Cloud VR Video Streaming Processing Algorithm Based on Edge Cloud Collaboration
Wenhao Zou, Zongshuai Zhang, Lin Tian, Jiaying Huang · 2023
This paper concerns a Virtual Reality(VR) video stream processing based on edge-cloud collaboration. Since VR video streaming demands high resolution and frame rates, current VR terminals often struggle to achieve optimal decoding speeds. This paper proposes a video stream processing algorithm that leverages edge cloud collaboration, combining the strong computing power and large delay of cloud servers with the weak computing power and fast response of edge servers. According to the current communication situation and the computing power distribution situation of the server, the video decoding task is split and allocated to the cloud server and the edge server respectively. The algorithm utilizes the different features of both servers to minimize latency. In the process of task splitting, it is necessary to allocate computing resources to different tasks of edge servers with different computing power. In this paper, the problem of computing resource allocation is modeled as a constrained problem and transformed into an unconstrained problem, which is then optimized by applying the Sequential Least SQuares Programming optimization algorithm(SLSQP). The simulation results demonstrate that the proposed algorithm outperforms existing methods in reducing video transmission time.