MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641080999 A) filed by G. Rajendra Kannammal; Dr. S. Sathya; Dr, B. Gunasundari; Dr K Phalguna Rao; and Dr, N. B. Mahesh Kumar on July 01, 2026, for Multi-Layer Privacy Shield Using Fedrated Intelligence And Homomorphic Encryption.
Inventors include G. Rajendra Kannammal; Dr. S. Sathya; Dr, B. Gunasundari; Dr K Phalguna Rao; and Dr, N. B. Mahesh Kumar.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: ABSTRACT The increasing adoption of cloud computing, loT, and distributed artificial intelligence has intensified concerns regarding data privacy and security. The present invention proposes a Multi-Layer Privacy Shield Using Federated Intelligence and Homomorphic Encryption to enable secure and privacy-preserving distributed machine learning without sharing raw user data. The framework integrates Federated Learning (FedAvg), Paillier Homomorphic Encryption (PHE), and Differential Privacy to safeguard sensitive information during model training and aggregation. Data remain on local client devices, where a deep learning model is trained locally. The resulting model updates are encrypted using the Paillier cryptosystem and securely aggregated at a federated server through the FedAvg algorithm, ensuring that raw data never leave the device. To further strengthen security, a Hyperledger Fabric blockchain layer maintains tamper-resistant records of model transactions, while an Isolation Forest-based anomaly detection module identifies malicious participants and prevents poisoned model updates from affecting the global model. The proposed framework enhances privacy, security, transparency, and trustworthiness while maintaining high learning performance. It is suitable for healthcare, finance, loT, smart city, and industrial applications requiring secure collaborative intelligence and regulatory compliance.
Disclaimer: Curated by HT Syndication.
