QoE-Aware Mobile VR HAS Cache Management With Coding Helper

Abstract

Virtual reality (VR) video has attracted substantial research interests with the explosion of mobile VR devices in recent years. In this paper, we propose a novel cache management architecture with coding helper to meet the VR’s high bit rate requirement and the multi-rate feature in the HTTP adaptive streaming (HAS) transmission. The coding helper has not only storage capacity just as traditional caching helper, but also video coding function to encode the VR video into different rates. In this new architecture of VR HAS cache management, we consider the problem that how to distribute the bit rate of different VR videos in a certain cache size by taking into account the quality of experience (QoE) in different situations. In order to decrease the computing time of VR video service and guarantee a high level QoE as well, we propose a new maximum QoE increase (MQI) cache management method to obtain the solution of the optimal problem in multi-video and multi-cache (MVMC) situation. Furthermore, we propose a scheduling algorithm of coding helper to decrease the backhaul traffic and develop four different request scenarios in the experiments. Simulation results are finally provided to show that the coding helper scheme has about 16%–100% gain of usefulness than the traditional caching helper scheme. The proposed MQI cache allocation algorithm has more than 50% decreasing of computing time and almost has the same QoE compared with the exhaustive search. The framework can provide useful guidelines to distribute rate of VR video in cache over wireless networks and get better QoE.