MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641089742 A) filed by Vuddanti Sandeep; P Ajay Sai Kiran; and L N Sastry Varanasi on July 23, 2026, for A Fidelity-Certified Digital Twin System And Method For Cross-Layer Cyber-Physical Attack Generation And Dataset Certification In Vehicle-To-Grid Networks.
Inventors include Vuddanti Sandeep; P Ajay Sai Kiran; and L N Sastry Varanasi.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The present invention discloses a fidelity-certified Digital Twin system and method for cross-layer cyber-physical attack generation and dataset certification in Vehicle- to-Grid (V2G) networks, designed to provide a secure, realistic, and standardized platform for cybersecurity evaluation in intelligent electric mobility infrastructures. The proposed system comprises a physical-grid modelling module, a communication-network simulation module, a protocol and control module, an attack injection and orchestration engine, a telemetry synchronization module, an evasion evaluation module, a fidelity certification module, and a certified dataset generation engine configured to emulate coordinated cyber physical interactions across communication, protocol, network, and power system layers. A structured attack tuple represented by (A=(phi, pi, eta, epsilon)), corresponding respectively to communication/session manipulation, protocol/control manipulation, network disturbance, and physical- grid impact, is generated and synchronously injected into the Digital Twin to reproduce realistic attack propagation scenarios. The generated attacks are evaluated using bad-data detection residual analysis and physical feasibility verification to classify attack instances as evasive, detectable, or infeasible while preserving operational realism. Subsequently, simulated telemetry is compared with reference system behaviour using multidimensional fidelity metrics, including Wasserstein distance and other statistical similarity measures, to certify the generated datasets prior to deployment. The invention further supports Hardware-in-the-Loop (HIL) validation, co- simulation frameworks, and intelligent visualization for comprehensive cybersecurity experimentation without affecting live charging infrastructure. The certified datasets generated by the disclosed system facilitate the development, validation, benchmarking, and explainability of artificial intelligence and machine learning-based intrusion detection, anomaly detection, resilient control, cyber-forensic analysis, and decision-support systems for next generation Vehicle-to-Grid ecosystems, thereby significantly enhancing the reliability, reproducibility, and cyber resilience of smart charging and intelligent power network operations.
Disclaimer: Curated by HT Syndication.
