Privacy Enhancing Technologies for Censorship-resistant and Secure Internet Communications

在过去的几年中，互联网已经看到由网站、消息传递应用和其他形式的在线通信提供的安全和隐私技术的显著增加的部署。 然而，这些目前部署在隐私保护通信领域的技术只能解决部分问题。 大多数方法都专注于保护消息的内容，但会泄露元数据（谁与谁通信，何时何地）。 这些元数据可以重建一个人的社交网络、位置和活动，其本身可以揭示敏感信息，例如属于少数群体或边缘化群体。 此外，如果通信本身受到审查，信息根本无法通过，保护通信内容和元数据的效用有限;国家规模的互联网审查在世界各地越来越普遍，严重损害了言论自由等人权。互联网的最初设计没有考虑到政府操纵和审查通信，因此没有提供适当的机制来保证可用性：人们可能希望使用的互联网服务实际上是可以获得的。 同样，目前以数字方式共享的大量高度敏感的个人信息也没有得到充分保护。 对手，从轻微的罪犯到政府审查员，都能够观察和操纵互联网通信。 这种监视和操纵既影响到个人，例如由于个人信息泄露，也影响到社会，例如由于全国范围的审查。 因此，必须开发技术来抵制监视和审查。不幸的是，手动设计和评估隐私增强技术（PET）以实现安全和抗审查通信是繁琐的，并且众所周知难以完美执行。此外，单个设计缺陷可能会给用户带来灾难性的现实后果。通过实现自动化的方案构建和验证，我们将确保未来对PET的研究将更加有效，并降低错误风险。我们研究的长期目标是开发原创和创新的PET，克服现有最先进方法的局限性，并提供更广泛的保护，免受监视和审查。根据上述讨论，我们确定了三个短期目标：i）保护通信元数据，ii）自动化协议构建，iii）抗审查的互联网访问。 我们用于安全和抗审查通信的PET解决了安全性的所有三个传统方面：机密性、完整性和可用性。除了推进科学知识，我们还专注于部署我们的系统，使世界各地的用户受益。

In the last few years, the Internet has seen significantly increased deployment of security and privacy technologies offered by websites, messaging applications, and other forms of online communication. However, these currently deployed technologies in the area of privacy-preserving communication only solve part of the problem. Most approaches focus on protecting the contents of a message but reveal the metadata (who communicated with whom, when, and where). Such metadata allows the reconstruction of a person's social network, location, and activities, and by itself can reveal sensitive information such as being part of a minority or marginalized group. Furthermore, protecting communication contents and metadata is of limited utility if the communication itself is censored and the messages are not getting through at all; country-scale Internet censorship is increasingly common around the world, drastically undermining human rights such as freedom of speech.******The original design of the Internet did not account for governmental parties manipulating and censoring communication and hence does not provide adequate mechanisms to guarantee availability: that the Internet services one might wish to use are in fact reachable. Similarly, the abundance of sensitive highly personal information that is currently shared digitally is not adequately protected. Adversaries, ranging from petty criminals to governmental censors, are able to observe and manipulate Internet communication. Such surveillance and manipulation affects both individuals, for instance due to leaked personal information, as well as societies, for instance due to country-wide censorship. It is therefore essential to develop technologies to resist surveillance and censorship.******Unfortunately, manually designing and evaluating privacy-enhancing technologies (PETs) for secure and censorship-resistant communication is cumbersome and notoriously difficult to perform perfectly. Moreover, a single design flaw can lead to catastrophic real-world consequences for users. By enabling automated protocol construction and verification, we will ensure future research on PETs will be more efficient and have lower risk for errors.******The long-term goal of our research is to develop original and innovative PETs that overcome the limitations of the existing state-of-the-art approaches and provide more extensive protection from surveillance and censorship. In light of the above discussion, we identified three short-term goals: i) protecting communication metadata, ii) automated protocol construction, and iii) censorship-resistant Internet access. Our PETs for secure and censorship-resistant communication address all three traditional aspects of security: confidentiality, integrity, and availability. In addition to advancing scientific knowledge, we focus on deploying our systems to benefit users all over the world.