JP5420887B2 - Distortion compensation device - Google Patents

Description

  The present invention relates to a distortion compensator used for a radio transmission apparatus, and more particularly, to a distortion compensator using a DPD (Digital PreDistortion) for compensating distortion generated in an amplifier circuit.

A digital signal processor (DSP) is used in a digital distortion compensation apparatus that feeds back a part of an output from an amplifier circuit, obtains a distortion compensation value from an input signal to the amplifier circuit and a feedback signal, and operates adaptively. The thing is proposed (for example, refer patent document 1).
JP 2001-267850 A

  In the conventional distortion compensator, since the division process is performed when calculating the inverse distortion characteristic, the circuit scale has been increased. Therefore, there is a possibility that the processing is delayed and the cost is increased.

  Accordingly, an object of the present invention is to provide a distortion compensation apparatus that reduces the circuit scale by eliminating the division process when calculating the inverse distortion characteristic.

In order to solve the above problem, the distortion compensating apparatus according to the present invention includes a distortion inverse characteristic calculation unit that exits Risan by distortion inverse characteristic in LMS (Least Mean Square) algorithm, distortion compensation value of each input signal power Is updated to a new distortion compensation value according to the inverse characteristic of the distortion calculated by the distortion inverse characteristic calculating unit .
By using the LMS algorithm, the distortion inverse characteristic can be calculated without performing division processing.

  According to the present invention, since the inverse distortion characteristic is calculated without performing division processing, the circuit scale can be reduced.

  Embodiments of the present invention will be described with reference to the accompanying drawings. The embodiment described below is an example of the configuration of the present invention, and the present invention is not limited to the following embodiment.

(Embodiment 1)
FIG. 1 is a schematic configuration diagram of a distortion compensation apparatus according to the present embodiment. The distortion compensation apparatus according to this embodiment includes a D / A 21, a Q-MOD 22, a HYB 23, an oscillator 24, and an amplifier circuit 11. The distortion compensation output signal d (t) is orthogonally modulated and transmitted. . Furthermore, an address calculation unit 16, a reference table (LUT) 15, and a distortion compensation unit 12 are provided to compensate for nonlinear distortion of the amplifier circuit 11. Further, the Q-DEM 25, the HYB 23, the oscillator 24, the A / D 26, the distortion inverse characteristic calculation unit 13, the distortion compensation value update unit 14, and the address calculation unit 17 are provided to perform adaptive distortion compensation.

The address calculation unit 16 calculates the power | x (t) | ² of the input signal x (t). This is the address a ′ of the LUT 15. The address calculation unit 17 calculates the power | y (t) | ² of the feedback signal y (t). This is an address a for calculating distortion inverse characteristics. The address calculation unit 16 and the address calculation unit 17 can have the same configuration.

The LUT 15 is a reference table in which a distortion compensation value w (n) is defined. Further, the distortion compensation value corresponding to the address a ′ calculated by the address calculation unit 16 is read out. The distortion compensation unit 12 is connected to the previous stage of the amplifier circuit 11 and is based on a reference table in which a distortion compensation value w _{a ′} (t) of an address a ′ corresponding to the input signal power | x (t) | ² is defined. A complex multiplication of the distortion compensation value w _{a ′} (t) is performed on the input signal x (t) input to the amplifier circuit 11. For example, the input signal x (t) is complex-multiplied by the distortion compensation value w _{a ′} (t) read from the LUT 15 to output the signal d (t).

  The signal d (t) is converted into an analog signal by the D / A 21, orthogonally modulated by the Q-MOD 22, HYB 23 and the oscillator 24, and then amplified by the amplifier circuit 11. As a result, it is possible to transmit a signal in which nonlinear distortion generated in the amplifier circuit 11 is compensated.

  A part of the output signal from the amplifier circuit 11 is demodulated as a feedback signal by the Q-DEM 25, the HYB 23, and the oscillator 24, converted into a digital signal by the A / D 26, and then sent to the distortion inverse characteristic calculation unit 13. Entered.

The distortion inverse characteristic calculation unit 13 calculates the distortion inverse characteristic r _b (M _b ) of the feedback signal y (t) from the amplifier circuit 11 with respect to the input signal x (t) using the LMS algorithm. The distortion compensation value update unit 14 converts the distortion compensation value w _b (n) to a new distortion compensation value w _b (n + 1) according to the inverse distortion characteristic r _b (M _b ) calculated by the distortion inverse characteristic calculation unit 13. Update to

  By using the LMS algorithm in the distortion inverse characteristic calculation, the division process and the averaging process in the distortion inverse characteristic calculation unit 13 become unnecessary, and the circuit scale of the distortion inverse characteristic calculation unit 13 can be reduced.

Hereinafter, details of the distortion inverse characteristic calculation unit 13, the distortion compensation value update unit 14, and the distortion compensation unit 12 will be described.
Equations 1 and 2 are equations representing distortion inverse characteristic calculation processing in the distortion inverse property calculator 13. The inverse distortion characteristic calculation unit 13 calculates the inverse distortion characteristic r _a (K _a ) by repeatedly calculating Expression 1 and Expression 2. Equations 1 and 2 are LMS algorithms. However, the point to be obtained is different for each level (address a). Calculating the address a from the feedback signal y (t), calculates the coefficient r _a for respective address a.

However, x is the input signal, y is the feedback signal, _{r a} is a coefficient, _{e a} is the error, mu ₁ is the step size, a is an integer satisfying 0 ≦ a <A at the address corresponding to the feedback signal y _{(k a)} numerical, a is the number of addresses, _{k a} is an integer satisfying 0 ≦ _{k a <K} a in the number of updates of the coefficient _{r a} corresponding to the address a. Here, the initial value of _{k a} is 0. e _a , x, y, and r _a are complex numbers, and * indicates a complex conjugate.

Equation 3 is an equation representing the distortion compensation value update processing in the distortion compensation value update unit 14. The distortion compensation value update unit 14 calculates Equation 3 for each address b, thereby updating the distortion compensation value w _b (n) in the reference table to a new distortion compensation value w _b (n + 1). After repeating the processing of Equation 1 and Equation 2 any number of times, w _b is updated for all addresses b by Equation 3.

However, _{w b} is the distortion compensation value (the value of the LUT), μ ₂ is the step size, n is the number of updates _{w b,} the value of r _{b (M} b) is at the time of update _{r a (k} a), b update Is an integer value satisfying 0 ≦ b <A. Here, A is the number of addresses. In _addition, w _{b, r} b is a complex number.

  The above-described Formula 1, Formula 2, and Formula 3 are a single update process of the distortion compensation value. The distortion compensation value is calculated by repeating these mathematical expressions.

Equation 4 is an equation representing distortion compensation processing in the distortion compensator 12. The distortion compensation value w _{a ′} (t) of the LUT 15 is read out by the address a ′, and the signal d (t) is obtained by complex multiplication with the input signal x (t).

(Embodiment 2)
FIG. 2 is a schematic configuration diagram of the distortion compensation apparatus according to the present embodiment. In the distortion compensation apparatus according to the present embodiment, the address a of the distortion inverse characteristic calculation unit 13 is the power | x (t) | ^{2 of} the input signal x (t). In this case, a = a ′. By repeating the processing of Equation 1, Equation 2, and Equation 3, the distortion inverse characteristic becomes linear, and the difference between | x (t) | ² and | y (t) | ² is eliminated. Further, by omitting the address calculation unit 17 in the first embodiment, the circuit scale can be further reduced.

  FIG. 3 shows an example of the distortion compensation value stored in the LUT. FIG. 3 is not limited to the AM-AM distortion compensation value, the AM-PM distortion compensation value, or any of a real value and an imaginary value. When frequency distortion is included in the feedback signal (including the PA memory effect), there is no distortion in the address area with low power, or the error of the distortion compensation value increases despite the small distortion. Although the error of all addresses is large, a small address area is particularly large. All addresses are unstable, but a small address area is particularly unstable. This is improved by setting the small address to the same value as the adjacent address as shown in FIG. Since a small address area has a small distortion (distortion compensation value), even if the value is the same, the influence is small.

  The present invention can be used for a wireless transmission device.

1 is a schematic configuration diagram of a distortion compensation apparatus according to Embodiment 1. FIG. FIG. 3 is a schematic configuration diagram of a distortion compensation apparatus according to a second embodiment. An example of the distortion compensation value stored in the LUT is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Amplifier circuit 12 Distortion compensation part 13 Distortion reverse characteristic calculation part 14 Distortion compensation value update part 15 LUT
16 Address Calculation Unit 17 Address Calculation Unit 21 D / A
22 Q-MOD
23 HYB
24 oscillator 25 Q-DEM
26 A / D

Claims (1)

In a distortion compensation device that compensates for nonlinear distortion of an amplifier circuit,
A distortion compensation unit that is connected to the previous stage of the amplifier circuit and that performs complex multiplication of the distortion compensation value on the input signal based on a reference table in which a distortion compensation value is defined in accordance with input signal power input to the amplifier circuit; ,
A distortion inverse characteristic calculation unit that calculates an inverse characteristic of distortion of the feedback signal from the amplification circuit with respect to the input signal using an LMS (Least Mean Square) algorithm;
A distortion compensation value updating unit that updates the distortion compensation value of each input signal power defined in the reference table to a new distortion compensation value according to the distortion inverse characteristic calculated by the distortion inverse characteristic calculation unit ;
Equipped with a,
The distortion inverse characteristic calculation unit calculates the distortion inverse characteristic r _b (M _b ) obtained by updating the coefficient r _a (k _a +1) calculated from Equations 1 and 2 an arbitrary number of times ,
The distortion compensation value update unit calculates Equation 3 for each address b, thereby updating the distortion compensation value w _b (n) in the reference table to the new distortion compensation value w _b (n + 1). A distortion compensation device.

However, x is the input signal, y is the feedback signal, _{r a} is a coefficient, _{e a} is the error, μ _1, μ ₂ is the step size, a is the address of the reference table corresponding to the power of the feedback signal y _{(k a)} integer satisfying 0 ≦ a <a in, a is the number of addresses of the lookup table, _{k a} is the number of updates of the coefficients _{r a, w b} is distortion compensation values, n represents the number of updates of _{w b,} r _{b (M b} ) Is the value of r _a (k _a ) at the time of update , and b is an integer value satisfying 0 ≦ b <A with the address of the reference table of the coefficient to be updated.

Images