Joint Optimization for Quality Selection and Resource Allocation of Live Video Streaming in Internet of Vehicles
Penglin Dai,Meiting Wu,Ke Li,Xiao Wu,Yan Ding
DOI: https://doi.org/10.1109/tsc.2023.3349051
IF: 11.019
2024-01-01
IEEE Transactions on Services Computing
Abstract:Live Video Streaming (LVS) services are critical in supporting real-time applications in Internet of Vehicles (IoV) by transmitting real-time generated video content from streaming server to vehicles. Due to restricted spectrum resources and high vehicle mobility, LVS suffers from notable performance degradation. Moreover, existing strategies such as buffer size control and edge caching, are designed for video-on-demand service, which is ineffective for LVS in IoV. Accordingly, we investigate the problem of LVS-IoV by synthesizing multicasting and Scalable Video Coding-based encoding with the goal of maximizing Quality of Experience (QoE), which is defined as the weighted sum of video quality, rebuffering time, and quality variation. The LVS-IoV is decoupled into three sub-problems: vehicle grouping, quality selection, and resource allocation. Firstly, we propose a K-means-based vehicle grouping method that considers geographical distribution, velocity, and dynamic channels. Secondly, we determine the quality selection of each group based on the Value Decomposition Network for maximizing overall video quality. This network utilizes global value function decomposition and centralized training to achieve fast convergence, followed by distributed execution. Lastly, we propose a sub-gradient algorithm to achieve optimal resource allocation. We build simulation model and perform extensive evaluation, which demonstrates its superiority compared to other competitive methods.
computer science, information systems, software engineering