DoA estimation using channel state information (CSI) is an important technology for localization in Wi-Fi network. Existing direction of arrival (DoA) estimation algorithms such as multiple signal classification (MUSIC) algorithm may provide feasible DoA estimation accuracy, if the array gain-phase error can be well calibrated in commodity Wi-Fi devices. However, for multipath propagation environment, classical subspace-based calibration methods cannot be directly deployed, due to the coupling effect between gain-phase error parameters and DoA parameters. Therefore, most works deploy wired calibration methods with specialized devices, which also require professional and trivial deployments. This paper offers an iterative pairwise antenna gain-phase error calibration method, which is easy deployment and low cost. Furthermore, we combine spatial smoothing technique with pairwise antenna calibration schemes to guarantee calibration accuracy and parameters decoupling, respectively. Experimental results in realistic indoor environments to verify the effectiveness of the proposed calibration method.

使用信道状态信息（CSI）进行DoA估计是Wi-Fi网络中定位的一项重要技术。如果可以在商用Wi-Fi设备中很好地校准阵列增益相位误差，则诸如多信号分类（MUSIC）算法之类的现有到达方向（DoA）估计算法可能会提供可行的DoA估计精度。然而，对于多径传播环境，由于增益相位误差参数和DoA参数之间的耦合效应，传统的基于子空间的校准方法无法直接部署。因此，大多数工作使用专用设备来部署有线校准方法，这也需要专业且琐碎的部署。本文提供了一种迭代的成对天线增益相位误差校准方法，该方法易于部署且成本低廉。此外，我们将空间平滑技术与成对天线校准方案相结合，以分别确保校准精度和参数去耦。在现实的室内环境中进行的实验结果验证了所提出的校准方法的有效性。