As a new area of technology, the Internet of Things (IoT) is a flagship and promising paradigm for innovating society. However, IoT-based critical infrastructures are an appealing target for cybercriminals. Such distinctive infrastructures are increasingly sensitive to cyber vulnerabilities and subject to many cyberattacks. Thus, protecting these infrastructures is a significant issue for organizations and nations. In this context, raising the cybersecurity posture of critical cyber infrastructures is an extremely urgent international issue. In addition, with the rapid development of adversarial techniques, current cyber threats have become more sophisticated, complicated, advanced and persistent. Thus, given these factors, prior to implementing efficient and resilient cybersecurity countermeasures, identification and in-depth mapping of cyber threats is an important step that is generally overlooked. Therefore, to solve cybersecurity challenges, this study presents a critical analysis of the most recent cybersecurity issues for IoT-based critical infrastructures. We then discuss potential cyber threats and cyber vulnerabilities and the main exploitation strategies adopted by cybercriminals. Further, we provide a taxonomy of cyberattacks that may affect critical cyber infrastructures. Finally, we present security requirements and some realistic recommendations to enhance cybersecurity solutions.

Malware, a lethal weapon of cyber attackers, is becoming increasingly sophisticated, with rapid deployment and self-propagation. In addition, modern malware is one of the most devastating forms of cybercrime, as it can avoid detection, make digital forensics investigation in near real-time impossible, and the impact of advanced evasion strategies can be severe and far-reaching. This makes it necessary to detect it in a timely and autonomous manner for effective analysis. This work proposes a new systematic approach to identifying modern malware using dynamic deep learning-based methods combined with heuristic approaches to classify and detect five modern malware families: adware, Radware, rootkit, SMS malware, and ransomware. Our symmetry investigation in artificial intelligence and cybersecurity analytics will enhance malware detection, analysis, and mitigation abilities to provide resilient cyber systems against cyber threats. We validated our approach using a dataset that specifically contains recent malicious software to demonstrate that the model achieves its goals and responds to real-world requirements in terms of effectiveness and efficiency. The experimental results indicate that the combination of behavior-based deep learning and heuristic-based approaches for malware detection and classification outperforms the use of static deep learning methods.

Cybercriminals are becoming increasingly intelligent and aggressive, making them more adept at covering their tracks, and the global epidemic of cybercrime necessitates significant efforts to enhance cybersecurity in a realistic way. The COVID-19 pandemic has accelerated the cybercrime threat landscape. Cybercrime has a significant impact on the gross domestic product (GDP) of every targeted country. It encompasses a broad spectrum of offenses committed online, including hacking; sensitive information theft; phishing; online fraud; modern malware distribution; cyberbullying; cyber espionage; and notably, cyberattacks orchestrated by botnets. This study provides a new collaborative deep learning approach based on unsupervised long short-term memory (LSTM) and supervised convolutional neural network (CNN) models for the early identification and detection of botnet attacks. The proposed work is evaluated using the CTU-13 and IoT-23 datasets. The experimental results demonstrate that the proposed method achieves superior performance, obtaining a very satisfactory success rate (over 98.7%) and a false positive rate of 0.04%. The study facilitates and improves the understanding of cyber threat intelligence, identifies emerging forms of botnet attacks, and enhances forensic investigation procedures.

Distributed Denial-of-Service (DDoS) attacks represent a pervasive and critical threat to autonomous vehicles, jeopardizing their operational functionality and passenger safety. The ease with which these attacks can be launched contrasts sharply with the difficulty of their detection and mitigation, necessitating advanced defensive solutions. This study proposes a novel deep-learning framework for accurate DDoS detection within automotive networks. We implement and compare multiple artificial neural network architectures, including Convolutional Neural Networks, Recurrent Neural Networks, and Deep Neural Networks, enhanced with an active learning component to maximize data efficiency. The most performant model is subsequently deployed within a federated learning paradigm to facilitate collaborative, privacy-preserving training across distributed clients. The study is evaluated against three primary DDoS attack vectors: volumetric, state-exhaustion, and amplification. Experimental results on the CIC-DDoS2019 benchmark dataset demonstrate the efficacy of our approach, achieving a 99.98% accuracy in attack classification. This indicates a promising solution for real-time DDoS detection in the safety-critical context of autonomous driving.