Wideband Channel Estimation for mmWave Massive MIMO System with Off-Grid Sparse Bayesian Learning.

In this paper, we design a compressed sensing (CS) based channel estimation method for millimeter wave (mmWave) massive MIMO systems and investigate the impact of dual-wideband effect (frequency-wideband and spatial-wideband) that appears in large array communications. Specifically, we adopt the off-grid sparse Bayesian learning (SBL) that directly works on the continuous angle-delay parameter domain and avoids the basis mismatch problem. Hence, the proposed method achieves better channel estimation accuracy compared to most state-of-the-art algorithms that rely on on-grid CS approach. Moreover, the proposed method could successfully handle the spatial-wideband effect (sometimes known as beam squint effect) for wideband massive MIMO communications that was previously ignored by many existing literatures. Simulation results are provided to demonstrate the superior performance of the proposed method.