Amplifiers – Hum or noise or distortion bucking introduced into signal...
Reexamination Certificate
2002-08-15
2003-11-04
Ton, My-Trang Nu (Department: 2816)
Amplifiers
Hum or noise or distortion bucking introduced into signal...
Reexamination Certificate
active
06642786
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a linearization method and apparatus for a high power amplifier (HPA) used for data transmission in digital communication systems; and, more particularly, to a piecewise polynomial predistortion method and apparatus for compensating nonlinear distortion of the high power amplifier.
DESCRIPTION OF RELATED ARTS
Power efficiency and spectrum efficiency are the main concern in modern digital communication systems. The spectrum efficiency requires the system to be more linear while the power efficiency requires the system using an amplifier with nonlinear characteristics. So how to tradeoff between these two conflict factors is a challenge to communication engineers.
Generally, a nonlinear characteristic of a high power amplifier (HPA) is represented by a range of power-in vs. power-out curve. The power-in vs. power-out curves for HPAs are also called amplitude-dependent amplitude distortion (AM/AM) transfer curves.
FIG. 1
shows a typical power-in vs. power-out characteristic of a nonlinear HPA. In addition to the AM/AM transfer curves, manufacturers also provide phase shift properties. A phase shift is imparted to a signal depending on a drive level. The phase shift curve is also called an amplitude-dependent phase distortion (AM/PM) conversion curve.
Ideally it is required to operate the HPA at saturation because the most power is obtained from the HPA operating at saturation. But it is found that the nonlinear characteristic of the HPA causes significant signal distortion at saturation, so the communication engineers back off from this desired optimum, resulting in inefficient use of power.
The linearization of the HPA is the most effective method to overcome this difficulty. Over the years, a number of linearization technologies have been developed and predistortion has been the most common approach developed in new systems today.
Basically, these techniques aim to introduce “inverse” nonlinearities that can compensate the AM/AM and AM/PM distortions generated by the nonlinear HPA. The predistortion device uses a memoryless nonlinear device placed usually between a shaping filter and the HPA. This nonlinear device can be designed digitally by using a mapping predistorter (a look-up table) or a polynomial function based on Cartesian or Polar representation.
The mapping predistorter has an advantage that it can be applied to any order of nonlinearity and any modulation technique. However, to get an acceptable accuracy, the size of the look-up table has to be high enough. So a major drawback with the mapping predistorter is the large size of the look-up table, which results in a long adaptation time.
Therefore, in order to increase the rate of convergence and to reduce the complexity, several adaptive polynomial predistorters have been proposed.
For example, U.S. Pat. No. 6,236,837, entitled “Polynomial Predistortion Linearizing Device, Method, Phone and Base Station”, discloses a polynomial based predistortion method. The key idea of this patent is that the inverse characteristic of the HPA is approximated by the orthogonal polynomial functions expansion and the coefficient of the expansions is estimated by an orthogonal estimation algorithm. The primary objective of this patent is to realize the predistortion with medium or high accuracy depending on the application while minimizing the computational requirements. One feature of this predistortion is the orthogonal polynomial decomposition based on input signal statistical distribution. The computational complexity is increased by the use of orthogonal polynomial compared to methods that use a single non-decomposed polynomial function.
U.S. Pat. No. 5,900,778, entitled “Adaptive Parametric Signal Predistorter for Compensation of Time Varying Linear and Nonlinear Amplifier Distortion”, discloses a polynomial based predistortion method. In this patent, the predistortion is implemented by two look-up tables whose contents are updated by an off-line adaptive predistorter and the forward and inverse characteristics of the HPA are approximated by the polynomial functions. The primary objective of this patent is to overcome some problems of prior art; utilizing the smallest number of HPA output data to determine the best inverse to the output of the HPA where it is most needed, sensitivity to measurements error, and absence of gradients of AM/AM functions. However, the computational complexity of this patent is much larger than the others.
U.S. Pat. No. 6,072,364, entitled “Adaptive Digital Predistortion for Power Amplifiers with Real Time Modeling of Memoryless Complex Gain”, discloses a polynomial based predistortion method. In this patent, the forward function of HPA and its inverse functions are estimated by a line or cubic spline interpolation method and the update operation is performed in a block manner, so that it is possible to eliminate the convergence time and the need for an iterative algorithm. However, this patent is sensitive to the measurements error.
Korean laid-open patent publication No. 2001-064260, entitled “Adaptive Predistorter for Non-linear Distortion Compensation”, discloses a polynomial based predistortion method. In this patent application, the inverse function of HPA is approximated by the polynomial function and the coefficients of the polynomial are estimated by a least mean-square (LMS) algorithm. The primary objective of the this application is to reduce the required memory size and to speed up the convergence time
The performance of polynomial predistortion is highly dependent on the accuracy of the function models used. The order of polynomial must be increased for the performance improvement, resulting in increasing the computational complexity. In addition to this, the errors generated by the polynomial predistortion function are mainly contributed by the signal in the regions of saturation or cutoff.
FIG. 1
shows a plot of the AM/AM transfer function of the HPA and generally a large signal power amplifier device may be characterized by three regions of operations: cut-off, linear or active and saturation regions.
In general, the properties of the amplifier exploit distinct characteristics in each of the three regions and it is very difficult to approximate the function of AM/AM (or AM/PM) by using a single polynomial in the whole input signal regions with a small order of polynomial.
SUMMARY OF THE INVENTION
It is, an objective of the present invention to provide a method and apparatus for modeling a high power amplifier with piecewise polynomial functions in different regions of an input signal to thereby overcome the above-discussed and other problems of the prior art.
To increase the performance of the conventional polynomial based predistortion, the degree of the polynomial functions must be increased and this results in increasing computational complexity.
Generally, most of the power amplifiers may be characterized by three regions of operations: cut-off, linear or active and saturation regions and the amplifier exploits distinct properties in each region. Therefore, it is very difficult to approximate a transfer function of the amplifier with a lower degree of polynomial functions and in some cases it is almost impossible to do so.
To overcome this problem, in accordance with the present invention, the input signal region is divided into several sub-regions. In practical applications, it is very important to select the number of sub-regions, the sub-region points and the degree of the polynomial functions in each sub-region to tradeoff between the performance and complexity. In accordance with the present invention, the selection of these parameters is based on the input signal distribution and characteristics of the amplifier. The LMS algorithm is used in order to achieve the minimum approximation error.
In accordance with one aspect of the present invention, there is provided a polynomial predistortion apparatus, comprising: an amplitude predistortion circuit for predistorting an input amplitude signal based on first polynomial coe
Jin Minglu
Kim Soo-young
Oh Deockgil
Blakely & Sokoloff, Taylor & Zafman
Electronics and Telecommunications Research Institute
Nu Ton My-Trang
LandOfFree
Piecewise polynomial predistortion method and apparatus for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Piecewise polynomial predistortion method and apparatus for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Piecewise polynomial predistortion method and apparatus for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3141204