A privacy-preserving asynchronous Byzantine fault-tolerant consensus method and system based on Dumbo BFT.

Abstract

This invention discloses a privacy-preserving asynchronous Byzantine fault-tolerant consensus method and system based on Dumbo BFT, belonging to the field of blockchain and distributed system technology. First, this invention divides node states into public and private sub-states using an initial state decomposition algorithm, achieving privacy isolation at the data source. Second, it employs a local state difference communication mechanism, transmitting state difference values ​​instead of original data to prevent information leakage. Then, it introduces a dual verification system of threshold homomorphic encryption and zero-knowledge proofs to achieve ciphertext computation and legality verification during the transaction aggregation stage. Finally, it designs a dynamic adaptive sharding strategy, automatically adjusting the consensus group size based on network latency prediction. This invention achieves privacy protection for transaction content, node states, and communication processes in asynchronous Byzantine fault-tolerant consensus, effectively reducing the risk of privacy leakage. Furthermore, through dynamic sharding and hardware acceleration, it supports expansion to thousands of nodes, making it suitable for high-concurrency, privacy-sensitive scenarios such as finance and healthcare.

Images