基于数据场的总体布线拥挤度计算模型  

Congestion Estimation Model for Global Routing Based on Data Fields

作　　者：孟畅[1,2] 蔡懿慈[1] 

机构地区：[1]清华大学计算机科学与技术系,北京100084 [2]南京陆军指挥学院作战试验中心,南京210045

出　　处：《微电子学》2013年第2期296-300,共5页Microelectronics

基　　金：国家自然科学基金资助项目(60976035)

摘　　要：现有的拥挤度评估方法都是基于布线边的当前使用量和历史使用量,无法评估布线边周围布线资源使用情况对拥挤度的影响。提出了一种数据场拥挤度建模方法,根据数据场中势能和场强分布情况,对布线边所在区域拥挤度和布线边周围不同方向的拥挤度差异进行计算,给出了基于数据场拥挤度模型的布线代价计算方法,并应用于模式布线算法和迷宫布线算法的改进。实验结果表明,该方法能够有效降低布线拥挤度。The existing methods for congestion estimation are all based on current usage or historical usage of the routing edge,which cannot estimate effects of usage of routing resource around the edge on congestion.A data field method to model routing congestion was presented.Difference between congestion within edge region and congestion around edge region was calculated based on distribution of potential and strength in the data field.A routing cost calculation method was proposed based on data field congestion model to improve pattern routing and maze routing.Experiment results showed that this method was capable of reducing edge congestion effectively.

关 键 词：总体布线 拥挤度估计 数据场 

分 类 号：TP302.1[自动化与计算机技术—计算机系统结构]