Capturing the Sparsity for Massive MIMO Channel with Approximate Message Passing

In this work, we propose a low-overhead characteristic learning mechanism for the time-varying massive MIMO channels. Specially, we exploit the common sparsity and temporal correlation of the channel. Firstly, using VCR and modeling the temporal correlation as an autoregressive process, we formulate the dynamic massive MIMO channel as a sparse signal model. Then, an expectation maximization (EM) based sparse Bayesian learning (SBL) framework is developed to learn model parameters. To achieve the posteriors of model parameters, approximate message passing (AMP) is utilized in the expectation step. Finally, we demonstrate the performance through numerical simulations.