Robust threshold compressed sensing based sparse multipath channel estimation in OFDM system

In this paper, the Compressed Sensing (CS) based channel estimation scheme for the sparse multipath channel in OFDM system is considered. The CS based multipath delay channel model is investigated by using only a small amount of pilots to get the measurement matrix and the orthogonal matching pursuit (OMP) is adopted to reconstruct the channel impulse response. In order to overcome the drawback of not knowing the optimal number of iterations for OMP which reflects the number of optimal significant taps of a sparse multipath channel, this paper proposes a robust threshold CS-OMP based sparse channel estimation scheme, which uses noise coefficients among the detail wavelet coefficients to get a robust threshold, and achieves an effective channel estimation performance. Simulations show that the proposed method can significantly improve the performance which approximates closely the performance of optimal iteration number for CS-OMP based channel estimation.