Study On The Task Offloading Of Edge Computing System Based On Target Problem

In the era of the Internet of everything,massive data generated from devices need to be processed.Due to the insufficient performance of local devices and the high latency of the current cloud computing scene,more and more emerging businesses cannot be satisfied,so edge computing emerges at the right moment.Local devices can be unloaded to solve the problems encountered by these emerging businesses.In edge computing,computing offloading technology offloads the computing tasks of mobile terminals to the edge network,which solves the shortcomings of equipment in resource storage,computing performance,energy efficiency and other aspects.At the same time,compared with computing unloading in cloud computing,edge computing solves the problems of network resource occupation,high latency and extra network load.In recent years,with the continuous development of industrial Internet of Things,business needs are becoming more and more diversified.Some emerging technologies,such as edge computing,industrial Internet of Things,5G technology and software-defined network(SDN)technology,have reached the state of mutual fusion and mutual promotion.Therefore,it is very important to study the relationship between these business scenarios and technologies.1)This paper first introduces the edge computing reference architecture 3.0proposed by the Edge computing Industry Alliance,and summarizes the application scenarios of edge computing technology.At present,edge computing is mainly used in intelligent manufacturing,smart city,Internet of vehicles and industrial scenarios.Although the application scenarios are constantly expanding,edge computing technology still faces some challenges in standards,technology and safety.Therefore,how to determine unloading decision and rationally allocate computing resources has become a hot topic in current research.This paper mainly studies the task offloading technology for different targets in edge computing applications.2)Based on the existing research,this paper proposes a local + edge + edge cloud collaborative edge system model aiming at power stability and time delay,which can be applied to the deployment of edge system with energy collection function.After the system model is established,an optimized integer linear programming algorithm is proposed to solve task unloading and computational resource allocation problems.The algorithm consists of two parts: Task priority-based Lyapunov optimization algorithm and CPU utilization optimization integer programming algorithm.The experimental results show that the model and algorithm can improve the CPU utilization of edge server,further improve the system performance and reduce the delay while the local power remains stable.3)For the application of computing the lower edge of the industrial iot scenarios,this paper puts forward the target cost and delay of local + pretreatment edge +application edge + cloud collaboration model,the edge server is divided into edges pretreatment server and application server,used for to unload tasks,resources allocation strategy and edge server deployment issues.As this problem is NP-hard,it is finally solved by using intlin Prog function in MATLAB by transforming into integer linear programming.Through experimental analysis,the system model and solution method proposed in this paper can obtain a good calculation unloading and resource allocation strategy,and analyze the influence of preprocessing server capacity parameters and delay weight on the deployment of the edge system,which can provide an analysis method and deployment scheme for the application of the edge system in real scenes.