International Journal of Video & Image Processing and Network Security IJVIPNS-IJENS Vol: 11 No: 06 1

112106-6868 IJVIPNS-IJENS December 2011 IJENS

I J E N S

AbstractBit dependency of image compression algorithms
introduce error extension effects in transmission of compressed
images and considerably reduce the received image quality.
Channel coding provides protection against these errors but at
the cost of increased bandwidth. Asymmetric modulation methods
like QAM and QPSK provide alternative means of equal error
protection to all the coded bits without increasing the bandwidth.
This paper examines the suitability of Hierarchical QAM
(HQAM) where non-uniform signal constellation is used to
provide different degrees of protection to the significant and non-
significant bits in the compressed image data at lower channel
Signal to Noise Ratio (SNR) and compares it with that of QAM.
The performance of HQAM is evaluated using gray test image for
different values of the modulation parameter.

Index TermsError protection, HQAM, Image coding,
Wireless channel.

I. INTRODUCTION
or transmission of images over wireless communication
channels, bandwidth limitation and high probability of
error are two major concerns. Therefore, compression is
applied to the transmitted data in order to conserve the
bandwidth. However, compression increases bit dependency
that in turn introduces error extension effects. Another
problem unique to wireless networks is the extremely hostile
and random nature of the channels that introduce distortion
and considerably degrades the image quality. Error control
coding is used to control the errors, however, the addition of
check bits that carry no information further increases the data
rate and consequently the bandwidth. [1-6]




This work was supported by the Ministry of Science, Technology and
Innovation (MOSTI) under grant 01-01-15-SF0115.
Md. Abdul Kader is with the School of Computer and Communication
Engineering, University Malaysia Perlis, Malaysia. (e-mail:
kdr2k4@rediffmail.com).
Prof. Dr. Farid Ghani was with University Saint Malaysia, Malaysia. He is
now with the School of Computer and Communication Engineering,
University Malaysia Perlis, Malaysia. (e-mail: faridghani@unimap.edu.my).
Assoc. Prof. R. Badlishah Ahmed is with the School of Computer and
Communication Engineering, University Malaysia Perlis, Malaysia. (e-mail:
badli@unimap.edu.my).
This paper proposes the use of an asymmetric modulation
method known as Hierarchical Quadrature Amplitude
Modulation (HQAM) for the transmission of images over
wireless mobile channels. It is a modification of Quadrature
Amplitude Modulation (QAM) and provides unequal error
protection (UEP). This is a simple and efficient approach in
which non-uniform signal-constellation is used to give
different degrees of protection to the transmitted bits. The
advantage of this method is that different degrees of protection
are achieved without an increase in bandwidth in contrast to
channel coding that increases the data rate by adding
redundancy to the transmitted signal [2, 7-9]. A comparison of
HQAM and QAM is carried out in order to bring out the
advantages of HQAM over QAM for transmission of image
data over wireless erroneous channels. BER analysis is
presented for different values of the modulation parameter and
performance comparison is carried out through computer
simulation using 16-HQAM and 16-QAM techniques with
gray image as test image.
The paper is organized as follows: In Section II, general
model of image transmission system is briefly described.
Section III considers the effects of noise and distortion in
image transmission. In Sections IV & V, an overview of QAM
and HQAM is given. Based on computer simulation results,
the performance of HQAM is considered in Section VI.
II. MODEL OF IMAGE TRANSMISSION SYSTEM
The essentials of the image transmission system considered
here are shown in Fig. 1. The source encoder encodes the
source image using appropriate image compression technique.
For the protection of coded image in Fig. 1 channel encoder
add redundancy to the coded image by using appropriate
channel coding technique. Modulator modulates the coded
image and transmits through wireless channel. QAM is
invariably used as the modulation technique [14]. The channel
introduces noise and distortion to the transmitted image. The
demodulator receives the image data with error and
demodulates it. After channel decoding, the coded image is
decompressed.
There are two major constraints in transmission of images
over wireless channels. First there are fluctuations in the
channel bandwidth for this reason the image data must be
compressed. Second, there is a high probability of channel
Hierarchical Quadrature Amplitude Modulation
For Image Transmission Over Erroneous
Wireless Channels
Farid Ghani, Md. Abdul Kader, and R. Badlishah Ahmed
F
International Journal of Video & Image Processing and Network Security IJVIPNS-IJENS Vol: 11 No: 06 2

112106-6868 IJVIPNS-IJENS December 2011 IJENS
I J E N S
error for this reason the image data must be protected from
errors in order to maintain image quality.


Fig. 1. Model of image transmission system
III. EFFECTS OF NOISE AND DISTORTION IN IMAGE
TRANSMISSION
The most common type of noise that is encountered in
image transmission system is the Salt and pepper noise
(special case of impulse noise) [12, 13], where a certain
percentage of individual pixels in digital image are randomly
digitized into two extreme intensities (maximum and
minimum). Faulty memory locations or transmission through
erroneous channels can result in the received image being
corrupted with this type of noise [8]. The effect of salt and
pepper noise on the received image is shown in Fig. 2(a).


(a)



(b)


(c)



(d)

Fig. 2. (a) Effects of Salt & pepper noise (b) Effects of error in critical bits
(c) Effects of error in sign bit (non-critical bit)

Distortion in images occurs when errors cause local
variations in image scale and coordinate location of the image
pixels. The distortion is more severe when the errors occur in
critical (significant/high priority) bits of the received signal.
For errors in non-critical bits (sign bits/low priority bits and
refinement bits/low priority bits) the distortion is not that
severe. This can be seen from Fig. 2(b), 2(c) and 2(d). Fig.
2(b) corresponds to the case when the errors are in significant
bits while Fig. 2(c) and 2(d) result when the error is in sign or
refinement bits. Thus in image transmission it is necessary to
give more protection to significant bits as compared to the
insignificant bits rather than giving equal protection to all the
bits.
IV. QUADRATURE AMPLITUDE MODULATION
Quadrature Amplitude Modulation (QAM) is a big name for
a relatively simple technique. For a given channel bandwidth,
QAM transmits more information as compared to BPSK and
QPSK. However, it is more susceptible to noise because the
states are closer together so that a lower level of noise is
needed to move the signal to a different decision point.
Receivers for use with phase or frequency modulation are both
able to use limiting amplifiers that are able to remove any
amplitude noise and thereby improve the noise reliance. This
is not the case with QAM. When a phase or frequency
modulated signal is amplified in a transmitter, there is no need
to use linear amplifiers, whereas when using QAM that
Fading and
Noise
Channel
with Errors
Received Data
with Error
Compressed
Data

Transmitted
Image
Channel
Encoder
Modulator
Source
Encoder
Compression
De-
Modulator
Channel
Decoder
Received
Image
Source
Decoder
De-Compression
International Journal of Video & Image Processing and Network Security IJVIPNS-IJENS Vol: 11 No: 06 3

112106-6868 IJVIPNS-IJENS December 2011 IJENS
I J E N S
contains an amplitude component, linearity must be
maintained. Unfortunately linear amplifiers are less efficient
and consume more power, and this makes them less attractive
for mobile applications [3].
The signal constellation diagram for 16-QAM is shown in
Fig. 3(a). It can be seen from the Figure that QAM provides
equal error protection to the transmitted bits by assigning
equal priority to both the significant and non-significant bits of
data; hence it is classified as equal error protection (EEP)
method of modulation [6, 11]. This, however, is not desirable
in case of image transmission where unequal error protection
(UEP) is needed. Performance can be increased even more by
putting extra protection on the highly sensitive bits than low
sensitive bits using unequal error protection (UEP) schemes.
Hierarchical QAM which is a modification of QAM, has the
property of providing un-equal protection and can, therefore,
be used to advantage in transmission of images over wireless
erroneous channels. In this paper, a modified form of QAM
known as Hierarchical Quadrature Amplitude Modulation
(HQAM) is proposed. HQAM provides UEP to the transmitted
image data and is described in the following Section.
V. HIERARCHICAL QUADRATURE AMPLITUDE MODULATION
Hierarchical Quadrature Amplitude Modulation (HQAM) is
more spectrally efficient and dc-free modulation scheme [10].
It provides different degree of protection to the transmitted
data bits, in which the high priority (HP) data bits are mapped
to the most significant bits (MSB) and the low priority (LP)
data bits are mapped to the least significant bits (LSB) of the
modulation constellation points. Using HQAM will, therefore,
result in improved image quality compared with QAM
specially at low channel SNR conditions, since the highly
sensitive HP data bits are mapped to the MSBs with low bit
error rate (BER) in HQAM. For the sake of simplicity only 16-
HQAM is considered in this paper.


(a)

(b)


(c)

Fig. 3. Constellation diagram (a) 16-QAM (b) 16-HQAM for = 2, k=9 and
(c) 16-HQAM for = 1, k=9

In Hierarchical QAM, it is possible to give higher protection
to the most important data (significant bits) by changing the
value of modulation parameter . is the ratio of the distance
b between quadrants to the distance c between the points
within a quadrant in the constellation diagram with b/2+c
being a constant k. Fig. 3(b) shows the constellation diagram
for =2 i.e. b=2c and k=9. Moreover, the value of should not
exceed the square root of the carrier power p
c
, otherwise, the
constellation points of the same quadrant will overlap. When
= 1, i.e. b=c HQAM results in QAM as can be seen from Fig.
3(c) for k=9.
Referring to Fig. 3(c), the two bits represent the HP bits
which have lower BER than the last two bits. Last two bits are
representing LP bits. It can be seen that the four symbols in
every quadrant have the same HP bits but different LP bits;
this is also called constellation overlapping and ensure that the
HP bits to be transmitted correctly [14].
When increases, the distance of constellation points

1001 1011 0001 00 11

1000 1010 0000 0010

1101 1111 0101 0111

1101 1111 0100 0110

9



3



-3




-9
Q
-9 -3 3 9
I

HP LP

1001 1011 0001 0011

1000 1010 0000 0010

1101 1111 0101 0111

1101 1111 0100 0110
Q
9
4.5
-4.5
-9
c b
-9 -4.5 4.5 9
I





Q

I

International Journal of Video & Image Processing and Network Security IJVIPNS-IJENS Vol: 11 No: 06 4

112106-6868 IJVIPNS-IJENS December 2011 IJENS
I J E N S
between the quadrants increases and within a quadrant
decreases. Constellation diagrams are shown in Fig. 3(b) when
= 2 and Fig. 3(c) when = 1. It may be noted that by
increasing the degree of non-uniformity (b > c) the
improvement of the HP performance is significant, at the
expense of the LP sub-channel. This is a simple UEP method
without introducing any redundancy to the modulated data.

VI. PERFORMANCE COMPARISON OF 16-HQAM AND 16-
QAM
A simulation for gray scale image transmission and
reception was carried out using 16-QAM and the proposed 16-
HQAM. Results from 16-HQAM were then compared with
those obtained with 16-QAM. The flow diagram of the
proposed simulation is shown in Fig. 4 and simulated using
MATLAB 7.8 (R2009a). This simulation calculates the BER
of the image data by comparing the HP and LP bit streams
before modulation and after demodulation.


Fig. 4. Simulation of image transmission and reception using 16-HQAM

Fig. 5 shows the BER vs. channel SNR (for AWGN
channel) for HP and LP data using hierarchical 16-QAM for
= 2 to 5 and without hierarchical mode (16-QAM) where =
1. BER of HP and LP data are same for 16-QAM because of
same protection to all the transmitted bits. It is seen that as
expected by increasing the value of , BER of the HP bits
decreases and LP bits increases. However, HP (significant)
data are highly protected for the larger value of .

Fig. 5. BER vs. Channel SNR for HP and LP Data Using 16-HQAM for =
2 to 5 and 16-QAM ( =1)

Gray scale test image House [9] shown in Fig. 6(a) was used
to obtain the performance of 16-HQAM and 16-QAM using
MATLAB. Different values of modulation parameter were
used to see the improvement in the distortion of the received
image with . The SNR was kept at a fixed value of 17 dB.
The results are shown in Fig. 6(b) to 6(d). It can be seen that
image quality of the Fig. 6(c) where = 3 is better and less
corrupted by noise than Fig. 6(b) where = 1. By increasing
the value of , the quality of image in Fig. 6(d) is better than
Fig. 6(c). Fig. 6(d) shows best quality of image in proposed
simulation when = 5.



(a) Original image

Transmitted
Image

Read Bits From
Coded Image File
Add Noise
with
Transmitted
Signal
Two Level Bit
Stream Partitioning
16-HQAM
Modulation





MSBs
LSBs
Signal
Transmission
HP LP
Bit Stream
Merging
16-HQAM
Demodulation





MSBs
LSBs
Signal
Reception
HP LP
Write Bits to Coded
Image File
Received
Image

International Journal of Video & Image Processing and Network Security IJVIPNS-IJENS Vol: 11 No: 06 5

112106-6868 IJVIPNS-IJENS December 2011 IJENS
I J E N S

(b) =1


(c) =3


(d) =5

Fig. 6. (a) Original image and (b) to (d) Decoded images
VII. CONCLUSIONS
In image transmission over mobile channels, QAM provides
higher transmission efficiency by utilizing both amplitude and
phase variations. However, it requires higher transmission
bandwidth due to its equal protection of high and low priority
data bits. Hierarchical QAM (HQAM) overcomes this
disadvantage by providing more protection to the higher
priority bits and less protection to the lower priority bits of the
image data. Thus HQAM being a simple modification of QAM
technique provides a more efficient means of image
transmission over erroneous wireless channel without any
additional hardware.
REFERENCES
[1] Tongtong Li, Huahui Wang, Qi Lang, Source Aware Non-Uniform
Information Transmission for Minimum Distortion Signal Processing
Letters, IEEE, Vol.14, No.2, pp. 85-88, February, 2007.
[2] B.Barmada, M.M.Gandhi, E.V.Jones, M.Ghanbari, Prioritized
transmission of Data Partitioned H.264 Video with HQAM Signal
Processing Letters, IEEE, Vol.12, No.8, pp.577-580, August, 2005.
[3] Fuqin Xiong, Digital Modulation in Digital Modulation Techniques,
2
nd
Edition, 2006.
[4] Seamus OLeary, Hierarchical transmission and COFDM systems,
IEEE transactions on broadcasting, vol.43, no.2, June, 1997.
[5] Chee-Siong Lee, Thoandmas Keller and Lajos Hanzo, OFDM-Based
Turbo-Coded hierarchical and Non-Hierarchical terrestrial mobile
digital video broadcasting, IEEE Transactions on Broadcasting,
vol.46, no.1, March 2000.
[6] Yoong Choon Chang; Sze Wei Lee; Komiya, R.; , "A low-complexity
unequal error protection of H.264/AVC video using adaptive
hierarchical QAM", Consumer Electronics, IEEE Transactions on,
vol.52, no.4, November, 2006.
[7] M. Mahdi Ghandi and M. Ghanbari, "Layered H.264 video transmission
with hierarchical QAM", Elsevier J. Visual Communication, Image
Representation, Special issue on H.264/AVC, vol.17, no.2, pp.451 -
466, April, 2006.
[8] B. Barmada, E.V. Jones, Adaptive mapping and priority assignment for
OFDM, 3G Mobile Communication Technologies, 2002. Third
International Conference on (Conf. Publ. No. 489), pp. 495- 499, May,
2002.
[9] K. K. V. Toh, H. Ibrahim, and M. N. Mahyuddin, Salt-and-pepper
noise detection and reduction using fuzzy switching median filter,
IEEE Trans. Consumer Electron, vol.54, no.4, pp. 19561961,
November, 2008.
[10] Mirabbasi, S.; Martin, K.;, "Hierarchical QAM: a spectrally efficient dc-
free modulation scheme," Communications Magazine, IEEE, vol.38,
no.11, pp. 140- 146, November, 2000.
[11] John G. Proakis, Digital Communications Through Fading Multipath
Channels in Digital Communications, 3rd Edition, 2000.
[12] S. Zhang and M.A. Karim, A new impulse detector for switching
median filter, IEEE Signal Processing Letters, vol.9, no.11, pp.360-
363, Nov., 2002.
[13] W. K. Pratt, Image Enhancement in Digital Image Processing, 4
th

Edition, Wiley, New York, 1978.
[14] L. Hanzo, W. Webb, T. Keller, Basic QAM Techniques in Single and
multi carrier quadrature amplitude modulation: principles and
applications for personal communications, WLANs and broadcasting,
John W. & S., 2000.