LAUSR

Channel estimation is of great challenge for millimeter-wave (mmWave) multiple-input multiple-output (MIMO) systems. The underlying reason can be attributed to two facts. One is the huge dimension of MIMO channel matrix, since mmWave systems typically employ large antenna arrays at both the transmitter and receiver to realize sufficient link margin. The other fact is the limited number of radio-frequency chains available for transmitter/receiver array, due to the power and cost limitations. This paper proposes a two-stage accurate channel estimation scheme for mmWave MIMO. Initially, we utilize sparse Bayesian learning to obtain a coarse channel estimate. Then we refine the channel estimate by iteratively maximizing the likelihood function of channel parameters. It is shown through simulations that the proposed scheme significantly outperforms the counterparts, and can achieve accurate channel estimation performance.