LCB-CV-UNet: Enhanced Detector for High Dynamic Range Radar Signals
Yanbin Wang, Xingyu Chen, Yumiao Wang, Xiang Wang, Chuanfei Zang, Guolong Cui, Jiahuan Liu
TL;DR
This work tackles the challenge of detecting radar targets under high dynamic range HDR signals, where strong targets overwhelm weak ones. It introduces LCB-CV-UNet, a lightweight detector that fuses a Logarithmic Connect Block for phase coherent HDR processing with a Complex-Valued UNet backbone and a CFAR-style decider, along with Dual Hybrid Dataset Construction to generate balanced semi synthetic HDR data. The Logarithmic Connect Block achieves $O(N)$ complexity with $11N$ MACs while preserving phase, and DHDC creates training sets that mix strong and weak targets across controllable distributions to improve HDR robustness. Across simulations and urban-vehicle real data, the method yields about a 1% increase in detection probability with under 0.9% additional computational load and a notable 5% boost for 11--13 dB SNR cases, validating practical edge-side deployment for HDR radar sensing.
Abstract
We propose the LCB-CV-UNet to tackle performance degradation caused by High Dynamic Range (HDR) radar signals. Initially, a hardware-efficient, plug-and-play module named Logarithmic Connect Block (LCB) is proposed as a phase coherence preserving solution to address the inherent challenges in handling HDR features. Then, we propose the Dual Hybrid Dataset Construction method to generate a semi-synthetic dataset, approximating typical HDR signal scenarios with adjustable target distributions. Simulation results show about 1% total detection probability improvement with under 0.9% computational complexity added compared with the baseline. Furthermore, it excels 5% over the baseline at the range in 11-13 dB signal-to-noise ratio typical for urban targets. Finally, the real experiment validates the practicality of our model.