Parallelization Byzantine fault tolerance method applied to block chain consensus mechanism

Abstract

The invention discloses a parallelization Byzantine fault tolerance method applied to a blockchain consensus mechanism, and the consensus algorithm comprises a plurality of consensus partitions running in parallel at a bottom layer and independent verification partitions running at an upper layer, and the consensus partitions and the independent verification partitions jointly form a layered parallelization consensus verification system. The underlying consensus partitions respectively, independently and parallelly receive task requests (a pre-preparation stage, namely a Pre-prepare stage); and then, a message request service (a preparation stage, namely a Prepart stage) is executed, and meanwhile, signature of an execution result is achieved by adopting an aggregation signature algorithmin an execution process so as to achieve the purposes of reducing message complexity and improving system throughput. Besides, according to the algorithm design, the Commit stage message of the current consensus process and the Prepare stage message of the next consensus process are combined and sent, so that the purposes of compressing the communication rounds and reducing the communication overhead are achieved, the throughput of the system is further improved, and the average time delay is reduced.

Description

technical field [0001] The invention relates to the technical field of block chains, in particular to a parallel Byzantine fault-tolerant method for block chain consensus mechanisms. Background technique [0002] Consensus algorithm is one of the core technologies of blockchain. In the classic PBFT consensus algorithm, all consensus nodes (that is, a single consensus partition) participate in the consensus execution process of a request, thereby forming a final consensus on the entire node set. The execution of the PBFT algorithm requires message communication through the full communication of all consensus nodes, so its message complexity is O(n 2 ). Subsequently, new consensus algorithms based on Byzantine fault tolerance continued to emerge, which greatly improved consensus efficiency and message complexity. Among them, the Zyzzvya protocol, as a representative of the consensus algorithm using speculative thinking, realizes the direct interaction between the consensus ...

AI Technical Summary

Problems solved by technology

[0004] Combining the above two typical improved consensus algorithms based on Byzantine fault tolerance, the consensus efficiency has been significantly improved. However, when error nodes occur frequently, both algorithms will fall into repeated fault-tolerant processing, resulting in a sharp drop in consensus performance

In addition, when the scale of consensus nodes increases linearly, the consensus efficiency also shows a linear downward trend

Therefore, the current improved consensus algorithm based on Byzantine fault tolerance (for example, Zyzzyva and Query / Update protocol), will lead to a significant decrease in consensus efficiency when frequent error nodes or expansion of the consensus node set

Images