A cross-layer view of wireless multicast optimization

Abstract

We consider a system of multiple multicast transmitters with possibly overlapping receivers. We introduce a stationary policy that allocates the transmission rates and powers of every transmitter with the objective to maximize the overall user utility, which is measured in terms of the total average rate of each receiver. We show optimality of this policy by employing the theory of stochastic approximation for any utility function that is strictly concave and increasing in the average rate. The introduced policy is opportunistic in nature as it capitalizes on the variations of the wireless channel that are induced by fading. Specifically, we show through a set of simulations that the average throughput under unicast traffic observes a benefit under fading. Our results further indicate that this benefit is mitigated when multicast traffic is considered.