| 引用本文： |
-
王毅诚,徐艳云,张萌,黄伟庆.文印一体机个体及工作状态识别方法研究[J].信息安全学报,已采用 [点击复制]
- WANG Yicheng,XU Yanyun,Zhang Meng,HUANG Weiqing.Working state identification method of all-in-one printing machine[J].Journal of Cyber Security,Accept [点击复制]
|
|
| 摘要: |
| 电子信息设备在工作时会散发无意识的电磁辐射信号，该信号会引发信息泄漏风险，为信息安全带来隐患。电磁辐射信号不仅包含信息设备正在处理的明文信息，还包含设备本身的硬件信息及工作状态信息。不同的信息设备由于内部元器件构成不同，其散发的电磁辐射信号特征不同，同一台信息设备在不同工作状态下，其内部工作的电路元器件不同，电磁辐射信号特征也不同。本文针对这一特性对文印一体机的电磁辐射进行研究，提出了一种基于设备低频信号特征的文印一体机个体及工作状态识别算法，首先基于频谱特征与多分类SVM算法实现文印一体机个体识别，判断设备型号，而后基于个体识别结果对该文印一体机进行射频信息统计特征量提取，并运用K近邻及朴素贝叶斯进行分类，实现文印一体机工作状态精细化识别。经仿真与实验结果证明，本算法能够有效识别文印一体机个体型号及其工作状态，个体及工作状态识别率均达到90%以上。本研究可用于加强对敏感信息设备的监测，从而针对性的对设备进行精细化管控与防护，在保密领域具有广阔的应用前景，在电子对抗的方向上也具有较高的应用价值。 |
| 关键词: 电磁泄漏 电磁信息安全 个体识别 文印一体机 TEMPEST SVM K近邻 |
| DOI： |
| 投稿时间：2023-02-05修订日期：2023-04-12 |
| 基金项目:中国科学院C类战略性先导科技专项 XDC02040300 |
|
| Working state identification method of all-in-one printing machine |
|
WANG Yicheng, XU Yanyun, Zhang Meng, HUANG Weiqing
|
| (1Institute of Information Engineering,Chinese Academy of Sciences) |
| Abstract: |
| Electronic information equipment will emit unconscious electromagnetic radiation signals during operation, which can cause information leakage risks and pose hidden dangers to information security. The electromagnetic radiation signal not only contains the plaintext information being processed by the information device, but also includes the hardware information and working status information of the device itself. Due to the different composition of inter-nal components, different information devices emit different characteristics of electromagnetic radiation signals. Under different working conditions, the same information device has different internal working circuit components and electromagnetic radiation signal characteristics. This article studies the electromagnetic radiation signal charac-teristics of the text printing integrated machine based on this characteristic, and proposes a recognition algorithm for the individual and working status of the text printing integrated machine based on the low-frequency signal characteristics of the device. The algorithm is divided into two steps. Firstly, based on spectral features and multi classification SVM algorithm, the individual recognition of the printing machine is achieved, and the device model is determined. Then, based on the individual recognition results, the radio frequency information statistical feature quantity of the printing machine is extracted, and K-nearest neighbors and naive Bayes are used for classification. The algorithm ultimately achieves fine recognition of the working status of the text printing integrated machine through the above two steps. Through simulation and experimental results, it has been proven that the algorithm used in this article can effectively identify the individual model and working status of the text printing integrated machine, and the recognition rate of both individual and working status reaches over 90%. This study can be used to strengthen the monitoring of sensitive information devices, thereby providing targeted and refined control and pro-tection of devices. It has broad application prospects in the field of confidentiality and high application value in the direction of electronic countermeasures. |
| Key words: Electromagnetic leakage radiation Information security Multifunctional printer TEMPEST SVM K-nearest neighbor |
