KR101470817B1 - Apparatus and method of predistortion of multiple nonlinear amplifiers with single feedback loop - Google Patents

Abstract

The present invention relates to a predistortion apparatus and method for a nonlinear amplifier, and more particularly, to an apparatus and method for predistorting each nonlinear amplifier using a single feedback signal for a plurality of nonlinear amplifiers.
The present invention relates to a nonlinear amplifier having a nonlinear amplification characteristic according to the magnitude of an input signal, a single feedback circuit for combining signals branched from a plurality of output signals and feeding back the combined signals into one signal, And a predistortion compensator for predicting amplification characteristics of the amplifier, estimating a predistortion characteristic according to the amplification characteristic, and predistorting each input signal based on the predistortion characteristic, and using a single feedback circuit for a plurality of nonlinear amplifiers And a plurality of nonlinear amplifiers are constituted by a plurality of feedback circuits to precompensate nonlinear characteristics of the respective amplifiers. According to the present invention, by using a single feedback circuit for a plurality of nonlinear amplifiers This is a device that compensates the nonlinear characteristics of the amplifier The system can be implemented using only a single number of components such as a frequency down converter, an analog-digital converter, and a filter, thereby improving the hardware complexity and reducing the implementation cost It is effective.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a predistortion of multiple nonlinear amplifiers with a single feedback loop using a single feedback circuit for a plurality of nonlinear amplifiers,

The present invention relates to a predistortion apparatus and method for a nonlinear amplifier, and more particularly, to an apparatus and method for predistorting each nonlinear amplifier using a single feedback signal for a plurality of nonlinear amplifiers.

In a mobile communication system using a radio frequency (RF) signal, amplifiers can be roughly divided into a low noise amplifier and a high power amplifier. In the case of a high power transmission amplifier, the efficiency is more problematic than the noise characteristic. Therefore, the high power transmission amplifier often operates close to the nonlinear operating point in order to obtain high efficiency, which causes various problems. For example, if a power amplifier, which is a typical example of a high power transmission amplifier, is considered, adjacent channel interference due to intermodulation distortion, signal constellation, EVM (Error Vector Magnitude) And the like may be caused.

The simplest way to solve this problem is to reduce the output of the power amplifier to operate in a linear region. However, this causes a problem that the efficiency of the power amplifier is greatly reduced. In particular, when the average power ratio to the peak value of the signal is high, such as an OFDMA signal or a CDMA (Code Division Multiple Access) signal, the efficiency of the power amplifier is further lowered.

The digital predistortion scheme is used to compensate for nonlinearity without degrading the efficiency of the power amplifier. The digital predistortion scheme is an inverse function of the nonlinear amplification characteristic function of the power amplifier at the digital baseband stage And linearizing the power amplifier by implementing the corresponding predistortion. A study on digital predistortion has been carried out on various schemes and structures. Among them, a method of modeling and implementing a power amplifier and a digital predistorter using a polynomial is widely used.

However, various applications using multiple channels such as Multiple Input Multiple Output (MIMO) have been implemented for signal attenuation, interference, and efficient use of limited frequency resources in a wireless communication environment according to the development of a mobile communication environment. In this case, usually one power amplifier is used for each channel. In this case, since the signals of the respective channels are independent, it is common to configure a separate feedback circuit for improving the linearity for each power amplifier.

However, in order to construct a separate feedback circuit for each power amplifier and improve its linearity through predistortion, a frequency down converter (Down Converter), an analog-to-digital converter The cost of system implementation is greatly increased and the hardware configuration becomes complicated. In contrast to this, there is a prior art in which a separate feedback circuit is configured for each power amplifier, and the components of the feedback circuit are implemented as one module or an integrated circuit to constitute a system. However, It is difficult to conclude that there is a large difference in terms of configuration, complexity, or cost of the system.

SUMMARY OF THE INVENTION The present invention has been made in order to solve the problems of the prior art as described above, and it is an object of the present invention to provide a nonlinear amplifier which is capable of using a single feedback circuit to calculate amplification characteristics of a plurality of nonlinear amplifiers, And an apparatus and method for performing precompensation for each nonlinear amplifier.

According to an aspect of the present invention, there is provided a predistortion apparatus using a single feedback circuit for a plurality of nonlinear amplifiers, including: a plurality of nonlinear amplifiers whose amplification characteristics are non-linear according to the magnitude of an input signal; A single feedback circuit for branching and merging a plurality of signals and feeding back the signals as one signal; And a precompensator for predicting amplification characteristics of the nonlinear amplifiers from the feedback signal, estimating a predistortion characteristic according to the amplification characteristics, and predistorting each input signal based on the predistortion characteristic.

Here, the single feedback circuit may include a signal distributor for splitting a part of the signal from the signal, and a signal combiner for merging a plurality of signals into a single signal.

The precompensator includes a predistortion controller for determining an amplification characteristic of the nonlinear amplifier and calculating a predistortion characteristic according to the amplification characteristic, and a predistortion controller for predistorting an input signal based on the predistortion characteristic received from the predistortion controller, And a predistortion filter for outputting the predistortion filter.

According to still another aspect of the present invention, there is provided a predistortion method using a single feedback circuit for a plurality of nonlinear amplifiers, comprising: (a) combining output signals of a plurality of nonlinear amplifiers or signals derived therefrom and feeding back the combined signals to one signal; Calculating amplification characteristics of each of the nonlinear amplifiers based on the one feedback signal and calculating a predistortion characteristic according to the amplification characteristics; And (c) performing predistortion on an input signal of each of the nonlinear amplifiers based on the predistortion characteristic.

In the calculating of the amplification characteristics of the nonlinear amplifiers in the step (b) or the calculation of the respective precompensation characteristics, the amplification characteristic or the predistortion characteristic of the nonlinear amplifier is approximated by a sum of polynomials having a series of degrees A polynomial model can be used.

In the step (b), an adaptive algorithm may be used to grasp the amplification characteristics of the nonlinear amplifiers or to calculate the respective precompensation characteristics by converging the calculation results to the optimal values as the calculation is repeated.

According to the present invention, compared to the prior art in which a plurality of feedback circuits are constituted for a plurality of nonlinear amplifiers and the nonlinear characteristics of the amplifiers are predistorted, a nonlinear characteristic of the amplifier is inverted by using a single feedback circuit for a plurality of nonlinear amplifiers The system is implemented using only a single number of components such as a frequency down converter, an analog-digital converter, and a filter, thereby improving the hardware complexity and reducing the implementation cost There is an effect that can be saved.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.
FIG. 1 illustrates a structure of a predistorter using a single feedback circuit for a plurality of nonlinear amplifiers according to an embodiment of the present invention. FIG.
FIG. 2 is a flowchart for calculating predistortion characteristics for a plurality of nonlinear amplifiers according to an embodiment of the present invention; FIG.
3 is a block diagram of a verification simulation model in accordance with an embodiment of the present invention.
FIG. 4 is a graph showing a convergence curve of a precompensation cost function as a result of verification simulation according to an embodiment of the present invention.
5 is a graph of spectrum performance comparison of a power amplifier according to an embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is capable of various modifications and various embodiments, and specific embodiments will be described in detail below with reference to the accompanying drawings.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

The terms first, second, etc. may be used to describe various components, but the components are not limited by the terms, and the terms are used only for the purpose of distinguishing one component from another Is used.

In view of the fact that the prior art realizes only individual predistortion of amplifier units based on a plurality of feedback circuits for a plurality of nonlinear amplifiers, the output signals of a plurality of nonlinear amplifiers are merged and fed back as one signal From this, it is possible to understand the amplification characteristics of each nonlinear amplifier, calculate the precompensation characteristics of each nonlinear amplifier, and perform precompensation based on it, thereby realizing precompensation for a plurality of nonlinear amplifiers using a single feedback circuit .

1 shows the structure of a predistortion apparatus 100 using a single feedback circuit for a plurality of nonlinear amplifiers according to an embodiment of the present invention. Referring to FIG. 1, a predistorter 100 using a single feedback circuit for a plurality of nonlinear amplifiers according to an embodiment of the present invention includes a predistorter 110, a plurality of nonlinear amplifiers 120 A non-linear amplifier, a plurality of signal distributors 130, one or more signal converters 140, a plurality of antennas 150, a plurality of frequency up converters 160, a frequency down converter 170, A down converter, and a local oscillator 180. [0033] FIG. In addition to the above-described module, a filter, a signal delay, an AD / DA converter (Analog / Digital), and the like may be added. have.

The predistortion unit 110 obtains amplification characteristics of a plurality of nonlinear amplifiers 120 from a single feedback signal, estimates each precompensation characteristic according to the amplification characteristics, and then precompensates and outputs each input signal based thereon. Here, the amplification characteristic of the nonlinear amplifier 120 refers to the nonlinear characteristic of the output signal magnitude with respect to the magnitude of the input signal of the nonlinear amplifier 120, and the precompensation characteristic refers to a characteristic of improving the amplification characteristic of the nonlinear amplifier 120 Distortion characteristic of the output signal with respect to the input signal of the predistorter 110 required to compensate the distortion of the output signal.

The precompensator 110 further includes a predistortion controller 114 for grasping the amplification characteristics of the nonlinear amplifier 120 and calculating a predistortion characteristic according to the amplification characteristic and a precompensation controller 114 for receiving the predistortion characteristic from the predistortion controller 114, And a predistortion filter 112 for predistorting the input signal and outputting it. The signal output from the predistorter 110 is applied to a frequency up-converter 160.

The frequency up converter 160 converts the baseband signal into a high frequency band allocated for wireless communication. A mixer or the like may be used as the frequency up converter 160. The mixer mixes the local oscillation signal generated from the local oscillator 180 with the baseband signal for frequency conversion, . The local oscillator 180 may be configured using a phase locked loop using a reference clock. The high-frequency converted signal is applied to the non-linear amplifier 120.

The nonlinear amplifier 120 amplifies the input signal and emits the amplified signal through the antenna 150. A typical example of the nonlinear amplifier 120 is a power amplifier 120. Since the power amplifier 120 is generally operated close to a nonlinear operating point in order to obtain high efficiency, Intermodulation distortion may be a problem. In order to solve this problem, the linearity of the power amplifier 120 is improved through predistortion or the like.

The signal distributor 130 divides a part of the signal amplified and output by the nonlinear amplifier 120. The branched signal can be used for improving the linearity of the nonlinear amplifier 120 through a feedback circuit or the like. The signal distributor 130 may be configured using a directional coupler or the like for branching a part of a high-frequency signal flowing in a specific direction.

The signal combiner 140 combines a plurality of signals branched by the signal distributor 130 into a single signal. So that the output signals of the plurality of power amplifiers 120 can be fed back using a single feedback circuit. The signal combiner 140 may include an N-way combiner for summing and outputting a plurality of high frequency input signals.

The frequency down converter 170 converts the high frequency band signal into the baseband signal. Frequency up converter 160. The local oscillator 180 receives the local oscillation signal from the local oscillator 180 and mixes the local oscillation signal with a high frequency band signal to convert the signal into a desired frequency signal do. The baseband signal passed through the frequency down converter 170 is again applied to the precompensator 110.

Hereinafter, a predistortion method using a single feedback circuit for a plurality of nonlinear amplifiers 120 according to an embodiment of the present invention will be described.

FIG. 2 illustrates a calculation flowchart of predistortion characteristics for a plurality of power amplifiers 120 according to an embodiment of the present invention. (B) a predistorter (c) using a single feedback circuit according to an embodiment of the present invention; and (c) a predistorter using a single feedback circuit according to an embodiment of the present invention. Figure 2 shows a block diagram of a simplified model of a predistortion device using a single feedback circuit according to an embodiment. 3C is a block diagram of a simplified model for clearly explaining and verifying the gist of the present invention. The frequency up-converter 160 and the frequency down-converter 170, which are generally included in a mobile communication system, . However, the frequency up-converter 160 and the frequency down-converter 170 are not necessarily included in a general non-linear amplifier circuit. In the case of a mobile communication system, the frequency up-converter 160 and the frequency down- The linearity of the output signal with respect to the magnitude of the input signal becomes a problem rather than the characteristic depending on the frequency. However, even if the amplifier is considered as a simplified model of the present invention, And thus it is not only a problem in applying it again, but also helps to understand the gist of the present invention. Therefore, an embodiment of the present invention will be described below using the simplified model of FIG. 3 (c).

는 각 전치보상기(110)를 거쳐 복수의 전치보상 신호

로 출력되고, 이는 다시 복수의 전력증폭기(120)에 인가되어 복수의 출력 신호

로 출력되며, 각 출력 신호들은 신호 분배기(130)에 의하여 분기된 후 다시 신호 병합기(140)에 의하여 병합되어 하나의 피드백 신호

로 전치보상기(110)로 인가되게 된다.As shown in FIG. 3 (c), a plurality of input signals

The pre-compensator 110 generates a plurality of pre-

Which is again applied to a plurality of power amplifiers 120 to produce a plurality of output signals < RTI ID = 0.0 >

And the output signals are branched by the signal distributor 130 and then merged by the signal combiner 140 to form a single feedback signal

To the precompensator 110. The pre-

(i=1,2,...,M)이라고 할 때 상기 각 전력증폭기(120)의 출력 신호

은 아래의 <수학식 1>과 같이 표현될 수 있다.The amplification characteristics of the power amplifier 120 can be modeled as polynomials, and the amplification characteristics of each power amplifier 120

(i = 1, 2, ..., M), the output signal of each of the power amplifiers 120

Can be expressed as Equation (1) below.

Equation 1

과

은 크기가 L x 1인 벡터가 된다. 이때

이고, 이때

은 m번째 전력증폭기(120)의 증폭 특성 계수를 나타내며

이 된다. In Equation (1), the maximum polynomial degree of the amplification characteristic polynomial of the power amplifier 120 is 2L-1,

and

Is a vector of size L x 1. At this time

Lt; / RTI &gt;

Represents an amplification characteristic coefficient of the m-th power amplifier 120

.

In addition, although the case of not having the memory effect in Equation (1) is considered, it can be easily coped with using the memory polynomial.

As can be seen from Equation (1), since the power amplifier 120 has a nonlinear amplification characteristic, it is difficult to obtain the inverse function in order to obtain the predistortion characteristic, and the inverse function is also approximated by a polynomial. In this case, the inverse function of the power amplifier 120 can be expressed as Equation (2) below.

Equation 2

과

은 크기가 Q x 1인 벡터가 된다. 이때

은 m번째 전치보상 특성 다항식의 계수를 나타내는데

을 만족해야 한다. 상기 <수학식 2>에서

이며

가 된다.In this case, the inverse function of the power amplifier 120, that is, the maximum degree of the predistortion characteristic polynomial becomes 2Q-1,

and

Is a vector of size Q x 1. At this time

Represents the coefficient of the m-th predistortion compensation polynomial

. In Equation (2)

And

.

In the following, a predistortion method using a single feedback circuit for a plurality of nonlinear amplifiers 120 according to an embodiment of the present invention, based on the polynomial model of the nonlinear amplifier 120 and the predistorter 110, . First, when the predistortion method starts (S210), a step S220 of branching a part of the output signal of each nonlinear amplifier 120 is performed. In this case, it is assumed that a signal having a magnitude of 1 / K ₀ of the output signal is branched. The feedback signal a (n) has the same value as in Equation (3) below through the process of merging the output signals branched again to produce a single feedback signal (S230).

Equation 3

)를 계산한 후, 상기 실측치와 비교하여 그 차이를 최소화하는 증폭 특성 다항식 계수의 추정치

를 구하고, 이로부터 전력증폭기(120)의 증폭 특성에 가장 근접하는 다항식을 얻게 된다.Then, the amplification characteristics of each power amplifier 120 are grasped based on the feedback signal (S240). Since the feedback signal a (n) is applied to the predistorter 110 and its measured value can be known, the amplification characteristic of the power amplifier 120 is obtained by using the feedback signal a (n) . The amplification characteristic of the power amplifier 120 is expressed as a polynomial model The estimated value of the feedback signal (Equation (4)) < RTI ID = 0.0 >

), And then calculates an estimated value of the amplification characteristic polynomial coefficient that minimizes the difference with the measured value

And the polynomial closest to the amplification characteristic of the power amplifier 120 is obtained from this.

Equation 4

이며

이고

는 전력증폭기(120)의 이득을 나타낸다. 따라서 본 발명의 실시를 위해서는 전력증폭기(120)의 이득을 미리 알고 있거나 추정할 수 있어야 한다. 상기

는 M개의 전력증폭기(120) 다항식 모델의 계수의 추정치(

)를 모두 연결하여 만든 벡터이며 크기는 ML x 1 이다. 마찬가지로

도 M개의 전치보상기(110) 출력 신호(

)를 연결하여 만든 벡터이며 크기도 ML x 1 이다. 즉, 전력증폭기(120)를 하나만 사용하는 경우에 비해 안테나 개수만큼 벡터의 크기가 커진다. <수학식 3>은 병합된 신호가 전력증폭기(120) 이득만큼 감쇄되어 피드백됨을 의미한다.In Equation (4)

And

ego

Represents the gain of the power amplifier 120. Therefore, in order to implement the present invention, the gain of the power amplifier 120 must be known or estimated. remind

Is an estimate of the coefficients of the M power amplifier 120 polynomial models (

), And the size is ML x 1. Likewise

The M pre-compensator 110 output signals (

) And the size is ML x 1. That is, the size of the vector increases by the number of antennas as compared with the case where only one power amplifier 120 is used. Equation (3) means that the merged signal is attenuated by the gain of the power amplifier 120 and fed back.

는 전치보상기(110)가 출력하는 값이므로 이미 알려진 값이므로, 이를 이용하여

를 구함으로써 M개의 전력증폭기(120)의 모든 증폭 특성 다항식 모델의 계수를 알아낼 수 있고, 이는 전력증폭기(120)의 증폭 특성을 구할 수 있음을 의미한다. 이를 구하기 위하여 다음 <수학식 5>와 같은 비용함수(Error function)를 정의한다.In Equation (3)

Is a known value since it is a value outputted by the predistorter 110,

The coefficients of all the amplification characteristic polynomial models of the M power amplifiers 120 can be obtained, which means that the amplification characteristics of the power amplifier 120 can be obtained. In order to obtain this, an error function such as the following Equation (5) is defined.

Equation 5

을 나타낸다. <수학식 5>의 비용함수를 최소화하는

를 찾기 위한 적응형 알고리즘은 하기 <수학식 6>과 같이 쓸 수 있다. 여기서 적응형 알고리즘이라 함은 계산이 반복됨에 따라 그 계산 결과치가 최적치에 수렴하게 되는 알고리즘으로서, 전력증폭기(120)의 증폭 특성은 시간이나 온도에 따라서 변하기 때문에 적응형 알고리즘을 사용하여 그 값을 구하고, 또한 지속적으로 갱신하여 주는 것이 바람직하다. Where E [] is the mean

. Minimizing the cost function of Equation (5)

An adaptive algorithm for searching for &quot; Equation 6 &quot; Here, the adaptive algorithm is an algorithm in which the calculation result is converged to an optimum value as the calculation is repeated. Since the amplification characteristic of the power amplifier 120 changes with time or temperature, the adaptive algorithm is used to obtain the value using an adaptive algorithm , And it is also desirable to continuously renew it.

Equation 6

는 적응형 알고리즘의 각 계수별로 서로 다른 가중치를 주기 위한 대각행렬이다. 이 대각행렬을 통해 각 계수별로 서로 다른 가중치를 주는 이유는

의 각 원소가 서로 다른 비선형 차수를 가지고 있어서 그 값의 크기가 원소별로 큰 편차를 보이기 때문이다. 이때 μ는 가능한 작은 값을 사용하는 것이 좋다. 전력증폭기(120)의 증폭 특성은 시간에 따라 빠르게 변하지 않기 때문에 천천히 수렴하더라도 추정 정확도를 높이는 것이 중요하기 때문이다.

에 대한 대각행렬을 구하는 명확한 규칙은 없고 실험적으로 찾아야 한다. 적응형 알고리즘이 수렴한 후의 증폭 특성 다항식 계수의 추정치 값을

라고 하면 이로부터 각 전력증폭기(120)의 증폭 특성을 얻을 수 있는데 이는 전력증폭기(120)의 증폭 특성 다항식 계수

는 M개의 증폭 특성 다항식의 계수로 이루어져 있기 때문이다(

). 즉

에서 순서대로 L개씩 원소를 추출하면 이는 각 전력증폭기(120)의 증폭 특성 다항식의 계수가 된다. 따라서 각 전력증폭기(120)의 증폭 특성 다항식 계수

을 각각 구분해서 추출할 수 있고, 결국 첫 번째 전력증폭기(120)의 증폭 특성을

로부터 얻을 수 있고 같은 방법으로 M개의 전력증폭기(120)의 증폭 특성을 각각 구할 수 게 된다.Where μ is a coefficient that determines the update step size

Is a diagonal matrix for giving different weights to each coefficient of the adaptive algorithm. The reason why we give different weights to each coefficient through this diagonal matrix is

Since each element of the matrix has a different nonlinear order, the magnitude of the value shows a large variation by element. In this case, μ should be as small as possible. Since the amplification characteristic of the power amplifier 120 does not change rapidly with time, it is important to increase the estimation accuracy even if it slowly converges.

There is no clear rule for finding diagonal matrices for. The estimated value of the amplification polynomial coefficients after the adaptive algorithm converges is

The amplification characteristics of each power amplifier 120 can be obtained. This is because the amplification characteristic polynomial coefficients of the power amplifier 120

Because it consists of the coefficients of the M amplification polynomials (

). In other words

The L amplitudes of the power amplifiers 120 are the coefficients of the amplification characteristic polynomials. Therefore, the amplification characteristic polynomial coefficient of each power amplifier 120

The amplification characteristics of the first power amplifier 120 can be extracted

And amplification characteristics of the M power amplifiers 120 can be obtained in the same manner.

Since the amplification characteristics of the respective power amplifiers 120 have been extracted, a step S250 is performed to grasp the respective pre-compensation characteristics based on the amplification characteristics. In order to estimate the predistortion characteristic, a cost function such as Equation (7) is defined with respect to the m-th power amplifier 120.

Equation 7

이다. 즉, m번째 전치보상기(110)의 입력 신호와 m번째 전력증폭기(120) 출력신호를 이득으로 나눈 값 사이의 차이를 에러로 정의하여 이 에러가 최소가 되는 전치보상 특성을 찾는다. 그런데 피드백 경로가 하나밖에 없으므로 m번째 전력증폭기(120)의 출력인

을 얻을 수 없다. 따라서,

대신에 그 추정값을 사용할 수 밖에 없고, 이 추정값은 다음 <수학식8>과 같이 표현될 수 있다.In Equation (7)

to be. That is, a difference between a value obtained by dividing an input signal of the mth pre-compensator 110 and an output signal of the mth power amplifier 120 by an error is defined as an error, and a predistortion characteristic in which the error is minimized is searched. However, since there is only one feedback path, the output of the m-th power amplifier 120

Can not be obtained. therefore,

Instead, the estimated value must be used, and this estimated value can be expressed as Equation (8).

Equation 8

과

은 각각 m번째 전력증폭기(120)의 증폭 특성에 대한 추정 다항식 계수와 입력 신호 벡터로 <수학식 1>에서 정의 되었다. 따라서 상기 <수학식 7>의 비용함수는 다음 <수학식 9>와 같이 변형하여 나타낼 수 있다.here

and

silver The estimated polynomial coefficients and the input signal vector for the amplification characteristics of the m-th power amplifier 120 are defined in Equation (1). Therefore, the cost function of Equation (7) can be expressed by the following Equation (9).

Equation 9

은 m번째 전치보상기(110)의 출력임을 고려하면 (<수학식2> 참조), <수학식 9>로부터 <수학식 7>의 비용함수를 최소화 하는 계수

를 구하는 적응형 알고리즘은 다음 <수학식 10>과 같이 정리될 수 있다.Input of the mth power amplifier

(7) is a coefficient that minimizes the cost function of Equation (7) from Equation (9), considering that it is the output of the mth pre-compensator (110)

Can be summarized as Equation (10). &Lt; EMI ID = 10.0 >

Equation 10

은 갱신 계단 크기,

은 가중 대각행렬이고,

,

로 정의되는 벡터로서 전치보상기(110)의 출력 신호의 조합이므로 알 수 있는 값들이고,

은 전치보상기(110)에 대한 입력 신호 벡터로서 역시 알 수 있는 값이므로, 결국 앞서 살핀 전력증폭기(120)의 증폭 특성의 파악과 같은 과정을 통하여 상기 <수학식 10>의 적응형 알고리즘을 사용하여, 상기 <수학식 7>의 비용함수를 최소화할 수 있는

을 구할 수 있고, 또한 전력증폭기(120)의 증폭 특성의 파악과 같은 방법을 통하여 이로부터 각 전력증폭기(120)에 대한 전치보상 특성을 찾을 수 있게 된다.At this time,

The updated step size,

Is a weighted diagonal matrix,

,

Is a combination of output signals of the precompensator 110, which are known values,

silver The adaptive algorithm of Equation (10) is used to determine the amplification characteristic of the power amplifier 120, which is also known as the input signal vector to the predistorter 110, The cost function of Equation (7) can be minimized

In addition, it is possible to find the predistortion characteristic for each of the power amplifiers 120 through a method such as grasping the amplification characteristics of the power amplifier 120.

Finally, a step S260 of applying the predistortion characteristic obtained through the series of procedures to each input signal and outputting the predistortion signal is applied. The predistortion signal is applied to the power amplifier 120 and amplified. The linearity of the power amplifier 120 can be improved by passing through the predistortion compensation.

In order to verify the performance of the predistortion apparatus and method using a single feedback circuit for a plurality of nonlinear amplifiers according to an embodiment of the present invention, a computer simulation is performed, and the experimental environment is as follows.

Two transmit antennas 150 and two independent data streams are transmitted through the two antennas 120. [ Each stream is modulated by 16-QAM (Quadrature Amplitude Modulation) and the waveform shaping filter is a square root raised cosine (RRC) filter with a roll-off index of 0.25. The sampling clock at which predistortion operates is ten times the symbol rate. The Saleh model according to Equation (11) is used as the m-th power amplifier 120 model.

Equation 11

과

은 전력증폭기(120)의 특성을 결정하는데, 첫 번째 전력증폭기(120)에는

,

의 값을 사용하였고 두 번째 전력증폭기(120)에는

,

의 값을 사용하였다. 전력증폭기(120)의 이득은 1이라고 가정하였다 (

). 전력증폭기(120)의 모델과 전치보상기(110)의 모델로는 모두 차수 7의 다항식을 적용하였다 (L=Q=4). 전치보상 특성을 구하는데 사용된 갱신 계단 계수는

이고, 대각행렬

이 사용되었다. 전력증폭기(120) 또는 전치보상기(110)의 모델에 있어 다항식의 차수를 높이면 정밀도를 높일 수 있으나, 이 경우 계산이 요구되는 다항식 계수 벡터의 크기가 증가하여 복잡도가 증가하기 때문에 실험을 통하여 적절한 값을 선정해야 한다.

and

Determines the characteristics of the power amplifier 120, the first power amplifier 120

,

And the second power amplifier 120 uses a value of

,

Was used. It is assumed that the gain of the power amplifier 120 is 1 (

). The polynomial of order 7 is applied to the model of the power amplifier 120 and the model of the predistorter 110 (L = Q = 4). The updated stepping coefficient used to obtain the predistortion characteristic is

, A diagonal matrix

Was used. In the model of the power amplifier 120 or the predistorter 110, if the order of the polynomial is increased, the accuracy can be increased. In this case, since the complexity increases due to an increase in the size of the polynomial coefficient vector required for calculation, .

FIG. 4 shows a convergence curve graph of the precompensation cost function (error function) according to the verification simulation result according to an embodiment of the present invention. FIG. 4A shows a convergence curve of the first power amplifier 120 precompensation cost function (Equation 7) used in the simulation, FIG. 4B shows a convergence curve of the second power amplifier 120 precompensation cost function And FIG. 4 (c) shows the convergence curve of the precompensation cost function when a plurality of feedback circuits are used. As can be seen from the respective figures, the precompensation cost function converges It can be seen that the state after stabilization shows a slightly worse convergence characteristic as compared with the structure using a plurality of feedback circuits according to an embodiment of the present invention. the mean square error value is about 10 ^-10 , and according to the embodiment of the present invention, the MSE value is about 10 ^-9 . However, when the difference is not large and the error value is 10 ^-9 And the spectral performance, which is a more important indicator as shown below, shows excellent characteristics.

5 is a graph illustrating a spectral performance comparison of a power amplifier according to an embodiment of the present invention. FIG. 5 (a) shows the spectral characteristics of the first power amplifier 120 used in the simulation, and FIG. 5 (b) shows the second power amplifier 120. In the case where the predistorter 110 is absent (FIG. 5A and FIG. 5B (1)), the output of the power amplifier 120 is severely distorted and the spectrum is spread widely, have. On the other hand, it can be confirmed that such distortions are effectively removed when predistortion is added (FIGS. 5A and 5B (2)). Specifically, adjacent channel interference of about 20 dB or more is reduced. It can be confirmed that almost similar performance is obtained even when a plurality of feedback circuits are used (FIG. 5 (a) and FIG. 5 (b) (3)). 5 (a) and 5 (b) (4) show input signals of the power amplifier 120. FIG.

Also, EVM (Error Vector Magnitude) characteristics are compared in order to compare the quality of the transmission signal among the problems due to the nonlinearity of the power amplifier 120. EVM is defined as Equation (12).

Equation 12

Without the predistortion compensation, the EVMs of the two power amplifier 120 outputs were 9.49% and 9.00%, respectively. It can be seen that the quality of the signal is severely distorted in view of the fact that the EVM standard suggests about 5% within the mobile communication standard. The EVM of the two power amplifiers were 0.056% and 0.061% when predistortion was applied through multiple feedback circuits. On the other hand, according to the embodiment of the present invention, when the precompensation is applied, 0.078% and 0.081% are shown, respectively. It can be confirmed that the EVM according to an embodiment of the present invention shows a slightly poor performance as compared with the method using a plurality of feedback circuits, but satisfies the general EVM standard satisfactorily.

의 각 L 개 원소들의 블록이 과연 실제 전력증폭기(120)를 제대로 모델링하고 있는지, 다시 얘기하면 다중 되먹임 회로를 구성하여 얻을 수 있는 값과 같은지에 대하여 살핀다면 아래와 같다.In addition, in order to confirm that the amplification characteristics of the plurality of power amplifiers 120 can be correctly estimated through the present invention, the estimation value of the vector created by enumerating the coefficients of the amplification characteristic polynomial of each power amplifier 120

If the block of each of the L elements of the power amplifier 120 is properly modeling the actual power amplifier 120 and again, it is equal to the value obtained by constructing the multiple feedback circuit.

A signal fed back from the output of each power amplifier 120 can be expressed by Equation (13).

Equation 13

은 전력증폭기(120)의 증폭 특성의 추정치 에러에 기인한 전력증폭기(120) 출력 신호의 추정치 에러를 나타내는데 평균이 0이고 서로 다른 안테나(150)에서 이 값들은 서로 상관관계가 없다고 할 수 있다. 안테나 개수를 2개라고 가정하고(M=2), <수학식 13>을 이용하면 <수학식 5>의 비용함수는 다음 <수학식 14>와 같이 다시 쓰여질 수 있다.At this time

May represent an estimate error of the power amplifier 120 output signal due to an estimation error of the amplification characteristic of the power amplifier 120, which may be said to be zero on average and that these values are not correlated to each other at different antennas 150. Assuming that the number of antennas is 2 (M = 2) and using Equation (13), the cost function of Equation (5) can be rewritten as Equation (14).

Equation 14

과

는 각각

의 첫 L개의 원소로 만든 벡터, 두 번째 L개의 원소로 만든 벡터를 의미하고 각각 첫 번째, 두 번째의 전력증폭기(120)의 증폭 특성의 다항식 모델 계수를 나타낸다. 그리고

,

로 정의될 수 있다.

과

는 각각 복수의 피드백 회로를 구성한 기존의 방식에서 첫 번째 전력증폭기(120)의 비용함수, 두 번째 전력증폭기(120)에서의 비용함수를 나타내므로 본 발명은 기존 방식의 두 비용함수의 합을 최소화 하는 해를 구하고 있음을 알 수 있다. 따라서 상기 <수학식 5>의 비용함수를 최소화 하는 해는 각 전력증폭기(120)에서의 비용함수를 모두 최소화함을 알 수 있고, 이와 같은 분석을 통해 복수개의 전력증폭기(120) 신호가 더해지더라도 이로부터 각 전력증폭기(120)의 증폭 특성을 추정할 수 있음을 알 수 있다.In Equation (14)

and

Respectively

And a polynomial model coefficient of the amplification characteristics of the first and second power amplifiers 120, respectively. And

,

. &Lt; / RTI >

and

The present invention minimizes the sum of the two cost functions of the conventional method because the cost function of the first power amplifier 120 and the cost function of the second power amplifier 120 are represented in the conventional method in which a plurality of feedback circuits are formed. It can be seen that the Therefore, it can be seen that the solution minimizing the cost function of Equation (5) minimizes the cost function in each power amplifier 120. Even if a plurality of power amplifier 120 signals are added through the analysis It can be seen from this that the amplification characteristics of each power amplifier 120 can be estimated.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments described in the present invention are not intended to limit the technical spirit of the present invention but to illustrate the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents thereof should be construed as being included in the scope of the present invention.

Claims (6)

A plurality of nonlinear amplifiers having nonlinear amplification characteristics according to the magnitude of an input signal;
A single feedback circuit for branching the output signals of the plurality of nonlinear amplifiers, merging the output signals of the plurality of nonlinear amplifiers into one signal and feeding back the signals through a single line;
And a predistorter for predicting an amplification characteristic of each of the nonlinear amplifiers from the feedback signal and estimating a predistortion characteristic according to the amplification characteristic, and predistorting each input signal based on the predistortion characteristic. A predistorter using a single feedback circuit for a nonlinear amplifier. The method according to claim 1,
Wherein the single feedback circuit comprises:
A signal distributor for branching a part thereof from the signal,
And a signal combiner for combining the plurality of signals into a single signal. 2. The predistortion apparatus according to claim 1, wherein the plurality of nonlinear amplifiers comprise a single feedback circuit. 3. The method according to claim 1 or 2,
Wherein the predistorter comprises:
A predistortion controller for grasping an amplification characteristic of the nonlinear amplifier and calculating a predistortion characteristic according to the amplification characteristic;
And a predistortion filter that receives the predistortion characteristic from the predistortion controller and predistort the input signal based on the predistortion characteristic and outputs a predistortion signal. The present invention relates to a predistorter using a single feedback circuit for a plurality of nonlinear amplifiers, . (a) branching output signals of a plurality of nonlinear amplifiers or signals derived therefrom, merging the signals into a single signal, and feeding the signals through a single line;
(b) grasping amplification characteristics of the nonlinear amplifiers based on the one feedback signal and calculating a predistortion characteristic according to the amplification characteristics; And
(c) performing predistortion on an input signal of each of the nonlinear amplifiers based on the predistortion characteristic, wherein a single feedback circuit is used for a plurality of nonlinear amplifiers. 5. The method of claim 4,
Calculating the amplification characteristics of each of the nonlinear amplifiers in the step (b) or approximating the amplification characteristics or the precompensation characteristics of the nonlinear amplifier by a sum of polynomials having a series of orders, Wherein a single feedback circuit is used for a plurality of nonlinear amplifiers. 5. The method of claim 4,
Wherein in the step (b), an adaptive algorithm is used in order to grasp the amplification characteristics of the nonlinear amplifiers or to calculate the respective precompensation characteristics by converging the calculation result to an optimum value as the calculation is repeated. A precompensation method that uses a single feedback circuit for a nonlinear amplifier of.

Images