Research On Resource Allocation Strategy For Video Service Over Cloud Radio Access Network

As one of the key technologies of 5G systems,Cloud Radio Access Network(C-RAN)has the advantages of energy-saving,high spectrum utilization,good extensibility,resource pool sharing and so on.Meanwile,the use of Passive Optical Network(PON)as the fronthaul link of C-RAN can provide users with large-capacity,high-reliability,and lowlatency access services.One hand,with the wide application of multimedia terminals,video services have become the main business in the network.On the other hand,due to the increasing user demands,wireless network resources are in short supply.In order to improve the Quality of Experience(QoE)for users when watching videos,it is very important to design an efficient video distribution strategy under network conditions with limited resources.Therefore,this thesis focuses on resource allocation strategy for video service over cloud radio access network.The main research contents are as follows:1.To solve the problems of the uneven distribution of bandwidth and poor quality of the user experience in optical and wireless network,a virtualized cloud radio access network video transmission resource allocation strategy with QoE-aware is proposed.Establish the virtualized cloud radio access network utility function with the fronthaul bandwidth efficiency,tuning overhead,and load imbalance as parameters,and construct the virtualized cloud radio access network by using the coalition formation algorithm based on the merge and split rules.Furthermore,analyze the user's buffer size and requested video quality to construct the video interruption risk parameter,and the user and the virtualized cloud radio access network are modeled as a two-stage Stackelberg game.The allocation of wireless resources is achieved by solving the Nash equilibrium.The results show that the proposed strategy can not only reduce video interruption,but also improve the resource utilization in the optical and wireless domains.2.To solve the problems of the contradictory relationship between the strict delay requirements and limited network resources in the transmission of video in the fog radio access network,a delay-aware cloud and fog cooperative video caching,transcoding and radio resource allocation strategy are proposed.Build a video request scheduling model through the different locations of the user request video segments.Build a delay optimization model by analyzing the length of the user's playback buffer.To achieve the purpose of minimizing the interruption time,use the Actor-Critic Deep Reinforcement Learning algorithm to allocate wireless and computing resources to the user,and determine the location of the video cache.The results show that the proposed strategy can not only improve the cache hit rate,but also reduce the interruption time.