Expedient Prediction of Eye Opening of High-Speed Links with Input Design Space Dimensionality Reduction

Hanzhi Ma, Er Ping Li, Andreas C. Cangellaris, Xu Chen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We propose a new method, named Support Vector Regression-based Active Subspace, for the reduction of the dimensionality of the high-dimensional input space of design parameters pertinent to the predictive assessment of the eye opening prediction of high-speed links with IBIS-AMI transmitter and receiver equalization. We compare the method with Support Vector Regression model and Principal Component Analysis-based dimensionality reduction algorithm. Numerical results show that proposed method exhibits the best accuracy in predicting eye height, eye width, and eye width at 10-12 BER in the presence of correlated design variability.

Original languageEnglish (US)
Title of host publication2020 IEEE International Symposium on Electromagnetic Compatibility and Signal/Power Integrity, EMCSI 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages236-240
Number of pages5
ISBN (Electronic)9781728174303
DOIs
StatePublished - Jul 2020
Event2020 IEEE International Symposium on Electromagnetic Compatibility and Signal/Power Integrity, EMCSI 2020 - Reno, United States
Duration: Jul 28 2020Aug 28 2020

Publication series

Name2020 IEEE International Symposium on Electromagnetic Compatibility and Signal/Power Integrity, EMCSI 2020

Conference

Conference2020 IEEE International Symposium on Electromagnetic Compatibility and Signal/Power Integrity, EMCSI 2020
Country/TerritoryUnited States
CityReno
Period7/28/208/28/20

Keywords

  • Support vector regression
  • active subspace
  • dimensionality reduction
  • high-speed link analysis
  • principle component analysis

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Safety, Risk, Reliability and Quality
  • Radiation
  • Signal Processing
  • Information Systems and Management

Fingerprint

Dive into the research topics of 'Expedient Prediction of Eye Opening of High-Speed Links with Input Design Space Dimensionality Reduction'. Together they form a unique fingerprint.

Cite this