Threat Model Research Articles

Anomaly Detection Against GPS Spoofing Attacks on Connected and Autonomous Vehicles Using Learning From Demonstration

GPS spoofing attacks pose great challenges to connected vehicle (CVs) safety applications and localization of autonomous vehicles (AVs). In this paper, we propose to utilize transportation and vehicle engineering domain knowledge to detect GPS spoofing attacks towards CVs and AVs. A novel detection method using learning from demonstration is developed, which can be implemented in both vehicles and at the transportation infrastructure. A computational-efficient driving model, which can be learned from historical trajectories of the vehicles, is constructed to predict normal driving behaviors. Then a statistical method is developed to measure the dissimilarities between the observed trajectory and the predicted normal trajectory for anomaly detection. We validate the proposed method using two threat models (i.e., attacks targeting the multi-sensor fusion system of AVs and attacks targeting the intersection movement assist application of CVs) on two real-world datasets (i.e., KAIST and Michigan roundabout dataset). Results show that the proposed model is able to detect almost all of the attacks in time with low false positive and false negative rates.

Efficient Robustness Assessment via Adversarial Spatial-Temporal Focus on Videos.

Adversarial robustness assessment for video recognition models has raised concerns owing to their wide applications on safety-critical tasks. Compared with images, videos have much high dimension, which brings huge computational costs when generating adversarial videos. This is especially serious for the query-based black-box attacks where gradient estimation for the threat models is usually utilized, and high dimensions will lead to a large number of queries. To mitigate this issue, we propose to simultaneously eliminate the temporal and spatial redundancy within the video to achieve an effective and efficient gradient estimation on the reduced searching space, and thus query number could decrease. To implement this idea, we design the novel Adversarial spatial-temporal Focus (AstFocus) attack on videos, which performs attacks on the simultaneously focused key frames and key regions from the inter-frames and intra-frames in the video. AstFocus attack is based on the cooperative Multi-Agent Reinforcement Learning (MARL) framework. One agent is responsible for selecting key frames, and another agent is responsible for selecting key regions. These two agents are jointly trained by the common rewards received from the black-box threat models to perform a cooperative prediction. By continuously querying, the reduced searching space composed of key frames and key regions is becoming precise, and the whole query number becomes less than that on the original video. Extensive experiments on four mainstream video recognition models and three widely used action recognition datasets demonstrate that the proposed AstFocus attack outperforms the SOTA methods, which is prevenient in fooling rate, query number, time, and perturbation magnitude at the same time.

Security Provided by the Physical Layer in Wireless Communications

This article provides an overview of five physicallayer security mechanisms: Physical-Layer Authentication (PLA), Physical-Layer Confidential Communications (PLCC), Physical- Layer Key Generation (PLKG), Physical-Layer Communications with Low Probability of Detection (PLC-LPD), and Physical- Layer Steganography (PLS). We explain the relationship among five physical-layer security mechanisms. Then, we review the features of various physical-layer security mechanisms in terms of the basic idea, properties, assumptions, classification, threat model, and applications, respectively. At last, we introduce five promising plans by combining different physical-layer security mechanisms to further improve the final performance.

Beyond detection: Uncovering unknown threats

Threat management is essential for ensuring an organisation’s security, but traditional strategies often only address known threats, leaving the organisation vulnerable to unknown threats. To be well equipped against advanced cyberattacks, a proactive approach beyond detection that uncovers unknown and emerging threats is necessary. This paper proposes a comprehensive approach to threat management involving the partnership between the threat detection, threat hunting, threat intelligence and threat exposure teams. Various approaches for hunting unknown threats are explored, including simulation, forensics, threat modelling, incident pivoting, deception, and a process to hunt once and automate. Insights detailed in this paper will also help organisations make informed decisions on resources and practices around threat hunting. The proposed strategy emphasises the need for a proactive and iterative approach to threat management, allowing organisations to stay ahead of adversaries and be prepared for unknown threats.

Programmable Name Obfuscation Framework for Controlling Privacy and Performance on CCN

Consumer privacy leakage from data names poses a serious threat to Content-Centric Networking (CCN) networks. Obfuscating names is a promising countermeasure, and anonymizers with deterministic encryption schemes have been proposed to provide data privacy while enabling CCN features, such as in-network caching. Existing studies assume a weak threat model in which anonymizers are honest, and their obfuscation schemes are not resilient against privacy attacks such as name guessing attacks. This paper designs a name obfuscation framework based on the realistic assumption that anonymizers are semi-honest. The framework strengthens data privacy using multiple keys and separates obfuscation for prefixes and suffixes, and is implemented on a P4 switch to provide Tbps forwarding speed.

StratDef: Strategic defense against adversarial attacks in ML-based malware detection

Over the years, most research towards defenses against adversarial attacks on machine learning models has been in the image recognition domain. The ML-based malware detection domain has received less attention despite its importance. Moreover, most work exploring these defenses has focused on several methods but with no strategy when applying them. In this paper, we introduce StratDef, which is a strategic defense system based on a moving target defense approach. We overcome challenges related to the systematic construction, selection, and strategic use of models to maximize adversarial robustness. StratDef dynamically and strategically chooses the best models to increase the uncertainty for the attacker while minimizing critical aspects in the adversarial ML domain, like attack transferability. We provide the first comprehensive evaluation of defenses against adversarial attacks on machine learning for malware detection, where our threat model explores different levels of threat, attacker knowledge, capabilities, and attack intensities. We show that StratDef performs better than other defenses even when facing the peak adversarial threat. We also show that, of the existing defenses, only a few adversarially-trained models provide substantially better protection than just using vanilla models but are still outperformed by StratDef.

Moderating the Relationship Between Ulcerative Colitis and Crohn's Disease and Symptoms of PTSD: The Role of Remission.

This study examined whether inflammatory bowel disease (IBD) patients endorse clinically significant symptoms of post-traumatic stress disorder (PTSD), and tested whether remission status and remission expectations effectively moderate the relationship between endorsements of PTSD symptoms and aspects of IBD. The enduring somatic threat model speaks to the presentation of symptoms of trauma that result from ongoing somatic concerns rather than discrete external events. Literature shows patients living with acute conditions experience symptoms of PTSD; however, few studies extend this to the IBD population. In addition, literature suggests remission may serve as a protective factor for the impact of IBD; as such, aspects of remission may serve as moderators in the relationship between aspects of IBD and PTSD symptoms. Among a sample of adults with IBD, results showed that 32.8% of participants met the established cutoff for PTSD symptoms warranting further diagnostic evaluation. The findings further showed several aspects of remission moderated the relationship between multiple distinct IBD related concerns and PTSD symptoms. Given the results of this study, it would be beneficial for providers to maintain awareness of the potential impact of PTSD symptoms, including the ways in which these symptoms may influence patient engagement/presentation. Ultimately, these results inform efforts to continue appropriate referral to mental health professionals for follow up.

Emotional state as a modulator of autonomic and somatic nervous system activity in postural control: a review.

Advances in our understanding of postural control have highlighted the need to examine the influence of higher brain centers in the modulation of this complex function. There is strong evidence of a link between emotional state, autonomic nervous system (ANS) activity and somatic nervous system (somatic NS) activity in postural control. For example, relationships have been demonstrated between postural threat, anxiety, fear of falling, balance confidence, and physiological arousal. Behaviorally, increased arousal has been associated with changes in velocity and amplitude of postural sway during quiet standing. The potential links between ANS and somatic NS, observed in control of posture, are associated with shared neuroanatomical connections within the central nervous system (CNS). The influence of emotional state on postural control likely reflects the important influence the limbic system has on these ANS/somatic NS control networks. This narrative review will highlight several examples of behaviors which routinely require coordination between the ANS and somatic NS, highlighting the importance of the neurofunctional link between these systems. Furthermore, we will extend beyond the more historical focus on threat models and examine how disordered/altered emotional state and ANS processing may influence postural control and assessment. Finally, this paper will discuss studies that have been important in uncovering the modulatory effect of emotional state on postural control including links that may inform our understanding of disordered control, such as that observed in individuals living with Parkinson's disease and discuss methodological tools that have the potential to advance understanding of this complex relationship.

A survey on secure metadata of agile software development process using blockchain technology

AbstractScrum is an important and essential technique for developing software. Scrum framework is a project management technique that can emphasize software development teamwork. If you are starting, consider scrum as a method for accomplishing tasks as a team in manageable chunks simultaneously, with ongoing experimentation and feedback loops to learn and improve as you go. Scrum enables individuals and groups to collaboratively and gradually generate value. Being an agile framework, scrum gives people and teams the perfect structure to fit into their workflows while still including the best practices to cater to their requirements. There are some important considerations for security in the agile software development process such as threat modeling, secure coding practices, continuous security testing, authentication, and authorization. In this work, we reviewed various scrum developing techniques to find out the research gaps. This article consider the scrum technique with blockchain technologies for review because it is preferred by most of the software industries.

Anonymous federated learning framework in the internet of things

AbstractWith the continuous development of the internet of things (IoT), federated learning is being widely applied. This technology keeps data locally to protect data security. However, during the process of uploading gradients from local clients, there is a possibility of leaking sensitive information such as identities. To address this issue, this paper presents a secure and efficient anonymous federated learning framework. In our proposed, we first design a lightweight key‐sharing protocol based on elliptic curve cryptography (ECC) to generate shared keys and ensure secure communication. We then describe the improved process of federated learning, where clients communicate with the cloud server using pseudonyms to achieve anonymity. The security of our protocol is analyzed from both formal and informal perspectives, demonstrating that our proposed protocol satisfies “session key security.” We also employ the formal verification tool ProVerif to validate the security of the protocol in terms of mutual authentication and key configuration in the Dolve–Yao threat model. Finally, the computational and communication costs of the proposed anonymous federated learning framework are evaluated, showing that both the computational and communication expenses are relatively low. A comparison was made between the proposed federated learning process and three other federated learning processes, demonstrating the clear advantages of our proposed.

Use of Fuzzy Logic in Cyber Security System: Cyber Threat Intelligence

Purpose of the study: Fuzzy logic is a mathematical concept that allows for the handling of uncertain or imprecise data. In cyber security, fuzzy logic can be used to improve the accuracy and efficiency of security systems. Cyber threat intelligence plays a crucial role in modern cybersecurity by enabling organizations to proactively identify, analyse, and respond to potential cyber threats. With the increasing complexity and sophistication of cyber threats, effective cyber threat intelligence has become a critical component of modern cybersecurity. Traditional approaches often struggle to handle the inherent uncertainties and complexities in threat intelligence data.
 Methodology: Through a comprehensive literature review, examines the current state of cyber threat intelligence and the principles of fuzzy logic. The paper presents a detailed analysis of how fuzzy logic can be applied to various aspects of cyber threat intelligence, including threat modelling, detection and classification, risk assessment, and decision support systems.
 Main Findings: The findings highlight the benefits of using fuzzy logic in cyber threat intelligence and pave the way for further research and development in this promising field. The future prospects and challenges of integrating fuzzy logic with other advanced technologies such as machine learning and artificial intelligence.
 Applications of this study: The proposed approach has the potential to significantly improve the analysis and decision-making processes in agriculture, helping farmers to make more informed decisions about crop treatments and ultimately increasing crop yields.
 Novelty/Originality of this study: The outcomes of this research contribute to advancing the field of cyber threat intelligence and provide valuable insights into the practical implementation of fuzzy logic for better threat analysis and mitigation strategies.

Towards evaluating robustness of violence detection in videos using cross-domain transferability

Recent studies have demonstrated the applicability of classification models based on deep learning for detecting human action. For this reason, automatic violence detection from videos has become a crying need to prevent the spreading of vulnerable content on digital platforms. In spite of the remarkable success of these neural networks, they are prone to fail against adversarial attacks and thus, highlighted the need for evaluating the robustness of these state-of-the-art violence detection classifiers. Here, we propose a transferable logit attack for binary misclassification of video data which can evade the system by our spatially perturbed synthesized adversarial samples. We utilize the adversarial falsification-based threat model to validate a non-sparse white-box attack setting which will generate cross-domain adversarial video samples by perturbing only spatial features without affecting the temporal features. We carry out extensive experiments on the validation set of two popular violence detection datasets: Hockey Fight Dataset, and Movie Dataset and verify that our proposed attack method has high attack success rate for these datasets against the state-of-the-art violence detection classifier. This work aims to make future violence detection models more resistant to adversarial examples.

Which information locations in covered entities under HIPAA must be secured first? A multi-criteria decision-making approach.

Creating adequate safeguards for physical and online locations (e.g., desktop computers, network servers) where protected health information (PHI) may be breached is critical for management within entities compliant with the Health Information Portability and Accountability Act (HIPAA). With the increasing complexity of cyber breaches and budgetary issues, prioritizing which locations require the most immediate attention by top management through a data-driven model is more important than ever. Using CORAS threat modeling and five methods for multi-criteria decision-making, these locations were ranked from greatest to least risk of data breaches. Statistical methods were subsequently used for consistency and robustness checks. The findings illustrate that each type of covered entity under HIPAA must prioritize a different set of locations to safeguard first: health care providers must focus on the security of network servers, other portable electronic devices, and category of others (i.e., miscellaneous locations); health plans must focus on the security of paper and films, network servers, and others; and business associates must focus on the security of category of others, network servers, and other portable electronic devices. Combined with data on the source of the breaches (external vs. internal) and type of threats (e.g., hacking, theft), these findings provide recommendations for risk identification for privacy officers across health care.

Coming of Age

Stressing the importance of threat models.

Leveraging self-sovereign identity & distributed ledger technology in renewable energy certificate ecosystems

Renewable Energy Certificates (RECs) are tradable units that represent the commodity in the form of environmental attributes generated for each unit of electricity produced by a renewable energy source. Furthermore, the energy sector's digitalization ushers in new crucial enablers like Distributed Ledger Technology (DLT), which may be used for REC record tracking and trading. Unfortunately, there are a number of outstanding issues such as the lack of a common standard for the representation, communication, and verification of REC, reliance on centralized entities, and others. In order to harness the true potential of energy DLT and REC, it is imperative to address these issues. In this visionary article, we propose a holistic approach which leverages a novel decentralized identity mechanism called Self-sovereign Identity (SSI) and DLT. We present its architecture, based on a rigorous threat model and requirement analysis as well as detailed use-cases to illustrate how the architecture can be used in different REC use-cases.

DisguisedNets: Secure Image Outsourcing for Confidential Model Training in Clouds

Large training data and expensive model tweaking are standard features of deep learning with images. As a result, data owners often utilize cloud resources to develop large-scale complex models, which also raises privacy concerns. Existing cryptographic solutions for training deep neural networks (DNNs) are too expensive, cannot effectively utilize cloud GPU resources, and also put a significant burden on client-side pre-processing. This article presents an image disguising approach: DisguisedNets, which allows users to securely outsource images to the cloud and enables confidential, efficient GPU-based model training. DisguisedNets uses a novel combination of image blocktization, block-level random permutation, and block-level secure transformations: random multidimensional projection (RMT) or AES pixel-level encryption (AES) to transform training data. Users can use existing DNN training methods and GPU resources without any modification to training models with disguised images. We have analyzed and evaluated the methods under a multi-level threat model and compared them with another similar method—InstaHide. We also show that the image disguising approach, including both DisguisedNets and InstaHide, can effectively protect models from model-targeted attacks.

ADDA: An Adversarial Direction-Guided Decision-Based Attack via Multiple Surrogate Models

Over the past decade, Convolutional Neural Networks (CNNs) have been extensively deployed in security-critical areas; however, the security of CNN models is threatened by adversarial attacks. Decision-based adversarial attacks, wherein an attacker relies solely on the final output label of the target model to craft adversarial examples, are the most challenging yet practical adversarial attacks. However, existing decision-based adversarial attacks generally suffer from poor query efficiency or low attack success rate, especially for targeted attacks. To address these issues, we propose a query-efficient Adversarial Direction-guided Decision-based Attack (ADDA), which exploits the advantages of transfer-based priors and the benefits of a single query. The transfer-based priors provided by the gradients of multiple different surrogate models can be utilized to suggest the most promising search directions for generating adversarial examples. The query consumption during the ADDA attack is mainly derived from a single query evaluation of the candidate adversarial samples, which significantly saves the number of queries. Experimental results on several ImageNet classifiers, including l∞ and l2 threat models, demonstrate that our proposed approach overwhelmingly outperforms existing state-of-the-art decision-based attacks in terms of both query efficiency and attack success rate. We show case studies of ADDA against a real-world API in which it is successfully able to fool the Google Cloud Vision API after only a few queries.

A Formal Verification of a Reputation Multi-Factor Authentication Mechanism for Constrained Devices and Low-Power Wide-Area Network Using Temporal Logic.

There are many security challenges in IoT, especially related to the authentication of restricted devices in long-distance and low-throughput networks. Problems such as impersonation, privacy issues, and excessive battery usage are some of the existing problems evaluated through the threat modeling of this work. A formal assessment of security solutions for their compliance in addressing such threats is desirable. Although several works address the verification of security protocols, verifying the security of components and their non-locking has been little explored. This work proposes to analyze the design-time security of the components of a multi-factor authentication mechanism with a reputation regarding security requirements that go beyond encryption or secrecy in data transmission. As a result, it was observed through temporal logic that the mechanism is deadlock-free and meets the requirements established in this work. Although it is not a work aimed at modeling the security mechanism, this document provides the necessary details for a better understanding of the mechanism and, consequently, the process of formal verification of its security properties.

Practices and challenges of threat modelling in agile environments

Facing the increasing annual cybersecurity costs, threat modelling (TM) is a method to consider security as early as possible in the software development life cycle (SDLC). Thereby, TM helps to identify and address security-related design flaws in information systems. As the original TM approach is based on sequential development, it is not aligned with today’s predominantly agile environments. This results in several challenges. However, TM’s implementation in an agile development approach lacks the recommendations on how to tackle these challenges. Therefore, we assess the state-of-the-art of TM challenges and practices in agile environments by conducting a literature review covering 220 papers. Thereby, we identify nine categories of challenges and six categories of practices. We propose a valuable artefact for practitioners by mapping challenges and practices to the agile SDLC and by creating a matrix highlighting how the practices address the challenges of TM in an agile environment.

Defending Adversarial Attacks on Deep Learning-Based Power Allocation in Massive MIMO Using Denoising Autoencoders

Recent work has advocated for the use of deep learning to perform power allocation in the downlink of massive MIMO (maMIMO) networks. Yet, such deep learning models are vulnerable to adversarial attacks. In the context of maMIMO power allocation, adversarial attacks refer to the injection of subtle perturbations into the deep learning model’s input, during inference (i.e., the adversarial perturbation is injected into inputs during deployment after the model has been trained) that are specifically crafted to force the trained regression model to output an infeasible power allocation solution. In this work, we develop an autoencoder-based mitigation technique, which allows deep learning-based power allocation models to operate in the presence of adversaries without requiring retraining. Specifically, we develop a denoising autoencoder (DAE), which learns a mapping between potentially perturbed data and its corresponding unperturbed input. We test our defense across multiple attacks and in multiple threat models and demonstrate its ability to (i) mitigate the effects of adversarial attacks on power allocation networks using two common precoding schemes, (ii) outperform previously proposed benchmarks for mitigating regression-based adversarial attacks on maMIMO networks, (iii) retain accurate performance in the absence of an attack, and (iv) operate with low computational overhead.