Sybil node detection in peer-to-peer networks using indirect validation

Abstract

Peer- to- peer networks are used extensively today. Due to this wide use P2P networks is a target of malicious attacks. The most mentionable of them is the Sybil attack. Existing approaches for detection and mitigation of Sybil nodes are either computationally costly or are dependent on belief models found in social networks. It has been found that these belief models are themselves vulnerable to other attacks. In this paper, we propose a new type of indirect validation where we have a two stage validation in place to check that if a suspected node is Sybil or not. We crosscheck belief data from local monitor nodes and community detection data from randomly selected global monitor nodes, and then validate a suspected node. The proposed approach is found to be with less computation overheads and less vulnerable to malicious attacks.