边缘环境下交互式视频应用的资源分配算法研究

Research on resource allocation algorithm of interactive video application in edge environment

边缘计算在靠近用户的网络边缘提供计算服务,促进了众多交互式视频应用的发展,如增强现实/虚拟现实、视频直播、云电脑等。为了更有效地利用边缘节点的资源,尽可能地减少网络时延和视频帧率对用户体验的不良影响,文章提出了基于李雅普诺夫优化的在线内外两层算法。该算法能够实时监测当前的系统状态,包括用户位置、设备资源、网络状态等,并据此动态调整服务放置和带宽分配决策,高效利用资源,保证用户体验。文章进行了模拟实验,实验结果验证了该算法的可行性和性能。

Edge computing provides computing services at the edge of the network close to users,which promotes the development of many interactive video applications,such as augmented reality/virtual reality,live video,cloud computing,etc.In order to make more effective use of the resources of edge nodes and minimize the adverse effects of network delay and video frame rate on user experience,this paper proposes an online inner and outer two-layer algorithm based on Lyapunov optimization.The algorithm can monitor the current system status in real time,including user location,device resources,network status,etc.,and dynamically adjust the service placement and bandwidth allocation decisions to make efficient use of resources and ensure user experience.In this paper,simulation experiments are carried out,and the experimental results verify the feasibility and performance of the algorithm.