BER Performance Comparison of the Markov-Middleton and Middleton Class A Noise Models in the MIMO Power Line Communication Channel

Abstract

Power Line Communication (PLC) for indoor use has been of significant interest in recent decades. The availability and topology of power lines in every building besides advantages such as broadband data services and no need for new infrastructure, increase the appeal of this technology. However, power lines are a harsh environment for the transmission of data due to signal attenuation, multipath propagation, interference and noise, especially impulsive noise. In this work, an indoor broadband PLC channel is modelled which uses Multiple Input Multiple Output (MIMO) techniques along with OSTBC with an objective to improve the bit error rate performance (BER). For the bursty behaviour of impulsive noise, the Markov-Middleton noise model which has its parameters derived from the Middleton Class A noise model is considered. Both impulsive noise models are implemented in the MIMO PLC channel. It has been found that MIMO along with Orthogonal Space Time Block Codes (OSTBC) is greatly effective in reducing multipath propagation and frequency-selective fading as a gain of14.93 dB at a BER of $10^{-5}$ is observed when compared to the Single Input Single Output (SISO) counterpart. It was also found that Markov-Middleton noise degrades the system slightly more than Middleton Class A noise as the SNR difference between the systems increases from 0.99 dB to 4.04 dB when the impulsiveness increases.