MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202621028174 A) filed by Rabindranath Tagore University, Raisen, Madhya Pradesh, on March 10, for 'advanced intrusion detection system based on ae-cnn-mhatt-bigru and dualfusion-convbilstm architectures.'
Inventor(s) include Ramnaresh Sharma; Dr. Jitendra Singh Kushwah; Dr. Pritaj Yadav; Dr. Sanjeev Kumar Gupta; and Dr. Pratima Gautam.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The current disclosure pertains to a system and method for detecting, analyzing, predicting, and mitigating network intrusions within a communication environment through an adaptive and intelligent multi-layer architecture. The system has a multi-layer traffic acquisition module that can collect packet-level and flow-level network data, a preprocessing module that can normalize and structure traffic attributes, and a quantum-inspired feature optimization module that can do probabilistic feature selection and dimensionality reduction. The system also has a blockchain-secured traffic provenance module that checks the data's authenticity and stops bad actors from changing it. A self-evolving hybrid deep learning engine is set up to do representation learning, feature extraction, attention-based anomaly prioritization, and temporal sequence analysis for finding intrusions. The system also includes a neuro-symbolic threat reasoning module, a federated collaborative learning framework, and an explainable forensic intelligence module to improve detection accuracy, scalability, privacy protection, and real-time threat mitigation. This makes it a more advanced and reliable network security solution."
Disclaimer: Curated by HT Syndication.
