Privacy-Preserving Multiobjective Task Assignment Scheme with Differential Obfuscation in Mobile Crowdsensing

Abstract

In Mobile crowdsensing (MCS) applications, the system often performs optimal task assignments based on the participants’ information, e.g., the user’s precise location or bids to improve stability and efficiency. However, the collection and analysis of users’ data may jeopardize their privacy. Existing privacy protection methods of task assignment focus primarily on single objective constraint conditions and have rarely considered the comprehensive protection of user privacy. In this paper, we propose a privacy-preserving multiobjective task assignment scheme with differential obfuscation (PMTA) by comprehensively considering multi-constraint conditions and multi-dimensional privacy preserving of sensor users. Without the need for any Trusted Third Party (TTP), PMTA enables participants to obfuscate their sensitive data by utilizing differential privacy techniques. Based on game theory, PMTA uses the non-dominated sorting genetic algorithm II (NSGA-II) to optimize multiple objectives that minimize the expected travel distance of selected workers and the cost of task publishers. Extensive experimental results verify the effectiveness and efficiency of our scheme. Particularly, PMTA is superior to the scheme that only considers a single objective, reducing 41.3% average travel distance and 64% average cost, and is also superior to Laplace’s obfuscation scheme reducing 25.6% average travel distance and 63.3% average cost of expenditure.

Keywords: Mobile crowdsensing, Task assignment, Multiobjective optimization, Privacy preservation, Differential privacy