Temporally and spatially correlated uplink channel estimation for massive MIMO systems

It is well known that most wireless channels are sparse. The channel impulse responses (CIRs) of certain size antenna array have a partially shared support due to the spatial correlation. Moreover, the CIRs during adjacent symbols have a common support due to the temporal correlation. In this paper, we propose a novel compressed channel estimation scheme for the uplink massive multi-input multi-out (MIMO) systems based on the assumption of time division duplexing (TDD) mode by exploiting the above mentioned temporal and spatial correlations. Then a joint subspace pursuit (JSP) algorithm is proposed. Simulation results demonstrate the proposed scheme using the JSP algorithm outperforms the classical channel estimation techniques.