Achieving Secure and Privacy-Preserving in Mobile Social Networks

Introduction

A mobile social network fueled with heterogeneous wireless infrastructures (e.g., cellular/ WiFi) and mobile devices (e.g., smartphones, tablets), which can facilitate multimedia services by providing ubiquitous connections between service providers and users in a mobile environment (Zhang et al., 2014). Mobile social networks still face many security and privacy challenges, including private information leakage, cheating detection, Sybil attacks, DDoS attacks and so on (Liang et al., 2014). Based on spoofed identities, pseudonyms, locations, and profiles, an adversary can launch active or passive attacks (deliberately delays, drops, corrupts, or modifies messages) in order to steal the social data as well as to damage P2P communications (Ferrag et al., 2016). According to the work (Ferrag et al., 2017), the privacy preservation models for mobile social networks can be divided into location privacy, identity privacy, anonymity, traceability, interest privacy, backward privacy, and content-oriented privacy.

In this chapter, to address both security and performance challenges in ad hoc social networks, we propose a security framework for achieving secure and privacy preserving in mobile social networks, called ASPP, for ad hoc social communications. With the proposed ASPP scheme, each node user can be privacy-preserving authenticated before joining other nodes using routing protocol. The contributions of this chapter are fourfold.

• •

  First, we formalize the system model where we consider the social characteristics, i.e., human mobility, human group and preferences in a typical MANET which consists of trusted authority (TA), some stationary social unit (SU) deployed at the social space, and a large number of mobile equipped with WiFi technology moving on a social space. Next, we improve the AODV routing protocol basing on some concepts of social theory to be suitable for ad hoc social communications, i.e., degree centrality, closeness centrality, and betweenness centrality.

• •

  Second, we propose an efficient certificate scheme, where the TA issues the private key and certificate using the Schnorr signature algorithm (Schnorr, 1991). The node n_i can verify the certificate by the procedure S.check and cannot use these certificates directly in ad hoc social communication. Based on the proxy re-signature cryptography technology (Toshiyuki et al., 2013), the node request N_y resignature key from S_x and then re-signs the certificates issued by the TA to be the same as those issued by S_x itself. With this method of key distribution, the proposed ASPP scheme guarantees the node identity confidentiality.

• •

  Third, we provide conditional privacy preservation to the nodes with demand response. Although the SU act as certificate issuers in ASPP, they do not know what certificates are held by a node. Therefore, the adversaries cannot trace the interested nodes although they had compromised all SU. Then, to detect and verify the attack against routing protocol, based on three control messages {Detectreq, Detectrep, Notifreq}, each node initiates a response request and send Detectreq signed to nodes in its routing table at 1-hop, and waits the response of his request for runs the notification phase. After receiving a request response, the node checks its signature. If valid and , it considers that the link with the node is proved. Otherwise, it returns suspicious and starts the notification phase.

• •

  Finally, to validate the efficiency and effectiveness of the proposed ASPP, we integrate in the AODV implementation, being the modified protocol designated AODV-ASPP. Extensive simulation results in the first scenario show that the proposed ASPP scheme can detect the black hole attack more in the configuration where attack is launched on a number of more hops. Thus, in the second scenario, we focus on the transmission delay of ASPP at the node with extensive performance evaluation, which further convinces its practicality.