Automatic Recognition of General LPI Radar Waveform Using SSD and Supplementary Classifier

For low probability of intercept (LPI) radars, frequency-modulated and phase-modulated continuous waveforms are widely used because of their low peak power compared to that of pulse waves (PW). However, there has been a limited number of studies on recognizing continuous wave (CW) LPI radar, in spite of its importance and popularity. In this paper, in order to recognize both PW and CW LPI radar waveforms, we propose an LPI radar waveform recognition technique (LWRT) based on a single-shot multi-box detector (SSD) and a supplementary classifier. It is demonstrated with Monte Carlo simulations that the proposed LWRT achieves classification performance similar to that of the current LWRT with the highest classification performance for PW LPI radar waveforms, even without the prior condition used in existing LWRTs. For CW LPI radar waveforms, on the other hand, with the combination of the SSD and the supplementary classifier, the proposed LWRT achieves extraordinary recognition performance for all 12 LPI radar modulation schemes (i. e., BPSK, Costas, LFM, Frank, P1, P2, P3, P4, T1, T2, T3, and T4) considered in the literature.