Cyclostationary impulsive noise mitigation algorithm for narrowband powerline communications

The dominant noise in narrowband powerline communication (NB-PLC) systems is cyclostationary impulsive noise (IN), which severely degrades communication performance. Conventionally, a frequency shift (FRESH) filtering approach exploits the cyclostationary effect to extract the IN component in the received signal. In this paper, we propose a three-stage adaptive FRESH-based cyclostationary noise estimator to make a reliable IN estimation. Inspired by the active noise cancellation method, we firstly employ a FRESH filter to generate a reference IN signal. In addition, considering the cyclostationarity of the IN, we exploit the time-averaged objective function to adjust the coefficients of the FRESH filter. Then, this reference IN signal is fed into an adaptive estimator to further refine the estimation of IN. Lastly, the refined IN estimations are rectified by a hard thresholding device such that significant and reliable IN estimations can be leached out. By subtracting the reliable IN estimations from the received signal, we mitigate the impact of IN on the data transmission for NB-PLC systems. Computer simulation is conducted based on the IEEE 1901.2 specification and we use real measured IN data to verify the effectiveness of our proposed method. The results have shown that the improvement in terms of the normalized mean squared error can be up to 7.57 dB. (C) 2019 Published by Elsevier Ltd on behalf of The Franklin Institute.