RAKSHAM: Responsive approach to Knock‐off scavenging hackers and attack mitigation

Abstract

AbstractWith the increasing interconnectivity of devices and systems in the Internet of Things (IoT) landscape, the importance of cybersecurity has grown significantly. This work presents a comprehensive solution that combines deceptive techniques and AI‐based detection to strengthen the security of IoT networks against sophisticated cyber threats. Deceptive techniques, such as the utilization of honeypots, deceptive IoT devices, and deceptive network traffic, are deliberately employed with the aim of misleading potential attackers. However, the effectiveness of these systems is often challenged by highly skilled attackers and the complexities associated with allocating resources. To address these challenges, “RAKSHAM” is a unique approach to improving the security of IoT networks. It combines deceptive techniques with AI‐based detection mechanisms, minimizing false positives and operational outages in IoT networks. This approach utilizes Eigenvector Centrality (EVC) to identify nodes with a higher risk level. It also evaluates the performance of an AI‐based attack detection model. The AI model, which utilizes a Convolutional Neural Network (CNN), demonstrates exceptional efficacy in the detection of attacks with 99% accuracy and a minimal attack success probability rate of 0.10. RAKSHAM offers a pragmatic and efficient solution for securing IoT networks by accurately detecting attacks and minimizing false positives and operational outages.