Enhancing Autonomous System Security and Resilience With Generative AI: A Comprehensive Survey

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 109470 - 109493

Published: Jan. 1, 2024

This survey explores the transformative role of Generative Artificial Intelligence (GenAI) in enhancing trustworthiness, reliability, and security autonomous systems such as Unmanned Aerial Vehicles (UAVs), self-driving cars, robotic arms. As edge robots become increasingly integrated into daily life critical infrastructure, complexity connectivity these introduce formidable challenges ensuring security, resilience, safety. GenAI advances from mere data interpretation to autonomously generating new data, proving complex, context-aware environments like robotics. Our delves impact technologies—including Adversarial Networks (GANs), Variational Autoencoders (VAEs), Transformer-based models, Large Language Models (LLMs)—on cybersecurity, decision-making, development resilient architectures. We categorize existing research highlight how technologies address operational innovate predictive maintenance, anomaly detection, adaptive threat response. comprehensive analysis distinguishes this work reviews by mapping out applications, challenges, technological advancements their on creating secure frameworks for systems. discuss significant future directions integrating within evolving landscape cyber-physical threats, underscoring potential make more adaptive, secure, efficient.

Language: Английский

---

HIDIM: A novel framework of network intrusion detection for hierarchical dependency and class imbalance

Computers & Security, Journal Year: 2024, Volume and Issue: unknown, P. 104155 - 104155

Published: Oct. 1, 2024

Language: Английский

---

Real-Time Mouse Data Protection Method Using GANs for Image-Based User Authentication Based on GetCursorPos() and SetCursorPos() Functions

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(2), P. 977 - 977

Published: Jan. 20, 2025

In online services, password-based authentication, a prevalent method for user verification, is inherently vulnerable to keyboard input data attacks. To mitigate these vulnerabilities, image-based authentication methods have been introduced. However, approaches also face significant security challenges due the potential exposure of mouse data. address threats, protective technique that leverages SetCursorPos() function generate artificial has developed, thereby concealing genuine inputs. Nevertheless, adversaries employing advanced machine learning techniques can distinguish between authentic and synthetic data, leaving insufficiently robust. This study proposes an enhanced countermeasure utilizing Generative Adversarial Networks (GANs) produce closely emulate real input. approach effectively reduces efficacy learning-based adversarial Furthermore, counteract real-time proposed dynamically generates based on historical sequences integrates it with Experimental evaluations demonstrate classification accuracy by approximately 62%, validating its in strengthening

Language: Английский

---

A Heterogeneity-Aware Semi-Decentralized Model for a Lightweight Intrusion Detection System for IoT Networks Based on Federated Learning and BiLSTM

Sensors, Journal Year: 2025, Volume and Issue: 25(4), P. 1039 - 1039

Published: Feb. 9, 2025

Internet of Things (IoT) networks’ wide range and heterogeneity make them prone to cyberattacks. Most IoT devices have limited resource capabilities (e.g., memory capacity, processing power, energy consumption) function as conventional intrusion detection systems (IDSs). Researchers applied many approaches lightweight IDSs, including energy-based machine learning/deep learning (ML/DL)-based federated (FL)-based IDSs. FL has become a promising solution for IDSs in networks because it reduces the overhead process by engaging during training process. Three architectures are used tackle networks, centralized (client–server), decentralized (device-to-device), semi-decentralized. However, none solved while considering lightweight-ness performance at same time. Therefore, we propose semi-decentralized FL-based model IDS fit device capabilities. The proposed is based on clustering devices—FL clients—and assigning cluster head each that acts behalf clients. Consequently, number communicate with server reduced, helping reduce communication overhead. Moreover, helps improving aggregation sends average model’s weights one round. distributed denial-of-service (DDoS) attack main concern our model, since easily occurs configured three deep techniques—LSTM, BiLSTM, WGAN—using CICIoT2023 dataset. experimental results show BiLSTM achieves better suitable resource-constrained size. We test pre-trained datasets—BoT-IoT, WUSTL-IIoT-2021, Edge-IIoTset—and highest most classes, particularly DDoS attacks.

Language: Английский

---

Adversarial Sample Generation Based on Model Simulation Analysis in Intrusion Detection Systems

Electronics, Journal Year: 2025, Volume and Issue: 14(5), P. 870 - 870

Published: Feb. 22, 2025

The explosive development of artificial intelligence technology is profoundly affecting the strategic landscape cyberspace security, demonstrating enormous potential in field intrusion detection. Recent research has found that machine learning models have serious vulnerabilities, and adversarial samples derived from this vulnerability can significantly reduce accuracy model detection by adding slight perturbations to original samples. In our article, we propose a novel method called sample generation based on simulation quickly generates short period time improves model’s generalization robustness after training. Extensive experiments CICIDS-2017 dataset demonstrated consistently outperforms other current methods.

Language: Английский

---

ADHS-EL: Dynamic ensemble learning with adversarial augmentation for accurate and robust network intrusion detection

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2025, Volume and Issue: 37(1-2)

Published: March 24, 2025

Language: Английский

---

NIDS-CNNRF Integrating CNN and random forest for efficient network intrusion detection model

Internet of Things, Journal Year: 2025, Volume and Issue: unknown, P. 101607 - 101607

Published: April 1, 2025

Language: Английский

---