HIT: A Human-Inspired Trust Model
Résumé
This paper presents a new approach to assign trust levels in ad hoc networks. Our system is inspired by the human concept of trust. The trust level considers the recommendation of trustworthy neighbors and their own experience. For the recommendation computation, we take into account not only the trust level, but also its accuracy and the relationship maturity. We also propose the Recommendation Exchange Protocol (REP), which minimizes the number of exchanged messages. The results show the efficacy of the system and the influence of main parameters.