sAVSS: Scalable Asynchronous Verifiable Secret Sharing in BFT Protocols
This provides a scalable threshold trusted third party for applications like fair exchange, but it is incremental as it builds on existing VSS and BFT methods.
The paper tackled the problem of incorporating verifiable secret sharing (VSS) into Byzantine Fault Tolerance (BFT) protocols to add privacy to threshold guarantees, resulting in sAVSS, which enables writing secret-shared values in a key-value store with 30-50% throughput overhead and under 35 ms latencies.
This paper introduces a new way to incorporate verifiable secret sharing (VSS) schemes into Byzantine Fault Tolerance (BFT) protocols. This technique extends the threshold guarantee of classical Byzantine Fault Tolerant algorithms to include privacy as well. This provides applications with a powerful primitive: a threshold trusted third party, which simplifies many difficult problems such as a fair exchange. In order to incorporate VSS into BFT, we introduced sAVSS, a framework that transforms any VSS scheme into an asynchronous VSS scheme with constant overhead. By incorporating Kate et al.'s scheme into our framework, we obtain an asynchronous VSS that has constant overhead on each replica -- the first of its kind. We show that a key-value store built using BFT replication and sAVSS supports writing secret-shared values with about a 30% - 50% throughput overhead with less than 35 millisecond request latencies.