ETDE-Net: An End-to-End Time-Domain Enhancement Network for LPI Radar Signals

Abstract

Low probability of intercept (LPI) radar signals are widely used in modern electromagnetic warfare due to their exceptional anti-interception capabilities. A defining characteristic of LPI radar signals is their low peak power, which makes them highly susceptible to being masked by additive white Gaussian noise (AWGN), causing low signal-to-noise ratio (SNR) levels challenging their detection, recognition and parameter estimation. To recover the original LPI signals from the AWGN background, this paper proposes a novel deep neural network (DNN) for end-to-end time-domain enhancement of LPI radar signals, named ETDE-Net, which consists of a feature extraction module (FEM) and a signal restoration module (SRM). The FEM acquires the representative signal features with reshape operation, channel attention and linear layers, while the SRM can recover the waveform of LPI signals by capturing the pulse information using convolutional neural networks (CNNs) and state space models (SSMs). ETDE-Net is the first DNN to achieve end-to-end time-domain LPI radar signals enhancement with superior performance compared to typical filter-based and DNN-based methods. Simulation results show that ETDE-Net has excellent signal enhancement performance at low SNRs, validating its feasibility and effectiveness.