Compressive Sensing Sparse Channel Estimation in FDD Massive MIMO Systems

Precise channel estimation (CE) is essential to fully realize the potential performance benefits of massive MIMO technology. However, the pilot overhead required by the conventional CE schemes is unacceptable due to the massive number of antennas, especially for FDD massive MIMO. To address this issue, this chapter introduces a structured compressive sensing-based joint CE scheme exploiting the spatio-temporal common sparsity of delay-domain MIMO channels. It starts by developing non-orthogonal pilots at the base station, guided by compressive sensing theory, to reduce pilot overhead. Subsequently, an Adaptive Structured Subspace Pursuit (ASSP) algorithm is considered at the user end to jointly estimate channels of multiple OFDM symbols, which leverages the spatio-temporal common sparsity of MIMO channels to enhance CE accuracy. By exploiting temporal channel correlation, a space-time adaptive pilot scheme is introduced to further reduce pilot overhead. Additionally, the discussion on the CE scheme is extended from the single-cell scenario to the multi-cell scenario for wider applications. Numerical results confirm that the considered scheme achieves accurate channel estimation with limited pilot overhead, approaching the performance of the optimal oracle least-square estimator.