BROWSE
Detail
Conference Paper
Analyzing the Effects of Raised-Cosine Tapering on WVD-CNN Detection of Pulsed Chirp Signals
Cited - time in 
Share 
Share
- Authors
- Yunbae Kim, Jonghun Yoon, Bongsu Kim
- Issue Date
- 2025-12
- Citation
- Asia-Pacific Microwave Conference (APMC) 2025, pp.1-3
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- This paper investigates the application of the Wigner-Ville Distribution (WVD) for detecting pulsed chirp radar signals in spectrum sharing scenarios. Using a publicly available dataset containing synthetic linearly frequency modulated signals, we construct WVD-based time–frequency images and classify them using convolutional neural networks. Compared to spectrogram inputs, WVD yields higher detection accuracy, particularly when full resolution is preserved. To evaluate robustness in constrained settings, we also examine reduced-resolution inputs, which result in only minor degradation. Furthermore, we analyze the impact of raised-cosine window tapering on detection performance, showing how different levels of spectral leakage and mainlobe attenuation affect classification. These findings offer insights into designing efficient sensing pipelines for radar detection in dynamic spectrum environments.
- Keyword
- Wigner-Ville distribution, radar detection, convolutional neural network, raised-cosine window, LFM signal
- KSP Keywords
- Chirp signals, Convolution neural network(CNN), Cosine window, Detection accuracy, Frequency modulated, LFM signal, Linear frequency modulation(LFM), Modulated signals, Wigner-Ville distribution(WVD), chirp radar, detection performance