Research And Implementation Of Pipeline-Based Blockchain Consensus Algorithm

Blockchain,as a novel distributed computing paradigm,has garnered significant academic and industrial attention due to its technical advantages,such as immutability,self-proving,and full traceability.It can construct trust in untrusted environments at a low cost,and is therefore widely applied in vertical domains such as supply chain management,certification,traceability,the Internet of Things,livelihood services,and social governance,becoming a new computing paradigm and collaborative mode.The consensus algorithm is the core component of the blockchain system,which determines the block generation rate and accuracy and plays a crucial role in ensuring the consistency and security of the blockchain system.Within the pipeline consensus mechanism,nodes are grouped and categorized based on different functionalities,which reduces the time overhead of the consensus mechanism by enabling different nodes to vote at different points in time.Accordingly,this paper,in conjunction with blockchain and pipeline execution architecture,investigates the following primary research topics:(1)We propose a pipeline-based chain structure Byzantine Fault Tolerance(PCS-BFT)consensus algorithm to address the issue of block disorder in high-concurrency parallel pipeline consensus algorithms.PCS-BFT achieves a high system throughput by parallel processing different transaction block requests using the pipeline consensus architecture.The committee selection process in each round employs a random seed,thus reducing the vulnerability of Byzantine nodes.Concurrent block disorder is addressed by synchronizing block dependency order between nodes,and all validated blocks in the period ultimately form a chain structure to achieve deterministic consistency.We conduct performance tests on the proposed PCS-BFT consensus algorithm through experiments.Compared with the Hot Stuff consensus algorithm,when the number of nodes exceeds eight,PCS-BFT achieves a 20%higher throughput than Hot Stuff.On the basis of ensuring Byzantine fault tolerance safety,PCS-BFT can effectively improve system performance and reduce communication overhead while increasing block generation efficiency.(2)Building on the pipeline-based chain structure,we propose a graph structure Byzantine Fault Tolerance(PGS-BFT)consensus algorithm for pipeline execution.PGS-BFT adopts a leaderless pipeline execution architecture from the perspective of sorting,randomly selecting21)+ 1 nodes to form a committee each round to accelerate block generation efficiency.By synchronizing block dependency relationships between nodes,all validated blocks in the period ultimately form a graph structure,combined with the DAG graph structure to optimize block ordering and storage structure,enabling nodes to achieve final consistency.We conduct algorithm theoretical analysis and experimental validation on the proposed algorithm.The experimental results demonstrate that the consensus algorithm can guarantee consistency under a fault tolerance model of 31)+ 1,and PGS-BFT is 15% more efficient than the Hot Stuff consensus algorithm when the number of nodes exceeds eight.