CN103312372A - Frequency correcting method - Google Patents

Abstract

The invention relates to a frequency correcting method, which can be used for correcting the frequency of reference oscillation signals from the initial oscillation frequency to the correction oscillation frequency. The frequency correcting method comprises the following steps that a frequency scanning section region is divided into M scanning frequencies; receiving signals are subjected to frequency reduction according to the M scanning frequencies, and in addition, M receiving signals subjected to frequency reduction are obtained; the M receiving signals subjected to frequency reduction are respectively subjected to correlation operation, and M correlation operation results are obtained; N frequency groups are distinguished from the M scanning frequencies, wherein each frequency group comprises P selecting frequencies, and the P selecting frequencies correspond to P adjacent scanning frequencies; the N frequency groups are respectively subjected to group operation, and N group operation results are obtained; selected frequency groups are selected from the N frequency groups according to N group operation results, and in addition, the correction oscillation frequency is obtained from the selected frequency groups.

Description

Frequency calibrating method

Technical field

The present invention is relevant for a kind of frequency calibrating method, and particularly relevant for a kind of frequency calibrating method of oscillating reference signal.

Background technology

Wideband-CDMA access (Wideband Code Division Multiple Access is referred to as WCDMA) is a kind of digitized 3G (Third Generation) Moblie technology.In the WCDMA system, before transmission end (base station) sends data, need to use exhibition frequently to encode (Spreading Code) with narrow frequency signal generate broadband signal, see through again give after disturbing code (Scrambling Code) confuses receiving terminal wherein the unit of data be called bit (bit), the unit of sequence then is called chip (chip) after the exhibition frequently.

In order to make receiving terminal can reply the fundamental frequency signal that is transmitted by the transmission end, mobile phone must be finished synchronously with the base station first, otherwise the cause that receiving terminal and transmitting terminal will go on foot because of asynchronism(-nization), and obtains the output of noise shape waveform.The WCDMA system see through initial cell searching procedure (initial cell search procedure) reach mobile phone and base station the code territory and temporal synchronously.

The initial cell search refers in the WCDMA system, all all are in the situation of the unknown when mobile phone is just started shooting, mobile phone need be finished the action that search the base station, find out the strongest base station of signal, first so that mobile phone and base station clock pulse reach synchronously, find out again the employed disturbing code in base station, just can be correct and base station communication, and solve original fundamental frequency signal.For the ease of identification, with f _bRepresent the frequency of fundamental frequency signal, and with f _cRepresent the frequency of carrier signal (Carrier).

Usually all some errors can be arranged at mobile phone or the employed oscillator of base station end (oscillator), cause existing carrier frequency shift (carrier frequency offset is referred to as CFO) between both frequency, be called again frequency error f _ErrorIn order to make mobile phone can normally receive signal, need to adjust the frequency f of the oscillating reference signal that the local side oscillator of mobile phone produces _Ref, make the frequency f of oscillating reference signal _RefCan be more near the frequency f of carrier signal _c

When the WCDMA system carries out initialization cell searching program, in case the absolute wireless frequency number of channel (absolute radio-frequency channel number, referred to as ARFCN) be selected after, next step is that the frequency of oscillating reference signal that the local side oscillator is produced is revised.The permissible frequency error range of WCDMA system is approximately 3-13ppm, namely when the frequency of carrier signal is 2GHz, and the frequency f of oscillating reference signal _RefFrequency f with carrier signal _cBetween have the frequency error range of 6-26kHz.

In the initial cell searching procedure, how to allow the oscillating reference signal of mobile phone terminal and the carrier signal of base station reach Frequency Synchronization, to receive signal again and do suitable correction to produce correct receive frequency, be to allow the important step that the WCDMA system can normal operation.

Common technology can see through first a rough automatic frequency regulator (Coarse AFC) and carry out the acquisition of initialization frequency when carrying out the initial cell search.Revise the frequency f of the oscillating reference signal that the local side oscillator produces according to the result of initialization frequency acquisition _Ref, make the frequency f of oscillating reference signal _RefFrequency f with carrier signal _cBetween frequency error f _Error=f _c-f _Ref, can be adjusted within the scope of positive and negative 3ppm.

According to the planning of WCDMA system, the length of a code frame is 10ms, is comprising 15 time slots in each yard frame, comprises 2560 subcodes in each time slot.In order to assist mobile phone to reach the purpose of cell searching, the WCDMA system provides main synchronizing channel (Primary Synchronization Channel is referred to as PSCH) to assist mobile phone to finish slot synchronization.The length of PSCH passage only has 1/10 of original slot length, and the time slot of other positions does not then comprise any information or data.Therefore, only comprise one group through the main synchronizing sequence (primary synchronizationsequence) of design in incipient 256 subcodes of each time slot.

Because all base stations all use identical PSCH sequence, and mobile phone terminal also prestores the PSCH sequence, and therefore, whether mobile phone receives the PSCH sequence accurately just becomes receiving terminal and use the method for orienting boundary of time slot (slotboundary).Further, the main synchronizing sequence of receiving terminal utilization carries out correlation calculations (correlation calculation), and comes determination frequency error f according to the result of correlation calculations _Error

Frequency error f between oscillating reference signal and carrier signal _ErrorLarger, also just less by the numerical value of PSCH correlator (correlator) output (output).Therefore, the size of exporting according to the PSCH correlator and the practice of determination frequency error often is used in the initialization frequency acquisition in the WCDMA system.

See also Figure 1A, it is that signal in the WCDMA system transmits with the signal of receiving terminal and processes schematic diagram.The accompanying drawing left side represents the fundamental frequency signal of base station and the carrier signal that basestation oscillator 106 produces.

Fundamental frequency signal sees through frequency mixer 102 and after utilizing carrier signal to increase frequently, will produce and transmit signal.Frequency mixer 102 can be considered as two signals are multiplied each other herein, therefore, the frequency f of fundamental frequency signal _bThrough behind the frequency mixer 102, just producing frequency at the output of frequency mixer 102 is f _b± f _cThe smear signal.The smear signal is selected f via filter (not illustrating) _b+ f _cOr f _b-f _cTransmit.After the antenna of base station will transmit signal and see through see through mobile communication network 10 and transmit, again by the antenna reception of mobile phone terminal.

After receiving signal and being received by mobile phone, the frequency mixer 101 of mobile phone terminal can utilize the oscillating reference signal that local side oscillator 105 produces and it is carried out the frequency reducing operation.

In theory, the frequency f of oscillating reference signal _RefFrequency f with carrier signal _cCan be equal to each other, thereby answer obtains fundamental frequency signal in the certainly reception signal that can be complete.But the various variations when in fact not so, the difference between receiving terminal and transmission end is attributable to transport process and signal processing.

For instance, in transport process, the fundamental frequency signal x (n) that is sent by the base station may be subject to the interference (N) of noise or signal.Suppose that fundamental frequency signal is x (n), and 101 solutions of the frequency mixer of mobile phone terminal signal out is y (n), can be equivalent to the combination of fundamental frequency signal x (n) and noise N on the signal theory of frequency mixer 101 outputs this moment.

The noise in transport process, the frequency f of oscillating reference signal _Ref, and frequency (the carrier frequency f of the carrier signal of basestation oscillator generation _c) can't be in full accord, so that there is error in both frequencies.Suppose the frequency f of the oscillating reference signal that the local side oscillator produces _RefBe initial oscillation f _Orig(f _Ref=f _Orig), and the frequency of carrier signal is f _c, this frequency error f then _Error(f _Error=f _Ref-f _c=f _Orig-f _c) will affect and separate the drawn y (n) of modulation.

In order to judge the frequency error f between oscillating reference signal and the carrier signal _Error, the practice of oscillating reference signal being carried out frequency correction can be divided into coarse adjustment and fine setting two classes.

Wherein, the order ground that frequency is carried out coarse adjustment is first to the frequency f of oscillating reference signal _RefDo preliminary correction, make frequency error f _ErrorCan be contracted to first in the scope of frequency error class interval Δ f.Coarse adjustment proofread and correct finish after, then again to the frequency f of oscillating reference signal _RefFinely tune correction.This sentences frequency rough and transfers as the object of discussing.

Owing to after the coarse adjustment of the oscillating reference signal of local side process frequency, need to finely tune again correction.Therefore, still may the occurrence frequency error excessive if during the poor effect of frequency coarse adjustment, the phenomenon that can't correctly calibrate.Therefore detected ratios (detection rate) represented follow-up when finely tuning, can success for further adjustments and proofread and correct the frequency error f that sends as an envoy to _ErrorIt is 0 ratio.

The practice that common technology carries out the frequency coarse adjustment is, in the frequency scanning section, with different scanning frequency (sweep frequency) f _iTest.Scanning frequency f wherein _iI the scanning frequency of representative in the frequency scanning section.

Each scanning frequency f in the frequency scanning section _iOne default frequency error class interval Δ f is arranged to each other, according to each the scanning frequency f in the frequency scanning section _iCan draw correlativity calculation result y _iAfterwards, recycling correlativity calculation result y _iComparison, and draw the maximum y of correlativity calculation result _Max

Further, as correlativity calculation result y _iDuring for maximum, the scanning frequency f corresponding with it _iTo make frequency error f _ErrorThe minimum value that the frequency rough timing that levels off to can reduce.

Further, with each correlativity calculation result y _iNumerical value compare after, draw first maximum y wherein _Max, just can draw accordingly and revise frequency of oscillation (f _Ref=f _i).

Below utilize formula its practice of deriving, at first, receive signal r (t) with the representative of the 1st formula.

R (t)=α (t) s (t-t _b) exp (j2 π f _cT)+n (t), (the 1st formula)

The 1st formula utilizes s (t) to represent main synchronizing sequence (primary synchronization sequence), and it is 256 chips in front in the time slot (slot).α (t) represents Rayleigh fading (Rayleigh fading), supposes for simplicity its numerical value and frequency change have nothing to do (frequency-flat) herein.t _bThe timing off-set (timingoffset) of representative between system sequence (system timing) and aerial boundary of time slot (air slot boundary); f _cFrequency for carrier signal; And n (t) represents the summation of noise and other influences (interferences).

Then, utilize the frequency f of oscillating reference signal through frequency mixer 101 _RefR (t) carries out frequency reducing to received signal.It is noted that the frequency f of oscillating reference signal _RefCan be according to scanning frequency f _iDifference and change (f _Ref=f _i).

Further, the frequency f of oscillating reference signal _RefBe initial oscillation f at the beginning _Orig, and initial oscillation f _OrigAnd be not equal to the frequency f of carrier signal _cFor such problem, the use of the PSCH correlator of just arranging in pairs or groups utilizes different scanning frequency f _iAnd test draws different correlativity calculation result.Along with scanning frequency f _iDifference, the collocation frequency mixer uses and the frequency of carrying out to received signal the oscillating reference signal of frequency reducing conversion (down conversion) also changes thereupon, this also so that frequency mixer 101 frequency reducing Output rusults to received signal change along with the difference of scanning frequency.

Then, because scanning frequency f _iFor known, different scanning frequency f _iThe corresponding correlativity calculation result that draws can learn too.The 2nd formula is with correlativity calculation result y _iRepresentative is at different sequential border option (slotboundary candidates) t _mIn, make the PSCH correlator be output as peaked situation.

$y_i=\underset{t_m}{\max }|{\&Integral;}_{t_i}^{t_i+T}r\left(t\right)\&times;s^{*}(t-t_m)\exp (-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA0000143858560000051.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}{f}_{i}t)\mathrm{dt}|$ (the 2nd formula)

T wherein _iBe and scanning frequency f _iThe zero-time (starting time) of correlation.By t _iCan find out, as scanning frequency f _iDuring change, the signal that is used for calculating correlation is not identical yet.T represents (integrationperiod) between integration period, namely is equivalent to a WCDMA time slot.

Can learn according to the 2nd formula, be under the prerequisite of constant at Rayleigh fading α (t), as scanning frequency f _iNear the frequency f of carrier signal _cThe time, by the correlativity calculation result y of PSCH correlator output _iTo have maximum.Therefore, first to each scanning frequency f _iAll carry out first the PSCH correlation operation, afterwards again according to correlativity calculation result judge which scanning frequency can be near the frequency of carrier signal

Therefore, the frequency f of oscillating reference signal _Ref=f _iCan draw according to the 3rd formula.

$\widehat{f_i}=\underset{f_i}{\max }\left\{y_i\right\}$ (the 3rd formula)

In case draw comparatively the frequency f near carrier signal _cScanning frequency f _iAfter, common technology just can utilize the frequency f of this scanning frequency correction oscillating reference signal _RefThat is the oscillating reference signal that the oscillator 107 of local side produces will be by initial oscillation f _Ref=f _OrigProofread and correct and be aforementioned scanning frequency f corresponding to the maximum correlation result _Ref=f _i

That is to say, common technology is utilizing the result of calculation estimation initial oscillation f of PSCH correlator _OrigWith carrier frequency f _cBetween frequency error f _ErrorThe time, its practice is as follows:

First the frequency scanning section is divided into a plurality of scanning frequency f _iTrial is with different scanning frequency f _iFrequency f as the reference oscillator signal _Ref, and observe and adopt which scanning frequency f _iWith the correlativity calculation result y that the PSCH correlator is produced _iHas maximum.

Therefore, the practice of common technology is, if a certain scanning frequency f _iMake the correlativity calculation result y of PSCH correlator output _iDuring for maximum, just judge this scanning frequency f _iIn the frequency scanning section, comparatively near carrier frequency f _cFrequency.The frequency of the oscillating reference signal that just the local side oscillator is produced afterwards, changes this scanning frequency f into by initial oscillation _i(correction frequency of oscillation).

For instance, the minimum frequency in the frequency scanning section and peak frequency may be defined as the frequency f with initial oscillator signal _OrigDiffer-15.4ppm is to the scope of+15.4ppm.Wherein the size (step size) of each frequency scanning class interval Δ f is 1.4ppm.Therefore, in the frequency f with initial oscillator signal _OrigCentered by-15.4ppm is to the frequency scanning section scope of+15.4ppm, if with 1.4ppm during as the scope of a frequency scanning class interval Δ f, the frequency scanning section comprises (15.4*2)/1.4+1=23 scanning frequency altogether.

See also Figure 1B, it is the schematic diagram that common technology carries out frequency scanning side in proper order.For convenience of explanation, not complete 23 scanning frequencys of painting in the formula frequency scanning section herein.This sentences the different different scanning frequencys of numbering representative, and each scanning frequency is numbered according to same way as via the result that the PSCH correlator calculates.For example: the first correlativity calculation result y1 that draws according to the first scanning frequency f1, according to the second correlativity calculation result y2 that the second scanning frequency f2 draws, the rest may be inferred by analogy for it.

In simple terms, the practice of common technology is in the frequency scanning section, with different scanning frequency f _iScan.This kind is according to scanning frequency f _iChange and the mode of tested and checking one by one, the frequency coarse adjustment that is used as oscillating reference signal is used.

Yet, be that the mode of the coarse adjustment that compares of the correlation calculations to scanning frequency but may make detected ratios undesirable with the side that confirms one by one.That is to say, the mode of this kind coarse adjustment might not make frequency error be reduced to 0 during for follow-up the fine setting.

When adopting this kind mode to estimate correction frequency of oscillation near the frequency of carrier signal, the reason of correctly finding out frequency error in the time of possibly can't making fine setting is that a plurality of different scanning frequency f are arranged _iAll correspondence produces and has larger correlativity calculation result y _i, or maximum extreme value all has maximum correlativity calculation result corresponding to two scanning frequencys that differ greatly each other (for example: f1 and f23) simultaneously.Just be difficult for judging which scanning frequency f this moment _iJust this is used to the frequency f of calibration reference oscillator signal _RefIn addition, in when change, utilized the 3rd formula estimated frequency error f along with the time as Rayleigh fading α (t) _ErrorEffect unsatisfactory.

According to analog result even can find, when in the common technology mode oscillating reference signal being carried out coarse adjustment, the frequency correction result is dispersed.Under this kind situation, just be difficult to find out the value of suitable correction frequency of oscillation, therefore, common technology is in the frequency scanning section, after scanning for each scanning frequency, directly judge with correlativity calculation result and the practice of adjusting the frequency of oscillating reference signal still has disappearance.

Summary of the invention

The present invention is relevant for a kind of frequency calibrating method, be applied to an oscillating reference signal, making the frequency of this oscillating reference signal is a correction frequency of oscillation by initial oscillation correction, and this frequency calibrating method comprises following steps: a frequency scanning section is divided into M scanning frequency; Receive signal according to this M scanning frequency to one and carry out frequency reducing, and draw M the reception signal after frequency reducing; Individual after the reception signal after the frequency reducing carries out a correlation operation respectively to this M, draw accordingly M correlativity calculation result; Distinguish N frequency group in this M scanning frequency, wherein respectively this frequency group all comprises P selecting frequency, and this P selecting frequency is corresponding to P adjacent in this frequency scanning section scanning frequency; Respectively this N frequency group carried out group computing and draw N group's operation result; And according to this N group's operation result and certainly select a selected frequency group in this N frequency group, and certainly should selected frequency group draw this correction frequency of oscillation.

For there is better understanding above-mentioned and other aspect of the present invention, preferred embodiment cited below particularly, and cooperation accompanying drawing are described in detail below:

Description of drawings

Figure 1A is that the signal in the WCDMA system transmits and the signal of receiving terminal is processed schematic diagram.

Figure 1B is the schematic diagram that common technology carries out frequency scanning side in proper order.

Fig. 2 is the preferred embodiment of the conception according to the present invention, with the schematic diagram of five frequency groups as the Calculating Foundation of frequency coarse adjustment.

Fig. 3 is the preferred embodiment of the conception according to the present invention, with the schematic diagram of 23 frequency groups as the Calculating Foundation of frequency coarse adjustment.

Fig. 4 A is according to conception of the present invention, realizes drawing the schematic diagram of the correction frequency of oscillation of frequency coarse adjustment take frequency group as the basis with the first implementation.

Fig. 4 B is according to conception of the present invention, realizes drawing the schematic diagram of the correction frequency of oscillation of frequency coarse adjustment take frequency group as the basis with the second implementation.

Fig. 5 is conception and the flow chart of the frequency calibrating method of concluding according to the present invention.

Fig. 6 is in frequency group, changes the employed ratio letter of weight allocation formula schematic diagram.

Fig. 7 A be respectively common technology and the present invention conception is applied to 3GPP TS 34.12 forms D.2.2.1 the efficiency ratio during defined the first situation than schematic diagram.

Fig. 7 B be respectively common technology and the present invention conception is applied to 3GPP TS 34.12 forms D.2.2.1 the efficiency ratio during defined second case than schematic diagram.

Fig. 7 C be respectively common technology and the present invention conception is applied to 3GPP TS 34.12 forms D.2.2.1 the efficiency ratio during defined the third situation than schematic diagram.

The main element symbol description

Mixer 102,101,201,301,303

Basestation oscillator 106 mobile communications networks 10

Local side oscillator 105 PSCH correlators 103,203,303

Mobile communications network 10 weighted sums calculate 205,305

Determining device 207,307

Embodiment

According to conception of the present invention, the practice as the Calculating Foundation of frequency coarse adjustment with frequency group has been proposed.How following elder generation uses the practice of the present invention with the embodiment explanation, and is aided with flowchart text step of the present invention.

See also Fig. 2, it is the preferred embodiment of the conception according to the present invention, with the schematic diagram of five frequency groups as the Calculating Foundation of frequency coarse adjustment.

In this preferred embodiment, the frequency scanning section is divided into seven scanning frequencys (f1～f7).Utilize the PSCH correlator that each scanning frequency in the frequency scanning section is carried out correlation operation, thereby draw the first correlativity calculation result y1 according to the first scanning frequency f1, draw the second correlativity calculation result y2 according to the second scanning frequency f2, draw closing property of third phase result of calculation y3 according to the 3rd scanning frequency f3, draw the 4th correlativity calculation result y4 according to the 4th scanning frequency f4, draw the 5th correlativity calculation result y5 according to the 5th scanning frequency f5, draw the 6th correlativity calculation result y6 according to the 6th scanning frequency f6, and draw the 7th correlativity calculation result y7 according to the 7th scanning frequency f7.

In this preferred embodiment, per three adjacent scanning frequencys are classified as a frequency group, and are taken as the frequency options in frequency group.Therefore, as seen from Figure 2, seven scanning frequency (f in the frequency scanning section _i, i=1～7) and can be divided into altogether five (G of frequency group _j, j=1～5).Different from common technology is, these can't directly be used to comparison according to the correlativity calculation result (y1, y2, y3, y4, y5, y6, y7) that scanning frequency (f1, f2, f3, f4, f5, f6, f7) draws, and after can seeing through ranking operation, more further with the G of frequency group _jBe fundamental unit, relatively according to each G of frequency group _jDrawn weighted sum operation result.

For instance, the scanning frequency among Fig. 2 is divided into five frequency groups (G1, G2, G3, G4, G5), wherein each G of frequency group _jAll represent the scanning frequency f that comprised with braces _iIn this preferred embodiment, suppose each G of frequency group _jEqual each self-contained three selecting frequencies (F1, F2, F3).Certainly, the G of frequency group _jNumber (N) and the selecting frequency number (P) that comprises of each frequency group, all designs of adaptive system and changing.

The G of frequency group _jIn selecting frequency can be according to by low and high serial number.In order to separate at each G of frequency group _jInterior selecting frequency numbering, this sentences F _j(j=1,2,3) distinguish each selecting frequency in the same frequency scanning group, and with f _i(i=1,2,3,4,5,6,7) represent the scanning frequency in the whole frequency scanning section.

The first selecting frequency F1 that the G1 of first frequency group comprises, the second selecting frequency F2, the 3rd selecting frequency F3 correspond respectively to the first scanning frequency f1, the second scanning frequency f2 and the 3rd scanning frequency f3 in the frequency scanning section.

The first selecting frequency F1 that the G2 of second frequency group comprises, the second selecting frequency F2, the 3rd selecting frequency F3 correspond respectively to the second scanning frequency f2, the 3rd scanning frequency f3 and the 4th scanning frequency f4 in the frequency scanning section.

The first selecting frequency F1 that the 3rd G3 of frequency group comprises, the second selecting frequency F2, the 3rd selecting frequency F3 correspond respectively to the 3rd scanning frequency f3, the 4th scanning frequency f4 and the 5th scanning frequency f5 in the frequency scanning section.

The first selecting frequency F1 that the 4th G4 of frequency group comprises, the second selecting frequency F2, the 3rd selecting frequency F3 correspond respectively to the 4th scanning frequency f4, the 5th scanning frequency f5 and the 6th scanning frequency f6 in the frequency scanning section.

The first selecting frequency F1 that the 5th G5 of frequency group comprises, the second selecting frequency F2, the 3rd selecting frequency F3 correspond respectively to the 5th scanning frequency f5, the 6th scanning frequency f6 and the 7th scanning frequency f7 in the frequency scanning section.

Concluding above stated specification can find, the selecting frequency (F1, F2, F3) that each frequency group (G1, G2, G3, G4, G5, G6, G7) comprises will be increased progressively by low and high order corresponding to scanning frequency (f1, f2, f3, f4, f5, f6, f7).

For example: the selecting frequency that the G1 of first frequency group comprises (F1, F2, F3) is with continuous three scanning frequencys (f1, f2, f3) headed by the first scanning frequency f1; The selecting frequency that the G2 of second frequency group comprises (F1, F2, F3) is with continuous three scanning frequencys (f2, f3, f4) headed by the second scanning frequency f2.The selecting frequency (F1, F2, F3) that all the other each frequency groups comprise also has identical characteristic.

Moreover, according to the corresponding scanning frequency of each frequency group (G1, G2, G3, G4, G5, G6, G7), each frequency group is carried out the ranking operation of correlativity calculation result.With this preferred embodiment, the drawn correlativity calculation result (Y1, Y2, Y3) of three selecting frequencies (F1, F2, F3) of just distributing three different weights (w1, w2, w3) to give to comprise according to frequency group.

For example: with the first weight w1 corresponding to according to the drawn correlativity calculation result Y1 of the first selecting frequency F1 in the frequency group; With the second scanning frequency weight w2 corresponding to according to the drawn correlativity calculation result Y2 of the second selecting frequency F2 in the frequency group; And, with the 3rd scanning frequency weight w3 corresponding to according to the drawn correlativity calculation result Y3 of the 3rd selecting frequency F3 in the frequency group.

Therefore, in this preferred embodiment, the numbering of the selecting frequency in weight, the frequency group, and the corresponding relation between the correlativity calculation result corresponding with selecting frequency can be represented as: (w1, F1, Y1), (w2, F2, Y2), (w3, F3, Y3).The numerical value of the weighting weight here can elasticity definition, for example: the second weight w2 is maximum, and the first weight equates (w1=w3) etc. with the 3rd weight.

According to conception of the present invention, will be for each G of frequency group _jInterior selecting frequency F _kDrawn correlativity calculation result Y _k, with the weighting weight w of correspondence _kCarry out corresponding weighted sum computing.Drawing the G of each frequency group _jCorresponding weighted sum operation result SY _jAfter, find out and wherein have peaked selected frequency group.Wherein, j represents the numbering of each frequency group, and k represents numbering corresponding with each selecting frequency in the frequency group.

Take Fig. 2 as example, each institute of frequency group correspondence draws weighted sum operation result SY _jCan be defined as: SY _j=w1*Y1+w2*Y2+w3*Y3, j=1 wherein, 2,3,4,5.

Certainly, the number of the selection frequency that frequency group comprises also may be different, therefore, and weighted sum operation result SY _jDefinition mode also can change thereupon.

That the number of supposing the scanning frequency in the frequency scanning section is still kept is shown in Figure 2 (f1～f7), but each frequency group changes into when comprising four selecting frequencies (F1, F2, F3, F4), can draw four frequency scanning groups (G1, G2, G3, G4), and the G of each frequency scanning group _j(j=1～4) corresponding weighted sum operation result (SY _j, j=1,2,3,4) can be represented as according to weighting and the correlativity calculation result of each scanning frequency: SY _j=w1*Y1+w2*Y2+w3*Y3+w4*Y4.

Y wherein ₁Represent basis and the G of frequency group _jThe first interior selecting frequency F ₁Corresponding scanning frequency and the first selecting frequency correlativity calculation result, the Y that draw ₂Represent basis and the G of frequency group _jThe second interior selecting frequency F ₂Corresponding scanning frequency and the second selecting frequency correlativity calculation result, the Y that draw ₃Represent basis and the G of frequency group _jThe 3rd interior selecting frequency F ₃Corresponding scanning frequency and the 3rd selecting frequency correlativity calculation result that draws, Y ₄Represent basis and the G of frequency group _jThe 4th interior selecting frequency F ₄Corresponding scanning frequency and the 4th selecting frequency correlativity calculation result that draws.

Same, when frequency group comprises four selecting frequencies (F1, F2, F3, F4), the scanning frequency that each frequency group actual packet contains also will have different combinations according to the different corresponded manners between the selecting frequency of each frequency group and scanning frequency.For example: the corresponding scanning frequency of the selecting frequency among the G1 of first frequency group (F1, F2, F3, F4) is (f1, f2, f3, f4), and the corresponding scanning frequency of the selecting frequency among the G5 of second frequency group (F1, F2, F3, F4) is (f2, f3, f4, f5), repeats no more herein.

From the above, in preferred embodiment shown in Figure 2, each selecting frequency for each frequency group comprises utilizes SY _k=w _j* Y ₁+ w ₂* Y2+w ₃* Y ₃Be weighted the summation computing, j=1 wherein, 2,3,4,5.

Therefore, further draw the first ranking operation summation SY1 according to the drawn correlativity calculation result (y1, y2, y3) of the corresponding scanning frequency of selecting frequency (f1, f2, f3) of the G1 of first frequency group; The correlativity calculation result (y2, y3, y4) drawn according to the corresponding scanning frequency of selecting frequency (f2, f3, f4) of the G2 of second frequency group further draws the second ranking operation summation SY2; The correlativity calculation result (y3, y4, y5) drawn according to the corresponding scanning frequency of selecting frequency (f3, f4, f5) of the 3rd G3 of frequency group further draws the 3rd ranking operation summation SY3; The correlativity calculation result (y4, y5, y6) drawn according to the corresponding scanning frequency of selecting frequency (f4, f5, f6) of the 4th G4 of frequency group further draws the 4th ranking operation summation SY4; And, further draw slender acanthopanax power computing summation SY5 according to the drawn correlativity calculation result (y5, y6, y7) of the corresponding scanning frequency of selecting frequency (f5, f6, f7) of the 5th G5 of frequency group.

The account form of these corresponding ranking operation summations of frequency group can be expressed as:

According to the first ranking operation summation SY1, wherein SY1=w1*y1+w2*y2+w3*y3 that the G1 of first frequency group is drawn;

According to the second ranking operation summation SY2, wherein SY2=w 1*y2+w2*y3+w3*y4 that the G2 of second frequency group is drawn;

According to the 3rd ranking operation summation SY3, the wherein SY3=w1*y3+w2*y4+w3*y5 that the 3rd G3 of frequency group is drawn;

According to the 4th ranking operation summation SY4, the wherein SY4=w1*y4+w2*y5+w3*y6 that the 4th G4 of frequency group is drawn; And,

According to the slender acanthopanax power computing summation SY5 that the 5th G5 of frequency group is drawn, wherein SY5=w1*y5+w2*y6+w3*y7.

Then, the numerical value of these ranking operation summations relatively, the person that selects wherein have the greatest measure, and with the corresponding frequency of this maximum weighted computing sum value group as selected frequency group.

Suppose according to the comparative result after the ranking operation summation to be: when the first ranking operation summation SY1 has maximum, just with the G1 of first frequency group as selected frequency group, and with the corresponding scanning frequency of center selecting frequency in the G1 of first frequency group as revising frequency of oscillation.In like manner, if according to the comparative result after the ranking operation summation be: the numerical value of the second ranking operation summation SY2 is during for the maximum in all ranking operation summations, just with the corresponding scanning frequency of center selecting frequency among the G2 of second frequency group as revising frequency of oscillation.

Subsidiary one carry be, in this preferred embodiment, because the selecting frequency number that each frequency group comprises is three (F1, F2, F3), therefore the center selecting frequency is the second selecting frequency F2 in the middle of being positioned at, but, along with the number of the selecting frequency that comprises of scanning group is different, the numbering of the corresponding selecting frequency of center selecting frequency also may be different.

For instance, when the Gj of scanning frequency group comprises five selecting frequencies, just with the 3rd selecting frequency F placed in the middle ₃Corresponding scanning frequency is as the center selecting frequency.Moreover, when the Gj of scanning frequency group comprises four scanning frequencys, then with the second selecting frequency F placed in the middle ₂With the 3rd selecting frequency F ₃The mean value of corresponding scanning frequency is as revising frequency of oscillation.The practice of remaining scanning frequency number also can be analogized according to this.

Hold, the present invention proposes a kind of frequency calibrating method, and being applied to be proofreaied and correct by initial oscillation with reference to the frequency of oscillator signal is to revise frequency of oscillation.Take Fig. 2 as example, the frequency calibrating method that this preferred embodiment is adopted can be summarized as following steps:

The frequency scanning section is divided into 7 scanning frequencys (f1, f2, f3, f4, f5, f6, f7); Carry out to received signal frequency reducing according to 7 scanning frequencys, and draw 7 reception signals after frequency reducing; To 7 after the reception signal after the frequency reducing carries out correlation operation respectively, draw accordingly 7 correlativity calculation result (y1, y2, y3, y4, y5, y6, y7); Distinguish 5 frequency groups (G1, G2, G3, G4, G5) in 7 scanning frequencys, wherein respectively this frequency group all comprises 3 adjacent scanning frequencys; Respectively 5 frequency groups (G1, G2, G3, G4, G5) are carried out group's computing and draw 5 group's operation results (SY1, SY2, SY3, SY4, SY5); And select a selected frequency group according to 5 group's operation results and in 5 frequency groups (G1, G2, G3, G4, G5), and draw the correction frequency of oscillation in the scanning frequency that comprises of certainly selected frequency pool group.

In order to illustrate further conception of the present invention, below utilize in addition preferred embodiment of explanation of Fig. 3, the present invention is described, and conduct draws the basis of revising frequency of oscillation such as how frequency pool group G, draws the better frequency coarse adjustment practice of detected ratios.

For the ease of comparing with common technology, in Fig. 3, with Figure 1B equally to utilize 1.4ppm as frequency scanning class interval Δ f.Difference then is that the frequency scanning section is divided into 31 scanning frequencys, and therefore, the scope of frequency scanning section herein can be larger than the frequency scanning section of common technology.

Further relatively can find, the frequency scanning section of Figure 1B is (1.4*11=15.4) in positive and negative 11 the frequency scanning class interval Δ f of initial oscillation.That is to say, scanning frequency can be between (initial oscillation-15.4ppm) is to (between the initial oscillation+15.4ppm).On the other hand, the frequency scanning section of Fig. 3 A is (1.4*15=21) in positive and negative 15 the frequency scanning class interval Δ f of initial oscillation.That is to say, scanning frequency can be between (initial oscillation-21ppm) is to (between the initial oscillation+21ppm).

Can learn referring to aforesaid explanation, the present invention is take frequency group as the basis, and the center selecting frequency of selection frequency group is as revising frequency of oscillation, therefore, the scanning frequency (such as f1, f7 etc., the selecting frequency number that video frequency group comprises and different) that is positioned at the edge in the frequency scanning section just can not become selected for revising frequency of oscillation.Also therefore, if wish to revise gap between frequency of oscillation and the initial oscillation when the scope of-15.4ppm to 15.4ppm, then the present invention in fact be used as the frequency scanning section scope then-the error frequency scope of 15.4ppm to 15.4ppm is larger.

See also Fig. 3, the preferred embodiment that it is conceived according to the present invention is with the schematic diagram of 23 frequency groups as the Calculating Foundation of frequency coarse adjustment.In this preferred embodiment, the frequency scanning section is divided into the individual scanning frequency of M (M=31).When comparing with previous preferred embodiment, if the scope of frequency scanning section keep when identical, if when the scanning frequency number that comprises is more, relatively represent frequency level less apart from Δ f.

In this preferred embodiment, suppose that each frequency group has comprised the individual selecting frequency of P (P=9), these selecting frequencies correspond to nine the adjacent scanning frequencys (adjacent frequencies) in the frequency scanning section.This sentences F1～F9 and represents nine selecting frequencies in the arbitrary frequency group.Wherein the 5th selecting frequency F5 in the frequency group is the center selecting frequency in the frequency group, and remaining selecting frequency (F1, F2, F3, F4, F6, F7, F8, F9) is treated as the subordinate selecting frequency.Certainly, during practical application, the number of the selecting frequency that frequency group comprises is not as limit.

Moreover the corresponding correlativity calculation result of aforementioned respectively this scanning frequency is respectively: the first selecting frequency correlativity calculation result Y1, the second selecting frequency correlativity calculation result Y2, the 3rd selecting frequency correlativity calculation result Y3, the 4th selecting frequency correlativity calculation result Y4, the 5th selecting frequency correlativity calculation result Y5, the 6th selecting frequency correlativity calculation result Y6, the 7th selecting frequency correlativity calculation result Y7, the 8th selecting frequency correlativity calculation result Y8, the 9th selecting frequency correlativity calculation result Y9.

After drawing whole correlativity calculation result, according to the order of each scanning frequency in each frequency scanning group, provide the ranking operation weight of each selecting frequency correlativity calculation result (weighting) in each frequency group (w1, w2, w3, w4, w5, w6, w7, w8, w9).

Namely with the first ranking operation weight w ₁Corresponding to the first selecting frequency correlativity calculation result Y1 in each frequency group, with the second ranking operation weight w ₂Corresponding to the second selecting frequency correlativity calculation result Y2 in each frequency group, the rest may be inferred by analogy for it.Thus, just can obtain weighted sum operation result SY such as the 4th formula _i

$S{Y}_i=\underset{j=1}{\overset{0}{\mathrm{\&Sigma;}}}{y}_j\&times;w_j$ (the 4th formula)

W wherein _jRepresent correlativity calculation result y _jCorresponding weight.

Because the output y of correlator _iCan utilize the parameter (normally distributedrandom variable) of a normal distribution to be similar to, particularly when its signal noise is larger than (signal to noise power ratio is referred to as SNR).In other words, because the output y of correlator _jRoughly with exp (((y ₁-E[y ₁]) ²+ ...+(y _j-E[y _j]) ²+ ...+(y _o-E[y _o]) ²)) proportional.Therefore, can be according to the different scanning frequency f _iCalculating, make the weighted sum operation result and find out Be peaked scanning frequency f _i

The mode of the correction frequency of oscillation that the present invention is used for estimating is according to maximum weighted computing summation and corresponding scanning frequency.That is, as the G of frequency group _jCorresponding ranking operation summation SY _jNumerical value when larger, represent selecting frequency that this group comprises more near the actual frequency of carrier signal.Therefore, estimate that the correction frequency of oscillation that draws belongs to the scanning frequency in the frequency group with maximum weighted sum.Just represent maximum likelihood estimator (maximum likelihood estimator) with the 5th formula herein.

$\widehat{f_0}=\underset{f_i}{\max }\left\{S{Y}_i\right\}$ (the 5th formula)

According to above stated specification, the present invention utilizes the conception of frequency group to be learnt by Fig. 2,3 explanation, with the next practice of further utilizing Fig. 4 A, 4B explanation to realize above-mentioned conception.

See also Fig. 4 A, it realizes drawing the schematic diagram of the correction frequency of oscillation of frequency coarse adjustment take frequency group as the basis with the first implementation according to conception of the present invention.

Receive signal and carry out frequency reducing via frequency mixer 201 with oscillating reference signal, and the frequency f of oscillating reference signal _RefWill be by initial oscillation f _Ref=f _OrigBe adjusted into and revise frequency of oscillation f _Ref=f _{I '}

Afterwards, utilize frequency mixer 201 to export the result of frequency reducing to PSCH correlator 203.Because the frequency scanning section is divided into M scanning frequency, therefore change oscillating reference signal and be this M scanning frequency, and be used for respectively carrying out to received signal frequency reducing, can draw altogether the individual reception signal after frequency reducing of M.

Then, for each reception signal after frequency reducing, carry out related operation by the PSCH correlator after, draw M correlativity calculation result.Then distinguish N frequency group according to this M correlativity calculation result, wherein respectively this frequency group all comprises P selecting frequency, and this P selecting frequency is by selecting P adjacent scanning frequency in the frequency scanning section.And, according to the division of frequency group, and respectively this N frequency group (for example: weighted sum computing 205) draw N group's operation result carried out group computing.

In N frequency group, select again at last a selected frequency group according to this N group's operation result, and certainly selected frequency pool group draws the correction frequency of oscillation.Namely the result with the weighted sum computing compares, by determining device 207 judge draw the frequency group with maximum weighted total value after, with this frequency group as selected frequency group, and with reference to the frequency f of oscillator signal _RefBe modified to the center selecting frequency in the selected frequency group.

See also Fig. 4 B, it realizes drawing the schematic diagram of the correction frequency of oscillation of frequency coarse adjustment take frequency group as the basis with the second implementation according to conception of the present invention.In this implementation, data storage element further is provided, and can have reduced the time that receives signal of waiting for, and also be minimized for the power consumption of switching radio frequency (radio frequency is referred to as RF) module.

In simple terms, the second preferred embodiment is divided into two stages with the step of frequency reducing, sees through respectively frequency mixer 301 and carries out with frequency mixer 302.

At first utilize frequency mixer 301 to carry out to received signal frequency reducing, the mode of its frequency reducing is to utilize initial oscillation to carry out to received signal frequency reducing, draws an initial frequency reducing signal through frequency mixer 301.

After drawing initial frequency reducing signal, data storage element 309 stores it.Because initial frequency reducing signal is stored, the follow-up frequency reducing action that can further utilize again test signal initial frequency reducing signal to be made the second degree.Because initial frequency reducing signal is stored in the data storage element 309 with the form of numerical digit, when allowing follow-up calculating, can with the mode of numerical digit, directly carry out correlation operation to the reception signal through down conversion process.

That is to say, the frequency mixer 302, PSCH correlator 303, weighted sum calculating 305, the determining device 307 that are used for carrying out for the second time frequency reducing among Fig. 4 B all are to carry out in numerical digit state (digital domain).

Then, utilize respectively again M frequency-splitting between each scanning frequency and initial oscillation (that is, each frequency scanning class interval) and initial frequency reducing signal is carried out frequency reducing, and then draw M the reception signal after frequency reducing.

When utilizing test signal that initial frequency reducing signal is carried out frequency reducing, the frequency of test signal also changes along with the variation of frequency scanning class interval at every turn, and just control data storage element 309 this moment, makes it that content of previous stored initial frequency reducing signal is provided again.

That is, according to scanning frequency f _iChange, this kind implementation is divided into two stages with the frequency reducing process, the phase I is directly to carry out to received signal frequency reducing with initial oscillator signal; Second stage then is initial frequency reducing signal to be done the frequency reducing of the second degree with the difference between different scanning frequencys and the initial oscillator signal.The result of calculation of the second degree frequency reducing will be output to PSCH correlator 203.

Because carry out the flow process of correlation operation for the result after each frequency reducing all similar with the first implementation, no longer describes in detail herein.

In brief, in the second implementation, receive signal at first with initial oscillation f _RefThis reception signal is carried out frequency reducing (down-convert) and draws initial frequency reducing signal.After receiving signal process frequency reducing (downconversion), just initial frequency reducing signal is stored.Namely the signal of several WCDMA time slots is recorded in the buffering area that data storage element provides.That is frequency reducing, PSCH correlation calculations, weighted sum that the second implementation is carried out each sweep signal in the mode of numerical digit calculate, and the determination frequency error, and then significantly improve whole processing speed.

Subsidiary one what carry is that the record and broadcast (record-and-replay) mode that adopt according to the second implementation are if obtain a reliable correlativity calculation result y _iNeed to use receive the many time slots in the signal, these time slots can be divided into a plurality of sections (segments), and are recorded respectively and replay.In addition, along with the variation that receives signal, the stored content of buffering area also can be by the signal update of other sections.

As example, the second implementation can be divided into a plurality of sections with 30 time slots with the reception signal that uses 30 time slots, and the signal of each section is stored in the memory, just can reuse.That is will receiving signal, to be recorded in buffering area (buffer) inner, can be repeated for different scanning frequencys to play.Because after receiving signal and being stored and carrying out frequency reducing, carry out computing in the mode of numerical digit during follow-up use, its processing speed significantly improves.

It is noted that no matter whether adopt data storage element, aforementioned two kinds of realization sides are that it selects the mode of scanning frequency to choose from M scanning frequency according to a predefined procedure when carrying out correlation calculations selecting scanning frequency.Wherein the mode of predefined procedure arrangement scanning frequency may be to increase progressively to choose sequentially, successively decrease and choose sequentially, choose at random order (interleaved) etc.

Adopt this kind to choose at random advantage sequentially and be, can reduce channel change (channel variation) for the impact of correlation calculations, make the result of correlation calculations more accurate.For instance, 9 selecting frequencies in same frequency scanning group just can utilize F ₁, F ₅, F ₉, F ₂, F ₆..., F ₈Order choose.

Conclude aforesaid explanation, bearing calibration proposed by the invention can be summarized as following step and flow process.

See also Fig. 5, it is conceived according to the present invention and the flow chart of the frequency calibrating method of concluding.In order to conclude conception of the present invention, below will represent each scanning selecting frequency number altogether that group was comprised with number, the P that M represent scanning frequency number in the frequency scanning section, N representative scanning group.M=N+ (P-1) wherein, and P is more than or equal to 2.

In addition, in any two the adjacent frequency groups in N frequency group, comprise identical (P-1) individual selecting frequency.With Fig. 2 example, each frequency group all comprises three selecting frequencies, and wherein the G1 of first frequency group and the G2 of second frequency group all comprise the second scanning frequency f2 and the 3rd scanning frequency f3 in the frequency scanning section.

At first, the frequency scanning section is divided into M scanning frequency (step S501), that is, according to frequency scanning class interval Δ f (as: 1.4ppm) the frequency scanning section is divided into this M scanning frequency.And frequency scanning class interval Δ f defines according to the adjusting range of local side oscillator.

Because the frequency scanning section centered by initial oscillation, is chosen at the scanning frequency at initial oscillation two ends with the frequency scanning class interval Δ f of integral multiple, therefore, initial oscillation is the mean value of this frequency scanning section.Moreover M is odd number, and initial oscillation is (M+1)/2 scanning frequency.

Receive signal according to M scanning frequency to one and carry out frequency reducing, and draw M the reception signal (step S502) after frequency reducing.That is, according to predefined procedure and one by one after M scanning frequency chosen one, with the frequency of selected scanning frequency as the reference oscillator signal; And after utilizing oscillating reference signal to carry out frequency reducing to received signal, draw M the reception signal after frequency reducing.Predefined procedure described herein can be to increase progressively to choose sequentially, successively decrease and choose sequentially, choose at random order

Afterwards, individual after the reception signal after the frequency reducing carries out a correlation operation respectively to this M again, draw accordingly M correlativity calculation result (step S503).

Further analyzing step S503 can learn that its practice is: choose sequentially according to one and select one first compensated scanning frequency from these scanning frequencys; See through this first compensated scanning frequency and oscillating reference signal and to received signal frequency reducing; And this reception signal after the frequency reducing carried out one first correlation calculations.It is noted that, the selection of predefined procedure can change, for example: increase progressively and choose sequentially, successively decrease and choose sequentially, choose at random sequential scheduling.

Choose at random the order representative when scanning frequency is carried out correlation operation, select scanning frequency with interlace mode (interleave), when adopting this kind practice, the interference cases between signal can comparatively alleviate.

Subsidiary one carry be, can learn according to aforesaid the second execution mode, can see through storage element and record the reception signal, and accelerate the efficient of frequency correction.When adopting the second execution mode, step S503 can further be subdivided into following steps:

Utilize initial oscillation to carry out frequency reducing to received signal and draw initial frequency reducing signal; Store initial frequency reducing signal; And utilize respectively M frequency-splitting between each scanning frequency and initial oscillation and initial frequency reducing signal is carried out frequency reducing, and then draw M the reception signal after frequency reducing.

Then, distinguish N frequency group in M scanning frequency, wherein each frequency group all comprises an adjacent P selecting frequency (step S504).

Concerning M scanning frequency, individual other scanning frequency all can correspondence draw correlativity calculation result, therefore total M correlativity calculation result.

Concerning N frequency group, a plurality of selecting frequencies that each Gj of frequency group comprises (for example: nine) are all corresponding to scanning frequencys different in the frequency scanning section.And these selecting frequencies can be divided into center selecting frequency and subordinate selecting frequency.

Step S504 refers to, and first take frequency group as unit, calculates the corresponding correlativity calculation result of each selecting frequency in each frequency pool.

Then, respectively N frequency group carried out group's computing and draw N group's operation result (step S505).

The thin section flow process of further probing into step S505 comprises: choose the frequency group that carries out computing in N frequency group; In M correlativity calculation result, select P correlativity calculation result according to selected frequency group; P weight one by one corresponding to after the P that the selects correlativity calculation result, is weighted the summation computing; And, repeat aforementioned each step and draw and N N the weighted sum operation result that frequency group is corresponding.

For instance, for belonging to the selecting frequency of same frequency group, be distributed in the correlativity calculation result y of each selecting frequency in the frequency group _jCorresponding weight, and carry out the weighted sum computing.Certainly, before assigning weight, must determine in advance the selecting frequency F that frequency group comprises _jNumber.In addition, the correlativity calculation result y of each selecting frequency in frequency group _jCorresponding weight w _j, can determine according to a ratio letter formula.

Afterwards, certainly select selected frequency group in this N frequency group according to this N group's operation result again, and certainly should selected frequency group draw this correction frequency of oscillation (step S506).

Step S506 refers to (for example: the maximum in these group's operation results) draw an extreme value according to the comparison of this N group's operation result; Selection is selected frequency group with the frequency group with this extreme value as being somebody's turn to do; And, in P the corresponding scanning frequency of selecting frequency that this selected frequency group comprises certainly, draw this correction frequency of oscillation.

Wherein, draw from selected frequency pool group and to revise frequency of oscillation and refer to draw the center selecting frequency according to P the selecting frequency that selected frequency group comprises, and with the corresponding scanning frequency of center selecting frequency as the correction frequency of oscillation.

When P was odd number, the center selecting frequency referred to the corresponding scanning frequency of (P+1)/2 selecting frequency in the selected frequency group; And when P was even number, the center selecting frequency referred to (P/2) individual selecting frequency in the selected frequency group and the mean value of the corresponding scanning frequency of [(P/2)+1] individual selecting frequency.

According to conception of the present invention, in the process of carrying out the frequency coarse adjustment, can use fully 256 chips in the PSCH sequence.Therefore, even in the situation that channel condition is undesirable, the practice that the present invention's conception is adopted still can be guaranteed its stability.

The conception of the object when in sum, the present invention's proposition is judged as extreme value with frequency group.This kind practice has avoided single scanning frequency easily to be subject to the phenomenon of noise interference effect.In addition, analog result has verified that also this kind conception can provide better detected ratios.

Moreover, also can find out according to aforesaid preferred embodiment, if when further providing data storage element, can utilize first the oscillating reference signal with initial oscillation to carry out to received signal frequency reducing.Then initial frequency reducing signal is stored, so that follow-up when utilizing test signal to carry out frequency reducing calculating, can significantly promote processing speed.

Subsidiary one what carry is corresponding to the corresponding P of the correlativity calculation result of each a scanning frequency weight, can also decide through a ratio letter formula in the frequency group.For instance, the ratio letter formula with the 6th formula represents ratio beta:

β=| " _TS (t-t _b) * s ^*(t-t _b) exp (j2 π (f _i-f _I+n) t) dt|/| " _TS (t-t _b) * s ^*(t-t _b) dt| (the 6th formula)

See also Fig. 6, it changes the employed ratio letter of weight allocation formula schematic diagram in frequency group.It is noted that, conception of the present invention is, the ranking operation weight in the time of can determining that weighted sum calculates according to ratio beta, but the ratio letter formula that defines ratio beta is not limited with the 6th formula, also gives an example for another such as the 7th formula of below.

β=| " _TS (t-t _b) * s ^*(t-t _b) exp (j2 π (f _i-f _I+n) t) dt| ²/ | " _TS (t-t _b) * s ^*(t-t _b) dt| ²(the 7th formula)

In addition, if when α (t) can draw according to modes such as assessment, calculating, predictions, then also it can be arranged in pairs or groups in ratio letter formula, following the 8th formula and the 9th formula for example.

β=| " _Tα (t) s (t-t _b) * s ^*(t-t _b) exp (j2 π (f _i-f _I+n) t) dt|/| " _Tα (t) s (t-t _b) * s ^*(t-t _b) dt| (the 8th formula)

β=| " _Tα (t) s (t-t _d) * s ^*(t-t _d) exp (j2 π (f _i-f _I+n) t) dt| ²/ | " _Tα (t) s (t-t _d) * s ^*(t-t _d) dt| ²(the 9th formula)

Except determining according to ratio letter formula the corresponding ranking operation weight of each correlativity calculation result, the corresponding ranking operation weight of each selecting frequency can also be adjusted according to the needs of system design, give a larger ranking operation weight etc. for example all with 1 as weight, or to the selecting frequency of particular order.

Definition according to 3GPP TS 34.12 specifications can be learnt, when carrying out efficiency measuring in multipath fading environment (multi-pathfading environment), can test according to propagation conditions (propagation condition).Referring to the form of the 493rd page of this part specification file D.2.2.1, six kinds of representative situations have wherein been defined.

Below utilize Fig. 7 A～7C to represent respectively first three kind situation in these six kinds of situations, and with common technology and this case practice it is carried out efficiency measuring respectively.Because the test comparison result of other situations is all similar, is listed as no longer in detail all situations herein.In these accompanying drawings, delivering power lo is equivalent to the summation (lo=lor+loc) of actual delivering power lor and noise power loc.For different passages (channel), lor represents the ratio of shared power, so the value of lor can change along with the difference of passage.

In these accompanying drawings, all represent the power (Ec/lo) that transmits with transverse axis, and represent when carrying out fine tuning error frequency f with the longitudinal axis _ErrorCan be less than the detected ratios of frequency level apart from Δ f (1.4ppm).

See also Fig. 7 A, its respectively common technology and the present invention conception is applied to 3GPP TS 34.12 forms D.2.2.1 the efficiency ratio during defined the first situation than schematic diagram.

The first multichannel situation refers to the first frequency band (Band I), the second frequency band (Band II) is per hour three kilometers (3km/h) with the speed of the 3rd frequency band (Band III), and the speed of the 5th frequency band (Band V) and the 6th frequency band (VI) is per hour seven kilometers (7km/h).The signal that wherein comprises two paths (Path), and the signal in these two paths has different average powers.The relative average power in one of them path (Relative meanPower) is 0 decibel (dB), and relative delay (Relative Delay) is 0 how second (nanosecond is referred to as ns); The relative average power in another path is-10dB, and relative delay is 976ns.

Fig. 7 A is with the solid line representative conception and utilize the result of calculation of maximum weighted summation according to the present invention, and utilizes with dotted line and represent the result that common technology utilizes maximum correlation to export.When the transverse axis position is identical, corresponding to the longitudinal axis position of each point on the solid line of the practice of the present invention all than high corresponding to the longitudinal axis position of each point on the dotted line of the common technology practice.Hence one can see that, when delivering power is identical, when proportion group mode is revised the estimation of frequency of oscillation, has better detected ratios.

See also Fig. 7 B, its respectively located by prior art and the present invention conception is applied to 3GPP TS 34.12 forms D.2.2.1 the efficiency ratio during defined the second multichannel situation than schematic diagram.

In the second multichannel situation, and each frequency band (Band I～VI) all identical with the first multichannel situation, and comprise the signal in three paths.These three paths have identical average power (being 0dB), and wherein the relative delay in the first path is 0ns, and the relative delay in the second path is 976ns: the relative delay in the third path is 20000ns.

Fig. 7 B is equally with the solid line representative conception and utilize the result of calculation of maximum weighted summation according to the present invention, and utilizes with dotted line and represent the result that common technology utilizes maximum correlation to export.When delivering power is identical (when the transverse axis position is identical), the longitudinal axis position of each point is all high than the longitudinal axis position of each point on the dotted line on the solid line, and the detected ratios that represents solid line is good than the detected ratios of dotted line.

See also Fig. 7 C, its respectively located by prior art and the present invention conception is applied to 3GPP TS 34.12 forms D.2.2.1 the efficiency ratio during defined the third situation than schematic diagram.

Situation when the third multichannel situation is equivalent to the user and is in fast moving, therefore, the first frequency band (Band I), the second frequency band (Band II) are 120km/h with the speed of the 3rd frequency band (Band III), and the speed of the 5th frequency band (Band V) and the 6th frequency band (VI) is 282km/h.The signal that wherein comprises four paths (Path).

The relative average power in first path is 0dB, and relative delay is 0ns; The relative average power in second path is-3dB, and relative delay is 260ns; The relative average power in the 3rd path is-6dB, and relative delay is 521ns; And the relative average power in the 4th path is-9dB, and relative delay is 781ns.

Fig. 7 C is with the solid line representative conception and utilize the result of calculation of maximum weighted summation according to the present invention, and utilizes with dotted line and represent the result that common technology utilizes maximum correlation to export.When the transverse axis position is identical, the longitudinal axis position of each point on the solid line all than dotted line on the longitudinal axis position of each point high.Hence one can see that, and when delivering power was identical, the detected ratios of solid line was all good than the detected ratios of dotted line.

In summary, these description of drawings according to 3GPP TS 34.12 forms D.2.2.1 in the situation of defined three kinds of multipath fading environment, the detected ratios when transmitting with different delivering powers, and the corresponding difference that has compared common technology and this case technology.

In each accompanying drawing, some representative the closer to the right side is transmitted with higher delivering power, by also can finding out in the accompanying drawing, in the accompanying drawing the closer to the point (corresponding to higher delivering power) on right side all relatively the point (corresponding to lower delivering power) in left side have higher detected ratios.That is, when delivering power be-detected ratios during 19dB is for the poorest; The detected ratios of delivering power during for-17.5dB taken second place; And when delivering power be-detected ratios during 14.5dB is for best.

Also can be found out by Fig. 7 A～7C, adopt the comparison of maximum weighted summation operation result proposed by the invention, can compare the practice than the maximum correlation output of adopting common technology has better detected ratios.That is to say, if when transmitting signal as prerequisite take the condition of equal-wattage, the invention provides and to estimate comparatively accurately the practice of revising frequency of oscillation.

At last, it is noted that, although aforesaid explanation is that as the example of frequency correction, conception of the present invention is not limited with the application of WCDMA system for main syncul sequence (primary synchronization sequence) with the WCDMA system.

When practical application, be one and have the sequence of automatic correlation (auto-correlation property) through design as long as receive signal s (t).Therefore, even the communication system of other types also can adopt conception of the present invention.

For instance, the DVB-T system uses phase quadruple counter-rotating modulation (Quadrature Phase ShiftKeying, quadriphase PSK is referred to as QPSK) to carry out the transmission of signal, and after this kind modulation mode need to carry out frequency correction equally, transmission DVB-T signal that could be correct.Perhaps, for the mobile communication of other types, can follow above-mentioned conception equally and carry out frequency correction.

In sum, although the present invention discloses as above with preferred embodiment, so it is not to limit the present invention.The persond having ordinary knowledge in the technical field of the present invention, without departing from the spirit and scope of the present invention, when being used for a variety of modifications and variations.Therefore, protection scope of the present invention is as the criterion when defining with claims.

Claims (16)

1. a frequency calibrating method is applied to an oscillating reference signal, and making the frequency of this oscillating reference signal is a correction frequency of oscillation by initial oscillation correction, and this frequency calibrating method comprises following steps:

One frequency scanning section is divided into M scanning frequency;

Receive signal according to this M scanning frequency to one and carry out frequency reducing, and draw M the reception signal after frequency reducing;

Individual after the reception signal after the frequency reducing carries out a correlation operation respectively to this M, draw accordingly M correlativity calculation result;

Distinguish N frequency group in this M scanning frequency, wherein respectively this frequency group all comprises P selecting frequency, and this P selecting frequency is corresponding to P adjacent in this frequency scanning section scanning frequency;

Respectively this N frequency group carried out group computing and draw N group's operation result; And

Certainly select a selected frequency group in this N frequency group according to this N group's operation result, and certainly should selected frequency group draw this correction frequency of oscillation.

2. frequency calibrating method as claimed in claim 1 is characterized in that, in any two the adjacent frequency groups in this N frequency group, comprises identical (P-1) individual selecting frequency.

3. frequency calibrating method as claimed in claim 1 is characterized in that, M is odd number, and this initial oscillation is (M+1)/2 scanning frequency.

4. frequency calibrating method as claimed in claim 1 is characterized in that, this initial oscillation is the mean value of this frequency scanning section.

5. frequency calibrating method as claimed in claim 1 is characterized in that, M=N+ (P-1), and P is more than or equal to 2.

6. frequency calibrating method as claimed in claim 1 is characterized in that, the step of this frequency scanning section being divided into this M scanning frequency refers to: according to a frequency scanning class interval this frequency scanning section is divided into this M scanning frequency.

7. frequency calibrating method as claimed in claim 1 is characterized in that, according to this M scanning frequency and this initial oscillation this reception signal is carried out frequency reducing, and the step that draws M the reception signal after frequency reducing refers to:

According to a predefined procedure and one by one after this M scanning frequency is chosen one, with the frequency of selected scanning frequency as this oscillating reference signal; And

After utilizing this oscillating reference signal that this reception signal is carried out frequency reducing, draw this M the reception signal after frequency reducing.

8. frequency calibrating method as claimed in claim 7 is characterized in that, this predefined procedure be one increase progressively choose the order, one successively decrease choose the order, one choose at random the order.

9. frequency calibrating method as claimed in claim 1 is characterized in that, respectively this N frequency group is carried out this group's computing and the step that draws N group's operation result comprises following steps:

In this N frequency group, choose the frequency group that carries out computing;

According to selected frequency group and certainly in this M correlativity calculation result, select P correlativity calculation result;

P weight one by one corresponding to after this P correlativity calculation result of selecting, carried out a weighted sum computing; And,

Repeat aforementioned each step and draw and this N N the weighted sum operation result that frequency group is corresponding.

10. frequency calibrating method as claimed in claim 9 is characterized in that, this P correlativity calculation result draws according to P the selecting frequency that selected frequency group comprises.

11. frequency calibrating method as claimed in claim 9 is characterized in that, this P weight determines according to a ratio letter formula.

12. frequency calibrating method as claimed in claim 1 is characterized in that, according to this N group's operation result and certainly selecting in this N frequency group should selected frequency group, and the step that certainly should selected frequency group draws this correction frequency of oscillation comprises following steps:

Comparison according to this N group's operation result draws an extreme value;

Selection is selected frequency group with the frequency group with this extreme value as being somebody's turn to do; And,

In P the selecting frequency that this selected frequency group comprises, draw this correction frequency of oscillation.

13. frequency calibrating method as claimed in claim 12 is characterized in that, this extreme value refers to the maximum in these group's operation results.

14. frequency calibrating method as claimed in claim 12, it is characterized in that, in this selected corresponding P of frequency group selecting frequency, the step that draws this correction frequency of oscillation refers to: draw a center selecting frequency according to this P selecting frequency, and revise frequency of oscillation with this center selecting frequency as this.

15. frequency calibrating method as claimed in claim 14 is characterized in that:

When P was odd number, this center selecting frequency referred to (P+1)/2 selecting frequency in this selected frequency group; And

When P was even number, this center selecting frequency referred to the average of (P/2) individual selecting frequency in this selected frequency group and [(P/2)+1] individual selecting frequency.

16. frequency calibrating method as claimed in claim 1 is characterized in that, according to this M scanning frequency and this initial oscillation this reception signal is carried out frequency reducing, and the step that draws this M the reception signal after frequency reducing comprises following steps:

Utilize this initial oscillation that this reception signal is carried out frequency reducing and draws an initial frequency reducing signal;

Store this initial frequency reducing signal; And

Utilize respectively respectively M frequency-splitting between this scanning frequency and this initial oscillation and this initial frequency reducing signal is carried out frequency reducing, and then draw the individual reception signal after frequency reducing of this M.

Images