Summary of the invention
The object of the present invention is to provide a kind of baseband signal preprocess method,, obtain the relevant parameter factor of distortion, bring among the QMC baseband signal is carried out preliminary treatment, reach the purpose of correcting distortion signal by the QM modulation.
Purpose of the present invention is achieved through the following technical solutions:
A kind of baseband signal preprocess method of the present invention may further comprise the steps:
(a) on FPGA, pass through logical algorithm, set up the computing circuit that receives the down-conversion output signal;
(b) the down-conversion output signal is decomposed into in-phase component I and quadrature component Q;
(c) by computing circuit in-phase component I and quadrature component Q are handled, eliminate the amplitude phase error and the direct current biasing of analog channel;
(d) output signal.
Further, the computing circuit set up of described step (a) comprises QM quadrature modulation circuit and QMC quadrature modulation correcting circuit.
Further, described step (c) may further comprise the steps:
(c1) at the idle time slot of system or the appointed interval of agreement, the training sequence during software trigger FPGA transmission is built-in, training sequence length is:
Consider that according to the GSM standard FPGA guarantees to send training sequence at the GP time slot, with the monotony frequency fc that determines to send, 1024 sampled points of sampling, 1024 sampled points are in the cycle of integral multiple that with 2 π is the cycle;
(c2) FPGA catches the training sequence through distortion that feeds back
, catch length and be
FPGA catches the training sequence S(t of the distortion that feeds back), software is sampled to it, and its sampling number is the sampling number in arbitrary cycle in 1024 sampled points;
(c3) to this distortion sequence of catching
, carry out following processing:
In the formula
Be the distortion IQ modulation I signal of base band in the quadrature modulation QM model,
Be the distortion IQ modulation Q signal of base band in the quadrature modulation QM model, the sampled point that extracts carried out the 1/4Fs mixing, extracts 2 times after, the baseband I Q signal of acquisition distortion, low-pass filtering is mainly used in and leaches image signal herein;
(c4) sampled signal is carried out Fourier transform, the direct current biasing in the picked up signal:
In the formula
DC deviation for I passage in the quadrature modulation QM model;
Be the DC deviation of Q passage in the quadrature modulation QM model, j is the imaginary symbols in the plural number,
For reducing measure error, select the length of sampling
, make
M is any natural number in the formula,
Pi=π;
(c5) from
With
Eliminate direct current biasing in the signal, obtain
With
,
In the formula
Be I, the normalization of Q channel amplitude error in the quadrature modulation QM model,
For supposition in the quadrature modulation QM model produces additional phase error;
(c6) right
Carry out Fourier transform:
Try to achieve direct current biasing according to following formula;
(c7) right
Signal carries out energy integral
(c8) difference following relationship
Try to achieve respectively
,
(c9) according to obtaining
,
,
,
,
According to above relational expression, try to achieve the calibration model parameter respectively, in the formula
DC deviation for I passage among the quadrature modulation calibration model QMC;
Be the DC deviation of Q passage among the quadrature modulation calibration model QMC,
Be I, the normalization of Q channel amplitude error among the quadrature modulation calibration model QMC,
For supposition among the quadrature modulation calibration model QMC produces additional phase error.
Further, described step (d) is following calculation procedure:
In the formula
With
Be baseband I signal and Q signal through exporting after the preliminary treatment, in the QMC model of FPGA, configuration multiplier and adder on FPGA are finished QMC and are proofreaied and correct, the output baseband signal with the calibration model parameter configuration of trying to achieve
With
The present invention compared with prior art has following advantage and beneficial effect:
A kind of baseband signal preprocess method of 1 the present invention, in each different working carrier frequency, QMC adjusts parameter and all is suitable for, and under different environment such as working temperature, QMC all can carry out in real time;
A kind of baseband signal preprocess method of 2 the present invention, quadrature modulation emission local-oscillator leakage and quadrature modulation transmitting mirror picture leak and minimize;
A kind of baseband signal preprocess method of 3 the present invention, the not obvious operate as normal process that influences of QMC trimming process.
Embodiment
The present invention is described in further detail below in conjunction with embodiment, but embodiments of the present invention are not limited thereto.
Embodiment
A kind of baseband signal preprocess method of the present invention may further comprise the steps:
(a) on FPGA, pass through logical algorithm, set up the QM quadrature modulation circuit and the QMC quadrature modulation correcting circuit that receive the down-conversion output signal.The FPGA type selecting can be considered according to logic scale and device cost, recommends Statix II GX and the Statix III GX series of ALTERA, perhaps the Virtex-4 SX of XILINX and Virtex-5 SX series;
(b) the down-conversion output signal is decomposed into in-phase component I and quadrature component Q;
(c1) at the idle time slot of system or the appointed interval of agreement, the training sequence during software trigger FPGA transmission is built-in, training sequence length is:
Consider that according to the GSM standard FPGA guarantees to send training sequence at the GP time slot, with the monotony frequency fc that determines to send, 1024 sampled points of sampling, 1024 sampled points are in the cycle of integral multiple that with 2 π is the cycle;
(c2) FPGA catches the training sequence through distortion that feeds back
, catch length and be
FPGA catches the training sequence S(t of the distortion that feeds back), software is sampled to it, and its sampling number is the sampling number in arbitrary cycle in 1024 sampled points;
(c3) to this distortion sequence of catching
, carry out following processing:
In the formula
Be the distortion IQ modulation I signal of base band in the quadrature modulation QM model,
Be the distortion IQ modulation Q signal of base band in the quadrature modulation QM model, the sampled point that extracts carried out the 1/4Fs mixing, extracts 2 times after, the baseband I Q signal of acquisition distortion, low-pass filtering is mainly used in and leaches image signal herein;
(c4) sampled signal is carried out Fourier transform, the direct current biasing in the picked up signal:
In the formula
DC deviation for I passage in the quadrature modulation QM model;
Be the DC deviation of Q passage in the quadrature modulation QM model, j is the imaginary symbols in the plural number, for reducing measure error, select the length of sampling
, make
M is any natural number in the formula,
Pi=π;
(c5) from
With
Eliminate direct current biasing in the signal, obtain
With
,
In the formula
Be I, the normalization of Q channel amplitude error in the quadrature modulation QM model,
For supposition in the quadrature modulation QM model produces additional phase error;
(c6) right
Carry out Fourier transform:
Try to achieve direct current biasing according to following formula;
(c7) right
Signal carries out energy integral
(c8) difference following relationship
Try to achieve respectively
,
(c9) according to obtaining
,
,
,
,
In the formula
DC deviation for I passage among the quadrature modulation calibration model QMC;
DC deviation for Q passage among the quadrature modulation calibration model QMC.
Be I, the normalization of Q channel amplitude error among the quadrature modulation calibration model QMC,
For supposition among the quadrature modulation calibration model QMC produces additional phase error.,, try to achieve the calibration model parameter respectively according to above relational expression;
(d) according to following calculation procedure:
In the formula
With
Be baseband I signal and Q signal through exporting after the preliminary treatment;
In the QMC model of FPGA, configuration multiplier and adder on FPGA are finished QMC and are proofreaied and correct, the output baseband signal with the calibration model parameter configuration of trying to achieve
With
Operation principle of the present invention is as follows:
The DC deviation of supposing the I passage is
The DC deviation of Q passage is
, simultaneously I, Q channel amplitude error are normalized to
, produce additional phase error thus and be assumed to
, then after considering the mismatch affects factor, the real output signal of the passage of QM modulation can be expressed as:
Wherein:
Its matrix form is:
In the formula
Baseband signal for final actual output.
The target of QMC calibration promptly is to obtain by estimation
,
,
,
To base-band input signal
With
Carry out preliminary treatment, to eliminate the amplitude phase error and the direct current biasing of analog channel.The QMC calibrating patterns as shown in the formula:
Therefore, then bring QMC precorrection model into the QM modulation pattern, can see:
Can see and working as
The promptly final actual baseband signal of exporting.
As mentioned above, just can realize the present invention well.