An Improved Dimensionality Reduction Fitting Algorithm based on 3D Cellular Model

Plasma etching processing plays a key role in integrated circuits (IC) fabrication technology. For exploring the optimization scheme of etching process, a novel three-dimensional (3D) cellular model with improved parametric equation group is proposed to study the evolution effect in etching processing. The dimensionality reduction fitting algorithm is also introduced to decompose original 3D cellular space into two 2D planes, then calculate the angle of ion incidence quickly by least squares fitting, and improve time-space efficiency of etching processing. Additionally, the cellular selection and the incident angle calculation algorithms are improved as well. After simulation results are compared with the experimental results, the high accuracy and efficiency of our algorithms in etching processing are verified.