loading page

Attention Mechanism Based Bidirectional LSTM Model for Broadband Power Amplifier Linearization
  • +1
  • rina Su,
  • Jiacheng Wang,
  • Gaoming Xu,
  • taijun liu
rina Su
Ningbo University of Technology

Corresponding Author:srn@nbut.edu.cn

Author Profile
Jiacheng Wang
Ningbo University
Author Profile
Gaoming Xu
Ningbo University
Author Profile
taijun liu
Ningbo University
Author Profile

Abstract

In this letter, a novel model for broadband power amplifier (PA) linearization is proposed, namely Attention Mechanism based Bidirectional Long Short-term Memory network (AM-BiLSTM). In order to verify the linearization performance of the AM-BiLSTM model, a 100MHz bandwidth 5G new radio (5G NR) signal is employed to test the sub-6G PA operating at 2.6-GHz. The experimental results show that the adjacent channel power ratio (ACPR) of the PA with AM-BiLSTM can be improved by 24dB which is 6-dB better than the generalized memory polynomial (GMP) and 3-dB better than the Chebyshev polynomials LSTM (CP-LSTM) in ref[1]. Therefore, the proposed AM-BiLSTM is very effective for the linearization of broadband PA.
10 May 2023Submitted to Electronics Letters
13 May 2023Submission Checks Completed
13 May 2023Assigned to Editor
16 May 2023Reviewer(s) Assigned
30 May 2023Review(s) Completed, Editorial Evaluation Pending
02 Jun 2023Editorial Decision: Revise Major
25 Jun 20231st Revision Received
27 Jun 2023Submission Checks Completed
27 Jun 2023Assigned to Editor
27 Jun 2023Review(s) Completed, Editorial Evaluation Pending
27 Jun 2023Editorial Decision: Accept