MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641093919 A) filed by Vardhaman College Of Engineering on August 03, 2026, for Autonomous Cyber Defense System Using Reinforcement Learning-Based Intrusion Response.
Inventors include Dr. Swati Sucharita; Ms. Budavarapu Pravallika; Mr. Allama Prabhu; Ms. Vijaylaxmi; Mr. Majjari Sudhakar; and Mr. Sai Madhav Raj Alakuntla.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: Autonomous Cyber Defense System Using Reinforcement Learning-Based Intrusion Response is the proposed invention. The proposed invention provides an autonomous cyber defence system using Deep Reinforcement Learning based on Proximal Policy Optimisation (PPO) for intelligent intrusion response and adaptive cyber threat mitigation. The framework includes distributed telemetry collection, AI-powered threat detection, cyber state encoding, a PPO policy optimisation agent, adaptive response orchestration, a digital twin cyber simulator, an explainable security module, and a continual learning repository. The PPO agent consumes security events from network, endpoint, cloud, and user behaviour sources and translates them into a comprehensive cyber state representation. The agent then uses the representation to decide on the optimal mitigation actions. Some of the selected responses are intelligent firewall reconfiguration, software defined networking policy updates, endpoint isolation, privilege restriction, deception deployment, cloud workload migration and automated recovery. A reward evaluation mechanism constantly updates the reinforcement learning policy based on containment success, response latency, system availability and operational impact. The invention provides a scalable, self-learning, explainable and resilient cyber defence framework that can autonomously mitigate advanced and zero-day cyber assaults in modern distributed computing environments.
Disclaimer: Curated by HT Syndication.
