Application Of Circuit Electromagnetic Scattering Model In Hardware Trojan Detection

With the globalization of the integrated circuit industry,integrated circuits are increasingly vulnerable to malicious attacks by redundant circuits.This redundant unit that performs malicious functions inside integrated circuits is a hardware Trojan.The integrated circuit is designed,manufactured,or used.Hardware Trojan circuits are embedded in the system.Nowadays,due to the increasing integration of integrated circuits,the area of??hardware Trojan circuits is continuously decreasing,which makes it more difficult to detect hardware Trojans.Hardware Trojans pose a growing threat to integrated circuits.Searching for an efficient and accurate hardware Trojan detection technology has become a hot topic for researchers at home and abroad.The study mainly carries out related research work based on electromagnetic side channel analysis technology.Firstly,the basic principle of electromagnetic radiation inside the integrated circuit is analyzed: the internal current change of the integrated circuit is related to the change of the frequency and amplitude of the electromagnetic radiation signal;then it is theoretically analyzed that the embedded hardware Trojan circuit will inevitably lead to the change of the internal current of the integrated circuit,and then The electromagnetic radiation information of the integrated circuit is changed.Finally,the method of establishing the dipole equivalent model of the electromagnetic radiation information is expounded,and the characteristics of each parameter in the equivalent model are analyzed to further explain the electromagnetic radiation information realized by the integrated circuit.The feasibility of hardware Trojan circuit detection.According to the characteristics of electromagnetic radiation information,the hardware Trojan circuit and carrier circuit are designed.At the same time,the reference circuit of ISCAS'89 series containing hardware Trojans based on FPGA is detected to obtain the electromagnetic radiation information of the reference circuit and the circuit embedded in the hardware Trojan.Aiming at the problem of poor data preprocessing in hardware Trojan detection,a method of processing data by convolutional neural network is proposed,and the process of data processing by convolutional neural network is introduced.Firstly,the applicableconvolutional neural network is created according to this topic.Then,the experimental set of the female circuit and the electromagnetic data embedded in the hardware Trojan circuit are used as a training set for model training,and then another set of the parent circuit and the electromagnetic data embedded in the hardware Trojan circuit are used as a test set.Test the classification and get the results of the classification.On this basis,this study selects and integrates the electromagnetic radiation information of the partially acquired female circuit and the circuit embedded in the hardware Trojan,and uses the integrated electromagnetic radiation information as a new training set and test set,and then performs test classification.The classification results are obtained,which further validates the feasibility and effectiveness of the hardware Trojan detection method based on convolutional neural network.The experimental results show that the hardware Trojan detection method described in this paper can effectively detect the hardware Trojan which accounts for 0.16% of the parent circuit area.