QoE-Orientated Resource Allocation for Wireless VR over Small Cell Networks

Abstract

For wireless virtual reality (VR) over small cell networks (SCN), the latency between user’s dynamic head rotation and the synchronous change in head-mounted display (HMD) influences quality of experience (QoE). In this paper, to assess VR user’s QoE, we specify mean opinion score (MOS) as a metric of latency. With the goal of maximizing system-wide MOS, the stochastic game approach is leveraged for investigating resource allocation problem. For problem solution, a distributed multi-agent learning algorithm is proposed, which can converge to a pure-strategy Nash equilibrium (NE). Numerical results demonstrate the excellent performance of our proposed algorithm.