AraTRM: Attack Resistible Ant-based Trust and Reputation Model

Trust and reputation models play an important part in ensuring the security of routing or data forwarding in networks, where the service consumers have no or little knowledge of the service providers. Trust and reputation models generate an accurate assessment for each identity in networks to provide enough information about the service provider and about how good the services offered. However the performance of a trust and reputation model may decline in potentially adversarial environments. In this paper, firstly, we listed a number of security threats applicable in the field of trust and reputation management. Secondly, we proposed statistical anomalies detection algorithms to cope with these threats. Finally, based on ant colony theory we designed a trust and reputation model, named AraTRM, and integrated these algorithms into AraTRM to evaluate the effectiveness of our proposed anomalies detection algorithms. Experimental results have shown that the proposed model is robust in adversarial environments and achieves high accuracy as well.