Mobile Crowdsourcing Task Allocation with Differential-and-Distortion Geo-Obfuscation

In mobile crowdsourcing, organizers usually need participants’ precise locations for optimal task allocation, e.g., minimizing selected workers’ travel distance to locations. However, the exposure of users’ raises privacy concerns. this paper, we propose a location privacy-preserving allocation framework with geo-obfuscation protect during assignments. More specifically, make participants obfuscate their reported under guarantee two rigorous schemes, differential and distortion privacy, without involve any third-party trusted entity. order achieve differential-and-distortion geo-obfuscation, formulate mixed-integer non-linear programming problem minimize expected workers constraints privacy. Moreover, worker may be willing accept multiple tasks, organizer concerned utility objectives such as acceptance ratio in addition distance. Against background, also extend our solution multi-task multi-objective optimization cases. Evaluation results on both simulation real-world user mobility traces verify effectiveness framework. Particularly, outperforms Laplace obfuscation, state-of-the-art mechanism, by achieving up 47 percent shorter average data same level protection.