Impulsive Noise Cancellation for MIMO-OFDM PLC Systems: A Structured Compressed Sensing Perspective

In this paper, a novel impulsive noise (IN) cancel- lation scheme based on structured compressed sensing (SCS) for multiple input multiple output orthogonal frequency division multiplexing (MIMO-OFDM) power line communication systems is proposed. The SCS theory is introduced to IN recovery for the first time in this paper to the best of the authors' knowledge, and the gap of lack of research on the IN mitigation for MIMO PLC systems is filled. First, the measurements matrix of the IN is obtained, and the SCS optimization framework is formulated through the proposed spatially multiple measuring method, by fully exploiting the spatial correlation of the IN signals at different receive antennas. To efficiently reconstruct the IN signal, an enhanced SCS-based greedy algorithm, structured a priori aided sparsity adaptive matching pursuit (SPA-SAMP), is proposed, which significantly improves the accuracy and robustness compared with the state-of-art methods. Theoretical analysis and computer simulations validate that the proposed scheme outperforms conventional methods in the typical MIMO PLC system.