Performance analysis of OFDM mmWave communications with compressive sensing based channel estimation and impulse noise suppression

Millimeter Wave (mmWave) channel estimation can be treated as a sparse signal recovery problem due to the sparse multipath characteristics of the channel. Utilizing compressive sensing (CS) theory, sparse signal recovery algorithms can be designed for channel estimation. However, these algorithms are sensitive to impulse noise, and hence the estimation accuracy degrades under non Gaussian noise. In this paper, we propose a novel algorithm for suppression of impulse noise to improve the system performance. The CS based channel estimation algorithms using pilot subcarriers of orthogonal frequency division multiplexing (OFDM) symbols in conjunction with the proposed impulse noise suppression method is utilized to evaluate the system performance over 60 GHz sparse channels. The mean square error (MSE) and bit error rate (BER) are considered as performance metrics. Comparison with the least squares (LS) estimation method using 4 quadrature amplitude modulation (QAM) OFDM symbols proves the efficacy of the proposed technique.