CN115882970A - Method and system for correcting received IQ imbalance - Google Patents
Method and system for correcting received IQ imbalance Download PDFInfo
- Publication number
- CN115882970A CN115882970A CN202110978397.1A CN202110978397A CN115882970A CN 115882970 A CN115882970 A CN 115882970A CN 202110978397 A CN202110978397 A CN 202110978397A CN 115882970 A CN115882970 A CN 115882970A
- Authority
- CN
- China
- Prior art keywords
- mismatch
- pha
- imbalance
- frequency
- amplitude
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
The invention discloses a method and a system for correcting received IQ imbalance, wherein the correction method comprises the following steps: s1: selecting specific M frequency points within a broadband signal frequency band range at equal intervals, wherein the selected frequencies are integral multiples of the frequency spectrum resolution; s2: estimating two paths of amplitude and phase imbalance information MAG (f) and PHA (f) of I/Q under M frequency points; s3: polynomial fitting to MAG (f) and PHA (f); s4: calculating mismatch parameters of each mismatch source; s5: constructing a calibration filter bank; s6: and compensating the RX receiving signal by using the obtained calibration filter bank. The method can improve the correction performance of the algorithm, simultaneously directly carries out polynomial fitting on I/Q amplitude and phase mismatch, only needs to carry out fitting on two parameters, greatly reduces the operation complexity and improves the working performance.
Description
Technical Field
The present invention relates to the field of radio frequency transceiving, and in particular, to a method and a system for correcting IQ mismatch.
Background
The zero if architecture in wireless communication systems is well known for its simple structure, low cost, low power consumption, and wide application in RF, but it also introduces a series of problems. The inevitable use of analog devices such as amplifiers, filters, mixers, etc. in communication systems is limited by problems of cost, power consumption, chip size, etc. Analog device fabrication is not ideal and results in amplitude mismatch of the I/Q two paths or a phase difference of the two paths of not 90 °, such deviations are collectively referred to as quadrature errors. Non-ideal analog parts affect system performance and are difficult to remove by analog methods, so proper correction strategies are designed in the digital processing part.
The existence of the quadrature error can introduce the image signal, which leads to the improvement of the error rate and greatly influences the communication quality. Therefore, the Rx end needs to perform the quadrature error correction process to suppress the image signal, reduce the error rate, and improve the communication quality. In the orthogonal error correction, an Image Rejection Ratio (IRR) is commonly used as an evaluation index of algorithm performance, and the IRR is defined as a ratio of the corresponding spectral energy of a signal of the signal itself to the spectrum of an image signal of the signal. The higher the IRR, the better the correction performance of its corresponding algorithm.
The Rx-side I/Q mismatch model is shown in fig. 1. As can be seen, the main sources of quadrature (I/Q) mismatch are three: (1) Radio frequency signal path mismatch (PD ERROR), including phase and amplitude ERRORs, before the received signal is demodulated to baseband by the mixer; (2) Analog baseband signal channel mismatch (BB ERROR), which is caused by channel delay and gain imbalance such as TIA and ADC, is after demodulation to baseband; (3) The I/Q local oscillator signal mismatch (LO ERROR) causes imbalance of the two paths I, Q when the local oscillator is not ideal, that is, the input I/Q signals have different amplitudes and the phase difference is not strictly 90 °. Where LOERROR is a frequency independent I/Q imbalance, and PD ERROR and BB ERROR are frequency dependent I/Q imbalances. In the narrow-band case only the frequency independent I/Q imbalance needs to be taken into account, but in the high-bandwidth case the frequency dependent I/Q imbalance needs to be taken into account as well. Based on this principle, the frequency independent and frequency dependent two-part imbalance is modeled separately, the model being shown in fig. 2. In this model, g and
respectively, amplitude and phase mismatch due to non-ideal Local Oscillators (LOs), the mismatch being independent of frequency. The sources of mismatch (PD ERROR and BB ERROR) that cause the frequency dependent I/Q imbalance are combined and modeled as H I (f) And H Q (f)。
The method in the prior art comprises the following steps: firstly, the amplitudes at different frequencies output by the I/Q mismatch analysis moduleDegree and phase mismatch MAG (f) and PHA (f) to directly calculate parameter information of different mismatch sources, namely LO PHA (f)、LO MAG (f)、BB PHA (f)、BB MAG (f)、PD PHA (f) And PD MAG (f) Wherein LO PHA (f) And LO MAG (f) Independent of frequency and therefore constant. Then, a polynomial fit is performed after obtaining mismatch parameters for different sources of mismatch. Therefore, four parameters in total need to be subjected to polynomial fitting, so that the calculation amount of the algorithm is greatly improved, more calculation resources are occupied, and the performance of the algorithm is influenced.
Disclosure of Invention
The present invention aims to overcome the defects of the prior art and provide a method for correcting IQ imbalance, and the purpose of the present invention is realized by the following technical scheme:
a method for correcting IQ imbalance includes the following steps:
s1: selecting specific M frequency points within a broadband signal frequency band range at equal intervals;
s2: estimating two paths of amplitude and phase imbalance information MAG (f) and PHA (f) of I/Q under M frequency points;
s3: polynomial fitting to MAG (f) and PHA (f);
s4: calculating mismatch parameters of each mismatch source;
s5: constructing a calibration filter bank;
s6: and compensating the RX receiving signal by using the obtained calibration filter bank.
Preferably, the selected frequency in S1 is an integer multiple of the spectral resolution.
Preferably, if the transmission signal in S2 is a sinusoidal signal r (t) =2Asin (2 pi f) r t), then:
for f t ＞f r
For f t ＜f r
The corresponding amplitudes are accordingly:
suppose that: the intermediate frequency being defined as f BB ＝abs(f t -f r ) (ii) a Namely: the phase imbalance caused by the baseband channel is E BB ＝2π·f BB τ, of course E BB Not only is the relative delay of two paths of I/Q generated;
wherein, f t And f r Local oscillator frequencies of transmission and reception, respectively, E LO Phase imbalance for receiving the local oscillator.
Preferably, in S3 of S3, a response function of the I/Q imbalance parameter in the whole frequency band is calculated by a polynomial coefficient a, where:
known at point x 1 ，x 2 ，…，x N Corresponding values are respectively y 1 ，y 2 ，…，y N Let k-th order polynomial be:
wherein a is 0 ，a 0 ，…，a k K +1 polynomial coefficients;
its corresponding polynomial coefficient a can be expressed as:
A＝(X·X T ) -1 ·X·Y，
wherein:
preferably, the mismatch source mismatch parameters in S4 are as follows:
PD PHA (f)＝PHA(f)-BB PHA (f)-LO PHA
wherein LO PHA Corresponding to the phase mismatch caused by the receiving local oscillator, the amplitude mismatch caused by the receiving local oscillator is irrelevant to the frequency, so the amplitude mismatch is often irrelevant to the amplitude mismatch BB caused by a baseband channel MAG (f) General consideration, BB PHA (f) For phase mismatch, PD, caused by baseband channel PHA (f) And PD MAG (f) Respectively, phase and amplitude mismatch caused by the radio frequency channel.
Preferably, the calibration filter bank comprises a real filter QFIR and a complex filter CFIR for the Q-path and a constant B.
A receiving IQ imbalance correction system includes an I/Q imbalance analysis module, an error analysis module and a correction module, wherein:
the I/Q imbalance analysis module is used for preprocessing an input single-tone test signal and estimating I/Q two-path amplitude and phase imbalance information MAG (f) and PHA (f) under different frequencies;
the error analysis module is mapped to more frequency points through polynomial fitting according to the output of the I/Q mismatch analysis module, namely amplitude and phase mismatch under different frequencies, response functions of I/Q two-path amplitude and phase mismatch in the whole frequency band are obtained, then parameter information from each mismatch source is calculated according to unbalance information of the amplitude and the phase, and finally a filter bank is constructed according to the mismatch parameter information;
the correction module performs initialization correction on the mismatch signal by using the filter banks QFIR and CFIR output by the error analysis module and a constant B.
Preferably, the I/Q imbalance analysis module includes a frequency point selection sub-module, an FFT/DFT sub-module and an estimation sub-module, which are connected in sequence, wherein:
the frequency point selection submodule selects the frequency of the input single-tone signal and is used for equally spacing signal frequencies in a frequency band, and the selected frequencies are integral multiples of the frequency spectrum resolution;
the FFT/DFT sub-module performs any one of fast Fourier transform or discrete Fourier transform on the signal to obtain a frequency spectrum value of the signal;
the estimation submodule is used for analyzing the frequency spectrum values of the corresponding positive and negative frequency pairs output by the FFT/DFT submodule to obtain amplitude and phase mismatch parameters under different frequencies.
Preferably, the error analysis module includes a polynomial fitting submodule, a data processing submodule, and a filter bank generating submodule, wherein:
the polynomial fitting submodule is used for carrying out polynomial fitting on the I/Q amplitude and phase mismatch MAG (f) and PHA (f) output by the I/Q mismatch analysis module to obtain a response function of the I/Q amplitude and phase mismatch MAG (f) and PHA (f) in the whole frequency band;
the data processing submodule is used for calculating mismatch parameters LO of each mismatch source according to the I/Q amplitude and phase mismatch MAG (f) and PHA (f) after polynomial fitting PHA 、BB PHA (f)、BB MAG (f)、PD PHA (f)、PD MAG (f)；
And the filter bank generation submodule calculates the filter banks QFIR and CFIR and the constant B according to the mismatch parameter tan/IFFT of each mismatch source.
Preferably, the correction module includes a BB ERROR correction sub-module, an LO ERROR correction sub-module, and a PD ERROR correction sub-module, wherein:
BB ERROR syndrome block: performing BB ERROR correction on the passing Q-path signal through QFIR;
LO ERROR syndrome module: multiplying the I path signal after fixed time delay processing by a constant B, and then adding the I path signal after being multiplied by the constant B to the Q path signal for correcting the LO ERROR;
PD ERROR syndrome module: and receiving and combining the Q path signal processed by the BB ERROR correction submodule and the LOERROR correction submodule and the I path signal after delay compensation, and sending the combined signal to the CFIR for correcting the PD ERROR.
The invention has the beneficial effects that:
1. the invention provides a frequency selection scheme for testing and estimating a single-tone signal, namely, a plurality of frequencies are selected at equal intervals in a frequency band, and the selected frequencies must meet integral multiples of the frequency spectrum resolution, so that an actual value can be obtained when the frequency domain value of the signal is extracted, the energy of a frequency point is prevented from being leaked to an adjacent point due to the problem of frequency spectrum leakage, and the correction performance of an algorithm is improved.
2. The invention provides a method for directly carrying out polynomial fitting on I/Q amplitude and phase mismatch MAG (f) and PHA (f) output by an I/Q mismatch analysis module to obtain a response function of the I/Q amplitude and phase mismatch MAG (f) and PHA (f) in the whole frequency band, and then calculating mismatch parameters of each mismatch source according to the I/Q amplitude and phase mismatch MAG (f) and PHA (f) after the polynomial fitting; is arranged in advance (X. X) T ) -1 And the table is stored, and can be directly called during calculation so as to reduce the operation amount and save resources.
Drawings
FIG. 1 is a diagram of a prior art Rx-side I/Q mismatch model.
Fig. 2 is a diagram of an I/Q mismatch model of a zero if receiver according to the prior art.
Fig. 3 is a flow chart of the method of the present invention.
FIG. 4 is a block diagram of an Rx-side I/Q imbalance initialization correction algorithm of the present invention.
Fig. 5 is a diagram of the simulation results of the initialization signal spectrum of the present invention.
FIG. 6 is a diagram illustrating an initial calibration simulation result according to the present invention.
FIG. 7 is a diagram of a simulation result of initialization correction according to the present invention.
FIG. 8 is a three-schematic diagram of the simulation result of initialization correction according to the present invention
Fig. 9 is a diagram illustrating an initial calibration result IRR according to the prior art.
Detailed Description
In order to clearly understand the technical features, objects and effects of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments.
In this embodiment, a method for correcting IQ imbalance is provided, where an initial correction of an I/Q imbalance at an Rx end is completed after power-on, a flowchart of a correction algorithm is shown in fig. 3, and the method includes:
s1: selecting specific M frequency points within a broadband signal frequency band range at equal intervals, wherein the selected frequencies are integral multiples of the frequency spectrum resolution;
because the frequency domain values correspond to the frequencies one by one, accurate frequency domain values can be obtained in a theoretical level. However, in a specific implementation, due to the limitations of the input test signal length and the sampling rate, the frequency domain value corresponding to each specific frequency cannot be obtained in the calculation, so this problem should be taken into consideration when selecting the test signal frequency.
In order to reflect the information of the whole frequency band, the present embodiment proposes to take the signal frequencies at equal intervals within the frequency band, and the selected frequencies are all integer multiples of the spectral resolution. When the selected frequency does not meet the integral multiple condition of the frequency spectrum resolution, the obtained frequency spectrum can be leaked, so that the frequency spectrum energy corresponding to the signal is not concentrated at the position of the signal frequency, the obtained frequency spectrum value is inaccurate, and the subsequent I/Q mismatch parameter estimation and the construction of a correction filter are influenced. Therefore, in order to solve this problem, several frequencies with integer times of the spectrum resolution should be selected at equal intervals in the frequency band to improve the algorithm correction performance.
S2: estimating two paths of amplitude and phase imbalance information MAG (f) and PHA (f) of I/Q under M frequency points;
if the transmitted signal is a sinusoidal signal r (t) =2Asin (2 π f) r t), the signal after receiving the local oscillator (LO ERROR) is:
I tmp ＝LPF{2Asin(2πf t t)·cos(2πf r t)}
＝Asin(2π(f t -f r )t)
wherein, after passing through the baseband BB ERROR, the signal is:
y I ＝sin(2π·(f t -f r )·t)＝sin(2π·(f t -f r )·t)
suppose that: f. of BB ＝abs(f t -f r ) (ii) a Namely: e BB ＝2π·f BB τ, of course E BB Not only does the I/Q two-way relative delay occur. If f is t ＞f r I.e., the received signal is at a positive frequency,
y I ＝A sin(2π·f BB ·t)
if f is t ＜f r I.e., the received signal is at a negative frequency,
y I ＝A sin(-2π·T BB ·t)＝A sin(2π·T BB ·t+π)
will y I And y Q Converting to a frequency domain through DFT; then:
therefore, the method comprises the following steps: for f t ＞f r
For f t ＜f r
The corresponding amplitudes are accordingly:
wherein g is BB For the amplitude imbalance caused by the baseband channel, assume: the intermediate frequency being defined as f BB ＝abs(f t -f r ) (ii) a Namely: the phase imbalance caused by the baseband channel is E BB ＝2π·f BB τ, of course E BB Not only the relative delay of the two paths of I/Q is generated;
wherein f is t And f r Respectively transmit and receive local oscillator frequencies, g LO For receiving amplitude imbalance caused by local oscillation, E LO Phase imbalance for receiving the local oscillator.
S3: polynomial fitting of MAG (f) and PHA (f);
in the embodiment, polynomial fitting is directly carried out on the I/Q amplitude and phase mismatches MAG (f) and PHA (f) output by the I/Q mismatch analysis module, so as to obtain the response function of the I/Q amplitude and phase mismatches MAG (f) and PHA (f) in the whole frequency band. The polynomial fitting procedure is as follows:
known at point x 1 ，x 2 ，…，x N Correspond toRespectively is y 1 ，y 2 ，…，y N Let k-th order polynomial be:
wherein a is 0 ，a 1 ，…，a k Is k +1 polynomial coefficients. The loss function based on the least squares method is the squared loss:
to find the value of a meeting the conditions, a is calculated k The partial derivatives are calculated and made equal to 0:
a simplification of the above equation can result:
the above equation set can be solved, and can be written in a matrix form:
assuming matrix X is a Van der Mond matrix:
simultaneously, the following requirements are met:
that is to say:
X·X T ·A＝X·Y
the polynomial coefficient a can be expressed as:
A＝(X·X T ) -1 ·X·Y
the response function of the I/Q imbalance parameter in the whole frequency band can be obtained according to the polynomial coefficient A. The polynomial fitting process has large computation amount, and the computation amount can be greatly reduced by directly carrying out polynomial fitting on the I/Q two-path amplitude and phase mismatch parameters. In addition, as is clear from the above analysis, polynomial fitting involves a matrix inversion process, which is a large amount of computation. While the matrix to be inverted (X. X) T ) -1 Only with the frequency vector of the input signal and the polynomial fitting order, independent of the actual output signal. Therefore, (X. X) can be arranged in advance T ) -1 And the table is stored, and can be directly called during calculation so as to reduce the calculation amount and save resources.
S4: calculating mismatch parameters of each mismatch source:
calculating parameter information from each mismatch source according to the imbalance information of the amplitude and the phase, wherein the theoretical reasoning process is as follows:
then:
PD PHA (f)＝PHA(f)-BB PHA (f)-LO PHA
wherein LO PHA Corresponding to the phase mismatch caused by the receiving local oscillator, the amplitude mismatch caused by the receiving local oscillator is irrelevant to the frequency, so the amplitude mismatch BB caused by the baseband channel is often related to MAG (f) General consideration, BB PHA (f) For phase mismatch, PD, caused by baseband channel PHA (f) And PD MAG (f) Phase and amplitude mismatches respectively caused by the radio frequency channels;
s5: constructing a calibration filter bank:
the compensation filter bank can be obtained by the mismatch parameters:
B＝tan(LO PHA )
the correction module performs initial correction on the mismatch signal by using the filter banks QFIR and CFIR output by the error analysis module and a constant B.
Wherein QFIR is a real filter for Q path, and CFIR is a complex filter.
S6: and compensating the RX receiving signal by using the obtained calibration filter group.
In this embodiment, the block diagram of the initialization correction algorithm is shown in fig. 4, and includes three modules: the device comprises an I/Q imbalance analysis module, an error analysis module and a correction module.
The I/Q imbalance analysis module is used for preprocessing an input single-tone test signal so as to estimate I/Q two-path amplitude and phase imbalance information MAG (f) and PHA (f) under different frequencies. The I/Q imbalance analysis module comprises a frequency point selection module, an FFT/DFT module and an estimation module which are connected in sequence, wherein the frequency point selection module selects the frequency of an input single-tone signal, preferably selects signal frequencies at equal intervals in a frequency band, and the selected frequencies are integral multiples of the frequency spectrum resolution. The FFT/DFT module performs any one of Fast Fourier Transform (FFT) or Discrete Fourier Transform (DFT) on the signal to obtain its spectral values. The estimation module analyzes the frequency spectrum values of the corresponding positive and negative frequency pairs output by the FFT/DFT module to obtain amplitude and phase mismatch parameters under different frequencies.
The error analysis module maps to more frequency points through polynomial fitting according to the output of the I/Q mismatch analysis module, namely amplitude and phase mismatch under different frequencies, obtains response functions of I/Q two-path amplitude and phase mismatch in the whole frequency band, calculates parameter information from each mismatch source according to unbalance information of the amplitude and the phase, and finally constructs a filter bank according to the mismatch parameter information. The error analysis module comprises a polynomial fitting module, a data processing module and a filter bank generation module. And the polynomial fitting module directly carries out polynomial fitting on the I/Q amplitude and phase mismatch MAG (f) and PHA (f) output by the I/Q mismatch analysis module to obtain a response function of the I/Q amplitude and phase mismatch MAG (f) and PHA (f) in the whole frequency band. The data processing module fits according to a polynomialThe subsequent I/Q amplitude and phase mismatches MAG (f) and PHA (f) are used to calculate the mismatch parameters LO of the respective mismatch sources PHA 、BB PHA (f)、BB MAG (f)、PD PHA (f)、PD MAG (f) In that respect And the filter bank generating module calculates the filter banks QFIR and CFIR and the constant B according to the mismatch parameter tan/IFFT of each mismatch source.
The correction module performs initial correction on the mismatch signal by using the filter banks QFIR and CFIR output by the error analysis module and a constant B. Since the center tap of the correction filter in some cases is not in the first bit, the I-path signal needs to be subjected to fixed delay processing according to practical situations, so as to avoid introducing additional I/Q imbalance in the correction process. The Q path signal is passed through QFIR for correction of BB ERROR. The LO ERROR can be corrected by multiplying the I signal by a constant B through a multiplier and adding the result to the Q signal through an adder. And finally, combining the I, Q signals into a complex signal, and then passing the complex signal through the CFIR for correcting the PD ERROR.
In this embodiment, channel modeling is performed according to the model shown in fig. 2, then initial orthogonal correction is performed according to the correction block diagram shown in fig. 4, performance test is performed using a single tone signal, and the signal spectrum before and after correction is shown in fig. 5.
In this embodiment, in order to test the calibration effect of the initialization algorithm at different frequency points, a plurality of single tone signals in a frequency band are used for testing to obtain IRRs at different frequencies, and the simulation result is shown in fig. 6. According to the simulation result, the IRR after the initialization correction is obviously improved compared with the IRR before the correction, and the IRR after the correction reaches more than 85 dB.
In this embodiment, when the frequency of the selected monophonic estimation signal does not satisfy the integral multiple of the spectral resolution, the performance of the algorithm will be affected, and the simulation results are shown in fig. 7 and 8. From fig. 6 to 8, three different cases were simulated, i.e., the deviation between the frequency and the spectral resolution of the selected tone signal was gradually increased from 0. Comparing the simulation results of the three graphs, it can be known that the correction performance of the algorithm is related to the deviation of the selected estimation frequency point and the spectral resolution, and the larger the deviation is, the worse the correction performance of the algorithm is.
Meanwhile, in the scheme provided by the prior art, after mismatch parameters of different mismatch sources are obtained, polynomial fitting is performed on the obtained parameters, and then four parameters need to be fitted, namely BB PHA (f)、BB MAG (f)、PD PHA (f) And PD MAG (f) Phase and amplitude imbalance parameters caused by BB ERROR and PD ERROR, respectively. The present embodiment proposes to prefix the polynomial fitting module, that is, to directly perform polynomial fitting on the amplitude and phase mismatch parameters, instead of the scheme originally proposed to perform polynomial fitting on the mismatch parameters of different mismatch sources. The simulation result of the scheme proposed by the prior art is shown in fig. 9, and it can be known from comparing fig. 6 and fig. 9 that the scheme proposed by the present embodiment has no influence on the correction performance of the algorithm, and the amount of computation can be greatly reduced and the algorithm performance can be improved by using this method.
In addition, because the matrix to be inverted in the polynomial fitting process is only related to the selected frequency point and the polynomial fitting order and is not related to the actual received signal, the matrix can be inverted in advance and then stored in a table.
When the I/Q imbalance initialization correction is performed on the Rx signal, the frequency of the test estimation tone signal will affect the estimation of the channel parameters, and this embodiment proposes a frequency selection scheme for the test estimation tone signal, that is, M frequencies are selected at equal intervals in a frequency band, and the selected frequencies must satisfy integer multiples of the spectral resolution, so that an actual value can be obtained when the frequency domain value of the signal is extracted, thereby avoiding the energy leakage of the frequency point to its neighboring points due to the problem of spectral leakage, and improving the correction performance of the algorithm.
In the prior art, after mismatch parameters of different mismatch sources are obtained, polynomial fitting is performed on the obtained parameters, and then four parameters need to be fitted. The embodiment proposes that the polynomial fitting module is arranged in front, polynomial fitting is directly carried out on I/Q amplitude and phase mismatch, only two parameters need to be fitted, the operation complexity is greatly reduced, and the working performance is improved.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (10)
1. A method for correcting IQ mismatch, comprising the steps of:
s1: selecting M specific frequency points within a broadband signal frequency band range at equal intervals;
s2: estimating two paths of amplitude and phase imbalance information MAG (f) and PHA (f) of I/Q under M frequency points;
s3: polynomial fitting to MAG (f) and PHA (f);
s4: calculating mismatch parameters of each mismatch source;
s5: constructing a calibration filter bank;
s6: and compensating the RX receiving signal by using the obtained calibration filter bank.
2. The IQ imbalance correction method according to claim 1, wherein the selected frequency in S1 is an integer multiple of the spectral resolution.
3. The IQ imbalance correction method according to claim 1, wherein if the transmitted signal in S2 is sinusoidal r (t) =2Asin (2 π f) r t), then:
for f t ＞f r
For f t ＜f r
The corresponding amplitudes are accordingly:
suppose that: the intermediate frequency being defined as f BB ＝abs(f t -f r ) (ii) a Namely: the phase imbalance caused by the baseband channel is E BB ＝2π·T BB τ, of course E BB Not only the relative delay of the two paths of I/Q is generated;
wherein, f t And f r For transmitting and receiving, respectively, local oscillator frequencies, E LO Phase imbalance for receiving the local oscillator.
4. The IQ imbalance correction method according to claim 1, wherein the response function of the I/Q imbalance parameters over the entire frequency band is calculated in S3 by polynomial coefficients a, wherein:
known at point x 1 ，x 2 ，…，x N Corresponding values are respectively y 1 ，y 2 ，…，y N Let k-th order polynomial be:
wherein a is 0 ，a 1 ，…，a k K +1 polynomial coefficients;
its corresponding polynomial coefficient a can be expressed as:
A＝(X·X T ) -1 ·X·Y，
wherein:
5. the IQ imbalance correction method according to claim 1, wherein mismatch source mismatch parameters in S4 are as follows:
PD PHA (f)＝PHA(f)-BB PHA (f)-LO PHA
wherein LO PHA Corresponding to the phase mismatch caused by the receiving local oscillator, the amplitude mismatch caused by the receiving local oscillator is irrelevant to the frequency, so the amplitude mismatch BB caused by the baseband channel is often related to MAG (f) General consideration, BB PHA (f) Is a basebandChannel induced phase mismatch, PD PHA (f) And PD MAG (f) Respectively, phase and amplitude mismatch caused by the radio frequency channel.
6. The IQ imbalance correction method according to claim 1, wherein the calibration filterbank comprises a real filter QFIR and a complex filter CFIR for the Q-path and a constant B.
7. A receiving IQ imbalance correction system, comprising an I/Q imbalance analysis module, an error analysis module and a correction module, wherein:
the I/Q imbalance analysis module is used for preprocessing an input single-tone test signal and estimating I/Q two-path amplitude and phase imbalance information MAG (f) and PHA (f) under different frequencies;
the error analysis module is mapped to more frequency points through polynomial fitting according to the output of the I/Q mismatch analysis module, namely amplitude and phase mismatch under different frequencies, response functions of I/Q two-path amplitude and phase mismatch in the whole frequency band are obtained, then parameter information from each mismatch source is calculated according to unbalance information of the amplitude and the phase, and finally a filter bank is constructed according to the mismatch parameter information;
the correction module performs initial correction on the mismatch signal by using the filter banks QFIR and CFIR output by the error analysis module and a constant B.
8. The IQ imbalance correction system according to claim 7, wherein the I/Q imbalance analysis module comprises a frequency point selection sub-module, an FFT/DFT sub-module and an estimation sub-module connected in sequence, wherein:
the frequency point selection submodule selects the frequency of the input single-tone signal and is used for equally spacing signal frequencies in a frequency band, and the selected frequencies are integral multiples of the frequency spectrum resolution;
the FFT/DFT sub-module performs any one of fast Fourier transform or discrete Fourier transform on the signal to obtain a frequency spectrum value of the signal;
the estimation submodule is used for analyzing the frequency spectrum values of the corresponding positive and negative frequency pairs output by the FFT/DFT submodule to obtain amplitude and phase mismatch parameters under different frequencies.
9. The received IQ imbalance correction system according to claim 7, wherein the error analysis module comprises a polynomial fitting submodule, a data processing submodule, and a filter bank generating submodule, wherein:
the polynomial fitting submodule is used for carrying out polynomial fitting on the I/Q amplitude and phase mismatch MAG (f) and PHA (f) output by the I/Q mismatch analysis module to obtain a response function of the I/Q amplitude and phase mismatch MAG (f) and PHA (f) in the whole frequency band;
the data processing submodule is used for calculating mismatch parameters LO of each mismatch source according to the I/Q amplitude and phase mismatch MAG (f) and PHA (f) after polynomial fitting PHA 、BB PHA (f)、BB MAG (f)、PD PHA (f)、PD MAG (f)；
And the filter bank generation submodule calculates and obtains filter banks QFIR and CFIR and a constant B according to the mismatch parameter tan/IFFT of each mismatch source.
10. The IQ imbalance correction system according to claim 7, wherein the correction module comprises a BB ERROR correction sub-module, a LO ERROR correction sub-module and a PD ERROR correction sub-module, wherein:
BB ERROR syndrome block: performing BB ERROR correction on the passing Q path signal through QFIR;
LO ERROR syndrome module: multiplying the I path of signals after fixed time delay processing by a constant B, and then adding the I path of signals after being multiplied by the constant B to the Q path of signals for correcting LO ERROR;
PD ERROR syndrome module: and receiving and combining the Q path signal and the I path signal after delay compensation after being processed by the BB ERROR correction submodule and the LO ERROR correction submodule, and sending the combined signal to the CFIR for correcting the PD ERROR.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110978397.1A CN115882970A (en) | 2021-08-24 | 2021-08-24 | Method and system for correcting received IQ imbalance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110978397.1A CN115882970A (en) | 2021-08-24 | 2021-08-24 | Method and system for correcting received IQ imbalance |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115882970A true CN115882970A (en) | 2023-03-31 |
Family
ID=85762313
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110978397.1A Pending CN115882970A (en) | 2021-08-24 | 2021-08-24 | Method and system for correcting received IQ imbalance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115882970A (en) |
-
2021
- 2021-08-24 CN CN202110978397.1A patent/CN115882970A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9749172B2 (en) | Calibration method and calibration apparatus for calibrating mismatch between first signal path and second signal path of transmitter/receiver | |
| US9036753B1 (en) | Calibration method and calibration apparatus for calibrating mismatch between I-path and Q-path of transmitter/receiver | |
| US8699552B2 (en) | Mechanism for measuring transmitter impairments using offset local oscillators | |
| KR101234963B1 (en) | I/q calibration of transmit and receive paths in ofdm fdd communication systems | |
| CN109560825B (en) | Quadrature error correcting method for zero intermediate frequency receiver | |
| US20040082305A1 (en) | Sideband suppression method and apparatus for quadrature modulator using magnitude measurements | |
| US8638893B2 (en) | Mechanisms for the correction of I/Q impairments | |
| CN109617560A (en) | A kind of I/Q signal compensation for calibrating errors method | |
| EP2487854A1 (en) | I/Q imbalance estimation and correction in a communication system | |
| TW201301818A (en) | Devices of IQ mismatch calibration, and methods thereof | |
| CN114374593B (en) | IQ imbalance compensation method for WiFi broadband transceiving path and application | |
| US9385656B2 (en) | Low complexity frequency selective IQ mismatch digital RX balancer and TX inverse balancer for non-ideal RF front-end | |
| US9281907B2 (en) | Quadrature error correction using polynomial models in tone calibration | |
| CN104601259A (en) | Wireless communication receiver with i/q imbalance estimation and correction techniques | |
| CN103873416B (en) | EVM (Error Vector Magnitude) phase estimating and compensating method | |
| US8811538B1 (en) | IQ error correction | |
| US11528179B1 (en) | System, apparatus, and method for IQ imbalance correction for multi-carrier IQ transmitter | |
| TW200901691A (en) | Channel estimation using frequency smoothing | |
| CN116781470A (en) | Broadband signal in-phase quadrature imbalance correction method and system | |
| US20100034250A1 (en) | IQ imbalance equalization system and method. | |
| CN107819720B (en) | IQ imbalance compensation method and device, time domain compensator and communication equipment | |
| CN115882970A (en) | Method and system for correcting received IQ imbalance | |
| CN113438039B (en) | Transmitter quadrature mismatch calibration method and device based on self-mixing | |
| CN109474553B (en) | Method and system for estimating signal distortion parameters of terminal transmission path | |
| KR20010076760A (en) | A Method and Apparatus for Multi-channel Calibration |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |