Dictionary-Based Tensor Decomposition-Aided Time-Varying Channel Estimation for Millimeter Wave MU-MIMO Systems

This research focuses on the multiple-input multiple-output (MIMO) mmWave systems' uplink time-varying channel estimation challenge. We convert the time-varying channel es-timation in the hybrid beamforming structure into a param-eter estimation problem that takes angles of arrival/departure (AoA/AoDs), Doppler shift, and path gains into account. We also suggest a two-stage channel estimation method. We describe the receiving signals as a third-order tensor in the initial stage and provide an algorithm that combines Matching Pursuit (MP) and Alternating Least Squares (ALS) called Dictionary-based Tensor Decomposition to estimate AoA/AoDs. In the second stage, we recover the Doppler shifts and paath gains using the estimated AoA/AoDs. The entire time-varying channel is finally restored. The simulation results demonstrate that the suggested method performs better than the traditional approach in terms of complexity and estimation accuracy.