Multi-frequency parallel compensation method for nonlinear phase of radio frequency device

Abstract

The invention belongs to the technical field of digital signal processing in a radio frequency / microwave communication system, and particularly discloses a multi-frequency parallel compensation method for a nonlinear phase of a radio frequency device, which is characterized by comprising the following steps of: S1, acquiring a phase-frequency curve p (f) of an output signal of a radio frequency end of the radio frequency device; s2, converting the phase frequency curve p (f) into a plurality of amplitude-frequency responses amp (f), performing inverse Fourier transform on the amp (f) to generate a filter coefficient hp, and performing M-order truncation and N-bit fixed-point quantization on the hp to generate an all-pass filter coefficient hp (m); s3, performing orthogonal down-conversion on a digital intermediate frequency signal y (n) to be compensated to a baseband, inputting the digital intermediate frequency signal y (n) into the all-pass filter, and simultaneously completing phase compensation on all frequency components contained in the y (n) by utilizing a group delay characteristic; and the compensated baseband signal is recovered to the original intermediate frequency through quadrature up-conversion, and an output signal with a corrected nonlinear phase is obtained. Phase modulation can be carried out on multiple frequency signals at the same time, frequency measurement is not needed, and frequency and data alignment does not need to be considered.

Description

Technical Field

[0001] This invention belongs to the field of digital signal processing technology in radio frequency / microwave communication systems, and specifically relates to a multi-frequency parallel compensation method for the nonlinear phase of radio frequency devices. Background Technology

[0002] Radio frequency devices internally employ infinite impulse response (IIR) filters. After the signal passes through, it acquires a non-linear phase, making it difficult to maintain phase consistency across harmonics. This results in phase disturbances in the output signal, potentially causing waveform distortion. Phase compensation can be performed within an FPGA to restore a roughly linear phase relationship. The steps are as follows: Step 1: Determine the phase frequency relationship of the nonlinear phase, use a vector network analyzer to measure the phase frequency curve of the RF device, and save the phase frequency data; Step 2: Use BRAM in the FPGA to store the phase frequency data of the RF device. The phase frequency data is transmitted via SPI. Read the compensation phase value at the current frequency according to the measured frequency. Step 3: Intra-pulse frequency measurement. The compensated phase data is obtained based on the frequency. The frequency needs to be measured in real time. Phase difference is used to measure the frequency, and the frequency measurement and data are aligned in time.

[0003] Step 4: Use the read phase data to perform phase adjustment.

[0004] This phase compensation method requires strict timing alignment to ensure that the read compensation phase value corresponds one-to-one with the data. Moreover, when multiple frequency signals arrive at the same time, only one phase compensation data can be read at this time, and phase compensation cannot be performed on multiple frequency signals at the same time. Summary of the Invention

[0005] Purpose of the invention: The purpose of this invention is to provide a multi-frequency parallel compensation method for the nonlinear phase of radio frequency devices that can simultaneously modulate the phase of multiple frequency signals.

[0006] Technical solution: The present invention provides a multi-frequency parallel compensation method for the nonlinear phase of radio frequency devices, comprising the following steps: S1. Obtain the phase frequency curve p(f) of the output signal from the RF terminal of the RF device; S2. Convert the phase frequency curve p(f) into a complex amplitude frequency response amp(f), perform an inverse Fourier transform on amp(f) to generate filter coefficients hp, and perform M-order truncation and N-bit fixed-point quantization on hp to generate all-pass filter coefficients hp(m). S3. After the digital intermediate frequency signal y(n) to be compensated is orthogonally downconverted to baseband, it is input into the all-pass filter. The group delay characteristic is used to simultaneously complete phase compensation for all frequency components contained in y(n). The compensated baseband signal is restored to the original intermediate frequency by orthogonal upconversion, resulting in an output signal whose nonlinear phase has been corrected.

[0007] A further limitation of the technical solution of the present invention is that the process of performing phase compensation on the phase frequency curve p(f) in step S2 to generate an all-pass filter specifically includes the following steps: S2.1 Convert the phase frequency curve p(f) into a complex amplitude frequency response amp(f): , where j is the imaginary unit of complex number operations; S2.2 performs an inverse Fourier transform on the complex amplitude-frequency response amp(f) to generate the time-domain filter coefficients hp, and then performs M-order truncation and N-bit fixed-point quantization on hp to generate the programmable all-pass filter coefficients hp(m). , In the formula, m represents the number of the filter coefficients, m=1,2,…,M+1; Represents the real part of the filter coefficients. represents the imaginary part of the filter coefficient; i represents the filter coefficient; floor[·] represents rounding down; the number of ifft points is M+1; and the quantization bit width N is determined by the fixed-point operation bit width of the FPGA in the RF device.

[0008] Furthermore, in step S3, the orthogonal downconversion has a two-stage structure, specifically: S3.1 The first-stage down-conversion shifts the signal spectrum to fs / 4. The input intermediate frequency signal is represented as The signal after downsampling by a quarter Represented as , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. This represents the intermediate frequency signal frequency, where j is the imaginary unit for complex number operations; Down-conversion of the first-level down-conversion signal Frequency, the filtered signal is represented as In the formula, n = 1, 2, 3...

[0009] S3.2 The second-stage down-conversion shifts the signal to zero intermediate frequency and introduces a cancelable frequency error Δf1 into the second-stage local oscillator. Signal and The frequency signal is mixed, and the resulting second-stage down-converted signal is... Represented as , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. Indicates the intermediate frequency signal frequency. Indicates DDS generation The frequency error generated when dealing with frequency signals.

[0010] Furthermore, in step S3, the orthogonal upconversion has a two-stage structure, specifically: S3.3 performs the first-stage up-conversion on the down-converted baseband signal, shifting the signal back to fs / 4. The frequency of the first-stage up-conversion is fs / 4, and the signal after the first-stage up-conversion is represented as follows: In the formula, n = 1, 2, 3…; S3.4 performs a second-stage up-conversion on the signal after the first-stage up-conversion. By inverting the imaginary part of the DDS output signal, it makes the equivalent output local oscillator signal of fc-fs / 4+Δf1, thereby canceling the error Δf1 of the previous stage and ensuring the phase compensation accuracy.

[0011] The frequency of the second-stage upconversion is: The input signal frequency of the DDS is At this point, since the frequencies are the same, the frequency error of the signal generated by the DDS is also [missing value]. According to Euler's formula , Inverting the imaginary part of the DDS output signal will change the DDS output... The frequency signal becomes The signal of frequency, the first-stage up-converted signal and the frequency of Signal mixing, the mixed signal Represented as , Right now , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. Indicates the intermediate frequency signal frequency; S3.5 passes the digitally up-converted signal through a phase-compensated all-pass filter. .

[0012] Beneficial effects: Compared with the prior art, the present invention transforms the phase frequency characteristics into filter coefficients through the design of an all-pass filter. It can simultaneously complete phase compensation for all frequency components in the signal without frequency measurement, overcoming the problem that traditional compensation schemes can only perform phase compensation for signals of one frequency at a time and cannot be applied to scenarios where multiple frequencies exist simultaneously. It can simultaneously perform phase modulation on multiple frequency signals without frequency measurement or consideration of frequency and data alignment. Detailed Implementation

[0013] The technical solution of the present invention will be described in detail below, but the scope of protection of the present invention is not limited to the embodiments described.

[0014] This embodiment provides a multi-frequency parallel compensation method for the nonlinear phase of radio frequency devices. In scenarios involving phase compensation, it can simultaneously perform phase compensation on multiple frequency signals. The detailed steps are described below.

[0015] S1. Use a vector network analyzer to obtain the phase frequency curve p(f) of the output signal of the RF terminal of the RF device, where f is the operating frequency of the RF device.

[0016] S2. Perform phase compensation on the phase frequency curve p(f), convert the phase frequency characteristics into filter coefficients, and generate all-pass filter coefficients hp(m). Phase compensation can be completed simultaneously for all frequency components in the signal without frequency measurement.

[0017] S2.1 Convert the phase frequency curve p(f) into a complex amplitude frequency response amp(f): , where j is the imaginary unit of complex number operations, used to convert the phase corresponding to a computable complex number form.

[0018] S2.2 performs an inverse Fourier transform on the complex amplitude-frequency response amp(f) to generate the time-domain filter coefficients hp, and then performs M-order truncation and N-bit fixed-point quantization on hp to generate the programmable all-pass filter coefficients hp(m). , In the formula, m represents the number of the filter coefficients, m=1,2,…,M+1; Represents the real part of the filter coefficients. represents the imaginary part of the filter coefficients; i represents the filter coefficients; floor[·] represents rounding down; the number of ifft points is M+1; and the quantization bit width N is determined by the fixed-point operation bit width of the FPGA.

[0019] This embodiment directly converts phase-frequency distortion data into all-pass filter coefficients, using the filter group delay characteristics to equivalently compensate for nonlinear phase, thus avoiding the timing bottleneck of the traditional lookup table method (BRAM+SPI).

[0020] S3. After the digital intermediate frequency signal y(n) to be compensated is orthogonally down-converted to baseband, it is input into the all-pass filter. The group delay characteristic is used to simultaneously perform phase compensation on all frequency components contained in y(n). The compensated baseband signal is restored to the original intermediate frequency by orthogonal upconversion, resulting in an output signal whose nonlinear phase has been corrected.

[0021] Both the quadrature downconversion and the quadrature upconversion are two-stage structures: S3.1 The first-stage downconversion shifts the signal spectrum to fs / 4.

[0022] The input intermediate frequency signal is represented as The signal after downsampling by a quarter Represented as , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. This represents the intermediate frequency signal frequency, and j is the imaginary unit for complex number operations.

[0023] Down-conversion of the first-level down-conversion signal Frequency, the filtered signal is represented as In the formula, n = 1, 2, 3...

[0024] S3.2 The second-stage downconversion shifts the signal to zero intermediate frequency and introduces a cancelable frequency error Δf1 into the second-stage local oscillator.

[0025] Signal and The frequency signal is mixed, and the resulting second-stage down-converted signal is... Represented as , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. Indicates the intermediate frequency signal frequency. Indicates DDS generation The frequency error generated when dealing with frequency signals.

[0026] S3.3 performs the first-stage up-conversion on the down-converted baseband signal, bringing the signal back to fs / 4.

[0027] The frequency of the first-stage up-conversion is fs / 4, and the signal after the first-stage up-conversion is represented as follows: In the formula, n = 1, 2, 3...

[0028] S3.4 performs a second-stage up-conversion on the signal after the first-stage up-conversion. By inverting the imaginary part of the DDS output signal, it makes the equivalent output local oscillator signal of fc-fs / 4+Δf1, thereby canceling the error Δf1 of the previous stage and ensuring the phase compensation accuracy.

[0029] The frequency of the second-stage upconversion is: The input signal frequency of the DDS is At this point, since the frequencies are the same, the frequency error of the signal generated by the DDS is also [missing value]. According to Euler's formula , Inverting the imaginary part of the DDS output signal will change the DDS output... The frequency signal becomes The signal of frequency, the first-stage up-converted signal and the frequency of Signal mixing, the mixed signal Represented as , Right now , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. Indicates the intermediate frequency signal frequency; S3.5 passes the digitally up-converted signal through a phase-compensated all-pass filter. .

[0030] This embodiment adopts a two-stage downconversion + two-stage upconversion architecture. After the signal is moved to the baseband, phase compensation is completed in the digital domain using an all-pass filter, and then it is restored to the original frequency.

[0031] Existing compensation schemes can only perform phase compensation for signals of a single frequency at the same time, and cannot be applied to scenarios where multiple frequencies coexist. The technology of this invention can simultaneously modulate the phase of multiple frequency signals, and does not require frequency measurement or consideration of frequency and data alignment.

[0032] As described above, although the invention has been shown and described with reference to the specified preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A multi-frequency parallel compensation method for the nonlinear phase of a radio frequency device, characterized in that, Includes the following steps: S1. Obtain the phase frequency curve p(f) of the output signal from the RF terminal of the RF device; S2. Convert the phase frequency curve p(f) into a complex amplitude frequency response amp(f), perform an inverse Fourier transform on amp(f) to generate filter coefficients hp, and perform M-order truncation and N-bit fixed-point quantization on hp to generate all-pass filter coefficients hp(m). S3. After the digital intermediate frequency signal y(n) to be compensated is orthogonally downconverted to baseband, it is input into the all-pass filter. The group delay characteristic is used to simultaneously complete phase compensation for all frequency components contained in y(n). The compensated baseband signal is restored to the original intermediate frequency by orthogonal upconversion, resulting in an output signal whose nonlinear phase has been corrected.

2. The multi-frequency parallel compensation method for the nonlinear phase of a radio frequency device according to claim 1, characterized in that, Step S2 involves performing phase compensation on the phase frequency curve p(f) to generate an all-pass filter. The specific steps are as follows: S2.1 Convert the phase frequency curve p(f) into a complex amplitude frequency response amp(f): , where j is the imaginary unit of complex number operations; S2.2 performs an inverse Fourier transform on the complex amplitude-frequency response amp(f) to generate the time-domain filter coefficients hp, and then performs M-order truncation and N-bit fixed-point quantization on hp to generate the programmable all-pass filter coefficients hp(m). ， In the formula, m represents the number of the filter coefficients, m=1,2,…,M+1; Represents the real part of the filter coefficients. represents the imaginary part of the filter coefficient; i represents the filter coefficient; floor[·] represents rounding down; the number of ifft points is M+1; and the quantization bit width N is determined by the fixed-point operation bit width of the FPGA in the RF device.

3. The multi-frequency parallel compensation method for the nonlinear phase of a radio frequency device according to claim 1, characterized in that, In step S3, the orthogonal downconversion has a two-stage structure, specifically: S3.1 The first-stage down-conversion shifts the signal spectrum to fs / 4. The input intermediate frequency signal is represented as Signal after downsampling by a quarter Represented as , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. This represents the intermediate frequency signal frequency, where j is the imaginary unit for complex number operations; Down-conversion of the first-level down-conversion signal Frequency, the filtered signal is represented as In the formula, n = 1, 2, 3…; S3.2 The second-stage down-conversion shifts the signal to zero intermediate frequency and introduces a cancelable frequency error Δf1 into the second-stage local oscillator. Signal and The frequency signal is mixed, and the resulting second-stage down-converted signal is... Represented as ， Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. Indicates the intermediate frequency signal frequency. Indicates DDS generation The frequency error generated when dealing with frequency signals.

4. The multi-frequency parallel compensation method for nonlinear phase of radio frequency devices according to claim 3, characterized in that, In step S3, the orthogonal upconversion has a two-stage structure, specifically: S3.3 performs the first-stage up-conversion on the down-converted baseband signal, shifting the signal back to fs / 4. The frequency of the first-stage up-conversion is fs / 4, and the signal after the first-stage up-conversion is represented as follows: In the formula, n = 1, 2, 3…; S3.4 performs a second-stage up-conversion on the signal after the first-stage up-conversion. By inverting the imaginary part of the DDS output signal, it makes the equivalent output local oscillator signal of fc-fs / 4+Δf1, thereby canceling the error Δf1 of the previous stage and ensuring the phase compensation accuracy; The frequency of the second-stage upconversion is: The input signal frequency of the DDS is At this point, since the frequencies are the same, the frequency error of the signal generated by the DDS is also [missing value]. According to Euler's formula ， Inverting the imaginary part of the DDS output signal will change the DDS output... The frequency signal becomes The signal of frequency, the first-stage up-converted signal and the frequency of Signal mixing, the mixed signal Represented as ， Right now , Right now In the formula, n = 1, 2, 3, ... Indicates the sampling rate. Indicates the intermediate frequency signal frequency; S3.5 passes the digitally up-converted signal through a phase-compensated all-pass filter. .

Images