MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202641017971 A) filed by Sr University, Warangal, Telangana, on Feb. 18, for 'machine learning-based rf signal classification for wireless communication systems.'
Inventor(s) include Dr. K. Shashidhar; and Dr. Chakradhar Adupa.
The application for the patent was published on Feb. 27, under issue no. 09/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses a machine learning-based radio frequency (RF) signal classification system and method for wireless communication environments. The system employs a complex-valued deep neural network architecture comprising complex convolutional feature extraction layers, multi-head self-attention aggregation modules, and multi-task output heads for simultaneous signal type classification and signal parameter estimation from raw in-phase and quadrature (I/Q) sample data. A Bayesian uncertainty quantification module provides calibrated classification confidence estimates through Monte Carlo dropout inference and out-of-distribution anomaly detection, enabling reliable operation in dynamic and contested spectrum environments. A few-shot adaptation module based on model-agnostic meta-learning enables rapid extension of the classifier to novel signal types using minimal labeled examples, while a continual learning framework preserves classification performance on established signal classes as the operational environment evolves. A hardware compression pipeline applying structured pruning and mixedprecision quantization enables deployment on resource-constrained embedded platforms with submillisecond inference latency. The system supports distributed spectrum sensing through a multisensor fusion architecture employing Dempster-Shafer evidence combination. Applications include cognitive radio spectrum management, electronic intelligence, interference mitigation, and dynamic spectrum access in 4G, 5G, and next-generation wireless networks."
Disclaimer: Curated by HT Syndication.
