Integrated Elliptic Curve Cryptography and Feedback Incentive Scheme for Dynamic Trust Model in Multi- Agent Systems

Multi-agent systems are open, anonymous and dynamic in nature. Trust and reputation models fail to properly evaluate the trust of an agent if malicious agents behave in an unpredictable way. These models are ineffective in providing quick response to malicious agent’s oscillating behavior and critical for sustaining good service quality .The dynamic trust computation model called Secured Trust analyze different factors related to evaluate the trust of an agent but it fails to address Sybil attack and man in middle attacks. In this paper Elliptic curve cryptography and Feedback Incentive scheme is presented in-order to improve the trust model of multi-agent systems and also the load balancing algorithm is used to improve the quality of service among the service providers. This integrated scheme resists more influential malicious adversaries in the network and also provides efficient distribution of workload among service providing agents under stable condition. Keywords— Multi-agent System, Trust Management, Reputation Model, Load Balancing, Malicious Behavior.