Background technology
In recent years, rapidly anti-war along with wireless communication technology, the wireless communication system product is more and more universal, becomes the important component part of mankind nowadays development of information.Radio-frequency transmitter is positioned at wireless communication system foremost, and its structure and performance directly affect whole communication system.Optimal design structure and the suitable manufacturing process of selection to improve the ratio of performance to price of system, are the directions that Radio Frequency Engineer is pursued.
Zero intermediate frequency reciver is not because of needing the outer high Q value band pass filter of sheet, can realizing the integrated people's of being subjected to of monolithic extensive attention.Particularly, see also the structured flowchart of zero intermediate frequency reciver shown in Figure 1.The structure of zero intermediate frequency reciver is simply a lot of than superheterodyne receiver.Its course of work is as follows: the radiofrequency signal that receives is after band pass filter (RF BPF) and low noise amplifier (LNA) amplification, with the two-way local oscillation signal mixing of quadrature each other, produce homophase and quadrature two-way baseband signal respectively, because the local oscillation signal frequency is identical with radio frequency signal frequency, so directly produce baseband signal after the mixing.Homophase and quadrature two-way baseband signal are passed through each self-corresponding low pass filter (LPF) and variable gain amplifier (VGA) respectively, and channel is selected and function is adjusted in gain to finish.This shows, the most attractive part of zero intermediate frequency reciver do not need in the downconversion process to be through intermediate frequency, and image frequency is exactly a radiofrequency signal itself, do not exist image frequency to disturb, can save image-reject filter essential in the former super-heterodyne architecture and intermediate-frequency filter all can omit, like this, cancelled outer member on the one hand, the monolithic that helps system is inherited, reduce cost, on the other hand, the minimizing of counting of circuit module that system is required and outside segments, reduce the power consumption of receiver, and reduced the chance that radiofrequency signal is subjected to external disturbance.
But also there is defective in zero intermediate frequency reciver itself, and for example DC deviation, local oscillator are revealed and IQ is unbalance etc.
Particularly, reveal for local oscillator, as shown in Figure 2, the local frequency of zero-if architecture is identical with signal frequency, if the isolation performance between the local oscillator mouth of frequency mixer and the radio frequency mouth is bad, local oscillation signal just is easy to the radio frequency mouth output from frequency mixer, leaks into antenna by low noise amplifier again, be radiated the space, form interference neighboring trace.
For DC deviation, DC deviation is the distinctive a kind of interference of zero intermediate frequency scheme, it is caused by self-mixing (Self-Mixing), the local oscillation signal that leaks can reflect from the output of LNA, the output and the antenna end of filter respectively, or the signal that leaks receives by antenna, enters the radio frequency mouth of frequency mixer.It and the local oscillation signal mixing mutually that the local oscillator mouth enters, difference frequency is zero, is direct current.Equally, the high reject signal that enters LNA also can be owing to the bad local oscillator mouth that bleeds of each port isolation performance of frequency mixer, next with the radio frequency mouth conversely strong jamming mixing mutually, and difference frequency is a direct current.These direct current signals will be superimposed upon on the baseband signal, and baseband signal is constituted interference, be called as DC deviation.The DC deviation often noise than radio-frequency front-end is also big, makes the signal to noise ratio variation, and big DC deviation of while may make the amplifiers at different levels behind the frequency mixer saturated, can't amplify useful signal.When self-mixing changed in time, it is very complicated that the DC deviation problem will become.This situation can take place under the condition below: receive when the local oscillation signal that leaks into antenna reflects by antenna from the object that moves again after antenna is launched, enter frequency mixer by LNA, become when the DC deviation of mixing generation will be.
For the I/Q mismatch, when adopting the zero intermediate frequency scheme to carry out digital communication, if homophase and quadrature two branch roads are inconsistent, the gain difference of frequency mixer for example, two local oscillation signal phase differences are not strict 90 °, can cause the variation of baseband I/Q signal, promptly produce the I/Q mismatch problems.The performance of I/Q mismatch on base band also is a direct current biasing to occur.Major obstacle when the I/Q mismatch problems is digital Design in the past along with the raising of integrated level, though the I/Q mismatch has obtained corresponding improvement, must cause enough attention during design.
At DC deviation, method commonly used in the prior art is AC coupled (AC Coupling).Baseband signal after the down-conversion is coupled to baseband amplifier with the method for electric capacity stopping direct current, eliminates the interference of DC deviation with this.For having concentrated the relatively baseband signal of macro-energy near the direct current, this method can increase the error rate, should not adopt.Therefore the effective ways that reduce the DC deviation interference are that the baseband signal of desire emission is carried out suitable coding and selected suitable modulation system, to reduce near the energy of baseband signal direct current.Can eliminate DC deviation and not lose dc energy this moment with the method for AC coupled.Shortcoming is to use big electric capacity, has increased area of chip.But this method is significant discomfort usefulness in the debugging of spread spectrums such as WCDMA, and in the spread spectrum debugging in broadband, zero-frequency and near zone thereof are occupied by useful signal.Also have some based on the method for calibrating signal independently, but these methods are had to interrupt the transmission of useful signal and are come the non-equilibrium of IQ calibrated.Such as using special signal source input receiver, be used to detect the unbalance degree of IQ, compensate again, so just can not receive the signal of mobile phone normally.
Summary of the invention
The object of the present invention is to provide a kind of carrier adaptive filtering method and system and zero intermediate frequency reciver of zero intermediate frequency, can effectively suppress DC component and I/Q mismatch, and do not need useful signal is interrupted, and use in can in WCDMA downstream signal or CDMA downstream signal receiver system, using.
For achieving the above object, the invention provides a kind of carrier adaptive filtering method of zero intermediate frequency, comprise: to carrying out despreading through the I signal of a FIR filter filtering with through the Q signal of the 2nd FIR filter filtering, utilize the EVM method to calculate I signal after the despreading and Q signal then in the coordinate of planisphere and the error between the both ideal constellation coordinate, adjust the coefficient of the filter of a described FIR filter and the 2nd FIR according to described error, the difference of the coefficient of the filter of a FIR filter and the 2nd FIR is all less than predetermined threshold value before and after adjusting.
On the other hand, the embodiment of the invention also provides a kind of carrier adaptive filtering system of zero intermediate frequency, comprising: a FIR filter, the 2nd FIR filter, despread unit, EVM computing unit, coefficient adjustment unit;
Wherein, a described FIR filter is used for according to filter coefficient I signal being filtered;
Described the 2nd FIR filter is used for according to filter coefficient Q signal being filtered;
Described despread unit is used for I signal and Q signal through a described FIR filter and the filtration of the 2nd FIR filter are carried out despreading;
Described EVM computing unit is used to utilize I signal and the error of Q signal between coordinate on the planisphere and both ideal constellation coordinate after the EVM method is calculated described despreading;
Described coefficient adjustment unit is used for adjusting according to described error the coefficient of a described FIR filter and the 2nd FIR filter;
Comparing unit, the difference of the coefficient of the filter of a FIR filter and the 2nd FIR is all less than predetermined value, if then trigger the adjustment that the coefficient adjustment unit stops filter coefficient before and after being used for relatively adjusting; If not, then trigger the coefficient adjustment unit and continue to adjust filter coefficient.
On the one hand, the embodiment of the invention also provides a kind of zero intermediate frequency reciver, comprises the carrier adaptive filtering system of above-mentioned zero intermediate frequency again.
Pass through the embodiment of the invention, can effectively suppress the DC component in I signal and the Q signal, and can suppress the I/Q mismatch, the embodiment of the invention is stated in the process of inhibition in realization, do not need to interrupt the reception of useful signal, adopt adaptive working mechanism, and can be applicable in the band spectrum modulation.
Embodiment
For the purpose, technical scheme and the advantage that make the embodiment of the invention clearer, below in conjunction with the accompanying drawing in the embodiment of the invention, technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.
The embodiment of the invention provides a kind of carrier adaptive filtering method of zero intermediate frequency, and as shown in Figure 3, this method comprises:
Step S301: to carrying out despreading through the I signal of a FIR (finite impulse response filter) filter filtering with through the Q signal of the 2nd FIR filter filtering.
Here, I and Q signal are that original I signal in the zero intermediate frequency reciver and Q signal are carried out the digital signal that demodulation also obtains through analog-to-digital conversion.DC component and unbalanced component have been carried in the I signal of digital form and the Q signal.
Step S302: utilize the EVM method to calculate I signal after the despreading and Q signal in the coordinate of planisphere and the error between the both ideal constellation coordinate.
Step S303: adjust the filter coefficient of a described FIR filter and the 2nd FIR filter according to described error, the difference of the FIR filter before and after adjusting and the filter coefficient of the 2nd FIR filter is less than predetermined value.
The purpose of adjusting is, eliminates the deterioration of error.Calculating adjusted filter coefficient can be based on LMS (Least Mean Square, lowest mean square root adaptive approach) algorithm.This algorithm is realized easily, can reduce cost.
The difference of the FIR filter before and after adjusting and the filter coefficient of the 2nd FIR filter is during less than predetermined value, the error of calculating among the step S302 will be very little, can think that a FIR filter and the 2nd FIR filter have carried out sufficient inhibition to DC component in I signal and the Q signal and unbalanced component.
Below describe the specific implementation of the carrier adaptive filtering method of zero intermediate frequency provided by the invention in detail with an embodiment.
Embodiment one
Present embodiment one provides a kind of carrier adaptive filtering method of zero intermediate frequency, in the present embodiment, the one FIR filter is used to filter I signal, the 2nd FIR filter is used to filter Q signal, and two FIR filters all are high pass filters, and the filter coefficient of two filters all is adjustable.
As shown in Figure 4, this method comprises:
Step S401: the filter coefficient of initialization the one FIR filter and the 2nd FIR filter.
Above-mentioned initialization can be that filter coefficient is reset to predefined fixed value, also can be to be set to be the last filter coefficient value of adjusting in the last task.
Step S402: I signal and Q signal to a FIR filter and the 2nd FIR filter filtration output carry out despreading;
Step S403: calculate the I signal after the step S402 despreading and the constellation coordinate of Q signal and the distance between the both ideal constellation coordinate in planisphere, the distance that aforementioned calculation is obtained is as error.
At numerical portion, I signal after the despreading and Q signal be combined to form so-called planisphere.See also Fig. 5, the ideal position that I signal in zero intermediate frequency reciver and Q signal are combined in planisphere should be 4 points, and the coordinate of these four positions (being called the both ideal constellation coordinate) is respectively (1,1), (1 ,-1), (1 ,-1) and (1,1).But in practice, as shown in Figure 5, because DC component and the unbalance coordinate of I signal and Q signal that causes of IQ depart from above-mentioned both ideal constellation coordinate.The constellation coordinate that calculates I signal after the despreading and Q signal and the distance between the both ideal constellation coordinate can be by horizontal range square and vertical range square summations, getting root again obtains, for example, if I signal after the despreading and Q signal are respectively (1.1,1.5), then the distance of the constellation coordinate of I signal after the despreading and Q signal and both ideal constellation coordinate (1,1) is [(1.1-1)+(1.5-1)
2]
1/2Here the error that the distance that calculates promptly utilizes the EVM method to calculate.
The purpose of the method that present embodiment provides is to adjust gradually the filter coefficient of FIR filter, so that the distance value of aforementioned calculation reduces gradually, also promptly eliminate above-mentioned error, make the constellation coordinate of I signal after the despreading and Q signal level off to desirable constellation coordinate as far as possible.
Step S404: the error that step S403 is calculated, the current I signal after filtering through a FIR filter and approach product between the coefficient and the current filter coefficient of a FIR filter and, as the adjusted filter coefficient of a FIR filter; Current Q signal after filtering with the error calculated among the step S403, through the 2nd FIR filter and approach product between the coefficient and the current filter coefficient of the 2nd FIR filter and, as the adjusted filter coefficient of the 2nd FIR filter.
The iteration that said process is actually in the iterative process realizes, if the FIR filter before will adjusting and the filter coefficient of the 2nd FIR filter are designated as wi (n-1) and wq (n-1) respectively, the adjusted FIR filter that calculates and the filter coefficient of the 2nd FIR filter are designated as wi (n) and wq (n) respectively, the error that this iterative process step S403 is calculated is designated as e (n), the I signal and the Q signal that filter through the FIR filter in this iterative process are designated as I and Q respectively, then calculate adjusted coefficients w i (n) and wq (n) and can represent with following two formulas:
Wi (n)=wi (n-1)+Mu*e (n) * I (formula 1)
Wq (n)=wq (n-1)+Mu*e (n) * Q (formula 2)
Wherein, Mu is for approaching coefficient, and span is (0,1), i.e. 0<Mu<1.
Step S405: adjust the filter coefficient of a FIR filter and the filter coefficient of the 2nd FIR filter respectively according to the filter coefficient of an adjusted FIR filter that calculates among the step S404 and the filter coefficient of the 2nd FIR filter.
Step S406: before and after relatively adjusting the difference of the filter coefficient of a FIR filter and the 2nd FIR filter all less than predetermined threshold value, if, execution in step S407 then: finishing iteration, the fixedly coefficient of FIR filter; If not, then return continuation execution in step S402.
Each iterative process all by step S406 relatively before and after the filter coefficient that obtains of twice iteration whether less than predetermined threshold value, like this, if less than predetermined threshold value, then the error that calculates among the step S403 also just can be thought fully little, thereby can think and adjust after the filter coefficient that the DC component of I signal and Q signal and IQ are unbalance to have obtained sufficient inhibition.
The method that provides by present embodiment one, effectively suppress the DC component in I signal and the Q signal, and can suppress the I/Q mismatch, this method is stated in the process of inhibition in realization, do not need to interrupt the reception of useful signal, adopt adaptive working mechanism, and can be applicable in the spread spectrum debugging.
Embodiment two
The embodiment of the invention two corresponding carrier adaptive filtering systems that a kind of zero intermediate frequency is provided, as shown in Figure 6, this system comprises: a FIR filter 601, the 2nd FIR filter 602, despread unit 603, EVM computing unit 604, coefficient adjustment unit 605 and comparing unit 606;
Wherein, a FIR filter 601 is used for according to filter coefficient I signal being filtered; The 2nd FIR filter 602 is used for according to filter coefficient Q signal being filtered; The one FIR filter 601 and the 2nd FIR filter 602 all are high pass filters.
Despread unit 603 is used for I signal and Q signal through a FIR filter 601 and 602 filtrations of the 2nd FIR filter are carried out despreading;
EVM computing unit 604 is used to utilize I signal and the error of Q signal between coordinate on the planisphere and both ideal constellation coordinate after the EVM method is calculated above-mentioned despreading; The error here is the I signal after the despreading and the constellation coordinate of Q signal and the gap between the both ideal constellation coordinate in planisphere.
Coefficient adjustment unit 605 is used for adjusting according to described error the coefficient of a described FIR filter and the 2nd FIR filter;
Comparing unit 606, the difference of the filter coefficient of the filter 602 of a FIR filter 601 and the 2nd FIR is all less than predetermined value, if then trigger the adjustment that the coefficient adjustment unit stops filter coefficient before and after being used for relatively adjusting; If not, then trigger the coefficient adjustment unit and continue to adjust filter coefficient.
Wherein, as shown in Figure 7, coefficient adjustment unit 605 can comprise:
Multiplying subelement 6051 is used to calculate that described EVM computing unit 604 obtains error, the current I signal after filtering through a FIR filter 601 and approach product between the coefficient; The described EVM computing unit 604 that also is used to calculate obtains error, the current Q signal after filtering through the 2nd FIR filter 602 and approach product between the coefficient;
Add operation subelement 6052, be used for calculate described error, current I signal after filtering through a FIR filter 601 and approach product between the coefficient and the current filter coefficient of a FIR filter 601 and, obtain the filter coefficient of an adjusted FIR filter 601; Also be used to calculate described error, the current Q signal after filtering through the 2nd FIR filter 602 and approach product between the coefficient and the current filter coefficient of the 2nd FIR filter 602 and, obtain the filter coefficient of adjusted the 2nd FIR filter 602;
Coefficient adjustment trigger element 6053 is used to trigger the filter coefficient that a FIR filter 601 is adjusted into its filter coefficient in a described adjusted FIR filter 601 that calculates; Also be used to trigger the filter coefficient that the 2nd FIR filter 602 is adjusted into its filter coefficient in described adjusted the 2nd FIR filter 602 that calculates.
Above-mentioned span of approaching coefficient can be (0,1), i.e. 0 to 1 open interval.
The system that provides by the embodiment of the invention, can effectively suppress the DC component in I signal and the Q signal, and can suppress the I/Q mismatch, native system is stated in the process of inhibition in realization, do not need to interrupt the reception of useful signal, adopt adaptive working mechanism, and can be applicable in the spread spectrum debugging.
Embodiment three
Present embodiment is corresponding to provide a kind of zero intermediate frequency reciver, this zero intermediate frequency reciver to comprise the carrier adaptive filtering system of the zero intermediate frequency of describing in the foregoing description two.In practice, the carrier adaptive filtering system of zero intermediate frequency can realize on the DSP in zero intermediate frequency reciver.
The above only is a preferred implementation of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.