DISTORTION COMPENSATION DEVICE FOR USE IN HIGH-FREQUENCY POWER AMPLIFIER

Abstract

A non-linear distortion signal contained in output signal components from a high-frequency power amplifier is detected, while a non-linear distortion signal is generated from a high-frequency signal to be amplified. Correlation between this non-linear distortion signal and the detected non-linear distortion signal is determined, and a control variable is calculated based on the signal thus obtained. The obtained control variable is multiplied by the generated non-linear distortion signal to generate a distortion compensation signal, which is added to the signal to be amplified. This provides a distortion compensation device capable of automatically maintaining optimum distortion compensation characteristics and rapidly achieving desirable distortion compensation.

Description

TECHNICAL FIELD

The present invention relates to a device for compensating non-linear distortion occurring during power amplification of high-frequency signals.

BACKGROUND ART

On amplifying high-frequency signals, linearity is required between input and output signals when an envelope of the input signal is varied with time. This is because absence of linearity will cause occurrence of so-called non-linear distortion (hereafter, referred to as “distortion”). Such non-linear distortion brings about radiating unnecessary power outside a desired band for original signals and results in interference with an adjacent channel.

Actual power amplifiers however have not a little non-linearity, and it is impossible to obtain perfect linearity in them. In general, power efficiency of the power amplifier, namely, a rate of power of high-frequency signals to the power of the power source tends to be reduced when placing importance on the linearity. In contrast, the non-linearity will be increased if it is tried to increase the power efficiency. Proposal has been made about techniques of compensating the non-linear distortion caused to occur in the amplifiers, so as to alleviate a trade-off between these conflicting characteristics of the amplifiers.

DISCLOSURE OF THE INVENTION

Problems to be Solved by the Invention

The related techniques to compensate the non-linear distortion are largely classified into a negative feedback technique, a feed-forward technique, and a predistorter technique. The negative feedback technique has an advantage that a high speed operation can be accomplished, but has a disadvantage that such an operation becomes instable as the distortion compensation is increased. The feed-forward technique can realize a stable operation, but has a shortcoming that the overall power efficiency is low due to the necessity of a secondary amplifier.

The predistorter technique is advantageous in these respects and therefore, has come under attention. However, this technique still has a problem that the characteristics of the power amplifier are varied by a fluctuation in the central frequency of the signals, voltage variation, temperature change, or aging. Therefore, it is important to determine how optimum distortion compensation characteristics can be automatically maintained. It is also a problem to be solved in association with the conventional predistorter technique that a longer period of time is required to achieve optimum distortion compensation characteristics compared to the negative feedback technique.

The present invention has been made in view of the problems as mentioned in the above, and seeks to solve the problem and to provide a predistorter-type distortion compensation device which is capable of automatically maintaining optimum distortion compensation characteristics, and rapidly achieving distortion compensation.

Means for Solving the Problems

In order to solve the problems above, one embodiment of the present invention is characterized by including:

distortion detection means for detecting a distortion detection signal in an amplifier output signal from the power amplifier;

distortion signal generation means for receiving a signal to be amplified, to generate an n-th (n is an integer of two or more) non-linear distortion signal;

distortion compensation signal generation means for generating a distortion compensation signal by multiplying the non-linear distortion signal by a control variable;

a correlation circuit for determining correlation between the n-th non-linear distortion signal and the distortion compensation signal; and

phase compensation means, arranged upstream of the distortion detection means, for compensating phase rotation of the signal,

the control variable being determined by an output signal of the correlation circuit.

According to another aspect of the present invention, there is obtained a method of compensating non-linear distortion generated in a power amplifier, comprising the steps of:

detecting a distortion detection signal in an amplifier output signal from the power amplifier;

generating an n-th (n is an integer of two or more) non-linear distortion signal; and

generating a distortion compensation signal by using the distortion detection signal and the n-th non-linear distortion signal.

ADVANTAGEOUS EFFECTS OF THE INVENTION

According to the present invention, even if the amplifier characteristics are varied, optimum distortion compensation characteristics can be automatically achieved, and it is possible to shorten the period of time required for obtaining the optimum distortion compensation characteristics.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a circuit block diagram of a distortion compensation device according to a first embodiment of the present invention;

FIG. 2 is a block diagram showing an example of a distortion signal generation circuit;

FIG. 3 is a block diagram showing a first example of a distortion detection circuit;

FIG. 4 is a block diagram showing a second example of a distortion detection circuit;

FIG. 5 is an explanatory diagram illustrating relation between filter characteristics of the distortion detection circuit shown in FIG. 4 and power spectrum of signals including distortion signals;

FIG. 6 is a block diagram showing a third example of a distortion detection circuit;

FIG. 7 is an explanatory diagram showing an example of distortion compensation characteristics;

FIG. 8 is a circuit block diagram showing a distortion compensation device according to a second embodiment of the present invention; and

FIG. 9 is a circuit block diagram of a distortion compensation device according to a third embodiment of the present invention.

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a circuit block diagram for explaining a distortion compensation device according to a first embodiment of the present invention. Overall operation of the device will be described using FIG. 1. A high-frequency signal to be amplified is input through an input terminal 11 to a distortion compensation signal generation circuit 12 and a distortion signal generation circuit 16. The distortion compensation signal generation circuit 12 supplies a power amplifier 13 with a signal to be amplified, as an amplifier input signal, and generates a distortion compensation signal for canceling the distortion component generated by a power amplifier 13 within the distortion compensation signal generation circuit 12 in a manner to be described later.

An output signal of the power amplifier 13 is supplied to an output terminal 14 and is partly given to a distortion detection circuit 15 through a phase compensation circuit 19, so as to extract a distortion signal component (may be called a distortion detection signal). Alternatively, the high-frequency signal to be amplified input through the input terminal 11 may be directly given to the distortion detection circuit 15, as indicated by the broken line in FIG. 1.

The distortion signal generation circuit 16 receives the high-frequency input signal and generates a high-order (n-th) distortion signal. “n” may take a value of 3, 5, 7, or the like. The n-th distortion signal of the distortion signal generation circuit 16 is partly input to the distortion compensation signal generation circuit 12 and multiplied by an n-th control variable generated by a control variable generation circuit 18 so as to produce the distortion compensation signal. Another part of the n-th distortion signal is input to a correlation circuit 17 where correlation calculation is conducted between the n-th distortion signal and an output signal from the distortion detection circuit 15. An output signal from the correlation circuit 17 is input to the control variable generation circuit 18, so as to produce a control variable.

The phase compensation circuit 19 compensates phase rotation that might occur in a signal from the input terminal 11 to an output terminal of the distortion detection circuit 15. Operation of the phase compensation circuit 19 will be described later in detail.

FIG. 2 is a circuit diagram showing an example of a distortion signal generation circuit according to an embodiment of the present invention. Description will be made using mathematical expressions. An arbitrary high-frequency signal can be represented by the following equation.

s(t)=A(t)cos(ω_ct+φ(t))  (1)

where A(t) denotes an amplitude, ω_c denotes a carrier wave frequency, and φ(t) denotes a phase. Information in the signal is reflected in A(t) and φ(t). The equation above can be rewritten as follows using a trigonometric function formula.

$\begin{array}{cc}\begin{array}{c}s\left(t\right)=A\left(t\right)\cos \varphi \left(t\right)\cos {\omega }_ct-A\left(t\right)\sin \varphi \left(t\right)\cos {\omega }_ct\\ =x\left(t\right)\cos {\omega }_ct-y\left(t\right)\sin {\omega }_ct\end{array}& \left(2\right)\\ \mathrm{where}x\left(t\right)=A\left(t\right)\cos \varphi \left(t\right),\mathrm{and}y\left(t\right)=A\left(t\right)\sin \varphi \left(t\right).& \end{array}$

The equation (1) can be represented as follows, using complex numbers.

$\begin{array}{cc}\begin{array}{c}s\left(t\right)=\mathrm{Re}\left[A\left(t\right){}^{\mathrm{j\varphi }\left(t\right)}{}^{{\mathrm{j\omega }}_ct}\right]\\ =\mathrm{Re}\left[z\left(t\right){}^{{\mathrm{j\omega }}_ct}\right]\end{array}& \left(3\right)\\ \mathrm{where}\text{:}& \\ z\left(t\right)=x\left(t\right)+jy\left(t\right)& \left(4\right)\end{array}$

When a component of the n-th distortion signal of s(t) around the carrier wave frequency ω_c is represented as s_dⁿ(t), it can be represented as follows.

$\begin{array}{cc}\begin{array}{c}{s}_d^n\left(t\right)=A^n\left(t\right)\cos \left({\omega }_ct+\varphi \left(t\right)\right)\\ =\mathrm{Re}\lfloor A^n\left(t\right){}^{\mathrm{j\varphi }\left(t\right)}{}^{{\mathrm{j\omega }}_ct}\rfloor \end{array}& \left(5\right)\end{array}$

The value of the carrier wave frequency ω_c is irrespective of the present invention and, is therefore neglected, and the (ω_c=0) signal is represented using complex numbers. A complex signal z(t) contains all information in the signal. If an actual high-frequency signal is to be represented, the equation (3) may be used.

A complex representation z_dⁿ(t)=|z(t)|^n-1z(t) is obtained for the distortion signal s_dⁿ(t) obtained by the equation (5). In this complex representation, it is given that Aⁿ=|z(t)|ⁿ (the symbol |•| denotes an absolute value).

The signals represented as z and z_dⁿ in FIG. 2 are complex representations from which the time factor t is omitted for simplification of description. It should be noted that the output from the distortion signal generation circuit 16 in FIG. 1 consists of a plurality of these distortion signals. The control variable generation circuit 18 also generates the same number of control variable α_n as the number of the distortion signals. An example of operation of the distortion compensation signal generation circuit 12 can be represented as follows by using mathematical expression. The output signal z′ of this circuit 12 is represented as:

$\begin{array}{cc}\begin{array}{c}{z}^{\prime }=z+{\alpha }_3z_d^3+{\alpha }_5z_d^5+\dots +{\alpha }_Nz_d^N\\ =z+{\alpha }_3{z}^2z+{\alpha }_5{z}^4z+\dots +{\alpha }_N{z}^{N-1}z\end{array}& \left(6\right)\end{array}$

The output signal z′ is obtained by adding the signal z to be amplified and the n-th distortion compensation signal, as shown in FIG. 1. The principle thereof is that distortion in the amplifier output is canceled by inputting, as an internal signal, a signal having the same amplitude as the distortion component and opposite phase to that of the distortion component generated in the amplifier and by adding the internal signal to the signal to be amplified. The n-th distortion compensation signal is obtained by multiplication of the n-th distortion signal z_dⁿ and each control variable α_n.

A first example of the distortion detection circuit is shown in FIG. 3. This circuit finds a difference between the signal z to be amplified and a signal z″ obtained from an output signal of the amplifier 13.

A second example of the distortion detection circuit is shown in FIG. 4. This circuit extracts a distortion signal component by allowing the signal obtained from an output signal of the amplifier to pass through a filter. FIG. 5 is a conceptual diagram of a spectrum for explaining operation of this circuit. The solid line in FIG. 5 represents the spectrum of the signal to be amplified, and the broken line represents a spectrum of a signal generated due to distortion. The amplitude characteristics |H(ω)| of the filter shown in FIG. 4 is ideally such that only the distortion signal component shown in FIG. 5 is allowed to pass through.

FIG. 6 shows a third example of the distortion detection circuit. In this circuit, a fast Fourier transform (FFT) is performed prior to extracting only the frequency component due to distortion by a distortion component detection circuit. After that, an inverse fast Fourier transform (IFFT) may be performed to return the signal to a signal of time region. Alternatively, a correlation may be obtained while keeping the signal in frequency region.

As shown in FIG. 1, the correlation circuit 17 according to the present invention calculates a correlation value between an n-th distortion signal z_dⁿ generated by the distortion signal generation circuit 16 and a distortion signal e output from the distortion detection circuit 15, and the correlation value is denoted by zⁿ_d* (the symbol * indicates a complex conjugate). The control variable generation circuit 18 generates a control variable α_n based on the correlation value. Any method can be used for this purpose as long as the distortion signal e is approximated to zero. For example, when the distortion compensation device is embodied as a discrete time system, a method of least mean squares (LMS) as follows can be used.

α_n[m+1]=α_n[m]+μzⁿ_d*[m]e[m]  (7)

where m represents a discrete time (m=0, 1, 2, . . . ), and μ is a small constant. The control variable α_n starts from an arbitrary initial value (usually zero) and automatically converges over time to such a value that an average value of |e[m]|² becomes a minimum value. This can be expressed mathematically that the distortion signal is relatively small. However, description of this will be omitted here since the description will not be concise and will become lengthy. Instead, the validity of the operation is proven by a computer simulation experiment. FIG. 7 shows the distortion compensation characteristics obtained by the control variable generation algorithm described above, in the form of power spectrum. It can be seen that the out-of-band power component is decreased by the reduction of distortion.

FIG. 8 shows a second embodiment of the distortion compensation device according to the present invention. This second embodiment differs from the first one shown in FIG. 1 in that a quadrature modulation circuit 811, a local oscillation circuit 812, and a quadrature detection circuit 813 are additionally provided, and a phase compensation circuit 814 is altered. The remaining circuit configuration is the same as the first embodiment, and the numerical references 1 to 8 are changed to numerical reference 81 to 88, respectively. The quadrature modulation circuit 812 performs calculation expressed by the equation (2) when a signal to be transmitted is represented by an in-phase component x(t) and a quadrature component y(t). The local oscillation circuit 812 is a circuit generating sinusoidal waves (cos ω_ct and sin ω_ct) of the carrier wave frequency ω_c. Further, the quadrature detection circuit 812 functions to extract in-phase and quadrature components x(t) and y(t) from the high-frequency signal s(t) obtained by the equation (2). Since particulars of operation of the local oscillation circuit are well known to those skilled in the art, the description thereof will be omitted here. The phase compensation circuit 814 is a circuit to rotate the phase of the input signal based on the phase and quadrature component signals. Operation of the phase compensation circuit 814 can be represented by mathematical expressions as follows.

x′″(t)=x″(t)cos Δφ−y″(t)sin Δφ  (8a)

y′″(t)=x″(t)sin Δφ+y″(t)cos Δφ  (8b)

where Δφ denotes an amount of phase rotation. The value of Δφ is set so as to cancel the phase rotation that is caused when a signal z=x+jy input to the input terminal 81 passes through the quadrature modulation circuit 811, the amplification circuit 83, and the quadrature detection circuit 813. More specifically, this can be done by detecting the phase difference between the signals z=x+jy and z″=x″+jy″. When a digital signal processing is to be carried out, an iteration method as follows can be used.

$\begin{array}{cc}\begin{array}{c}\mathrm{\Delta \phi }\left[m+1\right]=\mathrm{\Delta \phi }\left[m\right]+\mathrm{\mu Im}\left\{z\left[m\right]z^{\mathrm{\prime \prime \prime }*}\left[m\right]\right\}\\ =\mathrm{\Delta \phi }\left[m\right]+\mu \left(-x\left[m\right]y^{\mathrm{\prime \prime \prime }}\left[m\right]+y\left[m\right]x^{\mathrm{\prime \prime \prime }}\left[m\right]\right)\end{array}& \left(9\right)\end{array}$

where μ is a small constant, x[m] and y[m] respectively denote an in-phase and quadrature component of a discrete time signal of the input signal z, and x′″[m] and y′″[m] are outputs from the phase compensation circuit 814.

Operation of the second embodiment of the present invention is the same as the operation of the first embodiment when the in-phase and quadrature components in the complex representation z(t)=x(t)+jy(t) of the signal are replaced by x(t) and y(t) in the first embodiment. Therefore, further description will be omitted here. In the embodiment of the present invention, the portion enclosed by the broken line is preferably embodied by so-called digital signal processing. In this case, although not shown in FIG. 8, a digital-analogue modulation circuit is used for the part of the outputs x′ and y′ from the broken line portion, and an analogue-digital conversion circuit is used for the part of the inputs x″ and y″ to the broken line portion. This embodiment has advantages that the signal processing can be performed in the base band, and the use of the digital signal processing is facilitated.

A third embodiment of the distortion compensation device according to the present invention is shown in FIG. 9. This embodiment is the same as the second one except that the numerical references 81 to 88 and 811 to 814 are changed to numerical references 91 to 98 and 911 to 914, and a negative feedback circuit is added. The negative feedback circuit is formed by multiplying a part of the output of the phase compensation circuit 914 by a constant (β) of a coefficient multiplication circuit 915, and then subtracting the product from a signal z=x+jy input through the input terminal 91. The subtracted signal is input to the distortion compensation signal generation circuit 92 instead of the high-frequency signal input through the input terminal 91. The constant β is determined in view of a degree of distortion compensation and stability of the circuit operation. A low-pass filter (LPF) 916 is inserted between the distortion compensation signal generation circuit 92 and the quadrature modulation circuit 911 to ensure stability of the negative feedback circuit. The remaining circuit configuration of this embodiment is the same as that of the second embodiment, and therefore further description will be omitted. In this embodiment, it is essential to input a negative feedback signal to the input of the distortion compensation signal generation circuit 92. This embodiment has an advantageous effect that the problem intrinsic to the predistorter-type distortion compensation method that it takes a long time to converge to optimum distortion compensation characteristics can be improved by way of the negative feedback method. In another respect, it can be said that the disadvantage of the negative feedback method that operation becomes instable when trying to conduct large distortion compensation is solved by causing the predistorter-type distortion compensation to share the amount of distortion compensation.

Claims

1. A distortion compensation device for compensating non-linear distortion generated in a power amplifier, the distortion compensation device comprising: a distortion detection means for detecting a distortion detection signal in an amplifier output signal from the power amplifier;a distortion signal generation means for receiving a signal to be amplified, to generate an n-th (n is an integer of two or more) non-linear distortion signal;a distortion compensation signal generation means for generating a distortion compensation signal by multiplying the non-linear distortion signal by a control variable;a correlation circuit for determining correlation between the n-th non-linear distortion signal and the distortion compensation signal;phase compensation means, arranged upstream of the distortion detection means, for compensating phase rotation of the signal, andthe control variable being determined by an output signal of the correlation circuit.

2. The distortion compensation device according to claim 1, wherein the distortion detection means comprises: error detection means for detecting an error between the amplifier output signal and the signal to be amplified, to generate the distortion detection signal.

3. The distortion compensation device according to claim 1, wherein the distortion detection means comprises: a filter for allowing a signal obtained from the amplifier output signal to pass through to detect a frequency component generated due to distortion and to generate the distortion detection signal.

4. The distortion compensation device according to claim 1, wherein the distortion detection means comprises: means for carrying out Fourier transform of a signal obtained from the amplifier output signal; andmeans for detecting a frequency component generated due to distortion after Fourier transform.

5. The distortion compensation device according to claim 1, further comprising: a quadrature modulation circuit arranged before the power amplifier; and a quadrature detection circuit receiving a signal obtained from the amplifier output signal, wherein a base band signal of in-phase and quadrature components is used as the signal to be amplified to perform phase compensation, distortion detection, correlation calculation, generation of a distortion signal, generation of a control signal, and generation of a distortion compensation signal.

6. The distortion compensation device according to claim 5, further comprising: a negative feedback circuit arranged between an input terminal of the distortion compensation device and the distortion compensation signal generation means.

7. A method of compensating non-linear distortion generated in a power amplifier, comprising the steps of: detecting a distortion detection signal in an amplifier output signal from the power amplifier;generating an n-th (n is an integer of two or more) non-linear distortion signal; andgenerating a distortion compensation signal by using the distortion detection signal and the n-th non-linear distortion signal.

8. The method according to claim 7, wherein the step of generating the distortion compensation signal comprises the steps of: multiplying the non-linear distortion signal by a control variable;determining correlation between the n-th non-linear distortion signal and the distortion compensation signal; anddetermining the control variable with reference to the correlation.