Research Article
PrivaSense: Privacy-Preserving and Reputation-Aware Mobile Participatory Sensing
@INPROCEEDINGS{10.4108/eai.7-11-2017.2273602, author={Hayam Mousa and Sonia Ben Mokhtar and Omar Hasan and Lionel Brunie and Osame Younes and Mohiy Hadhoud}, title={PrivaSense: Privacy-Preserving and Reputation-Aware Mobile Participatory Sensing}, proceedings={14th EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services}, publisher={ACM}, proceedings_a={MOBIQUITOUS}, year={2018}, month={4}, keywords={participatory sensing privacy reliability reputation re-identification attacks}, doi={10.4108/eai.7-11-2017.2273602} }- Hayam Mousa
Sonia Ben Mokhtar
Omar Hasan
Lionel Brunie
Osame Younes
Mohiy Hadhoud
Year: 2018
PrivaSense: Privacy-Preserving and Reputation-Aware Mobile Participatory Sensing
MOBIQUITOUS
ACM
DOI: 10.4108/eai.7-11-2017.2273602
Abstract
The integration of privacy into reputation systems is a crucial need for building secure and reliable participatory sensing applications. Participants are given the assurance that their privacy is preserved even if they contribute some personal sensitive data. In addition, reputation systems allow an application server to monitor participants' behaviors and evict those who provide the system with corrupted data. However, this integration requires achieving seemingly conflicting objectives. Reputation systems monitor participants behaviors along subsequent interactions. Whereas, one of the major objectives of privacy preserving systems is to unlink subsequent interactions. In this paper, we define a new attack (RR attack), which exploits this conflict in order to detect the succession of contributions provided by the same participant and to subsequently re-identify his original identity. We show that using this attack, more than 35% of contributions can be associated to their successive contributions in each campaign. We then propose PrivaSense as a new privacy preserving reputation system that integrates both reputation and privacy such that their objectives are simultaneously achieved. Experimental results are conducted using a real data-set. These results show that PrivaSense decreases by up to 80% the number of contributions linked to their original providers.
