An abstract model for incentive-enhanced trust in P2P networks

Abstract

Peer-to-Peer (P2P) networks have emerged as a prime research topic, partly due to the vast unexploited possibilities unrestricted distribution of the workload provides. The main hindrance for unrestricted exploitation of the P2P topology is, due to lack of security-related issues, the gullible attitude taken towards unknown agents. Therefore, the severity of the vulnerabilities caused by gullibility must be mended by other means, for example, by an effective incentive scheme encouraging agents to trustworthy behaviour. This paper presents an abstract model for incentive enhanced trust, to progressively assign the participating agents rights for accessing distributed resources, emphasising consistent behaviour. The model consists of a degrading formula, an illustrative incentive triangle and a best-effort distributed supervision model. Moreover, the same incentive model facilitates anticipation of future behaviour concerning any given agent founded on several distinct agents' opinion, suggesting that any knowledge concerning the counterpart is better than none. © IFIP International Federation for Information Processing 2005.