一种基于决策树的比特币不可达节点发现方法

李锐光; 朱佳伟; 吴阜东; 高家奇; 徐大伟; 祝烈煌

doi:10.13700/j.bh.1001-5965.2022.0558

一种基于决策树的比特币不可达节点发现方法

doi: 10.13700/j.bh.1001-5965.2022.0558

李锐光^{1, 2},
朱佳伟²,
吴阜东³,
高家奇³,
徐大伟^{1, 3},
祝烈煌^1, ,

1.
北京理工大学网络空间安全学院，北京 100081
2.
国家计算机网络应急技术处理协调中心，北京 100029
3.
长春大学网络安全学院，长春 130022

基金项目: 国家重点研发计划(2020YFB1006100); 国家自然科学基金(62106060)

详细信息

通讯作者:
E-mail：liehuangz@bit.edu.cn

中图分类号: TP393；TN919
计量
- 文章访问数: 196
- HTML全文浏览量: 73
- PDF下载量: 6
- 被引次数: 0
出版历程
- 收稿日期: 2022-06-30
- 录用日期: 2022-08-26
- 网络出版日期: 2022-12-08
- 整期出版日期: 2024-06-27

A decision tree-based discovery method for Bitcoin unreachable nodes

LI Ruiguang^{1, 2},
ZHU Jiawei²,
WU Fudong³,
GAO Jiaqi³,
XU Dawei^{1, 3},
ZHU Liehuang^{1
, ,}

1.
School of Cyberspace Science and Technology，Beijing Institute of Technology，Beijing 100081，China
2.
National Computer Network Emergency Response Technical Team Coordination Center of China，Beijing 100029，China
3.
School of Cyber Security，Changchun University，Changchun 130022，China

Funds: National Key R & D Program of China (2020YFB1006100); National Natural Science Foundation of China (62106060)

More Information

Corresponding author: E-mail：liehuangz@bit.edu.cn

摘要

摘要:
不可达节点是指比特币网络中不接收外部连接请求的网络工作节点，发现、验证均较为困难。现有研究大多集中于可达节点，而对不可达节点的研究较少。为此，提出一种基于决策树算法的不可达节点发现方法，可以从大量比特币地址中自动分类发现不可达节点。实验结果表明：所提方法在实验数据集上分类准确率为95.73%，召回率为91.97%；在真实数据上进行实测，并利用网络空间搜索引擎进行验证，所提方法实际分类准确率为53.75%，召回率约为76.86%。对实验中发现不可达节点的总量、地理分布、所属网络服务商等进行统计分析，为比特币监管工作提供有力技术支撑。
- 比特币 /
- 不可达节点 /
- 可达节点 /
- 决策树 /
- 网络空间搜索引擎
Abstract:
Unreachable nodes refer to nodes that don't accept connection requests in the Bitcoin network, which are difficult to detect and verify. The existing studies mostly focused on the reachable nodes, but less on the unreachable nodes. A new approach is proposed to find the unreachable nodes based on a decision tree model, which can automatically classify unreachable nodes from a large numberof Bitcoin addresses. The results show that the proposed approach got an accuracy of 95.73% and a recall of 91.97% on the experimental dataset. The author applied the approach to the real dataset and verified it by the cyberspace search engines. The proposed approach’s accuracy was 53.75% and the recall was about 76.86%. The distribution of network providers, geographical areas, and the overall number of Unreachable nodes were discussed, which provided technical support for Bitcoin supervision.
- Bitcoin /
- unreachable nodes /
- reachable nodes /
- decision tree /
- cyberspace search engine

HTML全文

图 1 BNS系统结构

Figure 1. System structure of BNS

下载: 全尺寸图片幻灯片

图 2 比特币地址分类

Figure 2. Classification of Bitcoin addresses

下载: 全尺寸图片幻灯片

表 1 决策树深度与分类效果

Table 1. Decision tree depth and classification effect

树深度	准确率/%	召回率/%	F₁
1	85.84	99.26	0.8625
2	88.86	94.54	0.8836
3	89.10	97.20	0.8887
4	92.85	95.39	0.9227
5	95.72	91.98	0.9506
10	97.99	95.93	0.9771
15	99.32	98.83	0.9924
20	99.73	99.57	0.9969
22	99.76	99.66	0.9973
25	99.75	99.64	0.9972
30	99.75	99.64	0.9972

下载: 导出CSV

表 2 节点最小样本数与分类效果

Table 2. Node's minimum sample and classification effect

最小样本数	准确率/%	召回率/%	F₁
100	95.72	91.98	0.9506
200	95.73	91.97	0.9381
300	95.67	92.19	0.9501
400	95.61	9.17	0.9492
500	95.44	91.27	0.9472
600	95.27	92.15	0.9458

下载: 导出CSV

表 3 优化前后分类效果

Table 3. Classification effect before and after optimization

分类算法	准确率/%	召回率/%	F₁
决策树（默认参数）	85.84	86.67	0.8625
决策树（优化后）	95.73	91.97	0.9381

下载: 导出CSV

表 4 检索记录与相应的节点类型

Table 4. Searching records and corresponding node types

检索记录	可达节点	不可达节点	离线可达节点	离线不可达节点	虚假节点
是否可以检索到记录	是	是	是	是	否
是否提供比特币服务	是	否	是	否	否
是否超出扫描周期	是	是	否	否	否

下载: 导出CSV

表 5 真实数据分类

Table 5. Real data classification

可达节点	不可达节点	离线可达节点	离线不可达节点	虚假节点	总计
0	5688	113	2338	2444	10583

下载: 导出CSV

表 6 分类效果

Table 6. Classification results

数据集	准确率/%	召回率/%	F₁
实验数据集	95.73	91.97	0.9381
真实数据	53.75	76.86	0.6326

下载: 导出CSV

表 7 节点地理分布

Table 7. Geographical distribution of nodes

地区	所占比例/%
地区	可达节点^[4]	不可达节点
欧洲	58.07	50.21
美洲	30.91	29.91
亚洲	9.43	18.18
大洋洲	1.36	1.16
非洲	0.22	0.54

下载: 导出CSV

表 8 所属网络服务商统计

Table 8. Statistics of networkservice providers

网络服务商	不可达节点数量	占比/%
amazon.com	372	7.3
hetzner.com	293	5.7
digitalocean.com	276	5.4
China Telecom	178	3.5
comcast.com	162	3.2

下载: 导出CSV

参考文献(17)

[1]	NAKAMOTO S. Bitcoin: A peer-to-peer electronic cash system[EB/OL]. (2009-03-19)[2022-04-22]. https://cdn.nakamotoinstitute.org/docs/bitcoin.pdf.
[2]	蔡晓晴, 邓尧, 张亮, 等. 区块链原理及其核心技术[J]. 计算机学报, 2021, 44(1): 84-131. CAI X Q, DENG Y, ZHANG L, et al. The principle and core technology of blockchain[J]. Chinese Journal of Computers, 2021, 44(1): 84-131 (in Chinese).
[3]	CASTROS. Bitcoin P2P network sniffer[EB/OL]. (2012-08-26)[2022-04-22]. https://github.com/sebicas/bitcoin-sniffer.
[4]	LI R G, ZHU J W, XU D W, et al. Bitcoin network measurement and a new approach to infer the topology[J]. China Communications, 2022, 19(10): 169-179. doi: 10.23919/JCC.2022.00.030
[5]	LI R G, SHEN M, YU H, et al. A survey on cyberspace search engines[C]//Proceedings of the China Cyber Security Annual Conference. Berlin: Springer, 2020: 206-214.
[6]	DONET J A , PÉREZ-SOLA C, HERRERA-JOANCOMARTÍ J. The bitcoin P2P network[C]//Proceedings of the International Conference on Financial Cryptography and Data Security. Berlin: Springer, 2014: 87-102.
[7]	FADHIL M, OWENSON G, ADDA M. A Bitcoin model for evaluation of clustering to improve propagation delay in Bitcoin network[C]//Proceedings of the IEEE International Conference on Computational Science and Engineering and IEEE International Conference on Embedded and Ubiquitous Computing and 15th International Symposiumon Distributed Computing and Applications for Business Engineering. Piscataway: IEEE Press, 2016: 468-475.
[8]	PARK S, IM S, SEOL Y, et al. Nodes in the Bitcoin network: Comparative measurement study and survey[J]. IEEE Access, 2019, 7: 57009-57022. doi: 10.1109/ACCESS.2019.2914098
[9]	BIRYUKOV A, KHOVRATOVICH D, PUSTOGAROV I. Deanonymisation of clients in Bitcoin P2P network[C]//Proceedings of the ACM SIGSAC Conference on Computer and Communications Security. New York: ACM, 2014: 15-29.
[10]	NEUDECKER T, ANDELFINGER P, HARTENSTEIN H. Timing analysis for inferring the topology of the Bitcoin peer-to-peer network[C]//Proceedings of the International IEEE Conferences on Ubiquitous Intelligence & Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People, and Smart World Congress. Piscataway: IEEE Press, 2016: 358-367.
[11]	WANG L, PUSTOGAROV I. Towards better understanding of Bitcoin unreachable peers[EB/OL]. (2017-09-20)[2022-04-27]. http://arxiv.org/abs/1709.06837.pdf.
[12]	GRUNDMANN M, AMBERG H, HARTENSTEIN H. On the estimation of the number of unreachable peers in the Bitcoin P2P network by observation of peer announcements[EB/OL]. (2021-02-25)[2022-04-27]. http://arxiv.org/abs/2102.12774.pdf.
[13]	FRANZONI F, DAZA V. Improving Bitcoin transaction propagation by leveraging unreachable nodes[C]//Proceedings of the IEEE International Conference on Blockchain. Piscataway: IEEE Press, 2020: 196-203.
[14]	BIRYUKOV A, PUSTOGAROV I. Bitcoin over tor isn’t a good idea[C]//Proceedings of the IEEE Symposium on Security and Privacy. Piscataway: IEEE Press, 2015: 122-134.
[15]	MASTAN I D, PAUL S. A new approach to deanonymization of unreachable Bitcoin nodes[C]//Proceedings of the International Conference on Cryptology and Network Security. Berlin: Springer, 2018: 277-298.
[16]	PAPPALARDO G, CALDARELLI G, ASTE T. The bitcoin peers network[EB/OL]. (2016-06-30)[2022-04-28]. http://blockchain.cs.ucl.ac.uk/wp-content/uploads/2016/11/P2PFISY2016_paper_32.pdf.
[17]	SALLAL M, DE FRÉIN R, MALIK A, et al. An empirical comparison of the security and performance characteristics of topology formation algorithms for Bitcoin networks[J]. Array, 2022, 15: 100221. doi: 10.1016/j.array.2022.100221