Impact of Message Authentication on Braking Distance in Vehicular Networks Jonathan Petit, Zoubir Mammeri IRIT - Paul Sabatier University Toulouse, France {Jonathan.Petit@irit.fr, Zoubir.Mammeri@irit.fr} Abstract: Transport safety applications aim at avoiding vehicular accidents by using secure broadcast vehicle-to-vehicle (V2V) communications. However, any security mechanism used for authenticating broadcast V2V messages comes with overhead in terms of computation and communications. The IEEE1609.2 standard for vehicular ad hoc networks is based on the ECDSA algorithm for supporting the authentication mechanism. This paper provides an assessment of the processing and communication overhead of ECDSA. We analyze the impact of this mechanism on VANET performance. Then we focus on the impact of authentication communication on the braking distance. Keywords: VANET, ECDSA, authentication overhead, braking distance, message authentication 1. Introduction Due to the huge life losses and the economic impacts resulting from vehicular collisions, many governments, automotive companies, and industry consortia have made the reduction of vehicular fatalities a top priority [1]. On average, vehicular collisions cause 102 deaths and 7900 injuries daily in the United States, leaving an economic impact of $230 billion [2]. The damage is similarly devastating in the European Union, where there are more than 110 deaths and 4600 injuries daily, costing €160 billion annually [3]. A major evolution for the automotive industry is the context awareness, meaning that a vehicle is aware of its neighborhood. Modern cars now include a set of processors connected to a central computing platform that provides many wired and wireless interfaces. Smart vehicles are those vehicles that are equipped with On-Board Unit (OBU), which has recording, processing, positioning, and location capabilities and that supports wireless security protocols. Roads can be made smart, too. Road- Side Units (RSU) installed along a road can inform passing vehicles about the road traffic conditions. With more smart cars and roads [4], we can expect many changes. Particularly, it is expected that the number and severity of accidents should decrease. Automotive safety applications aim to assist drivers in avoiding vehicular accidents, by providing advisories and early warnings to drivers, using broadcast vehicle-to-vehicle (V2V) communications. Vehicles typically communicate as per the Dedicated Short Range Communication standard (DSRC) [5], and broadcast messages in response to certain notified events (emergency message) or periodically (beacon message) [6]. V2V communications enable an entire space of applications, in addition to automotive safety, as infotainment and commercial. Since drivers of vehicles participating in V2V communications are expected to act on messages received from other participants, it is clearly necessary that these messages be transmitted in a secure fashion. In order to secure vehicular communications, Wireless Access in Vehicular Environments (WAVE) architecture mandates the use of PKI mechanisms, where service application messages are encrypted and vehicle safety messages are digitally signed. All implementations of IEEE1609.2 standard [7] shall support the Elliptic Curve Digital Signature Algorithm (ECDSA) [8] over the two NIST curves P-224 and P-256. Unfortunately, security mechanisms come with overhead that affects the performance of the V2V communications, and hence that of the safety applications. Many of the envisioned safety and driver-assistance applications require tight deadlines for message delivery. Consequently, security mechanisms must take these constraints into consideration and impose low processing and communication overhead. In this paper, we assess the processing and communication overhead of the authentication mechanism provided by ECDSA. As the braking distance is an important metric in emergency braking application, we investigate the impact of the authentication key size on the braking distance. We analyze the effects of signature overhead and network density. The paper is organized as follows. First, we survey previous work. The overhead of authentication mechanism is discussed in section 3. In section 4, we present simulations results of safety message and discuss the impact of ECDSA on the braking distance. Section 5 concludes the paper.