基于网络流量的挖矿行为检测识别技术研究

针对国内现有挖矿行为检测识别技术的准确性不高、缺少具体挖矿行为证据等问题，提出了一种基于网络流量的挖矿行为检测识别模型和多维度挖矿指纹特征提取方法，该模型通过对Stratum、Getwork等矿池协议的指令特征提取分析，能够高效准确地实现对挖矿行为的自动检测识别，并提炼出多维度的挖矿指纹特征，包括挖矿指令、矿池币种、软件型号、挖矿账号、算力、能耗等信息。实验结果表明，该模型能够准确识别相关挖矿行为并分析得出其多维度挖矿指纹特征，算力特征识别的准确性为95%左右，总体的挖矿行为检测识别的准确性为91.73%，为虚拟货币挖矿检测提供一种精准有效的解决方案。

Research on mining behavior detection and identification technology based on network traffic

In view of the low accuracy of the existing domestic mining behavior detection and identification technology, lack of specific mining behavior evidence, and unfavorable supervision, this paper proposed a network traffic-based mining behavior detection and identification model and multi-dimensional mining behavior. Mining fingerprint feature extraction method, this model could efficiently and accurately realize automatic detection and identification of mining behavior by extracting and analyzing the command features of mining pool protocols such as Stratum and Getwork, and extracted multi-dimensional mining fingerprint features, including mining mining instructions, mining pool currency, software model, mining account, computing power, energy consumption and other information. The experimental results show that the model can accurately identify relevant mining behaviors and analyze its multi-dimensional mining fingerprint features. The accuracy of computing power feature recognition is about 95%, and the overall mining behavior detection and recognition accuracy is 91.73%, to provide an accurate and effective solution for regulating virtual currency mining.