Truthful Incentive Mechanism for Vehicle-based Nondeterministic Crowdsensing

Nowadays, vehicles have shown great potential in crowdsensing. To guarantee a good Quality of Service (QoS), stimulating enough vehicles to participate in crowdsensing is very necessary. In this paper, we focus on the incentive mechanism design in the vehicle-based nondeterministic crowdsensing. Different from existing works, we take into consideration that each vehicle performs sensing tasks along some trajectories with different probabilities, and each task must be successfully performed with a joint probability no less than a threshold. Designing an incentive mechanism for such a nondeterministic crowdsensing system is challenging, which contains a non-trivial set cover problem with non-linear constraints. To solve the problem, we propose a truthful incentive mechanism based on reverse auction, including an approximation algorithm to select winning bids with a nearly minimum social cost, and a payment algorithm to determine the payments for all participants. Through theoretical analysis, we prove that our incentive mechanism is truthful and individual rational, and we give an approximation ratio of the winning bid selection algorithm. In addition, we conduct extensive simulations, based on a real vehicle trace, to validate the performances of the proposed incentive mechanism.