Research On Cooperative Scheduling Technology Based On Edge Computing

The proliferation of mobile devices and universal access to wireless networks enable many new mobile applications such as augmented reality,mobile gaming and so on.Because these applications have very high latency requirements,the traditional cloud computing approach always offload all complex computations of the application to the cloud server.Since the location of traditional cloud computing server or cloud computing center is often very remote from the user,long-distance data transmission will cause high latency.With the rise of edge computing,researchers have considered the complex computation of these applications to be partially offloaded to a nearby edge server through a certain computational offloading scheme to reduce latency.The existing works mainly consider the offloading of complex computations between mobile terminal devices and traditional cloud computing with abundant resources.However,under edge computing,there are limitations in computing resources,storage resources,and network resources.The method proposed in the past works is not fully applicable at this time.Under resource-constrained conditions,it becomes important to perform computational offloading,task scheduling,and resource allocation to achieve lower latency.In view of the above aspects,the main work of this paper is as follows:1)Different from previous works,given a computational partitioning scheme of an application,this paper uses cooperative scheduling technology to reduce the delay,which is the average completion time of sensitive applications,in edge computing by using task scheduling,computational offloading and resource allocation techniques in the edge computing environment.2)A novel heuristic algorithm,namely Multi-dimension Search Adjust(MDSA)algorithm,is proposed.By comparing MDSA with the benchmark algorithm,the experimental results show that MDSA has superior performance than the benchmark algorithm in overall application delay.3)In addition,we have designed a Cooperative Online Scheduling(COS)algorithm and proposed a corresponding system architecture diagram,which can be easily deployed in the practical system.Through the evaluation of the experiment,we found that the performance of COS is better than the benchmark method.