Economics of Peer-to-Peer Mobile Crowdsensing

Abstract

Mobile crowdsensing is a new sensing paradigm relying on computation and storage capabilities of mobile devices. However, traditional server-client mobile crowdsensing models suffer from a high operational cost on the server, and hence a poor scalability. Peer-to- peer (P2P) mobile crowdsensing models can effectively reduce the server's operational cost, by leveraging the mobile devices' under-utilized computation and storage resources. In a P2P mobile crowdsensing model, the sensing data is saved and processed in mobile users' devices in a distributed fashion, and is shared among mobile users directly in a P2P manner. In this work, we focus on the incentive issue in such a P2P mobile crowdsensing model. Specifically, we propose a data market and a generic pricing scheme for the data sharing among data sensors and requesters. We analyze the user interactions in such a data market from a game theoretic perspective, and prove the existence and uniqueness of the market equilibrium. We further propose a generalized best response dynamics to reach the market equilibrium. Our theoretic analysis and numerical results indicate that the equilibrium social welfare decreases with the data transfer cost and data prices, while the ratio of the equilibrium social welfare to the maximum social welfare benchmark increases with the data transfer cost and data prices.