Channel Training Algorithms for Two-Way MIMO Relay Systems

Two-way relay systems are known to be capable of providing higher spectral efficiency compared with conventional one-way relay systems. However, the channel estimation problem for two-way relay systems is more complicated than that of one-way relay systems. In this paper, we propose and compare two channel estimation algorithms, namely the superimposed channel training scheme and the two-stage channel estimation algorithm, for two-way multiple-input multiple-output (MIMO) relay communication systems, where the individual channel state information (CSI) for the first-hop and second-hop links is estimated. For both algorithms, we derive the optimal structure of the source and relay training sequences which minimize the mean-squared error (MSE) of channel estimation. In the superimposed channel training scheme, the power allocation between the source and relay training sequences is optimized. For the two-stage channel estimation algorithm, we optimize the power allocation at the relay node between two stages to improve the performance of the algorithm. Numerical examples are shown to demonstrate and compare the performance of the proposed channel training algorithms.