Convex optimization strategies for precoding of broadcast channels

This work tackles the non-convex maximization of throughput for a multiuser MISO wireless system where the transmitter is provided with long-term channel state information and leverages joint multiuser linear precoding and orthogonal space-time block coding (OSTBC). The original problem is reduced here to convex optimization by means of two different strategies. The first solution is based on constrained optimization so that the global problem is split into multiple simple and convex optimizations. Convexity is achieved by setting additional constraints on the mutual interference between the users. The second solution is based on game theory and reaches the Nash equilibrium (i.e., the equilibrium point for non-cooperative games). Numerical results show that the optimized system based on OSTBC and multiuser linear precoding outperforms the conventional TDMA system when the users to be simultaneously served are carefully chosen according to their channels properties.