SybilLimit: A Near-Optimal Social Network Defense against Sybil Attacks

TLDR

Decentralized peer‑to‑peer systems are highly vulnerable to Sybil attacks, and without a trusted central authority defending against them is difficult; prior work such as SybilGuard uses social networks but can admit many Sybil nodes and relies on an unverified fast‑mixing assumption. This paper introduces SybilLimit, a protocol that builds on SybilGuard’s insight but provides dramatically improved, near‑optimal guarantees. SybilLimit leverages the same social‑network‑based insight as SybilGuard, but employs a refined construction that yields tighter bounds on accepted Sybil nodes. Experiments on a million‑node system show SybilLimit reduces accepted Sybil nodes by about 200×, its theoretical guarantee is within a log n factor of optimal, and analysis of three large real‑world networks confirms that social networks are fast‑mixing, validating the protocol’s core assumption.

Abstract

Decentralized distributed systems such as peer-to-peer systems are particularly vulnerable to sybil attacks, where a malicious user pretends to have multiple identities (called sybil nodes). Without a trusted central authority, defending against sybil attacks is quite challenging. Among the small number of decentralized approaches, our recent SybilGuard protocol [H. Yu et al., 2006] leverages a key insight on social networks to bound the number of sybil nodes accepted. Although its direction is promising, SybilGuard can allow a large number of sybil nodes to be accepted. Furthermore, SybilGuard assumes that social networks are fast mixing, which has never been confirmed in the real world. This paper presents the novel SybilLimit protocol that leverages the same insight as SybilGuard but offers dramatically improved and near-optimal guarantees. The number of sybil nodes accepted is reduced by a factor of ominus(radicn), or around 200 times in our experiments for a million-node system. We further prove that SybilLimit's guarantee is at most a log n factor away from optimal, when considering approaches based on fast-mixing social networks. Finally, based on three large-scale real-world social networks, we provide the first evidence that real-world social networks are indeed fast mixing. This validates the fundamental assumption behind SybilLimit's and SybilGuard's approach.