Design of downlink training sequences for FDD massive MIMO systems

In the massive Multiple-Input Multiple-Output (MIMO) system, a large number of transmit antennas are deployed at base stations. The massive MIMO has been proposed as one of the key technologies for the next generation of wireless communication systems. A major problem in practice to realize the capacity improvement promised by the massive MIMO technology is the availability of the channel state information in downlink at the base station. In frequency division duplexing (FDD) massive MIMO systems, the overhead of the training sequences could overwhelm the precious downlink resources due to the large number of transmit antennas. Therefore, a training sequences design approach that exploits the channel spatial correlation is proposed to reduce the overhead of training sequences. The optimal training sequence structure is analyzed and the length of optimal training sequence which could maximize the achievable data rate is evaluated by the Monte Carlo simulations in this paper. It is shown that the overhead is significantly reduced and a trade-off between accuracy of channel estimation and the system achievable data rate is achieved simultaneously by the proposed training sequences design method.