Research On Computing Offloading Strategy In IoT Edge Computing

With the exponential growth of smart mobile devices and the continuous development of Fifth-Generation Mobile Communication technology,Internet of Things(IoT)have significantly increased the demand for connection,computing,and storage for tasks generated by large-scale devices.At the same time,it put forward higher requirements for time delay,computing power,energy efficiency and other indicators.Internet of Things Mobile Devices(IMDs)is limited by mobility,size and other factors,and has great constraints on battery capacity,computing capacity and storage capacity.Cloud sever processing tasks will cause large latency,so it is difficult to meet the processing requirements of increasingly complex tasks.The emergence of MEC provides a solution to the above problems.Tasks can be offloaded to edge nodes to solve the problems of insufficient local processing capacity and large cloud delay.At the same time,after the combination of edge computing and cloud computing,more computing and storage resources can be provided,and tasks can be offloaded to the optimal computing device for processing according to task types.In this situation,it is very necessary to design a reasonable scheduling and computing offloading strategy for tasks that are time-sensitive and computationally intensive according to the actual scene requirements.To solve the above problems,the main work of this paper includes the following two parts:(1)Proposed a Maximum Average Energy Efficiency(MAEE)algorithm based on MEC.For the high-speed Internet of Things business,IMDs are equipped with energy collection devices to randomly walk in fixed areas and capture renewable energy and convert it into direct current.This algorithm is used to solve the problems of mobility management,energy harvesting,offloading decision and resource allocation among multiple IMDs.We coordinate the allocation of wireless resources and computing resources to adaptively make dynamic offloading decisions.It reduces the execution delay while maximizing the average energy efficiency of the system,realizing the trade-off between system latency and energy efficiency.Experimental results show that the MAEE algorithm has higher energy efficiency than the traditional greedy-based offloading algorithms while ensuring latency.(2)Proposed a Delayed Feedback Computing Offloading(DFCO)algorithm based on Edge computing collaborative cloud computing.Aiming at the joint optimization problems of multiple types of delay-sensitive tasks,task scheduling strategies and task offloading strategies,DFCO is algorithm proposed.Firstly,the algorithm divides the tasks with different delay requirements into three levels.For the queues with medium and low delay levels,the queue model of first-in first-out is used to store the tasks;and for high delay level queues,a preemptible heap queue model is proposed.At the same time,the algorithm selects the optimal computing sever for each task according to the resource availability and transmission delay,and jointly optimizes the packet loss rate of delay-sensitive tasks,the waiting delay and offloading delay of various tasks.Simulation results show that the proposed algorithm can better meet the multi-task application scenarios of IoT application scenarios.In the Mobile Edge Computing scenario of Internet of Things,The MAEE algorithm and DFCO algorithm proposed in this paper have improved the energy efficiency of the system,met the multi-task type of IoT application scenarios,and it can provide feasibility research ideas for the development of high-rate IoT applications.