Challenges of Anonymous Communication: Bridging Gaps between Theory and Practice

Abstract

Anonymous communication is an important building block for privacy enhancing technologies and in addition to this, the deployed anonymity networks have become important tools to protect the users’ privacy in the Internet where hundreds of thousands of users rely on them. During the last decades, researchers have investigated the security aspects of anonymous communication networks. One important research methodology is the investigation of attacks which helps to understand the security properties of the analyzed systems. Most of the research has been focused on the Tor network, which is the most popular anonymity network. Tor is however not the only deployed anonymity network. Several others have been deployed during the years, e.g., AN.ON, I2P, and MixMaster. Each of them is based on different concepts which entail different security properties as well as attack vectors. Therefore, it is also important to investigate the security aspects of the other networks. To this end, the main part of this work deals with the security aspects of AN.ON and attacks against AN.ON. Moreover, improvements for AN.ON are discussed and analyzed. Besides the investigation of the security of AN.ON, two further aspects are researched in this thesis. The first aspect, basically being the starting point, is the question of how to reduce the complexity of anonymity networks. The last aspect of this thesis deals with future challenges of anonymous communication networks. Here, it is investigated how the combination of proposed extensions and enhancements respectively impact the anonymity properties of the Tor network.

Overall, the work can be divided in three parts. All of them address challenges in the area of anonymous communication. The first part contributes to the research in this field by proposing an easy-to-develop and easy-to-understand anonymity network. It has recently been used to investigate the impact of different countermeasures against website fingerprinting attacks. The results of the main part have contributed to a significant improvement of the level of security that AN.ON can provide for its roughly 100 000 users. The last part of this work contributes to the discussion of how to overcome a potential scalability problem of anonymity networks, as it highlights some of the ramifications caused by one proposal, i.e., splitting the Tor network