Performance Analysis of Nonlinear Pre-Processors for Markov-Middleton Impulsive Noise Mitigation in OFDM Based Power-Line Communications

Abstract

Power line communication (PLC) is a hostile communication medium and is distinctively affected by burst impulsive noise which prevails from electrical appliances in the PLC system. Nevertheless, impulsive noise models replicating a burst structure are not widely used in literature as opposed to canonical models such as Middleton Class A and others. Hence, this work focuses on a bursty impulsive noise model, Markov-Middleton which is established by using Markov chain representations and parameters derived from its reference statistical-physical Middleton Class A impulsive noise model. Moreover, since burst impulsive noise causes significant performance degradation in PLC systems, it is worthwhile to apply mitigation techniques to reduce the interferences caused at the receiver side of the PLC system. Consequently, the aim of this work is to investigate the performance of noise suppression techniques such as Clipping and Blanking nonlinear preprocessors in mitigating Markov- Middleton impulsive noise in a convolutionally coded multipath OFDM PLC system. Results show that for a moderate impulsive case scenario used, the mitigation techniques are more effective at low Eb/No, especially Blanking technique which yields an Eb/No gain of 7.13 dB at BER for Markov-Middleton noise mitigation.