MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641058086 A) filed by Parthasarathy R; P Anand Krishna; Praveen Durai; and Fancy. C on May 07, 2026, for Intelligent Sdn Intrusion Detection Using Flow Traffic Features With Ml- Based Real-Time Alerts.
Inventors include Parthasarathy R; P Anand Krishna; Praveen Durai; and Fancy. C.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: Intelligent SDN Intrusion Detection Using Flow Traffic Features with ML-Based Real-Time Alerts The SDN enables centralized control and network programmability, but it also introduces critical security risks since the SDN controller becomes a prime target for attacks. Traditional signature-based IDS fuil to identify novel, polymorphic, or evolving threats. This work proposes a flow-based intelligent intrusion detection framework using ML to enhance SDN security. The system extracts key flow traffic features such as flow duration, packet count, protocol type, byte statistics, and inter-arrival time patterns. Four machine learning classifiers- SVM, Logistic Regression, Gradient Boosting, and Random Forest-are trained and evaluated using labeled datasets containing benign traffic and five categories of attacks. Experimental results show that Random Forest achieves the highest performance with 99.31% accuracy and 0.99 precision, recall, and Fl-score. A Flask-based web application supports realtime and batch predictions through CSV uploads, includes a monitoring dashboard, and sends automated email alerts. Intrusions are detected and reported within two seconds, ensuring timely response.
Disclaimer: Curated by HT Syndication.
