Precoding design for cognitive two-way relay networks

We study precoding design for cognitive two-way relay networks (C-TWRNs). In C-TWRN, the multi-antenna secondary transmitter not only transmits its own signal to the secondary receiver, it also acts as relay to help forwarding signals of two primary users via two-way relaying in the licensed frequency band. Our objective is to design linear relay transceiver or precoder such that the achievable rate of the secondary user is maximized while maintaining rate requirements of the primary users. To achieve this goal, different relay strategies, i.e., amplify-and-forward (AF), bit level XOR based decode-and-forward (DF-XOR) and symbol level superposition coding based decode-and-forward (DF-SUP), are considered. By transforming these non-convex design problems into suitable forms, efficient optimization tools are used to find the optimal solutions of all the schemes. Closed-form solutions are also obtained under certain conditions.