Low complexity SDNLMS Adaptive Channel Estimation for MIMO-OFDM systems

Abstract

Channel Estimation process is an important technical issue for Multiple Input Multiple Output (MIMO)-Orthogonal Frequency Domain Multiplexing (OFDM) wireless communication systems. Adaptive Channel Estimation (ACE) is a widely used estimation technique for MIMO-OFDM systems. Different types of channel estimator like least mean square (LMS), normalized least square (NLMS) are applied to ACE. LMS algorithm has low complexity but it suffers from high minimum mean square error (MMSE) performance. Normalized least mean square algorithm (NLMS) provides low MMSE performance with poor convergence rate. In this paper, an improved sign data normalized least mean square (SDNLMS) channel estimation technique is proposed which improves the convergence rate and computational complexity of NLMS algorithm, while maintaining low MMSE performance for channel estimation (CE). Simulation results show that SDNLMS CE method provides faster convergence rate with low MMSE and computational complexity, which is clearly better than NLMS and LMS method in a MIMO-OFDM system. Therefore SDNLMS CE method can be highly compatible for ACE process.