Research On Online Participants Selection Mechanisms In Mobile Crowdsourcing Sensing

Mobile smartphones have grown rapidly in recent decades.The ubiquitous smartphones and its owners have built a huge sensor network through which we can perform various data perceptions and acquisitions.We call this approach to mobile crowdsourcing sensing.Because the participants in the mobile crowdsourcing perception have the characteristics of mobility,randomness and so on.Therefore,choosing appropriate participants to complete the tasks is a very important issue in mobile crowdsourcing sensing.At present,many studies have begun to focus on solving the task allocation problem in mobile crowdsourcing sensing.However,existing researches ignored the arrival of participants,but only focuses on the selection process of participants.In order to solve this problem,this paper first considers a more realistic online scenario,and then uses linear programming to define our problem.Then,we propose a predictive model to predict the arrival of the participants.Firstly,we design the mechanism OSMPAZ through the predictive model to solve the problem under the zero arrival/departure model.Then,we design mechanism OSMPAG to solve the problem in a non-zero/arrival departure model.On the other hand,because the mobile crowdsourcing subtasks are usually small and simple.This makes it easy for participants with affinity to copy and transfer data to each other.Therefore,participants with affinity may collude to improve their benefit.In order to improve the quality of the data,we define the task assignment problem based on the affinity of the participants.Then,we first design mechanism AOSMZ at the zero arrival/departure model and then extend to the non-zero arrival departure model and design the mechanism AOSMG to solve the defined problem.Through rigorous theoretical analysis,we prove that the four mechanisms we proposed satisfy the expected properties such as computational efficiency,personal rationality,fairness,and cost-truthfulness.In addition,OSMPAG and AOSMG also meet the time-truthfulness.Finally,a large number of experiments can prove that the proposed mechanism can achieve better performance.