MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202611110318 A) filed by Sharda University on September 14, 2026, for A Graph Neural Network-Based Dynamic Cache Management System And Method For Distributed Microservice Architectures.
Inventors include Rahul Yadav; Ms. Sheenam Naaz; and Ms. Ashu Goyal.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A computer-implemented intelligent cache management system and method for a distributed cloud-native microservice architecture are disclosed. The system constructs and continuously updates a dynamic directed runtime dependency graph representing relationships among microservices, application programming interfaces, database entities, cached data objects, user sessions, and execution paths. Node and edge features extracted from runtime metadata are processed by a Graph Neural Network to generate graph embeddings and predict future access probabilities of cached data objects. An adaptive cache utility score is determined using the predicted probability and runtime parameters including recency, access frequency, graph centrality, latency impact, and memory footprint, thereby enabling intelligent cache eviction and prefetching responsive to evolving workload conditions.
Disclaimer: Curated by HT Syndication.
