Performance Analysis And Optimization Of Blockchain System For The Internet Of Things

With the vigorous development of the Internet of Things,its problems in data security and device reliability are gradually exposed.Blockchain technology can effectively solve these problems,but huge resource consumption and limited data processing performance limit its deployment in Io T scenarios.In order to solve the above problems,this paper carries out multi-dimensional analysis on the performance of different blockchain systems,and conducts targeted research on consensus mechanism optimization and node storage optimization.The main contributions of this paper are summarized as follows:(1)In order to deeply study the performance of blockchain systems in Io T scenarios,a comprehensive set of performance metrics and a blockchain benchmarking tool are proposed to analyze the performance of Ethereum,Fabric v1.4 and IOTA from the aspects of availability,resource utilization and stability.Experiments show that the proposed metrics can better measure the performance of blockchain systems,and the benchmarking tool also has good real-time performance,compatibility and scalability.(2)In order to solve the problems of low transaction throughput and high confirmation latency of blockchain,a Po W consensus mechanism optimization algorithm based on reputation incentive is proposed.Firstly,the algorithm basically realizes the identity judgment of miner nodes based on reputation interval and reputation fluctuation.Then,the difficulty value of hash calculation is dynamically adjusted according to the reputation value of the miner node.Finally,two different working strategies are defined to incentivize miner nodes to engage in honest mining.Experiments prove that the improved Po W consensus mechanism is feasible,effectively improving the performance of the blockchain system.(3)In order to solve the problem of insufficient storage capacity of blockchain nodes,a storage optimization strategy of blockchain nodes based on grouping-sharding is proposed.On the basis of node grouping,this strategy selects a full node and lightweight nodes based on the node function allocation strategy,proposes a block sharding storage strategy for lightweight nodes,and introduces virtual nodes to allocate the storage data of nodes.At the same time,a full node replacement strategy is designed to deal with the sudden offline of the full node.In addition,it also discusses the changes of nodes when nodes join and exit,as well as the verification process of node transactions.Experiments show that this strategy significantly reduces the data storage capacity of nodes and effectively saves the storage space of nodes.