A Whittle's Index Based Approach for QoE Optimization in Wireless Networks

Abstract

The design of schedulers to optimize heterogeneous users' Quality of Experience (QoE) remains a challenging and important problem for wireless systems. Our paper ([1]) explores three inter-related aspects of this problem: 1) non-linear relationships between a user's QoE and flow delays; 2) managing load dependent QoE trade-offs among heterogeneous application classes; and 3), striking a good balance between opportunistic scheduling and greedy QoE optimization. To that end we study downlink schedulers which minimize the expected cost modeled by convex functions of flow delays for users with heterogeneous channel rate variations. The essential features of this challenging problem are modeled as a Markov Decision Process to which we apply Whittle's relaxation, which in turn is shown to be indexable. Based on the Whittle's relaxation we develop a new scheduling policy, Opportunistic Delay Based Index Policy (ODIP). We then prove various structural properties for ODIP which result in closed form expressions for Whittle's indices under different scheduler scenarios. Using extensive simulations we show that ODIP scheduler provides a robust means to realize complex QoE trade-offs for a range of system loads.