Research On Joint Optimization Strategy For Cross Domain Streaming Media Edge Cloud

With the maturity of cloud computing technology,cross-domain edge cloud mode of streaming media services are gradually rising.In this mode,the storage and computing of streaming media services will be delivered to the sub-cloud in the edge network.Compared with the central streaming media service,the transmission delay can be reduced and the quality of service can be guaranteed.The edge cloud of the streaming media is required to schedule limited cloud resources to ensure the quality of the user's service under circumstance of user requests fluctuation.At the same time,due to differences in the requirements of the regional users and the edge of the cloud service,it is necessary to utilize the cross-domain forwarding requested by the user to avoid local congestion and frequent video redeployment to ensure the overall quality of service.Finally,the overall number of user requests fluctuates over time,in order to avoid cloud resource idle and guarantee user service quality at requests peaks,it is needed to dynamically adjust the cloud resources to handle the user request fluctuations.Therefore,this dissertation studies the internal resource scheduling of the edge cloud,the cross-domain user request allocation and the dynamic adjustment of the overall cloud resource in the cross-domain streaming media edge cloud co-optimization,and completes the following work:1)In this dissertation,a joint optimization strategy is proposed,which can effectively utilize the cloud resources to guarantee the quality of user service and dynamically adjust the leasing cost of cloud resource through cooperation of the sub-cloud internal resource scheduling,user request cross-domain forwarding and cloud resource dynamic adjustment.And the implementation cycle and cooperation mode of each strategy are in accordance with the rules of arrival of the actual user request,the charging scheme of the cloud service provider and the cloud service process of the cross-domain streaming media.In addition,this dissertation proposes a sub-cloud internal resource scheduling strategy for cross-domain joint optimization.The simulation results show that the strategy can make full use of the cloud resource to respond to user requests and minimize the response delay of the sub cloud.2)For the cross-domain request allocation problem,this dissertation takes account of the influence of the sub-cloud service model,the impact of the fluctuation on the copy of the single sub-cloud,the response of the sub-cloud to the local request and the influence of the service delay model on the allocation of the cross-domain user requests allocation.The simulation results show that the policy can make full use of the global cloud resources and ensure the quality of the users' services.3)In this paper,the cloud resource adjustment strategy quantifies the resource consumption and related QoE indexes of each service stage,and can decide the cloud resources needed by the request allocation strategy and the sub-cloud resource scheduling strategy according to the given relevant QoE indicators.The simulation results of the joint optimization strategy show that this strategy can dynamically adjust the leasing scheme of cloud resources according to the fluctuation of the user request,and ensure the global user service quality.Therefore,the joint optimization strategy proposed in this dissertation can make full use of cloud resources to ensure the quality of user service,and can dynamically adjust the cloud resources according to the user request fluctuation,and can be applied to the actual cross-domain streaming media edge cloud service system.