Terahertz Technology with AI-Powered Expert-System for Ensuring Hardware Security & the Reliability in Monolithic and Bilithic VLSI Chips for Bio-Medical Image Processing Applications

Abstract

In this paper, we provide a concise overview of a novel approach that combines Terahertz technology with an AI dependent expert system to safeguard the hardware securities & the reliabilities of monolithic and bilithic VLSI based IC chips. Ensuring the securities of hardware components has become a critical concern, particularly for intricate integrated circuits. Defective integrated circuits pose a significant threat to system security and dependability. With the growing complexity of digital signal systems, the task of developing effective methods for analyzing and authenticating the trustworthiness and legitimacy of integrated circuits has become more challenging yet indispensable. We propose a new terahertz-based inspection process for non-destructive and subtle identification of counterfeit or malfunctioning integrated circuits. This process involves sensing the response of the circuit to incident terahertz radiation and analyzing this response using artificial intelligence techniques. Terahertz technology is employed for its ability to provide a unique fingerprint feature that can distinguish between genuine and counterfeit integrated circuits. To facilitate the training of the convolutional neural network (CNN) model, a secure image dataset is created using data augmentation techniques. Additionally, an insecure image dataset is generated by modifying the original image data to represent altered integrated circuits. Remarkably, the trained models achieved a success rate of 90% in correctly identifying secure devices..