Research On Incentive Optimization Strategy In Mobile Crowd Sensing

The recent surge in the number of mobile devices equipped with sensors,coupled with the inherent mobility of their users,has led to the rise of mobile crowd sensing(MCS).In a typical MCS system,the platform recruits users to perform the sensing tasks issued by data requesters.In order to stimulate users to participate in MCS,the MCS incentive mechanism should compensatively select the right set of workers to perform tasks,while optimizing platform profit or social welfare.In this paper,we consider a realistic MCS scenario,which relies on probabilistic collaboration between workers and assumes that the computing capacity of the platform is limited.Firstly,in this MCS scenario,we focus on the platform profit maximization(PPM),which is a NP-hard problem.In this paper,we propose the greedy algorithm MaxG for PPM.Because of the non-monotonicity of the objective function caused by the probabilistic collaboration of workers and the quality-based payment of data demanders,we propose RandG,which embeds the stochastic strategy into the greedy framework to seek the opportunity to skip the local optimum.This paper proves that RandG can reach a constant approximation ratio of 1/e.For general PPM problems,this paper proposes RandCom algorithm,which combines MaxG and RandG to achieve the highest platform profit.Finally,the effectiveness of the approximation algorithms in the three polynomial time is verified by experiments:the platform benefits of MaxG and RandG are higher than that of the baseline algorithm rand-K(randomly selected among the former K workers),and RandG will be better than MaxG in the case of multiple repetitions.Secondly,from the perspective of government departments,aiming at maximizing social welfare,this paper uses double auction mechanism and simply proposes MaxS algorithm.Firstly,the algorithm makes a preliminary screening of the worker's task completion rate,and then chooses the task demander with the greatest social welfare as the winner and joins the corresponding worker set into the winner's worker set.Finally,a simple experiment was conducted to evaluate the higher social welfare of MaxS.