TrustVis: visualizing trust toward attack identification in distributed computing environments



The establishment of trust relations is recognized as an important approach to initiate cooperation between unfamiliar entities in distributed computing environments. Particularly, trust relations can play a key role to help nodes make decisions on service discoveries in wireless environments. Visualizing trust relations can help us understand and analyze the real-time threats that the distributed system is facing and consequently identify the attacks in order to deploy corresponding countermeasures. The main challenges to visualize trust in such distributed environments are as follows: (i) visually reorganizing and presenting trust relations to show underlying cooperation between attackers; (ii) mapping attacker behaviors to visual patterns in an ambiguity-free manner; (iii) properly organizing and encoding other contextual information to analyze how these factors play along with trust. In this paper, we introduce TrustVis, a tool that helps users visually analyze trust relations to identify attacks in a semi-automatic fashion. In parallel with a running trust evaluation engine that actively monitors trust relations, TrustVis reorganizes and presents trust relations with a matrix to map the cooperative attack schemes to visual patterns. By coordinating the trust matrix and other contextual information into a multi-faceted view, TrustVis incorporates the intelligence of domain experts to interactively monitor the networked system toward identifying both the attackers' identities and the adopted attack schemes. Copyright © 2012 John Wiley & Sons, Ltd.