Bayesian network based trust model with time window for Pure P2P computing systems

In recent years P2P computing systems gain popularity and impact due to the anonymity, dynamism, scalability, and cost sharing. Due to autonomy of peers and open environment, security is big issue in P2P computing systems. Traditional security mechanisms used in centralized systems are not suited for P2P computing systems because there is no central point of authority. The working of a P2P computing system is highly dependent on collaboration of peers. To improve the collaboration between peers many trust management systems have been developed recently. This study proposed a Bayesian network based trust model with time window to measure the trustworthiness of task processor peer in pure P2P computing framework. The proposed model is compared with Bayesian network based trust model using JADE 4.3.0. This shows that proposed model identify malicious peers earlier than the Bayesian network based trust model in a P2P computing system.