Truthfully coordinating participation routes in informative participatory sensing

Shaofei Chen, Dengji Zhao, Alexandros Zenonos, Jing Chen, Lincheng Shen

Abstract: Participatory sensing is a promising approach with which people contribute sensory information to form a body of knowledge. In practice, people may have different ways to engage in a participatory sensing campaign. For example, there are several possible routes from a participant’s home to her office, where a route can be seen as a set of space-temporal coordinates, and measurements can be taken at these coordinates. To coordinate participation routes to collect more valuable information with a limited number of participants, a further concept, informative participatory sensing (IPS) has been developed recently. However, existing IPS systems lack incentive mechanisms to fight against the strategic behaviours of self-interested users. Hence, we propose a formal model of IPS where a service provider can coordinate individual schedules of self-interested participants. As the problem of informative path coordination is NP-hard, the well-known mechanism, Vickrey-Clarke-Groves (VCG) will be computationally inefficient to solve our problem. Given this, we design a sequentially sorting mechanism (SSM) for the model to allocate the schedules and determine the bonuses for these participants, and we then theoretically prove that SSM is computationally efficient, individually rational, profitable and truthful. Furthermore, we empirically evaluate our route allocation method in simulations and show that it significantly outperforms several benchmark approaches.