Multiuser Massive MIMO Channel Estimation and BER Analysis Using Convex Combination Based Algorithms

摘要

Massive multiple-input multiple-output (MIMO) employs a large antenna array and is used as a key technology in 5G communication standards. In this work, time-division duplex (TDD) is used for uplink (UL) and downlink (DL) channels as the same bandwidth can be utilized for bidirectional data transfer. This paper proposes a computationally efficient adaptive algorithm based channel estimation model. These models are applicable to multiuser MIMO (MU-MIMO) systems, which are comprised of a large number of channel coefficients. Channel state information (CSI) is acquired from estimated channel coefficients of the UL channel, and CSI of DL is obtained through channel reciprocity. The dynamic characteristics of MIMO channels are generally sparse in nature, and the degree of sparseness varies over time. Convex combination-based modeling with the introduction of l0 norm penalty to Least mean squares (LMS) algorithm develops an efficient and stable sparse channel estimation model, which is the most significant contribution of the paper. The research work presented in the paper has simultaneously focused on channel estimation using the convex combination of momentum-fractional LMS approach and equalization for massive MIMO channels. The simulation result exhibits that the proposed method outperforms the existing conventional compressed sensing (CS) based channel estimation. The UL and DL performance of the system was investigated in terms of bit error rate (BER), mean square error (MSE), and channel capacity calculated from estimated CSI.