Probabilistic reasoning in intelligent systems: networks of plausible inference
Probabilistic reasoning in intelligent systems: networks of plausible inference
NSPW '97 Proceedings of the 1997 workshop on New security paradigms
Performance analysis of the CONFIDANT protocol
Proceedings of the 3rd ACM international symposium on Mobile ad hoc networking & computing
Core: a collaborative reputation mechanism to enforce node cooperation in mobile ad hoc networks
Proceedings of the IFIP TC6/TC11 Sixth Joint Working Conference on Communications and Multimedia Security: Advanced Communications and Multimedia Security
Decentralized trust management
SP'96 Proceedings of the 1996 IEEE conference on Security and privacy
Loopy belief propagation for approximate inference: an empirical study
UAI'99 Proceedings of the Fifteenth conference on Uncertainty in artificial intelligence
| Hi-index | 0.00 |
Trust management, broadly intended as the ability to maintain belief relationship among entities, is recognised as a fundamental security challenge for autonomous and self-organising networks. In this work, we focus on the evaluation process of trust evidence in distributed networks, where no pre-established infrastructure can be assumed. After casting the problem into the framework of estimation theory, a distributed maximum likelihood trust estimation algorithm is proposed. Strong parallels with spin glasses theory are shown, providing key insights about the algorithm performance and limitations, as well as useful formulas for parameters tuning. The problem is also formulated as an inference problem on a Markov random field, and an alternative fully distributed algorithm based on message passing techniques is then proposed. This work presents a mathematically rigorous analytical approach to the trust management problem, and proposes the use of statistical physics methods not only to understand the complex dynamics that arise from the interactions of peers in decentralised networks but also to design robust protocols whose performance can be rigorously evaluated.