| 摘 要: 文章研究分析了直播中多用户、多服务器场景下存在的直播用户的体验质量(Quality of Experience,QoE)不高的问题,为提升用户的QoE,将能耗和时延作为决策目标,设计一种经改进的NSGA-Ⅱ(非支配排序遗传算法),即L-NSGA-Ⅱ,用线性排名的方式进行父代的选择加速算法收敛。实验表明,与LUA、NSGA-Ⅱ和Random算法策略相比,所提方案的平均延迟降低约9.1%,用户QoE提升约4.39%。该方案应用于直播场景中,在减少延迟、提升吞吐量和降低能源开销方面表现出较好的效果。 |
| 关键词: 直播;边缘计算;卸载;用户体验;用户分配;L-NSGA-Ⅱ |
|
中图分类号: TP393
文献标识码: A
|
| 基金项目: 2024年度吉林省教育厅科学研究重点课题项目(JJKH20240749KJ);中国职业技术教育学会2023年度重点课题立项(ZJ2023A022) |
|
| Research on Edge Computing Task Offloading Scheme for Live Broadcasting |
|
XIE Song, WANG Wei
|
(College of Network Security, Changchun University, Changchun 130022, China)
1248904025@qq.com; wangwei@ccu.edu.cn
|
| Abstract: This paper studies and analyzes the problem of low Quality of Experience (QoE) for live streaming users in multi-user and multi-server scenarios. To improve the QoE of users, this paper proposes to design an improved NSGA-Ⅱ (Non-Dominant Sorting Genetic Algorithm), namely L-NSGA-Ⅱ, taking energy consumption and delay as decision-making objectives. The algorithm uses linear ranking to accelerate the convergence of the parent selection algorithm. The experiment shows that compared with the LUA, NSGA-Ⅱ, and Random algorithm strategies, the proposed scheme reduces the average latency by about 9.1% and improves user QoE by about 4.39% . This scheme has shown good performance in reducing latency, improving throughput, and reducing energy consumption when applied to live streaming scenarios. |
| Keywords: live streaming; edge computing; offloading; user experience; user assignment; L-NSGA-Ⅱ |
