سناریوهای تهدید امنیتی در الگوهای اعتبار و اعتماد به منظور سیستم های توزیع یافته (DS)

ABSTRACT Security threats scenarios in trust and reputation models for distributed systems

Trust and reputation models have been recently proposed by many researches as an innovative solution for guaranteeing a minimum level of security between two entities belonging to a distributed system that want to have a transaction or interaction.

Thus, many studies, works and models have been designed, carried out and developed in this direction, leading to a current solid research field on which both academia and industry are focusing their attention.

Many methods, technologies and mechanisms like fuzzy logic (Tajeddine et al., 2006), bayesian networks (Wang et al., 2006b) or even bio-inspired algorithms (Go´mez Ma´rmol, 2008) have been proposed in order to manage and model trust and reputation in systems such as P2P networks (Almena´rez et al., 2004), ad-hoc ones (Moloney and Weber, 2005), wireless sensor networks (Boukerche et al., 2007) (WSN) or even multi-agent systems (Sabater and Sierra, 2001).

Analyzing and studying some of these models (Josang et al., 2007; Sabater and Sierra, 2005) we realized that there are some security threats directly related to this specific kind of models, which are common and applicable to most of these approaches.

Nevertheless, we also noticed that each author proposed his/her own threats when testing their developed models, revealing the lack of a commonly agreed process of checking the robustness of a trust and reputation model against the mentioned risks.

This paper presents the most important security threat scenarios that can be found in the area of trust and reputation in a distributed system where some entities request some services and other ones provide those services. As far as we know, this is one of the first research works making such a thorough analysis.

We will describe each threat and propose a possible solution for tackling it. We will additionally study how some of the most representative models deal with those threats and analyze their proposed solutions.

Every accurate and robust trust and reputation model should have some mechanisms to effectively overcome all the threats that could be applied to it. Therefore, this work is intended to serve as a reference guide for developing secure trust and reputation models.

The rest of the paper is organized as follows: Section 2 makes a brief description of trust and reputation management and discusses the importance of dealing with the security threats exposed in Section 3, where some solutions are also proposed. A taxonomy of these threats is described in Section 4 and how some trust and reputation models face them is shown in Section 5. Finally, some conclusions and future work are depicted in Section 6.