Research Article
Location Differential Privacy Protection in Task Allocation for Mobile Crowdsensing Over Road Networks
@INPROCEEDINGS{10.1007/978-3-030-92635-9_2, author={Mohan Fang and Juan Yu and Jianmin Han and Xin Yao and Hao Peng and Jianfeng Lu and Ngounou Bernard}, title={Location Differential Privacy Protection in Task Allocation for Mobile Crowdsensing Over Road Networks}, proceedings={Collaborative Computing: Networking, Applications and Worksharing. 17th EAI International Conference, CollaborateCom 2021, Virtual Event, October 16-18, 2021, Proceedings, Part I}, proceedings_a={COLLABORATECOM}, year={2022}, month={1}, keywords={Mobile Crowdsensing Location privacy Task allocation Road network Linear programming}, doi={10.1007/978-3-030-92635-9_2} }- Mohan Fang
Juan Yu
Jianmin Han
Xin Yao
Hao Peng
Jianfeng Lu
Ngounou Bernard
Year: 2022
Location Differential Privacy Protection in Task Allocation for Mobile Crowdsensing Over Road Networks
COLLABORATECOM
Springer
DOI: 10.1007/978-3-030-92635-9_2
Abstract
Mobile Crowdsensing (MCS) platforms often require workers to provide their locations for task allocation, which may cause privacy leakage. To protect workers’ location privacy, various methods based on location obfuscation have been proposed. MCS over road networks is a practical scenario. However, existing work on location protection and task allocation few considers road networks and the negative effects of location obfuscation. To solve these problems, we propose a Privacy Protection Task Allocation framework (PPTA) over road networks. Firstly, we introduce Geo-Graph-Indistinguishability (GeoGI) to protect workers’ location privacy. And then we model a weighted directed graph according to the road network topology and formulate a linear programming to generate an optimal privacy mechanism, which aims to minimize the utility loss caused by location obfuscation under the constraint of GeoGI. We also improve the time-efficiency of the privacy mechanism generation by using a δ-spanner graph. Finally, we design an optimal task allocation scheme based on obfuscated locations via integer programming, which aims to minimize workers’ travel distance to task locations. Experimental results on Roma taxi trajectory dataset show that PPTA can reduce average travel distance of workers by up to 23.4% and increase privacy level by up to 21.5% compared to the existing differential privacy methods.
