Chengjie Gu

Pricing Incentive Mechanism based on Multi-stages Traffic Classification Methodology for QoS-enabled Networks

We propose a novel pricing incentive mechanism based on multi-satges traffic classification methodology supporting load balancing and allocating network resource efficiently for QoS-enabled networks in this paper. We integrate the pricing with QoS routing and present a novel pricing incentive mechanism meeting the QoS requirements of different applications. This mechanism provides an equitable pricing incentive for applications according to their service requests. A novel multi-stages traffic classification methodology that brings together the benefits of port mapping, signature matching and flow character classification techniques is motivated by variety of network activities and their QoS requirements of traffic. We study the pricing and different levels of services in detail and integrate admission control scheme and load balancing in our framework. By theoretical analysis and extensive simulations, we prove its effectiveness in making traffic load balance and providing an incentive for users to utilize network resources to users’ satisfaction.

A Novel Trust Management Model for P2P Network with Reputation and Risk Evaluation

A novel decentralized trust management model with reputation and risk evaluation is proposed (RFTrust). Trust is an essential ingredient of the transaction process. Trust and risk are two closely related factors to make transaction in an uncertain environment. RFTrust model establish the credibility by calculating current reputation and acquiring the risk evaluation which represents the accumulative assessment of the long-term transaction behavior. Simulation and theoretical system show that RFTrust has advantages in combating various malicious behaviors such as dishonest feedbacks from malicious peer, the collusion and the strategic attacks to the trust model itself, and surpasses the current trust models. This model is more robust on trust security and more predominant in successful transaction rate than other trust mechanisms.