Design And Research Of Consensus Algorithm To Improve Transaction Processing Capability

Data analysis technology stimulates the potential value of data and puts forward higher requirements for data security.As a decentralized distributed database technology that does not rely on third parties,blockchain technology can replace trust with technology and ensure the security,transparency and traceability of data on the chain.It is considered to be an important force for technological innovation and industrial transformation.However,performance problems such as weak transaction processing capability and serious storage space consumption restrict the application of blockchain technology to massive data processing fields such as the Internet of Things.In this regard,domestic and foreign scholars mostly focus on the improvement of the block transmission process,and rarely improve the system performance by processing transactions in advance.In response to the above problems,this thesis designs a consensus algorithm that improves transaction processing capabilities from the perspective of processing transactions in advance,and designs a chain structure that reduces system storage space overhead and optimizes query efficiency,so as to build a high-performance blockchain system.The specific content as follows:(1)In order to speed up block generation and reduce communication overhead,a dynamic multi-block double-round consensus algorithm is designed based on the nodes processing transactions in advance.The algorithm can use the node consensus waiting time to process transactions in advance,realizing the batch generation of blocks in each round of consensus,and effectively improve the block generation speed;adopting a double-round consensus strategy from a single cluster to the whole system to reduce the communication complexity of the algorithm;designing incentive mechanism to urge nodes to work honestly to maintain the stable operation of the system.(2)In order to reduce the system storage space overhead,a dual-chain block storage model is designed.The information chain that records the key results of the consensus is saved completely by all nodes,and the value chain that records the complete block information is grouped saved.Compared with the traditional blockchain system in which nodes store a complete chain structure,this dual-chain model reduces the storage overhead of the system.In addition,the transaction index structure on the information chain is constructed to optimize the data query efficiency on the chain,so as to avoid the problem of reduced query efficiency due to optimization of storage space.(3)Building a high-performance blockchain system based on the consensus algorithm and storage structure proposed in this thesis,and testing the transaction processing capability and data query performance of the system.Simulation experiments show that in the transaction processing capability test,compared with the traditional consensus algorithm,the algorithm proposed in this thesis can effectively improve the system transaction throughput,can improve the system performance degradation caused by network delay,and can also improve the system performance in the presence of network delay to guarantee considerable transaction processing capability;in the data query performance test,compared with the traditional linear structure,the structure proposed in this thesis can achieve efficient and stable on-chain data query.