Secure and Efficient AI-SDN-Based Routing for Healthcare-Consumer Internet of Things
Zabeeh Ullah, Nauman Ali Khan, Ikram Ud Din, Ahmad Almogren, Ayman Altameem, Mohsen Guizani · IEEE Transactions on Consumer Electronics · 2024
The advancement of communication technologies and cloud systems has led to the emergence of the Healthcare-Consumer Internet of Things (H-CIoT) as a significant domain. This emergence has transformed the traditional healthcare system into the next generation of H-CIoT, characterized by higher connectivity and intelligence. Software-Defined Networking (SDN) is currently being incorporated into H-CIoT, enabling it to meet the complex, dynamic, and heterogeneous requirements of H-CIoT networks. As the H-CIoT network expands, there is an increasing demand for secure, efficient, and optimal routing to ensure low latency and high throughput. In this paper, we propose an Artificial Intelligence (AI)-based approach that combines the strengths of Generative Adversarial Networks (GANs) and Deep Reinforcement Learning (DRL) to accurately detect anomalies in H-CIoT imbalance data and achieve optimum routing. The DRL model dynamically formulates the optimal routing policies through efficient adaptation to underlying network traffic patterns. It also comprehends the characteristics of imbalance data to enhance its routing decisions. Simulation-based results validate the effectiveness and superiority of our proposed model over OSPF routing optimization technique in term of throughput (12%), latency (20%), and the Probability of avoiding malicious minor class attacks (30%), confirming it as an outstanding suitability for the next-generation H-CIoT network.