非高斯杂波协方差矩阵估计新方法

针对雷达目标自适应检测中的复合高斯杂波协方差矩阵估计问题,提出了一种基于杂波分组的约束迭代估计方法.该方法在迭代过程中有效利用所有辅助数据,并对最终得到的估计矩阵进行关于迹的约束.在估计的杂波分组大小与实际情况匹配的条件下,约束迭代估计方法的估计精度与杂波功率水平无关.仿真实验表明,所提出的方法对不同的杂波分组大小失配情况具有很好的鲁棒性;与已有的两种协方差矩阵估计方法相比,约束迭代估计方法能极大的提高估计精度,加快迭代过程的收敛速率,且计算量更小.

A Novel Covariance Matrix Estimator for Non-Gaussian Clutter

In radar target detection applications, for the clutter-dominated disturbance modeled as compound-Gaussian, the unknown covariance matrix usually needs estimating by utilizing the secondary data free of target signal. A novel constrained recursive estimator based on clustered-clutter is proposed. The proposed estimator exploits all secondary data fully, and introduces a constraint that the trace of the final estimated covariance matrix is equal to the number of integrated pulses. Moreover, the novel estimator is independent of the clutter texture components for match between the estimated clutter group size and the actual one. The performance assessment conducted by Monte Carlo simulation shows that, the proposed estimator is very robust to different mismatch cases;furthermore, it converges more rapidly and possesses higher estimation accuracy with less computational burden than the two existing estimators.