Behavioral Modeling of Power Amplifiers With Modern Machine Learning Techniques

Dikmese, Sener; Anttila, Lauri; Campo, Pablo Pascual; Valkama, Mikko; Renfors, Markku

doi:10.1109/imc-5g47857.2019.9160381

Cited by 12 publications

(8 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Li et al established a boosting tree model for PA by parallelizing multiple decision trees and built a higher‐precision PA model with fewer resources 13 . However, decision trees are prone to overfitting, and simply parallelizing decision trees exacerbates this advantage 14 . Dikmense compared the effects of multiple machine learning models such as gradient boosting, SVR, decision trees, random forest, and ANN on PA modeling and found that gradient boosting is optimal in modeling accuracy and speed 14 .…”

Section: Introductionmentioning

confidence: 99%

“…However, decision trees are prone to overfitting, and simply parallelizing decision trees exacerbates this advantage 14 . Dikmense compared the effects of multiple machine learning models such as gradient boosting, SVR, decision trees, random forest, and ANN on PA modeling and found that gradient boosting is optimal in modeling accuracy and speed 14 . However, gradient boosting uses a serial iterative algorithm, which increases the variance of the overall model during the training process, making the model prone to overfitting 15 …”

Section: Introductionmentioning

confidence: 99%

“…14 Dikmense compared the effects of multiple machine learning models such as gradient boosting, SVR, decision trees, random forest, and ANN on PA modeling and found that gradient boosting is optimal in modeling accuracy and speed. 14 However, gradient boosting uses a serial iterative algorithm, which increases the variance of the overall model during the training process, making the model prone to overfitting. 15 In summary, the modeling history of RF PA is the application history of various modeling methods, 15,16 and any new modeling method will be quickly applied to the modeling of RF PA.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Particle swarm optimization‐XGBoost‐based modeling of radio‐frequency power amplifier under different temperatures

Wang,

Zhou

2023

Int J Numerical Modelling

View full text Add to dashboard Cite

XGBoost is the optimization of gradient boosting with the best overall performance among machine learning algorithms. By introducing a regularization term into the loss function of gradient boosting, XGBoost can effectively limit the complexity of the model, improve the generalization ability, and solve the overfitting problem. In this paper, XGBoost is first introduced into modeling radio‐frequency (RF) power amplifiers (PA) under different temperatures. Furthermore, the modeling effect of XGBoost is mainly dependent on hyperparameters. As traditional grid search is time‐consuming and labor‐intensive, this paper combines particle swarm optimization (PSO) with XGBoost for searching hyperparameters. The experimental results show that XGBoost can effectively suppress the overfitting problem in gradient boosting while modeling RF PAs in different ambient temperatures. In addition, compared to classic machine learning algorithms, including support vector regression (SVR), gradient boosting, and XGBoost, the proposed PSO‐XGBoost can increase the modeling accuracy by one order of magnitude or more while also increasing the modeling speed by more than one magnitude or more. The PSO‐XGBoost model proposed in this paper can be introduced into modeling other microwave/RF devices and circuits to improve modeling accuracy and reduce modeling time.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Particle swarm optimization‐XGBoost‐based modeling of radio‐frequency power amplifier under different temperatures

Wang,

Zhou

2023

Int J Numerical Modelling

View full text Add to dashboard Cite

show abstract

“…A residual SI signal generally remains after analog cancellation, which must be canceled in the digital domain. This digital cancellation requires modeling the non-linear distortion and memory effects, such as digital-to-analog converter (DAC) and analog-to-digital-converter (ADC) non-linearities [5], IQ imbalance [5], [6], phase-noise [7], [8], and power amplifier (PA) non-linearities [4]- [6], [9] in addition to the memory effects of the channel as illustrated in Fig 1 . To describe and adjust the parameters of these effects various methods have been proposed such as polynomial models [4], [6], [10], [11], support vector machines [12], neural networks (NNs) [13]- [15], and more traditional machine learning methods [16]. However, while the cancellation performance and complexity of time-invariant systems have already been investigated thoroughly, we note that the parameters of FD systems change over time due to, e.g., circuit temperature variations and motion in the surrounding environment.…”

Section: Introductionmentioning

confidence: 99%

On the Implementation Complexity of Digital Full-Duplex Self-Interference Cancellation

Kristensen¹,

Balatsoukas‐Stimming²,

Burg³

2021

Preprint

View full text Add to dashboard Cite

In-band full-duplex systems promise to further increase the throughput of wireless systems, by simultaneously transmitting and receiving on the same frequency band. However, concurrent transmission generates a strong self-interference signal at the receiver, which requires the use of cancellation techniques. A wide range of techniques for analog and digital self-interference cancellation have already been presented in the literature. However, their evaluation focuses on cases where the underlying physical parameters of the full-duplex system do not vary significantly. In this paper, we focus on adaptive digital cancellation, motivated by the fact that physical systems change over time. We examine some of the different cancellation methods in terms of their performance and implementation complexity, considering the cost of both cancellation and training. We then present a comparative analysis of all these methods to determine which perform better under different system performance requirements. We demonstrate that with a neural network approach, the reduction in arithmetic complexity for the same cancellation performance relative to a state-of-the-art polynomial model is several orders of magnitude.

show abstract

“…Traditional non-linear methods, which have been widely used in the FD literature, rely on polynomial models [5]- [7]. More recently, machine learning techniques have also been used to model transceiver non-linearities, mostly focusing on the use of black-box and model-based neural networks (NNs) [8]- [17], but also on other techniques such as support vector machines [18] and tree-based algorithms [19].…”

Section: Introductionmentioning

confidence: 99%

Non-Linear Self-Interference Cancellation via Tensor Completion

Jochems¹,

Balatsoukas‐Stimming²

2020

Preprint

View full text Add to dashboard Cite

Non-linear self-interference (SI) cancellation constitutes a fundamental problem in full-duplex communications, which is typically tackled using either polynomial models or neural networks. In this work, we explore the applicability of a recently proposed method based on low-rank tensor completion, called canonical system identification (CSID), to non-linear SI cancellation. Our results show that CSID is very effective in modeling and cancelling the non-linear SI signal and can have lower computational complexity than existing methods, albeit at the cost of increased memory requirements.

show abstract

Behavioral Modeling of Power Amplifiers With Modern Machine Learning Techniques

Cited by 12 publications

References 6 publications

Particle swarm optimization‐XGBoost‐based modeling of radio‐frequency power amplifier under different temperatures

Particle swarm optimization‐XGBoost‐based modeling of radio‐frequency power amplifier under different temperatures

On the Implementation Complexity of Digital Full-Duplex Self-Interference Cancellation

Non-Linear Self-Interference Cancellation via Tensor Completion

Contact Info

Product

Resources

About