A cross-layer view of wireless multicasting under uncertainty

In this paper we study the problem of scheduling through rate and power control for multicast traffic. Our objective is to maximize the total user utility where user utilities are measured in terms of the long term average received rates. Considering the fact that accurate knowledge of the wireless channel conditions may not be always available, we focus on policies that take decisions based only on inaccurate estimates of the wireless channel state. We introduce an on-line, stationary scheduling and rate control policy. We establish its optimality among all policies with the same knowledge regarding the current channel state through stochastic approximation arguments.