MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641097806 A) filed by Mohan Babu University on August 13, 2026, for Federated Learning-Based Privacy-Preserving Intrusion Detection System For Internet Of Things Networks.
Inventors include Dr. C. Sushama; Ms. Kooraparthi Harshitha; Mr. Kummetha Mallikarjuna Reddy; Mr. Vennapusa Vivek Reddy; and Mr. T Praveen Kumar Naidu.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: The present invention relates to a Federated Learning-Based Privacy-Preserving Intrusion Detection System for Internet of Things Networks, configured to provide secure and collaborative intrusion detection without transmitting raw data from IoT devices to centralized servers. The system comprises multiple IoT devices, local intrusion detection modules, a federated aggregation server, secure communication channels, and a global model update mechanism. Each IoT device independently trains a local intrusion detection model using locally available data and transmits only encrypted model parameters to the aggregation server. The aggregation server securely combines the received model updates to generate an optimized global intrusion detection model, which is redistributed to participating devices for subsequent learning iterations. The proposed framework preserves user privacy, minimizes communication overhead, improves intrusion detection accuracy, and eliminates dependence on centralized data storage. The modular architecture supports scalable deployment across heterogeneous IoT environments including healthcare, industrial automation, smart cities, transportation systems, cloud-edge infrastructures, and other distributed cyber-physical networks while providing adaptive protection against evolving cyber threats.
Disclaimer: Curated by HT Syndication.
