Markov chain Monte Carlo methods for statistical analysis of RF photonic devices

Molly Piels, Darko Zibar

Research output: Contribution to journalJournal articleResearchpeer-review

334 Downloads (Pure)

Abstract

The microwave reflection coefficient is commonly used to characterize the impedance of high-speed optoelectronic devices. Error and uncertainty in equivalent circuit parameters measured using this data are systematically evaluated. The commonly used nonlinear least-squares method for estimating uncertainty is shown to give unsatisfactory and incorrect results due to the nonlinear relationship between the circuit parameters and the measured data. Markov chain Monte Carlo methods are shown to provide superior results, both for individual devices and for assessing within-die variation
Original languageEnglish
JournalOptics Express
Volume24
Issue number3
Pages (from-to)2084-2097
Number of pages14
ISSN1094-4087
DOIs
Publication statusPublished - 2016

Fingerprint Dive into the research topics of 'Markov chain Monte Carlo methods for statistical analysis of RF photonic devices'. Together they form a unique fingerprint.

Cite this