RU2761856C1 - Method for increasing the linearity of high-frequency power amplifiers and apparatus for implementation thereof - Google Patents

Abstract

FIELD: radio engineering.

SUBSTANCE: invention relates to the field of radio engineering and can be applied in radio transmitting apparatuses to create power amplifiers with a low level of out-band emissions used in radio communication and television and radio broadcasting. To realise the technical result, the distortion signal is isolated by subtracting the input signal from the output signal, and the functional dependence of these distortions on the input signal is determined. Using this functional dependence in the form of a separate pre-distortion function, a compensating signal is synthesised, admixed to the input of the power amplifier. A look-up table (LUT) can be used as pre-distortion functions. The apparatus comprises a control circuit (3) for mutual alignment of the amplitudes and phases of the input and output signals, an adder (4) for isolating the distortion signal, a unit for calculating the parameters of the pre-distortion function (5), a switch (6), a pre-distortion function unit (7), at the output whereof a compensating signal is synthesised, supplied to the input of the power amplifier (2) after phase alignment (in unit 8), thereby compensating for distortions thereof. For power amplifiers with the memory effect, the parameters of the pre-distortion function are calculated over the entire memory interval.

EFFECT: reduction in the level of residual distortion at the output of the power amplifier.

6 cl, 3 dwg

Description

Technology area

The invention relates to the field of radio engineering, in particular to radio transmitting devices for creating power amplifiers with a low level of out-of-band emissions used in radio communication and television and radio broadcasting.

State of the art

Heavy use of the radio frequency spectrum exacerbates the EMC problem of radio equipment operating in adjacent frequency bands. The main source of interference in such conditions is out-of-band emissions, which occur mainly in the output high-frequency power amplifiers (PAs). PAs operating in class A mode have a reduced level of out-of-band emissions, but they have low efficiency and are not economically effective. Energy efficient PAs with high efficiency operate in nonlinear modes and create unacceptable levels of out-of-band emissions.

Known methods and devices for linearizing such amplifiers or compensating out-of-band emissions. Patent RU 2731135 "Radio transmission device with digital nonlinearity correction" describes a device for minimizing the amplitude and phase distortions of the transmitted signal under conditions of rapidly changing and a priori undefined nonlinearity of the transmission characteristic of the radio path. The device contains a digital nonlinearity corrector using a weighted sum of E-polynomials, and a combinational component meter that calculates the amplitude and phase of distortions using the Fourier transform. The obtained distortion amplitudes are normalized to the input signal, and the phases are rotated by 180 °, after which the distortions are fed to a digital nonlinearity corrector, in which, using E-polynomials, a transfer characteristic is formed that compensates for the distortions introduced by the PA. The formation of the transfer characteristic of the digital corrector is carried out on a test (two-tone) signal in the absence of an information signal. The disadvantages of the presented device are the need for measuring and preliminary description of the transfer characteristic of the digital nonlinearity corrector using E-polynomials, as well as the need for periodic calibration of the device with a test signal with interruption of the information signal transmission.

In the materials of the 2009 IEEE International Symposium on Radio-Frequency Integration Technology, a method for linearizing the PA using the method of direct communication (Power Amplifier Linearization Using Feedforward Technique for Wide Band Communication System) is described. In this method, PA distortion compensation is carried out in two stages. At the first stage, after aligning the amplitudes, phases and delays of the input and output signals of the PA, the input signal is subtracted from the output signal, as a result of which signal distortions are allocated at the output of the subtractor. At the second stage, the selected distortions are subtracted from the PA output signal, thereby excluding them from the PA output signal. For the successful implementation of this method at each stage, it is necessary to accurately align the amplitudes, phases and delays of the interacting signals. Devices that implement this method contain at least two servo control loops that ensure the alignment of signal parameters. In this method, distortion compensation occurs at the output of the PA, therefore, before subtraction, the distortions must be amplified to the required level by a linear PA, which leads to a decrease in the efficiency of the device as a whole. In addition, the need to use at least two circuits reduces the stability of the device.

The closest analogue of the present invention is the method and device described in US patent US8774314 B2 TRANSMITTER ARCHITECTURES. In this method, as in the previous one, PA distortion compensation is carried out in two stages. At the first stage, after aligning the amplitudes, phases and delays of the input and output signals of the PA, the input signal is subtracted from the output. As a result, signal distortions are highlighted at the output of the subtractor. Unlike the previous analogue, at the second stage, the selected distortions are added to the input signal, while the amplitude and phase of the distortions are set in such a way as to compensate for the distortions at the PA output.

The device that implements this method, at least contains the first adder and PA connected in series. The adder input is the device input, and the PA output is the device output. The device also contains a control loop for aligning the amplitudes and phases of the input and output signals, and an interface unit for correcting the amplitude and phase of distortions introduced to the input of the PA through the subtractive input of the first adder. In this case, one input of the control loop is connected to the output of the PA, and the other to the input of the device. The output of the control loop is connected to the subtractive input of the second adder. Another input of the second adder is connected to the output of the PA. And the output of the second adder is connected to the input of the interface unit, the output of which is connected to the subtractive input of the first adder. In essence, the operation of the device is carried out in the same way as in the analogue described above, but the input of the compensating signal is carried out at the input of the PA.

A circuit consisting of a series-connected first adder, a PA, a second adder and an interface unit, the output of which is connected to the first adder, form a closed control loop, thus there are two interconnected control loops in the device, which can lead to device instability. In addition, the compensating signal is formed from the output signal of the PA, i.e. with a delay in relation to the input signal, which reduces the efficiency of the device.

Disclosure of the essence of the invention

In accordance with the proposed method, the increase in the linearity of the PA is carried out in two stages. At the first stage, after aligning the amplitudes, phases and delays of the input and output signals of the PA, the input signal is subtracted from the output. As a result, distortion is highlighted. By observing the input signal and distortion, the functional dependence of the distortion signal on the input signal is determined. After the functional dependence is determined, it is isolated as a separate function. Then, using the envelope of the current input signal as an argument to this function, this function synthesizes a compensating signal that is a model of the distortion in the power amplifier. In a second step, the compensating signal is subtracted from the input signal. Because the formation of a compensating signal, and the determination of the functional dependence is carried out independently, then the control loop turns out to be open, which ensures the stability of the device as a whole. In addition, in the proposed method, the formation of the compensating signal is based on the current input signal, i.e. without delays in the control loop, which improves distortion compensation.

Analytically, the process can be described as follows.

The signal at the output of the PA has the form:

, (1)

, (one)

where

y (t) - output signal;

t is time;

K ₀ - constant amplification factor of the PA;

K (x (t-τ)) - nonlinear part of the PA output signal;

τ is the signal delay in the PA;

x (t) - PA input signal.

Moreover, in practice, the distortion power of the PA is always significantly less than the signal power, i.e. the condition is satisfied

, (2)

, (2)

where the bar above is the mean. ¬¬

Expression (1) can be represented as:

, (3)

, (3)

where

y ₀ (t) - linear part of the PA output signal;

y _n (t) - distortion of the PA output signal,

and correspondingly

, (4)

, (4)

. (5)

... (5)

After mutual equalization of the amplitudes, phases and delays of the input and output signals of the PA and subtracting the input signal from the output, we get:

. (6)

... (6)

As a result, distortions are highlighted with an accuracy of a constant coefficient 1 / K ₀ .

There is a functional dependence of distortion u _n (t) on the input signal x _in (t)

, (7)

, (7)

where F (x _in (t)) - functional dependence;

x _in - input signal of the device.

The functional dependence, up to a constant coefficient K _0, corresponds to the nonlinear part of the PA output signal in expression (5).

The functional dependence F (x _in (t)) after approximation in the form of a separate function is used to generate a compensating signal:

. (8)

... (eight)

The compensation signal is subtracted from the input signal:

. (9)

... (9)

Hence, taking into account (3) and (2), we get the signal at the output of the PA:

. (10)

... (10)

Subtracted in (10) characterizes the residual distortions at the output of the PA, caused by mixing the compensating signal at its input, and is the value:

. (11)

... (eleven)

Taking into account condition (2), it can be seen that the residual distortions decrease according to the quadratic dependence with respect to the initial value. For example, if the original PA had a distortion level of minus 30 dB, then when using the proposed method, the distortion level can be minus 60 dB.

The functional dependence of distortions on the input signal (7) for a PA that does not have a memory effect is approximated as the product of the input signal by a parameterized function of the amplitude of the input signal

, (12)

, (12)

where Λ is the vector of parameters of the function f.

To compensate for distortions of a PA with a memory effect, functional dependence (7) is approximated on the PA memory interval, which can be written in the following form:

(13)

(thirteen)

where M is the memory depth of the PA according to the input signal;

K is the depth of the PA memory by the amplitude of the input signal.

The function f (| x _in (n) |, Λ) can be represented in various forms, for example, (but not limited to) lookup table, interpolation polynomials, etc.

The device for implementing this method (see figure 1) contains at least the first adder (1) and the PA (2) connected in series. The adder input (1) is the device input, and the PA output (2) is the device output. The device also contains a control loop (3) for mutual equalization of amplitudes, phases and delays of the input and output signals. In this case, one input of the control loop (3) is connected to the output of the device, and the other - to the input of the device. The device contains a second adder (4), the subtractive input of which is connected to the first output of the control loop (3), and the other input of the adder (4) is connected to the second output of the control loop (3). In this case, respectively, at the first and second outputs of the control loop (3), the input and output PA signals are formed, balanced in amplitude, phase and delay.

The device differs in that it additionally includes a block for calculating the parameters of the predistortion function (5), a key (6) (optional element), a block of the predistortion function (7) and a phase equalization block (8). In this case, the input of the phase alignment block (8) is connected to the output of the predistortion function block (7), and its output is connected to the second input of the first adder (1). The second input of the predistortion function block (7) is connected to the input of the device. The first input of the block for calculating the parameters of the predistortion function (5) is connected to the first output of the control loop (3), and the second input is connected to the output of the second adder (4).

To compensate for distortions of the PA (2), which does not have a memory effect, the predistortion function block (7) contains a signal amplitude calculator (11), a parameterized function calculator (10) and a multiplier (9) connected in series, the output of the multiplier is the output of the predistortion function block, and the second input of the multiplier is connected to the input of the signal amplitude calculator and is the second input of the predistortion function block (7), while the first input of the predistortion function block is the second input of the parameterized function calculator (10).

The device operates as follows. After aligning the amplitudes, phases and delays of the input and output signals in the control loop (3), they go to the second adder (4), at the output of which a distortion signal is formed in accordance with expression (6). Using the distortion signal and the input signal that passed the control loop (3), in the block for calculating the parameters of the predistortion function (5), the vector of parameters of the functional dependence of the distortion signal on the input signal is calculated, which is transmitted through the closed switch (6) to the block of the predistortion function (7). In this case, the key (6) is an optional element, since the vector of the parameters of the predistortion function can be updated both continuously and occasionally, if necessary, or once during the manufacture of the device, if it is predefined for all modes of PA operation in the manner described above. The key (6) is used in accordance with the selected option of updating the vector of parameters in the block of the predistortion function. The second input of the predistortion function block (7) receives an input signal, which is used to calculate the value of the predistortion function at a given time. This is done by multiplying the input signal in the multiplier (9) by the value of the parameterized function f (…, Λ), obtained from the unit for calculating this function (10). In this case, the calculation of the parameterized function f (..., Λ) is carried out for the current value of the amplitude of the input signal obtained from the signal amplitude calculator (11).

At the output of the predistortion function block (7), a compensating signal is generated in accordance with expression (12). Then, in the phase equalization unit (8), the phase of the compensating signal is changed so that in the first adder (1) the compensating signal is subtracted from the input signal of the device in accordance with expression (9). As a result, a signal will be generated at the output of the PA in accordance with expression (10). In this case, the level of residual distortions decreases in a quadratic relationship with respect to the original value.

When compensating for distortions in PA (2) with a memory effect, in the predistortion function block (7), the functional dependence of the distortion on the input signal is calculated as the sum of the products of the delayed input signal and the parameterized function of the delayed input signal amplitude for all combinations of input signal delays and amplitude delays input signal at the memory interval of the power amplifier. In this case, the block of the predistortion function (7) contains M + 1 multiplied by K + 1 multipliers (9) and blocks for calculating the parameterized function f (10) with parameters Λ _mk , as well as M + K delay elements (12), the input signal amplitude calculator (11) and adder (13).

In the block for calculating the parameters of the predistortion function (5), a multidimensional set of parameters Λ _{mk is} calculated for the parameterized functions f (10) in block (7).

Brief Description of Drawings

In the figure, FIG. 1 shows a functional diagram of a device implementing a method for increasing the linearity of high-frequency power amplifiers.

In the figure, FIG. 2 shows an example of a circuit for calculating a distortion signal for a PA that does not have a memory effect.

In the figure, FIG. 3 shows an example of a functional diagram that calculates the distortion of a PA with a memory effect.

Implementation of the invention

The invention can be carried out in accordance with the circuit shown in FIG. 1. Signal conversion operations in accordance with the presented method can be performed at radio frequency, at an intermediate frequency, or in the baseband. In the figure, FIG. 1, the circuits necessary to transfer signals to the corresponding frequency range are not shown, but their use is assumed if necessary. Functional operations with signals in the block for calculating the parameters of the predistortion functions (5) and the block of the predistortion functions (7) are performed in digital form, however, other parts of the device can be performed both in digital and in analog form, therefore, in the description of embodiments of the method and device the operations of analog-to-digital and digital-to-analog conversion are not explicitly indicated, but their use is assumed where it is necessary, depending on the specific implementation.

Examples of the implementation of the procedure for aligning the amplitudes, phases and delays of the input and output signals in the control loop (3) can be devices for automatic gain control, phase-locked loop frequency control and adjustable delay lines, and it is also possible to implement them in digital form, for example, through mutual normalization operations signal amplitudes and search for delays by the correlation method.

The predistortion function block (7) can be executed, for example (but not limited to), on the basis of a programmable logic integrated circuit (FPGA), a computer, a digital signal processor (DSP-processor) and other computing devices.

For a PA without a memory effect, the block of the predistortion function (7) contains a multiplier (9), a block for calculating the parameterized function f (10) with parameters Λ, and a calculator of the input signal modulus (11) (see figure Fig. 2) and implements the expression (12 ).

For a PA with a memory effect in the figure, FIG. 3 shows an example of a scheme for calculating expression (13). In this case, the block of the predistortion function (7) contains M + 1 multiplied by K + 1 multipliers (9) and blocks for calculating the parameterized function f (10) with the parameters Λ _mk , M + K - delay elements (12), the calculator of the input signal module ( 11) and adder (13). The compensating signal generated by the predistortion function block (7) is subtracted in the first adder (1) from the input signal of the device in accordance with expression (9). As a result, a signal is generated at the PA output in accordance with expression (10). In this case, the level of residual distortions decreases in a quadratic relationship with respect to the original value.

The parameterized function calculator f (10) can be implemented in various ways, for example (but not limited to), as a p-degree interpolation lookup table, or as a polynomial calculation of degree q. In the case of a lookup table, the vector of parameters Λ is a vector of length L + 1, whose elements are the values of the function f at L + 1 points of the argument of the function f, usually (but not always) uniformly distributed over the domain of the function f. In the case of a polynomial of degree q, the vector of parameters Λ is a vector of length q + 1, whose elements are coefficients at the corresponding powers of the input argument of the function f.

The block for calculating the parameters of the predistortion function (5) contains a calculator, which can be executed, for example (but not limited to), on the basis of an FPGA, a computer, a DSP processor, and other computing devices.

One of the methods for calculating the parameters of the vector Λ is the linear regression method. For this, expression (12) can be represented in matrix form as

u _n =AΛ, (14)

where u _n = [u _n (0) u _n (1)… u _n (N-1)] ^T is a column vector of distortion counts of length N;

A is a matrix of input signals of size N by the length of the vector Λ.

Matrix A is formed from the elements of the input signal sample vector x _in = [x _in (0) x _in (1)… x _in (N-1)] ^T and zeros. The exact form of the matrix A depends on how the function f is calculated.

For example, for a function f, specified as a zero-degree interpolation lookup table, the vector Λ = [λ0) λ (1)… λ (L)]^T and the nth row of the matrixA has the forma _n = [a_n(0) a_n(1) ... a_n (L)]. In this case, all elements of the linea _n equal to zero, except for the element

, (15)

, (15)

where i _n is the index of the one closest to | x _in (n) | points.

Expression (14) can also be used for a model with memory (13), for this it is necessary to _{form one vector Λ from a multidimensional set of vectors of parameters Λ mk} , arranging the vectors Λ _mk , for example, by columns:

Λ = [Λ ₀₀ Λ ₁₀ … Λ _{M -1.0} Λ ₀₁ Λ _eleven ... Λ _{M -1.1} … Λ _{M -one, K -one} ]^T... (sixteen)

The row a _{n of the} matrix A can be formed from substrings in a corresponding way.

Thus, by solving the matrix equation (14), the vector of parameters Λ is found. One of the possible ways to solve (14) is the method of least squares (OLS).

Other possible options for calculating the Λ parameters are the calculation by the Least mean squares (LMS) or Recursive least squares (RLS) methods, known from the adaptive filtering theory (Simon Haykin: Adaptive Filter Theory, 5th edition, International Edition, Pearson Education, 2014, pages 266- 332 and pages 449-473).

The obtained parameters Λ through the closed key (6) are transmitted to the block of the predistortion function (7).

One of the options for the operation of the block for calculating the parameters of the function (5) is an iterative update of the parameters to track changes in the PA parameters over time. In this case, the initial values of the vector of parameters Λ (0) are set equal to zero or other predefined values and are transferred through the closed key (6) to the predistortion function block (7). Then, at each i-th iteration (starting from 1), the following actions are performed:

1. From the samples of signals arriving at the input of the block for calculating the parameters of the function (5), the vector Λ is calculated using one of the methods described above.

2. The current vector of parameters is updated Λ (i) = Λ (i - 1) + Λ.

3. The obtained current vector of parameters Λ (i) is transmitted through the closed key (6) to the block of the predistortion function (7).

Claims (6)

1. A method for increasing the linearity of a high-frequency power amplifier includes the process of isolating distortions of a power amplifier by subtracting the input signal from the output after their equalization in amplitudes, phases and delays, contains the process of forming a compensating signal corresponding to the selected distortions and mixing it in antiphase to the input signal of the power amplifier , characterized in that after the allocation of distortions, observing the input signal and distortions, approximate the functional dependence of the distortions on the input signal in the form of the product of the input signal by the parameterized function of the amplitude of the input signal, isolate this functional dependence in the form of a separate predistortion function and, using as an argument This function is an input signal, a compensating signal is synthesized. 2. A method for increasing the linearity of a high-frequency power amplifier according to claim 1, in which, when compensating for distortions in a power amplifier with a memory effect, the functional dependence of the distortion on the input signal is approximated as the sum of the products of the delayed input signal by the parameterized function of the delayed amplitude of the input signal for all combinations of delays the input signal and the delays of the amplitude of the input signal on the memory interval of the power amplifier. 3. A method for increasing the linearity of a high frequency power amplifier according to claim 1 or 2, wherein the parameterized function is implemented as a power p interpolation lookup table containing parameters indexed using the amplitude of the input signal. 4. The device for implementing the method according to claim 1 contains at least a first adder and a high-frequency power amplifier connected in series, while the input of the first adder is the input of the device, and the output of the power amplifier is the output of the device, the device also contains a control loop for mutual equalization of amplitudes, phases and delays of the input and output signals, while one input of the control loop is connected to the output of the device, and the other to the input of the device, the device also contains a second adder, the subtractive input of which is connected to the first output of the control loop, and the other input of the adder is connected to the second output of the loop in this case, respectively, at the first and second outputs of the control loop, the input and output signals of the power amplifier are formed, balanced in amplitude, phase and delay, characterized in that the device additionally includes a block for calculating the parameters of the predistortion function, and sequentially connected the predistortion function block and the phase alignment block, while the output of the phase alignment block is connected to the second input of the first adder, the second input of the predistortion function block is connected to the device input, the first input of the predistortion function parameter calculation block is connected to the first output of the control loop, and the second input to the output of the second adder, while the output of the block for calculating the parameters of the predistortion function is connected to the first input of the block of the predistortion function directly, thereby the predistortion function is updated continuously, or they are connected through a key, thereby the predistortion function is updated when the key is closed, and the second input of the predistortion function block is connected with device input. 5. The device according to claim 4, in which the predistortion function block contains a signal amplitude calculator, a parameterized function calculator and a multiplier connected in series, the multiplier output is the output of the predistortion function block, and the second multiplier input is connected to the signal amplitude calculator input and is the second input of the block the predistortion function, wherein the first input of the predistortion function block is the second input of the parameterized function calculator. 6. The device according to claim 4, in which, when compensating for distortions in a power amplifier with a memory effect, in the predistortion function block, the functional dependence of distortions on the input signal is calculated as the sum of the products of the delayed input signal by the parameterized function of the delayed input signal amplitude for all combinations of input delays signal and delays of the amplitude of the input signal on the memory interval of the power amplifier, while the block of the predistortion function contains M + 1 multiplied by K + 1 multipliers and blocks for calculating the parameterized function with the parameters Λ _mk , as well as M + K delay elements, the calculator of the amplitude of the input signal and an adder, where M and K are the memory depth for the input signal and its amplitude, respectively, in the block for calculating the parameters of the predistortion function, a multidimensional set of parameters Λ _{mk is} calculated for the parameterized functions in the block of the predistortion function.

Images