Frequency Estimations of Normal and Frequency Hopping Radar Signals using Wigner-Ville Distributions

This study presents a comprehensive analysis of frequency estimation for Normal Radar Signals (NRS) and Frequency Hopping Radar Signals (FHRS) using various Wigner-Ville Distributions (WVDs). The Normal Wigner-Ville Distribution (NWVD), Short-Time Wigner-Ville Distribution (STWVD), and Smoothed Wigner-Ville Distribution (SWVD) were employed to achieve high-resolution time-frequency representations. The methodology involved generating NRS and FHRS, adding noise, and applying these WVDs to estimate instantaneous power and frequency. Thereafter, frequencies per pulse were determined based on the estimated instantaneous powers and frequencies. For NRS, NWVD achieved 100% estimation accuracy at SNRs of -1 dB to -2 dB for different frequencies, while STWVD reached 100% accuracy at SNRs of -2 dB to 0 dB. SWVD outperformed both, attaining 100% accuracy even at lower SNRs of -6 dB to -4 dB. For FHRS, NWVD showed limited accuracy, with only one frequency achieving 100% at 8 dB. In contrast, STWVD provided significant improvement with all frequencies achieving 100% accuracy at higher SNRs of 0 dB to 11 dB. SWVD demonstrated superior performance, achieving 100% accuracy at SNRs as low as -7 dB. The results underscore the effectiveness of SWVD in providing accurate frequency estimations in noisy environments, highlighting its potential for airborne radar signal recognition.