基于电磁泄漏信号的电子设备行为识别与安全应用

Behavior Recognition and Security Application of Electronic Equipment based on Electromagnetic Leakage Signal

使用时域分析可以从芯片等集成电路的泄漏信号中提取出算法执行等操作行为,但该方法在信号空间捕获、时间对齐上有较大的局限性。使用频谱扫描和比对可以检测出具有较大电磁泄漏信噪比的设备行为特征,但该方法实时性差,容易受噪声影响。测试发现,结合上述两种方法,通过动态的频谱变化呈现可以得到更明显的操作行为相关的特征。通过采集并分析电磁泄漏信号的频谱变化特征,提出电子设备的行为识别新方法,探讨行为电磁特征库的建立方法和基于电磁泄漏信号的芯片级、设备级、系统级行为识别安全应用。

With time-domain analysis,operation behaviors such as algorithm execution can be extracted from the leakage signals of integrated circuits such as chips,but this method has great limitations in signal space acquisition and time alignment.Spectral scanning and comparison can detect device behavior characteristics with large electromagnetic leakage SNR,but the method is poor in real-time and susceptible to noise.Experiment test indicates that by combining the above two methods,more obvious operational behavior-related characteristics can be obtained by dynamic spectrum change presentation.By collecting and analyzing the spectrum change characteristics of electromagnetic leakage signals,a new method for behavior recognition of electronic equipment is proposed,and the establishment method of behavior electromagnetic characteristic database and the safety application of chip level,equipment level and system level behavior recognition based on electromagnetic leakage signals are discussed.