CN115102866B - A consensus algorithm-oriented block chain TPS estimation method and its verification method - Google Patents

Abstract

The invention discloses a consensus algorithm-oriented block chain TPS estimation method and a verification method thereof. The method analyzes the block chain system transaction process based on a Byzantine fault-tolerant algorithm, and establishes a mathematical model between the block chain TPS and node communication and computing capabilities to perform TPS estimation; the accuracy verification method uses block chain networks with different structures to perform TPS tests, and compares the actual test results with the prediction results of the mathematical model to verify the accuracy of the block chain TPS mathematical model. The present invention realizes the prediction of the performance of the blockchain network before the deployment of the blockchain system, and the prediction error does not exceed 10%. The mathematical model is convenient for the hardware selection of the blockchain nodes and the cost estimation of the network construction when the blockchain system is actually deployed, and the feasibility of actually deploying the blockchain system under a specific network structure can be verified by estimating the performance of the blockchain.

Description

A consensus algorithm-oriented block chain TPS estimation method and its verification method

technical field

The invention belongs to the field of communication networks, and in particular relates to a consensus algorithm-oriented block chain TPS estimation method and a verification method thereof.

Background technique

After years of development, IoT applications have made great strides. However, the large-scale development of the Internet of Things still faces some key challenges, especially how to balance construction/operating costs and service capabilities/efficiency issues. Blockchain technology helps to solve the problems of cross-industry, departmental, and cross-network service system resource sharing, data security transmission, and credit system construction of the Internet of Things. At the same time, large-scale Internet user groups are not only users of various application services, but also contributors to cyberspace big data and application services, which constitute the group intelligence resources that support a large number of successful applications.

Researchers from the fields of artificial intelligence, big data, computer collaborative work, and human-computer interaction have conducted research on this from different academic perspectives, and proposed research directions such as crowdsourcing, human computing, swarm intelligence, and social computing. At the same time, the industry has also launched a large number of platforms and applications that support data collaboration. Data collaborative computing provides great convenience for the development of new artificial intelligence technologies such as big data and statistical machine learning, and its core lies in the effective management and collaborative utilization of group intelligence resources in an open network environment to maximize group intelligence. If the alliance chain data collaboration framework is built based on technologies such as smart contracts and secure multi-party computing, blockchain technology can be used to reliably and safely perform data interactions, and ensure that all records are transparent and traceable.

However, with the rapid development of smart cities, sharing economy and other businesses, the unknown of the actual deployment performance of the blockchain has become one of the reasons why the blockchain system is not widely used. Therefore, there is an urgent need for a method to predict the performance of the blockchain system before the actual deployment of the blockchain system.

Contents of the invention

In order to solve the technical problems mentioned above in the background technology, the present invention proposes a consensus algorithm-oriented blockchain TPS estimation method and a verification method thereof.

In order to realize above-mentioned technical purpose, technical scheme of the present invention is:

A block chain TPS estimation method oriented to consensus algorithm, comprising the following steps:

S1, according to the transaction format of the blockchain and the smart contract used by the test TPS, calculate the actual transaction size S _tx sent by the blockchain system, using the same test transaction content and format;

S2, according to the actual deployed network nodes, obtain the time T _txbuild of each transaction generated at the client, and calculate the time T _txsend of each transaction generated by the client and sent to the blockchain node connected to the client;

S3, get the signature verification time T _txcheck of the transaction, the signature verification time of each transaction is the same, and the signature verification time is only considered once in the total transaction processing time;

S4, obtain the packaging time T _blockpk of the block, the size of the block header S _blockhead and the number n of transactions contained in each block, and calculate the total time T _blocksend of the transaction packaging block and sending it to other blockchain nodes;

S5, obtain the block verification time T _blockcheck and the execution time T _tx of each transaction, and calculate the time T _blockdone of each block verification and execution on the node;

S6, obtain the time T _blocksave of the transaction order;

S7, according to T _txbuild , T _txsend , T _blockpk , T _blocksend , T _blockcheck , T _blockdone , T _blocksave obtained in steps S2-S6, calculate the total time T _tot of completing a block; divide T _tot into communication time T _comm and calculation time T _comp ; obtain the average time T _txcomp of each transaction by performing multiple single-node transaction processing;

S8. Establish a blockchain TPS mathematical model, and use the communication capability of the node as an independent variable to import the model to draw a TPS estimation curve for the blockchain system.

Preferably, the step S2 specifically refers to: record the maximum uplink and downlink speed limit of each blockchain node as C, then the time for each transaction to be transmitted between two blockchain nodes is Get the time T _txbuild of each transaction generated at the blockchain node, then the time for each transaction generated by the client and sent to the blockchain node is />

Preferably, in the step S4, the total time for the transaction to package the block and send it to other blockchain nodes

The calculation formula of T _blocksend is as follows:

Where C is the maximum uplink and downlink rate limit of each blockchain node.

Preferably, in the step S5, the time T _blockdone of each block verification and execution on the node is counted

The calculation formula is as follows:

T _blockdone = T _blockcheck + nT _tx .

Preferably, the calculation formula of the total time T _tot for completing a block in the step S7 is as follows:

T _tot ＝n(T _txbuild +T _txsend +T _txcheck )+T _blockpk +T _blocksend +T _blockdone +T _consensus +T _blocksave

Among them, T _consensus refers to the consensus time of the block, which is ignored when calculating the communication time;

The formula for calculating the communication time T _comm is as follows:

In the formula, N is the number of blockchain nodes, and C is the maximum uplink and downlink rate limit of each blockchain node;

The calculation formula of the calculation time T _comp is as follows:

T _comp =n(T _txbuild +T _txcheck )+T _blockpk +T _blockcheck +nT _tx +T _blocksave .

Preferably, the calculation formula of the block chain TPS mathematical model in the step S8 is expressed as follows:

In the formula, N is the number of blockchain nodes, C is the maximum uplink and downlink speed limit of each blockchain node, and S _tx is the size of each transaction.

A method for verifying the accuracy of a consensus algorithm-oriented block chain TPS estimation method, comprising the steps of:

Step 1, build a blockchain network with the same network structure as that used in the blockchain TPS mathematical model analysis, and deploy and test smart contracts on the blockchain network;

Step 2, deploy the Caliper blockchain performance testing software on the Linux host outside the blockchain network;

Step 3, install system bandwidth limiting software on each blockchain node, modify the maximum uplink and downlink bandwidth of each node, and verify the error between the predicted results of the blockchain TPS mathematical model and the actual TPS at different rates;

Step 4: After modifying the maximum uplink and downlink bandwidth of the node, use the Caliper blockchain performance testing software to initiate multiple rounds of TPS tests on the deployed blockchain network through the Linux host outside the blockchain network, and record the actual TPS results of the test, and take the average of multiple sets of data at the same rate; Step 5, draw a curve for the actual results, compare it with the TPS theoretical value predicted by the blockchain TPS mathematical model, and calculate the errors predicted by the mathematical model under different node communication capabilities and the actual test to judge the accuracy of the blockchain TPS estimation method .

Preferably, in the first step, a block chain network with the same structure as the block chain TPS mathematical model is built, specifically including the following steps:

S101, provide multiple physical machines, and multiple physical machines access each other through a router, configure the static IP address of the physical machine, and use the Ping command in the command line to detect the communication between each blockchain node;

S102, configuring the environment required for the operation of the blockchain node on the physical machine;

S103, downloading the chain building script, modifying the ipconf configuration file according to the actual situation of the blockchain network, and correspondingly configuring the ip address of the blockchain node and the affiliated institution in the ipconf configuration file;

S104, execute the chain building script, automatically generate the node certificates required to build the blockchain network according to the configuration in the ipconf configuration file, and send the generated certificates to each blockchain node through SCP;

S105, writing a script, accessing the blockchain nodes one by one, respectively executing the blockchain node startup program on each Linux host in the blockchain network, the startup program will run the docker container under the blockchain node, and map the preset docker container port to the physical machine port to realize communication with other docker containers;

S106, prohibiting mutual access between two nodes through the TC command, when the prohibited node obtains the transaction or block sent by the other party, it needs to wait for the broadcast after receiving the transaction from other nodes in the blockchain network;

S107, after all the blockchain nodes are started, select any blockchain node, execute the view command, and observe whether all the nodes in the blockchain network can discover each other directly or in multiple ways.

Preferably, modifying the maximum uplink and downlink rate of each node in the step 3 specifically includes the following steps: S301, downloading the wondershaper bandwidth limiting software on each physical machine;

S302, writing the automation script SpeedControl.sh, using the uplink and downlink bandwidths to be limited as the two input values of the script respectively;

S303, installing iperf3 bandwidth testing software on all blockchain nodes;

S304, execute SpeedControl.sh, and limit the node bandwidth to a set value;

S305, start the iperf3 bandwidth testing software, and test the communication bandwidth between any two nodes to verify whether the bandwidth limitation works.

Preferably, said step four specifically includes the following steps:

S401, download the Caliper open source code from the Github website;

S402, configuring the dependent environment of Caliper on a Linux host outside the blockchain network;

S403, deploy Caliper in the Linux system, and integrate the Caliper software with the actual blockchain platform

bind;

S404, modify the configuration file config.yaml required for the Caliper test, and convert the file config.yaml

The test smart contract method in the test is changed to the actual deployed smart contract method;

S405, start the Caliper software to perform the TPS test.

The beneficial effect brought by adopting the above-mentioned technical scheme:

(1) Based on the analysis of the transaction process, this method establishes a mathematical model between the blockchain network TPS and the communication and computing capabilities of the blockchain nodes, and realizes the prediction of the performance of the blockchain network before the deployment of the blockchain system. This mathematical model is convenient for the hardware selection of the blockchain nodes and the cost estimation of the network construction when the blockchain system is actually deployed, and the feasibility of actually deploying the blockchain system under a specific network structure can be verified by estimating the performance of the blockchain.

(2) This method tests the TPS performance based on the actual blockchain platform built, and compares the test results with the mathematical model to verify the accuracy of the established mathematical model, and obtains that the maximum prediction error of the verification result does not exceed 10%.

Description of drawings

Fig. 1 is a block chain system node identity and function schematic diagram of the embodiment of the present invention;

FIG. 2 is a schematic flow diagram of a PBFT-based block chain according to an embodiment of the present invention;

Fig. 3 is a schematic diagram of the block chain system transaction structure of the embodiment of the present invention;

Fig. 4 is a schematic flow chart of transaction verification based on PBFT block chain according to an embodiment of the present invention;

Fig. 5 is a schematic flow diagram of a block chain consensus node packaging transaction according to an embodiment of the present invention;

Fig. 6 is a schematic flow diagram of the block chain internal transaction executed by the whole network node of the block chain according to the embodiment of the present invention;

Fig. 7 is a relationship diagram between the average time of single-node transaction processing and the number of transactions in the embodiment of the invention;

Fig. 8 is a block chain system TPS mathematical model diagram based on PBFT of the embodiment of the present invention;

Fig. 9 is an actual structural diagram for verifying the TPS mathematical model of the block chain system according to the embodiment of the present invention;

Fig. 10 is a comparison chart of the actual test TPS and the TPS value predicted by the mathematical model of the embodiment of the present invention.

Detailed ways

The technical solutions of the present invention will be described in detail below in conjunction with the accompanying drawings.

The PBFT-based blockchain network structure analyzed according to the present invention is shown in FIG. 1 . PBFT-based blockchain network nodes mainly have several roles: access node, consensus node, and network-wide node. Each role is responsible for different responsibilities.

Referring to Figure 2, the present invention proposes a TPS prediction method oriented to the PBFT consensus blockchain network. The method analyzes the transaction flow of the blockchain system based on the Byzantine fault-tolerant algorithm. The transaction processing flow includes: transaction generation, transaction broadcast, transaction packaging, transaction execution, transaction consensus, and transaction placement. The present invention proposes a TPS prediction method oriented to the PBFT consensus blockchain network, including:

S1. According to the transaction format of the blockchain and the smart contract used by the test TPS, calculate the actual transaction size sent by the blockchain system and record it as S _tx for easy analysis. The test transaction content and format used are the same;

See Figure 3, the above blockchain transaction format, including the content in the figure, such as transaction type, random number nonce, recipient address and other transaction-related information.

S2, according to the actual deployment of network nodes, record the maximum uplink and downlink rate limit of each node as C, then the transmission time of each transaction between two nodes is At the same time, it is assumed that the time for each transaction to be generated at the node is T _txbuild , so the time for each transaction to be generated by the client and sent to the blockchain node is />

S3, assuming that the transaction signature verification time is T _txcheck , because the test transaction content and format adopted are the same, so the signature verification time of each transaction is T _txcheck , and the signature verification time is only considered once in the total transaction processing time;

Referring to Figure 4, the above-mentioned blockchain transaction signature verification process includes verifying whether the transaction is legal or repeated. If it is illegal or the transaction is repeated, the transaction will be discarded directly; if the transaction is legal and there is no duplication, the transaction will be stored in the local transaction pool of the node.

S4, assuming that the packaging time of the block is T _blockpk , the size of the block header is S _blockhead , and each block contains n transactions, the total time for the transaction to be packaged into a block and sent to other nodes is:

Referring to Figure 5, the above-mentioned blockchain transaction packaging process includes taking out transactions from the head of the transaction pool. The Sealer takes several transactions as the main body of the block, attaches the information of the current node to the block header, packs the above information into a block, and sends the block to other blockchain nodes.

S5, assuming that the block verification time is T _blockcheck and the execution time of each transaction is T _tx , then the time for each block verification and execution on the node is T _blockcheck +nT _tx ;

S6. After the execution of a block is completed, the node will broadcast the execution result of the block to other blockchain nodes. The data size of the block execution result is negligible compared with the size of the block. It is assumed that the time for placing the transaction is T _blocksave .

See Figure 6. The rest of the blockchain nodes receive the block information sent by other nodes. First, the basic information of the block is verified by the block validator, including verifying whether the identity information and signature of the sending node are legal. If the block passes the legality test, the execution engine will take out transactions from the block one by one, and search for relevant smart contract codes from the local smart contract library according to the specific content of the transaction and execute them one by one. After the execution is completed, the current node will broadcast the execution results on the entire network.

S7, the total time to execute a block is

T _tot =n(T _txbuild +T _txsend +T _txcheck )+T _blockpk +T _blocksend +T _blockdone +T _consensus +T _blocksave .

S8, dividing the transaction processing time into two parts, the communication time T _comm and the calculation time T _comp . The calculation time is T _comp =n(T _txbuild +T _txcheck )+T _blockpk +T _blockcheck +nT _tx +T _blocksave , and the size of the block header can be ignored when calculating the communication time, so the communication time can be expressed as

S9, the calculation time of blockchain processing transactions is affected by many factors, and it is difficult to quantify accurately. Therefore, the average time T _txcomp of each transaction processing is obtained by performing multiple single-node transaction processing, and the test results under different transaction quantities are recorded to draw the relationship between the execution time of each transaction and the transaction volume.

See Figure 7. In this step, a single node is built on a single machine for testing. The purpose is to measure the more accurate total time of transaction creation, transaction execution, and transaction placement. Figure 7 shows the relationship between the number of different transactions sent to the blockchain and the execution time of each transaction. It can be seen that the total time for each transaction to be generated, verified, executed, and placed in the blockchain has no obvious relationship with the number of transactions entered in each test. The average time of all transactions in the figure is 0.7397ms. In the following, this value is used as the average time for each transaction to be processed by the blockchain. into the mathematical model for fitting.

S10, so the established blockchain TPS mathematical model is Using the Matlab program, the TPS curve of the blockchain system can be drawn with the communication capability of the node as the independent variable.

Referring to Figure 8, it can be seen that when the communication capability of the blockchain nodes is poor, the TPS value of the blockchain is mainly affected by the communication capabilities of the blockchain nodes. Increasing the maximum uplink and downlink rate of the nodes can effectively improve the TPS of the blockchain network. However, when the communication capabilities of the nodes increase to a certain size, the impact on the TPS of the blockchain will gradually decrease if the communication capabilities of the nodes continue to be improved.

S11, build an actual blockchain network with the same network structure as the mathematical model, conduct a TPS test on the actual blockchain network, and compare it with the above-mentioned theoretical analysis value.

S1101, provide multiple physical machines, and multiple physical machines access each other through a router, configure the static IP address of the physical machine, and use the Ping command in the command line to detect the communication between each blockchain node.

See Figure 9 to build an actual blockchain network.

For example: using one router to achieve mutual access between 10 physical machines, the structure of the blockchain network is shown in Figure 2. Start the Linux operating system on the physical machines, and configure the static IP addresses of 10 physical machines as 192.168.1.200-209.

S1102, configuring the environment required for running the blockchain node on the physical machine. For example: install openssl, curl, java 14.0.2, docker and other software.

S1103, download the chain building script build_chain.sh, modify the ipconf configuration file according to the actual situation of the blockchain network, and configure the ip address of the blockchain node and the organization to which it belongs (the organization in the alliance chain, press: the alliance chain is composed of multiple institutions) in the ipconf configuration file.

S1104, execute the chain building script build_chain.sh, automatically generate the node certificates required to build the blockchain network according to the configuration in the ipconf configuration file, and send the generated certificates to each blockchain node through SCP.

S1105, write a script, access the blockchain nodes one by one, execute the blockchain node startup program in the specified directory, the startup program will run the docker container under the current blockchain node, and map the preset docker container port to the physical machine port to realize communication with other docker containers.

S1106, if you want to prevent direct communication between two nodes in the blockchain network, you need to prohibit the mutual access between the two points through the TC command. If the prohibited node wants to obtain the transaction or block sent by the other party, it can only wait for other nodes in the blockchain network to receive the broadcast after the transaction;

S1107, after all the blockchain nodes are started, select any blockchain node, execute the view command, and observe whether all the nodes in the blockchain network can discover each other directly or in multiple ways;

S1108, configure the Caliper blockchain TPS testing software on a Linux host outside the blockchain network, and send transactions to any node of the blockchain network through the node for TPS testing;

S1109, modifying the maximum uplink and downlink rates of each node, and performing multiple tests respectively, so as to compare with the TPS prediction values at different rates of the mathematical model;

S1110, drawing and comparing the test results of the actual network and the prediction curve of the mathematical model through Matlab software, and calculating the maximum error.

Referring to Figure 10, it can be seen that the established mathematical model and the actual 10-node fully-connected blockchain system have high accuracy when the communication rate is low, and a large error occurs when the maximum communication rate of the nodes is greater than 10Mbyte/s. The reason for the error is mainly due to the influence of the test environment. The gigabit router used in the actually built blockchain system allows a maximum flow rate of 100Mbyte/s. If the uplink and downlink rate of a node is higher than 10Mbyte/s, queuing and sending will occur in the router, resulting in a decrease in communication efficiency and affecting the final TPS value of the experiment. It can also be seen from the TPS curve of the actual test that after the network bandwidth reaches 10Mbyte/s, the improvement of node communication capabilities has little effect on the actual network TPS. Taking the data with the highest communication rate of 50Kbyte/s-10Mbyte/s of nodes in the above figure as valid data, it can be obtained that the maximum error of the mathematical model in the global coupling network is 6.74%.

The present invention provides an electronic device, which includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

The present invention provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the steps of any one of the above methods are realized.

It should be understood that if the integrated units are realized in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention is essentially a part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, the computer software product is stored in a storage medium, and includes several instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, and other media that can store program codes.

The embodiment is only to illustrate the technical idea of the present invention, and can not limit the protection scope of the present invention. Any modification made on the basis of the technical solution according to the technical idea proposed in the present invention all falls within the protection scope of the present invention.

Claims (5)

1. A block chain TPS estimation method oriented to consensus algorithm, is characterized in that, comprises the following steps: S1, according to the transaction format of the blockchain and the smart contract used by the test TPS, calculate the actual transaction size S _tx sent by the blockchain system, using the same test transaction content and format; S2. According to the actual deployment of network nodes, obtain the time T _txbuild of each transaction generated at the client, calculate the time T _txsend of each transaction generated by the client and sent to the blockchain node connected to the client, and record the maximum uplink and downlink rate of each blockchain node as C, then the transmission time of each transaction between two blockchain nodes is Get the time T _txbuild of each transaction generated at the blockchain node, then the time for each transaction generated by the client and sent to the blockchain node is /> S3, get the signature verification time T _txcheck of the transaction, the signature verification time of each transaction is the same, and the signature verification time is only considered once in the total transaction processing time; S4. Obtain the packaging time T _blockpk of the block, the size of the block header S _blockhead and the number of transactions n contained in each block, and calculate the total time T _blocksend of the transaction packaging block and sending it to other blockchain nodes. The calculation formula is as follows: Where C is the maximum uplink and downlink rate limit of each blockchain node; S5, obtain the block verification time T _blockcheck and each transaction execution time T _tx , calculate the time T _blockdone of each block verification and execution on the node, the calculation formula is as follows: T _{block done} = T _{block check} + nT _tx ; S6, obtain the time T _blocksave of the transaction order; S7, according to T _txbuild , T _txsend , T _blockpk , T _blocksend , T _blockcheck , T _blockdone , T _blocksave obtained in steps S2-S6, calculate the total time T _tot of completing a block; divide T _tot into communication time T _comm and calculation time T _comp ; obtain the average time T _txcomp of each transaction processing by performing multiple single-node transaction processing, and the calculation formula is as follows: T _tot ＝n(T _txbuild +T _txsend +T _txcheck )+T _blockpk +T _blocksend +T _blockdone +T _consensus +T _blocksave Among them, T _consensus refers to the consensus time of the block, which is ignored when calculating the communication time; The formula for calculating the communication time T _comm is as follows: In the formula, N is the number of blockchain nodes, and C is the maximum uplink and downlink rate limit of each blockchain node; The calculation formula of the calculation time T _comp is as follows: T _comp ＝n(T _txbuild +T _txcheck )+T _blockpk +T _blockcheck +nT _tx +T _blocksave S8. Establish a blockchain TPS mathematical model, and use the communication capability of the node as an independent variable to import the model to draw a TPS estimation curve for the blockchain system. The calculation formula of the blockchain TPS mathematical model is expressed as follows: In the formula, N is the number of blockchain nodes, C is the maximum uplink and downlink speed limit of each blockchain node, and S _tx is the size of each transaction. 2. The accuracy verification method based on a consensus algorithm-oriented block chain TPS estimation method according to claim 1, characterized in that, comprising the steps of: Step 1, build a blockchain network with the same network structure as that used in the blockchain TPS mathematical model analysis, and deploy and test smart contracts on the blockchain network; Step 2, deploy the Caliper blockchain performance testing software on the Linux host outside the blockchain network; Step 3, install system bandwidth limiting software on each blockchain node, modify the maximum uplink and downlink bandwidth of each node, and verify the error between the predicted results of the blockchain TPS mathematical model and the actual TPS at different rates; Step 4: After modifying the maximum uplink and downlink bandwidth of the node, use the Caliper blockchain performance testing software to initiate multiple rounds of TPS tests on the deployed blockchain network through the Linux host outside the blockchain network, record the actual TPS results of the test, and average multiple sets of data at the same rate; Step 5: Draw a graph of the actual results, compare it with the TPS theoretical value predicted by the blockchain TPS mathematical model, and calculate the error between the results predicted by the mathematical model and the actual test under different node communication capabilities, and judge the accuracy of the blockchain TPS estimation method. 3. The accuracy verification method according to claim 2, characterized in that, in said step 1, a block chain network with the same structure as the block chain TPS mathematical model is built, specifically comprising the following steps: S101, provide multiple physical machines, and multiple physical machines access each other through a router, configure the static IP address of the physical machine, and use the Ping command in the command line to detect the communication between each blockchain node; S102, configuring the environment required for the operation of the blockchain node on the physical machine; S103, downloading the chain building script, modifying the ipconf configuration file according to the actual situation of the blockchain network, and correspondingly configuring the ip address of the blockchain node and the affiliated institution in the ipconf configuration file; S104, execute the chain building script, automatically generate the node certificates required for building the blockchain network according to the configuration in the ipconf configuration file, and send the generated certificates to each blockchain node through SCP; S105, write the script, visit the blockchain nodes one by one, and execute the blockchain node startup program on each Linux host in the blockchain network, the startup program will run the docker container under the blockchain node, and map the preset docker container port to the physical machine port to realize communication with other docker containers; S106, prohibiting mutual access between two nodes through the TC command, when the prohibited node obtains the transaction or block sent by the other party, it needs to wait for the broadcast after receiving the transaction from other nodes in the blockchain network; S107, after all the blockchain nodes are started, select any blockchain node, execute the view command, and observe whether all the nodes in the blockchain network can discover each other directly or in multiple ways. 4. The accuracy verification method according to claim 2, characterized in that, modifying the maximum uplink and downlink rates of each node in the step 3 specifically comprises the following steps: S301, download the wondershaper bandwidth limiting software on each physical machine; S302, writing the automation script SpeedControl.sh, using the uplink and downlink bandwidths to be limited as the two input values of the script respectively; S303, installing iperf3 bandwidth testing software on all blockchain nodes; S304, execute SpeedControl.sh, and limit the node bandwidth to a set value; S305, start the iperf3 bandwidth testing software, and test the communication bandwidth between any two nodes to verify whether the bandwidth limitation works. 5. the accuracy verification method according to claim 2, is characterized in that, described step 4 specifically comprises the following steps: S401, download the Caliper open source code from the Github website; S402, configuring the dependent environment of Caliper on a Linux host outside the blockchain network; S403, deploy Caliper in the Linux system, and bind the Caliper software with the actual blockchain platform; S404, modify the configuration file config.yaml required for Caliper testing, and change the test smart contract method in the file config.yaml to the smart contract method actually deployed; S405, start the Caliper software to perform the TPS test.