In this paper, improved signal processing techniques are developed for the analysis and classification of low probability of intercept (LPI) radar waveforms. The intercepted LPI radar signals are classified based on the type of pulse compression waveform. They are classified as linear frequency modulation, nonlinear linear frequency modulation, binary frequency shift keying, polyphase Barker, polyphase P1, P2, P3, P4, and Frank codes. The classification approach is based on the parameters measured from the preprocessed radar signal intercepted by electronic support (ES) or electronic intelligence (ELINT) system. First, signal embedded within the noise is estimated using Wigner Ville distribution to improve the signal-to-noise ratio (SNR). Next, features are extracted using the time-domain and frequency-domain techniques. Furthermore, parameters measured from the fractional Fourier transform are used for the classification. This type of techniques are required in various systems such as ES, electronic attack, radar emitter identification and multi input multi output (MIMO) radar applications. Extensive simulations are carried out with different LPI radar-modulated waveforms corrupted with additive white Gaussian noise of SNR up to –15 dB and impulse noise with 90% of noise density. The proposed algorithm outperforms the existing techniques of classification and can be used under strategic environment.