An Anomaly Detection Scheme for Blockchain Transactions Based on AdaBoost Model

doi:10.3969/j.issn.1671-1122.2024.01.003

Abstract

Abstract:

In response to potential anomalous behaviors, such as coin theft in the transaction records of the blockchain-based cryptocurrency, a detection scheme with privacy protection function based on the adaptive boosting (AdaBoost) model was proposed. This scheme integrated additive homomorphic encryption and matrix perturbation techniques, ensuring the preservation of transaction data privacy while effectively identifying and predicting anomalies. The scheme’s underlying protocol was designed and implemented in a cloud outsourcing environment, and its correctness and privacy protection properties were proven. Compared with similar protocols, this scheme has high detection accuracy and recall while ensuring privacy. The detection time for each record was at the millisecond level, making it suitable for real cryptocurrency transaction detection scenarios.

Key words: privacy protection, machine learning, anomaly detection, homomorphic encryption

CLC Number:

TP309

SONG Yuhan, ZHU Yuefei, WEI Fushan. An Anomaly Detection Scheme for Blockchain Transactions Based on AdaBoost Model[J]. Netinfo Security, 2024, 24(1): 24-35.

Figures/Tables 4

References 23

[1]	KUMAR N, SINGH A, SHUKLA S K. Detecting Malicious Accounts on the Ethereum Blockchain with Supervised Learning[C]// Springer. Cyber Security Cryptography and Machine Learning- Fourth International Symposium. Heidelberg: Springer, 2020: 94-109.
[2]	BARTOLETTI M, LANDE S, POMPIANU L, et al. A General Framework for Blockchain Analytics[C]// ACM. The 1st Workshop on Scalable and Resilient Infrastructures for Distributed Ledgers. New York: ACM, 2017: 1-6.
[3]	KHALILOV M C K, LEVI A. A Survey on Anonymity and Privacy in Bitcoin-Like Digital Cash Systems[J]. IEEE Communications Surveys & Tutorials, 2018, 20(3): 2543-2585.
[4]	JIA Jinyuan, SALEM A, BACKES M, et al. Memguard: Defending Against Black-Box Membership Inference Attacks via Adversarial Examples[C]// ACM.The 26th ACM Conference on Computer and Communications Security (CCS 2019). New York: ACM, 2019: 259-274.
[5]	TRAMER F, ZHANG Fan, JUELS A, et al. Stealing Machine Learning Models via Prediction Apis[C]// USENIX. 25th USENIX Security Symposium 2016. Berkeley: USENIX, 2016: 601-618.
[6]	HIRSHMAN J, HUANG Yifei, MACKE S. Unsupervised Approaches to Detecting Anomalous Behavior in the Bitcoin Transaction Network[EB/OL]. (2013-09-23) [2023-08-01]. https://cs229.stanford.edu/proj2013/HirshmanHuangMacke-UnsupervisedApproachesToDetectingAnomalousBehaviorInTheBitcoinTransactionNetwork.pdf.
[7]	PHAM T, LEE S. Anomaly Detection in Bitcoin Network Using Unsupervised Learning Methods[EB/OL].(2016-11-12) [2023-08-01]. http://arxiv.org/abs/1611.03941.
[8]	PHAM T, LEE S. Anomaly Detection in the Bitcoin System-A Network Perspective[EB/OL]. (2016-11-13) [2023-08-01]. http://arxiv.org/abs/1611.03942.
[9]	MONAMO P M, MARIVATE V, TWALA B. A Multifaceted Approach to Bitcoin Fraud Detection: Global and Local Outliers[C]// IEEE. The 15th IEEE International Conference on Machine Learning and Applications. New York: IEEE, 2016: 188-194.
[10]	AWAN M K, CORTESI A. Blockchain Transaction Analysis Using Dominant Sets[C]// Springer. The 16th IFIP TC8 International Conference. Heidelberg: Springer, 2017: 229-239.
[11]	KIM J, NAKASHIMA M, FAN Wenjun, et al. Anomaly Detection Based on Traffic Monitoring for Secure Blockchain Networking[C]// IEEE. IEEE International Conference on Blockchain and Cryptocurrency (ICBC 2021). New York: IEEE, 2021: 1-9.
[12]	SONG Yuhan, WEI Fushan, ZHU Kaijie, et al. Anomaly Detection as a Service: An Outsourced Anomaly Detection Scheme for Blockchain in a Privacy Preserving Manner[J]. IEEE Transactions on Network and Service Management, 2022, 19(4): 3794-3809. doi: 10.1109/TNSM.2022.3215006 URL
[13]	LI Qinbin, WU Zhaomin, WEN Zeyi, et al. Privacy-Preserving Gradient Boosting Decision Trees[EB/OL]. (2020-10-29) [2023-08-01]. https://arxiv.org/abs/1911.04209.
[14]	YANG Mengwei, SONG Linqi, XU Jie, et al. The Tradeoff Between Privacy and Accuracy in Anomaly Detection Using Federated XGBoost[EB/OL]. (2019-10-14) [2023-08-01]. https://arxiv.org/abs/1907.07157.
[15]	CHENG Kewei, FAN Tao, JIN Yilun, et al. SecureBoost: A Lossless Federated Learning Framework[EB/OL].(2021-04-07) [2023-08-01]. https://arxiv.org/abs/1901.08755.
[16]	CHEN Weijing, MA Guoqiang, FAN Tao, et al. SecureBoost+: A High Performance Gradient Boosting Tree Framework for Large Scale Vertical Federated Learning[EB/OL]. (2021-12-23) [2023-08-01]. https://arxiv.org/abs/2110.10927.
[17]	FREUND Y, SCHAPIRE R E. A Decision-Theoretic Generalization of on-Line Learning and an Application to Boosting[J]. Journal of Computer and System Sciences, 1997, 55(1): 119-139. doi: 10.1006/jcss.1997.1504 URL
[18]	MIVULE K, TURNER C, JI S. Towards A Differential Privacy and Utility Preserving Machine Learning Classifier[C]// Elsevier. The Complex Adaptive Systems 2012 Conference. Amsterdam: Elsevier, 2012: 176-181.
[19]	SHEN Siqian. Research on Differential Privacy Protection Classification Algorithm[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2017.
	沈思倩. 关于差分隐私保护分类算法的研究[D]. 南京: 南京航空航天大学, 2017.
[20]	MA Zhuo, LIU Yang, LIU Ximeng, et al. Lightweight Privacy-Preserving Ensemble Classification for Face Recognition[J]. IEEE Internet of Things Journal, 2019, 6(3): 5778-5790. doi: 10.1109/JIOT.2019.2905555
[21]	DAVOUDI M. Efficient and Privacy-Preserving Adaboost Classification Framework for Mining Healthcare Data over Outsourced Cloud[D]. New Jersey: University of New Brunswick, 2020.
[22]	PAILLIER P. Public-Key Cryptosystems Based on Composite Degree Residuosity Classes[C]// Springer. EUROCRYPT’ 99. Heidelberg:Springer, 1999: 223-238.
[23]	CHEN Binjie. Research on Blockchain Abnormal Transaction Detection for Privacy Protection[D]. Zhengzhou: Information Engineering University of PLA, 2022.
	陈彬杰. 隐私保护的区块链异常交易检测研究[D]. 郑州: 战略支援部队信息工程大学, 2022.

方法	特征维度	隐私保护	通信带宽	通信轮数	计算复杂度	准确率	精度	召回率	F1 分数
LOF^[9]	3	否	$O({{n}^{2}})$	1	$O({{n}^{2}}\log n)$	89.5%	0	0	—
MDB^[8]	3	否	$O({{n}^{2}})$	1	$O({{n}^{2}})$	89.6%	0.5%	0.6%	0.006
OCSVM^[8]	3	否	$O(nd)$	1	$O({{n}^{3}})$	82.8%	0.5%	0.2%	0.003
ADaaS^[12]	9	是	$O(nd)$	2	${{E}_{v}}+nI{{P}_{v}}$	97.8%	85.8%	87.5%	0.867
本文方法	9	是	$O({{d}^{2}})$	2	$d({{E}_{p}}+{{D}_{p}})$	97.3%	97.1%	100%	0.985