MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202641043437 A) filed by Mrs. B. Arulmozhi; Dr. J. I Sheeba; and Dr. S. Pradeep Devaneyan, Pondicherry, India, on April 5, for 'a system and method for adaptive federated learning with privacy-preserving distributed intelligence in heterogeneous network environments.'
Inventor(s) include Mrs. B. Arulmozhi; Dr. J. I Sheeba; and Dr. S. Pradeep Devaneyan.
The application for the patent was published on April 17, under issue no. 16/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a system and method for adaptive federated learning configured to enable privacy-preserving distributed intelligence across heterogeneous network environments. The system comprises a coordination server, a plurality of edge devices, and an adaptive orchestration module configured to dynamically select participating devices based on parameters including computational capability, network conditions, and data quality. Each edge device performs local training using private data and generates model updates without sharing raw data. A privacy module applies transformation techniques including noise injection and encryption to protect sensitive information. The coordination server includes a secure aggregation engine for combining updates while maintaining confidentiality. An adaptive weighted aggregation mechanism improves model convergence by assigning weights based on reliability metrics. The system further incorporates communication optimization through compression and asynchronous updates, thereby enabling efficient, scalable, and secure distributed learning suitable for real-time applications."
Disclaimer: Curated by HT Syndication.
