MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction

Matching constraints in an analog integrated circuit (IC) are critical to optimizing layout performance. To extract these matching constraints accurately and efficiently from the netlist, we propose the heterogeneous matching constraint extraction graph neural network (MCE-HGCN). First, the netlist...

Full description

Saved in:
Bibliographic Details
Main Authors: Yong Zhang, Yong Yin, Ning Xu, Bowen Jia
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Micromachines
Subjects:
Online Access:https://www.mdpi.com/2072-666X/16/6/677
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849425280815333376
author Yong Zhang
Yong Yin
Ning Xu
Bowen Jia
author_facet Yong Zhang
Yong Yin
Ning Xu
Bowen Jia
author_sort Yong Zhang
collection DOAJ
description Matching constraints in an analog integrated circuit (IC) are critical to optimizing layout performance. To extract these matching constraints accurately and efficiently from the netlist, we propose the heterogeneous matching constraint extraction graph neural network (MCE-HGCN). First, the netlist is mapped into a heterogeneous attribute multi-graph, and based on the characteristics of analog IC matching constraints, a mixed-domain attention mechanism is developed to leverage both the topology information and node attributes in the graph to characterize node embeddings. A matching classifier, implemented using the support vector machine (SVM), is then employed to classify different types of matching constraints from the netlist. Additionally, a matching filter is introduced to remove interference terms. Experimental results demonstrate that the MCE-HGCN model converges effectively with small datasets. In the matching prediction process, the mean <i>F</i>1 score reached 0.917 across different netlist processes and circuit types while maintaining a shorter runtime compared to other methods. Ablation experiments also show that incorporating the mixed-domain attention mechanism and the matching filter individually leads to significant performance improvements. Overall, MCE-HGCN excels at extracting matching constraints from various analog circuits and processes, offering valuable insights for placement guidance and enhancing the efficiency of analog IC layout design.
format Article
id doaj-art-d508b6f4eddf48d79a0494d450fb598a
institution Kabale University
issn 2072-666X
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Micromachines
spelling doaj-art-d508b6f4eddf48d79a0494d450fb598a2025-08-20T03:29:49ZengMDPI AGMicromachines2072-666X2025-06-0116667710.3390/mi16060677MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints ExtractionYong Zhang0Yong Yin1Ning Xu2Bowen Jia3School of Information Engineering, Wuhan University of Technology, Wuhan 430070, ChinaSchool of Information Engineering, Wuhan University of Technology, Wuhan 430070, ChinaSchool of Information Engineering, Wuhan University of Technology, Wuhan 430070, ChinaSchool of Information Engineering, Wuhan University of Technology, Wuhan 430070, ChinaMatching constraints in an analog integrated circuit (IC) are critical to optimizing layout performance. To extract these matching constraints accurately and efficiently from the netlist, we propose the heterogeneous matching constraint extraction graph neural network (MCE-HGCN). First, the netlist is mapped into a heterogeneous attribute multi-graph, and based on the characteristics of analog IC matching constraints, a mixed-domain attention mechanism is developed to leverage both the topology information and node attributes in the graph to characterize node embeddings. A matching classifier, implemented using the support vector machine (SVM), is then employed to classify different types of matching constraints from the netlist. Additionally, a matching filter is introduced to remove interference terms. Experimental results demonstrate that the MCE-HGCN model converges effectively with small datasets. In the matching prediction process, the mean <i>F</i>1 score reached 0.917 across different netlist processes and circuit types while maintaining a shorter runtime compared to other methods. Ablation experiments also show that incorporating the mixed-domain attention mechanism and the matching filter individually leads to significant performance improvements. Overall, MCE-HGCN excels at extracting matching constraints from various analog circuits and processes, offering valuable insights for placement guidance and enhancing the efficiency of analog IC layout design.https://www.mdpi.com/2072-666X/16/6/677analog ICmatching constraintsheterogeneous multi-graphmixed attentionssmall dataset
spellingShingle Yong Zhang
Yong Yin
Ning Xu
Bowen Jia
MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction
Micromachines
analog IC
matching constraints
heterogeneous multi-graph
mixed attentions
small dataset
title MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction
title_full MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction
title_fullStr MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction
title_full_unstemmed MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction
title_short MCE-HGCN: Heterogeneous Graph Convolution Network for Analog IC Matching Constraints Extraction
title_sort mce hgcn heterogeneous graph convolution network for analog ic matching constraints extraction
topic analog IC
matching constraints
heterogeneous multi-graph
mixed attentions
small dataset
url https://www.mdpi.com/2072-666X/16/6/677
work_keys_str_mv AT yongzhang mcehgcnheterogeneousgraphconvolutionnetworkforanalogicmatchingconstraintsextraction
AT yongyin mcehgcnheterogeneousgraphconvolutionnetworkforanalogicmatchingconstraintsextraction
AT ningxu mcehgcnheterogeneousgraphconvolutionnetworkforanalogicmatchingconstraintsextraction
AT bowenjia mcehgcnheterogeneousgraphconvolutionnetworkforanalogicmatchingconstraintsextraction