Scalable Bias-Resistant Distributed Randomness

Explore a conference talk that delves into the challenges and solutions for generating bias-resistant public randomness in large-scale distributed systems. Learn about two innovative protocols, RandHound and RandHerd, designed to provide publicly-verifiable, unpredictable, and unbiasable randomness against Byzantine adversaries. Discover how RandHound leverages an untrusted client to divide randomness servers into groups for scalability, while RandHerd implements an efficient, decentralized randomness beacon. Examine the structural similarities between RandHerd and BFT protocols, and understand how RandHound is utilized in a one-time setup to create unbiased random secret-sharing groups. Gain insights into the performance and failure probability of these protocols across hundreds of participants, and learn how proper parameter selection, such as group size and secret-sharing threshold, impacts their effectiveness. Understand the practical applications of these protocols in various distributed systems requiring secure and unbiased randomness generation.

Scalable Bias-Resistant Distributed Randomness