Natural gradient algorithm for neural networks applied to non‐linear high power amplifiers

Abdulkader, H.; Langlet, F.; Roviras, Daniel; Castanié, F.

doi:10.1002/acs.725

Cited by 21 publications

(8 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They have been successfully used in many applications related to digital communication such as coding and decoding [7], equalization [2], etc. In [1] and [6], NN have been successfully used to identify and predistort high power amplifiers. In [9], a NN is implemented on an ASIC for use in on-board regenerative satellites [9].…”

Section: Neural Network Structurementioning

confidence: 99%

Mitigation of Multipath Influence on Tracking Errors in LEO Navigation Applications

Abdulkader

Roviras²,

Chaggara³

et al. 2006

The 2006 IEEE International Joint Conference on Neural Network Proceedings

Self Cite

View full text Add to dashboard Cite

This paper presents a solution to enhance the navigation of LEO satellites by mitigating the multipath influence. In several cases, Navigation of LEO satellites needs a very high accuracy in positioning, while GPS receivers suffer from leak of accuracy. As multipath is a major source of inaccuracy, we propose to use a neural networks (NN) technique in the range domain in order to estimate and compensate for Phase and code (delay) tracking errors. In this paper we present a rich discussion of the NN application in the range domain and the major factors influencing its performance.

show abstract

Section: Neural Network Structurementioning

confidence: 99%

Mitigation of Multipath Influence on Tracking Errors in LEO Navigation Applications

Abdulkader

Roviras²,

Chaggara³

et al. 2006

The 2006 IEEE International Joint Conference on Neural Network Proceedings

Self Cite

View full text Add to dashboard Cite

show abstract

“…Let y k , I and y k , Q denote the (I) and (Q) components of the received signal at k ‐ th receive antenna. The output of the m ‐ th hidden neuron at the k ‐ th receive antenna is given by

\begin{matrix} falsenone nonefalsearray \\ arraycenter v_{k, m} = f (\sum_{j} w_{k, j, m} y_{k, j} + b_{k, m}) j = I, Q \end{matrix}

where w k , j , m is the weight connection between y k , j and the m ‐ th neuron and f ( ⋅ ) is a nonlinear activation function (hyperbolic tangent function). Consequently, the output of the NN z k , j at the k ‐ th receive antenna can be expressed as

z_{k MathClass-punc, j} MathClass-rel= {MathClass-op\sum}_{m} u_{k MathClass-punc, m MathClass-punc, j} v_{k MathClass-punc, m}

where u k , m , j is the weight connecting the m ‐ th neuron of the hidden layer to the j ‐ th neuron of the output layer.…”

Section: Architecture Of the Nln‐mrc Receivermentioning

confidence: 99%

“…According to Equation and using the central limit theorem, ϵ can be characterized by a Gaussian distribution. For a noiseless MIMO‐STBC system, the variance of ϵ will be represented by

\begin{matrix} aligned \\ right σ_{ϵ}^{2} & left = E \{| r - \overset{MathClass-op̂}{r} |^{2}\} & right \\ right & left = {[r - \sum_{j} (\sum_{m} u_{k, m, j} f MathClass-open(\sum_{j} (w_{k, j, m} r + b_{k, m}) MathClass-close))]}^{2} \end{matrix}

…”

Section: Performance Analysismentioning

confidence: 99%

“…Several models of HPA such as the travelling wave tube amplifier (TWTA), the solid state power amplifier (SSPA) and the soft‐envelope limiter (SEL) can be used for HPA analysis. The HPA is usually designed to operate near its saturation region in order to maximize the power efficiency . However, the HPA may operate in its nonlinear region especially for high‐power signal levels.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the effect of neural network compensation on MIMO‐STBC systems in the presence of HPA nonlinearity

Belkacem

Ammari

Zayani

et al. 2014

Trans. Emerging Tel. Tech.

View full text Add to dashboard Cite

In this paper, we focus on the effect of nonlinear high‐power amplifiers (HPA) on the multiple‐input‐multiple‐output space‐time block coded (MIMO‐STBC) systems. In order to compensate the HPA nonlinearity, we propose a new receiver scheme based on a neural network algorithm in conjunction with the maximal‐ratio combining (MRC) technique. The performances of the proposed nonlinear network (NLN), called NLN‐MRC receiver, are evaluated for a MIMO‐STBC systems over uncorrelated Rayleigh fading channels. Analytic expressions of the average symbol error rate and the error vector magnitude are delivered. We also analyse the channel capacity of the considered system assuming the perfect knowledge of the channel coefficients and the use of the water‐filling approach. Simulation results show that the proposed compensation technique can efficiently reduce the effect of HPA distortions. In addition, we note an excellent agreement between analytic expressions and Monte‐Carlo simulation curves. Furthermore, the proposed adaptive NLN‐MRC scheme has a low complexity, fast convergence, and best performance than its competitors given in the literature. Copyright © 2014 John Wiley & Sons, Ltd.

show abstract

“…The inverse modeling approach based on the Volterra series analysis has also been proposed to give a predistortion scheme for the static nonlinearity of HPA [2]. Furthermore, data-based learning techniques of neural network [3] and support vector machine approach [4] are also utilized to solve the nonlinear predistortion in the time domain. However, the previous approaches do not consider linear dynamics in HPA.…”

Section: Introductionmentioning

confidence: 99%

Time-domain adaptive predistortion for nonlinear amplifiers

Ding

Sano

2004 IEEE International Conference on Acoustics, Speech, and Signal Processing

View full text Add to dashboard Cite

This paper is concerned with a new time-domain adaptive predistortion scheme for compensating nonlinearity of high power amplifiers (HPA) in OFDM systems. A complex Wiener-Hammerstein model (WHM) is adopted to describe the input-output relationship of unknown HPA with linear dynamics, and a power series model with memory (PSMWM) is proposed to approximate the HPA expressed by WHM. By using the PSMWM, the compensation input to HPA is calculated in a real-time manner so that the linearization from the predistorter input to the HPA output can be attained even if the nonlinear input-output relation of HPA is uncertain and changeable. The effectiveness of the proposed predistortion method is validated by numerical simulation for 64QAM-OFDM systems.

show abstract

Natural gradient algorithm for neural networks applied to non‐linear high power amplifiers

Cited by 21 publications

References 26 publications

Mitigation of Multipath Influence on Tracking Errors in LEO Navigation Applications

Mitigation of Multipath Influence on Tracking Errors in LEO Navigation Applications

On the effect of neural network compensation on MIMO‐STBC systems in the presence of HPA nonlinearity

Time-domain adaptive predistortion for nonlinear amplifiers

Contact Info

Product

Resources

About