Adaptive filtering based 3D massive MIMO sparse channel estimation

Abstract

In the 3D massive MIMO system, both the transmitter encoding and the receiver signal detection require channel state information. The accuracy of channel state information will directly affect the overall performance of the system. Therefore, accurate channel estimation is the key to reliable communication in the 3D massive MIMO system. Due to the channel sparsity of 3D massive MIMO, existing works use compressive sensing to propose channel estimation schemes which exploit the sparse structure of the channel to improve the capability of channel estimation. This paper further excavates the sparse structure information of the 3D massive MIMO system and suppresses noise from the perspective of adaptive filter to further improve the performance of channel estimation. The basic idea of this approach is to transform 3D massive MIMO sparse channel estimation problem into CS sparse signal reconstruction problem and then employ adaptive filtering algorithm in CS reconstruction problem. The scheme of adaptive filtering based 3D massive MIMO sparse channel estimation (AFSCE) has many advantages over other typical CS algorithms, such as its probability of exact reconstruction against measurement numbers, etc., which are verified by simulation results.