MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641109546 A) filed by R. Kandasamy; Balasubramaniam C; Ayisha S; G. Sivarama Subramanium; R. Ragunath; P. Thirumoorthy; S. Renugadevi; P. Karunakaran; R. Gopalan; and Manoj Praveen V on September 11, 2026, for Federated Edge-Ai System For Real-Time Detection And Autonomous Mitigation Of Cyberattacks In Distributed Iot Networks.
Inventors include R. Kandasamy; Balasubramaniam C; Ayisha S; G. Sivarama Subramanium; R. Ragunath; P. Thirumoorthy; S. Renugadevi; P. Karunakaran; R. Gopalan; and Manoj Praveen V.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: The present invention discloses a federated Edge-AI system for real-time detection and autonomous mitigation of cyberattacks in distributed Internet of Things networks. The system comprises a plurality of IoT devices, multiple edge nodes, and a federated coordination server connected through secure communication channels. Each edge node collects network traffic, device behavior, authentication events, protocol information, and system logs and converts the collected data into security features. A local artificial-intelligence model analyses the features to generate an anomaly score, attack classification, and confidence value without transferring raw data outside the corresponding network domain. Edge nodes collaboratively train a shared threat-detection model through federated learning by transmitting protected model updates to the coordination server. The server validates the updates, rejects poisoned, corrupted, stale, or abnormal parameters, and aggregates accepted updates using a weighted, robust, or adaptive aggregation mechanism. A risk-assessment engine determines threat severity based on detection confidence, affected assets, service criticality, expected impact, propagation probability, and network conditions. An autonomous mitigation controller selects and executes proportionate responses including alert generation, traffic filtering, rate limiting, connection blocking, device isolation, credential revocation, route modification, quarantine, or service migration. The system maintains auditable records of detections, model versions, aggregation decisions, mitigation actions, and recovery status. Local detection and mitigation continue when connectivity to the coordination server is unavailable, and eligible updates are synchronized after communication is restored. The invention thereby provides scalable, privacy-preserving, low-latency, and coordinated cybersecurity protection for heterogeneous IoT deployments. It supports smart homes, healthcare, industry, transportation, agriculture, energy, and urban infrastructure.
Disclaimer: Curated by HT Syndication.
