PCB crosstalk prediction based on machine learning
Author:
Affiliation:

School of Electronic Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,China

Funding:

Ethical statement:

  • Article
  • |
  • Figures
  • |
  • Metrics
  • |
  • Reference
  • |
  • Related
  • |
  • Cited by
  • |
  • Materials
    Abstract:

    With the rapid improvement of clock frequency in electronic system, crosstalk has become one of the problems that Printed Circuit Board(PCB) designers must concern. Although the design cost has been cut to a certain degree, it still takes a lot of time to simulate the crosstalk on PCB even with the help of high-speed circuit simulation software. Aiming to improve the efficiency of PCB crosstalk prediction, a new data structure is proposed to describe PCBs. The factors that cause crosstalk on PCB are comprehensively analyzed, and a PCB crosstalk prediction system is built by using Natural Language Processing(NLP), which reduces the time for crosstalk prediction to the magnitude of seconds and achieves 73.2% accuracy.

    Reference
    Related
    Cited by
Get Citation

陈星宇,石丹,王云鹏.基于机器学习的PCB串扰预测[J]. Journal of Terahertz Science and Electronic Information Technology ,2023,21(6):819~825

Copy
Share
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
History
  • Received:May 09,2020
  • Revised:July 06,2020
  • Adopted:
  • Online: July 04,2023
  • Published: