Heavy Traffic Asymptotic Approach for Video Streaming over Small Cell Networks with Imperfect State Information

International audience; In this work, we address the problem of decentralized power allocation for satisfying individual QoS constraints in video streaming in a Small Cells Network.They QoS metric we use here is the probability that the queue length at each transmitter exceeds some threshold: we want this probability to be fixed at a desired value. We focus on a model with interfering transmitter-receiver pairs. Using heavy traffic asymptotic modelling we propose a power control algorithm for the case where each base station has access to local SINR feedback, information about the queue length of its user and delayed information of the queues in the other base stations. Simulation results suggest that the proposed algorithm is quite robust in the case of delayed information sharing.