Channel Estimation Based on Sparsity-Enhanced Penalty for MmWave MIMO Communications

In this paper, we present a novel sparsity-enhanced penalty-based fast block complex-valued least mean square (SEP-FBCLMS) algorithm for channel estimation within a single-carrier millimeter-wave hybrid Multiple-Input Multiple-Output (MIMO) system. Specifically, we propose a sparsity-enhanced penalty (SEP) scheme tailored for channel estimation which is integrated into the FBCLMS algorithm to enhance sparse representation capabilities. The SEP-FBCLMS algorithm then incorporates an equilibrium factor that intelligently balances the weights assigned to both the data fitting error and the channel sparsity. Simulation results validate the performance of the SEP-FBCLMS algorithm, demonstrating its capacity to achieve higher estimation accuracy compared to other existing LMS-based algorithms.