多层P/G网络内RLC信号线的参数提取和时序分析

Parameter extraction and temporal analysis techniques for RLC signal lines in multi-layered P/G networks

大规模集成电路特征尺寸的急剧缩小及系统功能日益复杂使得电源分配网络规模越来越大,越来越密集,长互连上的传输线效应日益显著.文中以一些布置在多层P/G网络内的关键信号线网为研究对象;利用智能查表工具(NNIE与NNCE),实现对长信号线静态分割后各分段频变寄生电感及电容参数的快速提取,从而建立起长信号线梯状分布RLC模型;数值模拟表明,该方法具有快速的特点和良好的精度.

Because the feature size in ultra large scale integrated circuits （ULSI） is sharply shrinking and chip functions are becoming complex, the scale of power supply network is larger and denser. It makes the inductive effects more remarkable in the interconnected long routes. In this paper, the key signal line located at the multi-layered P/G networks are taken as the research objects. An expert system based on BP neural network （so called NNCE and NNIE） is used to quickly extract parasitic capacitance and inductance in the segments divided from the long signal line. A ladderlike distributed RLC model is built. The simulation results show that the RLC system has a fast response and fine precision.