Parameterized Basis Functions for Numerically Stable
Behavioral Modeling of RF Power Amplifiers
Afef Harguem1,3, Noureddine Boulejfen2,3, Fadhel M. Ghannouchi3, and Ali Gharsalla1
2
Faculté des Sciences de Tunis, Université El-Manar, Tunis, Tunisia.
Institut Supérieure des Sciences Appliquées et de Technologie, de Kairouan, Kairouan, Tunisia. 3
iRadio Laboratory, Department of Electrical and Computer Engineering, Schulich School of
Engineering, University of Calgary, Calgary, AB, Canada T2N 1N4.
orthogonal ones. However, this requires the calculation of converting terms with exponentially increasing number as the polynomial order increases. Moreover, the orthogonality of the obtained set of basis functions is guaranteed for uniformly distributed and zero reaching input envelope signals only. To overcome this limitation, associated Laguerre polynomials have been proposed [5]. The resulting set of basis function is fully orthogonal for Rayleigh distributed envelope amplitudes, making it suitable for complex Gaussian envelopes. Once again the stability of the resulting model is guaranteed only for a Rayleigh distribution.
In this paper, a new polynomial model based on a set of parameterized basis functions is suggested to guarantee a maximum numerical stability regardless of the input signal statistics and the model nonlinearity order.
The organization of the paper is such that an introduction to conventional polynomial models is presented in Section II. In Section III, the proposed polynomial model is presented. Section IV is devoted to the validation of the proposed model; and, finally, Section V presents the conclusions.
1
Abstract— In this paper, we propose a new set of parameterized basis functions suitable for behavioral modeling of RF power amplifiers exhibiting memory effects. Depending on the input signal distribution, the proposed basis functions can be optimized for numerically stable model identification procedure. The efficiency and robustness of the resulting polynomial model are demonstrated and compared to the ones of previously published polynomial models. The obtained results revealed an exceptional numerical stability regardless of the input signal statistics, making the proposed new model suitable for multimode and broadband nonlinear wireless transmitters.
Index Terms—power amplifier; numerical stability; condition number; behavioral modeling; WCDMA.
I. INTRODUCTION
Power amplifiers are critical components in modern wireless communication systems. In fact, they represent the major source of nonlinearity that causes signal distortion leading to an undesired spectral regrowth. . For this reason, the accurate modeling of power amplifiers has become crucial, in order to estimate their effects and proceed to their linearization in some cases.
In recent years, memory polynomial models [1, 2] have been widely used. These models are composed of a finite II. MODELING OF RF POWER AMPLIFIERS WITH number of delay taps followed by nonlinear static CONVENTIONAL POLYNOMIAL MODELS functions. They can be seen as a truncation of the general
To predict the spectral regrowth at the output of RF Volterra series, where only the diagonal terms of the ~
Volterra kernel are considered. This truncation power amplifiers PAs, the complex envelopes x(t) and
~
significantly reduces the number of parameters in the y(t) of the input and output signals respectively are model. On the other hand, it can be interpreted as a special related by: case of the parallel Wiener model [3], where infinite ~y(t)=~x(t)G(r) (1) impulse response functions are reduced to single delay
taps. Many attempts have been made to improve the where r=~
x(t, and G(r) is the complex gain of the PA.
convergence of this model.
Polynomial models, however, suffer from a major G(r) can be represented by a complex power series of a problem related to numerical stability. In fact, the finite order, K, such that: parameter identification associated with these models K
l 1~necessitates inversion of matrices, which are generally y(t)=∑bl~x(t)~x(t) (2)
badly conditioned due to the set of basis functions used in l=1these models. This stability problem is more pronounced
To evaluate the coefficients, bl, N samples of the with high polynomial orders and memory depths. To
overcome the stability problem, Raich [4] has proposed the measured input and output envelopes can be used. Using conversion of the conventional basis functions to the least square estimation approach, a solution such as
Parameterized telescoping proves algebraic independence of sums Carsten Schneider Abstract: Usually creative telescoping is used to derive recurrences for sums. ...
Parameterized telescoping proves algebraic independence of sums Carsten Schneider Abstract: Usually creative telescoping is used to derive recurrences for sums. ...
在C#中ParameterizedThreadStart和ThreadStart区别_计算机软件及应用_IT/计算机_专业资料。在C#中ParameterizedThreadStart和ThreadStart区别...
Parameterized Suffix Trees – Motivations – Definitions – Searching & Example – Parameterized Duplications 18 April 2000 hsh@cs.umd.edu Tries & Trees ? ...
Beyond parameterized verication_专业资料。Abstract. We present a sound and fully automated method for the veri cation of safety properties of parameterized ...
Parameterized Computational Feasibility RODNEY G. DOWNEY and MICHAEL R. FELLOWS y 24 November 1992, revised 5 May 1994 Abstract Many natural computational ...
9, NO. 6, DECEMBER 2001 Parameterized RTL Power Models for Soft Macros Alessandro Bogliolo, Roberto Corgnati, Enrico Macii, Member, IEEE, and Massimo ...
MalformedParameterizedTypeException_专业资料。jdk1MalformedParameterizedTypeException (Java Platform SE 6) function windowTitle() { if (location.href.indexOf(...
In this chapter we present an approach to the verification of parameterizedIn this chapter we present an approach to the verification of parameterized隐藏>...
Fitting Parameterized Three-dimensional Models to Images Model-based recognition and motion tracking depends upon the ability to solve for projection and ...
热门文档