A Multipath AoA/AoD-Based Shared Dictionary Learning Framework for FDD Massive MIMO Channel Estimation

This paper addresses the compressive sensing (CS)-based frequency division duplexing (FDD) massive multiple-input multiple-output (MIMO) downlink channel estimation problem in dynamic scenarios. We propose a multipath angle of arrival (AoA) and angle of departure (AoD)-based shared dictionary learning (MASDL) algorithm, where the discriminative and shared features in the angular domain are exploited via supervised dictionary learning, enhancing the generalization ability of the model. Simulation results show that the proposed algorithm achieves better normalized mean square error (NMSE) performance and substantially reduces the pilot overhead compared with other channel estimation schemes.