摘要
针对集成电路中的硬件木马问题,开展了基于侧信道分析的硬件木马电路功耗模型的设计和优化工作.在完成建模的基础上,分析了木马电路对模型参数的影响.针对木马检测中的工艺偏差噪声问题,提出了一种基于主成分分析的模型优化算法.该方法利用协方差矩阵完成数据的投影变换,从而减小工艺偏差噪声对测试的影响.经仿真验证表明建立的模型与实验得到的数据基本符合.通过对测试电路进行蒙特卡罗分析,完成了工艺偏差噪声的仿真,同时验证了模型优化算法的有效性.
A power consumption model for the hardware Trojan was proposed and optimized based on side channel analysis. Then, the influence of the Trojan circuit on the model parameters was analyzed. An optimization algorithm through the principal component analysis was presented to deal with the process variation noise in the Trojan detection. This method which utilizes covariance matrix to archive data transformation can reduce the influence of process variation in the tests. The simulation results show that the proposed model is basically in line with the experimental data. Monte Carlo analysis was used to simulate the process variation of test circuits, and the effectiveness of the model optimization algorithm was validated.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2013年第2期53-57,共5页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
天津市自然科学基金重点资助项目(12JCZDJC20500)
关键词
电路建模
硬件木马
侧信道分析
工艺偏差
主成分分析
蒙特卡罗分析
circuit modeling
hardware Trojan
side channel analysis
process variation
principal component analysis
Monte Carlo analysis
