Performance of polar coding for the power line
communications in the presence of impulsive
noise
ISSN 1751-8628
Received on 21st February 2015
Revised on 12th June 2015
Accepted on 17th July 2015
doi: 10.1049/iet-com.2015.0141
www.ietdl.org
Liping Jin
1,2
, Youming Li
1
✉
, Bin Li
3
, Zhuanghun Wei
3
, Jiong Shi
2
1
College of Information Science and Engineering, Ningbo University, Ningbo 315211, People’s Republic of China
2
School of Electronic and Information Engineering, Zhejiang Wanli University, Ningbo 315100, People’s Republic of China
3
Beijing University of Posts and Telecommunications, Beijing 100876, People’s Republic of China
✉ E-mail: liyouming@nbu.edu.cn
Abstract: In this study, the authors propose a polar coding (PC) scheme for the power line communication (PLC) system to
cope with the impulsive noise and thereby promote the transmission performance. This new error-correcti ng coding
scheme is essentially inspired on a novel conception of channel polarisation. To b e specific, via recursively channel
combing and splitting, a g roup of chann els with i deal transmission co nditions, that is having a capacity of 1, will be
constructed to carry the useful information, while the other band sub-channels bear useless information. The decoding
performance of PC under realistic impulsive noises is investigated under the condition that the impulsive noise is
modelled by a well-known Middleton Class-A model. To mitigate the error propagation caused by sudden strong
impulsive noises and further enhance the decoding performance, a matrix interleave operator is integrated.
Simulations vali date the suggested PC scheme in PLC systems. Compared with another commonly used low-density
parity-check (LDPC) coding scheme, the suggested PC scheme, which has the low complexity, can significantly
improve the bit error rate (BER) performance of PLC transmissions with impulsive interference. The PC scheme, as
demonstrated by simulation results, can be of great importance to practical PLC systems.
1 Introduction
Power line communication (PLC), which transmits data information
over widely distributed power delivery networks, has drawn
extensive interests in recent years. It, hence, provides a convenient
and inexpensive communication medium [1–3]. A predominant
advantage of PLC is that its deployment costs could be reduced
significantly, as the information transmission will be implemented at
the same time of conveying electric energy [4, 5], yet without
building extra cabling as in most traditional wire-line or wireless
communication infrastructures. As is a double-edge sword,
nevertheless, PLC in practice will also face several major challenges
which are aroused from realistic adverse conditions in power lines,
such as frequency-selective fading [6] and impulsive noises [7].
For the commonly encountered frequency-selective fading,
various estimation schemes (e.g. the data-aided or blind estimation
techniques) can be employed to recover the unknown channel
response. As a consequence, the fading effects can be effectively
eliminated, in either high-speed orthogonal frequency division
multiplexing (OFDM) [8] or low rate single-carrier (SC) systems
[9]. For the second issue, that is impulsive noise, however, the
situation is more complex and challenging. The various estimation
schemes listed above will become impractical to impulsive noises,
which will degrade the data transmission performance
significantly. Therefore, one of the most important issues in
realistic PLC systems is how to effectively combat the negative
effects of impulsive noises and enhance the receiving performance.
Various methods (e.g. blanking, clipping or deep clipping etc.)
have been proposed to detect impulsive noise and further remove
it [10–12]. However, as the duration of impulsive noises is
extremely short and the complexity of detection methods is
relatively high, it is very difficult to detect them timely in practice.
Channel coding is considered as an effective approach to cope
with various noises and thereby promote the data detection
performance, which has been widely recommended in PLC
systems. In [13, 14], the well-known low-density parity-check
(LDPC) is fi rst applied to PLC channels, which is fully
characterised by a parity check matrix [15]. As has been
demonstrated, the LDPC scheme has achieved better performance
when it is further cascaded with a Luby transform (LT) code [16,
17], even in the presence of challenging impulsive noises. The
drawback with LDPC codes, however, is the exponential increase
with code length in implementation complexity.
In this investigation, the polar coding (PC), which has less
encoding and decoding complexity and achieves better
performance, is first applied to PLC systems in the presence of
impulsive noises. The essential idea behind PC, according to its
pioneer Arikan [18], is referred to channel polarisation. Premised
on an iterative and random approach, the independent channels
can be divided into a group of synthesised sub-channels. Some of
them have the ideal transmission conditions with a capacity of
approaching 1, while others have very poor quality with a capacity
of approaching 0. As shown, when the code length is sufficiently
large, almost all these synthesised channels will tend to two
extremes: the noisy channels and those almost free of noise. The
subsequent coding strategy is then obvious, that is the good
sub-channels are then employed in transmitting information, while
other band channels transmit useless message.
Although the applications of PC to mobile communications have
been studied under more commonly additive Gaussian noises [19],
to the best of our knowledge, there is no literature reported
explicitly in the context of impulsive noises. In this study, we
investigate the error-correcting performance of the PC scheme in
PLC and show that the PC scheme can be applied properly to
combat the impulsive interferences, which provides a more
attractive solution to promote the transmission performance in the
context of impulsive noises. As known, there are three essential
ideas behind the channel coding theorem, that is (i) the random
code selection; (ii) the joint asymptotic equipartition property
(AEP) (between the transmitted codeword and the received
sequence); and (iii) the optimal maximal likelihood decoding. It is
suggested that, except for the PC scheme [18], the joint AEP over
IET Communications
Research Article
IET Commun., 2015, Vol. 9, Iss. 17, pp. 2101–2106
2101
&
The Institution of Engineering and Technology 2015