See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271502431

Impulsive noise reduction using adaptive receiver structure technique

Conference Paper · October 2012

DOI: 10.1109/ICoSP.2012.6491614

CITATIONS READS

10 499

4 authors, including:

Shihab Jimaa Saleh R. Al-Araji

Khalifa University 136 PUBLICATIONS 761 CITATIONS
72 PUBLICATIONS 728 CITATIONS
SEE PROFILE
SEE PROFILE

Tetsuya Shimamura
Saitama University
288 PUBLICATIONS 1,765 CITATIONS

SEE PROFILE

All content following this page was uploaded by Shihab Jimaa on 05 February 2018.

The user has requested enhancement of the downloaded file.

ICSP2012 Proceedings

Impulsive Noise Reduction using Adaptive

Receiver Structure Technique

S. A. Jimaa, S. R. Al-Araji, and A. Al-Kaabi T. Shimamura

ECE Department, College of Engineering Department of Information and Computer Science
Khalifa University of Science, Technology, and Research Faculty of Engineering, Saitama University
Sharjah Campus, U.A.E. Japan
Email: {saj, alarajis}@kustar.ac.ae} Email: {[email protected]}

Abstract---This paper presents the design and simulation

of an Adaptive receiver structure using the Normalized
Least-mean-square (NLMS) for impulsive noise reduction
in Additive White Gaussian Noise (AWGN) based
communication systems. The NLMS algorithm is applied
for the adaptive control mechanism in order to reduce the
impact of impulsive noise. An AWGN system model that
uses BPSK, QPSK and 16-QAM modulation schemes was
designed, simulated, and tested. The Bit-Error-Rate
(BER) performance of the proposed system with and
without Impulsive noise was evaluated and compared
with the conventional receiver. The proposed system has
shown to give improved performance under the same
impulsive noise environment.

Keywords- Impulsive noise; NLMS; adaptive receivers;

I. INTRODUCTION
One of the main receiver’s limitations in making correct
Amplitude

symbol decisions is the presence of noise with the desired

Pulse Amplitude
signal. This unwanted electrical signal is always present in
electrical systems in the form of natural or manmade which Pulse
Duration
are generated from different sources. The presence of noise Inter-Arrival
Time
superimposed on a signal tends to obscure or mask the signal
[1]. The natural noise is generated by atmosphere and cosmic
Time
radiation such as sun noise. It has Gaussian distribution
which means that cannot be eliminated from the system [1, Figure1. An impulsive noise waveform and basic Impulsive noise
baseband model [4]
2]. However, the manmade type is non-Gaussian distribution
such as impulsive noise (IN) that can be eliminated and
removed from the system by using different techniques [3]. Several techniques [4-9] have been investigated to reduce
Impulsive noise (IN) is manmade noise and is usually the effect of IN. Some of these techniques involve the use of
generated by switching transients and other radiated a clipping to clip the received signal amplitude and thus
preventing very high amplitudes related to the IN to pass to
electromagnetic signals. IN can be analyzed in time or
the demodulator stage. Other techniques use gating to reduce
frequency domains. It has a flat spectrum in the frequency
the IN impact.
domain which makes it difficult to analyze. However, it is
easier to analyze and detect in the time domain. Fig. 1 shows These techniques were able to reduce the effect of IN but
a typical waveform of an impulsive noise in time domain [4]. at the cost of introducing harmful distortions effect to the
___________________________________ received signal [4]. In this paper, an adaptive receiver
978-1-4673-2197-6/12/$31.00 ©2012 IEEE


structure technique for impulsive noise reduction is proposed. III. THE PROPOSED METHOD
This technique uses the Normalized Least-mean-square The proposed method of using the NLMS adaptive
(NLMS) adaptive receiver structure to reduce the effect of receiver in impulsive noise reduction is shown in Fig. 4.
impulsive noise in an AWGN communication systems
environment. The BER performance is used to compare
between the proposed and the conventional methods.
Antenna

Following the above introduction, the rest of the paper is

organized as follows; Section 2 presents the IN channel Receiver Front
   
NLMS Adaptive    


O/P
End Filter   
model, Section 3 discusses the proposed IN reduction
receiver, Section 4 shows the results achieved through
_
simulation, and finally, Section 5 presents the conclusions. Adaptive Error

Algorithm
+
II. MODEL OF IMPULSIVE NOISE GENERATOR Reference Signal

Figure 4. IN reduction using the adaptive receiver structure

The block diagram of the communication system used in
this paper, to evaluate the impact of IN on an AWGN
communication systems’ performance, is shown in Fig. 2. First the received signal goes through the RF & IF stages
[10]. and then demodulated. The demodulated baseband signal is
fed into the adaptive system which consists of both the
adaptive filter and the adaptive algorithm. The output of the
Impulsive Noise filter is subtracted from the reference signal to obtain the
Channel Structure
error which is used by the NLMS adaptive algorithm to
update the filter coefficients. This consequently results in
reducing the error.
O/P1 O/P2 O/P4 O/P5

The reference signal used here is assumed to be the

Uniform Random
Number
Relay
X Zero-Order Hold

training sequence which is available at the receiver. This

signal will be the same as the transmitted BPSK, QPSK, or 16
O/P3

Random Number
+ QAM. The adaptive filter used here is a linear feed-forward
transversal filter with a length of w taps. The process of
Input updating the filter coefficients is explained in what follows:
Impulsive Noise Generator (ING)

Output The filter output is obtained as follows:

y n = Wn X n
Figure 2. Impulse Noise Generator T
(1)
Also shown in the figure the model used to test the
effectiveness of the proposed adaptive receiver in reducing Where xn is the filter input and wi, (i=0, 1, …, M-1) is the
the IN. The IN generator includes data source, zero-order weight vector which corresponds to the filter length. The
hold, and sign as a comparator. To generate IN with fixed
error signal e(n), involved in the adaptive process, is defined
amplitude and variable width, the output is multiplied with a by
random number, see Fig. 3.
e(n) = d (n) – y (n) (2)

= d (n) – wH(n) x (n) (3)

The NLMS algorithm adapts the tap-weights sequence

using a gradient descent algorithm that reduces the squared
estimation error at each time instant. The algorithm is given
below:

x[n]
w[n + 1] = w[n] + μ en (4)
λ + || x[n] || 2
Where µ is the algorithm’s step-size and λ is a very small
number added to avoid instability, and en is the error signal
Figure 3. ING output signal
which is defined in equation (3).


 IV. SIMULATION RESULTS
The effects of the mean amplitude and likelihood
probability of IN on the performance of BPSK, QPSK and
16-QAM modulated signals were investigated using
MATLAB / SIMULINK simulation tool. The results indicate
that IN has a significant effect on the BER performance of the
received signal as shown in Figs. 5 & 6. It can be seen that
the BER performance of the three modulation schemes
deteriorates as the mean amplitude of IN increases. Also
increasing the likelihood probability of IN degrades the BER
performance. The difference in BER performances in Fig. 5 is
noticable between IN mean amplitudes of 0.1 and 0.8 while
in Fig. 6 the difference in BER performances is noticable
between IN likelihood probabilities of 0.1 and 0.7.

Figure 7. The effect of step-size on BER Performance for BPSK

Figure 5. The Effect of changing the mean amplitude of

IN on BER performance of BPSK, QPSK, and 16QAM

Figure 8. The effect of step-size on BER Performance for QPSK

Figure 6. The effect of changing the likelihood probability of

IN on BER performance of BPSK, QPSK, and 16QAM

Also the effect of varying the step-size of the NLMS

Figure 9. The effect of step-size on BER Performance for 16QAM
algorithm on the BER performances of BPSK, QPSK, and 16
QAM at an SNR level of 20 dB were investigated and the


 V. CONCLUSION
simulation results are shown in Figs. 7-9. The main aim of In this paper an adaptive receiver structure for impulsive
this investigation is to find the value of the step size that noise reduction, that was designed and simulated in an
gives lower MSE. From these results, it can be noticed that AWGN based communication system, is presented. The
the value of those step-sizes were 0.0001, 0.001, and 0.5 for performance improvements in the NLMS adaptive receiver
BPSK, QPSK, and 16-QAM, respectively. were demonstrated through the simulation results presented in
this work. The proposed system was simulated and tested
After choosing the step-sizes with lower error rate, the using MATLAB/SMULINK software tool. According to the
BER against SNR performance for the 16 QAM modulation impulsive noise model parameters, the mean amplitude of the
scheme has been investigated using the conventional and the impulsive noise was found to be more effective on the BER
proposed adaptive NLMS receiver. Also the impact of IN on performance than its Likelihood probability for BPSK, QPSK
the BER performance was shown, see Fig. 10. It is clear that and 16-QAM modulation schemes. This paper also includes
the performance was degraded by more than 2 dB at BER of the results of investigating the value of the step-sizes that
10-2 after adding the IN to the AWGN channel. However, the give lower BER in BPSK, QPSK and 16-QAM. Finally, the
BER performances for the receiver using the NLMS adaptive results of the BER performance against the SNR for the 16-
filter improved by more than 1.5 dB at BER of 2x10-3 when QAM indicate that the impact of impulsive noise is
compared with the conventional receiver under the same minimized when using the NLMS receiver compared with the
conditions. conventional receiver.

REFERENCES
[1] B. Sklar, Digital Communications Fundamentals and Applications,
Prentice Hall PTR, 2nd Ed., Oct, 2001.
[2] D. Kim, J. Roh, S. Cho ,and J. Kim, “Effects of Impulsive Noise and
Self Co-channel Interference on the Bluetooth Scatternet”, IEICE
Trans. Comms., Vol.E85-B, No.10, Oct 2002.
[3] S. V. Zhidkov, “Impulsive Noise Suppression in OFDM Based
Communication Systems”, IEEE Transactions on Consumer
Electronics, Vol. 49, No. 4, NOVEMBER 2003
[4] M. S. McLaughlin, W. Henkel, R. Kirkby, and T. Kessler,“Impulse
generation with appropriate amplitude, length, inter-arrival, and
spectral characteristics”, IEEE J. Select Areas Commun., Vol. 20,
pp. 901-912, June 2002
[5] S. R. Al-Araji, M. A. Al-Qutayri, and M. S. Al-Tenaiji, “Impulsive
noise Reduction using Auto-Gating technique”, IEEE-GCC
Conference, UAE, Feb. 19-22, 2011
[6] K. Al-Mawali, A. Hussain, and K. Sadik,” Joint Time/Frequency
domains Impulsive Noise Reduction in OFDM based PLC, ATNAC
conference, pp 138-142, Australasian, 7-10 Dec. 2008.
[7] Smolka, B.,” Adaptive technique of impulsive noise removal in
color images”, ICCOMP’10 Proceedings of the 14th WSEAS CSCC
conference, Vol. II, Wisconsin, USA, 2010
[8] M. Celebi, H. A. Kingravi, and Y. A. Aslandogan,” Nonlinear vector
filtering for impulsive noise removal from color images”, Journal of
Electronic Imaging, Vol. 16, No. 3, 2007
[9] T. Kitamura, K. Ohno, and M. Itami,” Iterative impulsive noise
reduction by generating its replica signal in OFDM reception, 2011
IEEE Int. conference on Computer Electronics (ICCE), pp 389-390,
Las Vegas, NV, USA, 9-12 Jan. 2011
Figure 10. BER comparison of 16-QAM in AWGN channel [10] S.R. Al-Araji, M. A. Al-Qutayri, K. BelHaj, and N. Al-Shwawreh,
“Impulsive Noise Reduction Techniques Based on Rate of
Occurrence Estimation, Int. Symp. On Signal Processing and its
Applications 2007 (ISSPA 2007), Sharjah, UAE, February 12-15,
2007.



View publication stats