Showing papers on "Trojan published in 2022"

Trojan Nanobacteria System for Photothermal Programmable Destruction of Deep Tumor Tissues.

Abstract: Herein, we develop a novel bacteria-based drug delivery system, termed as the Trojan nanobacteria system, which is made of nanoagents internalized into engineered bacteria through bacteria-specific maltodextrin (MD) transporters. Compared with the system of attaching nanoagents to bacterial surfaces, the presented Trojan system features higher payloads and better stability. In cancer therapy, Trojan nanobacteria can specifically discriminate the tumor region and then penetrate deep tumor tissues. Afterwards, Trojan nanobacteria systems are able to destroy deep tumor tissues, which is due to the synergistic effects of antitumor protein ( e.g., tumor necrosis factor-α, TNF-α) expression and photothermal properties.

Detecting Hardware Trojans in PCBs Using Side Channel Loopbacks

Abstract: Malicious modifications to printed circuit boards (PCBs) are known as hardware Trojans. These may arise when malafide third parties alter PCBs premanufacturing or postmanufacturing and are a concern in safety-critical applications, such as industrial control systems. In this research, we examine how data-driven detection can be utilized to detect such Trojans at run-time. We develop a flexible and reconfigurable PCB test bed derived from the popular open-source programmable logic controller (PLC) platform “OpenPLC.” We then develop a Trojan detection framework, which utilizes and analyzes multimodal side channels (e.g., timing, magnetic signals, power, and hardware performance counters). We consider defender-configurable input/output (I/O) loopback test, comparison with design-document baselines, and magnetometer-aided monitoring of system behavior under defender-chosen excitations. Our approach can extend to golden-free environments. Golden (known-good) versions of the PCBs are assumed not available, but design information, datasheets, and component-level data are available. We demonstrate the efficacy of our approach on a range of Trojans instantiated in the test bed.

FLAW3D: A Trojan-Based Cyber Attack on the Physical Outcomes of Additive Manufacturing

Abstract: Additive manufacturing (AM) systems such as 3-D printers use inexpensive microcontrollers that rarely feature cybersecurity defenses. This is a risk, especially given the rising threat landscape within the larger digital manufacturing domain. In this work, we demonstrate this risk by presenting the design and study of a malicious Trojan (the FLAW3D bootloader) for AVR-based Marlin-compatible 3-D printers ( $>$ 100 commercial models). We show that the Trojan can hide from programming tools, and even within tight design constraints (less than 1.7 KB in size), it can compromise the quality of additively manufactured prints and reduce tensile strengths by up to 50%.

Detecting Hardware Trojans Using Combined Self-Testing and Imaging

Abstract: Hardware Trojans are malicious modifications in integrated circuits (ICs) with an intent to breach security and compromise the reliability of an electronic system. This article proposes a framework using self-testing, advanced imaging, and image processing with machine learning to detect hardware Trojans inserted by untrusted foundries. It includes on-chip test structures with negligible power, delay, and silicon area overheads. The core step of the framework is on-chip golden circuit design, which can provide authentic samples for image-based Trojan detection through self-testing. This core step enables a golden-chip-free Trojan detection that does not rely on an existing image data set from Trojan-free chip or image synthesizing. We have conducted an in-depth analysis of detection steps and discussed possible attacks with countermeasures to strengthen this framework. The performance evaluation on a 28-nm FPGA and a 90-nm IC validates its high accuracy and reliability for practical applications.

Hardware Trojan Insertion Using Reinforcement Learning

Abstract: This paper utilizes Reinforcement Learning (RL) as a means to automate the Hardware Trojan (HT) insertion process to eliminate the inherent human biases that limit the development of robust HT detection methods. An RL agent explores the design space and finds circuit locations that are best for keeping inserted HTs hidden. To achieve this, a digital circuit is converted to an environment in which an RL agent inserts HTs such that the cumulative reward is maximized. Our toolset can insert combinational HTs into the ISCAS-85 benchmark suite with variations in HT size and triggering conditions. Experimental results show that the toolset achieves high input coverage rates (100% in two benchmark circuits) that confirms its effectiveness. Also, the inserted HTs have shown a minimal footprint and rare activation probability.

Orbital stability analysis and photometric characterization of the second Earth Trojan asteroid 2020 XL5

Abstract: Trojan asteroids are small bodies orbiting around the L4 or L5 Lagrangian points of a Sun-planet system. Due to their peculiar orbits, they provide key constraints to the Solar System evolution models. Despite numerous dedicated observational efforts in the last decade, asteroid 2010 TK7 has been the only known Earth Trojan thus far. Here we confirm that the recently discovered 2020 XL5 is the second transient Earth Trojan known. To study its orbit, we used archival data from 2012 to 2019 and observed the object in 2021 from three ground-based observatories. Our study of its orbital stability shows that 2020 XL5 will remain in L4 for at least 4 000 years. With a photometric analysis we estimate its absolute magnitude to be [Formula: see text], and color indices suggestive of a C-complex taxonomy. Assuming an albedo of 0.06 ± 0.03, we obtain a diameter of 1.18 ± 0.08 km, larger than the first known Earth Trojan asteroid.

Design and Evaluation of a Multi-Domain Trojan Detection Method on Deep Neural Networks

Abstract: Trojan attacks on deep neural networks (DNNs) exploit a backdoor embedded in a DNN model that can hijack any input with an attacker’s chosen signature trigger. Emerging defence mechanisms are mainly designed and validated on vision domain tasks (e.g., image classification) on 2D Convolutional Neural Network (CNN) model architectures; a defence mechanism that is general across vision, text, and audio domain tasks is demanded. This work designs and evaluates a run-time Trojan detection method exploiting STR ong I ntentional P erturbation of inputs that is a multi-domain input-agnostic Trojan detection defence across Vi sion, T ext and A udio domains—thus termed as STRIP-ViTA. Specifically, STRIP-ViTA is demonstratively independent of not only task domain but also model architectures. Most importantly, unlike other detection mechanisms, it requires neither machine learning expertise nor expensive computational resource, which are the reason behind DNN model outsourcing scenario—one main attack surface of Trojan attack. We have extensively evaluated the performance of STRIP-ViTA over: i) CIFAR10 and GTSRB datasets using 2D CNNs for vision tasks; ii) IMDB and consumer complaint datasets using both LSTM and 1D CNNs for text tasks; and iii) speech command dataset using both 1D CNNs and 2D CNNs for audio tasks. Experimental results based on more than 30 tested Trojaned models (including publicly Trojaned model) corroborate that STRIP-ViTA performs well across all nine architectures and five datasets. Overall, STRIP-ViTA can effectively detect trigger inputs with small false acceptance rate (FAR) with an acceptable preset false rejection rate (FRR). In particular, for vision tasks, we can always achieve a 0 percent FRR and FAR given strong attack success rate always preferred by the attacker. By setting FRR to be 3 percent, average FAR of 1.1 and 3.55 percent are achieved for text and audio tasks, respectively. Moreover, we have evaluated STRIP-ViTA against a number of advanced backdoor attacks and compare its effectiveness with other recent state-of-the-arts.

Hardware Trojan Detection using Graph Neural Networks

Abstract: The globalization of the Integrated Circuit (IC) supply chain has moved most of the design, fabrication, and testing process from a single trusted entity to various untrusted third party entities around the world. The risk of using untrusted third-Party Intellectual Property (3PIP) is the possibility for adversaries to insert malicious modifications known as Hardware Trojans (HTs). These HTs can compromise the integrity, deteriorate the performance, and deny the functionality of the intended design. Various HT detection methods have been proposed in the literature; however, many fall short due to their reliance on a golden reference circuit, a limited detection scope, the need for manual code review, or the inability to scale with large modern designs. We propose a novel golden reference-free HT detection method for both Register Transfer Level (RTL) and gate-level netlists by leveraging Graph Neural Networks (GNNs) to learn the behavior of the circuit through a Data Flow Graph (DFG) representation of the hardware design. We evaluate our model on a custom dataset by expanding the Trusthub HT benchmarks trusthub1. The results demonstrate that our approach detects unknown HTs with 97% recall (true positive rate) very fast in 21.1ms for RTL and 84% recall in 13.42s for Gate-Level Netlist.

Golden-Free Hardware Trojan Detection Using Self-Referencing

Abstract: The globalization of the semiconductor supply chain has developed a new set of challenges for security researchers. Among them, malicious alterations of hardware designs at an untrusted facility, or Trojan insertion, are considered one of the most difficult challenges. While side-channel analysis-based hardware Trojan detection techniques have shown great potential, most solutions, proposed over the past decade, require the availability of golden (i.e., Trojan-free) chips and are susceptible to process variations. Few techniques that do not require a golden chip depend on simulation-based modeling of the side-channel signature, which may not be reliable for differentiating between process and Trojan induced variations. Furthermore, most of these techniques are evaluated either using very few Trojan inserted chips or simulation-based test setup. Spatial and temporal self-referencing-based detection mechanisms proposed earlier effectively eliminate the need for a golden chip and the impact of process variations. However, these techniques have not been adequately studied to achieve high detection sensitivity. In this article, we propose a golden-free multidimensional self-referencing technique that analyzes the side-channel signatures in both the time and frequency domains to significantly broaden the Trojan coverage and strengthen the detection confidence. We introduce a fully automated detection framework containing systematic methodologies for test generation, signature extraction, signal processing, threshold calculation, and metric-based decision-making that effectively enables the synergistic self-referencing approach. Finally, we evaluate the proposed technique through a comprehensive hardware measurement setup consisting of 96 Trojan-inserted test chips. Along with achieving a high detection coverage, we demonstrate that the analysis of spatial and temporal discrepancies in both frequency and time domains helps to reliably detect small hard-to-detect Trojans under process and measurement induced variations.

TrojanDetector: A Multi-Layer Hybrid Approach for Trojan Detection in Android Applications

Abstract: Trojan Detection—the process of understanding the behaviour of a suspicious file has been the talk of the town these days. Existing approaches, e.g., signature-based, have not been able to classify them accurately as Trojans. This paper proposes TrojanDetector—a simple yet effective multi-layer hybrid approach for Trojan detection. TrojanDetector analyses every downloaded application and extracts and correlates its features on three layers (i.e., application-, user-, and package layer) to identify it as either a benign application or a Trojan. TrojanDetector adopts a hybrid approach, combining static and dynamic analysis characteristics, for feature extraction from any downloaded application. We have evaluated our scheme on three publicly available datasets, namely (i) CCCS- CIC-AndMal-2020, (ii) Cantagio-Mobile, and (iii) Virus share, by using simple yet state-of-the-art classifiers, namely, random forest (RF), decision tree (DT), support vector machine (SVM), and logistic regression (LR) in binary—class settings. SVM outperformed its counterparts and attained the highest accuracy of 96.64%. Extensive experimentation shows the effectiveness of our proposed Trojan detection scheme.

Hardware Trojan Detection at Behavioral Level Using Inline Assertions and Verification Using UVM

Abstract: Recently, hardware Trojan (HT) is posing a significant challenge to the integrated circuit (IC) industry and has inspired various improvements in the Trojan identification plans. This research study presents the inline assertions for the detection of hardware Trojan at the behavioral level of a system on chip (SoC). In the proposed RTL design, a modified circuit design flow is suggested to incorporate inline assertions into a SoC. Flexible inline assertions are developed in the RTL block within the design module. The router IP design and inline assertions are synthesized and implemented in Xilinx Vivado and Aldec Rivera Pro using Verilog HDL. The universal verification methodology (UVM) is also used to verify the proposed design with the different test case scenarios. The functional coverage and code coverage are analyzed in Aldec Rivera Pro. Parameters such as power and area are analyzed in the Synopsys design compiler (DC).

Introduction to The Data Mining Techniques in Cybersecurity

Abstract: As a result of the evolution of the Internet and the massive amount of data that is transmitted every second, as well as the methods for protecting and preserving it and distinguishing those who are authorized to view it, the role of cyber security has evolved to provide the best protection for information over the network. In this paper, the researcher discusses the role of data mining methods in cyber security. Data mining has several uses in security, including national security (for example, surveillance) and cyber security (e.g., virus detection). Attacks against buildings and the destruction of key infrastructure, such as power grids and telecommunications networks, are examples of national security concerns. Cybersecurity is concerned with safeguarding computer and network systems from harmful malware such as Trojan horses and viruses. In addition, data mining is being used to deliver solutions such as intrusion detection and auditing.

Hardware Trojan Detection Using Unsupervised Deep Learning on Quantum Diamond Microscope Magnetic Field Images

Abstract: This article presents a method for hardware trojan detection in integrated circuits. Unsupervised deep learning is used to classify wide field-of-view (4 × 4 mm2), high spatial resolution magnetic field images taken using a Quantum Diamond Microscope (QDM). QDM magnetic imaging is enhanced using quantum control techniques and improved diamond material to increase magnetic field sensitivity by a factor of 4 and measurement speed by a factor of 16 over previous demonstrations. These upgrades facilitate the first demonstration of QDM magnetic field measurement for hardware trojan detection. Unsupervised convolutional neural networks and clustering are used to infer trojan presence from unlabeled data sets of 600 × 600 pixel magnetic field images without human bias. This analysis is shown to be more accurate than principal component analysis for distinguishing between field programmable gate arrays configured with trojan-free and trojan-inserted logic. This framework is tested on a set of scalable trojans that we developed and measured with the QDM. Scalable and TrustHub trojans are detectable down to a minimum trojan trigger size of 0.5% of the total logic. The trojan detection framework can be used for golden-chip-free detection, since knowledge of the chips’ identities is only used to evaluate detection accuracy.

Research and Development Priorities for Security of Embedded Hardware Devices

Abstract: Supply chains for embedded hardware devices are subject to a variety of security threats that are concerning to governments and industry. Such threats include counterfeit parts, cyber-attacks, Trojan attacks, reliance on offshore-manufactured and controlled hardware, software, and rapid increase of network technologies. Addressing these threats is vital to maintain and enhance system trust and security. This article develops a framework to address initiatives and success criteria which accounts for different perspectives of system stakeholders and identify different potential emergent conditions and scenarios which are most and least impactful to the trust and security of the supply chain. The article demonstrates a research and development priorities preferences method that accounts for evolving perspectives of supply chain stakeholders and describing the example application of embedded hardware devices as a system.

BppAttack: Stealthy and Efficient Trojan Attacks against Deep Neural Networks via Image Quantization and Contrastive Adversarial Learning

Abstract: Deep neural networks are vulnerable to Trojan attacks. Existing attacks use visible patterns (e.g., a patch or image transformations) as triggers, which are vulnerable to human inspection. In this paper, we propose stealthy and efficient Trojan attacks, BppAttack. Based on existing biology literature on human visual systems, we propose to use image quantization and dithering as the Trojan trigger, making imperceptible changes. It is a stealthy and efficient attack without training auxiliary models. Due to the small changes made to images, it is hard to inject such triggers during training. To alleviate this problem, we propose a contrastive learning based approach that leverages adversarial attacks to generate negative sample pairs so that the learned trigger is precise and accurate. The proposed method achieves high attack success rates on four benchmark datasets, including MNIST, CIFAR-10, GTSRB, and CelebA. It also effectively bypasses existing Trojan defenses and human inspection. Our code can be found in https://github.com/RU-System-Software-and-Security/BppAttack.

Golden Reference-Free Hardware Trojan Localization Using Graph Convolutional Network

Abstract: The globalization of the integrated circuit (IC) supply chain has moved most of the design, fabrication, and testing process from a single trusted entity to various untrusted third-party entities worldwide. The risk of using untrusted third-Party Intellectual Property (3PIP) is the possibility for adversaries to insert malicious modifications known as Hardware Trojans (HTs). These HTs can compromise the integrity, deteriorate the performance, deny the service, and alter the functionality of the design. While numerous HT detection methods have been proposed in the literature, the crucial task of HT localization is overlooked. Moreover, a few existing HT localization methods have several weaknesses: reliance on a golden reference, inability to generalize for all types of HT, lack of scalability, low localization resolution, and manual feature engineering/property definition. To overcome their shortcomings, we propose a novel, golden reference-free HT localization method at the pre-silicon stage by leveraging graph convolutional network (GCN). In this work, we convert the circuit design into its intrinsic data structure, graph, and extract the node attributes. Afterward, the graph convolution performs automatic feature extraction for nodes to classify the nodes as Trojan or benign. Our approach is automated and does not burden the designer with manual code review. It locates the Trojan signals with 99.6% accuracy, 93.1% $F1$ -score, and a false-positive rate below 0.009%.

Towards Effective and Robust Neural Trojan Defenses via Input Filtering

Abstract: Trojan attacks on deep neural networks are both dangerous and surreptitious. Over the past few years, Trojan attacks have advanced from using only a single input-agnostic trigger and targeting only one class to using multiple, input-specific triggers and targeting multiple classes. However, Trojan defenses have not caught up with this development. Most defense methods still make inadequate assumptions about Trojan triggers and target classes, thus, can be easily circumvented by modern Trojan attacks. To deal with this problem, we propose two novel “filtering” defenses called Variational Input Filtering (VIF) and Adversarial Input Filtering (AIF) which leverage lossy data compression and adversarial learning respectively to effectively purify potential Trojan triggers in the input at run time without making assumptions about the number of triggers/target classes or the input dependence property of triggers. In addition, we introduce a new defense mechanism called “Filtering-then-Contrasting” (FtC) which helps avoid the drop in classification accuracy on clean data caused by “filtering”, and combine it with VIF/AIF to derive new defenses of this kind. Extensive experimental results and ablation studies show that our proposed defenses significantly outperform well-known baseline defenses in mitigating five advanced Trojan attacks including two recent state-of-the-art while being quite robust to small amounts of training data and large-norm triggers.

An Efficient Framework with Node Filtering and Load Expansion for Machine-Learning-Based Hardware Trojan Detection

Abstract: The globalization of the integrated circuit (IC) industry has raised concerns about hardware Trojans (HT), and there is an urgent need for efficient HT-detection methods of gate-level netlists. Machine learning (ML) is a powerful tool for this purpose. A Trojan-detection framework is proposed in this paper to solve the data imbalance and low accuracy problems of existing ML-based HT-detection algorithms. To solve the problem of data imbalance, we propose the node-filtering algorithm, which extracts structure templates from HT circuits and removes most normal nodes based on them. To enhance the identification of unknown HT payload, we propose the load-expansion algorithm, which expands the identified HT nodes based on their fanout features. We evaluate the framework using different ML algorithms. The results show that the framework significantly improves the Trojan-detection rate of the original algorithms, and achieves a 10% improvement in true positive rate compared to the original algorithms.

Size Distribution of Small Jupiter Trojans in the L5 Swarm

Abstract: We present an analysis of survey observations of the trailing L5 Jupiter Trojan swarm using the wide-field Hyper Suprime-Cam CCD camera on the 8.2 m Subaru Telescope. We detected 189 L5 Trojans from our survey that covered about 15 deg2 of sky with a detection limit of m r = 24.1 mag, and selected an unbiased sample consisting of 87 objects with absolute magnitude 14 ≲ H r ≤ 17 corresponding to diameter 2 km ≲ D ≲ 10 km for analysis of size distribution. We fit their differential magnitude distribution to a single-slope power law with an index α = 0.37 ± 0.01, which corresponds to a cumulative size distribution with an index of b = 1.85 ± 0.05. Combining our results with data for known asteroids, we obtained the size distribution of L5 Jupiter Trojans over the entire size range for 9 ≲ H V ≤ 17, and found that the size distributions of the L4 and L5 swarms agree well with each other for a wide range of sizes. This is consistent with the scenario that asteroids in the two swarms originated from the same primordial population. Based on the above results, the ratio of the total number of asteroids with D ≥ 2 km in the two swarms is estimated to be N L4/N L5 = 1.40 ± 0.15, and the total number of L5 Jupiter Trojans with D ≥ 1 km is estimated to be 1.1 × 105 by extrapolating the obtained distribution.

ATTRITION: Attacking Static Hardware Trojan Detection Techniques Using Reinforcement Learning

Abstract: Stealthy hardware Trojans (HTs) inserted during the fabrication of integrated circuits can bypass the security of critical infrastructures. Although researchers have proposed many techniques to detect HTs, several critical limitations exist, including: (i) a low success rate of HT detection, (ii) high algorithmic complexity, and (iii) a large number of test patterns. Furthermore, as we show in this work the most pertinent drawback of prior (including state-of-the-art) detection techniques stems from an incorrect evaluation methodology, i.e., they assume that an adversary inserts HTs randomly. Such inappropriate adversarial assumptions enable detection techniques to claim high HT detection accuracy, leading to a "false sense of security." To the best of our knowledge, despite more than a decade of research on detecting HTs inserted during fabrication, there have been no concerted efforts to perform a systematic evaluation of HT detection techniques. In this paper, we play the role of a realistic adversary and question the efficacy of HT detection techniques by developing an automated, scalable, and practical attack framework, ATTRITION, using reinforcement learning (RL). ATTRITION evades eight detection techniques (published in premier security venues, well-cited in academia, etc.) across two HT detection categories, showcasing its agnostic behavior. ATTRITION achieves average attack success rates of 47x and 211x compared to randomly inserted HTs against state-of-the-art logic testing and side channel techniques. To demonstrate ATTRITION's ability in evading detection techniques, we evaluate different designs ranging from the widely-used academic suites (ISCAS-85, ISCAS-89) to larger designs such as the open-source MIPS and mor1kx processors to AES and a GPS module. Additionally, we showcase the impact of ATTRITION generated HTs through two case studies (privilege escalation and kill switch) on mor1kx processor. We envision that our work, along with our released HT benchmarks and models fosters the development of better HT detection techniques.

On Large and Small Data Blow-Up Solutions in the Trojan Y Chromosome Model

Abstract: The Trojan Y Chromosome Strategy (TYC) is the only genetic biological control method in practice in North America for controlling invasive populations with an XX–XY sex determinism. Herein a modified organism, that is a supermale or feminised supermale, is introduced into an invasive population to skew the sex ratio over time, causing local extinction. We consider the three species TYC reaction diffusion model, and show that introduction of supermales above certain thresholds, and for certain initial data, solutions can blow-up in finite time. Thus, in order to have biologically meaningful solutions, one needs to restrict parameter and initial data regimes, in TYC type models.

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

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

Securing Hardware Development Process using Blockchain

Abstract: Modern microprocessors contain millions of gates and finding a small hidden malicious hardware trojan (HT) is difficult. Additionally, these HTs may not need any additional external input pins to activate. Many solutions have been proposed to find these HTs, but none of the solutions gives promising result due to their limitations. Moreover, pre-silicon verification and post-silicon testing also don't address the issue of HTs. In this paper we present methodology to limit the possibily of inserting HTs based on blockchain technology. It is better to limit the opportunity for insertion of HT than finding them afterwards in the design and fabrication process. We proposed a solution for monitoring hardware development process files to maintain integrity and trustful relationship using encryption and smart contracts in a blockchain network.

A security method of hardware Trojan detection using path tracking algorithm

Abstract: Abstract Recently, the issue of malicious circuit alteration and attack draws more attention than ever before due to the globalization of IC design and manufacturing. Malicious circuits, also known as hardware Trojans, are found able to degrade the circuit performance or even leak confidential information, and accordingly it is definitely an issue of immediate concern to develop detection techniques against hardware Trojans. This paper presents a ring oscillator-based detection technique to improve the hardware Trojan detection performance. A circuit under test is divided into a great number of blocks, path assignment is optimized using a path tracking algorithm, and a high coverage is reached accordingly.

Design of AI Trojans for Evading Machine Learning-based Detection of Hardware Trojans

Abstract: The globalized semiconductor supply chain significantly increases the risk of exposing System-on-Chip (SoC) designs to malicious implants, popularly known as hardware Trojans. Traditional simulation-based validation is unsuitable for detection of carefully-crafted hardware Trojans with extremely rare trigger conditions. While machine learning (ML) based Trojan detection approaches are promising due to their scalability as well as detection accuracy, ML-based methods themselves are vulnerable from Trojan attacks. In this paper, we propose a robust backdoor attack on ML-based Trojan detection algorithms to demonstrate this serious vulnerability. The proposed framework is able to design an AI Trojan and implant it inside the ML model that can be triggered by specific inputs. Experimental results demonstrate that the proposed AI Trojans can bypass state-of-the-art defense algorithms. Moreover, our approach provides a fast and cost-effective solution in achieving 100% attack success rate that significantly outperforms state-of-the art approaches based on adversarial attacks.

Test-Time Detection of Backdoor Triggers for Poisoned Deep Neural Networks

Abstract: Backdoor (Trojan) attacks are emerging threats against deep neural networks (DNN). A DNN being attacked will predict to an attacker-desired target class whenever a test sample from any source class is embedded with a backdoor pattern; while correctly classifying clean (attack-free) test samples. Existing backdoor defenses have shown success in detecting whether a DNN is attacked and in reverse-engineering the backdoor pattern in a "post-training" regime: the defender has access to the DNN to be inspected and a small, clean dataset collected independently, but has no access to the (possibly poisoned) training set of the DNN. However, these defenses neither catch culprits in the act of triggering the backdoor mapping, nor mitigate the backdoor attack at test-time. In this paper, we propose an "in-flight" defense against backdoor attacks on image classification that 1) detects use of a backdoor trigger at test-time; and 2) infers the class of origin (source class) for a detected trigger example. The effectiveness of our defense is demonstrated experimentally against different strong backdoor attacks.

TrojanZoo: Towards Unified, Holistic, and Practical Evaluation of Neural Backdoors

Abstract: Neural backdoors represent one primary threat to the security of deep learning systems. The intensive research has produced a plethora of backdoor attacks/defenses, resulting in a constant arms race. However, due to the lack of evaluation benchmarks, many critical questions remain under-explored: (i) what are the strengths and limitations of different attacks/defenses? (ii) what are the best practices to operate them? and (iii) how can the existing attacks/defenses be further improved? To bridge this gap, we design and implement TROJAN-ZOO, the first open-source platform for evaluating neural backdoor attacks/defenses in a unified, holistic, and practical manner. Thus far, focusing on the computer vision domain, it has incorporated 8 representative attacks, 14 state-of-the-art defenses, 6 attack performance metrics, 10 defense utility metrics, as well as rich tools for in-depth analysis of the attack-defense interactions. Leveraging TROJANZOO, we conduct a systematic study on the existing attacks/defenses, unveiling their complex design spectrum: both manifest intricate trade-offs among multiple desiderata (e.g., the effectiveness, evasiveness, and transferability of attacks). We further explore improving the existing attacks/defenses, leading to a number of interesting findings: (i) one-pixel triggers often suffice; (ii) training from scratch often outperforms perturbing benign models to craft trojan models; (iii) optimizing triggers and trojan models jointly greatly improves both attack effectiveness and evasiveness; (iv) individual defenses can often be evaded by adaptive attacks; and (v) exploiting model interpretability significantly improves defense robustness. We envision that TROJANZOO will serve as a valuable platform to facilitate future research on neural backdoors.

A security method of hardware Trojan detection using path tracking algorithm

Abstract: Abstract Recently, the issue of malicious circuit alteration and attack draws more attention than ever before due to the globalization of IC design and manufacturing. Malicious circuits, also known as hardware Trojans, are found able to degrade the circuit performance or even leak confidential information, and accordingly it is definitely an issue of immediate concern to develop detection techniques against hardware Trojans. This paper presents a ring oscillator-based detection technique to improve the hardware Trojan detection performance. A circuit under test is divided into a great number of blocks, path assignment is optimized using a path tracking algorithm, and a high coverage is reached accordingly.