Sparse Channel Estimation for MIMO Systems based on Time-Domain Training Sequence Optimization

摘要

Recently, the technique of multiple input multiple output (MIMO) is becoming more and more widespread in the communication systems in order to significantly improve the system capacity. The channel estimation is still one of the most tricky problems because of the large amount of coefficients to be estimated for one receive antenna. This paper focuses on designing and optimizing the time-domain training sequences for MIMO systems and proposes a novel sparse channel estimation scheme taking advantage of the spatial correlation among both the transmit antennas and the receive antennas. The structured compressive sensing (SCS) framework is set up, which is the main approach to recover the sparse channel for different antennas. It is illustrated in the simulation results that the proposed scheme has a higher probability to recovery the nonzero support of the channel than that of the traditional schemes, while the mean square error is also smaller, which means our scheme is expected to have a superior performance.