Lune

USENIX Security2025顶会

Parallelizing Universal Atomic Swaps for Multi-Chain Cryptocurrency Exchanges

Danlei Xiao, Chuan Zhang, Haotian Deng, Jinwen Liang, Licheng Wang, Liehuang Zhu

出版方
2025年份

摘要

Universal atomic swap is an emerging technique for secure cryptocurrency exchanges across diverse blockchains, eliminating the need for custom scripting language support from blockchains. While existing schemes primarily focus on exchanges among two users, extending these to multiple users across multiple blockchains incurs significant overheads due to the need for performing multiple two-party swaps serially. An intuitive insight is to parallelize the universal processes, but this idea still faces two technical challenges: (i) avoid asset theft during parallel asset locking; (ii) ensure atomicity by preventing partial execution of transactions with a uniform refund time used to avoid asset deadlock in parallel. In this paper, we present ParaSwap, the first framework to parallelize universal atomic swaps for cryptocurrency exchanges among multiple users across multiple blockchains. We replace the serial multiple two-party swaps with a concurrent mechanism, where each participant concurrently locks and withdraws coins, achieving parallel execution. To prevent asset theft, the necessary witness for swaps is collaboratively determined by all participants. Then we introduce a novel re-lock approach to ensure atomicity with a uniform refund time, allowing participants to re-lock their assets to new addresses when the remaining time is insufficient to complete their withdrawal. Notably, ParaSwap employs adaptor signatures and verifiable timed discrete logarithm (VTD) technology, relying only on the bare minimum ability of blockchain to verify transaction signatures. We implement ParaSwap on four public blockchain test networks: Bitcoin, Ethereum, Avalanche, and Binance Smart Chain. Our evaluation demonstrates that ParaSwap reduces the exchange time complexity from O(n) to O(1), where n is the number of participants, and lowers gas costs by 26.2x to 46.8x, compared to existing methods.

问问这篇 Paper

智能体会读完全文。

Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。

可以从这些问题问起

智能体调用

Luneget_paper_fulltext

在 Lune 里问

免费开始,无需绑卡

它引用的顶会 Paper11

相关 Paper

黄昏的海面,两侧是细线勾勒的悬崖