Test Pattern Sampling Methodology for Model Tuning Efficiency and Robustness Improvement

Optical proximity effect correction (OPC) is an important part of the integrated circuit fabrication process to prevent the device electrical properties from being affected by pattern distortion caused by optical proximity effect. Accurate correction requires high simulation accuracy and robustness of the OPC model, which needs high coverage of the test patterns over the actual layout. However, in the industry, the test patterns are selected based on experience of engineers, which has the problem of a large number of similar patterns and insufficient coverage over the actual layout. Therefore, in this paper, a sampling method is proposed, which uses clustering algorithms to classify the geometric and optical features of the test patterns. According to the classification results, redundant patterns are removed and the coverage of the test patterns is increased, which will shorten the modeling period and improve the model robustness.