Secure and Usable Decentralized Storage

Abstract

Recent advancements in the Dynamic-Committee Proactive Secret Sharing (DPSS) protocol have introduced mechanisms for periodically refreshing shares and dynamically changing the parties holding them. DPSS is particularly relevant in modern applications such as blockchain systems, where committees are responsible for managing confidential information, and decentralised storage networks, where participant involvement is inherently dynamic. Despite its potential, DPSS faces significant challenges, particularly its high communication complexity of O(n3ℓ + λn3) when refreshing data, which becomes a bottleneck when dealing with large-scale data. In this work, we address these challenges by introducing the concept of Dynamic-Committee Proactive Information Dispersal (DPID), a novel approach that can be viewed as a relaxed version of DPSS without the confidentiality requirement. We provide a formal model for DPID and construct practical schemes that achieve substantially reduced communication complexity to O(ℓ + λn2). To bridge the gap between DPID and DPSS, we also present a general framework for compiling DPID into DPSS, preserving efficiency while reintroducing confidentiality. By integrating our DPID construction, we achieve the first DPSS scheme that is optimised for large-scale data, offering dramatically lower communication overhead. We validate the efficiency and scalability of our approach through comprehensive experiments, which demonstrate clear improvements over state-of-the-art methods. Our work not only advances the theoretical understanding of proactive information dispersal but also provides a practical foundation for deploying DPSS in real-world systems