KR20060098681A - Apparatus and method for compensating long tim memory effect of power amplifier in a wireless communication system - Google Patents

Abstract

The present invention relates to a pre-distortion device for linearizing the nonlinear distortion characteristics of a high output power amplifier used in a wireless communication system, and a memory effect compensation method of the method. A memory effect compensating device of a predistorter compensating device, comprising: a power average calculating a power average of an error signal of a predistorter, and calculating a coefficient of a predetermined compensation polynomial for thermal memory effect compensation using the power average; And a gain determiner for determining a gain value of the error signal by multiplying the calculated coefficient by a power of the power average, and a compensator for generating a compensation signal for a thermal memory effect using the determined gain value. It is characterized by.

Memory effect, thermal memory effect, long time memory effect, compensation, predistorter, power amplifier

Description

Apparatus and method for compensating the thermal memory effect of a power amplifier in a wireless communication system {APPARATUS AND METHOD FOR COMPENSATING LONG TIM MEMORY EFFECT OF POWER AMPLIFIER IN A WIRELESS COMMUNICATION SYSTEM}

1 is a block diagram illustrating a transmitter configuration of a wireless communication system for compensating for nonlinear characteristics of a power amplifier using a predistorter in the form of a general lookup table.

2 is a block diagram showing the configuration of an adaptive controller included in a general predistorter.

3 is a block diagram illustrating a transmitter configuration of a wireless communication system equipped with a memory effect compensation device of a power amplifier according to an embodiment of the present invention.

4 is a block diagram showing the configuration of a memory effect compensation device according to an embodiment of the present invention;

The present invention relates to a device and a method for compensating for the nonlinear characteristics of a power amplifier in a wireless communication system, and in particular to a pre-distortion device for linearizing the nonlinear distortion characteristics of a high output power amplifier used in a wireless communication system and It's about how.

In general, a power amplifier for transmitting and receiving in a radio communication system using a radio frequency (RF) signal is classified into a low power low noise amplifier (LNA) and a high power amplifier (HPA). . The high power amplifier is mainly used in transmitters where efficiency is more important than noise. Therefore, high power amplifiers, which are widely used in wireless communication systems, operate close to nonlinear operating points for high efficiency.

In this case, the output of the power amplifier produces an intermodulation distortion (IMD) component that affects the spurious signal in other bands as well as in-band. In order to remove the spurless component, a feed forward method is mainly used. The feedforward method almost completely eliminates the spurless component, but the disadvantage is that the amplification efficiency is lowered and the control is required at the radio stage (RF stage).

In the high power amplifier of such a wireless communication system, nonlinear characteristics are generated in the process of obtaining high efficiency, and a so-called pre-distortion method is being studied to compensate for the nonlinear characteristics. The predistortion method is classified into an analog method and a digital method. Hereinafter, the digital predistortion method will be described. The digital predistortion method is nonlinear by applying a signal having the same magnitude and antiphase to the horn modulated distortion (IMD) component of the power amplifier at the digital stage connected to the input of the power amplifier to the input of the power amplifier. Linearize the signal of the characteristic power amplifier.

Herein, the non-linear characteristics include AM / AM (Amplitude Modulation to AM) characteristics in which an output signal is changed according to an input signal size of a power amplifier, and AM / PM (AM to Phase Modulation) in which a phase of an output signal is changed according to an input signal size. ) Characteristics.

Meanwhile, a typical algorithm for calculating the inverse nonlinear distortion characteristics of the power amplifier in the predistorter is a complex polynomial. The complex polynomial predistorter can shorten the time taken to remove nonlinearity of the power amplifier. That is, the convergence speed is fast when calculating the complex polynomial. When the memory effect is not taken into consideration, the transposition compensation polynomial of the P-order is represented by Equation 1 below.

Here, d (n) is a precompensation signal, and the part of {} multiplied by the baseband input signal X (n) can be regarded as a precompensation gain of X (n).

To date, most predistorters have been studied a lot for signals with single-tone or narrowband frequencies, so they only compensate for the memoryless nonlinear nature of the power amplifier (i.e., only the current input affects the current output). Most of the way was. However, the non-linear characteristics of a power amplifier that handles radio frequencies significantly change the AM / AM and AM / PM characteristics by affecting the output of the current power amplifier, as well as the current input signal.

This phenomenon is called memory effects, and the nonlinearity of the power amplifier varies depending on the frequency bandwidth of the input signal. Recently, research and development in consideration of the memory effect of the nonlinear amplifier have been actively conducted with the increase in the power output of the wireless communication system.

As a pre-compensation technique for compensating the memory effect, there is a technique using a look-up table (LUT). Since the lookup table stores the precompensation gains for the range of possible amplitude levels of the input signal, typically only a very small amount of computation is needed. However, in order to accurately remove the nonlinearity of the power amplifier having the memory effect, an adaptive algorithm must be applied to each of the elements of the lookup table, that is, the precompensation gains, so that it takes a long time to linearize the power amplifier. I.e., the convergence speed is slow.

Therefore, in order to overcome the drawbacks described above, a predistortion method is proposed to linearize a power amplifier using a predistorter and simultaneously take advantage of the fast convergence of the polynomial method and the small calculation amount of the lookup table method. That is, the parameter extraction of the predistorter is performed through a polynomial method, and the extracted predistorter parameter is changed into a lookup table. This results in a fast convergence using a polynomial and a small amount of computation with the predistorter in the form of a lookup table.

That is, FIG. 1 is a block diagram illustrating the configuration of a transmitter 100 of a wireless communication system that compensates for nonlinear characteristics of a power amplifier using a predistorter having a general lookup table.

The transmitter having the configuration of FIG. 1 is used in, for example, a wireless communication system using Radio Frequency (RF). First, referring to the forward path of the transmitter 100 of FIG. 1, a baseband signal X (n), which is a digital signal, is applied as an input signal, and the input signal is a nonlinear characteristic of the power amplifier (PA) 115. Digital Pre-Distorter (DPD) 101, Digital Quadrature Modulator (DQM) 103, and Digital / Analog outputting signals having the same magnitude and antiphase as the IMD component It is output as an analog signal via a converter (DAC) 105.

The output signal of the digital-to-analog converter 105 is an RF local oscillator (LO) 109 and a mixer via an intermediate frequency (IF) band pass filter (BPF) 107. After conversion to the desired radio frequency through a frequency up converter (111) consisting of (111), and through the radio frequency (RF) band pass filter (BPF) 113 to remove the out-of-band components, Amplified by the power amplifier 115 is transmitted to the wireless network through the antenna.

Next, referring to the feedback path of the transmitter 100 of FIG. 1, a portion of an output signal of the power amplifier 115 is extracted through a directional coupler 117 to form an RF local oscillator (LO). 109 and another mixer 119 are passed to a frequency down converter. The frequency down converter converts the amplified signal to an intermediate frequency (IF), and the signal converted to the intermediate frequency is a feedback band pass filter (FB BPF) 121, an analog to digital converter (ADC) ( The digital quadrature demodulator (DQDM) 125 is inputted through 123. The digital quadrature demodulator 125 demodulates the input digital signal into an in-phase signal component and a quadrature signal component, and then transmits the digital signal to the precompensator P1.

The predistorter P1 includes a digital predistorter (DPD) 101 for outputting a predistortion signal, a delay compensator (DL) 129 for delaying an input signal, and a delay estimator (DLL) ( 127 and an adaptive controller (ADPT) 131 which generally controls the predistortion operation. The adaptive controller 131 also periodically monitors the output signal from the predistorter 101. As a result, the adaptive controller 130 receives an output signal (amplified input signal) from the predistorter 101 and an output signal (amplified output signal) from the digital quadrature demodulator 125, respectively. The inputs are used to determine coefficients for calculating a precompensation polynomial such as < Equation 1 > and the determined polynomial coefficients are used to calculate precompensation gains for all possible magnitudes of the input signal. The calculated precompensation gains are stored in a lookup table (not shown) in the predistorter 101.

In the above configuration, when the parameter (precompensation gains) of the predistorter is changed into a lookup table form, the predistorter multiplies the input signal by referring to a specific lookup table value according to the size of the input signal. In this case, several clocks are required to determine the lookup table address according to the magnitude of the input signal, and several clocks are required to take data from the lookup table. In other words, an FPGA configuration that satisfies the predistortion polynomial equation requires processing with a clock approximately 10 times faster than the input sample rate of the predistorter.

Since the FPGA's maximum operating clock rate is currently about 200 MHz, when using the existing IIR structure, the input signal of the predistorter must have a sample rate of about 20 MHz or less, which means that the base station of a communication system using a high output power amplifier is used. Becomes an insufficient sample rate.

In addition, the memory effect is classified into an electric memory effect called a short time memory effect and a long time memory effect (or a thermal memory effect) by heat, and the short-term memory effect is generally compensated for. This is easy, but in the case of thermal memory effect compensation, the logic configuration of the predistortion device is complicated. Therefore, there is a need for a new structure of the predistortion device which can solve both the problem of insufficient sample rate and the thermal memory effect while simply configuring the logic of the predistortion device.

It is an object of the present invention to provide a predistortion apparatus and method capable of improving the sample rate while compensating for the memory effect of a power amplifier in a wireless communication system.

Another object of the present invention is to provide a predistortion apparatus and method capable of compensating for the thermal memory effect of a power amplifier in a wireless communication system.

It is another object of the present invention to provide a predistortion apparatus and method that can improve the sample rate while compensating for the thermal memory effect of a power amplifier with a simple logic configuration in a wireless communication system.

An apparatus of the present invention for achieving the above object is a memory effect compensation device of a predistortion device for compensating for nonlinear characteristics of a power amplifier in a wireless communication system, comprising: a power averaging device that calculates a power average for an error signal of a predistorter; A gain determiner for calculating a coefficient of a predetermined compensation polynomial for thermal memory effect compensation using the power average, and multiplying the calculated coefficient by a power of the power average to determine a gain value of the error signal; And a compensator for generating a compensation signal for the thermal memory effect by using the determined gain value.

According to an aspect of the present invention, there is provided a method of compensating a memory effect of a predistorter for compensating for a nonlinear characteristic of a power amplifier in a wireless communication system, the method comprising: calculating a power average of an error signal of a predistorter; Calculating a coefficient of a predetermined compensation polynomial for thermal memory effect compensation using the power average, multiplying the calculated coefficient by a power of the power average, and determining a gain value of the error signal; And generating a compensation signal for a thermal memory effect by using the determined gain value.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

In the following description, a process of generating a thermal memory effect in a general predistortion device will be briefly described to help understanding of the present invention related to thermal memory effect compensation.

FIG. 2 is a block diagram showing the configuration of an adaptive controller (ADPT) 131 included in a general predistortion apparatus. In FIG. 2, the same components as those shown in FIG. 1 are denoted by the same reference numerals. do. Referring to FIG. 2, the error of the conventional predistortion algorithm is based on the output of the predistorter 101 and the output of the quadrature demodulated power amplifier for the input signal X (n), respectively. It is defined as the difference 131a obtained by applying the same predistortion function (f (·)) 131b. Here, the predistortion algorithm refers to an algorithm for calculating precompensation gains provided to the predistorter 101, and the error of the predistortion algorithm is thermally compensated by the predistorter 101 of the existing configuration. Contains distortion signals due to memory effects. In FIG. 2, pdo_i and pdo_q mean an in-phase component and a quadrature component of the predistorter 101 output, respectively.

On the other hand, with respect to the thermal memory effect, the present applicant, the first application patent application No. 2004-0000148 discloses a precompensation device using a finite impulse response (FIR) filter that can improve the sample rate while compensating the memory effect of the power amplifier It is. However, in the case of the patent application, the electrical memory effect can be properly compensated, but since the predistorter is composed of a FIR filter having a few sample clock length, compensation of the thermal memory effect is relatively difficult. That is, if the time constant of the thermal memory effect is longer than the delay time of the FIR filter, the tap length of the FIR filter should be increased, but this greatly increases the size of the FPGA logic amount constituting the predistorter, There is a problem of increasing numerical instability due to accumulation of round-off errors. This thermal memory effect is more severe in narrowband signals than in wideband input signals, and must be considered especially for signals with frequency bands of tens of KHz, such as GSM.

Hereinafter, the predistortion apparatus and method of the present invention for performing thermal memory effect compensation will be described with reference to FIGS. 3 to 5.

3 is a block diagram showing the configuration of a transmitter 300 of a wireless communication system equipped with a memory effect compensation device of a power amplifier according to an embodiment of the present invention. Performs the same operation as the configuration of FIG.

The transmitter having the configuration of FIG. 3 is used in a wireless communication system using, for example, high frequency (RF). First, referring to the forward path of the transmitter 300 of FIG. 3, a baseband signal X (n), which is a digital signal, is applied as an input signal, and the input signal is a nonlinear type of a power amplifier (PA) 115. A digital predistorter (DPD) 301 for outputting a signal having the same size as that of the IMD component having a characteristic and having an antiphase, a memory effect compensator P3 for compensating thermal memory effects in particular according to the present invention; Desired radio frequency via a frequency up converter consisting of an RF local oscillator 309 and a mixer 311 via an orthogonal modulator (DQM) 103, a digital-to-analog converter (DAC) 105, and an IF BPF 307. The signal is converted into, and amplified by the power amplifier 315 via the RF BPF 313 and transmitted to the wireless network through the antenna.

Next, referring to the feedback path of the transmitter 300 of FIG. 3, the output signal of the power amplifier 315 is extracted from the RF local oscillator (LO) 309 by extracting a part of the signal through the directional coupler 317. Is input to the digital quadrature demodulator 325 via a frequency down converter composed of a different mixer 319, an FB BPF 321, and an analog / digital converter 123. The digital quadrature demodulator 125 demodulates the amplified output signal into an in-phase signal component and a quadrature signal component, and then transfers the amplified output signal to the precompensator P2.

The precompensator P2 of the present invention includes a digital predistorter (DPD) 301 for outputting a predistortion signal, a delay compensator (DL) 329 for delaying an input signal to compensate for a memory effect, and a delay estimator (DLL). 327 and an adaptive controller (ADPT) 331 for overall control of the precompensation operation, as well as a memory effect compensator connected to an output terminal of the digital predistorter 301 to compensate for thermal memory effects. P3) is further provided. The adaptive controller 331 receives the pre-amplified signal input from the predistorter 301 and the amplified signal from the power amplifier 315, respectively, and determines coefficients for calculating the precompensated polynomial, thereby determining all possible signals of the input signal. Compute the precompensation gains for the magnitudes. The calculated precompensation gains are stored in a lookup table (not shown) in the predistorter 301.

In addition, in the precompensator P2 of the present invention, the memory effect compensator P3 inputs the output of the predistorter 301 and the quadrature demodulated power amplifier 315 to the input signal X (n), respectively. The error signal of the predistortion algorithm obtained as follows is calculated as the average power of the output signal of the predistorter 301 to obtain a gain value for compensating for the thermal memory effect. The memory effect compensator P3 multiplies the obtained gain value by the input signal transmitted to the power amplifier 315 to compensate for the thermal memory effect.

In other words, the thermal memory effect is proportional to the average value of the output power of the power amplifier, which can be expressed as a function of the average value of the input power. Thus, the error signal of the predistortion algorithm can be expressed as a function of the average power of the output signal of the predistorter 301. Here, the average power is divided into an arithmetic average and an exponential average, but since the thermal memory effect is proportional to the average power having a weight over time, the average power of the output signal of the predistorter 301 may be calculated as the exponential average power. Can be. Therefore, in the present invention, the compensation of the thermal memory effect is performed by multiplying the gain value of the error signal as a function of the average power by the input signal delivered to the power amplifier 315.

FIG. 4 is a block diagram showing the configuration of the memory effect compensating apparatus according to the embodiment of the present invention, and the same reference numerals (symbols) will be given to the same components as those shown in FIG.

Referring to FIG. 4, the error signal of the predistortion algorithm has an output of the predistorter 301 and an output of a quadrature demodulated power amplifier for the input signal X (n), respectively, The difference 431a obtained by applying the same predistortion function f (·) Here, the predistortion algorithm refers to an algorithm for calculating precompensation gains provided to the predistorter 301, and an error signal of the predistortion algorithm is not compensated by the predistorter 301. It includes a distortion signal by. The error signal is transmitted to the memory effect compensator P3 of the present invention.

Hereinafter, the configuration and operation of the memory effect compensator P3 will be described in more detail.

Referring to FIG. 4, the memory effect compensator P3 includes a power averager 401 that calculates an exponential power average for the error signal, and a compensation polynomial for thermal memory effect compensation using the calculated power average. A gain determiner 403 for determining a gain value of the error signal by multiplying the calculated coefficient by the power of the exponential power average, and integrating the determined gain value with in-phase / orthogonality of the input signal; And a compensator 405 that multiplies with the phase component to produce a compensation signal for the thermal memory effect. In this case, the gain determiner 403 may be converted into a lookup table by applying a method of implementing the lookup table of the precompensation polynomial.

The operations performed by the power averager 401, the gain determiner 403, and the compensator 405 constituting the memory effect compensator P3 of FIG. 4 are represented by Equations 2 below. To <Equation 5>. The operation of the power averager 401 is represented by Equation 2, and the operation of the gain determiner 403 is represented by Equation 3 and Equation 4, and the compensator 405 Is expressed by Equation 5 below.

FIG. 5 is a circuit diagram showing the configuration of the compensator 405 shown in FIG. 4, which includes a plurality of multipliers 501, 507, 509, and 511 and adders 503, 505, 513, and 515. Since the operation of the compensator 405 is expressed as Equation 5, a detailed description thereof will be omitted.

The following algorithm is a programmatic representation of the compensation operation performed in the memory effect compensator P3.

값이 커질수록 이전 평균 전력의 영향을 많이 받는다. 이 경우 장시간 평균 전력을 구한다는 의미이다. 그리고 "pdo_i" 및 "pdo_q"는 전치 왜곡기의 출력 신호의 동위상 성분 및 직교 위상 성분을 의미한다. 그리고 상기 "tuiq"는 상기 보상 다항식의 계수 "itv" 및 "qtv"와 곱하여져 열적 기억 효과의 보상 신호 "sum_i"와 "sum_q"를 출력한다. 상기 보상 신호 "sum_i"와 "sum_q"는 각각 열적 기억 효과의 보상 신호중 동위상 성분과 직교 위상 성분을 나타내며, 전치 왜곡기의 출력 신호에 더해져서 전력 증폭기의 입력 신호를 형성한다. In describing the variables defined in Equations 1 to 4 and the program equation, "thrm_p" is the maximum order of the compensation polynomial for compensating the thermal memory effect, and "mean_o" is the average of the predistorter output signals. Means power. In Equation 1, "mean_o" means the current average power, the average power previously obtained of "mean_o",

The larger the value, the more influenced by the previous average power. In this case, it means to find the average power for a long time. And &quot; pdo_i &quot; and " pdo_q " mean in-phase components and quadrature components of the output signal of the predistorter. The "tuiq" is multiplied by the coefficients "itv" and "qtv" of the compensation polynomial to output the compensation signals "sum_i" and "sum_q" of the thermal memory effect. The compensation signals " sum_i " and " sum_q " respectively represent in-phase components and quadrature phase components among the compensation signals of the thermal memory effect, and are added to the output signal of the predistorter to form an input signal of the power amplifier.

The program equation can also be summarized as follows.

Here, the coefficients "itv" and "qtv" of the compensation polynomial are equal to the error signal (ter_i, ter_q) and controlled by the adaptive algorithm of the gain determiner 403.

If the control operation of the gain determiner 403 is expressed programmatically, " mu " is a factor for determining the speed of convergence.

As described above, according to the present invention, a thermal memory effect can be compensated for in the predistortion operation of the high output power amplifier for a wireless communication system.

In addition, the present invention can provide a compensation device capable of compensating for the thermal memory effect while simplifying the configuration of the predistorter. The present invention can also improve the linearization operation of a high output power amplifier using a narrowband signal as an input. have.

Claims (8)

A memory effect compensation device of a predistortion device for compensating for nonlinear characteristics of a power amplifier in a wireless communication system, A power averaging device that calculates a power average for the error signal of the predistorter, A gain determiner for calculating a coefficient of a predetermined compensation polynomial for thermal memory effect compensation using the power average, and multiplying the calculated coefficient by a power of the power average to determine a gain value of the error signal; And a compensator for generating a compensation signal for a thermal memory effect using the determined gain value. The method of claim 1, The average power is an exponential power average. The method of claim 2, And the compensator generates the compensation signal by multiplying the determined gain value by an in-phase component and a quadrature phase component of the input signal. The method of claim 1, The gain values are generated in a lookup table. A memory effect compensation method of a predistorter for compensating for nonlinear characteristics of a power amplifier in a wireless communication system, Calculating a power average for the error signal of the predistorter, Calculating coefficients of a predetermined compensation polynomial for thermal memory effect compensation using the power average; Determining a gain value of the error signal by multiplying the calculated coefficient by a power of the power average; And generating a compensation signal for a thermal memory effect by using the determined gain value. The method of claim 5, wherein The average power is an exponential power average. The method of claim 5, wherein And generating the compensation signal by generating the compensation signal by multiplying the determined gain value by an in-phase component and a quadrature phase component of the input signal. The method of claim 5, wherein Generating the gain values as a lookup table.

Images