Piecing Together the Jigsaw Puzzle of Transactions on Heterogeneous Blockchain Networks

The Web3 ecosystem is increasingly evolving to include multiple chains, with decentralized applications (dApps) distributed across different blockchains. This has driven the need for cross-chain bridges to enable blockchain interoperability. However, this opens up novel attack surface, and media outlets have reported serious attacks related to cross-chain bridges. Nevertheless, few prior works have studied cross-chain bridges and their related transactions, especially from a security perspective. To fill the void, this paper presents the first comprehensive analysis of cross-chain transactions. We first build a cross-chain transaction dataset based on semantic analysis of popular cross-chain bridges, covering 13 decentralized bridges and 7 representative blockchains, with over 80 million transactions in total. Based on this comprehensive dataset, we study the landscape of cross-chain transactions from several angles, including token usage, user profile and the purposes of transactions. We further observe that cross-chain bridges can be abused for malicious/aggressive purposes. Thus we design an automated detector and deploy it in-the-wild to flag misbehaviors from millions of cross-chain transactions. We identify hundreds of abnormal transactions related to exploits and arbitrages, etc. Our research underscores the importance of the cross-chain ecosystem, and offers an effective detector for pinpointing security threats.