VOL. 14, NO. 21, NOVEMBER 2019 ISSN 1819-6608 ARPN Journal of Engineering and Applied Sciences ©2006-2019 Asian Research Publishing Network (ARPN). All rights reserved. www.arpnjournals.com 3694 IMPULSIVE NOISE REDUCTION IN POWER LINE COMMUNICATION USING ADAPTIVE FORWARD ERROR CORRECTION FILTER Joliz Anton 1 , Nair Madhavan 1 , Bok-Min Goi 1 , Ezra Morris 1 and Mohanaprasad Kothandaraman 2,3 1 Lee Kong Chian Faculty of Engineering, Universiti Tunku Abdul Rahman, Malaysia 2 SENSE, VIT University, Chennai, India 3 Post-Doctoral Research Fellow, Lee Kong Chian Faculty of Engineering, Universiti Tunku Abdul Rahman, Malaysia E-Mail: madhavan@utar.edu.my ABSTRACT In this article, noise reduction in power line communications (PLC) was studied with the aim to improve its bit error rate (BER) and throughput performance by mitigating the harmful effects of impulsive noise. To this end, a noise reduction technique based on hybrid forward error correction (FEC)-Least Mean Square (LMS) filtering is proposed. Simulation results show that the proposed hybrid FEC-LMS method shows significant improvement over the conventional FEC method and adaptive LMS filtering method in terms of BER and throughput. Keywords: adaptive least mean squares filter, forward error correction, impulsive noise, medium access control, network simulator-3, power line communication. 1. INTRODUCTION Broadband over power-lines has emerged as a promising technology with the advantage of using existing power-line networks for data transmission which potentially provides a huge cost reduction [1-5]. Such technology, also known as power-line communication (PLC), is capable of delivering data at a rate of up to 1 Gbps [6]. Moreover, PLC has been regarded as one of the key enabling technologies for the Internet of Things (IoT), where many devices can connect and communicate with each other over PLC networks. However, its high data rate is inhibited by factors such as attenuation and interference caused by various types of noises [7] [8]. The presence of these noises in PLC networks can result in significant performance degradation [9]. The noise can be categorized into two: background noise and impulsive noise. Each of these two noises can further be broken down into categories as shown in Figure-1. It was found out that impulsive noise can cause stronger interference to PLC networks as it had been shown that man-made noise and atmospheric noise demonstrate impulsive behaviours with heavy tail characteristics [10]. Therefore, an in-depth analysis of impulsive noise is needed to understand its behaviour and to improve the performance of PLC networks [11]. In [12], forward error correction (FEC) had been found to be a better noise mitigation technique as compared to other existing noise reduction techniques such as time-domain techniques (e.g., clipping, blanking, nulling), time/frequency-domain techniques, Bayesian learning techniques, recursive detection techniques. In [13], it had been shown that adaptive noise reduction techniques using LMS algorithm outperform the existing notch filter to suppress periodic impulsive noise. Figure-1. Types of noises. The existing FEC technique can be used to filter the aperiodic impulsive noise partially. To perform the effective aperiodic impulsive noise cancellation there is a need for an adaptive algorithm. In this article, we propose a hybrid FEC-LMS technique to mitigate the impulsive noise problems in PLC networks. The combination of FEC Noise Background noise Colored noise Aperiodic impulsive noise Synchronou s periodic impulsive noise Asynchrono us periodic impulsive noise Impulsive noise Narrow band noise