LIP3: A lightweighted fine-grained privacy-preserving profile matching mechanism for mobile social networks in proximity

Abstract

Recently, Device to Device (D2D) based mobile social networking in proximity (MSNP) has witnessed great development on smartphones, which enable actively/passively and continuously seek for relevant value in one’s physical proximity, through direct communicating with other individuals within the communication range, without the support of centralized networking infrastructure. Specially, a user would like to find out and interact with some strangers with similar interest in vicinity through profile matching. However, in matching process, individuals always have to reveal their personal and private profiles to strangers, which conflicts with users’ growing privacy concerns. To achieve privacy preserving profile matching (i.e., friend discovery), many schemes are proposed based on homomorphic and commutative encryption, which bring tremendous computation and communication overheads, and are not practical for the resource limited mobile devices in MSNP. In this paper we adapt Confusion Matrix Transformation (CMT) method to design a Lightweighted fIne-grained Privacy-Preserving Profile matching mechanism, LIP3, which can not only efficiently realize privacy-preserving profile matching, but obtain the strict measurement of cosine similarity between individuals, while other existing CMT-based schemes can only roughly estimate the matching value.