Research On Video-based Resource Allocation In Wireless Networks

With the rapid development of wireless networks,people's various demands for video services have also increased rapidly.Due to the differences in the types of user-side equipment,the video needs to be transcoded to meet the different needs of users.On the other hand,because wireless network resources have dynamic characteristics and limitations,it is a great challenge to provide users with high-quality and strict latency requirements for video services.This paper studies the resource allocation problem based on live video service and on-demand video service in fog network and multi-access point edge network respectively.Main tasks as follows:(1)For live video services,this paper designs a resource allocation scheme based on Actor-Critic(AC)deep reinforcement learning algorithm in fog network.This solution jointly optimizes user scheduling,video quality selection,and resource allocation.By making a compromise between high video quality and low playback latency,it maximizes video quality and minimizes playback delay in the live video service.Therefore,the quality of experience(Qo E)of the user is optimized.In this paper,we use the AC algorithm to solve the above joint optimization problems.Simulation results show that the proposed scheme can significantly improve the user experience,and by comparing the simulation results,we can see that the performance of the AC algorithm is better than the Policy Gradient(PG)algorithm.(2)For on-demand video services,this paper designs a resource allocation scheme based on Dynamic Deterministic Policy Gradient(DDPG)algorithm for dynamic adaptive streaming over HTTP(DASH)in a multi-access point edge computing network.The most common problem for DASH systems is that multiple users are competing for limited server bandwidth and storage resources.This paper will jointly optimize the cache of edge nodes,the quality of video received by users,latency,and operating costs to ensure user fairness and improve the quality of experience.This paper uses the DDPG algorithm to solve the above joint optimization problems.Simulation results show that the proposed solution can significantly improve the user's video quality experience while effectively controlling the operating cost.