A method for tracing the originator of a Bitcoin transaction

By configuring test nodes in the Bitcoin network and using probes to collect data, a training model was built and database analysis was performed. This solved the problem of high resource consumption for tracing in a full-traffic environment and enabled efficient tracing and accurate classification of Bitcoin transaction initiators.

CN116362727BActive Publication Date: 2026-11-10NANJING FIBERHOME STARRYSKY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202310115155.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-02-15
Publication Date
2026-11-10
Estimated Expiration
2043-02-15

AI Technical Summary

Technical Problem

Tracing the origin of a Bitcoin transaction in a full-traffic environment requires excessive computing resources, and the tracing results often still have value for further analysis. Existing technologies struggle to efficiently trace the initiator of a Bitcoin transaction within a limited area.

Method used

By configuring test nodes to establish communication relationships with nodes in the region, using probes to collect data, building test sets to train models, improving the accuracy of tracing, and classifying each IP address into individual maintenance nodes or institutional maintenance nodes based on the number of wallets associated with each IP address, and using open source and proprietary databases to trace the institutional names.

Benefits of technology

It improves the accuracy of tracing the originators of Bitcoin transactions, enabling more accurate classification of IP addresses into individual or institutional nodes, and determining the name of the institution through database analysis. The data coverage is broader, thus enhancing the effectiveness of tracing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116362727B_ABST
    Figure CN116362727B_ABST
Patent Text Reader

Abstract

This invention discloses a method for tracing the initiator of a Bitcoin transaction, belonging to the field of blockchain technology. The method includes: establishing probe nodes to collect data; parsing the obtained pcap raw messages and restoring them into a session format; storing traffic data related to the Bitcoin mainnet; forming structured data and reconstructing wallet addresses based on the corresponding transaction patterns; merging input addresses with the same username based on Bitcoin characteristics to increase the sample set available for statistics; establishing a model to construct the association between IP addresses and users; classifying users into institutions or individuals based on the IP-wallet address association; if an institutional node is suspected, tracing the node's corresponding institutional name can be done through collision analysis using open-source and proprietary databases; establishing communication relationships between test nodes and nodes within the region and collecting data using probes to build a test set for model training, thereby improving the accuracy of tracing.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of blockchain technology, and in particular relates to a method for tracing the originator of a Bitcoin transaction. Background Technology

[0002] Bitcoin is a digital cryptocurrency based on blockchain technology. Due to its decentralized and anonymous characteristics, exploring Bitcoin transactions has practical social value.

[0003] A common method for tracing transactions at the network layer is to collect transaction information broadcast in the Bitcoin network layer, analyze the propagation path of each transaction, and trace the originating IP address to achieve transaction anonymity.

[0004] However, tracing transactions in a full-traffic environment requires excessive computing resources, and the originating address obtained by tracing transactions still has value for further analysis in many cases. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method for tracing the originator of a Bitcoin transaction for conducting transaction tracing work on nodes within a limited area. This method involves configuring test nodes to establish communication relationships with each node in the area and using probes to collect data, thereby establishing a test set to train a model and improve the accuracy of tracing.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] A method for tracing the originator of a Bitcoin transaction includes the following steps:

[0008] Step 1: Establish probe nodes to collect data, parse the obtained raw packets in pcap format, and reconstruct them into session format by reassembling the packets;

[0009] Step 2: For each session in the raw message in pcap file format, use the feature string encoding f9beb4d9 as the filtering condition, and only retain the protocol data related to the Bitcoin mainnet.

[0010] Step 3: parse the Bitcoin transaction action commandtx message in the Bitcoin protocol data to form structured data, and then reconstruct the wallet address based on the transfer mode corresponding to the structured data.

[0011] Step 4: Based on the characteristics of Bitcoin, merge wallet addresses with the same input into the same username to increase the sample set that can be used for statistics;

[0012] Step 5: Based on the sender IP address, sender port number, time, transaction ID, sender wallet address, and user group number from Step 4, build a model and construct the association between the sender IP address and the user.

[0013] Step 6: Based on the IP address associated with the wallet address, classify the IP address as an organization or an individual. If it is suspected to be an organization node, the name of the organization corresponding to the node can be traced by comparing the open-source database and the proprietary database. If it is an individual node, the IP address can be identified as an individual.

[0014] As a further preferred embodiment of the Bitcoin transaction initiator tracing method of the present invention, in step 1, there are approximately 70,000 nodes connected to the Bitcoin blockchain across the entire network, and the present invention can map the node situation in a specific area by establishing probe nodes.

[0015] As a further preferred embodiment of the Bitcoin transaction initiator tracing method of the present invention, in step 2, in the Bitcoin p2p network, all peer-to-peer communication is carried out entirely through the Transmission Control Protocol TCP.

[0016] In Transmission Control Protocol (TCP) traffic, Bitcoin protocol sessions have a fixed hexadecimal signature string encoded as f9beb4d9, which appears at the beginning of all messages sent on the Bitcoin network.

[0017] As a further preferred embodiment of the Bitcoin transaction initiator tracing method of the present invention, step 3 is specifically as follows:

[0018] Step 3.1: Extract version exchange (version), get node information (getaddr), test node availability (ping), Bitcoin transaction action (tx), and mining action (miner) from the Bitcoin traffic to parse the original data, and perform traffic parsing and address reconstruction for the Bitcoin transaction action (tx).

[0019] Among them, the Bitcoin transaction action tx is used to trace the sender of the transaction;

[0020] Version exchange (version), node information retrieval (getaddr), node availability testing (ping), Bitcoin transaction action (tx), mining action (miner) are used for information exchange and synchronization between nodes;

[0021] Step 3.2: The Bitcoin transaction action tx message is compatible with both the old and new structures. Field extraction and address restoration are performed for the two versions respectively. The resulting structured fields include: source IP, destination IP, source port, destination port, transaction ID, time, input summary, and output summary.

[0022] The input profile includes currency, amount, and address, while the output profile includes currency, amount, and address.

[0023] As a further preferred embodiment of the Bitcoin transaction initiator tracing method of the present invention, step 4 is specifically as follows:

[0024] Step 4.1: Since the sender must know the private keys of all input addresses simultaneously when initiating a transaction, or use a special method such as multi-signature for the transaction, all input addresses for each transaction are attributed to the same owner. Based on this principle, the addresses are identified to user units; this increases the sample set that can be used for statistics, that is, statistics based on users are obviously better than statistics based on IP addresses, and the data coverage is wider.

[0025] Step 4.2: Assign all input addresses in the test data that have the same transaction hash and are not multi-signature transactions to the same owner;

[0026] Step 4.3: Based on the results of step 4.2, the owners identified in transactions with different transaction hashes but the same input address are unified.

[0027] As a further preferred embodiment of the Bitcoin transaction initiator tracing method of the present invention, step 5 is specifically as follows:

[0028] Step 5.1: Transferring funds multiple times using the same Bitcoin address, with a certain node always being among the first to broadcast the transactions, confirms the association between the IP address and the wallet address. Identifying the wallet to the user unit is equivalent to establishing an association between the IP address and the corresponding user.

[0029] Step 5.2: The data with fields such as source IP, destination IP, source port, destination port, transaction ID, timestamp, and id that have been sorted out are grouped according to transaction hash, and a list is built based on the collected transaction propagation paths, and the three records with the earliest starting point are retained according to the timestamp.

[0030] Step 5.3: If two of these three paths share a common starting IP, then establish an association between that IP and the ID.

[0031] Step 5.4: If there is no common starting point, group the above data by ID, and count the number of times each ID is collected as the earliest sender of a certain transaction, and the distribution of IPs associated with the first release.

[0032] Step 5.5, establish a test dataset: configure the test node to send transactions to each node in the region, and simultaneously use the probe node to collect data and statistically analyze the collected IP distribution. Calculate the matching accuracy rate and take the average value as the threshold. For matching results that are greater than the average value, establish an association between the IP and the ID.

[0033] As a further preferred embodiment of the Bitcoin transaction initiator tracing method of the present invention, step 6 is specifically as follows:

[0034] Step 6.1: Some IPs correspond to multiple IDs, which should belong to institutional maintenance nodes, while some addresses are only associated with 1-2 wallets, which should belong to individual maintenance nodes;

[0035] Step 6.2, for institutional maintenance nodes: Using methods such as IP source analysis, certificate analysis, and same network segment analysis, attempt to trace the institutional name based on both open-source and proprietary databases. For individual nodes, the IP address can be identified as the individual.

[0036] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects:

[0037] 1. The present invention provides a method for tracing the originator of a Bitcoin transaction, which establishes communication relationships between test nodes and nodes in the region and uses probes to collect data to build a test set to train a model and improve the accuracy of tracing.

[0038] 2. After tracing the originating IP of a transaction, this invention classifies each IP into an individual maintenance node or an institutional maintenance node based on the number of wallets associated with each IP.

[0039] 3. For institutional nodes, this invention attempts to supplement the node identification by associating institutional names with open-source and proprietary databases through methods such as source analysis. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the structure of the old and new types of transaction messages in Bitcoin transactions provided by the present invention;

[0041] Figure 2 This is a schematic diagram of an embodiment of the present invention that identifies addresses with the same transaction hash to the same user;

[0042] Figure 3 This is a schematic diagram of a process provided by the present invention to identify transactions with the same input address as the same user;

[0043] Figure 4 This is a flowchart illustrating a Bitcoin transaction tracing method provided by the present invention. Detailed Implementation

[0044] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings:

[0045] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention. Those skilled in the art should understand that various substitutions and modifications made to the present invention based on basic technical knowledge and means in the art without departing from the above-described technical concept should be included within the protection scope of the present invention.

[0046] This invention discloses a method for tracing the originator of a Bitcoin transaction. The invention primarily targets nodes within a defined area for transaction tracing: it establishes communication relationships between test nodes and other nodes within the area, and uses probes to collect data to build a test set for model training, thereby improving tracing accuracy. Once the originating IP of the transaction is traced, it classifies nodes into individually maintained nodes or institutionally maintained nodes based on the number of wallets associated with each IP. For institutional nodes, it attempts to associate the institutional name with the node using open-source and proprietary databases through methods such as source identification analysis.

[0047] A method for tracing the originator of a Bitcoin transaction, such as Figure 4 As shown, the specific steps include:

[0048] Step 1: Establish probe nodes to collect data, parse the obtained pcap (a file format) raw packets, and reconstruct the session format by reassembling the packets;

[0049] Step 2: For each session in the pcap (a file format) raw message, use the feature string f9beb4d9 (feature string encoding) as the filtering condition, and only retain the protocol data related to the Bitcoin mainnet;

[0050] Step 3: parse the commandtx (Bitcoin transaction action) in the Bitcoin protocol data to form structured data, and then reconstruct the wallet address based on the transfer pattern corresponding to the structured data.

[0051] Step 4: Based on the characteristics of Bitcoin, merge wallet addresses with the same input into the same username to increase the sample set that can be used for statistics;

[0052] Step 5: Based on the data in Step 4 (sender IP, sender port number, time, transaction hash (transaction ID), sender wallet address, user group number), build a model and construct the association between sender IP address and user;

[0053] Step 6: Based on the IP address associated with the wallet address, classify the ID address to an organization or an individual. If it is suspected to be an organization node, the organization name corresponding to the node can be traced through collision analysis of the open-source database and the proprietary database. If it is an individual node, the IP address can be identified as an individual.

[0054] In step 1, as of October 2022, the total number of Bitcoin network nodes was approximately 70,000. This invention can perform mapping work on nodes within a specific area. These are nodes connected to the entire Bitcoin blockchain network; we can map the node situation within a specific area by establishing probe nodes.

[0055] In step 2, in Bitcoin's p2p network, all peer-to-peer communication is conducted entirely through TCP (Transmission Control Protocol).

[0056] In TCP (Transmission Control Protocol) traffic, Bitcoin protocol sessions have a fixed hexadecimal signature string f9beb4d9 that appears at the beginning of all messages sent on the Bitcoin network.

[0057] Step 3 is described in detail below.

[0058] Step 3.1: Extract version exchange (version), get node information (getaddr), test node availability (ping), Bitcoin transaction action (tx), and mining action (miner) from the Bitcoin traffic to parse the original data, and perform traffic parsing and address reconstruction for the Bitcoin transaction action (tx).

[0059] Among them, the Bitcoin transaction action tx is used to trace the sender of the transaction;

[0060] Version exchange (version), node information retrieval (getaddr), node availability testing (ping), Bitcoin transaction action (tx), mining action (miner) are used for information exchange and synchronization between nodes;

[0061] Step 3.2, since the tx message is compatible with both the old and new structures, it is roughly as follows: Figure 1 As shown, Figure 1 The structure of two types of messages in Bitcoin transactions is described. Since the transaction messages are compatible with both the old and new structures, we need to perform field extraction and address restoration for each version separately.

[0062] We need to perform field extraction and address restoration for the two versions separately, resulting in structured fields including: source IP, destination IP, source port, destination port, transaction ID, time, input summary (currency, amount, address), and output summary (currency, amount, address).

[0063] Step 4 is described in detail below.

[0064] Step 4.1: Since the sender must know the private keys of all input addresses simultaneously when initiating a transaction, or use a special (multi-signature) transaction, it's almost possible to attribute all input addresses for each transaction to the same owner. Using this principle, we can assign addresses to user units. This is primarily to increase the sample set available for statistical analysis (analyzing by user obviously covers a wider range of data than by IP address).

[0065] Step 4.2, specifically, we will assign all input addresses in the test data that have the same transaction hash and are not multi-signature transactions to the same owner, such as... Figure 2 As shown;

[0066] Figure 2 This is an implementation diagram that assigns addresses with the same transaction hash to the same user. In the test data, addresses 1, 2, and 3 with the same transaction hash are all assigned to user 1.

[0067] Step 4.3, based on the above results, further unify the owners identified in transactions with different transaction hashes but the same input address, such as... Figure 3 As shown.

[0068] Figure 3 This is an implementation diagram that identifies transactions with the same input address as the same user. In the test data, we identify all addresses in transaction hash 1 and transaction hash 2 that have the same input address - address 3 as user 1.

[0069] Step 5 is described in detail below.

[0070] Step 5.1: Transferring funds multiple times using the same Bitcoin address, with a certain node always being among the first to broadcast the transactions, confirms the association between the IP address and the wallet address. Identifying the wallet to the user unit is equivalent to establishing an association between the IP address and the corresponding user.

[0071] Step 5.2: The data with fields such as source IP, destination IP, source port, destination port, transaction ID, timestamp, and id that have been sorted out are grouped according to transaction hash, and a list is built based on the collected transaction propagation paths, and the three records with the earliest starting point are retained according to the timestamp.

[0072] Step 5.3: If two of these three paths share a common starting IP, then establish an association between that IP and the ID.

[0073] Step 5.4: If there is no common starting point, group the above data by ID, and count the number of times each ID is collected as the earliest sender of a certain transaction, and the distribution of IPs associated with the first release.

[0074] For example: (id1, associated IPs: ip1 (10 times, 2022.10.11-10.16), ip2 (3 times, 2022.10.17-10.20), ip3 (1 time, 2022.10.18-10.20));

[0075] Step 5.5, establish a test dataset: configure the test node to send transactions to each node in the region, and simultaneously use the probe node to collect data and statistically analyze the collected IP distribution. Calculate the matching accuracy rate and take the average value as the threshold. For matching results that are greater than the average value, establish an association between the IP and the ID.

[0076] Step 6 is described in detail below.

[0077] Step 6.1: Some IPs correspond to multiple IDs, which should belong to institutional maintenance nodes, while some addresses are only associated with 1-2 wallets, which should belong to individual maintenance nodes;

[0078] Step 6.2, for institutional maintenance nodes: Using methods such as IP source analysis, certificate analysis, and same network segment analysis, attempt to trace the institutional name based on both open-source and proprietary databases. For individual nodes, the IP address can be identified as the individual.

[0079] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0080] The above embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of this invention. The embodiments of the present invention have been described in detail above, but the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. A method of tracing a Bitcoin transaction initiator, characterized in that: Specifically, it includes the following steps: Step 1: Establish probe nodes to collect data, parse the obtained raw packets in pcap format, and reconstruct them into session format by reassembling the packets; Step 2: For each session in the raw message in pcap file format, use the feature string encoding f9beb4d9 as the filtering condition, and only retain the protocol data related to the Bitcoin mainnet. Step 3: parse the Bitcoin transaction action command tx message in the Bitcoin protocol data to form structured data, and then reconstruct the wallet address based on the transfer pattern corresponding to the structured data. Step 4: Based on the characteristics of Bitcoin, merge wallet addresses with the same input into the same username to increase the sample set that can be used for statistics; Step 5: Based on the sender IP address, sender port number, time, transaction ID, sender wallet address, and user group number from Step 4, build a model and construct the association between the sender IP address and the user. The specific details of step 5 are as follows: Step 5.1, when funds are transferred out multiple times using the same Bitcoin address, a certain node is always among the first to broadcast the transactions, thus determining that the IP address and the wallet address are associated, and the wallet is identified to the user unit, which is equivalent to establishing an association between the IP address and the corresponding user. Step 5.2: Group the collected data (source IP, destination IP, source port, destination port, transaction ID, timestamp, and id field type) according to transaction hash, then build a list based on the collected transaction propagation paths, and retain the 3 records with the earliest starting point according to timestamp. Step 5.3: If two of these three paths share a common starting IP, then establish an association between that IP and the ID. Step 5.4: If there is no common starting point, group the above data by ID, and count the number of times each ID is collected as the earliest sender of a certain transaction, and the distribution of IPs associated with the first release. Step 5.5, establish a test dataset: configure test nodes to send transactions to each node in the region, and simultaneously use probe nodes to collect data and statistically analyze the collected IP distribution. Calculate the matching accuracy rate and take the average as the threshold. For matching results that are greater than the average, establish an association between the IP and the ID. Step 6: Based on the IP address associated with the wallet address, classify the IP address as an organization or an individual. If it is suspected to be an organization node, the name of the organization corresponding to the node can be traced by comparing the open-source database and the proprietary database. If it is an individual node, the IP address can be identified as an individual.

2. The method of claim 1, wherein: In step 1, There are 70,000 nodes connected to the Bitcoin blockchain across the entire network, and this invention can map the node situation in a specific area by establishing probe nodes.

3. The method of claim 1, wherein: In step 2, In Bitcoin's peer-to-peer network, all peer-to-peer communication is conducted entirely through the Transmission Control Protocol (TCP). In Transmission Control Protocol (TCP) traffic, Bitcoin protocol sessions have a fixed hexadecimal signature string encoded as f9beb4d9, which appears at the beginning of all messages sent on the Bitcoin network.

4. The method of claim 1, wherein: The specific details of step 3 are as follows: Step 3.1, extract version exchange, getaddr node information, ping node availability test, tx Bitcoin transaction action, and miner mining action from the Bitcoin traffic to parse the original data, and perform traffic parsing and address restoration for the Bitcoin transaction action tx. Among them, the Bitcoin transaction action tx is used to trace the sender of the transaction; Version exchange (version), node information retrieval (getaddr), node availability testing (ping), Bitcoin transaction action (tx), mining action (miner) are used for information exchange and synchronization between nodes; Step 3.2: The Bitcoin transaction action tx message is compatible with both the old and new structures. Field extraction and address restoration are performed for the two versions respectively. The resulting structured fields include: source IP, destination IP, source port, destination port, transaction ID, time, input summary, and output summary. The input profile includes currency, amount, and address, while the output profile includes currency, amount, and address.

5. The method of claim 1, wherein: Step 4 is described in detail below. Step 4.1: Since the sender must know the private keys of all input addresses simultaneously when initiating a transaction, or use a special method such as multi-signature for the transaction, all input addresses for each transaction are attributed to the same owner. Based on this principle, the addresses are identified to user units; this increases the sample set that can be used for statistics, that is, statistics based on users are obviously more comprehensive than statistics based on IP addresses. Step 4.2: Assign all input addresses in the test data that have the same transaction hash and are not multi-signature transactions to the same owner; Step 4.3: Based on the results of step 4.2, the owners identified in transactions with different transaction hashes but the same input address are unified.

6. The method of claim 1, wherein: The specific details of step 6 are as follows: Step 6.1, some IPs correspond to multiple IDs, which should belong to institutional maintenance nodes, while some addresses are only associated with 1-2 wallets, which should belong to individual maintenance nodes; Step 6.2, for institutional maintenance nodes: Try using IP same-origin analysis, certificate analysis, and same-network-segment analysis methods. Based on open-source and proprietary databases, the name of the originating organization can be traced; if it is an individual node, the IP address can be identified to the individual.

Citation Information

Patent Citations

  • Identity identification method in bitcoin transaction

    CN108009807A

  • Bitcoin transaction station address identification method, system and device

    CN109918584A