Hunting Sybils in Participatory Mobile Consensus-Based Networks

Abstract

We focus on detecting adversarial non-existent nodes, Sybils, in anonymized participatory mobile networks where nodes support both a node-to-server and peer-to-peer connection capabilities. As data-driven decisions within such networks typically rely on local consensuses, they are susceptible to adversarial injection attacks which impersonate honest nodes and overpower local data through forgery. First, we propose a scheme wherein nodes validate each other's presence through local peer-to-peer communication. We then observe a fundamental information asymmetry between Sybils and honest nodes, and argue that conventional Sybil detection techniques fail to exploit it. Thereupon, we propose a novel Sybil detection technique tailored to utilizing claimed location data, introduce a probabilistic framework for the problem, and design the statistical approach for finding Sybils. Finally, we compare our detection algorithm with existing methods through complex simulated Sybil scenarios.