To Örebro University

Secure safety monitoring is an intriguing and practical feature of current ride-hailing services. It enhances the riders safety after they are onboard a drivers vehicle while not violating location privacy. Existing work suffer from the problems of unnecessary trajectory uploading, centralized trajectory monitoring, redundant trajectory monitoring, and assuming static trajectories. In this work, we propose a decentralized, secure, and efficient safety monitoring scheme Artemis to guarantee rider safety while supporting dynamic trajectories, i.e., permitted trajectory deviations. In specific, we design a 2-out-of-n private threshold signature protocol to achieve trajectory authenticity, design a decentralized RHS platform and a secure trajectory similarity computation protocol to guarantee both efficiency and location privacy, and design a secure three-party computation protocol to ensure deviation authenticity. Formal security experiments and proofs are provided. Experimental results from extensive experiments based on Ethereum and Intel SGX2 demonstrate that Artemis is highly efficient.

Adversarial examples are small and often imperceptible perturbations crafted to fool machine learning models. These attacks seriously threaten the reliability of deep neural networks, especially in security-sensitive domains. Evasion attacks, a form of adversarial attack where input is modified at test time to cause misclassification, are particularly insidious due to their transferability: adversarial examples crafted against one model often fool other models as well. This property, known as adversarial transferability, complicates defense strategies since it enables black-box attacks to succeed without direct access to the victim model. While adversarial training is one of the most widely adopted defense mechanisms, its effectiveness is typically evaluated on a narrow and homogeneous population of models. This limitation hinders the generalizability of empirical findings and restricts practical adoption.

In this work, we introduce DUMBer, an attack framework built on the foundation of the DUMB (Dataset soUrces, Model architecture, and Balance) methodology, to systematically evaluate the resilience of adversarially trained models. Our testbed spans multiple adversarial training techniques evaluated across three diverse computer vision tasks, using a heterogeneous population of uniquely trained models to reflect real-world deployment variability. Our experimental pipeline comprises over 130k evaluations spanning 13 state-of-the-art attack algorithms, allowing us to capture nuanced behaviors of adversarial training under varying threat models and dataset conditions. Our findings offer practical, actionable insights for AI practitioners, identifying which defenses are most effective based on the model, dataset, and attacker setup.

Data augmentation techniques show potential in various domains, yet their application to enhance robustness in wireless anomaly detection remains underexplored. Wireless datasets often suffer from anomaly scarcity and class imbalance, hindering the training of reliable detection models. This work introduces GANSec, a novel conditional Generative Adversarial Networks (GAN) framework specifically designed to augment wireless time-series data. We investigate different neural network architectures (MLP, LSTM, CNN) and two conditional training objectives (Embedded Conditional, Classification Oriented) within GANSec, evaluating the framework using real-world 5G measurements for jamming anomaly detection. For evaluation, we train the downstream anomaly detector exclusively on GANSec-generated data and test its performance in a cross-scenario setting. Our evaluation demonstrates that models trained this way significantly outperform those trained on original or baseline augmentation data when tested under unseen network conditions. Specifically, our approach achieved up to 92.13% accuracy on the unseen dataset (i.e., data collected from a different distribution reflecting network conditions distinct from the training set), compared to 78% for models trained on raw data and 83.33% for the best-performing baseline, exhibiting substantially enhanced robustness and generalization.

Online voting enables individuals to participate in elections remotely, offering greater efficiency and accessibility in both governmental and organizational settings. As this method gains popularity, ensuring the security of online voting systems becomes increasingly vital, as the systems supporting it must satisfy a demanding set of security requirements. Most research in this area emphasizes the design and verification of cryptographic protocols to protect voter integrity and system confidentiality. However, other vectors, such as network traffic analysis, remain relatively understudied, even though they may pose significant threats to voter privacy and the overall trustworthiness of the system. In this paper, we examine how adversaries can exploit metadata from encrypted network traffic to uncover sensitive information during online voting. Our analysis reveals that, even without accessing the encrypted content, it is possible to infer critical voter actions, such as whether a person votes, the exact moment a ballot is submitted, and whether the ballot is valid or spoiled. We test these attacks with both rule-based techniques and machine learning methods. We evaluate our attacks on two widely used online voting platforms, one proprietary and one partially open source, achieving classification accuracy as high as 99.5%. These results expose a significant privacy vulnerability that threatens key properties of secure elections, including voter secrecy and protection against coercion or vote-buying. We explore mitigations to our attacks, demonstrating that countermeasures such as payload padding and timestamp equalization can substantially limit their effectiveness.

Recent incidents have demonstrated the increasing vulnerability of Industrial Control Systems (ICSs) to sophisticated and targeted attacks orchestrated by adversaries with high motivation, resources, and domain knowledge. Among these threats, False Data Injection (FDI) attacks have emerged as one of the main security threats to ICSs, involving the deliberate manipulation or injection of false data into the control system to deceive or disrupt operations. FDI attacks pose a significant risk due to their high capacity of concealment and ability to evade intrusion detection systems that rely on accurate ICS models. In this paper, we present defclon, a novel Software-Defined Networking (SDN)-based Moving Target Defense (MTD) approach against FDI attacks. Defclon proactively replicates network packets across multiple network paths and adaptively selects a single path using a signaling game model to reach the destination end-device. We demonstrate the effectiveness of our approach through simulations, numerical analysis, and experiments on ICS network traffic and topologies. Experimental results show that defclon is able to not only mitigate the effects of FDI attacks, but also to introduce different levels of uncertainty without degrading network performance, significantly increasing the difficulty for adversaries to gather information and launch attacks.

Wildcard Keyword Searchable Encryption (WKSE) has grown into a ubiquitous tool. It enables clients to search desired files with wildcard expressions. Although promising, previous schemes confront three barriers: (1) An adversary can launch a correlation attack to acquire the similarity between keywords. (2) The WKSE schemes exhibit false positives which can lead to wrong search results. (3) Existing feature extraction strategies limit the flexibility of search expressions. In this paper, we propose a Multi-Character Searchable Encryption scheme (MCSE) that overcomes the aforementioned barriers. To resist correlation attacks, we design the randomize-pad model to encrypt the vector. To eradicate false positives, we apply the vector space model and complete feature extraction strategies so that a feature set uniquely identifies a keyword or expression. To enhance search flexibility, we introduce three distinct feature extraction strategies for keyword expressions, wildcard expressions, and logical expressions, enabling effective multi-character search. These strategies enable indexes to accommodate the search of diverse expressions. Finally, we prove that MCSE is indistinguishable against chosen-feature attacks and implement MCSE on two real datasets. Compared with state-of-the-art schemes, the experiment results show that MCSE achieves good performance.

Serverless computing has rapidly emerged as a prominent cloud paradigm, enabling developers to focus solely on application logic without the burden of managing servers or underlying infrastructure. Public serverless repositories have become key to accelerating the development of serverless applications. However, their growing popularity makes them attractive targets for adversaries. Despite this, the security posture of these repositories remains largely unexplored, exposing developers and organizations to potential risks. In this paper, we present the first comprehensive analysis of the security landscape of serverless components hosted in public repositories. We analyse 2,758 serverless components from five widely used public repositories popular among developers and enterprises, and 125,936 Infrastructure as Code (IaC) templates across three widely used IaC frameworks. Our analysis reveals systemic vulnerabilities including outdated software packages, misuse of sensitive parameters, exploitable deployment configurations, susceptibility to typo-squatting attacks and opportunities to embed malicious behaviour within compressed serverless components. Finally, we provide practical recommendations to mitigate these threats.

Machine learning and deep learning are transformative forces reshaping our networks, industries, services, and ways of life. However, the susceptibility of these intelligent systems to adversarial attacks remains a significant issue. On the one hand, recent studies have demonstrated the potential transferability of adversarial attacks across diverse models. On the other hand, existing defense mechanisms are vulnerable to advanced attacks or are often limited to certain attack types. This study proposes a random deep feature selection approach to mitigate such transferability and improve the robustness of models against adversarial manipulations. Our approach is designed to strengthen deep models against poisoning (e.g., label flipping) and exploratory (e.g., DeepFool, BIM, FGSM, I-FGSM, L-BFGS, C&W, JSMA, and PGD) attacks that are applied in both the training and testing stages, and Transfer Learning-Based Adversarial Attacks. We consider scenarios involving perfect and semi-knowledgeable attackers. The performance of our approach is evaluated through extensive experiments on the renowned UNSW-NB15 dataset, including both real-world and synthetic data, covering a wide range of modern attack behaviors and benign activities. The results indicate that our approach boosts the effectiveness of the target network to over 80% against label-flipping poisoning attacks and over 60% against all major types of exploratory attacks.

Tor is currently the most widely used anonymity network, providing anonymity to both users and hidden services. However, a wide range of attacks proposed in the literature claim the capability of de-anonymizing its users. Among the others, Flow Correlation Attacks aiming at discovering the activity of a specific target user present impressive results in the literature. Specifically, recent attacks such as DeepCoFFEA (S&P'22), and SUMo (NDSS'24) show more then 90% precision in correlating users and their traffic. However, when investigating Tor's past efforts in mitigating this kind of attack, we formulated the following research question: how realistic and effective are these attacks on state-of-the-art Tor networks?

In this paper, we answer this question by evaluating state-of-the-art attacks and mitigations. We create a novel dataset of Tor traffic in five different scenarios, leveraging both the latest available Tor implementation and state-of-the-art defenses. We also proposed and evaluated a novel countermeasure. We then used our collected data to test two recent and prominent flow correlation attacks, i.e., SUMo (NDSS'24) and ESPRESSO (2024), a refinement of DeepCoFFEA (S&P'22). Our experimental results demonstrate that both ESPRESSO and SUMo are ineffective at correlating Tor traffic, achieving 8% accuracy and 12% accuracy, respectively. By investigating the motivations behind these results, we discuss how mitigation developed by Tor substantially reduced the risks associated with such attacks, thereby strengthening user anonymity.

Fuzzing frameworks allow the detection of code bugs or errors by feeding the code with unconventional inputs. Recently, Ma et al. proposed LOKI (NDSS, 2023), the first fuzzing framework tailored to blockchain consensus protocols [1]. LOKI uses high-frequency messages to model the dynamics of consensus algorithms. However, low-frequency messages are fundamental to the consensus's state evolution. Thus, neglecting them hinders the in-depth exploration of vulnerabilities, leading to missed bugs or critical issues generated by low-frequency messages.

In this paper, we propose messAge-aWare hybrid markOved StatE model (AWOSE), a fuzzing frameworkmodeling highandlow-frequency messages to uncover vulnerabilities and bugs in consensus algorithms. AWOSEleverages Markov Chains (MCs) tomodel the general states of the consensus messages while leveraging Hidden Markov Models (HMMs) to model latent or hidden state transitions based on code mutations (as in LOKI) or external factors (e.g., malicious inputs). Such probabilistic state models are beneficial for capturing the uncertain nature of system behaviors, enabling a more realistic and flexible representation of consensus dynamics. AWOSE is the first model to introduce the message context-based state model generation for fuzzing purposes. Via a thorough evaluation, we compare AWOSE with LOKI. Results show that our proposed AWOSE provides 48.3% more accurate states and 83.2% better path coverage. Thanks to AWOSE we were able to uncover new bugs that LOKI did not detect. Our experiments lead to the identification of previously unknown vulnerabilities, currently being evaluated as CVEs.