Research On Task Allocation And Pricing Strategy In Edge Computing Based On Combinatorial Auction

As an emerging computing paradigm,edge computing can provide lower network latency,more efficient data transmission and faster computing processing,so more and more tasks are handed over to edge servers for processing.Different from cloud computing,edge computing servers are scattered and deployed near mobile devices,and users can offload computing-intensive,delay-sensitive tasks that cannot be carried by mobile devices to edge servers for processing.However,due to the small-scale and lightweight of edge servers,edge computing still faces the challenge of rationally and effectively utilizing edge server computing resources and network channel resources according to the characteristics of computing tasks.Specifically,first of all,there is often a certain correlation between computing tasks,and users have combined preferences for the execution of computing tasks,which will lead to the fact that the optimal allocation of computing tasks cannot be completed in polynomial time;secondly,the completion of computing tasks will be accompanied by the use of network channel resources,it is challenging to comprehensively consider computing task execution,network bandwidth allocation,and resource pricing.Existing research work shows that the combinatorial auction mechanism can well describe the bidirectional interaction between edge servers and mobile devices.Therefore,in response to the above problems,this paper conducts research on edge computing task allocation and resource pricing based on combinatorial auctions.The main research work is as follows:(1)A continuous computing task allocation and pricing method based on combinatorial auction is proposed.Activity on Edge Network(Ao EN)is introduced to model continuous tasks,and the computing task allocation problem in edge computing is formalized as a social welfare maximization problem.Based on the combinatorial auction approach,an approximate algorithm is designed to solve the winner determination problem in polynomial time,and a pricing strategy that guarantees the truthfulness and individual rationality of the auction participants is proposed.The approximate ratio of the proposed algorithm is proved theoretically and analytically.Experiments verify that the proposed auction mechanism can guarantee authenticity,individual rationality,and computational efficiency while considering continuous task allocation and pricing.(2)A computing task offloading strategy for joint computing resources and bandwidth allocation is studied.An online combinatorial auction algorithm mechanism based on pricing function is proposed.According to the network bandwidth and resource occupancy,a dynamic pricing function is designed to assist in the decision-making of network bandwidth allocation and computation offloading schemes.Introducing the primal-dual technique,in the absence of future information,real-time decisions are made for mobile users who arrive dynamically over time,which solves the problem of winner decision and payment decision in computation offloading.Through theoretical proof and comparative analysis of experiments,the approximate rate of the algorithm is obtained,and the proposed mechanism satisfies the properties of dominant-strategy incentive compatibility and computational efficiency.