一种低复杂度的鲁棒自适应波束形成算法

传统基于干扰噪声协方差矩阵（interference-plus-noise covariance matrix，INCM）重构的鲁棒自适应波束形成（robust adaptive beamformer，RAB）算法在多种样本数据协方差矩阵误差和信号导向向量误差的失配环境中具有较强的鲁棒性，但目前主流的INCM重构法都是对信号和干扰的导向向量通过建立凸优化模型来估计，这带来了很高的计算复杂度。为了解决这个问题，提出了一种低复杂度的基于INCM重构的RAB算法。该算法首先将干扰信号的导向向量分解为对应标称项和误差项的和，然后通过一种子空间方法估计得到误差项的单位向量。接下来对一个Capon空间谱功率最大问题进行求解，得到误差项的模值，以此得到重构的INCM。同时利用Capon空间谱中残差噪声的存在，使用交替投影法估计得到期望信号的导向向量，最后得到所提算法的权重向量。仿真实验表明所提算法在多种误差环境下具有较强鲁棒性的同时，还具有较低的计算复杂度。

Novel robust adaptive beamforming with low complexity cost

The traditional robust adaptive beamforming(RAB) algorithm based on interference-plus-noise covariance matrix(INCM) reconstruction has strong robustness in many mismatched environment with sample data covariance matrix errors and signal steering vector errors. However, the current main methods of INCM reconstruction estimate the steering vector of signal and interference by establishing convex optimization model, which leads to high computation complexity. To solve this problem, this paper proposed an RAB algorithm based on INCM reconstruction with low computation complexity. Firstly, it divided the steering vector of interference signal into two parts, namely the nominal vector and the mismatch vector, estimated the unit vector of the mismatch vector based on a subspace method, and solved a Capon spectrum power maximization problem to reconstruct the INCM. Meanwhile, using the existence of residual noise in the Capon spatial spectrum, this paper estimated the steering vector of desired signal by the alternating projection algorithm, and finally obtained the weight vector of the proposed method. Simulations demonstrate the robustness of the proposed RAB algorithm at a cost of low computational complexity in cases of various mismatches.