Web3 Sybil avoidance using network latency

Web3 is emerging as the new Internet-interaction model that facilitates direct collaboration between strangers without a need for prior trust network participants and central authorities. However, one of its shortcomings lack defense mechanism against ability single user to generate surplus identities, known Sybil attack. has attack problem because it uses peer sampling establish connections users. We evaluate promising but under-explored direction avoidance using latency measurements, according which two identities with equal latencies are suspected be operated from same node, thus likely Sybils. Network measurements have desirable properties: they only malleable by attackers adding latency, do not require any participants. Our basic SybilSys avoids if actively exploit their malleability. present an enhanced version protects targeted attacks variant flow correlation attack, we name TrafficJamTrigger. show how message flows Round-Trip Time can used expose patterns propose six classifiers recognize these patterns. experiments show, through both emulation real-world deployment, serve fundamental role Web3, effectively establishing real users even in face networks consisting 99%