Creating a Privacy-Respecting Anti-DDoS Mechanism for Rendezvous Channels
Pedro Filipe Cavaleiro · New University of Lisbon's Repository (New University of Lisbon) · 2026
With the increase of global attempts to block and censor the access to data that might be harmful to the political environments of certain countries, the desire to bypass said censorship rises equally. With that in mind, one of the main tools used to circumvent these virtually aggressive environments is the Tor Browser. Although Tor allows most users to successfully escape such environments, there are still plenty of malicious users whose intent is to overthrow Tor’s and other specialized software’s way of countering the censorship by launching Distributed Denial of Service (DDoS) attacks. Such attacks restrict a significant part of the user base as wished by those political and governmental entities (as well as some malicious private entities). In order to solve the latter, we propose a new bridge distribution architecture in which the amount of damage caused by said malicious users is vehemently reduced through the usage of cryptographical problems in RSA fashion, where a user has to solve a given computational problem involving random numbers, cryptographic keys and a strong enough puzzles algorithm, to access the bridge distribution nodes. We will also leverage the usage of trust levels to create a trustworthy bridge distribution node network as well as strong cypher suites to maintain the privacy of each individual and prevent any data leaks. In this thesis we will analyze the details of the literature, design and implementation of this system whose main objective is to allow the circumvention of censorship while simultaneously defending it from Distributed Denial of Service attacks from the censoring entities all the while protecting the privacy of all users. With that in mind, this solution aims at heavily weakening the prime censoring entities power by making it computationally unfeasible to enumerate and block many bridges in a short time window.