Dynamic Trust and Reputation-Based Framework for Opportunistic Networks
摘要
OppNets are a type of wireless network where the nodes interact with one another in an opportunistic manner without requiring a dedicated link to connect the source and the destination. It comes from the subclass of Delay-Tolerant Networks (DTN) known as mobile ad hoc networks (MANET). With OppNets, nodes are completely dependent on neighboring nodes for packet forwarding because there is no end-to-end communication. However, if these nodes are non-cooperative or malicious, it severely impacts the performance of the network, thus posing severe challenges in OppNet. This paper introduces DyTRON, (“Dynamic Trust and Reputation Framework designed for Opportunistic Networks”) that utilizes the Beta distribution’s probability density function (p.d.f.) to compute trust and reputation values, considering various network contexts. Trust is calculated dynamically by integrating two components: direct trust, which is the trust a node has in itself, and indirect trust is established based on recommendations or feedback received from adjacent nodes. Simulation results, conducted using the ONE Simulator, show that DyTRON outperforms established models such as Epidemic, BTRES, and PPHB+ across several key performance metrics, including average latency, message drops, overhead ratio, average hop count, and message delivery probability. The enhanced performance achieved by DyTRON contributes to the security of the network by reducing the likelihood of malicious nodes being involved in packet forwarding.