一种融合多维信息的电力物联网设备指纹识别技术
网络安全与数据治理
卢列文1,刘佳烨2,杨盛明1
1.工业和信息化部电子第五研究所; 2.中国人民解放军93131部队
摘要: 针对电力物联网无算力设备面临的恶意篡改、非法接入等安全问题,以及现有设备指纹技术在复杂环境下适应性不足,提出一种融合多维信息的设备指纹识别技术,提升设备识别安全性与准确性。方法上,融合硬件标识、通信协议特征及运行状态数据,通过自适应哈希编码、注意力增强编码、窗口算法等构建动态指纹模型,并与主流算法对比验证。研究结果表明,该技术在正常及异常网络环境和电力运行状态下,准确率、召回率和F1值均较高,适配无算力设备,且能适应一定程度的环境变化。该技术为电力物联网无算力设备的精准识别和安全管理提供了有效方案,未来可结合新技术进一步优化,以适应更多新场景。
中图分类号:TP309.2文献标识码:ADOI:10.19358/j.issn.2097-1788.2025.09.001
引用格式:卢列文,刘佳烨,杨盛明. 一种融合多维信息的电力物联网设备指纹识别技术[J].网络安全与数据治理,2025,44(9):1-7.
引用格式:卢列文,刘佳烨,杨盛明. 一种融合多维信息的电力物联网设备指纹识别技术[J].网络安全与数据治理,2025,44(9):1-7.
A fingerprint recognition technology for power IoT devices integrating multidimensional information
Lu Liewen1,Liu Jiaye2,Yang Shengming1
1. China Electronic Product Reliability and Environmental Testing Research Institute; 2. 93131 Troops of the Chinese People′s Liberation Army
Abstract: To address security issues such as malicious tampering and illegal access faced by computing-power-free devices in the power Internet of Things (IoT), and existing device fingerprint technology lacks adaptability in complex environments, this paper proposes a device fingerprint recognition technology that integrates multi-dimensional information to enhance the security and accuracy of device recognition. Methodologically, it combines hardware identifiers, communication protocol characteristics, and operational status data to construct a dynamic fingerprint model through adaptive hashing, attention-enhanced encoding, window algorithms, and other techniques, and verifies its performance through comparison with mainstream algorithms. The research results show that this technology achieves high accuracy, recall rate, and F1 score in both normal and abnormal network environments and power operation states. It is suitable for computing-power-free devices and can adapt to a certain degree of environmental changes. This technology provides an effective solution for accurate identification and security management of computing-power-free devices in the power IoT. In the future, it can be further optimized by combining new technologies to adapt to more new scenarios.
Key words : power Internet of Things; computing-power-free device; device fingerprint; identification technology; modeling; experimental simulation
引言
电力物联网作为智能电网发展的关键支撑,正逐渐改变着传统电力系统的运营模式[1]。在电力物联网架构中,无算力设备(通常是指不具备或仅具有极弱数据处理、计算能力的设备),如电表、数据终端单元(Data Terminal Unit,DTU)、馈线终端单元(Feeder Terminal Unit,FTU)、配电变压器监测终端(Transformer Terminal Unit,TTU)等,这类设备数量庞大且分布广泛,它们负责采集、传输电力数据,是电力物联网实现智能化的基础。然而,随着电力物联网的规模不断扩大,这些设备面临着严峻的安全挑战,如设备被恶意篡改、非法接入等问题,严重威胁电力系统的安全稳定运行。
设备指纹识别技术作为一种有效的设备管理和安全防护手段,能够为每台设备生成独一无二的身份信息,实现对设备的精准识别和追踪。通过准确识别无算力设备指纹,对于及时发现非法设备接入,提高设备管理的精细化程度,构建电力物联网安全防御体系具有重要意义。
本文详细内容请下载:
//www.51qz.net/resource/share/2000006700
作者信息:
卢列文1,刘佳烨2,杨盛明1
(1.工业和信息化部电子第五研究所,广东广州511300;
2.中国人民解放军93131部队,北京100068)

此内容为AET网站原创,未经授权禁止转载。
