An Adaptive Sparse Channel Estimation algorithm with fast convergence for broad band MIMO-OFDM systems

Accurate channel state information estimation in a time variant environment is a significant problem in Multi-Input Multi-Output (MIMO) — Orthogonal Frequency Division Multiplexing (OFDM) systems. Generally channel coefficients of a time variant wireless channel are obtained using Adaptive Channel Estimation (ACE) algorithms. One of the simple and stable ACE methods is based on the Normalized Least Mean Square (NLMS) algorithm. But, it cannot make use of the intrinsic sparsity of broadband MIMO wireless channel. This paper proposes a Variable Step Size-Sign Data Sign-Error Normalized Least Mean Square (VSS-SDSENLMS) algorithm for Adaptive Sparse Channel Estimation (ASCE) which is based on the application of sparse penalties to the cost function of SDSENLMS algorithm.