论文部分内容阅读
针对目前硬件木马的侧信道检测普遍采用基于降维与主特征提取的分类方法,该方法在选取有用信息过程中可能会损失包含木马特征的关键信息这一问题,提出了一种基于自组织竞争神经网络的硬件木马检测方法.该方法在不损失有用信息的基础上,采用无监督学习的方式建立数学模型,对母本信息与待测信息进行分类判别.基于FPGA搭建了验证系统并对侧信道电流信息进行采集.数据处理结果表明:该方法可以有效检测出占母本电路面积0.16%的硬件木马.
In view of the fact that the sidechannel detection of hardware Trojan generally adopts the classification method based on dimensionality reduction and main feature extraction, the method may lose the key information including the characteristics of the trojan during the process of selecting useful information, and proposes a self-organizing competition Neural network hardware Trojan detection method.This method does not lose useful information based on unsupervised learning mode to establish a mathematical model to classify the mother information and test information classification based on FPGA to build a verification system and the contralateral Channel current information is collected.The data processing results show that this method can effectively detect the hardware Trojan occupying 0.16% of the area of the female circuit.