MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202611065295 A) filed by Noida Institute Of Engineering And Technology Niet on May 23, 2026, for Cryptographic Framework For Privacy-Preserving Verifiable Fairness Metric Computation In Federated Learning Systems.
Inventors include Dr. Tripti Sharma; and Dr. Ahmad Neyaz Khan.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: The present invention relates to a cryptographic framework for privacy-preserving and verifiable fairness metric computation in federated learning systems. The framework includes a federated learning orchestration platform (100) coordinating institutional computing nodes (101, 102, 103), each having a local statistic computation module (104), Laplacian noise injection module (105), and Paillier encryption module (106). A secure aggregation server (200) includes a batched tree aggregation engine (201) performing hierarchical aggregation of encrypted statistics using homomorphic operations, and a verification proof validator (202) for validating zero-knowledge proofs. A fairness verification engine (300) comprises a metric computation processor (301), malicious participant detector (302) using Pedersen commitments and range proofs, and an attribute inference defense module (303). The framework enables efficient, privacy-preserving, and verifiable fairness aggregation while detecting manipulation, making it suitable for healthcare, finance, and regulatory compliance applications requiring algorithmic accountability.
Disclaimer: Curated by HT Syndication.
