Circuit Connectivity Inspired Neural Network for Analog Mixed-Signal Functional Modeling

被引：8

作者：

Hassanpourghadi, Mohsen ^{[1
]}

Su, Shiyu ^{[1
]}

Rasul, Rezwan A. ^{[1
]}

Liu, Juzheng ^{[1
]}

Zhang, Qiaochu ^{[1
]}

Chen, Mike Shuo-Wei ^{[1
]}

机构：

[1] Univ Southern Calif, Dept Elect & Comp Engn, Los Angeles, CA 90007 USA

来源：

2021 58TH ACM/IEEE DESIGN AUTOMATION CONFERENCE (DAC) | 2021年

关键词：

Analog Mixed-Signal; Modeling; Computer Aided Design; Artificial Neural Network; Modular NN; circuit-connectivity-inspired NN;

D O I：

10.1109/DAC18074.2021.9586236

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Among different types of regression methods to model Analog/Mixed-Signal (AMS) circuits, the Artificial Neural Network (ANN) is a promising candidate due to its reasonable accuracy and fast evaluation. However, for complex AMS circuits with wide specification ranges, creating an ANN model requires a large training dataset. To reduce the required training dataset's volume, we have proposed a circuit-connectivity-inspired ANN (CCI-NN), including multiple sub-ANNs linked according to the actual circuit connections. For validation, we have employed CCI-NN to model a three-stage amplifier and a current-steering digital-to-analog converter. For a certain modeling accuracy, the training dataset requirement is reduced by 3.5x-7.6x.

引用

页码：505 / 510

页数：6