KR20010064260A - Adaptive pre-distortion system for nonlinear distortion compensation in IMT-2000 system - Google Patents

Abstract

PURPOSE: An adaptation predistorter for compensating nonlinear distortion is provided to compensate nonlinear distortion of a high power amplifier so that a high power amplifier having high efficiency can be enabled, and to classify complex number of input signals through amplitude and phase for compensation so that correct linearity can be obtained. CONSTITUTION: A rectangular-to-polar coordinator conversion unit(201) classifies the amplitude and phase in complex number configured with actual number and imaginary number. An amplitude predistorter(202) receives a coefficient of an amplitude estimation polynomial, for predistortion of the amplitude of an inputted signal. A phase predistorter(203) receives a coefficient of a phase estimation polynomial, for predistortion of the phase of an inputted signal. A polar-to-rectangular coordinator conversion unit(204) classifies complex number configured with amplitude and phase into actual number and imaginary number.

Description

Adaptive pre-distortion system for nonlinear distortion compensation in IMT-2000 system}

The present invention relates to a predistorter, and more particularly, to a predistorter using an adaptive predistortion technique for nonlinear distortion compensation in a next generation mobile communication system.

In the next generation wireless communication system, since multimedia services are available, the amount of traffic transmitted between the base station and the terminal increases. This quickly consumes limited battery power, which shortens the usage time of the terminal. Therefore, if the power amplifier that consumes most of the terminal power is used as a highly efficient high power amplifier, the use time of the terminal can be extended, but the high power amplifier amplifies the output signal nonlinearly, and the response characteristic is Since it is time-variable, its use has been limited.

In particular, the next-generation mobile communication terminal based on multi-code CDMA (Code Division Multiple Access) has a large peak-to-average power ratio, which greatly degrades performance due to nonlinearity of the amplifier. Therefore, if the non-linear distortion of the high power amplifier is linearized by using the predistortion method, the usage time can be increased by increasing the power consumption efficiency of the terminal without performance degradation.

When the high power amplifier is used, the output signal is distorted in amplitude (AM / AM) and phase (AM / PM), and there is a predistortion method to compensate for the distortion of the signal.

In the predistortion method, the input signal of the amplifier is distorted in advance in order to obtain a linear response characteristic of the output signal. Accordingly, the output signal may be linearized. However, since the response characteristics of the high power amplifier vary depending on the usage time, the temperature, and the operating region of the signal, an adaptive predistortion method that estimates and compensates for the characteristics is required.

The conventional adaptive predistorter includes a lookup table method using a memory, a complex amplification method using a polynomial, and a method of estimating the coefficients of a filter, generating all possible signals, and storing the predistortion values thereof in the memory.

The lookup table method using the memory increases the accuracy of the predistorter as the memory size increases, but the update speed is slowed. Also, since the coefficients are updated independently, the values stored in the memory have discontinuous values. The output signal exhibits a phenomenon like background noise.

The complex amplification method is a method of predistorting a signal by multiplying an independent amplification ratio of a real part and an imaginary part of an input complex signal. The amplification factor multiplied by the real part and the imaginary part is expressed as a polynomial about the amplitude of the input signal, and the coefficients of the polynomial are updated by comparing the output value of the high power amplifier with the input value of the predistorter. However, this predistortion method does not accurately compensate for distortion by approximating the distortion characteristics of the amplitude and phase of the amplifier by the polynomials of the real part and the imaginary part.

The method of estimating the coefficient of the filter estimates the distortion characteristic of the power amplifier in the form of a polynomial, and estimates the inverse response characteristic of the power amplifier based on the polynomial. The predistorted signal is calculated for all possible signals and stored in the memory by using the estimated power response polynomial of the power amplifier. The predistorted value of the corresponding memory for the signal input to the predistorter is calculated. How to print Since this method estimates the response characteristics of the power amplifier in order to estimate the reverse response characteristics of the power amplifier, it requires two estimation processes, which increases the computational cost, decreases the update rate, and uses the memory. There is a problem.

SUMMARY OF THE INVENTION The present invention has been made to solve the problems of the prior art, and an object thereof is to provide a predistorter using an adaptive predistortion technique for nonlinear distortion compensation in a next generation mobile communication system.

1 is a schematic diagram of a case where an adaptive predistorter is added to a multi-code CDMA mobile communication system according to an embodiment of the present invention;

FIG. 2 is a configuration diagram of the predistorter shown in FIG. 1;

FIG. 3 is a detailed configuration diagram illustrating the functions of the delay unit and the estimator illustrated in FIG. 1 in more detail.

♠ Explanation of symbols for the main parts of the drawing ♠

101: data source generating data informations

102: Parallel-to-serial Conversion

103: multi-code spreading section

104: Pulse Shaping Filter

105: Predistorter

106: Delay Unit

107: estimator

108: digital to analog converter

109: analog to digital converter

110: Quadrature Modulation

111: Quadrature Demodulation

112: high power amplifier

201: Rectagular-to-Polar Coordinator Conversion

202: Amplitude Predistorter

203: Phase Predistorter

204: Polar-to-Rectangular Coordinator Conversion

301: rectangular coordinates-polar coordinates conversion unit

302: amplitude estimation unit

303: coefficient update unit of the amplitude estimation polynomial

304: first adaptive algorithm

305: quadrature modulation

306: coefficient updater of the phase estimation polynomial

307: high power amplifier

308: phase estimation unit

309: second adaptive algorithm

310: quadrature demodulator

According to the present invention for achieving the object as described above, in the predistorter using an adaptive predistortion technique for nonlinear distortion compensation, a rectangular coordinate that divides the amplitude and phase of the complex signal consisting of a real part and an imaginary part of the input signal -Rectangular-to-polar Coordinator Conversion (201) part; An amplitude predistorter 202 configured to receive a coefficient of an amplitude estimation polynomial and predistort the amplitude of the input signal; A phase predistorter unit 203 for predistorting a phase of an input signal by receiving coefficients of a phase estimation polynomial; And a polar-to-rectangular coordinator conversion (204) for dividing a complex signal consisting of an amplitude and a phase into a real part and an imaginary part. do.

In addition, a predistorter using an adaptive predistortion technique for nonlinear distortion compensation, the predistorter comprising: a rectangular coordinate-polar coordinate converter for dividing an input signal into a complex number consisting of a real part and an imaginary part by amplitude and phase; An amplitude estimator for generating an estimate of ρ _y using coefficients {α _i } and ρ _x of the polynomial for predistorting the amplitude; an amplitude estimation polynomial coefficient updating unit for adaptively updating coefficients of the n-th order amplitude estimation polynomial; a first adaptive algorithm for calculating a gain value using ρ _z ; A quadrature modulator for converting a complex number signal into a transmission frequency signal using an oscillator; a phase estimation polynomial coefficient updating unit for adaptively updating the coefficients of the n-th order phase estimation polynomial; A high power amplifier for amplifying the input signal; A phase estimator for generating an estimate of θ _y using the coefficients {β _i } and ρ _y of the polynomial for predistorting the phase; a second adaptive algorithm for calculating a gain value using ρ _y ; And a quadrature demodulator configured to demodulate a signal of the RF stage.

In the following, with reference to the accompanying drawings, a preferred embodiment according to the present invention will be described in detail.

In the present invention, the distortion caused by the high power amplifier is compensated by introducing a predistorter. The distortion of the amplitude and phase due to the high power amplifier is determined by the amplitude of the input signal. The distortion characteristics are different depending on the temperature, time, and the use area of the amplifier. Therefore, an adaptive algorithm is applied to compensate for time-varying characteristics, and simple and accurate linearization can be obtained by dividing the input signal into amplitude and phase and estimating each of them as a polynomial.

FIG. 1 is a schematic diagram illustrating a case where an adaptive predistorter is added to a multi-code CDMA mobile communication system according to an embodiment of the present invention.

As shown in FIG. 1, a data source 101 is input to a parallel-to serial conversion 102 and converted into parallel data, and then multi-code spread with a code assigned to each channel (Multi-code). A digital signal is converted into pulses to reduce intersymbol interference (ISI) during transmission by a spreading 103 and a pulse shaping filter 104.

The signal x (t) having passed through the pulse shaping filter 104 is input to a predistorter 105 according to an embodiment of the present invention, and the pulse shaped signal is predistorted in advance, thereby distorting the output signal of the high power amplifier. Do not occur. On the other hand, the digital signal output from the predistorter is converted into an analog signal by the D / A (Digital Analog Converter) 108, and then a signal composed of a complex number in the quadrature modulator 110 is generated using an oscillator. The signal is converted into a signal of a radio frequency (RF), and the signal is amplified in a high power amplifier.

Meanwhile, the quadrature demodulation unit 111 receives the transmission frequency converted by the quadrature modulation unit 110 and demodulates a signal of the RF stage, and inputs the signal from an A / D (analog digital converter) 109. After converting the analog signal into a digital signal, the estimator 107 compares the output signal of the predistorter 105 with the output signal of the high power amplifier 112 to calculate the coefficient value of the polynomial to be used in the predistorter. We obtain iteratively using an adaptive algorithm and input it to the predistorter 105. In addition, the delay unit 106 delays a signal to synchronize with a signal fed back from the high power amplifier 112.

FIG. 2 is a configuration diagram of the predistorter 105 shown in FIG. 1.

First, in the rectangular-to-polar coordinator conversion (201) part, a complex number consisting of a real part and an imaginary part is divided into amplitude and phase, and in the amplitude predistoter part 202, the estimator 107 is used. Predistortion of the amplitude of the input signal is received from the amplitude estimation polynomial coefficient from the phase predistorter unit 203. The phase predistorter 203 receives the coefficient of the phase estimation polynomial from the estimator 107 and predistorts the phase of the input signal. In the polar-to-rectangular coordinator conversion (204) unit, a complex signal consisting of amplitude and phase is divided into a real part and an imaginary part.

FIG. 3 is a detailed configuration diagram for describing in more detail the functions of the delay unit 106 and the estimator 107 shown in FIG. 1.

The rectangular coordinate-polar coordinate converter 301 divides a complex number consisting of a real part and an imaginary part into amplitude and phase, and the amplitude estimator 302 uses coefficients {α _i } and ρ _x of a polynomial to predistort the amplitude. generate an estimate of ρ _y , the amplitude estimation polynomial coefficient updater 303 adaptively updates the coefficients of the n th order amplitude estimation polynomial, and the first adaptive algorithm 304 calculates a gain value using ρ _z The quadrature modulator 305 converts a complex signal into a transmission frequency signal using an oscillator, and the phase estimation polynomial coefficient updater 306 adaptively updates coefficients of the n th phase estimation polynomial, The amplifier 307 amplifies the input signal, and the phase estimator 308 generates an estimate of θ _y using coefficients {β _i } and ρ _y of the polynomial that predistort the phase, and generates a second adaptive algorithm ( 309) calculating a gain value by using a ρ _y, and quadrature demodulator 310 serves to demodulate the signal from the RF stage.

1 to 3 will be described in more detail as follows.

When the signal x (t) passing through the pulse shaping filter 104 is input, the predistorter 105 predistorts the amplitude and phase of the input signal using the polynomial obtained from the estimator 107.

When a complex signal y (t) = ρ _y (t) exp [jθ _y (t)] is input to the high power amplifier 112, the output signal is expressed by Equation 1 below.

Y (n) = H (n) x _n + V (n)

Here, V (n) is a vector representing noise, and vector x _n is a parameter vector to be estimated.

On the other hand, since the predistorter 105 requires an inverse response M ⁻¹ (ρ) of AM / AM, the estimator 107 uses an adaptive algorithm using ρ _y and ρ _z , thereby providing M ⁻¹ ( ρ) is estimated. The inverse response of AM / AM can be represented by a polynomial having N _A numbers as shown in Equation 2 below.

Since ρ _y = M ⁻¹ (ρ _z ) can be obtained from Equation 2, ρ _z is an input signal for estimating an inverse response of AM / AM, and an estimated value of ρ _y is obtained. Obtain the response characteristics of the system whose output signal is. Since the expression ρ _y = M ^-1 (ρ _z ) must be satisfied, the following expression must be satisfied.

Estimates of ρ _y and M ^-1 (ρ _z ) The error value of the liver may be expressed as in Equation 4 below.

When n data samples (n ≧ N _A ) are obtained with respect to Equation 4, this may be represented by a matrix as shown in Equation 5 below.

In addition, through the definition of Equation 6 below, Equation 5 may be represented by Equation 7.

Y (n) = H (n) x _n + V (n)

In Equation 7, V (n) is a vector representing noise, and vector x _n is a parameter vector to be estimated. To obtain x _n , the cost function J (x _n ) is defined as in Equation 8 below.

J (x _n ) = {Y (n)-H (n) x _n } ² = Y ^T (n) Y (n)-2Y ^T (n) H (n) x _n + x _n ^T H ^T (n H (n) x _n

In other words, the cost function J (x _n ) represents the absolute value of the error resulting from the difference between the predicted value and the measured value. Can be obtained by the following [Equation 9].

Where the parameter that minimizes J (x _n ) about = 0, so Is expressed by Equation 10 below.

= {H ^T (n) H (n)} ^-1 H ^T (n) Y (n) = P (n) H ^T (n) Y (n)

P (n) = {H ^T (n) H (n)} ^-1

Coefficient of polynomial that minimizes the error for n data If you find, then from (n + 1) th iteratively Can be updated. P (n + 1) when the (n + 1) th data is included is expressed by Equation 11 below.

therefore, Is expressed by Equation 12 below.

Using Equation 12, a coefficient having an inverse response characteristic of AM / AM may be cyclically obtained.

AM / PM conversion estimation also applies a similar method to AM / AM estimation. However, since phase compensation is performed by addition and subtraction, it is necessary to estimate an approximated polynomial of φ (ρ) of the high power amplifier 112. Since a signal of y = ρ _y ∠θ _y is input to the high power amplifier 112 and a signal of z = ρ _z ∠θ _z is output, the phase change amount is θ _z -θ _y , and the phase change function Φ is applied to ρ _y . Since it is a related function, a relational expression as in Equation 13 below can be obtained.

That is, when the equation [13] is applied to n pieces of data, the equation is expressed by the following equation (14).

In addition, it can be expressed as shown in [Equation 16] through the following definitions.

Y (n) = H (n) x _n + V (n)

The parameter x _n is a vector of the coefficients of the polynomial. After that, if the above-described methods are applied in order, a response curve of AM / PM can be obtained.

Meanwhile, the predistorted signal y (t) = ρ _x A [ρ _x ] ∠ {θ _x -Φ (ρ _x A [ρ _x ])} is input to the high power amplifier 112 through a modulation process. It is sent to estimators 302 and 308 shown in FIG.

The signal input to the high power amplifier 112 becomes an amplified signal z (t) and is transmitted to the receiver and tracked by the estimators 302 and 308. The signal z (t) input to the estimators 302 and 308 is divided into an amplitude ρ _z and a phase θ _z . ρ _z combines with the previous polynomial to produce an amplitude estimate of the signal y (t) input from the predistorter 105 and compares it with the actual amplitude value ρ _y to generate an estimation error. Recursive Least Squares (RLS) algorithm is applied to minimize the estimation error, and the predistorter 105 is informed about the coefficient.

The estimation of the phase distortion uses the phase change amount θ _z -θ _y by the high power amplifier 112 and the amplitude value ρ _y of y (t). Estimate the phase change amount of the high power amplifier 112 using ρ _y , generate an estimation error by comparing the actual phase change amount (θ _z -θ _y ), and then repeat the algorithm so that the generated estimation error is minimized. To do it. Information about the updated coefficient is input to the predistorter 105 to predistort the phase of the input signal. Since the estimates of amplitude and phase can be obtained in the form of polynomials respectively, they can have a simple structure.

The present invention as described above is recorded on a computer-readable recording medium, and can be processed by a computer.

As described in detail above, the present invention provides a predistorter using an adaptive predistortion technique for nonlinear distortion compensation in a next generation mobile communication system.

First, by compensating for the nonlinear distortion of the high power amplifier, it is possible to use a high power amplifier with high efficiency in the wireless terminal, thereby extending the use time of the terminal.

Second, since the response of the amplifier is estimated using an adaptive algorithm, time-varying nonlinear distortion can be compensated for.

Third, accurate linearization can be obtained because complex input signals are compensated by being divided into amplitude and phase.

Fourth, since the response of the amplifier is estimated in the form of a polynomial, the structure of the predistorter is simplified.

The technical spirit of the present invention has been described above with reference to the accompanying drawings, but this is by way of example only and not by way of limitation. In addition, it is obvious that any person skilled in the art may make various modifications and imitations without departing from the scope of the technical idea of the present invention.

Claims (2)

In the predistorter using the adaptive predistortion technique for nonlinear distortion compensation, A rectangular-to-polar coordinator conversion unit for separating an amplitude and a phase from a complex number consisting of a real part and an imaginary part of an input signal; An amplitude predistorter for predistorting the amplitude of the input signal by receiving coefficients of the amplitude estimation polynomial; A phase predistorter unit for predistorting a phase of an input signal by receiving coefficients of a phase estimation polynomial; And A predistorter using an adaptive predistortion technique comprising a polar-to-rectangular coordinator conversion (204) unit for dividing a complex signal consisting of amplitude and phase into a real part and an imaginary part. In the predistorter using the adaptive predistortion technique for nonlinear distortion compensation, A rectangular coordinate-polar coordinate converter for dividing an input signal into a complex number consisting of a real part and an imaginary part by amplitude and phase; An amplitude estimator for generating an estimate of ρ _y using coefficients {α _i } and ρ _x of the polynomial for predistorting the amplitude; an amplitude estimation polynomial coefficient updating unit for adaptively updating coefficients of the n-th order amplitude estimation polynomial; a first adaptive algorithm for calculating a gain value using ρ _z ; A quadrature modulator for converting a complex number signal into a transmission frequency signal using an oscillator; a phase estimation polynomial coefficient updating unit for adaptively updating the coefficients of the n-th order phase estimation polynomial; A high power amplifier for amplifying the input signal; A phase estimator for generating an estimate of θ _y using the coefficients {β _i } and ρ _y of the polynomial for predistorting the phase; a second adaptive algorithm for calculating a gain value using ρ _y ; And A predistorter using an adaptive predistortion technique comprising a quadrature demodulator for demodulating a signal of an RF stage.