Research On P2P Network And Consensus Algorithm Improvement For Scalable Blockchain

With the rapid expansion of blockchain technology’s applications,from digital currencies to smart contracts,digital identity,Web3.0,and the metaverse,blockchain has provided powerful support for the development of the digital economy.However,with the expansion of the scale of nodes in the network,issues such as inadequate scalability and low-performance efficiency have become increasingly prominent,becoming obstacles to the large-scale implementation of blockchain applications.Challenges in network scale,transmission efficiency,consensus algorithm performance,and message redundancy are all significant.This dissertation aims to focus on improving the performance of blockchain systems and expanding the scale of nodes while ensuring security.The main research contents are as follows:(1)To address the common issues of low transmission efficiency and limited scalability of P2P networks in blockchain,the Mesh-and-Spoke network model is proposed.This model utilizes the heterogeneity of nodes to partition the blockchain network into multiple levels,decoupling the communication overhead of consensus from the size of ordinary nodes.This approach can avoid the problems of low transmission efficiency and reduced consensus performance due to the expansion of node scale,improving the scalability of nodes and effective transmission rates.Furthermore,this model also implements isolation,redundancy,recombination,incentive,and punishment mechanisms to ensure system security and fairness.(2)To address the communication complexity and scalability limitations of PBFT,a new consensus algorithm called H-PBFT has been proposed.This algorithm partitions the consensus of the entire network into consensus of sub-layer ranges based on the Mesh-and-Spoke network,reducing the communication complexity of the consensus algorithm and improving its efficiency and fault tolerance.Moreover,as the size of the ordinary network expands,the communication efficiency and throughput of consensus remain relatively stable.Therefore,H-PBFT has good consensus efficiency and scalability,supporting efficient consensus and communication for large-scale nodes.(3)Aiming at the problems of large message redundancy and high time delay in the block propagation process of the Gossip data transmission algorithm,an improved block data propagation algorithm LS-Gossip is proposed.The algorithm reduces the number of redundant messages by marking the latest block status of neighboring nodes for on-demand transmission.At the same time,the algorithm is also optimized for the block propagation path,reducing the propagation time delay caused by unreliable nodes.In short,the LS-Gossip algorithm has significantly improved the efficiency of disseminating block data,greatly reducing the propagation convergence time and communication overhead,and can be applied to actual scenarios with larger network scales.