+Advanced Search
Home > Archive>Volume 44, Issue 6, 2021 >76-80
PDF HTML XML Export Cite reminder
A Reverse Wave Based Method Enabling Rapid Extraction of Bayesian Inference-Based Behavioral Model
CSTR:
[cstr]
Author:
Affiliation:

1. Science and Technology on Electronic Test & Measurement Laboratory，Qingdao, 266000，China; 2. Key Laboratory of RF Circuits and Systems Ministry of Education，Hangzhou Dianzi University，Hangzhou 310018，China

Clc Number:

TN98

  • Article
    • | |
  • Metrics
    • |
  • Reference
    • |
  • Related [20]
    • |
  • Cited by
    • | |
  • Comments
    Abstract:

    Measurement speed is a key fact that limited the practical application of RF device behavior model. In this paper, a novel rapid extraction method based on reverse wave injection is proposed. This method changes the amplitude and phase of the reverse incident wave at the load side of the device under test, so that massive training data can be acquired for the Bayesian algorithm, without the need for a complete load-pull test. Experiment is conducted over a GaN device at 2GHz frequency, the results prove that the method has greatly increase the extracting speed without compromising the accuracy. Compared with the traditional scheme, this method can greatly improve the model extracting efficiency and can be widely applied in practice.

    Reference
    Cited by
Get Citation

Copy
Share
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
  • Cited by:
History
  • Online: October 18,2024
Article QR Code
You are the first119977 Visitors
Governing Body：Beijing Electronic Holding Co., LTD
Organizer：Beijing Radio Technology Research Institute
Address：No.79 Beiheyan Street, Dongcheng District, Beijing China, 100009
Phone：13810562337
Electric Measurement Technology ® 2025 All Rights Reserved