稀疏脉冲反演实际上是从含有噪声的雷达数据单道中计算出具有稀疏分布特性的反射系数。文中以范数约束为基础,提出L1 范数约束求解的方法,并且采用迭代重加权最小二乘法(IRLS)的优化算法,该算法具有精度高、速度快的优点,对穿墙侦查雷达数据处理后可以有效提高分辨率,获得反射系数,有助于进一步求得墙体的厚度和介电常数。最后针对不同数据的实验结果验证了算法的有效性。

The sparse reflectivity can be estimated from the noisy A-scan by sparse spike inversion. This paper is used L1 norm constrained method, and this method is optimized by Iteratively reweighted least squares (IRLS) which is high in terms of resolution and fast. The resolution can be improved obviously and the reflectivity can be obtained after processed the data of through wall detecting radar used this algorithm, then the thickness and dielectric constant can be obtained. The different experimental result is presented to illustrate the effectiveness of the algorithm at last.

TN957.52