**2.8. Reputation systems**

A reputation system is a type of cooperative filtering algorithm which attempts to determine ratings for a collection of entities that belong to the same community. Every entity rates other entities of interest based on a given collection of opinions that those entities hold about each other[5,60].

Reputation systems have recently received considerable attention in different fields such as distributed artificial intelligence, economics, evolutionary biology, etc. Most of the concepts in reputation systems depend on social networks analogy. As expected, reputation systems are complex in the sense that they do not have a single notion, but a single system will consist of multiple parts of notions. Thus, comparing reputation systems is, in fact, a very difficult problem. All known trials on such problem were based on qualitative approach. The work in [61] proposes an attempt on comparing reputation systems quantitatively based on game theory. The authors, thus, identify different notions of reputation systems like, contextualization, personalization, individual and group reputation, and direct and indirect reputation.

Reputation systems are often useful in large online communities in which users may frequently have the opportunity to interact with users with whom they have no prior experience. Such cases are clearly applicable to e-commerce applications and on line auctioning sites like eBay[62] and Epinions[63]. Another important field that derives the same concept of enforced interaction among entities that lack priori experience on each other is the field of ad hoc and wireless sensor network. This is because nodes in such networks need to route each others' packets. Thus, a trust relation should exist among themselves.

In the context of MANET and WSN [5, 11, 64], the reputation of a node is the amount of trust the other nodes grant to it regarding its cooperation and participation in forwarding packets. Hence, each node keeps track of each other's reputation according to the behavior it observes, and the reputation information may be exchanged between nodes to help each other to infer the accurate values. There is a trade-off between efficiency in using available information and robustness against misinformation. If ratings made by others are considered, the reputation system can be vulnerable to false accusations or false praise. However, if only one's own experience is considered, the potential for learning from the experiences of others goes unused, which decreases efficiency.

Any reputation system in the context of MANET and WSN should, generally, exhibit three main functions [1, 65]:

