The Impact of Increasing Number of Nodes on the Performance of Well-Known Trust and Reputation Models for Wireless Sensor Networks

Abstract

In recent years, several trust and reputation management models have been proposed to address the security issues of wireless sensor networks. In wireless sensor networks, trust and reputation management systems basically allow sensor nodes to make their own opinion about how trustworthy other nodes are so that a higher number of successful transactions can be obtained and the probability of being defrauded reduced. To assess the performance of trust and reputation management systems a number of performance metrics were proposed. In this study, with the aim of finding out the most suitable trust and reputation model when the number of sensor nodes involved in a wireless sensor network has been increased, the performance of EigenTrust, Linguistic Fuzzy Trust Mechanism, PeerTrust and PowerTrust is evaluated in terms of accuracy rate and path length. The reason for focusing on this is that if a trust and reputation model is able to achieve the same accuracy rate and path length performance without any performance degradation when more sensor nodes are involved in the network, it can be considered as scalable. The results of our simulation studies prove that compared to the other models, Linguistic Fuzzy Trust Mechanism provides higher accuracy and less path length scores and is more suitable for large-scale deployments of wireless sensor networks.