MXPA98006794A - Method and apparatus for the demodulation of signal and the combination in diversity of modulated signals ortogonalme - Google Patents

Abstract

The presinvon relates to combining diversity path signals comprising symbols, each modulated in accordance with a function of a plurality of orthogonal functions (eg, Walsh), a diversity combiner includes, for each path, a demodulator ( 40, 42, 44, 46) to demodulate each modulated signal symbol in accordance with a selected function (K) of the orthogonal functions, a phase estimator (56, 58) to estimate the phase rotation and the amplitude of the signal of the diversity path of the demodulated signal and a complex signal multiplier (52) for desroting the phase and weighting the amplitude of the modulated path signal in diversity depending on the estimated phase rotation and amplitude. The combiner sum (30) the real parts of the unlaced phase signals and the weighted modulated signals of the diversity paths, demodulates (32) the combined signal in accordance with all the orthogonal functions and selects (34, 36) the maximum demodulated signal of each symbol to determine in this way for the symbol the selected function (K) of the orthogonal functions

Description

METHOD AND APPARATUS FOR THE DEMODULATION OF SIGNAL AND COMBINATION IN DIVERSITY OF ORTHOGONALLY MODULATED SIGNALS TECHNICAL FIELD AND ICABI IDAD INDUSTRIAL APPLICATION This invention is related to the signal demodulation and the diversity combination in a communications system that uses orthogonal modulation. Although the invention is generally applicable to communication systems, it is particularly applicable and is described below by way of example in relation to the signal demodulation and the diversity combination for the reverse channel or the uplink (from a mobile station to a base station) of a direct sequence code division multiple access (DS-CDMA) cellular communications system that is compatible with the interim standard IS-95-A "Mobile Station-Base 'Station Compatibility Standard for Dual-Mode ideband Spread Spectrum Cellular System "of the TIA / EIA (Telecommunications Industry Association / Electronic Industries Association), referred to below for convenience simply as an IS-95 system. As is known, the reverse channel of an IS-95 system uses orthogonal 64-area modulation.

Pl3e / 98 X BACKGROUND OF THE INVENTION In a DS-CDMA mobile cellular communication system, two significant channel impairments are co-channel interference from other users and fading of the signal including the Doppler frequency shifts, when the Mobile station is in motion. The modulation and coding schemes used in the system allow a desirably low probability of error to be achieved or obtained for an SIR (signal to interference ratio) given that it corresponds to a certain capacity, that is, a certain number of users or mobile stations of the system. It is known that coherent detection techniques are preferable to non-coherent techniques for the optimization of the SIR and, hence, provide an optimum capacity of the system, because coherent detection techniques allow the different signals in diversity to be combined with others in phase. It is known to provide a communication system in which coherent detection is achieved by combining diversity signals after aligning them in phase using PLL (phase-locked) techniques. However, PLL techniques can not work properly in fading environments, such as in cellular communications systems, due to a P136 / 9TMX increase in the speed of the cycle jump. It is also known, for example, in an IS-95 system to provide a pilot signal in the direct channel or in the downlink (from the base station to the mobile stations) to facilitate coherent detection in the receivers of the mobile station. However, the use of this common pilot signal is not feasible in the reverse or uplink channel. The article by F. Ling entitled "Coherent Detection With Reference-Symbol Based Channel Estimation For Direct Sequence CDMA Uplink Communications ", IEEE Vehicular Technical Conference, VTC '93, pages 400-403, May 1993, proposes to insert reference symbols (pilot) at a relatively high rate or speed of 1 for every 6 symbols in the reverse channel to facilitate coherent detection. However, this technique is not compatible with an IS-95 system. In accordance with the above, it is presumed that only non-coherent detection techniques are practical for the reverse channel in an IS-95 system. An approach for the non-coherent demodulation of modulated PSK signals (Phase Displacement Manipulation) with a performance that tends towards that of coherent demodulation is described in an article by H. Leib et al. Entitled "The Phase Of A Vector Perturbed By P1364 / 98MX Gaussian Noise And Differentially Coherent Receivers, "IEEE Transactions on Information Theory, Vol. 34, No. 6, pages 1491-1501, November 1988. This approach is generalized in an article by H. Leib entitled "Data-Aided Noncoherent Demodulation of DPSK", IEEE Transactions on Communications, Vol. 43, No. 2/3/4, pages 722-725, dated in February / March / April 1995 and published on or about April 24, 1995. However, these approaches are not related to orthogonally modulated signals. The International Patent Application (WO-A-95/01018), published on January 5, 1995 and entitled "Receiver for a direct sequencce spread spectrum prthogonally encoded signal employing rake principie" reveals a method, known as the dual maximum metric generation and the apparatus for decoding a data signal orthogonally encoded in a non-coherent receiver system, such as for the reverse channel of an IS-95 system. An object of this invention is to provide methods of signal demodulation and diversity combining in a communications system using orthogonal modulation, especially for the reverse channel in an IS-95 system and with the apparatus for effecting the methods.

P1364 / 98MX SUMMARY OF THE INVENTION In accordance with one aspect, this invention provides a method for processing a modulated signal in a communication system using orthogonal modulation, comprising the steps of: demodulating the modulated signal in accordance with each of a plurality of orthogonal modulation functions to produce a plurality of demodulated signals; select an optimal signal from the demodulated signals; estimate a phase rotation of the selected demodulated signal; and desrotar the phase of the modulated signal depending on the estimated phase rotation. A related aspect of the invention provides the apparatus for processing a demodulated signal in a communications system that uses orthogonal modulation, comprising: a phase shunt that responds to the modulated signal and a phase estimate of the modulated signal to clear the phase of the modulated signal depending on the phase estimate to produce a demodulated phase-shifted signal; a scrambler a arranged to demodulate the phase-modulated phase signal retracted in accordance with each of a plurality of orthogonal modulation functions, to produce a plurality of demodulated signals; a selector arranged to select an optimum signal from the demodulated signals; P1364 / 98MX and a phase estimator that responds to the demodulated signal selected to produce the phase estimate. Another aspect of the invention provides a method for combining diversity path signals comprising symbols, each modulated in accordance with a function of a plurality of orthogonal functions, comprising the steps of: for each modulated path signal in diversity, producing a demodulated signal corresponding to the demodulated modulated signal according to a function selected from the orthogonal functions, estimating a phase rotation of the modulated signal from the demodulated signal and desroting the phase of the modulated signal depending on the phase rotation Dear; combining the modulated phase-shifted signals of the diversity trajectories to produce a modulated signal combined in diversity; demodulating the combined modulated signal in diversity according to each of the orthogonal functions, to produce a plurality of demodulated signals combined in diversity; and determining the maximum of the demodulated signals combined in diversity to determine in this way the selected function of the orthogonal functions. This method preferably also includes the steps of, for each modulated path signal in P1364 / 98MX diversity, estimate an amplitude of the modulated signal and weight the modulated trajectory signal in diversity, before the combination step, depending on the estimated amplitude. Advantageously, the combination step comprises summing only the real parts of the complex signals representing the modulated phase-shifted signals. The step of estimating a phase rotation conveniently comprises, for each diversity-path modulated signal, averaging the demodulated signal of a plurality of symbols. The invention also provides a diversity path signal combiner comprising: for each of a plurality of diversity path signals comprising symbols each modulated in accordance with a function of a plurality of orthogonal functions, an arranged demodulator or arranged to demodulate each modulated signal symbol in accordance with a function selected from orthogonal functions, to produce a demodulated signal, a phase estimator arranged or arranged to estimate a phase rotation of the modulated path diversity signal from the demodulated signal and, a phase derailter arranged or arranged to clear the phase of the modulated signal of P1364 / 98MX path in diversity depending on the estimated phase rotation to produce a modulated phase-shifted signal; a signal combiner arranged to combine the modulated phase-shifted signals of the diversity paths; a demodulator arranged to demodulate a combined signal, produced by the signal combiner, in accordance with the plurality of orthogonal functions to produce a plurality of demodulated signals combined in diversity; and a means for determining, for each symbol, an optimum signal of the demodulated signals combined in diversity to thereby determine the symbol of a selected function of the orthogonal functions.

BRIEF DESCRIPTION OF THE DRAWINGS OR FIGURES The invention will be further understood from the following description with reference to the accompanying drawings, in which: Figure 1 schematically illustrates a block diagram of a quadrature Walsh chip demodulator for a diversity path in a reverse channel receiver of an IS-95 cellular communications system; Figure 2 schematically illustrates parts of a diversity trajectory combiner and a P1364 / 98MX signal demodulator, in accordance with one embodiment of this invention, of the system; and Figure 3 illustrates a general block diagram of a signal demodulator in accordance with an embodiment of this invention.

FORMS FOR CARRYING OUT THE INVENTION As is known, in an IS-95 system, the data for transmission in the reverse channel, from a mobile station to a base station, at a rate or bit rate of 9.6 kilobits per second (kbps) they are encoded convolutionally in a 1/3 speed convolutional encoder to produce code ols at a rate or speed of 28.8 kilograms of code per second (ksps). After the block interleaving, the code ols at a rate of 28.8 ksps are modulated using 64-area orthogonal modulation, one of the 2 ^ = 64 mutually orthogonal waveforms generated using Walsh functions will be produced for each group of 6 code ols, to produce modulation ols at a rate of 4.8 ksps modulation. Each modulation ol is referred to as a Walsh ol and consists of 64 Walsh chips, a Walsh chip will be the shortest identifiable component of a Walsh function, and there will be 2N Walsh chips in a Walsh function of order N. Thus, the Walsh ols P1364 / 98MX have a Walsh chip rate of 307.2 kilochips per second (kcps). The Walsh chips at the speed of 307.2 kcps, undergo direct sequence dispersion by adding module 2 of each Walsh chip with the so-called long code to produce a chip rate of 1228.8 kcps and, the resulting chips are subjected to the quadrature dispersion using I and Q sequences (in phase and quadrature phase), PN (pseudo noise) and then filtering the baseband and transmission. A receiver is required at the base station to perform the complement of these operations. In order to facilitate reception, the receiver is implemented with four diversity trajectories, referred to as fingers, and each finger can be provided with an input from multiple sources, for example, from six sources comprising two antennas in each of three sectors of a cell. In order to maximize the capacity of the system, it is desired to combine the signals of the trajectories in diversity or fingers in an optimal way. As explained in the background of the invention, coherent detection, which includes the combination of finger signals in phase with one another, would be desirable but not feasible. In the prior art, the signals have been combined in a non- P1364 / 98MX consistent. More particularly, the magnitudes of the real and imaginary components of these signals have been squared and summed, the phase information is not used. Referring to the drawings, Figure 1 illustrates the known concept of a quadrature Walsh chip demodulator, which serves to produce Walsh chips in its output; this illustration is included for a complete understanding of the context of the invention. One of these demodulators is provided for each finger. Figure 2 illustrates a diversity trajectory combiner or finger combiner, comprising a plurality of units 10, one for each finger and in Figure 2 only one unit 10 is shown within a dotted line box and, the additional components which are common to the fingers and which are described further below. The output of the Walsh chip demodulator in quadrature of FIG. 1 constitutes the input to a respective unit of the units 10 of FIG. 2. Referring to FIG. 1, in the Quadrature Walsh chip demodulator two mixers 12 are supplied with quadrature signals of phase (cosine and sine) and a signal received on a line 14 and, its output signals are filtered by filters 16, comprising, for example, each, a low pass filter and a filter P1364 / 98MX compensated and sampled at the chip rate of 1228.8 kcps using the samplers 18 to produce respectively the phase quadrature signal samples X1 (m) and Xo- (m), where m is an integer Index of each sample and I and Q respectively denote the phase and quadrature phase components. The samples are multiplied in the multiplying units 20 by the sequences a1 (m) and aQ (m) which are constituted by the long code combined with the sequences I and Q PN, respectively and, the corresponding quadrature outputs of these units 20 are summed in summing units 22 to provide quadrature disperse chips. Groups of four successive chips that leave the summing units 22 are summed in the additional summing units 24 to produce the phase and phase (or real and imaginary) phase v (p) and vQ (p) components, respectively, of Walsh chips identified by a whole index p, at the Walsh chip rate of 307.2 kcps. Referring now to Figure 2, for each of the four fingers (in an IS-95 system) the Walsh chips produced by a demodulator, for example as described above with reference to Figure 1, are supplied to a respective unit of the four units 10, which are also all provided with an index P1364 / 98MX K that is discussed later. The outputs of the four units 10 are supplied to a summing unit 30 and summed therein, as shown in Figure 2, the adding unit 30 constitutes a combiner of fingers or of trajectories in diversity. The output of the summing unit 30 is supplied to a transformation unit 32, which performs an input conversion of 64 chips in series to parallel (SP) and a Hadamard 64-area transformation (Hg4>, thus constituting a demodulator of Walsh symbols for the combined signal in diversity The transformation unit 32 in accordance with the above has 64 parallel outputs, each of which represents a demodulation of the combined Walsh symbol in present diversity in accordance with a respective function of the 64 functions Walsh The maximum output of these outputs is determined by a unit of maximum (MAX) 34 that produces at its output the index K that identifies this maximum output, as shown in Figure 2, in addition to being supplied to each unit 10 , the index K is supplied as a control input to a selection unit 1 of 64 36 to which the outputs of the transformer unit are also supplied. n 32. The selection unit 36 in accordance with the above, supplies the determined maximum output of the transformation unit 32, which P1364 / 98MX constitutes the demodulated Walsh symbol, towards an output line 38. The output line 38 leads to a block deinterleaver and a subsequent convolutional decoder (not shown) which in known manner performs or performs the opposite of the block interleaving. and the convolutional coding discussed above. As an alternative, the entire vector transformed from the outputs of the transformation unit 32 can be supplied by the block deinterleaver to the convolutional decoder to provide an optimal combination of Walsh-Hadamard demodulation and convolutional decoding. Figure 2 shows in detail one of the fingers of the unit 10; the units 10 of the other fingers are identical to this. Each unit 10 comprises two transformation units 40 and 42, each of which is identical to the transformation unit 32 described above 'and, thus comprises a serial to parallel input conversion (SP) and a Hadamard transformation 64 - area (H54), to which the real and imaginary components v1 (p) and vQ (p) of the Walsh chips in the inputs of unit 10 are supplied respectively. The unit 10 further comprises two selection units 1 of 64 44 and 46, each of which is controlled by the index K supplied to the unit 10 from the unit of maximum 34.

P136 / 98MX to select a respective output of the 64 outputs of units 40 and 42, respectively. The outputs of the selection units 44 and 46 together constitute the real and imaginary components of a demodulated Walsh symbol of complex signal that is supplied to a gain and phase estimation unit 48 shown within a dashed line box and further described later. The estimation unit 48 produces at its output on lines 50, the real and imaginary components of a complex signal which is the complex conjugate of an estimate of the complex signal supplied to the estimation unit from selection units 44 and 46. In other words, if an estimate of the complex signal supplied by the selection units 44 and 46 is represented by AeJ ", where A denotes an estimated amplitude and? Represents an estimated phase, then the output of the estimation unit in the lines 50 is represented by AeJ. The unit 10 further comprises a complex signal multiplier 52 which is arranged or arranged to multiply the complex signal of incoming Walsh chips by the complex signal on lines 50 and a real function unit (REAL). 54 to which the product of the complex signal multiplier 52 is supplied.The unit 54 supplies the actual component of this signal product.

P1364 / 98MX complex to the adding unit 30 as the contribution of the respective finger, the adding unit 30 combines the contributions of the four fingers as described above. In the form of the estimation unit 48 illustrated in Figure 2, this unit is constituted by a plurality of delay elements 56, each providing a delay of Tw equal to the Walsh symbol time of 1 / 4.8 ksps = 208.3 μs . The delay elements 56 form derivative delay lines for the outputs of each of the selection units 44 and 46, the sockets or leads provide inputs to a complex averaging and conjugating unit 58. The unit 58 forms an average of the signals complexes represented by the real and imaginary components of the signal in the taps or derivations of the delay line, which constitute the complex signal AeJ "referred to above and produces the complex conjugate Ae ~ LJ? of this average in the form of real and imaginary components of the signal on the lines 50. In other words, if the outputs of the selection units 44 and 46 are the real and imaginary components of the signal zI (nl) and z ° - (nl), respectively, where n is a entire index of a normal or common Walsh symbol comprising Walsh chips that normally enter the unit P136 / 98MX 10, these together represent the complex signal z (n-l) + jzQ (n-l). If each delay line has a whole number L of delay elements 56, then the complex signal AeJ "is defined by the equation: 1 L + l AeJ? =. { } ? . { zI (n-r) + jzQ (n-r)} L + lr = l and the signal components in lines 50 are the real and imaginary parts of their complex conjugate Ae ~ J ". By way of example, L can conveniently be in the range of from about 10 to 12 o, can be a convenient power of two such as 8 or 16. In the operation, the transformation units 40 and 42 and the selection units 44 and 46 perform, in accordance with the recursively determined K index, the demodulation of each Walsh symbol of incoming complex signal to provide a complex signal demodulated to the gain and phase estimation unit 48. The unit 48 provides the complex output signal Ae "J" on the lines 50 having an amplitude A which is an estimate or an average of the amplitude of the previously demodulated Walsh symbols L + l and, hence, represents an amplitude of the signal received from the respective finger and, has a phase -? which is the inverse of a displacement of the carrier phase or the phase rotation? of the chips Pl3e4 / 98MX Walsh incoming from the real axis in a phase plane representation of the complex signal Walsh chips. The averaging serves in part to reduce the effects of noise and the value of L is selected in accordance with the above. In the multiplication unit 52Each Walsh chip of incoming complex signal is multiplied by the complex signal Ae ~ J® on lines 50. This communication provides the phase desrotation of each Walsh chip by the inverse -? of your phase rotation? of the real axis, with which each Walsh chip is aligned in phase with the real axis. At the same time, this multiplication provides a weighting of the amplitude of each Walsh chip in accordance with the amplitude of the average signal for the finger. Because the same processes occur in units 10 for all the fingers, the resulting complex signals in the outputs of the multiplication units 52 in the different units 10 are all substantially aligned in phase and with weighted amplitude in accordance with the amplitudes of the average signal of the respective fingers. Therefore, these can be combined by summing them to produce a result that is to an important degree, equivalent to a coherent combination of the Walsh chips of the different fingers. Because the outputs of the unit P1364 / 98MX multiplication 52 is a complex signal that includes noise and interference components, as well as the desired signal that is substantially aligned with the real axis, the actual function unit 54 is provided to pass through to the output of the unit 10 only the real part of this complex signal, whereby the improvement in the signal-to-noise ratio is supplied by suppressing the imaginary component of the complex signal comprising mainly noise and interference. The output of the actual function unit 54 then constitutes the output of the unit 10 and is a real signal that is supplied as described above to the adding unit 30, where it is combined with the other Walsh chip signals weighted from the other fingers by a simple sum of these real signals. Because the phase alignment and the weighting of the amplitude of the signals in the multiplication units 52, the output of the summing unit 30 constitutes an optimized combined sequence in diversity of Walsh chips aligned in phase with the real axis. As already described above, the Walsh symbols constituted by this sequence are demodulated by the transformation unit 32 in accordance with all the Walsh functions and the maximum is determined by the unit 34 to determine the index K.

P1364 / 98MX An advantage of this arrangement is that it provides a particularly good gain for situations in which the mobile transmission station is stopped or moves slowly, when the convolutional coding gain is relatively low. Conversely, for the rapid movement of the mobile station, when the convolutional coding gain is greater, the gain of that arrangement is smaller due to a less precise amplitude and phase estimate. This arrangement and convolutional coding together provide a complementary and particularly beneficial combination. The functions of the diversity combination array of Figure 2 may all be implemented conveniently, possibly together with other functions of the communication system, in one or more integrated circuits of the digital signal processor. It can be appreciated that different forms of gain and phase estimation in addition to the one described above in relation to unit 48 can be used. For example, instead of the simple averaging as described above, it can be seen that the array could instead use the recursive averaging, Kalman filtering or any other averaging or filtering process convenient to provide the desired amplitude and phase estimates. Additionally, although the P1364 / 98MX unit 48 as described above provides these amplitude and phase estimates in a single process, these estimates could instead be provided separately. For example, the amplitude could be estimated by adding the squared amplitudes of the demodulated signals at the outputs of units 44 and 46 or at the outputs of units 40 and 42 and, the phase could be estimated separately from this estimated amplitude and the amplitude at the output of the unit 46. Furthermore, although as previously described the Walsh chips of each finger are weighted or weighed directly depending on the estimated amplitude being multiplied by it in the multiplication unit 52, any other desired form of weighting, for example the dependent of a non-linear function such as the square of the estimated amplitude or, even without any weighting (although this is not what is preferred), it could be used instead. In addition, the pass and weight derotation functions, combined in the multiplication unit 52 as described above, can, if desired, be performed separately from one another. It can also be appreciated that the real function unit 54 can be omitted and that Walsh complex signal chips can be combined and demodulated instead of using only the actual parts of these signals P1364 / 98MX as described above, although, again this is not what is preferred, due to the noise cancellation benefits and the relative simplicity of the arrangement as described above. It can also be seen that, although as previously described, the Walsh chips of each Walsh symbol are demodulated in accordance with all orthogonal Walsh functions by the transformation units 40 and 42 in synchronism with the demodulation of the Walsh symbol combined in diversity by the unit of transformation 32, with the subsequent selection by the selection units 44 and 46 in accordance with the index K determined for the respective symbol, this is not necessarily the case. Instead, for example, a delay Tw can be provided by a Walsh symbol time between the incoming Walsh chips and the transformation units 40 and 42, these units will respond to the K Index (the separate selection units 44 and 46 will be dispatched). ) to demodulate the Walsh symbol of conformity only with the selected Walsh function determined by the K index. This alternative reduces the computation required for each unit 10 but, introduces a delay Tw in the estimation of the gain and the phase; however, this delay may not be significant if the gain and phase change relatively slowly.

P1364 / 98MX Furthermore, although the above description- relates specifically to an IS-95 system, it can be appreciated that the invention can also be applied to other communication systems that use orthogonal modulation in which the diversity path signals are combined, each diversity path signal are aligned in phase when performing the phase de-rotation dependent on a phase estimate of the incoming signal and, in a desirable but optional way, will also be weighted depending on an amplitude estimate of the incoming signal. It can be appreciated that the invention can also be applied to communication systems in which there is not necessarily any combination in diversity but simply the desire to phase-align an orthogonally modulated incoming signal before the demodulation, to achieve, for example, the reduction of the noise by using the actual function unit 54, as described above. With only one signal, the array would require only one unit 10 and, the above-described weighting and summing unit 30 would be dispatched with it. The array can then be simplified if desired by making it recursive, and units 32 and 36 dispatched, the maximum 34 unit instead of having its inputs supplied from the outputs of the P1364 / 98MX unit 40, and units 40 and 42 will be supplied with the outputs of the phase de-rotation unit (multiplication unit 52) instead of the incoming signal. This arrangement is illustrated in the form of a general block diagram in Figure 3. Referring to Figure 3, in which, for simplicity, simple lines are shown for transporting possibly complex signals, the incoming orthogonal modulation symbols are supplied by a phase shifter 60 to a demodulator 62 that demodulates each conformance symbol with each of the orthogonal modulation functions to produce a plurality of demodulated outputs, the optimum of these (e.g., the maximum) is determined by a unit 64 that controls a selector 66 for selecting this optimum demodulated signal as an output. The phase rotation of this output signal is determined by a phase estimator 68, whose output is used to control the phase shunt 60 to produce a reverse phase rotation. Thus, although the particular embodiments of the invention have been described in detail, it should be appreciated that these and numerous other modifications, variations and adaptations may be made within the scope of the claims.

P1364 / 98MX

Claims (17)

1. NOVELTY OF THE INVENTION Having described the present invention, it is considered as a novelty and, therefore, the content of the following CLAIMS is claimed as property: 1. A method to process a modulated signal in a communication system using orthogonal modulation , comprising the steps of: demodulating the modulated signal in accordance with each of a plurality of orthogonal modulation functions to produce a plurality of demodulated signals; select the optimal signal of the demodulated signals; estimate the phase rotation of the selected demodulated signal; and desrotar the phase of the modulated signal depending on the estimated phase rotation. A method according to claim 1, wherein the step of the desrotation of the phase of the modulated signal is effected in the modulated signal before the step of demodulating the modulated signal. 3. A method according to claim 1 or 2, wherein the modulated signal comprises Walsh symbols and the step of the demodulation comprises performing or effecting a P136 / 98MX Hadamard transformation. 4. A method for combining diversity path signals comprising symbols each modulated in accordance with one of a plurality of orthogonal functions, comprising the steps of: for each diversity-modulated path signal, producing a demodulated signal corresponding to the modulated signal, demodulated in accordance with a function selected from the orthogonal functions, estimate the phase rotation of the modulated signal from the demodulated signal and clear the phase of the modulated signal depending on the estimated phase rotation; combining the modulated phase-shifted signals of the diversity trajectories to produce a modulated signal combined in diversity; demodulating the combined modulated signal in diversity, in accordance with each of the orthogonal functions' to produce a plurality of demodulated signals combined in diversity; and determining the maximum of the demodulated signals combined in diversity to determine in this way the selected function of the orthogonal functions. 5. A method according to claim 4 and further comprising, for each modulated trajectory-diversity signal, the steps of estimating the amplitude of the signal P1364 / 98MX modulated and weighted the modulated trajectory signal in diversity, before the combining step, depending on the estimated amplitude. 6. A method according to claim 5, wherein, for each diversity-modulated path signal, the step of shunting the phase and of weighting the amplitude of the modulated signal comprises multiplying the signal modulated by a complex conjugate of an estimated derivative of the signal demodulated. 7. A method according to claim 4, 5 or the 6, wherein the step of combining comprises summing only the real parts of the complex signals representing the modulated signals of the unrooted phase. A method according to claims 4 to 7, wherein, for each diversity path modulated signal, the step of producing the demodulated signal comprises the steps of demodulating the modulated signal in accordance with each of the orthogonal functions to produce a plurality of demodulated signals and selecting a signal from the plurality of demodulated signals as the demodulated signal. 9. A method according to any of claims 4 to 8, wherein, for each diversity path modulated signal, the step of estimating the phase rotation comprises averaging and demodulating the signal P1364 / 98MX for a plurality of symbols. A method according to any of claims 4 to 9, wherein the modulated signals comprise Walsh symbols and the step of demodulating comprises performing a Hadamard transformation. 11. The apparatus for processing a modulated signal in a communications system that uses orthogonal modulation, comprises: a phase shunt that responds to the modulated signal and a phase estimate of the modulated signal to clear the phase of the modulated signal, depending of the phase estimate to produce a modulated phase-shifted signal; a demodulator arranged or arranged to demodulate the modulated phase-shifted signal in accordance with each of a plurality of orthogonal modulation functions to produce a plurality of demodulated signals; a selector arranged or arranged to select an optimum signal from the demodulated signals; and a phase estimator that responds to the demodulated signal selected to produce the phase estimate. 12. The apparatus according to claim 11, wherein the phase derailter comprises a multiplier of P1364 / 98MX complex signal. The apparatus according to claim 11 or 12, wherein the phase estimator comprises an averager. 14. A diversity path signal combiner comprising: for each of the plurality of diversity path signals comprising symbols, each modulated in accordance with a function of a plurality of orthogonal functions, a demodulator arranged to demodulate each modulated signal symbol in accordance with a function selected from orthogonal functions to produce a demodulated signal, an arranged phase estimator for estimating the phase rotation of the diversity path modulated signal from the demodulated signal and a phase derailor arranged to clear the phase of the modulated path signal in diversity depending on the estimated phase rotation to produce a modulated phase-shifted signal; a signal combiner arranged or arranged to combine the modulated phase-shifted signals of the diversity paths; a demodulator arranged or arranged to demodulate a combined signal, produced by the signal combiner, in accordance with the plurality of orthogonal functions to produce a plurality of signals P1364 / 98MX demodulated combined in diversity; and a means for determining, for each symbol, an optimum signal of the demodulated signals combined in diversity to thereby determine the symbol of the selected function of the orthogonal functions. 15. A diversity path signal combiner according to claim 14, and including, for each of the diversity path signals, a means for weighting the modulated phase-shifted signal depending on an estimate of the amplitude of the respective Diversity trajectory sign. 16. A diversity path signal combiner according to claim 15, wherein for each of the diversity path signals, the phase shifter and the weighting means are constituted by a complex signal multiplier. 17. A diversity path signal combiner according to claim 14, 15 or 16, wherein the signal combiner comprises a summing unit arranged to sum only the real parts of the complex signals representing the modulated phase-shifted signals. P1364 / 98MX