Dynamic Self-management of Autonomic Systems: The Reputation, Quality and Credibility (RQC) Scheme

In this report, we present a feedback-based system for managing trust and detecting malicious behavior in autonomically behaving networks. The two crucial insights that motivate our work are the notion of quality of a trust rating and the recognition as separate entities of the trust placed in a node and of the trust placed in the recommendations made by a node. These variables allow nodes to decide how much confidence they can place in the rating. We implement our scheme on a structured P2P network, Pastry, though our insights can be extended to unstructured systems and generic autonomic communication systems as well. Experimental results considering different models for malicious behavior indicate the contexts in which the RQC scheme has a better performance as compared to existing schemes in the literature.