MAD-HTLC：因为攻击HTLC非常便宜
MAD-HTLC: Because HTLC is Crazy-Cheap to Attack
摘要：
智能合约和交易允许用户在比特币和以太坊等加密货币区块链上实现精心设计的结构。许多这样的构造，包括操作支付渠道和原子交换，都使用了一种称为哈希时间锁定合同（HTLC）的构建块。在这项工作中，我们从HTLC中提取一个规范（HTLC-Spec），并提出了一个实现称为相互保证破坏哈希时间锁定合同（MAD-HTLC）。MAD-HTLC采用了一种新的方法，利用现有的区块链运营商（称为矿工）作为设计的一部分。如果用户行为不端，MAD-HTLC会激励矿工没收她的所有资金。我们证明了MAD-HTLC的安全性使用UC框架和博弈论分析。我们展示了MAD-HTLC的有效性，并通过在比特币和以太坊的运营区块链上实例化它来分析其开销。值得注意的是，目前的矿工软件在优化收入方面做得很少，因为优势相对较小。然而，随着需求的增长和其他收入组成部分的萎缩，矿工们更有动力充分优化他们的资金吸收。通过修补标准比特币客户端，我们证明了这种优化很容易实现，使矿工成为MAD-HTLC的自然执行者。最后，我们扩展了以前的结果HTLC的脆弱性贿赂攻击。攻击者可以通过提供高额交易费来激励矿工更喜欢她的交易。我们证明了这种攻击可以很容易地通过修补比特币客户端来实现，并使用博弈论工具在理性矿工存在的情况下定性地收紧这种贿赂攻击的已知成本范围。我们发现了比特币和以太坊主网络上发生的贿赂机会，其中几美元的贿赂可以产生数万美元的奖励（例如，超过2.5万美元的话2美元）。
英文摘要：
Smart Contracts and transactions allow users to implement elaborate constructions on cryptocurrency blockchains like Bitcoin and Ethereum. Many of these constructions, including operational payment channels and atomic swaps, use a building block called Hashed Time-Locked Contract (HTLC). In this work, we distill from HTLC a specification (HTLC-Spec), and present an implementation called Mutual-Assured-Destruction Hashed Time-Locked Contract (MAD-HTLC). MAD-HTLC employs a novel approach of utilizing the existing blockchain operators, called miners, as part of the design. If a user misbehaves, MAD-HTLC incentivizes the miners to confiscate all her funds. We prove MAD-HTLC's security using the UC framework and game-theoretic analysis. We demonstrate MAD-HTLC's efficacy and analyze its overhead by instantiating it on Bitcoin's and Ethereum's operational blockchains. Notably, current miner software makes only little effort to optimize revenue, since the advantage is relatively small. However, as the demand grows and other revenue components shrink, miners are more motivated to fully optimize their fund intake. By patching the standard Bitcoin client, we demonstrate such optimization is easy to implement, making the miners natural enforcers of MAD-HTLC. Finally, we extend previous results regarding HTLC vulnerability to bribery attacks. An attacker can incentivize miners to prefer her transactions by offering high transaction fees. We demonstrate this attack can be easily implemented by patching the Bitcoin client, and use game-theoretic tools to qualitatively tighten the known cost bound of such bribery attacks in presence of rational miners. We identify bribe opportunities occurring on the Bitcoin and Ethereum main networks where a few dollars bribe could yield tens of thousands of dollars in reward (e.g., \$2 for over \$25K).
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
