Precoding methods for multi-input multi-ouput interference channels with channel covariance Feedback

Usually, instantaneous channel information is required by the transmitter for the precoding in multi-input multi-output interference channels. In the case of fast fading channels, the resultant additional overhead of the system will be high. Against this, this study considers the scenario where the channel is spatially correlated and the transmitter only has the covariance information of the channel so that the system overhead can be reduced. On the basis of maximum signal-to-interference-plus-noise ratio (SINR) criterion, the author proposes two precoding methods, including the iterative precoding and bisection search precoding methods. The first method solves the optimal precoder of a given user provided that others´ precoders are known, so that the SINR can be increased gradually. Whereas the second method introduces an auxiliary variable, transforms the original non-convex problem into a convex one, and solves it by convex optimisation softwares. Both the proposed methods are proved to be convergent and their computation complexity and overheads are also analysed. Under various system settings, the proposed two methods are compared with a few existing precoding or joint transceiver methods by numerical method and computer simulation, and the superiority of the proposed methods is also shown in most cases.