Globalization trends in integrated circuit (IC) design are leading to increased vulnerability of ICs against hardware Trojans (HT). Recently, several side channel parameters based techniques have been developed detect these that require golden as a reference model, but due the widespread usage IPs, most system-on-chip (SoC) do not reference. Hardware intellectual property (IP)-based SoC designs considered major concern for future circuits. Most state-of-the-art runtime Trojan detection...

Autonomous vehicles (AVs) rely heavily on cameras and artificial intelligence (AI) to make safe accurate driving decisions. However, since AI is the core enabling technology, this raises serious cyber threats that hinder large-scale adoption of AVs. Therefore, it becomes crucial analyze resilience AV security systems against sophisticated attacks manipulate camera inputs, deceiving models. In paper, we develop camera-camouflaged adversarial targeting traffic sign recognition (TSR) in...

Globalization trends in integrated circuit (IC) design using deep submicron (DSM) technologies are leading to increased vulnerability of ICs against malicious intrusions. These intrusions referred as hardware Trojans. One way address this threat is utilize unique electrical signatures ICs. However, technique requires analyzing extensive sensor data detect the intruded circuits. In order overcome limitation, we propose combine signature extraction mechanism with machine learning algorithms...

The success of deep learning has fast paced the evolution current technology at unprecedented rate. In particular, convolutional neural networks (CNNs) gained a lot attention due to their extraordinary performance in wide range computer vision applications. While CNNs been excellent, implementation complexity has, however, always posed challenge computational and memory access intensive nature especially for resource constrained embedded platforms. this paper, we propose novel...

Leveraging deep convolutional neural networks (DCNNs) for various application areas has become a recent inclination of many machine learning practitioners due to their impressive performance. Research trends show that the state-of-the-art are getting deeper and such have shown significant performance increase. Deeper larger imply increase in computational intensity memory footprint. This is particularly problem inference-based applications on resource constrained computing platforms. On...

Internet of Things (IoT) devices have connected millions houses around the globe via internet. In recent past, threats due to hardware Trojan (HT) in integrated circuits (IC) become a serious concern, which affects IoT edge (IoT-ED). this paper, possibility IoT-ED with embedded HT that can cause security, privacy, and availability problems based Home Area Network (HAN) has been discussed. Conventional network attack detection techniques work at protocol layers, whereas lead peculiar...

Photoplethysomographic (PPG) signal is crucial for non-invasive monitoring of heart rate. It acquired by using pulse oximeter that are prone to artifacts. A major application this technique the rate during physical exertion. Extraction (HR) from PPG in case difficult due strong motion related This paper proposes an efficient method based on a reference generation singular value decomposition and then multistage filtered X-LMS removing artifacts signal. Simultaneous three-axis acceleration...

Hardware security is a major concern in the intellectual property (IP) centric integrated circuits (IC). 3-D IC design augments IP designs. However, ICs suffer from high temperatures their middle tiers due to long heat dissipation paths. We anticipate that this problem would exacerbate hardware issues ICs. Because, temperature leads undesired timing characteristics In paper we provide detailed analysis on how these delay variations can lead non-ideal behavior of control It demonstrated...

Additive manufacturing (AM), also known as three-dimensional (3D) printing, is the process of building a solid object in layer-wise manner. Cybersecurity prevalent issue that appears more and frequently AM becomes popular. This paper focuses on effect fan speed printing quality presents plugin called Fan Speed Attack Detection (FSAD) predicts monitors speeds throughout process. The goal to prevent cybersecurity attacks, specifically targeting speed. Using proposed FSAD, any changes during...

An investigation of the Single Event Transient (SET) characteristics (amplitude and width) variation while propagating through static True Phase Clock (TSPC) logic is presented. The dependencies SET on input patterns, propagation paths, pulse polarity, diverging re-converging paths are investigated. New insights induced broadening (PIPB) phenomenon in different combinations TSPC reported. worst best for attenuation identified. Our results demonstrate that pulses can lead to Byzantine faults...

Traditional learning-based approaches for run-time Hardware Trojan detection require complex and expensive on-chip data acquisition frameworks thus incur high area power overhead. To address these challenges, we propose to leverage the correlation between executing instructions of a microprocessor establish machine (HT) framework, called MacLeR. reduce overhead acquisition, single power-port current block using sensors in time-division multiplexing, which increases accuracy while incurring...

With significant potential improvement in device-to-device (D2D) communication due to improved wireless link capacity (e.g., 5G and NextG systems), a collaboration of multiple edge devices (called horizontal (HC)) is becoming reality for real-time Edge Intelligence (EI). The distributed nature HC offers an advantage against traditional adversarial attacks because the adversary does not have access entire deep learning architecture (DLA). Due involvement untrusted environment, possibility...

It is a consensus among the researchers, although not proven, that it close to impossible guarantee completely secure hardware design. Therefore, desired have run-time Trojan detection techniques. This paper toward developing framework of how achieve units. Although difficult predict stage circuit design at which intruder would insert as well methodology should be applied, behavior patterns certain units in can indicate malicious activities We propose translate such using formal verification...

Glitches due to soft errors can act as a severe deterrent asynchronous circuit operations. To mitigate in quasi delay insensitive (QDI) circuits, built-in error correction NULL convention logic (NCL) has been introduced [9]. However, this technique cannot detect during the phase of NCL pipeline, and also avoid propagation into pipeline after its detection. This paper provides modified approach overcome these limitations with, on average, comparable power latency costs. work analyzes...

Programmable logic controllers (PLCs) are heavy-duty computers used to control industrial systems. For many years these systems were physically separated from any other network making attacks extremely difficult. However, increasingly connected have not improved much in terms of security, leaving them vulnerable attacks. This paper attempts show that ladder programs for PLCs can be modeled NuSMV and verified using computational tree (CTL) specifications. also shows how simple changes the...