A hybrid pricing mechanism for data sharing in P2P-based mobile crowdsensing

Abstract

Mobile crowdsensing (MCS) is becoming more and more popular with the increasing demand for various sensory data in many wireless applications. In the traditional server-client MCS system, a central server is often required to handle massive sensory data (e.g., collecting data from users who sense and dispatching data to users who request), hence it may incur severe congestion and high operational cost. In this work, we introduce a peer-to-peer (P2P) based MCS system, where the sensory data is stored in user devices locally and shared among users in an P2P manner. Hence, it can effectively alleviate the burden on the server, by leveraging the communication, computation, and cache resources of massive user devices. We focus on the economic incentive issue arising in the sharing of data among users in such a system, that is, how to incentivize users to share their sensed data with others. To achieve this, we propose a data market, together with a hybrid pricing mechanism, for users to sell their sensed data to others. We first study how would users choose the best way of obtaining desired data (i.e., sensing by themselves or purchasing from others). Then we analyze the user behavior dynamics as well as the data market evolution, by using the evolutionary game theory. We further characterize the users' equilibrium behaviors as well as the market equilibrium, and analyze the stability of the obtained equilibrium.