MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202421098815 A) filed by Amity University Madhya Pradesh on December 13, 2024, for An Iot Based System With Sdn-Based Network Traffic Classification Using Optimized Machine Learning Techniques.
Inventors include Nidhi Bajpai; Madhavi Dhingra; and Nisha Chaurasia.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The proposed system represents an IoT-based system integrated with a software-defined networking (SDN) framework for intelligent network traffic classification using optimized machine learning techniques. IoT traffic is captured as flow-level metadata and processed through a structured pipeline comprising Bag-of-Words preprocessing, principal component analysis–based feature extraction, and Egret Swarm Optimization–based feature selection. Classification is performed using hybrid models ABCRFA, HXGBHPO, and RVMSBO to identify benign, malicious, and priority traffic. The SDN controller utilizes classifier outputs to install flow rules in real time, improving routing, security, and QoS management. The system achieves enhanced accuracy, specificity, precision, recall, and F1-score while reducing controller latency and resource overhead, enabling scalable deployment in heterogeneous SDN–IoT environments.
Disclaimer: Curated by HT Syndication.
