CN1893408A - Method and circuit apparatus for determining receiving-signal frequency of demodulation signals - Google Patents

Abstract

Being related to a method and a circuit device, the invention is in use for determining difference of carrier frequency in procedure for demodulating received symbol (P1,P2) in complex phase space (I,Q;R,alpha) when quadrature modulation method (QAM) is adopted. In order to determine frequency, the invention carries out comparison between received symbol and symbol (S1,S2) at nominal position in complex signal space. In order to make measurement independent to rotation between coordinate system of receiving signal and coordinate system of symbol, the invention suggest to determine angle (<(P1,P2)) between two received signal values (P1,P2), and further compares it with conceivable nominal angle in quadrature modulation method. Angle difference between measured angle and nominal angle of received signal value can be used to measure frequency deviation (delta f) directly.

Description

Determine the Method and circuits device of received signal frequency for demodulated received signal

Priority information

The German patent application 10 2,004 023889.8 that the present invention submitted to on May 12nd, 2004 is a priority, here with it as a reference.

Background technology

The present invention relates to a kind of in order in the complex phase space of quadrature modulation method, the symbol or the signal component that receive to be carried out demodulation and the method for definite frequency.

Correct determine frequency, i.e. carrier frequency is a key factor for receiving with orthogonal signalling to the receiver of relevant digital signal or receiving circuit device synchronously.Symbol is to represent monobasic or polynary digital value through the form of coding.To carrying out transfer encoding, it is equivalent to a pointer with orthogonal signalling, is assumed to particular moment at right Descartes's amplitude of orthogonal signalling and the discrete location in the phase space.Quadrature amplitude modulation (qam) and phase shift keying (PSK) all are the transmission methods of the type.

In the receiver of conventional receiving digital signals, use the complex multiplier that triggers by local oscillator or frequency mixer that the QAM signal of the carrier modulation that receives is mixed in the base band of circuit arrangement with accurate phase place and frequency.Circuit arrangement has a phase-locked loop to be used for process control usually.In the digital processing process, phase-locked loop can be before or after modulus (A/D) conversion.Signal is sampled, with the multiple in symbol clock cycle or cycle with its digitlization, perhaps the digitlization clock cycle with respect to the required symbol clock cycle from turning round.In this case, signal is converted to the multiple in symbol clock cycle or cycle by a pure digi-tal sample rate conversion process.Gain-adjusted can be guaranteed to use specific modulation range, and the signal that receives accurately is mapped to the sign determination level.Adaptive equalizer can reduce the intersymbol interference that the linear distortion of transmitter, transmission channel or receiver causes.

In the high-order demodulator of QAM or psk signal, be used for the frequency of local oscillator and phase control, gain controlling, the symbol clock cycle recovers and the automatic control circuit of adaptive equalizer need receive symbol and predetermined symbol alphabet and think most possible symbol element by differentiating level.Control types known being called " decision-feedback " control of this use signal to be determined.

Because decision-feedback control has been coupled in the digital demodulator in the prior art, when being used for that received signal is sneaked into the frequency of local oscillator of base band and phase control when also unstable, phase-locked process is just very difficult.Usually only when characteristic frequency relative with phase place during, could success phase-locked near nominal value.

The demodulator of QAM or psk signal uses phase control usually, the complex signal after received signal and the sampling is compared their coordinate in distributing to the signal space of symbol.I and Q direction in complex signal space are often used equidistant discrimination threshold.Coordinate is the mid point that the phase of received signal point of symbol nominal value is positioned at the I/Q judgement rectangle on multiple I/Q plane.

Used the method in utmost point footpath and fan-shaped territory among the DE 36 19744, rather than the quadrature decision space.At first determine one group of the most contiguous utmost point footpath among the EP 0 281 652, directly come to determine only phase angle according to wherein utmost point of being discussed in the step subsequently.United States Patent (USP) 5,741 has used in 508 limited symbol to select the decision maker of (simplification planisphere) to avoid mistake rotation in the high order modulation type.

Phase capturing zone in the existing method is very little, especially for high order modulation.But except when outside the carrier phase of circuit arrangement controlled blocked the time, symbol to be discriminated was incorrect usually, therefore can go out the direction of rotation of a mistake to specific sign computation.If the summation signals of all correction signals is drawn along phase deviation, obtained unwanted zero point in the high order modulation approach, and this has caused the locking fault.

The wrong zero point that has many methods can amplify capture range and avoid proofreading and correct summation signals at present.US5 for example, 741,508; EP 0571788; DE 36 19 744; DE 41 00 099; DE 44 10 607 and DE 199 28 206, capture range are always from the fixedly nominal coordinate point of signal space internal symbol.Do not have special logic measurement means can not expand the phase capturing scope, for example EP 0571788.

But all methods that increase the phase control scope have only been done very little improvement for this common demand of frequency difference between corrected received signal and the local oscillator, and fundamentally do not solve the demand of control method for frequency.This frequency difference has represented that the input signal coordinate system has the skew of lasting variation with respect to the rotation of the coordinate system of circuit arrangement.

If must proofread and correct frequency difference, need be provided with the loop gain of carrier Control very highly, if when not differing between input signal and the local oscillator, just correct by acquisition phase after zero point, and only remain on and to make correct or when being in the main true judgement in the very little scope (also local oscillator must be positioned at correct frequency and phase place place) when determinant.But the necessary stability limit of automatic control loop possible loop gain.

Therefore need a kind of system and method, determine in the complex signal space of modulator approach the received signal frequency that receiving symbol or signal component are separated timing with better mode of operation.

Summary of the invention

During the orthogonal demodulation signal frequency determining to receive, suppose that the nominal position of symbol in the frequency control signal that will receive and the complex signal space is made comparisons.For easier and definite more accurately frequency difference, what have advantage is to determine that angle between at least two signal values that receive is as acceptance angle, and with the angle between the corresponding nominal position that may use in this angle and the quadrature modulation method, perhaps compare, select immediate nominal angle then with Nominal angle.Differential seat angle between acceptance angle and the next immediate nominal angle is to measure of frequency difference between the local oscillator of input signal and circuit arrangement to measure.

Therefore, this method is based on the consideration differential seat angle, rather than received signal value and the fixing deviation between the nominal symbol.Therefore the instantaneous angle position in received signal space is incoherent.It can tilt with respect to the nominal coordinate system with any desired form, perhaps even under certain limitation rotates.Make in this way and can both in bigger frequency acquisition scope, produce suitable frequency correction signal in any case.Capture range and catch reliability and can further improve with the cascade method of measurement.

Certainly, can continue to change each clock cycle, realize this process progressively to the actual measurement clock cycle of determined number.

Determine to continue the acceptance angle of received signal value, and compare with corresponding nominal angle.This method can be one by one or with given interval signal value to be carried out sequential processes, thereby identify last angle determine in symbol order aspect the angulation difference, because the various combination of nominal angle defines, and makes this searching method more reliable, searches for also faster.

Cascading Methods use the Viterbi algorithm also can handle several symbols, and determine most possible new symbol.

Use the detected utmost point footpath of correlated signal values and also be of great use for delimiting possible nominal angle, because have only these symbols in forming process, to depend on specific nominal utmost point footpath with the coordinate that the nominal utmost point is directly represented.Can utilize radial component and frequency deviation or irrelevant this fact of skew of received signal.Thereby can determine the interior signal angle in zone in different poles footpath in the quadrature modulation method, and compare with corresponding possible nominal angle.

Also can advantageously determine the angle of signal value according to the utmost point footpath of clearly determining in the symbols alphabet.This can prevent the symbol of incorrect link to the adjacent poles footpath, for example the footpath of the utmost point in 64-QAM method 2-4.For example preferred use first and sextupole are directly measured in the 64QAM method, because these utmost points footpaths are contiguous, and comparatively suitable angle are arranged between respective symbol.

It also is very favourable only allowing the interior nominal angle of the received signal angle and the position field of selected symbol consistent.This has guaranteed only to have these symbols to can be used for angle and has determined, especially can be used to analyze adjacent-symbol, utmost point footpath or the like reliably, and angular relationship.In the territory of selected nominal character position, determine the angle of received signal value thus.

Can determine a plurality of angles in a kind of advantageous method, compare that one be used from definite frequency deviation then, its advantage is that combination and/or iterate improvement have been carried out in frequency measurement.Judge and consequent wrong control variables that for fear of mistake the preferred differential seat angle that uses is less than half of differential seat angle between certain nominal angle and the contiguous most probable nominal angle.

Measure between acceptance angle and the nominal angle a plurality of differential seat angles by iteration or get average and improved frequency measurement.Especially can use the differential seat angle that records gradually between acceptance angle and the nominal angle input signal as the legacy frequencies control method, or the integral part of the conventional phase of local oscillator control.

Differential seat angle between acceptance angle and the nominal angle or also can be used for the disposable frequency correction of local oscillator by the enhancing value that a plurality of differential seat angles constitute, thereby frequency that can putative signal in measuring process subsequently is corrected, required whole work only are to adjust, and for example further optimize or loss balancing.

Especially in high order modulation approach, several symbols may have identical position angle in coordinate system.And different symbol combination also can constitute identical nominal declinate.On the other hand, be used to discern symbol wherein, preferably use Cascading Methods or above-mentioned received signal value to have the coordinate method in nominal utmost point footpath or this ambiguity is eliminated in the combination of these two kinds of methods if determine nominal angle.Therefore the symbol that identifies can be used for next stage.Demodulator makes in this way and works, and is locked even its frequency and phase control also do not have.

What have advantage is that the angle of all received signal values can be mapped in the quadrant of complex plane, to determine the angle between them.This has reduced the quantity of the nominal angle of required consideration.

What also have advantage is, checks that institute considers that the zone of received signal is used for tolerance zone and possible nominal character position, perhaps nominal utmost point footpath especially, thus in further analysis, can eliminate the loss of received signal, and measurement is not influenced by loss can.The received signal value has the coordinate in one group of utmost point footpath in the method for optimizing in the tolerance zone, and these utmost points directly can be used for defining the selection of nominal angle.

The information that is used for distinguished symbol that difference obtains if will take measurement of an angle is used for further handling these symbols, be preferably in after the only angular surveying in a quadrant and further before the signal processing, by the reverse rotation of complex signal, it suitably is corrected in the correct quadrant.Preferred this method that adopts realizes, wherein only after the angular surveying in a quadrant and further before the signal processing, sign or the rotation of this symbol of suitably returning special symbol turn back in the correct quadrant.

In case initial frequency difference is fully proofreaied and correct, be used for the information that the differential seat angle of distinguished symbol extracts and just the absolute angle of one or more symbols can be used for phase control.

This method or corresponding circuit arrangement are particularly useful for binary system or plural Ditital modulation method, for example PSK or QAM.These modulator approaches are usually used in many via cable, satellite and sometimes based on news broadcast, TV and the data, services on land.

As shown in drawings, these targets of the present invention, feature and the following detailed description of preferred embodiment of advantages will be more obvious.

Description of drawings

Fig. 1 is a schematic block diagram of determining the system of symbol from digitized signal, and at least one orthogonal signalling that digitized signal is coupled to modulator approach are right, and circuit arrangement has the equipment of determining the received signal frequency;

Fig. 2 is the graphic representation of selected intersymbol angular relationship signal and character position on complex plane;

Fig. 3 is differential seat angle possible between received signal and the symbol plane graph as angle function;

Fig. 4 is the flow chart that frequency is determined method; And

Fig. 5 and 6 is frequency or phase control switching device shifter.

Embodiment

Demodulator 1 is used for determining the symbol Se of digitized signal sd, S, and its orthogonal signalling that are coupled to modulator approach are right, and for example according to the QAM standard, it comprises a plurality of independent components.They can be designed or also be can be used as the part of integrating circuit separately.Especially, the component that describes below can omit or replenish with additional components, and this depends on application.Equally also can be with signal as real signal, complex signal or single complex signal components, this depends on uses and the special circuit device.

Demodulator 1 receives from signal source 2 analog signal of frequency modulator for example from receiving lines 102.This analog signal is positioned at the intermediate frequency layer of limited bandwidth usually, is sent to analog to digital converter 3, and the latter provides digital signal to circuit 104.This digital signal input tape bandpass filter 5, the latter removes from digital signal and stablizes component and harmonic distortion, then the signal behind the bandpass filtering is sent to circuit 106.

Signal on the circuit 106 behind the bandpass filtering is imported into quadrature converter 6, and the latter is a baseband signal with conversion of signals.Base band satisfies the requirement of demodulator 1 and used modulator approach.Quadrature converter 6 provides the in-phase signal and the quadrature signal component I of cartesian coordinate system, Q respectively.In order to carry out frequency inverted, quadrature converter 6 has two carrier waves usually, is provided by local oscillator 7, and phase difference is 90 °, and its frequency and phase place are subjected to 8 controls of carrier wave control unit.These are major parts of phase place Coupling Control process.Quadrature signal component I, Q are sent to low pass filter 9, and decayed harmonic distortion and limited signal bandwidth of the latter is so that sample subsequently.Therefore filtered orthogonal signalling are to I, Q or two quadrature signal component I, and Q is sent to symbol sampler equipment 10 subsequently, and the latter has a controlling of sampling unit.By input control character sample devices 10, sampled signal t is arranged wherein _iSymbol sampler number of times t during normal operating conditions _iBe fixed as symbol rate 1/T or its multiple, and be fixed as the position of the precise phase of receiving digital signals usually.Low pass filter 9 and symbol sampler equipment 10 carry out temporal interpolation between the sampled value of A/D converter 3, insert a clock cycle or its integral multiple of symbol rate.

The output signal of sample devices 10 is carried out filtering by a low pass filter 11 with Nyquist feature, and the signal of generation is input to the feedback oscillator control unit 12 that has adjuster 19.Adjusting gain control unit 12 can utilize the modulation range of symbol judgement unit 15 best.The output signal of gain control unit 12 is sent to equalizer 14, and the latter has removed two orthogonal signalling to component I, the distortion among the Q, and at circuit 108 generation signal S.Afterwards, symbol judgement unit 15 is according to the symbol Se on the above-mentioned preparatory signal formation circuit 110.

Symbol Se on the circuit 110 is sent to another digital signal processing appts 16, and signal S on the circuit 108 and symbol Se are sent to the automatic control circuit in the demodulator 1 directly or indirectly or have the device of differentiating feedback subsequently.Symbol Se is used by the adjuster 19 of equalizer 14, gain control unit 12, gain control unit, carrier Control unit 8 and symbol sampler equipment 10 thus.These automatic control circuits are according to circuit arrangement, with two quadrature signal component I of cartesian coordinate or polar form received signal S or symbol Se, Q.

Gain control unit 12 is driven by fader 19, the signal indication that the latter receives the poor Δ R in utmost point footpath that records of utmost point footpath comparator 18.Directly compare in order to carry out the utmost point, utmost point footpath comparator 18 is sent to the signal S on the receiving lines 108 before the symbol judgement unit at signal S, simultaneously symbol Se co-pending on the receiving symbol judgement unit 15 interior circuits 110.

Therefore gain control unit 12 has received the gain factor V that fader 19 produces according to utmost point footpath error delta R, and fader 19 preferably is set to proportional integral (PI) adjuster.

Alternatively, gain control unit 12 can utilize the reference parameter in the memory, thereby comparator input signal and output signal are regulated self.The advantage of this way is to have cancelled utmost point footpath comparator 18 and fader 19, perhaps can be partially integrated in the gain control unit 12.And, this device can prevent two independently control procedure become the part of automatic control loop, i.e. gain controlling in the gain control unit 12 and another automatic control process in the equalizer 14.

Signal S and symbol Se are input to a time error checkout equipment 20.Time error checkout equipment 20 is determined clocking error, so-called timing error Δ t.Timing error Δ t is sent to sample devices adjuster 21, and it preferably is set to pi regulator.The control voltage that will produce then is sent to sample devices 10, and the latter converts cartesian coordinate I at input signal, and Q is afterwards to its sampling.

And signal S and symbol Se be input to angular error checkout equipment 22, and the latter determines angular error Δ γ, P and I component that it is excited as local oscillator 7 by carrier Control unit 8.

Carrier Control unit 8 also receives frequency deviation Δ f, and preferably also receives a switching signal U.Switching signal U and frequency deviation Δ f are provided by frequency preparation unit 50.

Frequency preparation unit 50 receives Descartes's complex coordinates space I, receiving and pretreated signal S among the Q.This signal converts polar coordinates R to by coordinate converter 51, α, and utmost point footpath component is sent to another symbol judgement unit 52.This symbol judgement unit 52 output stages footpath components R e, and the angle component α of coordinate converter 51 is input to memory 53, M.The utmost point footpath components R e that judgement unit 52 is judged also is input in the comparator 55.Comparator 55 receives the utmost point to be determined footpath components R e ' and differential seat angle component Δ α from memory 53, i.e. poor between the angle component α ' among the angle component α of coordinate converter 51 and the memory M.

In this additional judgement unit 52, utmost point footpath or the utmost point directly organized make differentiation.Can select by all utmost point footpaths, perhaps not by any utmost point footpath.Last utmost point of determining footpath value and the angle of determining at last or last several utmost point footpath value and angle of determining are stored in the memory 53.

Placed in the form 54 in the symbols alphabet corresponding utmost point footpath between all possible differential seat angle, it is arranged to the additional storage part in an annex memory or the memory 53.Therefore the extention of form 54 or memory 53 is as the preparation device in nominal angle and/or nominal utmost point footpath.These nominal values, current utmost point footpath value re, current differential seat angle Δ β, and once get utmost point footpath value re ' the comparator 55 before the expression from form 54.Comparator 55 with the corresponding utmost point of nominal symbol footpath between all possible differential seat angle compare with form 54 interior respective value.Comparator 55 is the utmost point to be determined footpath re, re in re ' and the form 54, and the combination of re ' compares, and seeks the most similar Nominal angle difference Δ β as determining or the actual angle difference Δ α of measurement.Advantage is the arch differentiation that can realize crossing over a plurality of mark spaces.Also advantageously, when uncertainty is too high, (be higher or lower than limit value) and just refuse this result.Output frequency error or frequency deviation Δ f are as the actual angle difference Δ α that records according to coordinate converter 20 differential seat angles between the storing values on the one hand and before in the memory 53, the result between the differential seat angle between the most probable Nominal angle difference Δ β on the other hand and in the form 54, so frequency error Δ f=f _Symbol(Δ α-Δ β)/2 π.Frequency error Δ f is sent to carrier Control unit 8.

Preferably this frequency error Δ f also is input to select switch 56, the latter monitors the size of frequency error Δ f.Because bigger when frequency error Δ f begins usually, frequency error Δ f is as the control variables of carrier Control unit 8.In case it is very little that Δ f becomes, when preferably being lower than threshold value, use usually and the differentiation process or the angular error checkout equipment 22 resulting angular error Δ γ that are familiar with.

Pi regulator is arranged in preferred carrier Control unit 8.According to an embodiment, the switching signal U of select switch 56 simply switches to angular error Δ γ at the input of PI device from frequency deviation or frequency error Δ f.

Among the embodiment of carrier Control unit 8 shown in Figure 5, the switching signal U of select switch 56 is sent to select switch 81 herein.Select switch 81 switches between two input signals, and they are multiplied each other at multiplier 82 places obtains the P component, and multiplying each other at multiplier 83 places obtains I component.The output signal of I multiplier 83 is admitted to integrator 84, and the latter's output signal is sent to adder 85, second output signal that input is a P multiplier 82 of adder.The output signal of adder 85 is as the control signal of local oscillator 7.

According to the preferred embodiment among Fig. 6, frequency deviation or frequency error Δ f equal or are directly proportional in determined angular error Δ α-Δ β, and it is directly delivered to the I device of carrier Control unit 8, and multiplies each other with unit self coefficient F.Angular error Δ γ is delivered to I device and P device simultaneously, and multiplies each other with I value or P value.The switching signal U of select switch 56 is at switch 181 places, the latter the first time switch step between P product component and empty position, switch, and the second time switch step multiply by between the coefficient F at I product angular error Δ γ and frequency deviation Δ f and switch.In the above-mentioned carrier Control equipment, the switching output valve that has I component is sent into integrator 84 successively, and arrives adder 85 thus, and second input of the latter is P component or the zero-signal after switching.Its output signal is sent to local oscillator 7 successively.

For control circuit device 1 and make it that fundamental clock cycle is provided, except the essential device of other operations, it has a clock generator 23 and adjuster C especially.

The conversion equipment 6 of circuit arrangement is transformed into Descartes's complex space I with digital signal, and when Q and transducer 51 converted aforementioned polar coordinates to, circuit arrangement also can convert digital signal sd to first transducer and have polar coordinate, the complex signal of R.Other respective component of ifs circuit device is used for suitably handling the complex signal I of cartesian coordinate space, during Q, and also can omitted coordinate transducer 51.

In the coordinate converter 51, (orthogonal signalling of promptly sampling Q) convert polar coordinates R, α to I to the complex signal that provides in the cartesian coordinate system.Therefore, utilize polar coordinates according to I=Rcos (α) and I=Rsin (α), and according to relational expression $R=\sqrt{(I^2+Q^2)}$ And α=arctan (Q/I), can constitute utmost point footpath components R and angle component α.

As an alternative, also can use the coordinate converter of another kind of type.Often use so-called Cordic method during Digital Signal Processing, wherein addition and binary multiplication are only used in conversion, and it can be realized with the simple arithmetic shift of binary number.As an alternative, other similar approach or form also are feasible.Reverse conversion (promptly being transformed into quadrature component I and Q from polar coordinates components R and α) also can use Cordic transducer, form or similar approach to finish similarly.

Describe to use before symbol Se determines the method for the frequency of demodulated received signal or carrier frequency, at first describing and have coordinate x, the symbol distribution in the Descartes's complex space I of y, Q according to Fig. 2.Also show corresponding coordinate R in the polar coordinate space, α.

Orthogonal signalling are to I, and the symbol S of 64-QAM signal has been represented on the cartesian coordinate plane that Q determines _Xy, indicated its corresponding position in a quadrant.The index x of corresponding specific cartesian coordinate, y are used for determining the symbol in the desirable sampling.Nine the round Ka that also drawn simultaneously, Kb, Kc ..., Ki, the symbol S on the circle _XyRegularly arranged according to the 64-QAM method.Each circle Ka, it is respectively R that Kb, the utmost point of Kc directly are worth _a=1.41; R _b=3.16 and R _c=4.24, all be to calculate from the origin of coordinates.For the polar coordinates R according to symbol, α determines symbol S _Xy, need specific angle component α.Symbol S for example ₁₁And S ₃₃Be 45 °, and S ₁₃And S ₃₁It is 71.7 ° and 18.3 °.Polar calculating is finished in the coordinate converter 51 of foregoing circuit device.When considering the angular relationship between each symbol, it is very useful converting polar coordinates to.

For example Fig. 2 has shown nominal position S in the cartesian coordinate system ₁₁And S ₅₇On two nominal symbol S1 and S2.Drawn straight line from initial point to these nominal position.Angle between the straight line has provided nominal angle ∠ (S _Xy, S _{X ' y '}).

In the form (Fig. 1) of memory 54 stored, listed distinct symbols the nominal angle ∠ (S between might nominal position _Xy, S _{X ' y '}), (∠ (S1, S2)) for example.What have advantage is, the coordinate of each symbol also has about being provided with the utmost point footpath Ra of symbol, Rb ... the information of Ri.As one very big comprise nominal angle might combination of angles the alternative form of form, can certainly indicate the position angle information of all nominal position, like this by forming the nominal angle of difference simply with regard to any given combination of energy measurement.

In the method that describes below, suppose mark-hold S1, and the differential seat angle ∠ that existence is fixed between the S2 (S1, S2).The received signal space can be thought incoherent with respect to the instantaneous angle position of the coordinate system of nominal symbol, even because in the received signal space, the angular relationship between each received signal value has identical ratio with nominal angle between the nominal symbol.Therefore the acceptance angle ∠ between the position of two received signal value P1 and P2 (P1, P2) with the highest probability corresponding to two respective symbol S1, corresponding nominal angle ∠ (S between the S2 _Xy, S _Xy), its locus with received signal this moment is irrelevant.

Because what use is not absolute angular position, and two received signal value P1 have only been used, differential seat angle ∠ (P1 between the P2, P2) and two respective symbol S1, nominal angle ∠ (S1 between the S2, S2), for carrier frequency produces a suitable frequency correction signal, have a very large capture range.Among Fig. 2 in order to distinguish received signal P1, P2, the with dashed lines corresponding straight line that draws.

For example according to table entries, (S1 is S2) with (∠ (S as nominal angle ∠ ₁₁, S ₅₇)) when being 9.5 °, but (∠ (P1, P2)) is 12 ° for received signal value P1 that measures and the angle between the P2.Consider the possible angular relationship of present 64-QAM modulator approach, (P1 also has other possible symbol combination in margin of tolerance P2) for different angle ∠.For example, symbol S ₁₁With symbol S ₃Or S ₃₅Between nominal angle ∠ (S1, S3) or ∠ (S ₁₁, S ₃₅) be 14 °.Shown in example, in the margin of tolerance, still can not be definitely definitely (P1 P2) distributes to a nominal angle with the angle ∠ of received signal value P1 and P2.

What have advantage is, when the coordinate probability is not when fully determining, and can be with the utmost point footpath R of received signal value P1 and P2 as another discrimination standard.Use utmost point footpath Ra, Rb ..., Ri, form clearly show the coordinate is-symbol S of signal P2 ₅₇Or symbol S ₇₅Simultaneously use angle difference ∠ (P1, P2) and the utmost point of P2 footpath admissible mark S only ₅₇As (S1, S2) right second portion, so nominal declinate ∠ (S ₁₁, S ₅₇) be 9.5 degree, differential seat angle reaches 12 degree-9.5 degree=+ 2.5 degree.This differential seat angle is the ratio of full circle, and multiply by symbol frequency f _Symbol, be exactly the frequency deviation of 7 of the local oscillations of reception carrier and circuit arrangement 1, i.e. Δ f=2.5 degree/360 degree * f _Symbol

As shown in Figure 2, the second received signal value P2 accurately is not positioned on the nominal utmost point footpath.Therefore suggestion is provided with the margin of tolerance around each nominal utmost point footpath, supposes that received signal belongs to the symbol on this nominal utmost point footpath.In this case, for more measured different angles, need to reduce the quantity of the different nominal angles of being considered.The margin of tolerance can be arranged to apart from the distance in contiguous nominal utmost point footpath half to have special advantage, and has determined a decision boundaries in this position.It is poor to constitute a fixing or variable utmost point footpath, for example in order to adapt to condition of acceptance, and also also possible different on the direction in bigger and less utmost point footpath.The margin of tolerance that forms also can be overlapping, and this means needs more Nominal angle difference to be used for differentiating.Signal value or other the uncertain signal values in assessment of also having got rid of close adjacent poles footpath by assessment.And as another constraints under the high order modulation approach, suggestion only realizes this method in specific tolerance differential seat angle.

Determined after the frequency deviation Δ f, the second received signal value P2 directly can be corrected to the tram of correct signal value P2*.Also can continue to use fixed frequency deviation Δ f in the comparator 55, thus this frequency deviation Δ f from beginning as the correction value of subsequent processing steps.

Because the differential seat angle between the nominal angle of use received signal angle and respective symbol depends on the quadrant on the complex plane, can advantageously all received signals be rotated in the quadrant of complex plane, for example receiving symbol P1 ^*, position angle is mapped in the zone between 0 ° and 90 ° or-90 ° and+zone between 90 ° on corresponding differential seat angle.

Therefore can not be directly the complex signal that receives be judged to be a symbol in the alphabet, but at first determines the nominal utmost point footpath that may occur each signal value in the symbols alphabet.Determine two then directly continuously or the signal value P1 of apart, P2, angle ∠ (P1, P2).For this angle, preferably consider possible utmost point footpath coordinate, (S1 S2) compares, and for example can directly produce any given nominal dot combination for the utmost point of the first signal value P1 and secondary signal value P2 with the nominal angle ∠ that occurred.From form to a nominal angle ∠ (S1, S2) differentiate after, for example (P1, angle P2) identify receiving symbol and corresponding signal value P1 and P2 to the most close measured angular ∠, and this instantaneous position with phase space has nothing to do.

(P1, P2) (S1, differential seat angle S2) may be identical to the angle ∠ that different received signals are right, and have provided the method for direct measurement frequency deviation, as clearly shown in Figure 3 to the most similar nominal angle ∠.The absolute angular position of actual symbol or nominal symbol is incoherent.Fig. 3 has shown that all utmost points are directly gone up all possible angle ∠ (S1, mapping graph S2) between the interior point of quadrant in the 16QAM system.This mapping graph is the frequency distribution in angle domain, and expenditure is represented.Solid line is also represented relevant position in the nominal coordinate system, and dotted line represent to rotate the back receive acceptance angle ∠ in the coordinate system (P1, P2).Nominal angle is respectively-63.4 ° corresponding to 0 ° ,-53.1 °, and-36.9 ° ,-26.6 °, 0 °, 26.6 °, 36.9 °, 53.1 ° and 63.4 °, 90 °.(P1 P2) clearly has 31 ° angle to the sampled measurements angle ∠ of received signal P1 and P2.In this example, 26 ° nominal angle is included in the possible utmost point that is used for preliminary election directly collects.By means of this chart, also promptly be used for the arithmetic chart of frequency correction subsequently, (S1, S2) and received signal value P1, (P1, P2) differential seat angle between is 4.4 ° for 31 ° of measured angular ∠ of P2 can to determine 26.6 ° nominal angle ∠ at an easy rate.Consider several differential seat angles of continuous measurement if desired, single singular value also can be averaged or select a differential seat angle in order to ask on average or not consider.In addition, also need to consider differential seat angle between the right respective angles of special angle that the different measuring signal is right and coordinate nominal symbol.

To consider also simultaneously that single angular surveying is used for the relation of pairing continuously, for example uses Viterbi algorithm commonly used.Itself in addition can both produce correct result when having comprised several adjacent poles footpath in high frequency frequency deviation or the differentiation process.When frequency difference was very big, the mistake that wrong nominal angle may take place measured angular is distributed to was differentiated, and therefore the rotation of the coordinate system between two symbols is to depend on last two the adjacent symbols S2 of same utmost point footpath Rh, S2 ^*The order of magnitude of differential seat angle.

Thus obtained angle is differentiated also can discern the symbol Se that is comprised or may insert in preliminary treatment, will consider a plurality of measured angular or the utmost point footpath information that differentiation comprised when especially cascade is differentiated.Therefore, processing sequence or corresponding algorithm have caused having carried out in the Phase synchronization process demodulation or symbol to be determined, does not need at first to realize frequency correction.And, the absolute angle information of the received signal that Symbol recognition obtained in the process of formation difference can be used for the corresponding nominal symbol that phase place is adjusted.

Handle if only in a quadrant, finish angle, for example receiving symbol P1 ^*Move on to first quartile from second quadrant, in order further directly to handle determined symbol in treatment facility subsequently, suitable quadrant is got back in rotation in the Phase synchronization process of circuit arrangement 1 then.

After frequency difference between carrier wave to received signal and the local oscillator 7 and differing compensated, suggestion sent the phase difference signal Δ γ that obtains in the ordinary symbol differentiation processes to carrier Control unit 8, with locking carrier phase coupling circuit.Connect carrier Control circuit 8, even like this after angular error Δ γ switches to phase regulator as input signal, the frequency difference information of original frequency adjusting, for example the information in the integrator 84 can be kept perfectly.

Those skilled in the art can construct different replacement circuit devices and realize this method.For example clock controller 21 can provide central authorities to adjust sampling and digitlization clock cycle, and its form is sampled signal t _i, be sent to AD converter 3.If sampled signal t _iSynchronous with the clock cycle of receiving symbol or signal sa, just do not need to carry out interpolation arithmetic at sample devices 10 subsequently.Therefore low pass filter 9 after the quadrature converter 6 also just no longer needs.The available low pass 11 with nyquist characteristic of its qualification effect replaces.Especially in the device of alternate embodiment, can only after quadrature converter 6, provide digital interface, i.e. A/D converter, for example the intermediate frequency position of the input signal sa that provides of signal source 2 is too high.

In the processing procedure, and according to Fig. 4, beginning receiving symbol P1 in the step 401 after the beginning, P2.Afterwards, in step 402, determine one group of potential nominal utmost point footpath for receiving symbol P2.If first signal or symbol just turn back to query steps 404.If first signal, first symbol P1 is set to equal receiving symbol P2 in the step 406, and determined one group of nominal utmost point directly is defined as first utmost point and directly organizes in the 4th step 402.This method proceeds to step 406 then, received signal value or symbol P2.

If step 404 does not comprise first signal, and the differential seat angle ∠ in the step 408 between definite signal value or the symbol (S*7) (P1, P2).Then in step 410, determine first utmost point directly organize in nearest nominal angle ∠ in all combinations of symbol S2 directly organized of symbol S1 and second utmost point (S1, S2).((P1 P2)-∠ (S1, S2)), infers in the step 412 and draws frequency deviation Δ f ∠ according to differential seat angle at last.Frequency deviation Δ f is eliminated.After this, turn back to step 406 and the receiving symbol that basic parameter is set.

Although the present invention has shown and has described several preferred embodiments, its form and details are carried out various variations, omission and interpolation all do not depart from the spirit and scope of the invention.

Claims (13)

1. One kind in modulation technique the complex phase bit space (I, in Q) be the demodulated received signal value (P1 P2) determines the method for received signal frequency, wherein in order to determine frequency, the received signal value (P1, P2) and complex signal space (I, Q; R, α) Nei nominal position (S ₁₁And S ₅₇) symbol (S that locates ₁, S ₂) compare, the step that this method comprises has:

   Determine the angle (∠ (P1, P2)) between signal value P1 that first receives and second the signal value P2 that receives; And

   With determined angle (∠ (P1, P2)) and a plurality of nominal angles (∠ (S relevant with modulation technique _Xy, S _Xy)) compare, and determine the most approaching determined angle of which nominal angle in described a plurality of nominal angles, measure as of frequency deviation (Δ f).
2. 2. the method described in claim 1, it is characterized in that determined angle (∠ (P1, P2)) receives first and second signal value P1 according to predetermined interval, and P2 is determined, and with corresponding possible nominal angle (∠ (S _Xy, S _Xy)) compare.
3. 3. the method described in claim 1, it is characterized in that determined angle (∠ (P1, P2)) be to be used for each possibility utmost point footpath (Ra of quadrature modulation method (QAM), Rh) scope is interior to first and second received signal value P1, P2 is determined, and with nominal angle (∠ (S1, S2)) compares.
4. 4. the method described in claim 1, it is characterized in that determined angle (∠ (P1, P2)) be for the utmost point measured footpath the clearest and the most definite in the symbols alphabet (Ra, Rb, Rh, Ri) the received signal value on is determined.
5. 5. the method described in claim 1 is characterized in that (∠ (P1, P2)) is that (for received signal value P1, P2 is determined in the S1, position range S2) at selected nominal symbol for the angle of institute's exactness.
6. 6. the method described in claim 1 is characterized in that in order to unite and/or iteration strengthens frequency and determines, has determined that (∠ (P1, P2)) compares a plurality of angles, and unifiedly considers so that determine frequency deviation (Δ f).
7. 7. the method described in claim 1, it is characterized in that differential seat angle (∠ (and S1, S2)-∠ (P1, P2)) is used for disposable frequency correction and demodulation and/or as the input signal of FREQUENCY CONTROL in the future.
8. 8. the method described in claim 1 is characterized in that (information that obtains among the ∠ (P1, P2)) is used to discern relevant symbol from angle.
9. 9. the method described in claim 6 is characterized in that receiving symbol (P1 ^*) location map (I is used for determining angle in single quadrant Q) to complex plane.
10. 10. the method described in claim 9 is characterized in that only determining in a quadrant after the angle, and the sign of determined symbol suitably superposes and returns or this symbol rotates in the suitable quadrant once more.
11. 11. the method described in claim 9, it is characterized in that only considering in the margin of tolerance (Δ R) in possible nominal utmost point footpath the received signal value (P1, P2).
12. 12. the method described in claim 11 is characterized in that (P1 P2) has the coordinate that the utmost point is directly organized, and is used to define the selection of nominal angle the received signal value in the margin of tolerance (Δ R) in possible nominal utmost point footpath.
13. 13. one kind be used for determining received signal in the complex phase bit space (P1, the circuit arrangement of the frequency of demodulator circuit device P2) comprises:

    Frequency is determined equipment, and it determines two received signal values (P1, P2) angle between (∠ (P1, P2)) on the complex plane;

    The nominal angle management equipment, it has stored distinct symbols (S1, the nominal angle of combination S2) (∠ (S corresponding to nominal position in the quadrature modulation method (QAM) _Xy, S _Xy)), and optionally storing coordinate utmost point footpath (Ra ..., Rh); And

    Adjuster, (∠ (P1, P2)) distributes at least one contiguous corresponding nominal angle (∠ (S1 in the memory with determined angle, S2)), and determine between the above-mentioned angle differential seat angle (∠ (and S1, S2)-∠ (P1, P2)) measures as of frequency deviation (Δ f).

Images