A Novel Approach for Aerial Moving Target Fusion-Based Detection Method

Yu Li,Jinming Chen,Weiwei Wang,Yingbin Wang,Xiaoshuang Jiao,Xinyu Su,Chongdi Duan

DOI: https://doi.org/10.1109/cipcv61763.2024.00030

2024-01-01

Abstract:Notice that the illuminators of opportunity (IOs) of a passive distributed radar network (DRN) are from third-party sources and hence often out of control to the surveillance, and that the signal processing architecture in an active/passive DRN is somewhat sophisticated. In this paper, a contrast-enhanced detection method is proposed to detect aerial moving targets using an active LEO satellite constellation. The target echoes over a long-time observation period are integrated within individual radars first, and then integrated target signals among node radars are fused to enhance the aerial moving target detection performance. In the integration period, the observation period is further divided into several frames, with the employment of coherent integration inside the frame and noncoherent integration among frames. Especially, by applying Taylor-series expansion to the range difference is more accurate than to the satellite range itself, a range difference-based positive and negative second-order KT method is developed to improve the target motion compensation. In the detection period, we propose a contrast-enhanced detection method via DFR variance and smooth spatial filtering function weighting to identify the fake target by the property of approximately equal radar DFRs and to refine the potential target regions to make the detection more efficient, respectively. Simulation results verify the effectiveness of the proposed method.