Anonymous Multi-Hop Locks for Blockchain Scalability and Interoperability
Anonymous Multi-Hop Locks for Blockchain Scalability and Interoperability
复制标题
DOI：
10.14722/ndss.2019.23330
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Giulio Malavolta;Pedro A. Moreno-Sánchez;Clara Schneidewind;Aniket Kate;Matteo Maffei
中科院分区：
文献类型：
--
作者：
Giulio Malavolta;Pedro A. Moreno-Sánchez;Clara Schneidewind;Aniket Kate;Matteo Maffei
-加密货币使用的巨大增长暴露了无需许可的区块链技术固有的可扩展性问题。支付通道网络（PCN）已成为缓解可扩展性问题的最广泛部署的解决方案，允许两个用户之间的大部分支付在链外进行。不幸的是，正如文献报道和本文进一步证明的那样，当前的 PCN 没有提供有意义的安全和隐私保证 [30]、[40]。在这项工作中，我们研究和设计安全且保护隐私的 PCN。我们首先对现有 PCN 进行安全分析，报告一种适用于所有主要 PCN（包括闪电网络）的新攻击，并允许攻击者通过同一支付路径从诚实的中介机构窃取费用。然后，我们正式定义匿名多跳锁（AMHL），这是一种新颖的加密原语，可作为安全和隐私保护 PCN 设计的基石。我们提出了几种可证明安全的加密实例，使 AMHL 与绝大多数加密货币兼容。特别是，我们证明（线性）同态单向函数足以为支持脚本语言（例如以太坊）的 PCN 构建 AMHL。我们还提出了一种基于 ECDSA 签名的结构，不需要脚本，从而解决了该领域的一个突出的开放问题。
—Tremendous growth in cryptocurrency usage is exposing the inherent scalability issues with permis-sionless blockchain technology. Payment-channel networks (PCNs) have emerged as the most widely deployed solution to mitigate the scalability issues, allowing the bulk of payments between two users to be carried out off-chain. Unfortunately, as reported in the literature and further demonstrated in this paper, current PCNs do not provide meaningful security and privacy guarantees [30], [40]. In this work, we study and design secure and privacy-preserving PCNs. We start with a security analysis of existing PCNs, reporting a new attack that applies to all major PCNs, including the Lightning Network, and allows an attacker to steal the fees from honest intermediaries in the same payment path. We then formally define anonymous multi-hop locks (AMHLs), a novel cryptographic primitive that serves as a cornerstone for the design of secure and privacy-preserving PCNs. We present several provably secure cryptographic instantiations that make AMHLs compatible with the vast majority of cryptocurrencies. In particular, we show that (linear) homomorphic one-way functions suffice to construct AMHLs for PCNs supporting a script language (e.g., Ethereum). We also propose a construction based on ECDSA signatures that does not require scripts , thus solving a prominent open problem in the field.