Deep Learning Modeling Method for RF Devices Based on Uniform Noise Training Set

TL;DR

This work addresses the nonlinear modeling of RF devices amid growing IC complexity by proposing a uniform noise training set for deep learning-based characterization. The authors implement a four-part workflow—data acquisition, dataset generation (uniform noise, narrowband noise, and sine waves), dataset processing (delay compensation, normalization, extraction), and model training with AutoEncoder, ResNet, and Mamba—and validate it on the PW210 amplifier. Results show that models trained with uniform noise can predict unseen waveforms and perform well in both time- and frequency-domain tests, with ResNet frequently delivering the best accuracy and generalization. The approach offers a practical, generalizable path for RF IC modeling across diverse devices and operating conditions, potentially simplifying testing and design workflows in RF electronics.

Abstract

As the scale and complexity of integrated circuits continue to increase, traditional modeling methods are struggling to address the nonlinear challenges in radio frequency (RF) chips. Deep learning has been increasingly applied to RF device modeling. This paper proposes a deep learning-based modeling method for RF devices using a uniform noise training set, aimed at modeling and fitting the nonlinear characteristics of RF devices. We hypothesize that a uniform noise signal can encompass the full range of characteristics across both frequency and amplitude, and that a deep learning model can effectively capture and learn these features. Based on this hypothesis, the paper designs a complete integrated circuit modeling process based on measured data, including data collection, processing, and neural network training. The proposed method is experimentally validated using the RF amplifier PW210 as a case study. Experimental results show that the uniform noise training set allows the model to capture the nonlinear characteristics of RF devices, and the trained model can predict waveform patterns it has never encountered before. The proposed deep learning-based RF device modeling method, using a uniform noise training set, demonstrates strong generalization capability and excellent training performance, offering high practical application value.

Figures (18)

• Figure 1: Time-domain Histogram and Frequency Spectrum of Uniform Noise
• Figure 2: Test equipment and their connection relationships
• Figure 3: Abstract connection diagram of test equipment
• Figure 4: Delay correction of sequences
• Figure 5: SiLU Activation Function