隨著工業時代之演進,IC產業之發展日益繁雜,製程技術之日新月異致使積體電路內之電晶體數量驟增,有鑑於此,IC佈局之設計需更加嚴謹。本文提出以模糊C-Means分群法進行自動化佈局時元件之分群,輸入之數值藉由差值擴張法擴張其值域,群集之結果可決定電晶體元件是否可共用防護環,進而優化其類比積體電路之佈局面積使用率,降低寄生電容。 本文藉由輸入型態為結合差值擴張法之模糊C-Means分群法達成佈局時元件分群之目的,自動化佈局工具輸出之腳本語言(Script)藉由Laker完成自動佈局之設計。本文自動佈局分別以反相器(Inverter)、史密特觸發器(Schmitt Trigger)、前置放大器(Pre-Amplifier)做為驗證電路,藉著上述電路實現本文提及之分群演算法,經由自動化佈局之流程,即可節省佈局之時間、減少佈局所佔之面積,使IC設計流程能更加順暢。
With the evolution of the industrial era and the development of increasingly complex IC industry, the IC design technology enables the number of transistors in integrated circuit increase. In this study, a fuzzy C-Means clustering method was developed for automated IC layout. Based on the input values of gate length and width as clustering criteria, the results can be used as a cluster of shared protection ring and decided to reduce the layout area and reduced parasitic capacitance. In this research, the C-Means clustering by combining input patterns and the expansion method was used to achieve the goal of automated layout. The output of the scripting language (Script) of Automated layout tool is a Laker. In this study, Inverter, Schmitt Trigger, Pre-Amplifier were used as a validation circuit. Through the above-mentioned circuit verification, clustering algorithms used in this study can indeed reach the proposes of saving time, reducing the IC layout area., Hence, IC design process can more smoothly.