Deep Learning-Based Hardware Trojan Detection With Block-Based Netlist Information Extraction

With the globalization of semiconductor industry, hardware Trojans (HTs) are an emergent security threat in modern integrated circuit (IC) production. Research is now being conducted into designing more accurate and efficient methods to detect HTs. Recently, a number machine learning (ML)-based HT detection approaches have been proposed; however, most them still use knowledge-driven design features often engineering intuition carefully craft model improve accuracy. Therefore, this work, we propose data-driven system based on gate-level netlists. The consists four main parts: 1) Information extraction from netlist block; 2) Natural language processing (NLP) for translating information; 3) Deel (DL)-based model; 4) component final voter. In experiments, both long short-term memory networks (LSTM) convolutional neural network (CNN) used as our models. We performed experiments benchmarks Trust-hub K-fold crossing verification has applied evaluate different parameter settings training procedure. experimental results show that proposed can achieve 79.29% TPR, 99.97% TNR, 87.75% PPV 99.94% NPV combinational Trojan 93.46% 99.99% 98.92% 99.92% sequential after voting-based optimization using LEDA library-based ( logic_level =4, upsampling, LSTM, 5 epochs).