Online Incentive Mechanisms For Mobile Participatory Sensing

In recent years, with the continuous development of hardware technology and mobile devices built-in sensors constantly enrich the development of mobile collaboration to provide a good sense of the hardware foundation. To achieve good service quality for a participatory sensing application, incentive mechanisms are necessary to attract enough participants. However, most existing incentive mechanism mainly apply for the offline scene where all users have uploaded the auction information in advances. Therefore, this kind of offline incentive mechanism is only applicable to users who can accept long time waiting.For those participates who are more mobile or have limited waiting time, even if they want to participate in the task, they cannot tolerate the return time of feedback from the offline mechanism. That is not good for the recruitment of platform.In this paper, a series of online participatory sensing incentive mechanisms were designed on the basis of existing participatory sensing technology. With the arriving of users one by one, the incentive mechanism can immediately give the auction results without waiting for all users to arrive. In this paper, we first established an online auction system model. The goal was to maximize the cost-effective of platform purchasing sensor data. The constraint conditions that the incentive mechanism should meet were computational efficiency, individual rationality, truthfulness, consumer sovereignty and competitiveness. Then, we presented an online incentive mechanism based on reverse auction and analyzed its ideal characteristics. However, this kind of online incentive mechanism cannot guarantee the truthfulness. So this paper continued to design an online incentive mechanism that considered truthfulness. And we proved its bid-truthfulness characteristic. Since the length of time that users stay in the system are different,this paper gave out the concepts of "zero arrival-departure interval" model and"non-zero arrival-departure interval" model. We considered that the first two incentive mechanisms are based on the zero arrival-interval model. And on the basis of that, we studied the online incentive mechanism of non-zero arrival-departure interval model. Then, we designed an online incentive mechanism based on general arrival-departure model. We proved how the time-truthfulness when the departure time is not equal to the arrival time. On this basis, we discussed a special case - the limited budget online incentive mechanism. We aimed to maximize the value of sensing data with a limitation of budget. After the design of this mechanism, we proved its budget feasibility. We also emphasize that our main idea and framework can also be applied to design online incentive mechanisms with other objectives such as frugality and profit maximization,while guaranteeing both consumer sovereignty and time-truthfulness that are seldom considered by existing work. Finally, we summarized the whole paper and prospect the future work. We put forward three key points that could be further studied: data quality and accuracy, data privacy and user priority.