QoE-Oriented Rate Allocation for Multipath High-Definition Video Streaming Over Heterogeneous Wireless Access Networks

Abstract

Global mobile video traffic is predicted to grow explosively in the near future and high-definition (HD) video streaming applications are increasingly prevailing over the mobile Internet. To optimize the transmission performance of HD video streaming applications over the heterogeneous wireless access networks, multipath video streaming that exploits the multihoming features of mobile devices is proposed as a potential solution. However, the delivery of HD video has much higher bitrate and more stringent delay constraint than other services. It is thus challenging to transfer HD video through wireless networks. Conventionally, multipath video streaming is optimized by the video distortions or other network-level quality-of-service parameters, which do not correlate well with the human visual perceptions. To address this problem, we first propose a novel quality-of-experience (QoE) prediction model to prefer the QoE values of different rate allocation scenarios over heterogeneous wireless access networks. Then, the rate allocation problem over heterogeneous wireless networks is formulated as a nonlinear optimization problem of maximizing the perceived video quality at the receiver. A relaxing function called penalty function is exploited to transform the rate allocation problem into an unconstrained optimizing problem and the pattern search method is adopted to obtain the close-to-optimal solution in each rate allocation session. The extensive simulation results show that our proposal significantly improves the QoE in terms of mean opinion score compared with the reference schemes.