UC Riverside

In this paper, while considering the impact of the antenna correlation and the interference from neighboring users, we analyze channel estimation and training sequence design for multi-input multi-output (MIMO) two-way relay systems. To this end, we propose to decompose the bidirectional transmission links into two phases, i.e., the multiple access (MAC) phase and the broadcast (BC) phase. By considering the Kronecker-structured channel model, we derive the optimal linear minimum mean-square-error (LMMSE) channel estimators. The corresponding training designs for the MAC phase and the BC phase are then formulated and solved to improve channel estimation accuracy. For the general scenario of the training sequence design for both phases, two iterative training design algorithms are proposed that are verified to produce training sequences achieving near optimal channel estimation performance. Furthermore, for specific practical scenarios, where the covariance matrices of the channel or disturbances are of particular structures, the optimal training sequence design guidelines are obtained. The minimum required training lengths for channel estimation in both the MAC phase and the BC phase are also analyzed. Comprehensive simulations are carried out to demonstrate the effectiveness of the proposed training designs.