Joint Coding and Stochastic Data Transmission for Uplink Cloud Radio Access Networks

We study the uplink cloud radio access networks using quantize and forward as an underlying relaying scheme. In all previous works, each radio unit (RU) is assumed to choose a compression rate equal to an instantaneous backhaul capacity. In this letter, we consider a more general network model for which each RU is equipped with a queue whose arrival rate is equal to a compression rate and departure rate is equal to a current backhaul capacity. For each time slot, it can choose a higher compression rate than the given backhaul capacity as long as the queue is stable. Using the principle of stochastic network optimization, we present a distributed algorithm to find the compression rate of each RU such that a time-averaged sum-rate is maximized subject to the queue stability. Under the optimal compression policy, some of RUs are active with a higher compression rate and the others are silent, for a given time slot, which is completely different from the conventional approach that all RUs are always active.