基于最近邻域的弹道多目标分辨及micro-Doppler提取

Ballistic multi-target resolution and micro-Doppler extraction based on the nearest neighborhood

针对尺寸相近的弹道多目标分辨及micro-Doppler提取问题,文中提出了一种新的基于"选择"的弹道多目标分辨及micro-Doppler提取方法.在滑动散射模型的基础上,根据弹道目标micro-Doppler变化率的变化趋势,首先对时频谱中各点的最近邻域进行多尺度编码,通过分层处理实现多目标的分辨,然后对各分层区域进行方位编码序列配准,运用零点回溯的方法提取出目标的micro-Doppler信息.仿真结果表明,该方法能够有效地克服多散射点模型交叉区域对多目标分辨及曲线分离的干扰作用,适用于多种微动目标散射模型,且抗噪性好,较好地实现了弹道多目标分辨及micro-Doppler提取.

To implement ballistic multi-target resolution and micro-Doppler extraction with similar structures,in this paper, we propose a novel algorithm based on code selection. Based on the sliding scattering model, the nearest neighborhood of the points in the time-frequency spectrum is first disposed with multi-scale codes by analyzing the variations in the micro-Doppler rate of change, and multiple targets are then effectively distinguished by stratification processing. We obtain the azimuth-code sequences on the stratified region for advanced registration, and we obtain the micro-Doppler information based on the zero-point backtracking method. The simulation result indicates that the method is capable of anti-noise, and is suitable for a variety of micro-motion scattering models. However, above all, it can effectively overcome the interference of intersection regions for multi-target resolution and micro-Doppler curve separation in multi-point scattering models, thus implementing the multi-target resolution and micro-Doppler extraction well.