Joint Optimization Of Service Chain Caching And Task Routing In Mobile Edge Computing

Mobile edge computing is an emerging computing paradigm with great potential.By deploying computing and storage resources in the base station,tasks can be processed quickly at the edge of the network,thus meeting the needs of many emerging applications for low response latency.Current research on mobile edge computing mainly focuses on task offloading,routing and scheduling strategies.Although the relevant research has achieved certain results,most of the work assumes that the base station can handle any type of task.However,in an actual scenario,the base station needs to cache the application before performing the corresponding task.Moreover,most of these applications consist of multiple services or modules.Therefore,instead of the traditional modeling method that is not suitable for the actual scenario,we model the application as a service chain,and based on this model,dynamic service chain caching and efficient task routing planning for resourcelimited base stations.The main work of this paper is as follows:From the perspective of application service providers,we study the joint optimization of service chain caching and task routing under the constraints of long-term average cost.To make a more accurate decision of service chain caching and task routing,we use slotted mode to operate.Firstly,according to the knowledge of Open Jackson Queuing Network,we analyze the average sojourn time of service chain processing tasks and then quantify the service chain caching and task routing decisions as a long-term system average response delay minimization problem.Then,based on the long-term average cost set by the application service provider,we design an online cost adaptive algorithm based on Lyapunov optimization technology,which converts the long-term minimization problem of future global task information into only a single-slot problem that requires current slot task information.Finally,for service chain caching and task routing problem in a single slot,we design a single-slot service chain caching and task routing algorithm based on the construction method and an iterative search algorithm based on joint aware of cost and delay.The simulation results show that the proposed algorithm can effectively lower the average response delay of system processing tasks compared with other algorithms,and ensure that the cost of application service providers renting base station resources remains at a relatively low level.