3P3: Strong Flexible Privacy for Broadcasts
David Mödinger, Franz J. Hauck · 2020
Privacy concerns have reached the mainstream discourse in society and already had a significant impact on research and technology. Cryptocurrencies have adopted many transaction-level privacy mechanisms to provide privacy in the persisted blockchain. Unfortunately, these are insufficient as network-level attacks can also provide privacy-breaking insights into transactions and their origins. We proposed k-Dining Cryptographers and topological methods as a basis for a privacy-preserving broadcast protocol. In this work, we present 3P3, a three-phase privacy-preserving broadcast protocol. We transformed our approach to a stronger attacker model so that it provides strong base privacy against global attackers and malicious nodes and additional privacy against common attackers, e.g., botnets. Further, we provide mechanisms to transmit almost arbitrarily long messages, reduce overhead for zero-message rounds, a more extensive analysis, and simulation results of our enhanced protocol. Our simulations show the dissemination of a message to all nodes within 1000ms in 99.9% of instances. These results hold for all network sizes, including networks of up to 10,000 participants. Bandwidth estimates also show practical applicability with usual group sizes of 10 to 30 participants.