Explainable deep learning for attack intelligence and combating cyber-physical attacks

Article Computer Science, Information Systems

X-IIoTID: A Connectivity-Agnostic and Device-Agnostic Intrusion Data Set for Industrial Internet of Things

Muna Al-Hawawreh et al.

Summary: The Industrial Internet of Things (IIoT) is a high-value target for cyber attacks, and developing security solutions that fit its requirements is challenging due to the lack of accurate data. To address this, we propose X-IIoTID, an intrusion data set for IIoT that includes multi-view features of connectivity protocols, device activities, attack types, and protocols.

IEEE INTERNET OF THINGS JOURNAL (2022)

Add to Collection

Article Engineering, Electrical & Electronic

An Online Model to Minimize Energy Consumption of IoT Sensors in Smart Cities

Muna Al-Hawawreh et al.

Summary: The Internet of Things (IoT) is facing challenges related to data redundancy and energy consumption. To address this, we propose an AI-powered solution that utilizes autocorrelation and deep reinforcement learning to make smart decisions about transmitting data, thereby reducing data redundancy and minimizing sensor power consumption.

IEEE SENSORS JOURNAL (2022)

Add to Collection

Article Computer Science, Information Systems

Robust Botnet DGA Detection: Blending XAI and OSINT for Cyber Threat Intelligence Sharing

Hatma Suryotrisongko et al.

Summary: In this study, the researchers investigated 12 years of DNS query logs from their campus network and discovered the presence of malicious botnet domain generation algorithm (DGA) traffic. They found that DGA-based botnets are difficult to detect using traditional cyber threat intelligence systems and proposed the use of AI/machine learning-based systems for improved detection. The researchers developed a model to detect DGA-based traffic using statistical features and discussed the expansion of CTI using computable CTI paradigm. They also explored methods to improve the explainability of the model outputs using explainable AI (XAI) and open-source intelligence (OSINT). Experimental results showed the effectiveness of their models and the superiority of their random forest model against adversarial attacks compared to other deep learning models. The researchers demonstrated the potential of XAI-OSINT blending in improving trust for CTI sharing and validating computable CTI paradigm.

IEEE ACCESS (2022)

Add to Collection

Article Computer Science, Information Systems

Asynchronous Peer-to-Peer Federated Capability-Based Targeted Ransomware Detection Model for Industrial IoT

Muna Al-Hawawreh et al.

Summary: This paper introduces a targeted ransomware detection model tailored for IIoT systems, utilizing Asynchronous Peer-to-Peer Federated Learning and Deep Learning techniques to effectively detect known and unknown attacks in these systems with their heterogeneous and distributed nature.

IEEE ACCESS (2021)

Add to Collection

Article Computer Science, Artificial Intelligence