Grouping-Enabled and Privacy-Enhancing Communications Schemes for VANETs

Abstract

A vehicular ad hoc network (VANET) is also known as a vehicular sensor network [Zhang, Lu, Lin, Ho & Shen (2008)] by which driving safety is enhanced through inter-vehicle communications or communications with roadside infrastructure. It is an important element of the Intelligent Transportation Systems (ITSs) [Wang et al. (2006)]. In a typical VANET, each vehicle is assumed to have an on-board unit (OBU) and there are road-side units (RSU) installed along the roads. A trusted authority (TA) and maybe some other application servers are installed in the backend. The OBUs and RSUs communicate using the Dedicated Short Range Communications (DSRC) protocol [Oh et al. (1999)] over the wireless channel while the RSUs, TA, and the application servers communicate using a secure fixed network (e.g. the Internet). Based on this infrastructure, vehicles can broadcast safety messages (e.g. road condition, traffic accident information), referred to as ”ad hoc messages”, to other nearby vehicles and RSU such that other vehicles may adjust their travelling routes and RSU may inform the traffic control center to adjust traffic lights for avoiding possible traffic congestion. Like other communication networks, security issues have to be well addressed. For example, the message from an OBU has to be integrity-checked and authenticated. Otherwise, an attacker can replace the safety message from a vehicle or even impersonate a vehicle to transmit a fake safety message. For example, an attacker may impersonate an ambulance to request other vehicles to give way to it or request nearby RSUs to change traffic lights to green. Besides, privacy is another important issue. A driver may not want others to know its driving routes by tracing messages sent by its OBU. Thus an anonymous communications protocol is needed. While being anonymous, a vehicle’s real identity should be revealable by a trusted party when necessary. For example, the driver who sends out fake messages causing an accident should not be able to evade responsibility by using an anonymous identity. In terms of integrity-checking and authentication, digital signature in conventional public key infrastructure (PKI) [Housley et al. (1999)] is a well accepted choice. However, requiring a vehicle to verify the signatures of other vehicles by itself using such schemes as in [Tsang & Smith (2008)] induces two problems as mentioned in [Zhang, Lin, Lu & Ho (2008)]. First, the computation power of an OBU is not adequate to handle all verifications in a short time, especially in places where the traffic density is high. Second, to verify a message from an unknown vehicle involves the transmission of a public key certificate which causes heavy message overhead. Therefore, the general approach is to let the nearby RSU help a vehicle verify the message of another. The volume of signatures to be verified can 9