In radar systems, antenna arrays acquire direction-dependent information to localize targets or create images of the environment. However, because of unknown complex amplitudes per channel and mutual coupling, a calibration is necessary for good performance. Common calibration approaches measure to targets under known angles in a multipath-free far-field environment, which often can only be provided by anechoic chambers. Therefore, this process is not suitable for low-cost and frequent calibration. To overcome these limitations, this paper proposes a novel calibration approach using an unknown but sparse target scene. Multiple measurements at known relative positions of the radar are combined to a synthetic aperture. Then, the full mutual coupling matrix is estimated simultaneously with the unknown target scene. The method requires neither target position information, far-field conditions, nor an anechoic chamber because multipath propagation can be suppressed easily for targets located in the near-field. The proposed calibration approach is validated by measurements of a commercial 77GHz radar. The performance is evaluated by comparing the achieved image quality using the calibration results of this work and the radar’s ex-factory calibration data. The proposed novel calibration procedure improves the image quality, while considerably lowering the demands on the calibration measurements and environment.

中文翻译：

---

使用稀疏场景进行天线阵列校准

在雷达系统中，天线阵列获取与方向有关的信息以定位目标或创建环境图像。但是，由于每个通道的未知复振幅和相互耦合的原因，必须进行校准才能获得良好的性能。常见的校准方法是在无多路径的远场环境中以已知角度测量目标，而通常只能由消声室提供。因此，此过程不适用于低成本且频繁的校准。为了克服这些限制，本文提出了一种使用未知但稀疏的目标场景的新颖校准方法。将雷达已知相对位置的多次测量结果合并到一个合成孔径中。然后，与未知目标场景同时估计完整的互耦合矩阵。该方法既不需要目标位置信息，也不需要远场条件，也不需要消声室，因为对于位于近场的目标可以轻松地抑制多径传播。通过对商用77GHz雷达的测量验证了所提出的校准方法。通过使用这项工作的校准结果与雷达的出厂校准数据对获得的图像质量进行比较，来评估性能。所提出的新颖的校准程序改善了图像质量，同时大大降低了对校准测量和环境的要求。通过使用这项工作的校准结果与雷达的出厂校准数据对获得的图像质量进行比较，来评估性能。所提出的新颖的校准程序改善了图像质量，同时大大降低了对校准测量和环境的要求。通过使用这项工作的校准结果与雷达的出厂校准数据对获得的图像质量进行比较，来评估性能。所提出的新颖的校准程序改善了图像质量，同时大大降低了对校准测量和环境的要求。