A Differential Game Model for Data Utility and Privacy-Preserving in Mobile Crowdsensing

Hongjie Gao,Haitao Xu,Xianwei Zhou,Long Zhang

doi:10.1109/access.2019.2940096

Abstract

Mobile crowdsensing (MCS) is becoming an extremely pervasive sensing paradigm with the popularization of intelligent devices, which needs users to release their data to the sensing platform. But to the MCS system, user's privacy-preserving demands may be time-varying in the data releasing process. In addition, protecting data privacy and ensuring data utility is becoming a contradictory and critical issue, which results in a trade-off problem that needs to be solved. In this article, we construct a differential game model to solve the trade-off problem between the data utility and privacy preserving in mobile crowdsensing system, and solve the feedback Nash equilibrium solutions based on the dynamic programming in the MCS system. Based on the feedback Nash equilibrium solutions, users and the platform can achieve maximization of privacy requirement and data utility, respectively. Ultimately, a numerical simulation has been made to show the correctness of the proposed differential game model.

Highlights

Mobile crowdsensing (MCS) is becoming a very popular paradigm, which uses the tremendous sensing capability of various sensors to complete various sensing tasks [1] in a cost-effective approach
In order to solve the trade-off problem, we design a noncooperative differential game model [10], [11] to find if there exist an optimal strategies for the players, that is, to find Nash Equilibrium solutions for the trade-off problem to users and platform over time, no matter what noise addition mechanism has been used for data perturbation and what data aggregation mechanism has been adopted for data value to be implemented
FEEDBACK NASH EQUILIBRIUM SOLUTION OF THE GAME Based on the non-cooperative differential game model established for the mobile crowdsensing network in section IV, we try to solve the proposed model given in (3), (7), and (11), to find the feedback Nash equilibrium solutions, which will be considered as the optimal bounds for the data utility to platform, and the optimal bounds for the data privacy to the users

Summary

INTRODUCTION

Mobile crowdsensing (MCS) is becoming a very popular paradigm, which uses the tremendous sensing capability of various sensors (e.g., camera, tape-recorder, video camera, GPS) to complete various sensing tasks (e.g., personal health monitoring, pricing auditing, monitoring noise and ambiance, real-time traffic conditions) [1] in a cost-effective approach. H. Gao et al.: Differential Game Model for Data Utility and Privacy-Preserving in MCS adding some different grades noise before releasing the data to the platform according to the dynamic changing privacy demands, in order to protect their sensitive information [8]. In order to solve the trade-off problem, we design a noncooperative differential game model [10], [11] to find if there exist an optimal strategies for the players, that is, to find Nash Equilibrium solutions for the trade-off problem to users and platform over time, no matter what noise addition mechanism has been used for data perturbation and what data aggregation mechanism has been adopted for data value to be implemented. Different from the above scheme, we formulate a noncooperative differential game model to solve the trade-off problem between privacy and utility in MCS based on a dynamic privacy requirement condition.

RELATED WORK

LAPLACE MECHANISM

SYSTEM MODEL

FEEDBACK NASH EQUILIBRIUM SOLUTION OF THE GAME

NUMERICAL SIMULATIONS AND ANALYSIS

CONCLUSION AND FUTURE WORK