RIS-Aided Wireless Sensor Network in Presence of Bursty Impulsive Noise for Smart-Grid Communications

Wireless sensor networks (WSNs) in smart grid (SG) applications are greatly affected by the detrimental impact of bursty impulsive noise (IN) generated by various partial discharges from aging power equipments in Smart Grid’s power substations. Additionally, the deleterious impact of fading on the radio frequency (RF) links between the sensor nodes further deteriorates the performance. To alleviate the problem of fading on the RF links, in this paper, we propose to exploit reconfigurable intelligent surfaces (RISs). In addition, to deal with the bursty IN, we consider the maximal a posteriori (MAP) decoding of the received symbols using the Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm. The performance of the considered system is evaluated by deriving a novel closed-form expression for the bit error rate (BER). Furthermore, an asymptotic BER analysis is presented in order to highlight the diversity order achieved by the considered RIS-aided system. Finally, extensive numerical results are provided to demonstrate the significant gain that can be achieved through the proposed RIS-aided architecture for WSNs.

I. Introduction

WIRELESS SENSOR NETWORKS (WSNs) have recently gained much attention in various smart grid (SG) applications owing to their low-cost, flexibility, wide coverage, self-organization, and rapid deployment [1]. In this context, the WSNs usually consist of a large number of low-cost, low-power, and multi-functional sensor nodes used for monitoring the SG. Such monitoring helps to diagnose and rectify problems that arise and may hinder the SG functionality. However, various aging power equipment in the SG, such as transformers, busbars, circuit-breakers, and switches generate impulsive noise from various partial discharge sources that can affects the wireless links between the sensor nodes of the WSNs and thus, can drastically degrade the overall communication performance [2]. Memoryless models, such as the Middleton class-A [3] and Bernoulli-Gaussian [4] have been mostly used in the literature to represent the impulsive noise. However, these models do not capture the bursty nature of the impulses, i.e., the correlation among the noise samples in the time domain. In order to handle the correlation among the noise samples, Markov chain models have been proposed in the literature [5], [6]. Considering a two-state Markov Gaussian (TSMG) model for the characterization of bursty impulsive noise (IN), in [7], the authors analysed the performance of decode-and-forward (DF) cooperative relaying in the presence of bursty IN. In addition, the authors showed that the optimal maximal a posteriori (MAP) decoding of the received symbols considering the Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm performs significantly better than the conventional receivers developed for the additive white Gaussian noise (AWGN) and memoryless IN channels. In [8], the detrimental impact of bursty IN in non-orthogonal multiple access (NOMA) systems is mitigated by considering a MAP receiver-based approach. In [9], the authors proposed improved modeling of the correlation between continuous-valued sources in low density parity check (LDPC)-based distributed source coding (DSC). The authors in [10] analysed distributed source-channel coding based on real-field Bose-Chaudhuri-Hocquenghem (BCH) Codes. Although [7], [8] proposed the BCJR algorithm based optimal MAP receiver in order to mitigate the detrimental impact of bursty IN, the underlying work ignored the detrimental impact of fading on the radio frequency (RF) links between the senor nodes.