Research On Multi-star Edge Computing Offloading Strategy For Low Earth Orbit Satellite Systems

Currently,the research work on the sixth generation mobile networks(6G)is progressing in an orderly manner.The vision of 6G is to build a communication network that integrates land,sea,air,and space,and construct a world where all things are intelligently connected.However,terrestrial networks are limited by technology and cost,making it difficult to achieve seamless global coverage,in addition,natural disasters can easily lead to damage to terrestrial networks.Low Earth Orbit(LEO)satellite systems have the advantages of wide coverage,high stability,and low latency,and are expected to build a seamless global communication network.Meanwhile,with the rapid development of mobile communication and the substantial growth of smart devices,a large number of new applications have emerged.Its strict latency requirements make Mobile Cloud Computing(MCC)no longer applicable.Mobile Edge Computing(MEC),with its proximity to data sources,enables future networks to efficiently support new types of applications.Combining LEO satellites with MEC is expected to realize the vision of 6G.However,it also faces the problems of resource limitation and mobility of LEO satellites,uneven distribution of users,large number and diversity of tasks.The computation offloading strategy is the main reason that affects user’s task computation delay and energy consumption,so it is especially important to design a reasonable computation offloading strategy.This paper addresses the above issues,we study the computation offloading strategy for the multi-star MEC scenario of LEO satellite system based on static resource allocation and dynamic resource allocation,respectively.The specific research of the paper is as follows:1.Joint optimization Offloading Decision and Resource Allocation(ODRA)strategy is proposed based on static resource allocation,with the user’s task computation delay and energy consumption as the optimization objectives.A particle swarm algorithm is used to solve the offloading decision problem and a Lagrange multiplier method is used to solve the resource allocation problem.Simulation results show that this strategy can effectively reduce the user’s task computation delay and energy consumption compared to the comparison strategy.2.Based on dynamic resource allocation,the study is carried out based on Orthogonal Frequency Division Multiple Access(OFDMA)and Time Division Multiple Access(TDMA),respectively,with the user’s task computation delay and energy consumption as the optimization objectives.Firstly,the joint optimization Offloading Decision and Dynamic Resource Allocation(ODDRA)strategy is proposed based on OFDMA,which dynamically divides resources equally according to the number of tasks handled by LEO satellites at the current moment.The matching-coalition algorithm is proposed to optimize the offloading decision.The joint optimization Offloading Decision and Task Offloading Sequence(ODTOS)strategy is proposed based on the TDMA dynamic time slot allocation,and the task offloading sequence problem is modeled as a permutation flow shop scheduling problem with minimized total flowtime,the heuristic(LR)algorithm proposed by Liu and Reeves is used to minimize the total time delay,while the proposed joint matching-coalition algorithm is used to optimize the offloading decision.Simulation results show that based on the same multi-access approach,the proposed strategy can effectively reduce the user’s task computation delay and energy consumption compared to the comparison strategy.