/FFT complexity and DAC/ADC bandwidth for each ONU in depth.
will reduce as the number of subbands increases. Table III and In summary, today’s digital processing and DAC/ADC device
Fig. 23 illustrate this relationship for a PON with a total of at technologies are sufficiently mature to support the application of
least 500 subcarriers. DSP in high speed optical access networks. Exploiting DSP not
GIDDINGS: REAL-TIME DIGITAL SIGNAL PROCESSING FOR OPTICAL OFDM-BASED FUTURE OPTICAL ACCESS NETWORKS
only allows future network performance requirements to be met,
but can also enable cost-effective technology by allowing the
utilization of low cost, simple IMDD optical components. Fur-
thermore, the high equipment volumes associated with optical
access networks can inevitably lead to cost-effective electron-
ics. OOFDM is one of the leading DSP-based optical access
technologies which is perceived by many as one of the main
contenders for future optical access networks due to its poten-
tial for high cost-effectiveness, data capacity per wavelength
far beyond 10 Gb/s, adaptiveness to varying network charac-
teristics, and flexibility in terms of bandwidth allocation. It is
also envisioned that adopting a multiband OOFDM-based PON
will maximize total access network cost-effectiveness. Given
the exponentially growing demand for data capacity and the
operators’ need for flexible cost-efficient access networks, it is
believed that DSP-based optical access networks will emerge in
the future.
REFERENCES
[1] Z. Kostic and S. Seefharaman, “Digital signal processors in cellular radio
communications,” IEEE Commun. Mag., vol. 35, no. 12, pp. 22–35, Dec.
1997.
[2] S. Chaturvedi, “The role of digital signal processors (DSP) for 3 G mobile
communication systems,” Int. J. Emerging Technol., vol. 1, no. 1, pp. 23–
26, Feb. 2010.
[3] A. Gatherer, T. Stetzler, M. McMahan, and E. Auslander, “DSP-based ar-
chitectures for mobile communications: Past, present, and future,”. (2000,
Jan.). IEEE Commun. Mag., vol. 38, pp. 84–90, 2000.
[4] “Coherent DWDM technologies.” (2012). White Paper Infinera. [On-
line] Available: www.infinera.com/pdfs/whitepapers/Infinera_Coherent_
Tech.pdf
[5] T. Koonen, “Trends in optical access and in-building networks,” in Eur.
Conf. Opt. Commun., Brussels, Belgium, 2008, pp. 1–31.
[6] P. Vetter, “Next generation optical access technologies,” presented at Euro-
pean Conf. Optical Communication, Amsterdam, The Netherlands, 2012.
[7] Y. Luo, X. Zhou, F. Effenberger, X. Yan, G. Peng, Y. Qian, and Y. Ma,
“Time- and wavelength-division multiplexed passive optical network
(TWDM-PON) for next-generation PON stage 2 (NG-PON2),” J. Lightw.
Technol., vol. 31, no. 4, pp. 587–593, Feb. 2013.
[8] G. Engel, D. E. Fague, and A. Toledano, “RF digital-to-analog converters
enable direct synthesis of communications signals,” IEEE Commun. Mag.,
vol. 50, no. 10, pp. 108–116, Oct. 2012.
[9] C. Cox, E. Ackerman, R. Helkey, and G. E. Betts, “Techniques and per-
formance of intensity-modulation direct-detection analog optical links,”
IEEE Trans. Mircowave Theory Tech., vol. 45, no. 8, pp. 1375–1383, Aug
1997.
[10] A. Villafranca, J. Lasobras, and I. Garcés, “Precise characterization of
the frequency chirp in directly modulated DFB lasers,” in Spanish Conf.
Electron. Devices, 2007, pp. 173–176.
[11] M. Müller and M. C. Amann, “State-of-the-art and perspectives for long-
wavelength high speed VCSELs,” in 13th Int. Conf. Transp. Opt. Netw.,
Stockholm, Sweden, 2011, pp. 1–4, Paper Mo.C5.2.
[12] C. H. Yeh and C. W. Chow, “10-Gb/s upstream TDM-PON based on four
WDM signals with OFDM-QAM remodulation,” in 13th OptoElectron.
Commun. Conf., Hong Kong, 2009, pp. 1–2, Paper ThLP46.
[13] M. F. Huang, D. Qian, and N. Cvijetic, “A novel symmetric lightwave cen-
tralized WDM-OFDM-PON architecture with OFDM-remodulated ONUs
and a coherent receiver OLT,” presented at Europeon Conf. Optical Com-
munication, Geneva, Switzerland, 2011, Paper Tu.5.C.1.
[14] N. Cvijetic, “OFDM for next-generation optical access networks,” J.
Lightw. Technol., vol. 30, no. 4, pp. 384–398, Feb. 2012.
[15] J. L. Wei, J. D. Ingham, D. G. Cunningham, R. V. Penty, and I. H. White,
“Performance and power dissipation comparisons between 28 Gb/s NRZ,
PAM, CAP and optical OFDM systems for data communication applica-
tions,” J. Lightw. Technol., vol. 30, no. 20, pp. 3273–3280, Oct. 2012.
[16] “Stratix V device overview,” (2013). [Online]. Available: http://www.
altera.co.uk/literature/hb/stratix-v/stx5_51001.pdf
569
[17] R. Bouziane, P. Milder, R. Koutsoyannis, Y. Benlachtar, J. C. Hoe,
M. Püschel, M. Glick, and R. I. Killey, “Design studies for ASIC imple-
mentations of 28 GS/s optical QPSK- and 16-QAM-OFDM transceivers,”
Opt. Exp., vol. 19, no. 21, pp. 20857–20864, Oct. 2011.
[18] R. W. Chang, “Synthesis of band-limited orthogonal signals for multi-
channel data transmission,” Bell Syst. Technol. J., vol. 45, pp. 1775–1796,
1966.
[19] R.W Chang and R. A. Gibby, “A theoretical study of performance of an or-
thogonal multiplexing data transmission scheme,” IEEE Trans. Commun.
Technol., vol. 16, no. 4, pp. 529–540, Aug. 1968.
[20] M. Zimmerman and A. Kirsch, “The AN/GSC-10/KATHRYN vari-
able rate data modem for HF radio,” IEEE Trans. Commun. Technol.,
vol. COM-15, no. 2, pp. 197–205, Apr. 1967.
[21] J. Salz and S. B. Weinstein, “Fourier transform communication system,”
in Proc. ACM Symp. Problems Optim. Data Commun. Syst., New York,
NY, USA, 1969, pp. 99–128.
[22] W. L. Briggs and V. E. Henson, The DFT An Owner’s Manual for the
Discrete Fourier Transform. Philadelphia, PA: SIAM, 1995.
[23] J. Armstrong, “OFDM for optical communications,” J. Lightw. Technol.,
vol. 27, no. 3, pp. 189–204, Feb. 2009.
[24] A. Ragheb and H. Fathallah, “Candidate modulation schemes for next
generation-passive optical networks (NG-PONs),” presented at the High
Capacity Optical Networks and Enabling Technol., Istanbul, Turkey, 2012.
[25] C. Milion, T. Duong, N. Genay, E. Grard, V. Rodrigues, B. Charbonnier,
J. Le Masson, M. Ouzzif, P. Chanclou, and A. Gharba, “High bit rate trans-
mission for NG-PON by direct modulation of DFB laser using discrete
multi-tone,” presented at the European Conf. Optical Communication,
Vienna, Austria, 2009, Paper 7.5.4.
[26] L. A. Neto, A. Gharba, P. Chanclou, N. Genay, B. Charbonnier, M. Ouzzif,
C. Aupetit-Berthelemot, and J. Le Masson, “High bit rate burst mode
optical OFDM for next generation passive optical networks,” presented
at the European Conf. Optical Communication, Torini, Italy, 2010, Paper
Tu.3.B.5.
[27] E. Wong, “Next-generation broadband access networks and technologies,”
J. Lightw. Technol, vol. 30, no. 4, pp. 597–608, Feb. 2012.
[28] N. Cvijetic, “Recent progress in DSP-based optical access and OFDMA-
PON,” presented at the Opto-Electronic and Communication Conf., Busan,
Korea, 2012, Paper 4A4-1.
[29] B. Inan, S. Adhikari, O. Karakaya, P. Kainzmaier, M. Mocker, H. von
Kirchbauer, N. Hanik, and S. L. Jansen, “Realization of a real-time 93.8-
Gb/s polarization-multiplexed OFDM transmitter with 1024-point IFFT,”
in Proc. Eur. Conf. Opt. Commun., Geneva, Switzerland, 2011, Paper
We.9.A.7.
[30] S.G. Wilson, Digital Modulation and Coding. Englewood Cliffs, NJ:
Prentice-Hall, 1995.
[31] J. M. Tang, P. M. Lane, and K. A. Shore, “Transmission performance of
adaptively modulated optical OFDM signals in multimode fiber links,”
IEEE Photon. Technol. Lett., vol. 18, no. 1, pp. 205–207, Jan. 2006.
[32] J. M. Tang and K. Alan Shore, “30-Gb/s signal transmission over 40-km
directly modulated DFB-laser-based single-mode-fiber links without opti-
cal amplification and dispersion compensation,” J. Lightw. Tech., vol. 24,
no. 6, pp. 2318–2327, Jun. 2006.
[33] E. Giacoumidis, A. Kavatzikidis, A. Tsokanos, J. M. Tang, and I. Tomkos,
“Adaptive loading algorithms for IMDD optical OFDM PON systems
using directly modulated lasers,” J. Opt. Commun. Netw., vol. 4, no. 10,
pp. 769–778, Oct. 2012.
[34] X. Q. Jin, J. L. Wei, R. P. Giddings, T. Quinlan, S. Walker, and J. M. Tang,
“Experimental demonstrations and extensive comparisons of end-to-end
real-time optical OFDM transceivers with adaptive bit and/or power load-
ing,” IEEE Photon. J., vol. 3, no. 3, pp. 500–511, Jun. 2011.
[35] R. P. Giddings, X. Q. Jin, E. Hugues-Salas, E. Giacoumidis, J. L. Wei,
and J. M. Tang, “Experimental demonstration of a record high 11.25 Gb/s
real-time optical OFDM transceiver supporting 25 km SMF end-to-end
transmission in simple IMDD systems,” Opt. Exp., vol. 18, no. 6, pp. 5541–
5555, Mar. 2010.
[36] Y. Benlachtar, P. M. Watts, R. Bouziane, P. Milder, D. Rangaraj,
A. Cartolano, R. Koutsoyannis, J. C. Hoe, M. Püschel, M. Glick, and
R. I. Killey, “Generation of optical OFDM signals using 21.4 GS/s real
time digital signal processing,” Opt. Exp., vol. 17, no. 20, pp. 17658–
17668, Sep. 2009.
[37] Y. Benlachtar, P. M. Watts, R. Bouziane, P. Milder, R. Koutsoyannis,
J. C. Hoe, M. Püschel, M. Glick, and R. I. Killey, “21.4 GS/s real-
time DSP-based optical OFDM signal generation and transmission over
1600 km of uncompensated fibre,” presented at the 35th European Conf.
Optical. Communication, Vienna, Austria, 2009, Paper PD2.4.
570
[38] F. Buchali, R. Dischler, A. Klekamp, M. Bernhard, and D. Efinger, “Reali-
sation of real-time 12.1 Gb/s optical OFDM transmitter and its application
in a 109 Gb/s transmission system with coherent reception,” presented at
the 35th European Conf. Optical Communication, Vienna, Austria, 2009,
Paper PD2.1.
[39] R. Schmogrow, M. Winter, B. Nebendahl, D. Hillerkuss, J. Meyer,
M. Dreschmann, M. Huebner, J. Becker, C. Koos, W. Freude, and
J. Leuthold, “101.5 Gbit/s real-time OFDM transmitter with 16QAM
modulated subcarriers,” presented at the Optical Fiber Communication
Conf., Los Angeles, CA, USA, 2011, Paper OWE5.
[40] R. Schmogrow, R. Bouziane, D. Hillerkuss, P. A. Milder,
R. J. Koutsoyannis, Y. Benlachtar, P. M. Watts, P. Bayvel, R. I. Killey,
C. Koos, W. Freude, and J. Leuthold, “85.4 Gbit/s real-time OFDM signal
generation with transmission over 400 km and preamble-less reception,”
presented at the Conf. Lasers and Electro-Optics, San Jose, CA, USA,
2012, Paper CTh1H.4.
[41] Q. Yang, S. Chen, Y. Ma, and W. Shieh, “Real time reception of multi-
gigabit coherent optical OFDM signals,” Opt. Exp., vol. 17, no. 10,
pp. 7985–7992, May 2009.
[42] S. Chen, Y. Ma, and W. Shieh, “Multiband real-time coherent optical
OFDM reception up to 110 Gb/s with 600-km transmission,” IEEE Pho-
ton. J., vol. 2, no. 3, pp. 454–459, Jun. 2010.
[43] N. Kaneda, Q. Yang, X. Liu, S. Chandrasekhar, W. Shieh, and Y. Chen,
“Real-time 2.5 GS/s coherent optical receiver for 53.3-Gb/s sub-banded
OFDM,” J. Lightw. Tech., vol. 28, no. 4, pp. 494–501, Feb. 2010.
[44] X. Xiao, F. Buchali, M. Berhard, S. Chen, and H. Bülow, “Implementation
of a 10.5 Gb/s real-time CO-OFDM receiver,” presented at the 16th Op-
toeElectronics and Communications Conf. (OECC), Kaohsiung, Tiawan,
2011.
[45] D. Qian, T. T. Kwok, N. Cvijetic, J. Hu, and T. Wang, “41.25 Gb/s real-time
OFDM receiver for variable rate WDM-OFDMA-PON transmission,” pre-
sented at the Optical Fiber Communication Conf., San Diego, CA, USA,
2010, Paper PDPD9.
[46] R. P. Giddings, X. Q. Jin, H. H. Kee, X. L. Yang, and J. M. Tang, “Real-
time implementation of optical OFDM transmitters and receivers for prac-
tical end-to-end optical transmission systems,” Electron. Lett., vol. 45,
no. 15, pp. 800–802, 2009.
[47] R. P. Giddings, X. Q. Jin, and J. M. Tang, “Experimental demonstration of
real-time 3 Gb/s optical OFDM transceivers,” Opt. Exp., vol. 17, no. 19,
pp. 16654–16665, Sep. 2009.
[48] R. P. Giddings, X.Q. Jin, and J.M. Tang, “First experimental demonstra-
tion of 6 Gb/s real-time optical OFDM transceivers incorporating chan-
nel estimation and variable power loading,” Opt. Exp., vol. 17, no. 22,
pp. 19727–19738, Oct. 2009.
[49] R. P. Giddings, E. Hugues-Salas, X.Q. Jin, J. L. Wei, and J. M. Tang,
“Experimental demonstration of real-time optical OFDM transmission at
7.5 Gb/s over 25-km SSMF using a 1-GHz RSOA,” Photon. Technol. Lett.,
vol. 22, no. 11, pp. 745–747, Jun. 2010.
[50] R.P. Giddings, E. Hugues-Salas, B. Charbonnier, and J. M. Tang, “Exper-
imental demonstration of real-time optical OFDM transmission at 11.25
Gb/s over 500 m MMFs employing directly modulated DFB lasers,” IEEE
Photon. Technol. Lett., vol. 23, no. 1, pp. 51–53, Jan. 2011.
[51] R.P. Giddings, E. Hugues-Salas, and J.M. Tang, “Experimental demon-
stration of record high 19.125 Gb/s real-time end-to-end dual-band optical
OFDM transmission over 25 km SMF in a simple EML-based IMDD sys-
tem,” Opt. Exp., vol. 20, no. 18, pp. 20666–20679, Aug. 2012.
[52] Q. W. Zhang, E. Hugues-Salas, R. P. Giddings, M. Wang, and J. M. Tang,
“Experimental demonstrations of record high REAM intensity modulator-
enabled 19.25 Gb/s real-time end-to-end dual-band optical OFDM color-
less transmissions over 25 km SSMF IMDD systems,” Opt. Exp., vol. 21,
no. 7, pp. 9167–9179, Apr. 2013.
[53] R.P. Giddings, E. Hugues-Salas, and J. Tang, “30 Gb/s real-time triple sub-
band OFDM transceivers for future PONs beyond 10 Gb/s/λ,” presented at
the European Conf. Optical Communication, London, U.K., 2013, Paper
P.6.7.
[54] S. Cho, K. W. Doo, J. H. Lee, J. Lee, S. I. Myong, and S. S. Lee, “Demon-
stration of a real-time 16 QAM encoded 11.52 Gb/s OFDM transceiver for
IM/DD OFDMA-PON systems,” in 18th OptoElect. Comm. Conf. (OECC)
and Int. Conf. Photon. Switching (OECC/PS), Kyoto, Japan, 2013, Paper
WP2-3.
[55] IEEE Standard for Test Access Port and Boundary-Scan Architecture,
IEEE Standard 1149.1-2013, 2013.
[56] X.Q. Jin, E. Hugues-Salas, R.P. Giddings, J.L. Wei, J. Groenewald, and
J.M. Tang, “First real-time experimental demonstrations of 11.25 Gb/s
optical OFDMA PONs with adaptive dynamic bandwidth allocation,”
Opt. Exp., vol. 19, no. 21, pp. 20557–20570, Oct. 2011.
JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 32, NO. 4, FEBRUARY 15, 2014
[57] E. Hugues-Salas, R. P. Giddings, X. Q. Jin, Y. Hong, T. Quinlan, S. Walker,
and J.M. Tang, “REAM intensity modulator-enabled 10 Gb/s colorless
upstream transmission of real-time optical OFDM signals in a single-
fiber-based bidirectional PON architecture,” Opt. Exp., vol. 20, no. 19,
pp. 21089–21100, Sep. 2012.
[58] E. Hugues-Salas, Q. Zhang, R.P. Giddings, M. Wang, and J.M. Tang,
“Adaptability-enabled record-high and robust capacity-versus-reach per-
formance of real-time dual-band optical OFDM signals over various
OM1/OM2 MMF systems [Invited],” J. Opt. Comm. Netw., vol. 5, no. 10,
pp. A1–AA11, Jun. 2013.
[59] J. W. Cooley and J. W. Tukey, “An algorithm for the machine calculation
of complex fourier series,” Math. Comput., vol. 19, no. 90, pp. 297–301,
Apr. 1965.
[60] E. J. Kim and M. H. Sunwoo, “High speed eight-parallel mixed-radix FFT
processor for OFDM systems,” presented at the IEEE Int. Symposium on
Circuits and Systems, Rio de Janeiro, Brazil, 2011.
[61] X.Q. Jin, R.P. Giddings, and J.M. Tang, “Real-time transmission of 3 Gb/s
16-QAM encoded optical OFDM signals over 75 km SMFs with negative
power penalties,” Opt. Exp., vol. 17, no. 17, pp. 14574–14585, Aug. 2009.
[62] R.P. Giddings and J. M. Tang, “Real-time experimental demonstration of
a versatile optical OFDM symbol synchronisation technique using low-
power DC offset signalling,” in Eur. Conf. Exhibit. Optical Communica-
tion. (ECOC), Geneva, Switzerland, 2011, Paper We9A3.
[63] R. P. Giddings and J. M. Tang, “Symbol alignment in high speed optical
orthogonal frequency division multiplexing transmission systems,” U.K.
Patent 1105808.8, 2011.
[64] X. Q. Jin, R. P. Giddings, E. Hugues-Salas, and J. M. Tang, “Real-time
experimental demonstration of optical OFDM symbol synchronization in
directly modulated DFB laser-based 25 km SMF IMDD systems,” Opt.
Exp., vol. 18, no. 20, pp. 21100–21110, Sep. 2010.
[65] J. M. Tang and X. Q. Jin, “Synchronization process in optical frequency
division multiplexing transmission,” U.K. Patent 0919057.0, 2009.
[66] X. Q. Jin and J. M. Tang, “Optical OFDM synchronization with symbol
timing offset and sampling clock offset compensation in real-time IMDD
systems,” IEEE Photon. J., vol. 3, no. 2, pp. 187–196, Apr. 2011.
[67] D. K. Kim, S. H. Do, H. B. Cho, H. J. Choi, and K. B. Kim, “A new joint
algorithm of symbol timing recovery and sampling clock adjustment for
OFDM systems,” Trans. Consumer Electron., vol. 44, no. 3, pp. 1142–
1149, 1998.
[68] R. P. Giddings and J. M. Tang, “World-first experimental demonstration of
synchronous clock recovery in an 11.25 Gb/s real-time end-to-end optical
OFDM system using directly modulated DFBs,” presented at the Optical
Fiber Communication Conf., Los Angeles, CA, USA, 2011, Paper OMS4.
[69] R. P. Giddings and J. M. Tang, “Experimental demonstration and optimi-
sation of a synchronous clock recovery technique for real-time end-to-end
optical OFDM transmission at 11.25 Gb/s over 25 km SSMF,” Opt. Exp.,
vol. 19, no. 3, pp. 2831–2845, Jan. 2011.
[70] R. P. Giddings and J. M. Tang, “Synchronous clocking for optical orthog-
onal frequency division multiplexing transmission systems,” U.K. Patent
1 008 018.2, 2010.
[71] N. Cvijetic, M. Cvijetic, M. F. Huang, E. Ip, Y. K. Huang, and T. Wang,
“Terabit optical access networks based on WDM-OFDMA-PON,” J.
Lightw. Technol., vol. 30, no. 4, Feb. 2012.
Roger Giddings received the B.Eng. (Hons.) degree in engineering science and
technology from Loughborough University, Loughborough, U.K., in 1989, and
the Ph.D. degree in optical communications in 2011 from Bangor University,
Bangor, U.K.
He joined Nokia Networks UK in 1990, where he worked as a Hardware
Design Engineer and Senior Hardware Design Engineer. In 1998, he joined
Nokia Networks, Finland. In 2000, he joined Nokia’s Research Center, Finland,
as a Senior Research Engineer. In 2004, he joined the Nokia Ventures Unit,
Finland, as a Senior Hardware Specialist.
In 2007, he joined the Optical Communications Research Group, School of
Electronic Engineering, Bangor University, as an Experimental Researcher and
studying for the Ph.D. degree in the area of real-time DSP for OOFDM-based
next generation optical networks. In 2012, he was awarded a lectureship and is
continuing to research optical communication systems based on optical OFDM.
During his work at Bangor he has (co)authored more than 60 papers and has
filed 2 patents.