Loading [MathJax]/extensions/MathMenu.js
Large Signal Behavioral Modeling of Power Transistor from Active Load-pull Systems | IEEE Conference Publication | IEEE Xplore

Large Signal Behavioral Modeling of Power Transistor from Active Load-pull Systems


Abstract:

Large signal behavioral modeling for RF power transistor has been developed rapidly in the past decade, this paper compares and contrasts recent nonlinear large-signal be...Show More

Abstract:

Large signal behavioral modeling for RF power transistor has been developed rapidly in the past decade, this paper compares and contrasts recent nonlinear large-signal behavioral modeling techniques designed for power transistors. These techniques include traditional function based behavioral modeling techniques, e.g. X-parameters, Padé model, Cardiff model, load-pull X-parameters; and artificial neural network (ANN) and machine learning (ML) technique based methods, e.g. real value feed-forward neural network (RVFFNN) model, Bayesian inference model, support vector regression (SVR) based model, in the frequency domains. Besides, approaches to generating the data for modeling from active load-pull measurements are reviewed as well, including closed loop, feed-forward, and open loop active load pull system.
Date of Conference: 28-30 August 2019
Date Added to IEEE Xplore: 12 December 2019
ISBN Information:
Conference Location: Nanjing, China

I. Introduction

Increasingly demands of various semiconductor technologies e.g., GaAs HEMTs and InP HBTs, in terms of power, frequency, and applied high peak-to-average ratios broad band signals, have placed massive requirements on the accuracy of the device models used for design. New semiconductor material systems, e.g., GaN, have been developing at such a fast rate that conventional equivalent circuit modeling approaches may not be able to keep up.

References

References is not available for this document.