New Fractional Ambiguity Function Integrated with CNN-Based Machine Learning for Signal Classification
Researchers have developed a new fractional ambiguity function (NFrAF) based on the fractional Fourier transform, generalizing the classical ambiguity function. This NFrAF demonstrates superior time-frequency resolution and localization capabilities for detecting and classifying linear frequency modulated (LFM) signals. When integrated into a convolutional neural network (CNN) framework, the NFrAF significantly improved classification accuracy compared to traditional methods like spectrograms. AI
IMPACT Introduces a novel representation for signal analysis that enhances machine learning classification accuracy.