Research of the algorithm for bipolar analog IC mask recognition based on pattern identification principle

The bipolar analog IC(B-aIC) holds an important position in the RF IC which is widely used in personal mobile communication. Mask patterns recognition and mask verification of B-aIC are the basis of B-aIC CAD. The authors present a method for B-aIC mask pattern recognition based on the idea of pattern recognition and an expert system (ES). The five elemental mask pattern relationships are defined, the mask description vectors are structured and the algorithm system and the recognition system are organized. The advantage of the algorithm lies in the algorithm being foreign to a specific process, and this makes the recognition fully systemized. An experiment reveals that the algorithm and the recognition are effective, accurate and reliable.