Joint Channel and Impulsive Noise Estimation using Compressive Sensing for Powerline Communications

Channel estimation in conjunction with impulsive noise (IN) mitigation using compressive sensing is proposed for OFDM based power line communication (PLC) systems. Compressive sensing (CS) based IN mitigation algorithms use silent-pilots (or zero-subcarriers) to estimate and cancel the IN at receiver. To estimate the channel and IN jointly, non-zero pilots have to be used. However, the usage of non-zero pilots degrades the IN estimation due to channel frequency response terms appearing at the pilots’ positions. Therefore, we design a method which estimates the channel impulse response using CS, removes the channel frequency response terms appearing at the pilots’ positions and then estimates the IN using CS. The performance evaluation indicates that the proposed method improves the overall Bit Error rate (BER) of 1/2-rate codedOFDM system to almost 4 dB as compared to the ’Least Square’ (LS) method.