Threat Of Attacks Research Articles

Potential cyber threats of adversarial attacks on autonomous driving models

Autonomous Vehicles (CAVs) are currently seen as a viable alternative to traditional vehicles. However, CAVs will face serious cyber threats because many components of the driving system are based on machine learning models and are vulnerable to adversary attacks. We have reviewed the scientific literature and highlighted the main types of disruptive attacks on autonomous driving models that pose potential threats to CAVs. In this paper, we have compiled a dataset with traffic sign images obtained from public sources. We made experiments in which we distorted the original images and used them to train deep neural network-based classification models. The experiments demonstrated a possible threat to traffic sign recognition by autonomous vehicles. This work can give researchers and engineers a better understanding of the current state and trends in CAV security for their future use.

Ransomware Attacks Threat Modeling Using Bayesian Network

Ransomware is a dangerous malware that blocks access to data through encryption, and it exploits device vulnerabilities to perform chain attacks from one system to another. This study results in modeling the threat of ransomware attacks using Bayesian Network. The structure of the model is created using device vulnerabilities that can be exploited. As the basis for calculating the probability of the model, the EPSS vulnerability score is used. The risk exposure rating is calculated through the joint probability distribution formulation based on attack scenarios. Our model shows that ransomware attacks are most likely to exploit the chain of vulnerabilities CVE-2021-26855, CVE-2021-26857, CVE-2021-27065, CVE-2021-36942, and CVE-2017-0144 which has a probability value of 0.046534. In addition, the use of the EPSS also makes the risk assessment more factual, accurate, and effective. The threat modeling method can help in identifying ransomware attacks through a chain of vulnerabilities, making risk assessment more precise.

Robust Feature-Guided Generative Adversarial Network for Aerial Image Semantic Segmentation against Backdoor Attacks

Profiting from the powerful feature extraction and representation capabilities of deep learning (DL), aerial image semantic segmentation based on deep neural networks (DNNs) has achieved remarkable success in recent years. Nevertheless, the security and robustness of DNNs deserve attention when dealing with safety-critical earth observation tasks. As a typical attack pattern in adversarial machine learning (AML), backdoor attacks intend to embed hidden triggers in DNNs by poisoning training data. The attacked DNNs behave normally on benign samples, but when the hidden trigger is activated, its prediction is modified to a specified target label. In this article, we systematically assess the threat of backdoor attacks to aerial image semantic segmentation tasks. To defend against backdoor attacks and maintain better semantic segmentation accuracy, we construct a novel robust generative adversarial network (RFGAN). Motivated by the sensitivity of human visual systems to global and edge information in images, RFGAN designs the robust global feature extractor (RobGF) and the robust edge feature extractor (RobEF) that force DNNs to learn global and edge features. Then, RFGAN uses robust global and edge features as guidance to obtain benign samples by the constructed generator, and the discriminator to obtain semantic segmentation results. Our method is the first attempt to address the backdoor threat to aerial image semantic segmentation by constructing the robust DNNs model architecture. Extensive experiments on real-world scenes aerial image benchmark datasets demonstrate that the constructed RFGAN can effectively defend against backdoor attacks and achieve better semantic segmentation results compared with the existing state-of-the-art methods.

Deceptive Tricks in Artificial Intelligence: Adversarial Attacks in Ophthalmology.

The artificial intelligence (AI) systems used for diagnosing ophthalmic diseases have significantly progressed in recent years. The diagnosis of difficult eye conditions, such as cataracts, diabetic retinopathy, age-related macular degeneration, glaucoma, and retinopathy of prematurity, has become significantly less complicated as a result of the development of AI algorithms, which are currently on par with ophthalmologists in terms of their level of effectiveness. However, in the context of building AI systems for medical applications such as identifying eye diseases, addressing the challenges of safety and trustworthiness is paramount, including the emerging threat of adversarial attacks. Research has increasingly focused on understanding and mitigating these attacks, with numerous articles discussing this topic in recent years. As a starting point for our discussion, we used the paper by Ma et al. "Understanding Adversarial Attacks on Deep Learning Based Medical Image Analysis Systems". A literature review was performed for this study, which included a thorough search of open-access research papers using online sources (PubMed and Google). The research provides examples of unique attack strategies for medical images. Unfortunately, unique algorithms for attacks on the various ophthalmic image types have yet to be developed. It is a task that needs to be performed. As a result, it is necessary to build algorithms that validate the computation and explain the findings of artificial intelligence models. In this article, we focus on adversarial attacks, one of the most well-known attack methods, which provide evidence (i.e., adversarial examples) of the lack of resilience of decision models that do not include provable guarantees. Adversarial attacks have the potential to provide inaccurate findings in deep learning systems and can have catastrophic effects in the healthcare industry, such as healthcare financing fraud and wrong diagnosis.

A post-correlation graphical way for continuous GNSS spoofing detection

Global navigation satellite system (GNSS) receivers are under the threat of spoofing attacks. The signal quality monitoring (SQM) technique is effective in monitoring the distorted auto-correlation function (ACF). However, it has limited detection effectiveness in detecting spoofing signals that are greatly overpowered and with significant code phase shifts. This paper introduces a new metric called ACF similarity to characterize the ACF distortion and power abnormality in the tracking loop based on image feature extraction techniques. The spoofing detection algorithm and the optimal threshold utilizing the generalized likelihood ratio test (GLRT) are further developed. The performance of the proposed detector has been verified by utilizing the Texas Spoofing Test Battery (TEXBAT) dataset and Monte Carlo simulation tests. Results show that with 1% false alarm rate, the proposed detector achieves a detection probability of 87% in the TEXBAT datasets and provides better detection sensitivity than the conventional SQM metrics.

Privacy-Preserving Object Detection With Poisoning Recognition for Autonomous Vehicles

Object detection has achieved significant progress in attaining high-quality performance without leaking private messages. However, traditional approaches cannot defend the poisoning attacks. Poisoning attacks can make the predictive model unusable, which quickly causes recognition errors or even traffic accidents. In this paper, we propose a privacy-preserving object detection with poisoning recognition (PR-PPOD) framework via distributed training with the help of the CNN, ResNet18, and classical SSD network. Specifically, we design a poisoning model recognition algorithm to remove the uploaded local poisoning parameters to guarantee a trained model's availability based on given privacy-preserving progress. More importantly, the PR-PPOD framework can effectively prevent the threat of differential attacks and avoid privacy leakage caused by reverse model reasoning. Moreover, the effectiveness, efficiency, and security of PR-PPOD are demonstrated via comprehensive theoretical analysis. Finally, we simulate the performance of local poisoning model recognition based on the MNIST, CIFAR10, VOC2007, and VOC2012 datasets, which could achieve good performance compared with the case without poisoning recognition.

Combined Trauma and CBRN Exposure

Introduction:Whether it is COVID in New York, chemicals in Syria, or a radiological exposure from a dirty bomb or nuclear detonation, trauma is increasingly mixed with CBRN. Decontamination, isolation, and newer treatments all impact the patient as well as the physician. Military physicians regularly plan for these patients, but the civilian world has not. Civilian CBRN training has been spotty, at best. Our recent experience with COVID and the current threats of nuclear attack, emphasize the need to integrate CBRN exposure into our trauma treatment plans, both to help our patients and to protect the caregivers.Method:Literature review and US Department of Health and Human Services Guidelines.Results:A template for patient flow and treatment priorities is presented along with just-in-time references. Triage guidelines for trauma patients with Acute Radiation Syndrome (ARS) are also provided.Conclusion:Civilian preparation for managing the CBRN contaminated trauma patient starts with a mental model of treatment priorities and self-protection strategies. Just-in-time treatment references can provide additional support and confidence.

Blockchain Protocol-Based Secondary Predictive Secure Control for Voltage Restoration and Current Sharing of DC Microgrids

DC microgrids with distributed architectures inevitably encounter the threats of communication constraints and cyber attacks, and there is little research to address these two communication irregularities simultaneously. To provide DC microgrids with maximum resistance to communication constraints and cyber attacks, a distributed predictive secure control method based on blockchain protocol is proposed in this paper. The proposed control approach is the first attempt to address cyber attacks with a hybrid of data and models, improving the resilience and robustness of DC microgrids. With the help of practical Byzantine ideas, the proposed strategy preserves the resilience strengths of the blockchain against corruption and discards its shortcomings of low real-time performance, significantly enhancing the security without compromising the dynamic performance of the system by active intervention. Subsequently, a bound on the communication delay induced by the blockchain network under the expected security index is given based on the theoretical analysis. Unlike the existing practice of passively tolerating communication delays and packet dropouts, the predictive control method proposed in this paper actively compensates for them experienced by DC microgrids. Then, an analytical model of the closed-loop DC microgrid system is developed, and stability analysis is presented accordingly. Finally, several experimental tests performed on a PV-based DC microgrid hardware system are presented to demonstrate the effectiveness of the proposed control strategy.

Adversarial Attack Mitigation Strategy for Machine Learning-Based Network Attack Detection Model in Power System

The network attack detection model based on machine learning (ML) has received extensive attention and research in PMU measurement data protection of power systems. However, well-trained ML-based detection models are vulnerable to adversarial attacks. By adding meticulously designed perturbations to the original data, the attacker can significantly decrease the accuracy and reliability of the model, causing the control center to receive unreliable PMU measurement data. This paper takes the network attack detection model in the power system as a case study to analyze the vulnerability of the ML-based detection model under adversarial attacks. And then, a mitigation strategy for adversarial attacks based on causal theory is proposed, which can enhance the robustness of the detection model under different adversarial attack scenarios. Unlike adversarial training, this mitigation strategy does not require adversarial samples to train models, saving computing resources. Furthermore, the strategy only needs a small amount of detection model information and can be migrated to various models. Simulation experiments on the IEEE node systems verify the threat of adversarial attacks against different ML-based detection models and the effectiveness of the proposed mitigation strategy.

Pembelaan Terpaksa Melampaui Batas (Noodweer Exces) dalam Perkara Pidana Putusan Nomor 34/PID.B/2020/PN MII

This research examines Decision Number 34/Pid.B/2020/ PN MII in the case of loss of life of another person cannot be convicted because it is based on forced defense beyond the limit or noodweer exces. Noodweer exces is a defense when a person experiences an attack or threat of attack, the exceeding of the limits of a defense must be due to the influence of a great mental shock. The purpose of this study is to determine the applicability of noodweer exces in criminal case number 34/Pid.B/2020/ PN MII and to determine the criminal liability of the perpetrator in Decision Number 34/Pid.B/2020/ PN MII in terms of forced defense beyond the limit. The type of research used is normative juridical with a case approach method. The results showed that noodweer exces could not be applied because the shooting of the victim's head by the defendant Sumantri was not caused directly by severe mental shock. According to the Theory of necessary defense, the defense must be proportional to the nature of the attack. Because the element of necessity defense is not fulfilled and the defendant's actions fulfill the element of negligence, the defendant can be convicted as formulated in Article 359 of the Criminal Code.

앤드류 니콜의Good Kill에 나타난 드론 전쟁의 부당성에 대한 저항

Following the 9/11 attacks, the United States turned to drone warfare in its fight against terrorism, as drones proved to be both cost-effective and politically safer than manned fighter jets. The Obama administration, believing in the effectiveness and morality of drones, expanded and intensified drone operations compared to its predecessor. However, Andrew Niccol’s 2014 film, Good Kill, uncovers the injustices of the CIA-led drone strikes and the growing opposition against them. The movie highlights the psychological burden on drone pilots, who face an identity crisis due to their detachment from direct combat, and struggle to balance family life with the stress of participating in unequal warfare. Their families, unable to help alleviate the pilots’ psychological anxiety, instead experience crises brought on by the pilots’ stress.BR Additionally, the film emphasizes the undemocratic nature of CIA-led drone warfare, characterized by a lack of transparency and limited oversight by the public, media, and Congress. The CIA’s arbitrary application of signature strikes and threats of imminent attacks has led to civilian casualties, the division of communities, and the violation of international law. Drone strikes have been conducted within the territories of non-conflicting countries such as Pakistan and Yemen, undermining their sovereignty. These actions have fueled anti-American sentiment, support for extremist militants, and violent counterattacks against U.S. drone strikes, leading to a vicious cycle of violence and global instability.BR In response, the film’s protagonist, Tommy, sympathizes with his co-pilot, Suarez, and decides to sabotage the CIA’s operation arbitrarily applying the principles of signature strikes and proportionality. He makes an overt protest against drone warfare by leaving the military. Ultimately, Good Kill warns of the danger posed by drone warfare and calls for a ban on armed drones to prevent further proliferation and to protect the world from uncontrollable dangers.

TLR4-A Pertinent Player in Radiation-Induced Heart Disease?

The heart is one of the organs that is sensitive to developing delayed adverse effects of ionizing radiation (IR) exposure. Radiation-induced heart disease (RIHD) occurs in cancer patients and cancer survivors, as a side effect of radiation therapy of the chest, with manifestation several years post-radiotherapy. Moreover, the continued threat of nuclear bombs or terrorist attacks puts deployed military service members at risk of exposure to total or partial body irradiation. Individuals who survive acute injury from IR will experience delayed adverse effects that include fibrosis and chronic dysfunction of organ systems such as the heart within months to years after radiation exposure. Toll-like receptor 4 (TLR4) is an innate immune receptor that is implicated in several cardiovascular diseases. Studies in preclinical models have established the role of TLR4 as a driver of inflammation and associated cardiac fibrosis and dysfunction using transgenic models. This review explores the relevance of the TLR4 signaling pathway in radiation-induced inflammation and oxidative stress in acute as well as late effects on the heart tissue and the potential for the development of TLR4 inhibitors as a therapeutic target to treat or alleviate RIHD.

TAT: Targeted backdoor attacks against visual object tracking

Visual object tracking (VOT) is a fundamental computer vision task that aims to track a target in a sequence of video frames. It has been broadly adopted in safety- and security-critical applications, such as self-driving systems and traffic control systems. However, the VOT models (i.e., the trackers) that rely on third-party training resources face a severe threat of backdoor attacks, which refer to the type of the attacks that poison a portion of training data and mislead the tracker to track a wrong target. A surge of research interest has arisen in backdoor attacks in the domain of image classification, as a measure to expose the potential security risks of the classifiers and inspire new defense techniques. Despite the prosperity of the research in backdoor attacks in image classification, there still lacks investigation in backdoor attacks against VOT, due to their unique challenges: first, the architecture of a VOT model is much more complicated than that of an image classifier; second, VOT targets a sequence of video frames rather than individual images. To bridge the gap, we propose a novel and effective targeted backdoor attack approach TAT specifically against VOT tasks. In particular, TAT includes a basic version TAT-BA that can achieve effective and stealthy backdoor attacks against VOT trackers, and an advanced version TAT-DA that can evade two representative defense techniques. Our large-scale experimental evaluation demonstrates the effectiveness and the stealthiness of TAT. Moreover, we also demonstrate the performances of TAT-BA under real-world settings and the abilities of TAT-DA to counter defense techniques. The code will be available at https://github.com/MisakaZipi/TAT.

RELEVANCE OF CIVIL DEFENSE OF POPULATION OF RESIDENTIAL BUILDINGS IN THE CITIES OF UKRAINE

In this research highlight history of civil defense, actuality of providing of civil defense of population in residential building design, normative base they are created, modern state and problematic. Hybrid war and full-scale invasion showed weakness and unpreparedness of cities by civil defense of population. Possible solutions of this problem have to take into consideration modern trends of sustainability, adaptivity of development for fast changing actual reality, contribute to progress of urban spaces and people in conditions of permanent war threats of conflicts, terroristic attacks, natural and technogenic disasters etc. The new requirements for civil defense must take into account the urgent need for sustainable development, be economically feasible, and provide the possibility of multi-purpose use of bomb shelters. A detailed consideration of possible ways to meet these requirements takes into account economic, legal, environmental and social factors.

Firmware Attacks: The Silent Threat to Your IoT Connected Devices

Firmware attacks on IoT devices have become a growing concern in recent years. These attacks exploit vulnerabilities in the firmware, which is the low-level software that controls the hardware of the device, to gain access to sensitive data or control the device remotely. Firmware attacks are particularly dangerous because they can be difficult to detect and can allow attackers to maintain access to a device for an extended period of time. In this paper, we explore the silent threat of firmware attacks on connected devices and the potential consequences of a successful attack. We discuss the methods that attackers use to exploit firmware vulnerabilities and the impact these attacks can have on the security and privacy of users. We also provide recommendations for protecting against firmware attacks, including keeping firmware up to date, using strong passwords, and monitoring for suspicious activity on the network. By understanding the risks and taking proactive steps to protect against firmware attacks, users can help to ensure the security and integrity of their connected devices.

On some relations of SCA-related properties of S-box under the Hamming weight leakage mode

Abstract— Physical implementations of cryptographic algorithms are vulnerable to so-called side-channel attacks, in which sensitive information can be recovered through the analysis of the leakages produced by the operating device. In the particular case of block ciphers, substitution boxes are often the target of such attacks, as they are the main nonlinear component of the cipher. Literature survey contains some definitions of theoretical properties to measure the resistance provided by substitution boxes with respect to the imminent threat of side-channel attacks. The fundamental questions we board in this paper are: are all these properties related? And if so, what can we tell of their relation? We pretend to bring some clarification to this subject using some well-known side-channel related properties of S-box.

GNSS Spoofer Localization for Vehicles Based on Doppler and Clock Drift Double Difference

Global navigation satellite systems (GNSS) have been widely used in our daily lives. Meanwhile, GNSS are facing a severe threat of spoofing attacks, which can mislead users by false position, velocity and time (PVT) solutions, resulting in serious consequences. Especially, with the development of vehicle navigation, autonomous driving and unmanned aerial vehicles (UAV), increasing spoofing attacks have emerged towards vehicles, where anti-spoofing techniques are desperately needed. Among these techniques, spoofer localization is a necessary step to fundamentally address the spoofing attacks. Most of the conventional spoofer localization methods are based on time of arrival (TOA) or time difference of arrival (TDOA). These methods require measurements observed by multiple receivers at the same time, where time synchronization is indispensable. However, it is hard to apply these methods to the existing vehicles equipped with independent commercial receivers. To cope with this problem, in this paper we develop a flexible method for vehicles to locate a stationary spoofer based on Doppler and clock drift double difference (DCDD). DCDD is calculated using Doppler measurements and clock drift solutions of one vehicular receiver and another reference receiver. Generally, Doppler frequency of GNSS signals changes slowly with time, which releases the restriction for high-precision time synchronization. DCDD represents the Doppler frequency caused by the vehicle motion, from which the spoofer position can be estimated. In addition, with the elimination of irrelevant components in Doppler measurements, DCDD minimizes the user complexity required for localization, where at least two receivers and one vehicle is needed. Utilizing DCDD, an iterative weighted least squares (WLS) algorithm is used to solve the spoofer position, and a closed-form initial value is given to improve the convergence rate. We evaluate the localization performance by the derived Cramér-Rao lower bound (CRLB) in the presence of vehicle position and velocity errors. Simulations and field vehicle experiments are conducted to verify the feasibility and effective of the proposed method. The experimental results show that with an average vehicle velocity of 8 m/s, the horizontal localization error can be less than 40 m and 20 m, using DCDD observations of 4 epochs and 8 epochs respectively.

Threat Framing of CPEC In Pakistani Newspapers: Post Dasu Attach Analysis

This research intends to study geostrategic and eco-political threats attached to China Pakistan Economic Corridor (CPEC) concerning Pakistani media coverage of Dasu attack in 2021. While 13 workers, including Chinese engineers were killed which resultantly resurfaced the security threat attached to this project. To analyze the media’s apprehension about the incident and framing trends, editorials of four leading newspapers, two English; Dawn and The Nation, two Urdu: Daily Jang and Daily Express, were selected with a duration from July 1, 2021 to November 30, 2022. Previous studies helped in determining to incorporate ‘threat frame’ to be analyzed in editorials which further lead to quantify its coverage frequency. The findings revealed that newspapers indicated towards threats of security, insurgencies, internal and external terrorism attacks with serious apprehension to be resolved. The findings also concluded that editorial manuscripts help in steering and determining geopolitical issues in constructive ways.

The need for an individualized approach to what is considered a clinically significant reduction in seizure frequency: A patient's perspective.

The need for an individualized approach to what is considered a clinically significant reduction in seizure frequency: A patient's perspective.

Fuzzy Harmony Search Technique for Cyber Risks in Industry 4.0 Wireless Communication Networks

Industry 4.0 houses diverse technologies including wireless communication and shared networks for internal and external operations. Due to the wireless nature and remote operability, the exposure to security threats is high. Cyber risk detection and mitigation are prominent for secure industrial operations and planned outcomes. In addition, the system faces the threat of intelligence attacks, security standards issues, privacy concerns and scalability problems. The cyber risk related research problems influence overall data transmission in industry wireless communication networks. For augmenting communication security through cyber risk detection, this article introduces an Explicit Risk Detection and Assessment Technique (ERDAT) for cyber threat mitigation in the industrial process. A fuzzy harmony search algorithm powers this technique for identifying the risk and preventing its impact. The harmony search algorithm mimics the adversary impact using production factors such as process interruption or halting and production outcome. The search performs a mimicking operation for a high objective function based on production output for the admitted plan. The fuzzy operation admits the above factors for identifying the cyber impacting risk, either for its impacts or profitable outcome. In this process, the fuzzy optimization identifies the maximum or minimum objective output targeted for either outcome or risk interrupts, respectively. The fuzzy threshold is identified using a mediated acceptable range, computed as the ratio between minimum and maximum, mimicking occurrences between the risk and scheduled production outcomes. Therefore, the mimicking crossing or falling behind the threshold for the interruption/halting or production, respectively, are identified as risks and their source is detected. The detection communication source is disconnected from the industrial process for preventing further adversary impacts. The introduced system achieves 8.52% high-risk detection, 12.5% fewer outcome interrupts, 8.3% fewer halted schedules, 8.08% less interrupt span, and 7.94% less detection time compared to traditional methods.