Fuzzy Inference Trust in P2P Network Environment

Abstract

How to build the mutual trust among peers in P2P network is crucial to secure distributed P2P applications. In this paper, we propose to use Mamdani-type fuzzy inference techniques in the design of P2P trust systems based on collecting and aggregating peers' trust value, which can better handle uncertainty and fuzziness in peer trust. There are five parts in the process of P2P trust fuzzy inference: fuzzification of the input variables, application of the fuzzy operator in the antecedent, implication from the antecedent to the consequent, aggregation of the consequents across the rules, and defuzzification. Fuzzy logic inference is clearly effective for distributed trust management in P2P networks.