Research On Online Vehicle Recruitment Mechanism In Mobile Crowdsensing System

The mobile crowdsensing platform can assign sensing tasks to the intelligent networked vehicles in the form of vehicle recruitment,thus enhancing the dynamic sensing capability and scalability of the system.However,in scenarios of long-term sensing,such as population flow monitoring and environmental monitoring,the offline vehicle recruitment strategy is unable to guarantee continuous coverage of the target area and cannot accurately evaluate the long-term sensing coverage capability of the vehicle.To solve the above problems,in this paper,the intelligent networked vehicles with deterministic and non-deterministic trajectories are used as sensing nodes,and the vehicle trajectory model and spatiotemporal availability description model are designed to accurately evaluate the sensing coverage of the vehicles,The online vehicle recruitment framework and vehicle recruitment algorithm are designed to dynamically recruit intelligent networked vehicles to persistently sense the target areaFor the sensing scenario in which vehicles have deterministic trajectories,this paper presents a spatio-temporal segment model to describe the spatio-temporal characteristics of sensing tasks and vehicle trajectories.And it can be used to formulate a spatio-temporal availability evaluation scheme for vehicles,which effectively quantifies the task coverage and collection capability of vehicles.For the real-time participation of intelligent vehicles,we design an online vehicle recruitment framework for deterministic trajectories,which can periodically update the spatio-temporal availability of vehicles.Meanwhile,we design an effective online vehicle recruitment algorithm.The simulation results show that in the deterministic trajectory scenario,the online recruitment mechanism proposed in this paper achieves 98%coverage of the ideal offline algorithm,and the total number of recruited vehicles is smaller than the offline algorithm and the random recruitment algorithm.To address the sensing scenario where vehicles have nondeterministic trajectories,this paper presents the probabilistic space-time segment model to describe the non-deterministic trajectories of vehicles based on the space-time segment model.Then the paper formulates the problem of maximizing the expected recruitment benefits in this scenario,and designs a heuristic algorithm to solve this problem.In addition,this paper also considers the marginal pricing of vehicles under the expected recruitment cost to achieve effective incentives for vehicles,further optimizing the recruitment benefits.Finally,this paper improves the online recruitment framework to adapt the recruitment algorithm based on the probabilistic space-time segment model,which enables the system to dynamically recruit vehicles with non-deterministic trajectories.The simulation results show that in the non-deterministic trajectory scenario,the online recruitment mechanism proposed in this paper is about 81%of the ideal algorithm when the coverage rate is greater than 90%;in terms of average recruitment benefits,the proposed mechanism is much higher than the existing online algorithm,which is 93%of the ideal algorithm.