Robust Detection of Micro-UAS Drones with L-Band 3-D Holographic Radar
Robust Detection of Micro-UAS Drones with L-Band 3-D Holographic Radar
复制标题
使用 L 波段 3D 全息雷达对微型 UAS 无人机进行稳健检测
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
C. Baker
中科院分区：
文献类型：
--
作者：
M. Jahangir;C. Baker
无人机系统 (UAS) 是无人驾驶飞机（无人机），与有人驾驶飞机相比，其特点是雷达截面非常小、运动轨迹相对较慢以及工作高度较低。直接的后果是它们更难以检测和跟踪。在传统的二维扫描雷达中，这种情况更加严重，传统的二维扫描雷达很难在同时具有短重访时间和高多普勒分辨率的冲突需求之间找到折衷方案。在这里，我们使用全息雷达TM (HR)，它采用 2D 天线阵列和适当的信号处理来创建多波束、3D、广域、凝视监视传感器，能够实现高检测灵敏度，同时提供精细的多普勒分辨率和几分之一秒的更新率。在整个搜索范围内持续关注目标的能力使 HR 能够实现一定程度的处理增益，足以在复杂的静止和移动杂波背景下检测特征非常低的目标，例如微型无人机。本文介绍的试验结果显示使用 32 x 8 元件 L 波段接收器阵列检测小型六轴飞行器 UAS。必要的高检测灵敏度意味着可以检测和跟踪许多其他小型移动目标，而鸟类是杂波的主要来源。为了克服这个问题，需要进一步的处理阶段来区分无人机和其他移动物体。这里，机器学习决策树分类器用于拒绝非无人机目标，从而几乎完全抑制错误轨迹，同时保持无人机的高检测概率。
Unmanned Aerial Systems (UAS) are pilotless aircraft (drone) and are characterized by having very small radar cross-sections, relatively slow motion profiles and low operating altitudes compared with manned aircraft. As a direct consequence they are considerably more difficult to detect and track. This is exacerbated in traditional 2-D scanning radar which struggle to find a compromise between the conflicting needs to simultaneously have short re-visit times and high Doppler resolution. Here, we use Holographic RadarTM (HR) that employs a 2-D antenna array and appropriate signal processing to create a multibeam, 3-D, wide-area, staring surveillance sensor capable of achieving high detection sensitivity, whilst providing fine Doppler resolution with update rates of fractions of a second. The ability to continuously dwell on targets over the entire search volume enables HR to achieve a level of processing gain sufficient for detection of very low signature targets such as miniature UAS against a background of complex stationary and moving clutter. In this paper trials results are presented showing detection of a small hexacopter UAS using a 32 by 8 element L- Band receiver array. The necessary high detection sensitivity means that many other small moving targets are detected and tracked, birds being a principle source of clutter. To overcome this a further stage of processing is required to discriminate the UAS from other moving objects. Here, a machine learning decision tree classifier is used to reject non- drone targets resulting in near complete suppression of false tracks whilst maintaining a high probability of detection for the drone.
影响因子：
1.9
作者：
C. Clemente;A. Balleri;K. Woodbridge;J. Soraghan
通讯作者：
C. Clemente;A. Balleri;K. Woodbridge;J. Soraghan