Secure and Privacy-Preserving, Timed Vehicular Communications

Abstract.

We consider the problem of privacy (anonymity) and security in vehicular (V2V) communication, in particular securing routine safety messages. Traditional public key mechanisms are not appropriate for such applications because of the large number of safety messages that have to be transmitted by each vehicle, typically one message every 100-300 ms. We first show that a recently proposed V2V communication scheme, the TSVC, based on the Time Efficient Stream Loss-tolerant Authentication (TESLA) scheme, is subject to an impersonation attack in which the adversary can distribute misleading safety information to vehicles, and propose a modification that secures it against such attacks. We then address general concerns regarding the inappropriateness of TESLA for vehicular applications (caused by the delayed authentication and buffer overflows), and propose a V2V communication scheme based on a variant of TESLA, TESLA⁰, for which there is no delay and packets are self-authenticating. This is appropriate for applications in which vehicles are in close proximity. Finally we combine both schemes to get a hybrid communication scheme that addresses in a flexible way the mobility requirements of V2V communications.urrent research in location-based services (LBSs) highlights the importance of cryptographic primitives in privacy preservation for LBSs, and presents solutions that attempt to support the (apparently) mutually exclusive requirements for access control and context privacy (i.e., identity and/or location), while at the same time adopting more conservative assumptions in order to reduce or completely remove the need for trust on system entities (e.g., the LBS provider, the network operator, or other peer nodes). This paper surveys the current state of knowledge concerning the use of cryptographic primitives for privacy-preservation in LBS applications.

Keywords: Vehicle-to-Vehicle communication, security, privacy, TESLA, timed hash chains

Download: (PDF file)