Statistical Analysis for Field-to-Line Coupling Model With Noise Based on Least Squares Approximation

In this letter, three least squares approximation (LSA) algorithms are presented and applied to the statistical analysis of a field-to-line coupling model over a nonideal ground. These three LSA algorithms are developed by combining coherence-optimal sampling (COS), Latin hypercube sampling (LHS), and Sobol sequence (Sobol), respectively, with the l1-minimum (l1-M) problem and incorporating a noise handling technique. These three LSA algorithms are then compared.