Application of support vector machine for the fast and accurate reconstruction of nanostructures in optical scatterometry

TL;DR

This paper presents a combined support vector machine and Levenberg-Marquardt approach to improve the speed and accuracy of nanostructure parameter reconstruction in optical scatterometry, ensuring global solution convergence.

Contribution

It introduces a novel method integrating SVM with LM algorithm to reliably select initial solutions for nanostructure parameter extraction.

Findings

Demonstrated effectiveness on silicon grating data

Achieved faster convergence to global solutions

Validated through simulations and experiments

Abstract

Nonlinear regression methods, such as local optimization algorithms, are widely used in the extraction of nanostructure profile parameters in optical scatterometry. The success of local optimization algorithms heavily relies on the estimated initial solution. If the initial solution is not appropriately selected, it will either take a long time to converge to the global solution or will result in a local one. Thus, it is of great importance to developing a method to guarantee the capture of a globally optimal solution. In this paper, we propose a method that combines the support vector machine and Levenberg-Marquardt algorithm for the fast and accurate parameters extraction. The SVM technique is introduced to pick out a sub-range in the rough ranges of parameters, in which an arbitrary selected initial solution for the LM algorithm is then able to achieve the global solution with a higher possibility. Simulations and experiments conducted on a one-dimensional Si grating and a deep-etched multilayer grating have demonstrated the feasibility and efficiency of the proposed method.