Anomaly Detection Algorithm Based on Smart Contracts Behaviours in Ethereum Ecosystem

Abstract

Blockchain technology has attracted significant attention in recent years. One of the reasons for this new trend is the introduction of on-chain smart contracts, enabling the implementation of decentralized applications in trustless environments. Given the high amounts handled, smart contracts have become the targets of attacks. Along with its adoption, attacks are perpetrated via criminal smart contracts and vulnerable smart contracts by exploiting vulnerabilities. To thwart these attacks and avoid the exploitation of flaws, several approaches have been explored, such as the identification of vulnerabilities or the verification of models using a formal verification. However, these approaches have failed in solving the blockchain crime problem. This article proposes a new anomaly detection approach based on the behavior of smart contracts in the Ethereum blockchain. This formalism will be based on mathematical principles such as state mutation and an approach to verify the behavior of smart contracts.