Search

In this paper, an algorithm for extracting and localizing a radar pulse in a noisy environment is described. The algorithm combines two powerful tools: wavelet denoising and the short-time Fourier transform (STFT) analysis with statistical-based threshold. We aim to detect radar pulses transmitted by any radar in blind mode regardless of the intra-pulse modulation and parametric features. The use of the proposed technique makes the detection and localization of radar pulses possible under very low signal-to-noise ratio conditions (−18 dB), which leads to a reduction of the required signal power or alternatively extends the detection range of radar systems. Radar classes pattern-based analysis is used in blind mode to decrease the probability of false alarm. [ABSTRACT FROM AUTHOR]