An improved SAMP scheme for sparse OFDM channel estimation

Accurate channel estimation is an important prerequisite to ensure the performance of OFDM communication systems. Conventional pilot-based schemes such as Least Square (LS) and Minimum Mean Square Error (MMSE) suffer from low spectrum utilization. Taking advantage of the channel sparsity, compressed sensing (CS) is an effective method for estimating sparse channels. In this paper, an improved sparsity adaptive matching pursuit (SAMP) method is proposed. It optimizes the setting of initial sparsity and iteration threshold. The simulation results show that the proposed method shares better performance in BER and MSE than SAMP.