Front-Running Protection for Distributed Exchanges using Tamper-Resistant Round Trip Time Measurements

TL;DR

ODIN is a decentralized front-running protection system that estimates network latency indirectly to prevent manipulation by malicious peers in P2P exchanges.

Contribution

It introduces a novel, tamper-resistant RTT measurement algorithm for secure, decentralized front-running protection in hostile network environments.

Findings

ODIN's RTT estimation is accurate and practical.

The system effectively prevents front-running in P2P exchanges.

The indirect RTT measurement method resists tampering by malicious peers.

Abstract

In this paper we present ODIN, a front-running protection system that uses a novel algorithm to measure Round-Trip-Time (RTT) to untrusted servers. ODIN is the decentralized equivalent of THOR, a RTT-aware front-running protection system for trading on centralized exchanges. Unlike centralized exchanges, P2P exchanges have potentially malicious peers which makes reliable direct RTT measurement impossible. In order to prevent tampering by an arbitrarily malicious peer, ODIN performs an indirect RTT measurement that never interacts directly with the target machine. The RTT to the target is estimated by measuring the RTT to a randomized IP address that is known to be close to the target's IP address in the global routing network. We find that ODIN's RTT estimation algorithm provides an accurate, practical, and generic solution for collecting network latency data in a hostile network environment.