Article Preview
Top1. Introduction
Vehicular Adhoc NETworks (VANETs) have received considerable attention by academic communities and in-dustrial companies (Hu, 2016). They aim to establish communications and collaboration between cars for which different safety, commercial and entertainment applications (Hadded et al., 2015) are developed. These applications required specific and stringent QoS and security requirements. Therefore, different Medium Access Control (MAC) (Hadded et al., 2015) protocols have been proposed and developed in order to handle the network access and transmission with minimum packet loss. One of the best-known techniques used by these solutions is Time Division Multiple Access (TDMA) that allows several vehicles to share the same frequency channel by dividing the channel into different time slots. Each vehicle can access the channel during its dedicated time slot to send data messages, while it can only receive messages during the time slots reserved for other vehicles. Nevertheless, various challenges remain unsolved due to the mobility of the vehicles and the security issues related to VANETs.
Indeed, the presence of a malicious vehicle that may disturb the whole communication process or which sends false information is a real possibility and dangerous in this kind of network. This issue has been the focus of several research and industrial studies (Mejri et al., 2014) that aim to ensure the security of VANETs. Security solutions in VANETs can be categorized into two classes. The first class aims to define new approaches to ensure the three essential security properties: Confidentiality, Integrity and Availability (CIA). Besides, the second one deals with the trust evaluation of the VANET entities. For example, in (Mejri et al., 2014), the authors identify existing security problems in VANETs and classifies them from a cryptographic point of view. Another study (Zeadally et al., 2012) de-fines a set of possible attacks, then classifies them according to the security property addressed and finally, deals with authentication based on digital signature solutions as in (Hu & Laberteaux, 2006) and (Hartenstein & Laberteaux, 2008). Based on Zeadally et al. (2012), this kind of technique is one of the best choices to secure communication in VANETs. Researchers are currently seeking to establish a trust-based framework for VANETs (Yan et al., 2013). To do so they must evaluate and integrate the concept of trust in order to enhance the security and QoS. Indeed, being able to have a secure communication with a vehicle does not necessarily mean that this vehicle is a trustworthy node. Sending false messages or perturbing the network with DoS attacks may occur even if secure communication is provided. Therefore, different solutions have been proposed in (Yan et al., 2013) to ensure a set of privacy requirements in VANETs. Mainly, they propose using mandatory access control to govern access to driver data and a trust evaluation mechanism based on the certification authorities. In this system, they design a trust-based message propagation. Each message is assigned a trust level based on different metrics including opinion-based trust, experience-based trust and role-based trust. However, there are no details on how to evaluate the experience and the opinion concerned. The authors do not address this problem which is critical in any trust framework. All these studies highlight the importance of trust in VANETs. Some of them define trust based on a key management framework; others define a trust evaluation system based on different metrics related to a specific algorithm or layer (routing, geo-localization, application based, etc.).