Privacy considerations for cloud-based positioning

Abstract

Cloud-based positioning provides better support for resource-constrained mobile devices; however, the user's location information is exposed during positioning when the computation is performed on the cloud. The improper exposure of location information could result in severe consequences that make users the target of fraudulent attacks. This study proposes a privacy-preserving localization scheme based on homomorphic encryption techniques in order to protect user privacy from both imminent attackers and untrusted cloud servers. The proposed algorithm exposes unreliable cloud only an encrypted version of the measurements and allows positioning to be performed in an encrypted domain. This scheme prevents cloud servers from understanding the computed results and avoid an adversary monitoring the transmission to log user behavior. On-site experiments show the feasibility of our approach. The results show that positioning in an encrypted domain would not affect accuracy. Experimental results also show that the proposed algorithm requires less computational overhead and achieves higher privacy level simultaneously compared to traditional encryption approaches.