Sparse channel estimation for MIMO-OFDM systems using compressed sensing

Abstract

One of the major challenge for practical Multiple Output-Orthogonal Frequency Division Multiplexing (MIMO-OFDM)system is the accurate channel estimation which is very essential to guarantee the system performance. In this paper, the Subspace Pursuit (SP), Orthogonal Matching Pursuit (OMP) and Compressed Sampling Matching Pursuit(CoSaMP) techniques combined with Minimum Mean Square Error(MMSE) and Least Mean Square(LMS)tools are used to estimate the channel coefficients for MIMO-OFDM system. These algorithms are used for the channel estimation in MIMO-OFDM system to develop the joint sparsity of the MIMO channel. Simulation results shows that SP, OMP and CoSaMP techniques combined with MMSE and LMS tools provides significant reduction in Normalized Mean Square Error (NMSE) vs Signal to Noise Ratio (SNR) when compared to SP, OMP and CoSaMP technique with Least Square (LS) tool and also the conventional channel estimation methods such as LS, MMSE and LMS. Moreover CoSaMP combined with LMS tool performs better than SP and OMP techniques with LMS tool with less computational time complexity.