Security requirements for automotive on-board networks

Abstract

This paper considers security requirements for automotive on-board networks and describes the processes used for identifying and prioritizing such requirements. The security engineering process starts from use cases for automotive on-board networks that require wireless communication interfaces and involves an investigation of security threat scenarios and the assessment of the relative risks associated with the threats

Design and verification for transportation system security

“…Another way is to exploit the wireless communication interfaces such as Bluetooth and keyless entry system (e.g., the work in [4] shows unlocking a car without a car key is possible). Once the attackers have gained access to in-vehicle architecture platform, they can then exploit the vulnerabilities of aforementioned bus systems to deploy sensor spoofing attacks [12], replay attacks, masquerade attacks, DoS attacks, etc.…”

Attribute Decoration of Attack–Defense Trees

“…Edge et al [Edge et al, 2006] introduce protection costs; Byres et al [Byres et al, 2004] use detectability which illustrates how easily a defender can discover an attack and difficulty which specifies the needed skill of the attacker. As suggested by Henniger et al [Henniger et al, 2009], attack time -which models an amount of time needed to perform an attack -may be considered independent of an attacker's skill, costs or resources. Attacks that a defender cannot really protect himself against can be annotated as unmitigatable.…”

“…A similar approach has been used in [Jürgenson and Willemson, 2008], where the costs, success probability, gain and penalty attributes have been combined in order to define a new attribute called the exact expected outcome of the attacker. Henniger et al [Henniger et al, 2009] combine the attributes elapsed time, expertise, knowledge of system, window of opportunity and required equipment, in order to deduce the required attack potential. Finally, Fung et al [Fung et al, 2005] show how the difficulty level associated to the non-refined nodes can be used to estimate the survivability in the root node.…”

“…Difficulty [Byres et al, 2004, Fung et al, 2005, Tanu and Arreymbi, 2010, Henniger et al, 2009, Amenaza, 2001, Mauw and Oostdijk, 2005, Abdulla et al, 2010, Amoroso, 1994, Wang et al, 2011 The technical or social skill level needed for the attacker or defender to succeed.…”

“…False (F): No special skill is required. Time [Henniger et al, 2009, Schneier, 1999, Wang et al, 2011 For the attacker this is the time needed to perform the attack, independent of difficulty and costs of attack.…”