Research On Edge Server Deployment And Crowdsourcing Task Assignment In Mobile Computing

Mobile edge computing brings computing and storage resources to the edge of the mobile network,enabling mobile devices to run resource-intensive applications,such as augmented reality and online games,while meeting strict latency requirements.In recent years,many researches have been carried out on the hot spots of mobile edge computing,such as computing offloading,low latency and energy efficiency,and good experimental results have been obtained.However,few researchers study the deployment of mobile edge servers.The quality of edge server deployment directly affects capital expenditures,as well as the flexibility of deployment and configuration.The popularity of mobile devices has also promoted the development of spatial crowdsourcing.Spatial crowdsourcing is a popular distributed problem solution that uses the power of mobile participants to perform location-based tasks,such as checking product placement and taking photos of landmarks.Usually,a participant need to travel physically to the target location so as to finish the assigned task.Therefore,the participants' familiarity with the target location directly affects the completion quality of the task.In addition,from the perspective of the spatial crowdsourcing platform,it is hoped that all tasks can be completed with lower recruitment costs.This article will study the two hotspots of mobile edge computing and spatial crowdsourcing.The main work contains:(1)We propose a multimodal deployment scheme for edge servers in mobile edge computing,while optimizing task allocation to minimize the average system response time.In this scheme,a heuristic algorithm based on particle swarm algorithm is proposed,which improves the update scheme of the position of particle.At the same time,it combines the niching technology to obtain multiple competitive server deployment solutions in one operation.In the simulation experiment,through the comparison of three existing algorithms,the superiority of the algorithm proposed in this study on the real data set is demonstrated.(2)We propose a bi-objective optimization problem in spatial crowdsourcing to jointly optimize participants' location familiarity with assigned tasks and the recruitment cost of the spatial crowdsourcing platform.When modeling the problem,we analyze the participant's previous trajectory information and introduce the forgetting curve to measure the participant's location familiarity with the task.For this study,we respectively proposed a divide-and-conquer algorithm based on the constraint method and a heuristic algorithm based on the simulated annealing algorithm.In the simulation experiment,we use a real dataset to conduct experiments on the two proposed algorithms and the existing algorithms with different parameter settings.Experimental results show that the proposed algorithms can always obtain more competitive results regarding the two optimization goals.