MDPI - Publisher of Open Access Journals

21 pages, 19235 KB

Open AccessFeature PaperArticle

Exploring Public Data Vulnerabilities in Semi-Supervised Learning Models through Gray-box Adversarial Attack

by Junhyung Jo, Joongsu Kim and Young-Joo Suh

Electronics 2024, 13(5), 940; https://doi.org/10.3390/electronics13050940 - 29 Feb 2024

Cited by 3 | Viewed by 2578

Semi-supervised learning (SSL) models, integrating labeled and unlabeled data, have gained prominence in vision-based tasks, yet their susceptibility to adversarial attacks remains underexplored. This paper unveils the vulnerability of SSL models to gray-box adversarial attacks—a scenario where the attacker has partial knowledge of the model. We introduce an efficient attack method, Gray-box Adversarial Attack on Semi-supervised learning (GAAS), which exploits the dependency of SSL models on publicly available labeled data. Our analysis demonstrates that even with limited knowledge, GAAS can significantly undermine the integrity of SSL models across various tasks, including image classification, object detection, and semantic segmentation, with minimal access to labeled data. Through extensive experiments, we exhibit the effectiveness of GAAS, comparing it to white-box attack scenarios and underscoring the critical need for robust defense mechanisms. Our findings highlight the potential risks of relying on public datasets for SSL model training and advocate for the integration of adversarial training and other defense strategies to safeguard against such vulnerabilities. Full article

(This article belongs to the Special Issue AI Security and Safety)

► Show Figures

Figure 1

25 pages, 1030 KB

Open AccessArticle

Smarter Evolution: Enhancing Evolutionary Black Box Fuzzing with Adaptive Models

by Anne Borcherding, Martin Morawetz and Steffen Pfrang

Sensors 2023, 23(18), 7864; https://doi.org/10.3390/s23187864 - 13 Sep 2023

Cited by 2 | Viewed by 2622

Abstract

Smart production ecosystems are a valuable target for attackers. In particular, due to the high level of connectivity introduced by Industry 4.0, attackers can potentially attack individual components of production systems from the outside. One approach to strengthening the security of industrial control systems is to perform black box security tests such as network fuzzing. These are applicable, even if no information on the internals of the control system is available. However, most security testing strategies assume a gray box setting, in which some information on the internals are available. We propose a new approach to bridge the gap between these gray box strategies and the real-world black box setting in the domain of industrial control systems. This approach involves training an adaptive machine learning model that approximates the information that is missing in a black box setting. We propose three different approaches for the model, combine them with an evolutionary testing approach, and perform an evaluation using a System under Test with known vulnerabilities. Our evaluation shows that the model is indeed able to learn valuable information about a previously unknown system, and that more vulnerabilities can be uncovered with our approach. The model-based approach using a Decision Tree was able to find a significantly higher number of vulnerabilities than the two baseline fuzzers. Full article

(This article belongs to the Special Issue Technological Challenges and Trends in Sensor-Based Smart Production Ecosystems)

► Show Figures

Figure 1

18 pages, 3711 KB

Open AccessEditor’s ChoiceArticle

Adaptative Perturbation Patterns: Realistic Adversarial Learning for Robust Intrusion Detection

by João Vitorino, Nuno Oliveira and Isabel Praça

Future Internet 2022, 14(4), 108; https://doi.org/10.3390/fi14040108 - 29 Mar 2022

Cited by 36 | Viewed by 10609

Abstract

Adversarial attacks pose a major threat to machine learning and to the systems that rely on it. In the cybersecurity domain, adversarial cyber-attack examples capable of evading detection are especially concerning. Nonetheless, an example generated for a domain with tabular data must be realistic within that domain. This work establishes the fundamental constraint levels required to achieve realism and introduces the adaptative perturbation pattern method (A2PM) to fulfill these constraints in a gray-box setting. A2PM relies on pattern sequences that are independently adapted to the characteristics of each class to create valid and coherent data perturbations. The proposed method was evaluated in a cybersecurity case study with two scenarios: Enterprise and Internet of Things (IoT) networks. Multilayer perceptron (MLP) and random forest (RF) classifiers were created with regular and adversarial training, using the CIC-IDS2017 and IoT-23 datasets. In each scenario, targeted and untargeted attacks were performed against the classifiers, and the generated examples were compared with the original network traffic flows to assess their realism. The obtained results demonstrate that A2PM provides a scalable generation of realistic adversarial examples, which can be advantageous for both adversarial training and attacks. Full article

(This article belongs to the Topic Cyber Security and Critical Infrastructures)

► Show Figures

Graphical abstract

19 pages, 3624 KB

Open AccessArticle

Image Encryption Using Elliptic Curves and Rossby/Drift Wave Triads

by Ikram Ullah, Umar Hayat and Miguel D. Bustamante

Entropy 2020, 22(4), 454; https://doi.org/10.3390/e22040454 - 16 Apr 2020

Cited by 22 | Viewed by 4742

Abstract

We propose an image encryption scheme based on quasi-resonant Rossby/drift wave triads (related to elliptic surfaces) and Mordell elliptic curves (MECs). By defining a total order on quasi-resonant triads, at a first stage we construct quasi-resonant triads using auxiliary parameters of elliptic surfaces in order to generate pseudo-random numbers. At a second stage, we employ an MEC to construct a dynamic substitution box (S-box) for the plain image. The generated pseudo-random numbers and S-box are used to provide diffusion and confusion, respectively, in the tested image. We test the proposed scheme against well-known attacks by encrypting all gray images taken from the USC-SIPI image database. Our experimental results indicate the high security of the newly developed scheme. Finally, via extensive comparisons we show that the new scheme outperforms other popular schemes. Full article

(This article belongs to the Special Issue Entropy in Image Analysis II)

► Show Figures

Figure 1

14 pages, 2135 KB

Open AccessArticle

Quantum Image Encryption Scheme Using Arnold Transform and S-box Scrambling

by Hui Liu, Bo Zhao and Linquan Huang

Entropy 2019, 21(4), 343; https://doi.org/10.3390/e21040343 - 28 Mar 2019

Cited by 98 | Viewed by 6037

Abstract

The paper proposes a lossless quantum image encryption scheme based on substitution tables (S-box) scrambling, mutation operation and general Arnold transform with keys. First, the key generator builds upon the foundation of SHA-256 hash with plain-image and a random sequence. Its output value is used to yield initial conditions and parameters of the proposed image encryption scheme. Second, the permutation and gray-level encryption architecture is built by discrete Arnold map and quantum chaotic map. Before the permutation of Arnold transform, the pixel value is modified by quantum chaos sequence. In order to get high scrambling and randomness, S-box and mutation operation are exploited in gray-level encryption stage. The combination of linear transformation and nonlinear transformation ensures the complexity of the proposed scheme and avoids harmful periodicity. The simulation shows the cipher-image has a fairly uniform histogram, low correlation coefficients closed to 0, high information entropy closed to 8. The proposed cryptosystem provides 2²⁵⁶ key space and performs fast computational efficiency (speed = 11.920875 Mbit/s). Theoretical analyses and experimental results prove that the proposed scheme has strong resistance to various existing attacks and high level of security. Full article

(This article belongs to the Section Quantum Information)

► Show Figures

Figure 1

Search Results (5)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Saved Queries

Search Filter Reset All

Years

Feature Papers

Subjects

Journals

Article Types

Countries / Regions

Search Results (5)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI