Optimizing Cloud IoT Applications with Machine Learning Driven Hybrid Key Management Protocols for WSN Performance
Simhadri Mallikarjuna Rao, B. Adinarayana, Kasani Vaddikasulu, Satya Prasad Dontamsetti, Peesapati Surya Venkata Ranganadh · 2024
AI-IoT integration is significant, changing the future of networking, particularly in terms of performance and security enhancements. The collection and transmission of sensitive data across secure networks is crucial for the IoT, which combines embedded systems and Networks of Wireless Sensors (WSNs). Ensuring the efficacy and security of these networks is crucial. In this research, a novel hybrid key Hybrid management protocol for Internet of Things applications using WSNs is proposed. The maximizes network performance of efficiency and security by utilizing cloud services and machine learning, specifically XGBoost. The suggested protocol creates a secure communication framework for WSNs by combining symmetric and asymmetric key management strategies. Furthermore, the centralized functions and scalability of cloud-based services enable network-wide updates and effective key distribution. The protocol offers anomaly detection, intelligent network load balancing, dynamic key management, predictive key distribution, and intelligent network load balancing by employing XGBoost-based machine learning algorithms. By using network behavior modelling and historical data analysis, the system forecasts the optimal times and locations for important updates, reducing overhead and enhancing security. Additionally, anomaly detection based on XGBoost improves the protocol's ability to recognize anomalies and potential security risks. This hybrid architecture creates a highly flexible and effective protocol for IoT-enabled WSNs by combining cloud integration, centralized key management, and AI-driven intelligence. The proposed approach performs better on the network due to proven methods and algorithms.