Hardware Trojan horse detection system and information data processing method based on unsupervised learning

Abstract

The invention belongs to the technical field of hardware security, and discloses a hardware Trojan horse detection system and an information data processing method based on unsupervised learning, and the method includes analyzing a circuit structure and Trojan horse circuit operation logic, and proposing features required by Trojan horse detection; analzying the importance degrees of the features by combining a random forest, a correlation matrix and a parallel coordinate graph, the features are screened, and obtaining an optimal feature set; performing dimensionality reduction on the high-dimensional data features by adopting a principal component analysis (PCA) method; training an Isolation Forest unsupervised model by adopting the data subjected to dimension reduction to obtain an optimal training model; and testing by adopting test data, calculating parameters such as accuracy according to a test result, and evaluating the model. According to the invention, the majority of information of the data is reserved while the data dimension is reduced, the accuracy is effectively improved, the training time is shortened, and the problem that the label value in the hardware Trojan horse detection field is not easy to obtain or even cannot be obtained is solved by using an unsupervised learning method.

Description

technical field [0001] The invention belongs to the technical field of hardware security, and in particular relates to a hardware Trojan horse detection system and an information data processing method based on unsupervised learning. Background technique [0002] At present, with the rapid development of today's information society and the accelerated application of artificial intelligence technology, people's demand for integrated circuit chips is increasing day by day. However, due to the complexity of the chip design and manufacturing process, chip manufacturers cannot achieve complete autonomy and control over each link, which provides the possibility for some attackers to maliciously modify and destroy integrated circuits. This defect module, which is intentionally manufactured by attackers and inserted into the chip, changes the function or performance of the chip, and is triggered under certain special conditions is called a hardware Trojan. Hardware Trojan horses wi...

AI Technical Summary

Problems solved by technology

[0006] (1) The hardware Trojan horse does not contain any state information, the malicious attacker completely controls the hardware Trojan trigger, and implants various types of hardware Trojan horses, which is difficult to detect by traditional verification techniques

[0007] (2) The SoC in circulation is a complex heterogeneous system composed of multiple third-party IP cores. Due to the small size and concealment of hardware Trojans, it is difficult for the Trojan detection technology in third-party IP cores to completely distinguish Trojan networks. Malicious third-party suppliers even collude to manufacture hardware Trojans to evade detection

[0008] (3) Existing methods based on machine learning theory are all supervised learning methods that require a large amount of known information; the training process of supervised learning methods is time-consuming and usually requires a large amount of balanced training data

[0017] (3) A feature dimension reduction method is provided, which effectively reduces the feature dimension and retains more than 99% of the data information, and solves the problems of high algorithm complexity, long detection time, and detection problems caused by the high circuit feature dimension of the previous machine learning methods. Poor accuracy and other issues

Images