To promote honest service provision in multi-agent systems, a Computational Trustworthiness and Rating scheme collects service ratings and computes agents' trustworthiness levels (TLs). Studies of existing schemes fail to reflect closed-loop dynamics of TLs arising from the mutual influence of agents' behavior and TLs, and to capture relevant properties. Recent simulative and analytical models produce results tied to a particular attack scenario, or restricted to small-size systems or simplistic agent behavior. We analyze a class of parameterized skimp and slander attacks in a challenging setting featuring agents' virtual anonymity, collusion, and blind, i.e., TL-insensitive selection of service providers, as well as unpredictable service availability and receptivity. We derive closed-loop Markovian TL dynamics and their mean-value approximation to analytically characterize agents' steady-state TLs, and find a systematic defense against skimp and slander.
Authors
Additional information
- DOI
- Digital Object Identifier link open in new tab 10.1109/trustcom60117.2023.00107
- Category
- Aktywność konferencyjna
- Type
- publikacja w wydawnictwie zbiorowym recenzowanym (także w materiałach konferencyjnych)
- Language
- angielski
- Publication year
- 2023
