A Hotspot Detection Algorithm Based on Graph Family Modelling and Graph Neural Networks

Liang Longxi; Cai Shuting; Zhang Hui

doi:10.12052/gdutxb.250111

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

A Hotspot Detection Algorithm Based on Graph Family Modelling and Graph Neural Networks

更新时间：2026-05-27

- A Hotspot Detection Algorithm Based on Graph Family Modelling and Graph Neural Networks
- Vol. 42, Issue 6, Pages: 34-43(2025)
- 作者机构：
  
  1. 广东工业大学集成电路学院,广东,广州,510006
  2. 广东工业大学自动化学院,广东,广州,510006
- 作者简介：
- 基金信息：
- DOI：10.12052/gdutxb.250111
  CLC： TN407;TN707
- Published：2025
- 稿件说明：
移动端阅览
梁珑曦, 蔡述庭, 张慧. 基于图族建模与图神经网络的热点检测算法[J]. 广东工业大学学报, 2025, 42(6): 34–43.

Liang Longxi, Cai Shuting, Zhang Hui. A Hotspot Detection Algorithm Based on Graph Family Modelling and Graph Neural Networks[J]. 2025, 42(6): 34-43.
梁珑曦, 蔡述庭, 张慧. 基于图族建模与图神经网络的热点检测算法[J]. 广东工业大学学报, 2025, 42(6): 34–43. DOI： 10.12052/gdutxb.250111

Liang Longxi, Cai Shuting, Zhang Hui. A Hotspot Detection Algorithm Based on Graph Family Modelling and Graph Neural Networks[J]. 2025, 42(6): 34-43. DOI： 10.12052/gdutxb.250111.

摘要

随着集成电路工艺推进至先进节点，版图结构复杂度大幅上升，制造良率维持难度显著增加，对光刻热点检测提出更高要求。针对现有检测方法在版图建模表达力、增强样本区分性与模型检测能力方面的问题，本文提出一种基于图族建模与图神经网络的热点检测算法，实现结构表达力与模型判别性能的协同提升。首先，面向光刻版图片段提出一种多尺度图族建模方法，通过局部子图与全局大图联合构建图族结构，能有效提升版图结构与语义属性的表达能力，同时增强样本在数据增强处理下的可分性。其次，在图族结构作为样本输入的基础上，设计了分层图神经网络热点检测模型，提升了多层次特征提取能力与对核心区域的结构聚焦能力。最后，本文图族建模方法的t-SNE嵌入可视化实验结果表明，在样本分布均匀性与结构区分性方面具有显著优势，相比传统图像建模，其平均最近邻距离(Mean Nearest Neighbor Distance，MNND) 提升了206个百分点、Shannon分布熵提升了16个百分点。本文图神经网络模型使用ICCAD'19 TNSB竞赛数据集进行测试，在保持结构层次化的同时实现了优异的检测效果，召回率达到99.91%，误报率仅为1.12%，为光刻版图片段的数据增强表达能力、复杂热点感知能力的提升与光刻热点检测问题的解决提供创新算法。

Abstract

The rapid scaling of semiconductor technologies has led to increasingly complex integrated circuit layouts

which pose severe challenges to maintaining manufacturing yield and highlight the urgent need for accurate lithographic hotspot detection. Existing detection methods are often constrained by limited layout modelling capacity

insufficient separability of augmented samples

and suboptimal classification performance. To overcome these challenges

a hotspot detection algorithm was developed that combined a multi-scale graph family modelling strategy with a hierarchical graph neural network. The proposed graph family model constructed both local subgraphs and a global graph to represent geometric details and contextual semantics simultaneously

thereby improving structural representation and preserving data separability under various augmentation operations. On this basis

a hierarchical GNN was designed to extract multi-level features and strengthen the model’s focus on critical core regions. Experimental evaluations on the ICCAD'19 TNSB benchmark demonstrate that the proposed method substantially improves data distribution balance

achieving a 206 percentage point increase in MNND and a 16 percentage point gain in Shannon entropy. Moreover

the detection model attains a recall of 99.91% with only a 1.12% false alarm rate

outperforming state-of-the-art alternatives. These results confirm that the proposed graph-based approach offers both strong structural expressiveness and robust detection capability

providing a promising solution for addressing the challenges of lithographic hotspot detection in advanced technology nodes.

关键词

Keywords

references

YANG H, LIN Y, YU B, et al . Lithography hotspot detection: from shallow to deep learning[C ] //2017 30th IEEE International System-on-Chip Conference (SOCC) . Munich: IEEE, 2017: 233-238.

BORISOV V, SCHEIBLE J. Research on data augmentation for lithography hotspot detection using deep learning[C]//34th European Mask and Lithography Conference. Bellingham: SPIE, 2018, 10775: 204-209.

WANG Y, WANG J, FENG D, et al . Location is all you need: efficient lithographic hotspot detection using only polygon locations[C ] //2025 Design, Automation Test in Europe Conference (DATE) . Lyon: IEEE, 2025: 1-7.

SUN S, JIANG Y, YANG F, et al . Efficient hotspot detection via graph neural network[C ] //2022 Design, Automation Test in Europe Conference Exhibition (DATE) . Antwerp: IEEE, 2022: 1233-1238.

YAN H, WANG Y, GAO P, et al . A lightweight heterogeneous graph embedding framework for hotspot detection[J ] . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2025, 44(9): 3479-3489.

TONG Y, LIN C, CHEN S. An automatic optical-simulation-based lithography hotspot fix flow for post-route optimization[C]//2009 International Symposium on Physical Design (ISPD) . New York: ACM, 2009: 115-122.

YANG J, RODRIGUEZ N, OMEDES O, et al . DRCPlus in a router: automatic elimination of lithography hotspots using 2D pattern detection and correction[C ] //Design for Manufacturability through Design-Process Integration IV. Bellingham: SPIE, 2010: 7641-7650.

WUU J, SIMMONS M, MAREK-SADOWSKA M. Post-placement lithographic hotspot detection and removal in one-dimensional gridded designs[C]//13th International Symposium on Quality Electronic Design (ISQED) . Santa Clara, CA: IEEE, 2012: 193-199.

DING D, TORRES A, PIKUS F, et al . High performance lithographic hotspot detection using hierarchically refined machine learning[C ] //16th Asia and South Pacific Design Automation Conference (ASP-DAC 2011) . Tokyo: IEEE, 2011: 775-780.

GAO J, YU B, PAN D. Accurate lithography hotspot detection based on PCA-SVM classifier with hierarchical data clustering[C]//Design-Process-Technology Co-optimization for Manufacturability VIII. Bellingham: SPIE, 2014, 9053: 114-123.

LECUN Y, BOTTOU L, BENGIO Y, et al . Gradient-based learning applied to document recognition[J ] . Proceedings of the IEEE, 1998, 86(11): 2278-2324.

JIANG C, SUN H, FENG D, et al . LithoExp: explainable two-stage CNN-based lithographic hotspot detection with layout defect localization[J ] . ACM Transactions on Design Automation of Electronic Systems, 2025, 30(3): 1-25.

LIN M, ZENG F W Q, LIU X X, et al . Lithography hotspot detection method based on improved YOLOv3[J ] . Acta Optica Sinica, 2023, 43(23): 190-198.

WU Q Y, LIU J M, ZHANG S, et al . Lithography hotspot detection algorithm based on improved Yolov5s[J ] . Laser Optoelectronics Progress, 2023, 60(24): 251-259.

CHEN Y, WU Y, WANG J, et al . LLM-HD: layout language model for hotspot detection with GDS semantic encoding[C ] //61st ACM/IEEE Design Automation Conference. San Francisco: ACM/IEEE, 2024: 1-6.

SCARSELLI F, GORI M, TSOI A, et al . The graph neural network model[J ] . IEEE Transactions on Neural Networks, 2009, 20(1): 61-80.

YU Y, JIANG H, ZHANG Y, et al . Integrated layout method for polygon via shapes : US 9594867 [P ] . 2017-03-14.

FEI N, GAO Y, LU Z, et al . Z-score normalization, hubness, and few-shot learning[C ] //Proceedings of the IEEE/CVF International Conference on Computer Vision. New York: IEEE, 2021: 142-151.

GILMER J, SCHOENHOLZ S, RILEY P, et al . Message passing neural networks[M ] . SCHÖLKOPF B, TSAFARIS A, CHAVENT M. Machine learning meets quantum physics. Cham: Springer International Publishing, 2020: 199-214.

RUMELHART D, HINTON G, WILLIAMS R. Learning representations by back-propagating errors[J]. Nature, 1986, 323: 533-536.

NAIR V, HINTON G. Rectified linear units improve restricted Boltzmann machines[C]//Proceedings of the 27th International Conference on Machine Learning (ICML-10) . New York: ACM, 2010: 807-814.

RUMELHART D, MCCLELLAND J, PDP RESEARCH GROUP. Parallel distributed processing, volume 1: explorations in the microstructure of cognition: foundations[M]. Cambridge: The MIT Press, 1986: 423-464.

REDDY G, MADKOUR K, MAKRIS Y. Machine learning-based hotspot detection: fallacies, pitfalls and marching orders[C]//2019 IEEE/ACM International Conference on Computer-Aided Design (ICCAD) . New York: IEEE, 2019: 1-8.

ZHANG Q, ZHANG Y, LI J, et al . Litho-NeuralODE 2.0: improving hotspot detection accuracy with advanced data augmentation, DCT-based features, and neural ordinary differential equations[J ] . Integration, 2022, 85: 10-19.

MAATEN L, HINTON G. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008, 9: 2579-2605.

JIANG Y, YANG F, YU B, et al . Efficient layout hotspot detection via neural architecture search[J ] . ACM Transactions on Design Automation of Electronic Systems, 2022, 27(6): 1-16.

WANG C, FANG Y, ZHANG S. Feature fusion based hotspot detection with R-EfficientNet[C]//Proceedings of the Great Lakes Symposium on VLSI 2024. New York: ACM, 2024: 446-451.

GENG H, YANG H, ZHANG L, et al . Hotspot detection via attention-based deep layout metric learning[C ] //Proceedings of the 39th International Conference on Computer-Aided Design. New York: ACM/IEEE, 2020: 1-8.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Factor-level Feature and Attribute Preference Joint Learning Based Session Recommendation

Fuzzy Graph Neural Network-based Industrial Internet Service Component Orchestration Method

Modeling of 3D NAND Memory Cells with Taper Angle and Programming Strategy Optimization

Ground Deformation Monitoring and Influencing Factors Analysis in Fuzhou City Based on PSI Technology

Research on Foreign Object Detection in Power Transmission Lines Based on the AAGV-YOLOX Model

Related Author

Lin Hao

Chen Ping-hua

Liu Anxin

Cheng Lianglun

Wang Tao

Related Institution

School of Computer Science and Technology, Guangdong University of Technology

AI问答

⁰