In distributed environments, Reputation Management Systems (RMSs) aim to estimate agents' trustworthiness by exploiting different sources of information. The distributed nature of these systems makes them vulnerable to several types of security attacks, and the response provided by a specific RMS depends on various factors, such as the algorithms adopted for estimating the reputation values and the communication protocols used to enable the cooperation among agents. This work examines the most important security attacks against RMSs and proposes a set of metrics for a quantitative evaluation of the RMS vulnerabilities. A parallel simulation framework is used to automatically give a vulnerability score to a RMS according to the computed metrics. Experiments performed on a case-study RMS show the effectiveness of the metrics we defined, and the convenience of using a simulation environment to support the design of a secure RMS.
|Numero di pagine||8|
|Stato di pubblicazione||Published - 2017|
All Science Journal Classification (ASJC) codes