VLSI Circuit Measurement System Analysis Based on Random Fuzzy Variables

Data variation is one way of representing the test quality of very large scale integration(VLSI) circuits using automatic test equipment(ATE), especially by the quantitative parameters. Deep submicron integration may cause difference of IC products increase. In order to prevent part variation induced analysis accuracy drop, test fixture, test board and test program are combined and crossly designed with the three-way ANOVA is studied, instead of common operator factor. Based on the 3-way model, with triangular fuzzy number as input and trapezoidal fuzzy number as output, the MSA indices are calculated and verified by a typical memory test.