摘要
针对基于旁路分析的硬件木马检测,旁路信号数据在空间中呈现高维分布、非线性的问题,探究在降低信号维数方面核函数的可行性,提出了一种基于K_L变换与支持向量机(Support Vector Machine,SVM)的硬件木马检测模型。在FPGA加密芯片中植入所设计的硬件木马进行实验,实验结果对比表明,利用主特征向量构建的SVM模型在检测占电路2%的硬件木马时,测试准确率达到了98.1%,为规模更小的木马检测提供了新思路。
Aiming at the problem that side-channel signal data appears non -linear and high- dimensional distribution in hardware Trojan detection. For detecting the feasibility of reducing the dimension of the signal by kernel function, hardware Trojan detection mode based on K_L transform and SVM is proposed. Finally, the hardware Trojan is implanted in the FPGA encryption chip. The experimental results show that the SVM with radial basis function has accuracy of 98.1 % in detecting the hardware Trojan occupying 2 % of the circuit, provides a new idea for the detection of smaller hardware Trojans.
作者
魏延海
李雄伟
张阳
王晓晗
郭东昕
WEI Yan-hai;LI Xiong-wei;ZHANG Yang;WANG Xiao-han;GUO Dong-xin(Shijiazhuang Campus of Army Engineering University,Shijiazhuang 050003,China)
出处
《火力与指挥控制》
CSCD
北大核心
2019年第10期38-42,共5页
Fire Control & Command Control
基金
国家自然科学基金(61271152,51377170)
国家青年科学基金(61602505)
河北省自然科学基金资助项目(F2012506008)
关键词
旁路分析
硬件木马检测
K_L
变换
SVM
side-channel analysis
hardware trojan detection
K_L transform
SVM
