A Pulse Accumulation Method for LPI Radar Emitter Signal Recognition

Radar emitter signal recognition is a crucial means to distinguish unknown radars, and it depends on the high quality of received signals. However, signals of Low probability of intercept (LPI) radars are easily interfered with by noise, resulting in poor signal quality and low recognition accuracy. We propose a pulse accumulation method to improve the quality of pulse signals for LPI radar emitter signal recognition. Firstly, the pulse accumulation problem is described. Then we propose the time-domain alignment iterative method (TAIM) to improve the effect of pulse accumulation. Finally, time-frequency images of the accumulated signals are input to the deep residual network for recognition to verify the effect of the proposed method. This approach enhances the effect of pulse accumulation by employing the proposed TAIM, and ultimately improves the performance of signal recognition. Simulation results show that the proposed pulse accumulation method can improve the recognition accuracy by 7% more than that of single-pulse signals at −6dB.