TOWARDS DETECTION ATTRIBUTION OF CYBERATTACKS IN IOT ENABLED CYBER PHYSICAL SYSTEM
International Research Journal of Modernization in Engineering Technology and Science · 2025
The security of Internet of Things (IoT) integrated cyber-physical systems presents unique challenges that traditional IT security measures cannot adequately address.This research proposes a dual-layer ensemble framework specifically engineered for cyber threat detection and source identification in industrial control environments.The first layer employs a hybrid approach combining decision trees with advanced deep learning representation models to identify threats in imbalanced industrial datasets.The second layer utilizes ensemble neural networks for precise attack source attribution.Performance evaluation conducted on authentic gas pipeline and water treatment facility datasets demonstrates superior accuracy compared to existing methodologies while maintaining computational efficiency.The increasing deployment of IoT devices in cyber-physical infrastructures has enhanced operational capabilities but simultaneously created expanded vulnerability surfaces for sophisticated cyber threats.Current research predominantly addresses threat detection while neglecting the critical aspect of attack attribution -determining the origin and responsible entities behind security breaches.This work addresses this gap by developing comprehensive detection and attribution mechanisms suitable for distributed, resource-limited IoT-CPS environments.