MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641048537 A) filed by Keshav Memorial Institute Of Technology, Hyderabad, Telangana, on April 16, for 'network intrusion detection system using hybrid cnn-lstm deep learning models.'
Inventor(s) include Ms. Padmavathi Bandlamudi; Mr. Sonte Venkatramana; Mr. Urumadla Sai Chetha; Mr. Guduru Sai Kiran; and Ms. Sunnam Pralasi.
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 Hybrid CNN-LSTM Based Network Intrusion Detection System is an intelligent cybersecurity framework designed to detect malicious activities in computer networks with high accuracy and reliability. Traditional intrusion detection systems often suffer from limited adaptability and high false alarm rates. To address these issues, the proposed system integrates Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to learn both spatial and temporal patterns from network traffic data.The system follows an automated pipeline that includes data preprocessing, feature selection, deep feature extraction, and attack classification. CNN layers extract spatial correlations among network features, while LSTM layers capture sequential dependencies in traffic flows. The hybrid architecture is trained and evaluated on benchmark datasets such as CIC-IDS2017, UNSW-NB15, and WSN-DS, demonstrating superior performance compared to traditional and standalone models.The proposed invention offers a scalable, robust, and efficient solution for modern network security, enabling accurate detection of known and unknown cyber threats while reducing false alarms."
Disclaimer: Curated by HT Syndication.
