Intellectual Property India Publishes Patent Application for An Intelligent Machine Learning-Based System For Anomaly Detection In Network Traffic

MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641112703 A) filed by Mr. Rajakumar Perumal; Dr. Sivakumar Madeshwaran; Dr. Nitesh Gupta; Dr. Vinod Kumar Yadav; Mr. Anurag Shrivastava; Ms. Sonam Kumari; Ms. Indervati; Ms Parvesh; Mr. Yashwant Soni; and Mr. P. Rajakumar on September 20, 2026, for An Intelligent Machine Learning-Based System For Anomaly Detection In Network Traffic.

Inventors include Mr. Rajakumar Perumal; Dr. Sivakumar Madeshwaran; Dr. Nitesh Gupta; Dr. Vinod Kumar Yadav; Mr. Anurag Shrivastava; Ms. Sonam Kumari; Ms. Indervati; Ms Parvesh; Mr. Yashwant Soni; and Mr. P. Rajakumar.

The application for the patent was published on September 25, 2026, under issue no. 39/2026.

Abstract: The present system discloses an intelligent machine learning-based system for detecting anomalies in network traffic. The system comprises a network traffic acquisition module, preprocessing module, feature extraction module, behavioral profiling module, machine learning engine, anomaly scoring module, contextual correlation module, explainability module, response orchestration module, and feedback module. Network traffic acquired from one or more observation points is processed to generate spatial, temporal, statistical, and behavioral features. A dynamically maintained behavioral profile is generated to represent expected network behavior. One or more machine-learning models analyze the extracted features and generate a detection output. An anomaly score is generated by combining machine- learning output with behavioral, temporal, and contextual information. Detected anomalies may be classified and accompanied by an explanation identifying contributing traffic characteristics. Based on the anomaly score and contextual information, the system can generate alerts or initiate network security responses including traffic restriction, endpoint isolation, or enhanced monitoring. Feedback associated with detected events is used to update the behavioral profile and/or machine-learning model, thereby enabling continuous adaptation to changing network conditions. The invention is applicable to enterprise networks, cloud networks, data centers, IoT networks, industrial networks, campus networks, and telecommunications networks.

Disclaimer: Curated by HT Syndication.