Research On Shared Resource Allocation Algorithm In Mobile Edge Computing Environment

Cloud computing technology has been widely used in the past decade,and it relies heavily on the centralization of computing and data resources.Because users are far away from cloud data centers,there is a certain delay when accessing these data centers.In order to further reduce latency,the Mobile Edge Computing(MEC)paradigm has been proposed,which transfers network resources near mobile devices to reduce latency.In mobile ad-hoc MEC,multiple mobile devices cooperate with each other,and resource-rich mobile devices(MDs)share their resources to resource-poor mobile devices.The resource allocation algorithm is one of the key elements that determines system performance.Therefore,this paper studies how to propose an effective resource allocation mechanism to promote resource allocation among multi-user equipment in the MEC system.This paper assumes that there are multiple access points(APs)in the system to assist in the discovery and matching of idle resources.Specifically,research is performed on two different scenarios: 1)Resource sharing and allocation between mobile devices in multiple APs non-cooperative scenarios;2)Resource sharing and allocation of mobile devices in multiple APs cooperative scenarios.For the above two scenarios,this paper proposes effective resource allocation algorithms to optimize system performance and utility.The main work of this paper is as follows:1.For the scenario of multiple non-cooperative APs,this paper studies how to design an effective resource auction mechanism to match multiple supplying resource devices and demanding resource devices.Combining the difference in favor among resource demanders for different resource suppliers,this paper proposes a multi-user resource allocation algorithm based on double auctions,which aims to effectively utilize the resources provided by supplying resource devices.The algorithm can benefit from resource allocation.Furthermore,theoretical analysis proves that the demanding resource devices in the auction mechanism satisfy the economic characteristics such as truth,budget balance and individual rationality.2.For the scenario of multiple cooperative Aps,this paper studies the resource allocation problem coordinated by multiple Aps among multiple mobile devices.Specifically,in our model,each AP determines the optimal price and salary to attract demanding resource devices and supplying resource devices,respectively.Based on the total supply and demand of each AP,all APs perform a demand matching process to maximize social utility.In order to solve the problem of joint pricing and request matching,this paper first expresses the problem as an integer program problem,then proposes an approximate algorithm for this problem,and gives the upper limit of the gap between the performance of the proposed algorithm and the optimal solution through theoretical analysis.3.In this paper,the proposed algorithm is simulated and analyzed in detail.Aiming at the resource allocation between mobile devices in the scenario where multiple APs do not cooperate,we verify that the proposed auction-based algorithm can significantly improve the efficiency of MEC,and can maintain economics such as individual rationality.For the resource allocation of mobile devices in the collaboration scenario between multiple APs,verify the performance of joint pricing and request matching algorithms.