Adaptive impulsive noise mitigation using Multi Mode Compressive Sensing for powerline communications

We propose an adaptive algorithm based on Compressive Sensing (CS) to mitigate impulsive noise (IN) from Orthogonal Frequency Division Multiplexing (OFDM) based systems. In recent CS literature on IN mitigation, the sparsity of IN is assumed to be constant and thereby the number of pilots to reconstruct IN is also fixed. However, in practical systems especially in Power Line Communications (PLC) the sparsity of IN changes significantly over time. This limits the application of traditional CS algorithms. Therefore, we develop a Multi Mode Compressive Sensing (MMCS) scheme that adaptively changes the number of pilots used to reconstruct the IN depending on the IN's current severity. The scheme not only stabilizes the Bit Error Rate (BER) in stern IN conditions but also improves overall system throughput.