Precoding optimization for joint relay and QoS-constrained data transmission in cooperative wireless network

We consider the problem of precoding optimization for wireless MIMO nodes that are assigned both to serve its own destination and to relay another data transmission with a different destination. Unlike the traditional MIMO relay broadcast model, our cooperative node not only amplifies and relays another source signal to its destination but also sends its own information to another destination. Our goal is to maximize the capacity of the relay link while achieving a desired level of its own channel capacity. Under full channel knowledge, we develop two suboptimal solutions: one based on a two-step decomposition; the other based on a reduced gradient approach. We demonstrate negligible performance difference between the two solutions. In particular, both algorithms can approach sum rate bound at high signal to noise ration (SNR) with the two-step decomposition method being less complex.