Ethereum Throughput Bottleneck Analysis And Optimization Research

The emergence of Ethereum has greatly expanded the application scenarios of blockchain technology,so that blockchain technology is no longer limited to payment scenarios,and thus has become a representative of blockchain 2.0,which has received widespread attention.However,the current throughput of Ethereum is low,and a large number of transactions cannot be processed in time,resulting in transaction accumulation.Ethereum has gradually fallen into a dilemma that cannot meet current business needs and cannot support many decentralized application(DApp)in the future.At the same time,in the process of combining blockchain technology with many fields,the performance of the blockchain system is low and it cannot meet commercial applications.Low throughput has become the main obstacle to the promotion of blockchain technology.In response to the above two problems,this article takes Ethereum as the research object,and conducts throughput bottleneck research from its data layer and consensus layer,analyzes the factors that limit Ethereum throughput,explores Ethereum throughput bottlenecks,and studies Ethereum high throughput system architecture to solve the problem of low throughput of blockchain technology business applications.In response to the above two problems,this article takes Ethereum as the research object,and conducts throughput bottleneck research from its data layer and consensus layer,analyzes the factors that limit Ethereum throughput,explores Ethereum throughput bottlenecks,and studies Ethereum high throughput System architecture to solve the problem of low throughput of blockchain technology business applications.This paper studies the impact of the Ethereum data layer transaction processing flow,block capacity,and gas Limit adjustment mechanism on throughput.In the consensus layer,discuss the limits of the consensus algorithm and the difficulty of mining on throughput.Finally,after a targeted test,the bottleneck of Ethereum throughput was summarized,and a throughput performance optimization plan was proposed based on the test results.The main work and results of this article are as follows:(1)Through testing,we completed the research on Ethereum throughput bottleneck.First,the principle and source code of Ethereum transaction processing are analyzed.Secondly,an Ethereum test network was built to perform fine-grained tests on the performance of Ethereum transaction signature and verification,transaction broadcast,construction block,block broadcast and other stages.Then,at the consensus layer,the impact of mining difficulty and consensus algorithm selection on system throughput is studied.Finally,during the test,some performance problems such as inefficient Ethereum algorithm and slow growth of gas Limit were found.(2)Corresponding optimization schemes are proposed for the discovered performance problems.Experiments show that the optimization scheme can improve the throughput of Ethereum by orders of magnitude.Specifically,this paper proposes a single-chain optimization scheme and a multi-chain parallelization scheme.After testing and verification,the single-chain optimization scheme can steadily increase the Ethereum throughput of 15TPS(Transactions Per Second)to about 650 TPS.In addition,the isomorphic multi-chain parallelization scheme can linearly improve the performance of Ethereum under the experimental environment.