CN104486273A - Self-adaptive direct quadrature variable-frequency modulation error correction method - Google Patents

Abstract

The invention discloses a self-adaptive direct quadrature variable-frequency modulation error correction method. The method creates a new quadrature modulation error model, and puts forward a self-adaptive QMC (Quadrature Modulation Compensation) algorithm for direct quadrature variable-frequency modulation error correction on the basis of the model. Through antenna coupling and down-conversion mixing, by obtaining a feedback signal after intermediate-frequency high-speed sampling, a quadrature modulation error inverse distortion coefficient is estimated after the feedback signal and a transmission signal are aligned, and self-adaptive correction is implemented in a numeric field programmable logic device FPGA (field programmable gate array). The self-adaptive direct quadrature variable-frequency modulation error correction method can efficiently and accurately inhibit local oscillator leakage caused by direct-current bias and image interference caused by I/Q amplitude and phase imbalance, and has important significance for self-adaptive direct quadrature variable-frequency modulation error correction.

Description

A kind of self adaptation direct orthogonalization Frequency-variable Modulation error calibration method

Technical field

The present invention relates to a kind of self adaptation direct orthogonalization Frequency-variable Modulation error calibration method, belong to digital signal processing technique field.

Background technology

In digital-communications transmission system, need by upconvert baseband signals to radio frequency, be conducive to the high efficiency of transmission of signal and improve the utilance of frequency spectrum resource.Real signal secondary or three frequency conversions digital low intermediate frequency signal is converted to simulation low intermediate frequency signal after, after bandpass filtering, then carry out once or double conversion to specify frequency.Repeatedly frequency conversion system consistency of performance is high, and eliminates local oscillator leakage and image spectra by filtering, shortcoming be to circuit and requirement of system design high, and phase noise is poor.In recent years, a kind of novel Direct frequency conversion based on quadrature modulation, namely direct orthogonalization converter technique obtained and developed rapidly.This technology adopts the orthogonal principle that offsets to eliminate unwanted sideband signal, suppresses image spectra.Its outstanding advantages does not need that intermediate frequency amplifies, filtering and frequency conversion part, and relax the performance requirement of radio frequency part filter, even do not need to add radio-frequency filter, thus greatly reduce the volume of transmitter, weight, power consumption and cost.But width the balancing each other property of direct orthogonalization converter technique to baseband signal and local oscillation signal requires very high, the imbalance of width phase will reduce the image interference rejection ability of frequency converter, the adjacent-channel power rejection ratio of transmitter declines.Meanwhile, cause local-oscillator leakage owing to inevitably there is the various problems such as crosstalk, radiation, direct current offset modulation in circuit, and due to the local oscillation signal of useful signal and leakage lean on frequency spectrum very near, stripper filtering cannot be utilized.Local oscillator leakage signals can reduce the efficiency of transmitter on the one hand, receiver DC point can be caused to offset on the other hand, easily cause nonlinear distortion, even occur saturation jamming phenomenon.

Through the retrieval to prior art, Chinese Patent Application No. is CN200810207707, publication number is the technical literature of CN1707962A, disclose the adaptive equalization system of IQ amplitude in a kind of direct frequency conversion modulation, in the unbroken situation of main signal, automatically can monitor, follow the tracks of and compensate the IQ amplitude difference that in whole up-conversion link, all devices cause because of ambient temperature and humidity change, self adaptation feedback arrangement precision is high simultaneously.Chinese Patent Application No. is CN200810207707, publication number CN 101478287B discloses the adaptive elimination system of carrier wave leakage in a kind of direct frequency conversion modulation, in the unbroken situation of main signal, automatically can monitor, follow the tracks of and compensate the DC component caused because of ambient temperature and humidity change in all devices in whole up-conversion link, self adaptation feedback arrangement precision is high simultaneously.

Therefore, designing one can either suppress image disturbing signal, simultaneously can cause the direct orthogonalization Frequency-variable Modulation method of local oscillator leakage by DC-offset correction, become a kind of new technical need.

Summary of the invention

(1) technical problem that will solve

The present invention is directed to prior art above shortcomings, the reason of error is produced by analyzing quadrature modulation process, set up a kind of new error model, self adaptation QMC (the Quadrature Modulation Compensation) algorithm of a kind of direct orthogonalization Frequency-variable Modulation error correction is proposed, the image disturbing signal that this algorithm can suppress imbalance of amplitude and phase to cause, again can the simultaneously local oscillator leakage that causes of DC-offset correction.

At digital baseband to quadrature modulation errors founding mathematical models, desirable orthogonal demodulation signal is as follows:

$\mathrm{RF}=\mathrm{Re}\{(I+\mathrm{jQ})\×e^{j{\mathrm{\ω}}_{c}t}\}=\mathrm{Re}\{S\left(t\right)\×e^{j{\mathrm{\ω}}_{c}t}\}---\left(1\right)$

Wherein, Re represents the real part got in complex operation, and I represents in-phase signal, and Q represents orthogonal signalling, and j represents the imaginary unit in plural number, w _crepresent angular frequency, t represents continuous time signal, when baseband signal s (t) being superimposed with DC component, then exists:

$\begin{array}{c}{I}^{\′}=I+I_{\mathrm{DC}}\\ Q^{\′}=Q+Q_{\mathrm{DC}}\end{array}---\left(2\right)$

Wherein, I _dCrepresent homophase DC component, Q _dCrepresent orthogonal DC component, suppose that quadrature modulator is completely orthogonal, after quadrature modulation, so obtain radiofrequency signal is:

$\mathrm{RF}=\mathrm{Re}\{S^{\′}\left(t\right)\×e^{{\mathrm{j\ω}}_{c}t}\}=\mathrm{Re}\{s\left(t\right)\×e^{{\mathrm{jw}}_{c}t}\}+I_{\mathrm{DC}}{\cos \mathrm{\ω}}_{c}t+Q_{\mathrm{DC}}\sin {\mathrm{\ω}}_{c}t---\left(3\right)$

Can find out, when IQ being superimposed with DC component, modulation signal, except desired radiofrequency signal, also can produce an extra carrier wave leakage signal on a carrier frequency.

The unbalanced problem of IQ extensively occurs in a communications system, shows the real part imaginary part not strict orthogonal of signal in communication process, or power is not etc.To simplify the analysis, with I road for benchmark, only calculate the error entered at Q pass.

When only there is amplitude imbalance, if the amplitude gain on Q road is β, have:

Q′＝(1+β)Q (4)

The radiofrequency signal obtained after quadrature modulation there will be the double-sideband signal of a frequency centered by carrier wave:

$\mathrm{RF}=\mathrm{Re}\left\{\right[I+(1+\mathrm{\β})\mathrm{jQ}]\×e^{j{\mathrm{\ω}}_{c}t}\}=\mathrm{Re}\{s\left(t\right)\×e^{j{w}_{c}t}\}-\mathrm{\β}Q\sin {\mathrm{\ω}}_{c}t---\left(5\right)$

When only there is unbalance in phase, the offset error on Q road is that θ, Q road signal is modulated onto Q ' road, thus causes the changes in amplitude on actual Q road and enter the interference of Q road at I pass:

$\begin{array}{c}{I}^{\′}=I-Q\sin \mathrm{\θ}\\ Q^{\′}=Q\cos \mathrm{\θ}=Q+(\cos \mathrm{\θ}-1)Q\end{array}---\left(6\right)$

The baseband complex signal of unbalance in phase, after ovennodulation, can produce double-side band interference signal:

$\mathrm{RF}=\mathrm{Re}\left\{\right[I^{\′}+{\mathrm{jQ}}^{\′}]\×e^{j{\mathrm{\ω}}_{c}t}\}=\mathrm{Re}\{s\left(t\right)\×e^{j{w}_{c}t}\}-Q\sin \mathrm{\θ}\sin {\mathrm{\ω}}_{c}t-(\cos \mathrm{\θ}-1)Q\sin {\mathrm{\ω}}_{c}t---\left(7\right)$

Quadrature modulation errors Mathematical Modeling expression formula can be obtained by formula (2), (4) and (6):

$\begin{array}{c}{I}^{\′}=I+\mathrm{pQ}+I_{\mathrm{DC}}\\ Q^{\′}=I+\mathrm{qQ}+Q_{\mathrm{DC}}\end{array}---\left(8\right)$

Formula (8) is generalized to more general situation, and with matrix representation, following error model can be obtained:

$\left[\begin{array}{c}{I}^{\′}\\ Q^{\′}\end{array}\right]=\left[\begin{array}{ccc}{a}_1& a_2& a_3\\ a_4& a_5& a_6\end{array}\right]\left[\begin{array}{c}I\\ Q\\ 1\end{array}\right]---\left(9\right)$

Wherein, a _irepresent the coefficient of QMC, formula (9) directly describes the uneven gain of I/Q signal in transmitting procedure and the direct current biasing of superposition, and the distortion that indirectly unbalance in phase of I/Q signal causes with the formal description of IQ crosstalk.

(2) technical scheme

The present invention is in wireless transmitting system, pass through antenna-coupled, and obtain feedback signal by intermediate frequency high-speed sampling, to feedback signal with transmit and carry out registration process after, estimate that quadrature modulation errors is against distortion QMC coefficient, adaptively correcting is realized in numeric field chip FPGA, the method can the local oscillator leakage that caused by direct current biasing of the suppression of efficiently and accurately and the Image interference that caused by IQ imbalance of amplitude and phase, the self adaptation direct orthogonalization Frequency-variable Modulation error calibration method that the present invention proposes, described method comprises the steps:

S1:QMC coefficient initialization, and from the digital baseband signal of input, intercepted length is the base band data of L, is corrected by described base band data by QMC equalizer;

S2: digital-to-analogue conversion is carried out to the data after correcting, transmits after quadrature modulation;

S3: by antenna-coupled receiving feedback signals, carries out down-conversion mixing and if sampling to feedback signal, gives FPGA carry out quadrature demodulation by the digital signal y (n) obtained;

S4: Delay Estima-tion is carried out to the signal after demodulation, calculates time of delay, and time delay registration process is carried out to raw baseband data z (n);

S5: estimate QMC coefficient.

Preferably, the error model of described quadrature modulation is such as formula shown in (9):

$\left[\begin{array}{c}{I}^{\′}\\ Q^{\′}\end{array}\right]=\left[\begin{array}{ccc}{a}_1& a_2& a_3\\ a_4& a_5& a_6\end{array}\right]\left[\begin{array}{c}I\\ Q\\ 1\end{array}\right]$

Wherein, a ₁, a ₂, a ₃, a ₄, a ₅, a ₆represent QMC coefficient, described model directly describes the uneven gain of I/Q signal in transmitting procedure and the direct current biasing of superposition, and the distortion that indirectly unbalance in phase of I/Q signal causes with the formal description of IQ crosstalk, according to this error model, quadrature modulation errors can be obtained against distortion factor a _i.

Preferably, by antenna-coupled in described S3, and obtain feedback signal by intermediate frequency high-speed sampling, to feedback signal with transmit and carry out registration process after, estimate that quadrature modulation errors is against distortion factor.

Preferably, the linear equation of the estimation QMC coefficient in described S5 is as follows:

Z＝UA (10)

Wherein, Z represent column vector z (n), z (n-1) ..., z (n-L+1) } ^t, A represents QMC coefficient vector [a ₁, a ₂, a ₃, a ₄, a ₅, a ₆] ^t, the matrix U=Re{Re (Y) of U corresponding to coefficient A, Im (Y), ones (L, 1), jRe (Y), jIm (Y), jones (L, 1) }, wherein Re represents the real part got in complex operation, Im represents the imaginary part of getting in complex operation, and the ones element represented in matrix is all the imaginary unit that 1, j represents in plural number, adopt least-squares algorithm can separate the estimated value that linear equation obtains A, i.e. QMC coefficient.

Preferably, through quadrature modulation errors judgement after described S5, when error is less than 10 ^-4time negligible, error range can be determined according to system accuracies, if error is greater than 10 ^-4, then coefficient is passed to QMC equalizer and continues to correct, if error is less than 10 ^-4then stop correcting.

(3) beneficial effect

As can be seen from technique scheme, the Image interference that the self adaptation direct orthogonalization Frequency-variable Modulation error calibration method that the present invention proposes can suppress imbalance of amplitude and phase to cause, again can the simultaneously local oscillator leakage that causes of DC-offset correction, significant to the error correction of self adaptation direct orthogonalization Frequency-variable Modulation.

Accompanying drawing explanation

Fig. 1 shows the self adaptation direct orthogonalization Frequency-variable Modulation error calibration method theory diagram that the present invention proposes;

Fig. 2 shows the system block diagram of the specific embodiment of the invention;

Fig. 3 shows the QMC algorithm flow chart of the preferred embodiment of the present invention;

Fig. 4 shows the simulation of Signal Processing result figure of the specific embodiment of the invention.

Embodiment

For making the object, technical solutions and advantages of the present invention clearly understand, below in conjunction with embodiment also with reference to accompanying drawing, the present invention is described in more detail.Should be appreciated that, these describe just exemplary, and do not really want to limit the scope of the invention.In addition, in the following description, the description to known features and technology is eliminated, to avoid unnecessarily obscuring concept of the present invention.

Fig. 1 shows the self adaptation direct orthogonalization Frequency-variable Modulation error calibration method theory diagram that the present invention proposes.

As shown in Figure 1, in the self adaptation direct orthogonalization Frequency-variable Modulation error calibration method that the present invention proposes, first by QMC equalizer 100 after data input, then direct orthogonalization frequency conversion 101 is carried out, data transmission is carried out after modulating the signal to radio frequency, antenna-coupled 102 receiving feedback signals, down-conversion 103 mixing and if sampling 104 are carried out to feedback signal, the digital signal y (n) obtained is given FPGA to carry out demodulation and align 105 with numeral, carry out QMC parameter Estimation 106 afterwards, finally carry out quadrature modulation errors judgement 107, and determine whether transmit QMC equalizer coefficients according to court verdict.

Fig. 2 shows the system block diagram of the specific embodiment of the invention.

As shown in Figure 2, in the specific embodiment of the invention, by being the base band data of specific standard through coded modulation, inputing to FPGA, in FPGA, carrying out quadrature modulation errors correction.To QMC coefficient initialization 1, be set to [1,0,0,0,1,0], by base band data after QMC201 module, baseband signal does not change, still be divided into I and Q two paths of signals, then carry out digital-to-analogue conversion 202, convert digital baseband signal to analog signal, then quadrature modulation 203 is carried out, by modulates baseband signals to carrying in wave frequency, by amplifying, by antenna 205 emitting radio frequency signal through signal power after amplifier 204.This system has feedback control loop, feedback data is obtained by antenna-coupled, feedback signal carries out down-conversion Frequency mixing processing by frequency mixer 206, obtain intermediate-freuqncy signal, carry out analog-to-digital conversion 207 again, convert analog signal to digital signal, then filtering 208 and demodulation 209 is carried out, obtain signal y (n) containing error, now signal list is shown as I ' and Q ', Delay Estima-tion is carried out to signal y (n) after demodulation, calculate time of delay, and time delay registration process 210 is carried out to raw baseband data z (n), estimate that QMC coefficient estimates 211 with base band data z (n) after feedback signal y (n) and alignment, utilize the vector determination quadrature modulation errors 212 of z (n) and y (n), if there is error, QMC coefficient is passed to QMC module and corrects, if there is not error, then stop correcting.

Fig. 3 shows the QMC algorithm flow chart of the preferred embodiment of the present invention.

As shown in Figure 3, the QMC algorithm flow chart of the preferred embodiment of the present invention specifically comprises the steps:

S1:QMC coefficient initialization, for launching the signal without precorrection;

The span of S2: intercepted length is the base band data of L from the digital baseband signal of input, L is arranged according to signal bandwidth and sample rate, and general range is [512,8192];

S3: the data in S2 are carried out precorrection by QMC equalizer, offsets local oscillator leakage and the Image interference of radio frequency link;

S4: carry out digital-to-analogue conversion to corrected data, baseband digital signal is changed into analog signal, is beneficial to the high-speed transfer of signal;

S5: the data in S4 are carried out quadrature modulation;

S6: through antenna transmission radiofrequency signal;

S7: by antenna-coupled receiving feedback signals;

S8: the radiofrequency signal received is carried out down-conversion, and carries out intermediate frequency filtering, filter out-band external noise;

S9: if sampling is carried out to the signal of S8, continuous time signal is sampled into discrete time digital signal;

S10: carry out quadrature demodulation to digital signal in S9, obtains the I ' containing error and Q ' two paths of signals;

S11: Delay Estima-tion is carried out to signal y (n) after demodulation, calculates time of delay, and time delay registration process is carried out to raw baseband data z (n);

S12: estimate QMC coefficient with base band data z (n) after feedback signal y (n) and alignment, solve the linear equation of QMC coefficient such as formula shown in (10):

Z＝UA (10)

Wherein, Z represent column vector z (n), z (n-1) ... z (n-L+1) } ^t, A represents QMC coefficient vector [a ₁, a ₂, a ₃, a ₄, a ₅, a ₆] ^t.The matrix of U corresponding to coefficient A:

U＝Re{Re(Y)，Im(Y)，ones(L，1)，j·Re(Y)，j·Im(Y)，j·ones(L，1)}。

Adopt least-squares algorithm can separate the estimated value that linear equation (10) obtains A, i.e. QMC coefficient;

S13: the vector determination quadrature modulation errors utilizing z (n) and y (n); If there is error, proceed to correct, otherwise stop correcting;

S14: the QMC coefficient estimating in S13 to obtain is passed to QMC equalizer, performs S2, proceeds to correct.

Fig. 4 shows the simulation of Signal Processing result figure of the specific embodiment of the invention.

As shown in Figure 4, the signal of base band input is monolateral multiple multi-tone signal, carrier frequency is 40MHz, sample estimates length L=1024, radio frequency Output rusults as shown in Figure 4, the figure (a) on the left side is the RF signal amplitude-frequency response schematic diagram not adding QMC correction, the figure (b) on the right is the RF signal amplitude-frequency response schematic diagram adding QMC correction, can find out after adding that QMC corrects from the amplitude-frequency response schematic diagram of RF signal, essentially eliminate local oscillator leakage, the image disturbing signal simultaneously produced due to IQ imbalance obtains obvious suppression, and optimization amplitude is more than 40dB.

In sum, the present invention proposes a kind of self adaptation direct orthogonalization Frequency-variable Modulation error calibration method, the Image interference that the method can suppress imbalance of amplitude and phase to cause, again can the simultaneously local oscillator leakage that causes of DC-offset correction, significant to the error correction of self adaptation direct orthogonalization Frequency-variable Modulation.

Should be understood that, above-mentioned embodiment of the present invention only for exemplary illustration or explain principle of the present invention, and is not construed as limiting the invention.Therefore, any amendment made when without departing from the spirit and scope of the present invention, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.In addition, claims of the present invention be intended to contain fall into claims scope and border or this scope and border equivalents in whole change and modification.

Claims (6)

1. a self adaptation direct orthogonalization Frequency-variable Modulation error calibration method, is characterized in that, described method comprises the steps:

S1:QMC coefficient initialization, and from the digital baseband signal of input, intercepted length is the base band data of L, is corrected by described base band data by QMC equalizer;

S2: digital-to-analogue conversion is carried out to the data after correcting, transmits after quadrature modulation;

S3: by antenna-coupled receiving feedback signals, carries out down-conversion mixing and if sampling to feedback signal, the digital medium-frequency signal obtained is sent into FPGA and carries out quadrature demodulation;

S4: Delay Estima-tion is carried out to signal y (n) after demodulation, calculates time of delay, and carry out time delay registration process with raw baseband data z (n);

S5: estimate QMC coefficient.

2. self adaptation direct orthogonalization Frequency-variable Modulation error calibration method according to claim 1, is characterized in that: in described S1, the span of L is determined according to signal bandwidth and sample rate.

3. self adaptation direct orthogonalization Frequency-variable Modulation error calibration method according to claim 1, is characterized in that: the error model of described quadrature modulation is

I′＝a ₁I+a ₂Q+a ₃

Q′＝a ₄Q+a ₅Q+a ₆

Wherein, a ₁, a ₂, a ₃, a ₄, a ₅, a ₆represent QMC coefficient, described model directly describes the uneven gain of I/Q signal in transmitting procedure and the direct current biasing of superposition, and the distortion that indirectly unbalance in phase of I/Q signal causes with the formal description of IQ crosstalk, according to this error model, quadrature modulation errors can be obtained against distortion factor a _i.

4. self adaptation direct orthogonalization Frequency-variable Modulation error calibration method according to claim 1, it is characterized in that: in S3, pass through antenna-coupled, and obtain feedback signal by intermediate frequency high-speed sampling, to feedback signal with transmit and carry out registration process after, estimate that quadrature modulation errors is against distortion factor.

5. self adaptation direct orthogonalization Frequency-variable Modulation error calibration method according to claim 1, is characterized in that: the linear equation of the estimation QMC coefficient in described S5 is as follows:

Z＝UA

Wherein, Z represent column vector z (n), z (n-1) ... z (n-L+1) ^t, A represents QMC coefficient vector [a ₁, a ₂, a ₃, a ₄, a ₅, a ₆] ^t, the matrix U=Re{Re (Y) of U corresponding to coefficient A, Im (Y), ones (L, 1), jRe (Y), jIm (Y), jones (L, 1) }, wherein Re represents the real part got in complex operation, and Im represents the imaginary part of getting in complex operation, and the ones element represented in matrix is all 1, j represents the imaginary unit in plural number, adopt least-squares algorithm to solve linear equation, obtain the estimated value of A, i.e. QMC coefficient.

6. self adaptation direct orthogonalization Frequency-variable Modulation error calibration method according to claim 1, is characterized in that: carry out quadrature modulation errors judgement after described S5, when error is less than 10 ^-4in time, ignores, if error is greater than 10 ^-4, then coefficient is passed to QMC equalizer and continues to correct, if error is less than 10 ^-4then stop correcting.