MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202411103973 A) filed by Lovely Professional University on December 28, 2024, for Adaptive Network Security System With Reinforcement Learning For Anomaly Detection.
Inventor includes Gourav Patidar.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The invention pertains to the technical field of network security systems, specifically focusing on an adaptive network security system that employs reinforcement learning for anomaly detection. The system comprises a reinforcement learning (RL) agent that monitors and analyzes network traffic in real-time, utilizing a reward- based learning model to differentiate between normal and abnormal behaviors. A data processing unit is strategically positioned at network entry and exit points to capture, preprocess, and analyze traffic data, which is then provided to the RL agent. An alert and response system, integrated with the network's security infrastructure, generates alerts and initiates responses upon anomaly detection. The system features a simulated network for training the RL agent in diverse conditions and a modular design for scalability, customization, and data integration.An integration interface ensures compatibility with security frameworks for seamless data and alert exchange.
Disclaimer: Curated by HT Syndication.
