An incentive scheme based on heterogeneous belief values for crowd sensing in mobile social networks

Crowd sensing is a new paradigm which exploits pervasive mobile devices to provide complex sensing services in mobile social networks (MSNs). To achieve good service quality for crowd sensing applications, incentive mechanisms are indispensable to attract more participants. Most of existing research apply only for the offline sensing data collections, where all participants' information are known a priori. In contrast, we focus on a more real scenario requiring a continuous crowd sensing. We model the problem as a restless multi-armed bandit process rather than a regular auction, where users submit their bids to the server over time, and the server choose a subset of users to collect sensing data. In this paper, to maximize the social welfare for the infinite horizonal continuous sensing, we design an incentive scheme based on heterogeneous belief values for joint social states and realtime throughput. Analysis results indicate that our algorithm is not only near optimal and stable, but truthful, individually rational, profitable, and computationally efficient.