Online truthfully incentive mechanisms with budget constraint for multiple overlapped tasks crowdsourced sensing

Mobile Crowdsourced Sensing (MCS) enables mobile users (MU) equipped with mobile devices to upload their sensed data to MCS application/platform and share these data with other mobile users. To achieve high quality of service, incentive mechanisms are designed to attract MUs' participation. However the existing incentive mechanisms are designed for independent tasks instead of correlated multiple tasks. On the contrary, we focus on this more practical scenario where multiple tasks are partially overlapped. We investigate the problem that users submit their private cost to the MCS platform and the platform aims select a subset of users for maximizing the value of each task with budget constraint. We propose an incentive mechanism meeting computational efficiency, individual rationality, budget feasibility and truthfulness. We evaluate the performance on the wireless sensing data set. The results show our mechanism outperforms the state-of-the-art mechanism 19.6% in terms of quantity of information collected from MUs.