Skip to Main Content
Security is a significant factor in many sensor network systems, such as peer-to-peer networks. Most distributed systems are known to be particularly vulnerable to Sybil attacks. In a Sybil attack, a malicious node fakes multiple identities and pretends to be multiple, distinct nodes in the system. Then an edge may exit between a Sybil node and an honest node if a malicious node successfully fools a honest node into trusting it. In this paper, we propose a novel protocol to limit the influence of Sybil attacks by combining ant colony optimization (ACO) algorithm, where a node does random works; it will leave trails on the path. However, with the random work, the trails of the first node left on each node become diluted, at the end the routing it just fades away. Based on the nature of the ACO, we can limit the number of attack edge in an efficiently and conveniently. Hence, our defending system can ensure a honest node would accept and be accepted by other honest nodes in system with high probability, also, reject Sybil nodes with great probability.