使用zk-SNARKs的匿名对等内容共享系统的垃圾邮件预防
Spam Prevention Using zk-SNARKs for Anonymous Peer-to-Peer Content Sharing Systems
摘要：
去中心化的未经许可的对等网络在允许任意参与者将内容提交到公共索引或注册表时，本质上容易受到垃圾邮件的攻击;由于缺乏可以充当看门人的中央仲裁者，因此很难防止这种情况。出于这个原因，即使在分散的网络中（例如Bittorrent与DHT，IPFS），新内容的索引通常也留给集中式服务，如torrent网站。去中心化的垃圾邮件预防方法（例如Web of Trust）已经存在[1][2]，但它们要求提交者采用假名身份并随着时间的推移建立信任。在本文中，我们提出了一种方法的垃圾邮件预防工作的假设下，参与者是完全匿名的，不希望他们的不同的意见书被链接到彼此。我们所说的垃圾邮件并不是指未经请求的广告;而是指在短时间内添加大量内容，使网络容量饱和，并对对等方造成拒绝服务的做法。我们的解决方案的目的是防止用户饱和系统，可以被描述为速率限制。该系统应能抵制审查：提交的内容不应因内容本身而被排除，用户也不应因其提交的内容而被排除。我们首先讨论了一个基于单一的、集中的、抗审查和匿名的限速器的解决方案，然后我们将其扩展到一个完全分散的、基于区块链的系统，并提出了使其经济和可扩展的方法。
英文摘要：
Decentralized unpermissioned peer-to-peer networks are inherently vulnerable to spam when they allow arbitrary participants to submit content to a common public index or registry; preventing this is difficult due to the absence of a central arbitrator who can act as a gate-keeper. For this reason indexing of new content even in otherwise decentralized networks (e.g. Bittorrent with DHT, IPFS) has generally been left to centralized services such as torrent sites. Decentralized methods for spam prevention, such as Web of Trust, already exist[1][2] but they require submitters to assume pseudonymous identities and establish trust over time. In this paper we present a method of spam prevention that works under the assumption that the participants are fully anonymous and do not want different submissions of theirs to be linked to each other. By spam we do not specifically mean unsolicited advertising; rather it is the practice of adding a large amount of content in a short time, saturating the capacity of the network, and causing denial of service to peers. The purpose of our solution is to prevent users from saturating the system, and can be described as rate-limiting. The system should be censorship resistant: it should not be possible for submissions to be excluded because of the content itself, or for users to be excluded based on what they submit. We first discuss a solution based on a single, centralized rate-limiter that is censorship resistant and anonymous, then we extend this to a fully decentralized blockchain-based system and present methods to make it economical and scalable.
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。
