Deep Learning-Based Channel Estimation for Wideband Hybrid MmWave Massive MIMO

Hybrid analog-digital (HAD) architecture is widely adopted in practical millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) systems to reduce hardware cost and energy consumption. However, channel estimation the context of HAD challenging due only limited radio frequency (RF) chains at transceivers. Although various compressive sensing (CS) algorithms have been developed solve this problem by exploiting inherent sparsity structures, effects, such as power leakage beam squint, can still make real features deviate from assumed models result performance degradation. Besides, high complexity CS caused a large number iterations hinders their applications practice. To tackle these issues, we develop deep learning (DL)-based approach where sparse Bayesian (SBL) algorithm unfolded into neural network (DNN). In each SBL layer, Gaussian variance parameters angular domain are updated tailored DNN, which able capture complicated structures domains effectively efficiently. The measurement matrix jointly optimized for improvement. Then, proposed extended multi-block case correlation time further exploited adaptively predict facilitate update parameters. Simulation results show that approaches outperform existing terms both complexity.