FR3009911A1 - METHOD FOR TRANSMITTING A CDMA SIGNAL, TRANSMITTING DEVICE AND CORRESPONDING COMPUTER PROGRAM. - Google Patents

Abstract

The invention relates to a method for transmitting a CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M distinct data streams. According to the invention, the method comprises, for at least one block of M constellation symbols: a pre-construction phase (201) of SF complex temporal samples, representative of the M data streams multiplied by a spreading factor spreading code SF, a correction phase (202) of the constellation symbols, comprising for each data stream: - detection (2021) of P samples greater than a predetermined threshold (α) delivering said P samples and SF-P null samples, constituting SF samples to be corrected, - correction (2022) of the constellation symbol according to of said SF samples to be corrected, delivering a correction (Fm), - construction (2023) of SF samples associated with said correction, - updating (2024) of said SFs ec preconstructed gloves.

Description

Method for transmitting a corresponding CDMA signal, transmission device and computer program FIELD OF THE DISCLOSURE The field of the invention is that of radio frequency communications for which code division multiple access, CDMA or "Code Division Multiple Access": CDMA in English, is implemented. More specifically, CDMA modulation is one of the most used for digital transmission. This modulation technique makes it possible to make a narrow-band signal more resistant to the disturbances of a radio propagation channel by a Direct Sequence Spread Spectrum (DSSS) technique. Such a technique also allows several users to share the same band without disturbing, a significant increase in transmission rates is accordingly obtained.

Because of this advantage, CDMA modulation is particularly, but not exclusively, used for third-generation cellular mobile radiotelephony, in particular for CDMAone, CDMA 2000 in the United States, and UMTS in Europe standards ("Universal Mobile Telecommunications System"). ) based on Wideband Code Division Multiple Access Evalutation (WCDMA) broadband code technology and standardized by 3GPP (3rd Generation Partnership Project). 2. PRIOR ART 2.1 CDMA modulation CDMA modulation consists in simultaneously transmitting, within a so-called "CDMA" signal, the sum of data streams, each stream being for example associated with a distinct user or service and previously multiplied by a spreading code whose code length is called the Spreading Factor SF, the spreading codes used for each data stream being ideally orthogonal to each other. In addition, each data stream can be modulated by a different type of constellation from one data stream to another. As a result, the data streams are independent of each other and are able to be transmitted simultaneously in the same frequency band within the CDMA signal obtained. In reception, each particular data stream transmitted by the source is reconstructed by multiplying the received CDMA signal (the latter summing the orthogonal signals transmitted) by the particular code used for that particular data stream.

Furthermore, the CDMA modulation requires a separate power adjustment of each signal, realized by multiplying the complex signal by a real factor, in order to adapt to the various distinct propagation conditions for each transceiver pair. 2.2 Disadvantages of CDMA A major disadvantage of the CDMA technique is inherent in the summation of the different orthogonal signals associated with distinct users which produces large fluctuations in the instantaneous amplitude of the envelope of the modulated signal. Indeed, in the time domain, the summation of the data flows associated with distinct users respectively multiplied by separate spreading codes takes place in power most of the time, but also sometimes coherently, which leads to peaks of instantaneous amplitude. The peak factor, which characterizes the level of these amplitude peaks with respect to the effective value of the signal, is thus generally very high and increases with the number of users N. 2.3 Definition of the peak factor The peak factor C of the CDMA signal is expressed as the ratio between the maximum peak amplitude of the CDMA signal Ixpicl and its rms value xrms and therefore corresponds to the square root of the peak power to average power ratio (PAPR). Ratio "in English): = Ixpticl = PAPR Xi-MS In practice, signal peaks of a given amplitude occur with a certain probability of occurrence. In particular, it is unlikely that the amplitude of the signal is as large as N, especially since N will be large. Also, in a conventional way, to characterize the PAPR of a CDMA system, the complementary cumulative distribution function (CCDF) is used which provides the probability that the amplitude of the signal exceeds a certain threshold. For example, in the case of a WCDMA signal comprising 64 code-modulated data streams with a spreading factor of SF = 128, the corresponding CCDF curve indicates that for a spreading elementary elementary bit element (called "chip") In English), the peak amplitude exceeds the rms value by more than 7 dB. In other words, this result means that, for a spreading factor of 128 for example, such an overshoot occurs at each transmission period of a symbol of the CDMA signal which is considerable.

Consequently, the operation of the digital-to-analog converters, and especially the transmit power amplifiers, must be adapted to take into account such a peak factor. In particular, the power amplifiers then have the constraint of having to operate at average power levels well below their maximum power to avoid non-linearities and saturation. In addition, in addition to generating intermodulation scrambling adjacent channels, these distortions cause inter-channel interference that degrades the performance of different data links.

In addition, as indicated previously, the data streams used to form the CDMA signal are not always all transmitted with the same power. Interference between data streams will therefore be more penalizing for the data streams that will have been transmitted with the lowest power, and they will also be more penalizing for data streams that will have required high data rates and the use of modulation constellations. at large number of states much more sensitive to disturbances. 2.4 Prior art for the reduction of the crest factor In the literature, numerous techniques have already been proposed to overcome this problem related to the crest factor. A first approach is to select the spreading codes used so that in the space that defines them, they are regularly spaced and not contiguous. However, this technique is limited in terms of effectiveness. Indeed, this approach is effective only when the number of data streams is strictly less than the spreading factor. In other words, this approach is effective only when the number of codes necessary for the modulation is much lower than the number of spreading codes available, in order to select only those which are spaced and non-contiguous. Moreover, even when the constraint of this approach is respected, the reduction of the crest factor is limited and insufficient, which requires that this first approach be combined with another approach allowing the reduction of the crest factor, to obtain a crest factor acceptable.

A second approach is the "clipping" technique, which consists of clipping the amplitude of the CDMA signal when it exceeds a predefined threshold. But this clipping is by nature non-linear and introduces a distortion of the emitted CDMA signal resulting not only in a degraded bit error rate BER but also in a rise in the secondary lobes of the power spectral density, DSP.

A third approach is the compensation pulse technique ("Pulse compensation") which consists of adding a pulse signal in phase opposition with the amplitude peak to be reduced. However, this approach also generates distortions of the CDMA signal within the transmission band.

In this particular context, the inventors have therefore identified a need for a new technique to improve the reduction of the crest factor while remaining simple to implement. 3. DISCLOSURE OF THE INVENTION The invention proposes a new solution that does not have all of these disadvantages of the prior art, in the form of a method for transmitting a code division multiple access signal. said CDMA signal, said CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M distinct data streams forming said CDMA signal, each constellation symbol Xm being identified by an index m, m being an integer such that 1.rrKM .

According to the invention, such a method comprises the following steps, for at least one block, of M constellation symbols of said CDMA signal: a preprocessing step, comprising: a pre-construction phase of SF preconstructed complex temporal samples representative of said M data streams respectively multiplied by a spreading factor distinct SF spreading code, SF being an integer, and o a constellation symbol correction phase modulating the M data streams of said CDMA signal, said correction phase comprising the following steps, repeated for each data flow: detection of P samples, P being an integer, of said SF preconstructed complex temporal samples resulting from said pre-construction phase when m = 1, or among SF complex temporal samples derived from an update step implemented for a previous data stream when 1 <n-KM, having a higher power at a predetermined threshold (a), said detection step delivering said P samples and SF-P null samples, constituting SF complex time samples to be corrected, correction of the constellation symbol modulating said data flow as a function of said SF complex temporal samples to be corrected , delivering a complex correction data item (Fm) of said constellation symbol, construction of SF complex time samples associated with said complex correction data item, updating said SF preconstructed complex time samples, said update implementing an accumulation associating two by two said SFs complex temporal samples associated with said complex correction data with said SF preconstructed complex temporal samples resulting from said pre-construction phase when m = 1, or with said SF complex time samples resulting from an updating step implemented for a flu x of previous data when 1 <n-KM, delivering SF complex time samples used for the next data stream; a spreading step implementing a multiplication of each data stream of said modulated CDMA signal by a constellation symbol resulting from said preprocessing step by a spreading factor spreading code SF; a scrambling step of said CDMA signal resulting from said spreading stage, a step of transmission and / or storage of a CDMA signal originating from said scrambling step. Thus, the invention is based on a novel and inventive approach to the reduction of the peak factor of a CDMA signal. More specifically, the present invention improves the crest factor reduction performance with low computational complexity.

Indeed, the method according to the invention successively modifies and in a controlled manner the constellation symbols modulating the data streams of a CDMA signal before the implementation of a spreading step of multiplying a data stream of index m by a spreading code Sm, the spreading codes used for each data stream being ideally orthogonal to each other.

It should be noted that the term "pre-construction" means that, even though the spreading step has not yet been performed, the temporal samples of the CDMA signal which "could" be obtained beforehand from the transmission and or storage of the CDMA signal. Thus, the term "pre-built" is associated with the complex temporal samples possibly corrected before implementation of the spreading operations, scrambling and optionally power adjustment. By "complex" is meant "which may have a real value and / or imaginary such that this value is for example defined by v = a + j.b with j2 = -1". The invention uses, for the reduction of the PAPR, a real-time servocontrol of the correction of a data flow, referred to as the current data flow of index m, with respect to the previously corrected data streams of the CDMA signal. This slaving is notably based on the implementation for correcting the constellation symbol Xm modulating an index data flow m of a detection of P samples whose amplitude exceeds a threshold a, among the SF complex temporal samples representative of the sum of the temporal responses of the previously corrected and accumulated data flows. Detection occurs after each accumulation. In other words, among the SF preconstructed time samples, the method detects the peak or peaks of power greater than a predetermined threshold a. The value of the threshold a setting the desired final peak factor level.

Then, taking into account the SF complex time samples to be corrected, a complex correction data is obtained. This complex correction data is then used to define the correction to be made to the coordinates of the constellation symbol modulating the current data flow of index m that is to be corrected. In order for this correction to be taken into account for the processing of the following data streams the current data flow of index m, the SF complex time samples associated with the complex correction data are constructed and associated two by two with the SF temporal samples. preconstructed complexes resulting from the pre-construction phase when m = 1, or with SF complex temporal samples resulting from the processing of the previous data stream when 1 <n-KM.

Such a method therefore results in a global correction of the CDMA signal because each data stream of the CDMA signal is corrected, taking into account the power peaks detected successively for each data stream. This correction is optimized because the complex displacement of constellation coordinates is determined as a function of a complex correction data evolving for each data stream as a function of the correction performed on the previous data stream. According to a particular embodiment of the invention, said pre-construction phase implements at least one step of leveling according to which each constellation symbol Xm modulating one of said M data streams is multiplied by a gain factor Gm and in that prior to said step of transmitting and / or storing said method further comprises a power adjustment step whose input is fed by said scrambling step and whose output feeds said transmitting step and / or storage and said leveling step. Thus, optionally, the pretreatment of the data streams, making it possible to reduce the peak factor, takes into account the optional power adjustment of the data streams forming the CDMA signal carried out before the transmission and / or storage of the CDMA signal. . More precisely, a level weighting is implemented for the pre-construction of the temporal samples of the CDMA signal that "could" be obtained prior to the transmission and / or storage of the CDMA signal by multiplying each constellation symbol Xm modulating one of said M fluxes. of data by a gain factor Gm, but also during the updating step by multiplying the SF complex time samples associated with the complex correction data dm of the symbol Xm by the gain factor Gm. According to a particular embodiment of the invention, said complex correction data Fm results from the complex correlation of said SF complex temporal samples to be corrected with SF real time samples associated with spreading code Sm of spreading factor SF multiplying said index data flow m. Such a complex correlation operation indeed makes it possible to elaborate a complex correction data whose taking into account and the evolution at each data flow of the CDMA signal makes it possible to advance the updating of the CDMA signal in the direction of the overall decrease. P detected peaks. Thus, the complex correction data is the result of a complex correlation operation, and is also called complex correlation data. The implementation of the complex correlation according to the invention advantageously makes it possible to vary proportionally the correction of the constellation symbol modulating each data stream of the CDMA signal. Thus, a fixed correction of each constellation symbol is avoided, such a fixed correction being able to limit the reduction of the desired crest factor. Indeed, since the number P of peaks greater than a predetermined threshold may vary from one data stream to another, the complex correlation operation makes it possible to obtain a result different from a data stream at the same time. other taking into account in particular the correction of the correction symbol of the previous data flows. Thus, the implementation of the complex correlation according to the invention allows a fine optimization of the reduction of the PAPR with regard to the techniques of the prior art.

According to a feature of the invention, said transmission method uses a shaping filter whose interpolation factor L is equal to one. According to a feature of the invention, said pretreatment step applies at least one delay line of duration equal to said spreading factor SF to said block of M constellation symbols of said CDMA signal.

According to a particular aspect of the invention, the pre-construction phase comprises the following steps, repeated for each data flow of index m: pre-construction of SF complex temporal samples associated with said data flow of index m multiplied by a spreading code Sm, storing said SF complex time samples associated with said data stream, by accumulating two by two associating said SF complex temporal samples associated with said data stream with SF complex temporal samples associated with the previous data streams. The accumulation of the SF complex time samples associated with a current data flow of index m with SF complex time samples associated with the previous data flows makes it possible, once this accumulation processing applied to the M data streams, to obtain SF preconstructed complex temporal samples representative of the CDMA signal even though the spreading has not yet been carried out. Advantageously, the correction step implements a summation of the coordinates of said constellation symbol, called original coordinates of said constellation symbol, with coordinates representative of a complex displacement (dAm, dBm) of said constellation symbol on the axes. abscissa and ordinate of the complex plane of the constellation of said symbol, said complex displacement being selected by means of said complex correction data (Fm), among the complex displacements belonging to at least one of the following categories: real displacement, respectively imaginary, said complex displacement, negative when the sign of the real part, respectively of the imaginary part of said complex correction data (Fm) is positive; real or imaginary displacement of the positive complex displacement when the sign of the real part, respectively of the imaginary part, of said complex correction data item (Fm) is negative. By "real displacement of the complex displacement" is meant the displacement along the axis of the reals of the real part of the complex displacement. By "imaginary displacement of the complex displacement" is meant the displacement along the imaginary axis of the imaginary part of the complex displacement. Indeed, the movements of the real part and the imaginary part of the constellation symbol are independent of each other. For example, the real part of the constellation symbol can be moved positively with respect to the real axis. While the imaginary part can be moved negatively with regard to the imaginary axis.

Thus, the invention proposes a control of the complex displacement of each constellation symbol modulating a data flow on the abscissa and ordinate axes of the complex plan of the constellation of this symbol. We obtain a controlled complex displacement of constellation symbols that can be distinct, whether by its displacement value, or by its polarity of a constellation symbol modulating a data stream to another constellation symbol modulating another data flow. In other words, it is possible, for example, for the real and / or imaginary components of the constellation symbol modulating a data stream of index m + g, with g integer such that 1 <m + g .1 \ / 1 are corrected according to a real and / or imaginary positive displacement of value dg while the real and / or imaginary components of the constellation symbol modulating the flow of data of index m are corrected according to a real displacement and / or negative imaginary value of dm = dAm + j.dBm. The nature of the complex displacement being controlled, the complex displacements implemented can lead to keep the constellation points in their decision area or in the constellation of origin or to move them outside. Advantageously, the absolute value of the real part VAm, respectively imaginary VBm, of said complex displacement is proportional to the real part, respectively imaginary, of said complex correction data (Fm). A high accuracy of correction is thus obtained, the values VAm and VBm being proportional to the complex correction datum (also called complex correlation data) which results from the complex correlation previously mentioned. In other words, the actual components of two distinct constellation Sm and Sm + g symbols modulating two distinct data streams can be corrected by two distinct values of displacement dAm and dAm + g, as well as the imaginary components dBm and dBm + g are distinct from one symbol to another. Similarly for the same stream of data of index m given, the values VAm and VBm, and thus dAm and dBm which result, are distinct because the displacements of the real part and the imaginary part of the constellation symbol are independent l one of the other. The displacement value therefore varies from one data flow to another depending on the complex correlation taking into account the P power peaks detected among the SF preconstructed complex temporal samples from the pre-construction phase when m = 1, or among SF complex time samples from a refresh step implemented for a previous data stream when 1 <n-KM. According to a particular aspect of the invention, the correction step also implements a weighting (308) of the real or imaginary part, respectively, of said complex displacement as a function of said original coordinates of said constellation symbol. In fact, in the case of constellation changes defined by the ACE technique, for example, correcting displacements of the coordinates of the constellation symbols to the outside of the original constellation are only allowed for symbols located on the periphery of the constellation. constellation. The weighting implemented according to the invention therefore makes it possible to respect the rules specific to the techniques of the prior art, such as the ACE technique or a combination of these techniques, for example the CD and ACE techniques. According to a particular aspect of the invention, the transmission method further comprises a step of weighting said complex correction data item (Fm) by a weighting factor K. Such a weighting makes it possible in particular to homogenize the symbol correction amplitudes. of constellation. According to a particular embodiment of the invention, the transmission method further comprises a switching step for transferring said SF preconstructed complex time samples from said pre-construction phase input of said step of detecting said phase of correction. The invention also relates to a device for transmitting a code division multiple access signal, said CDMA signal, said CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M separate data streams forming said CDMA signal, each constellation symbol being identified by an index m, m being an integer such that 1.rrKM, according to the invention, the transmission device comprises, for at least one block, M constellation symbols of said CDMA signal: - a unit of preprocessing, comprising: a preconstructed complex temporal SF sample pre-construction module representative of said M data streams respectively multiplied by a spreading factor distinct factor SF, SF being an integer, and o a correction module constellation symbols modulating the M data streams of said CDMA signal, said correction module comprising the following entities, implemented for each data flow: a detection entity of P samples, P being an integer, of said SF preconstructed complex temporal samples from said pre-construction module when m = 1, or among SF complex temporal samples from an entity update method implemented for a previous data stream when 1 <n-KM, having a power greater than a predetermined threshold (a), said detection entity delivering said P samples and SF-P null samples, constituting SF temporal samples complex to be corrected, a correction entity of the constellation symbol modulating said data flow as a function of said SF complex time samples to be corrected, delivering a complex correction data Fm of said constellation symbol, an entity for constructing SF complex temporal samples associated with said complex correction data, an entity for updating said SFs preconstructed complex time frames, said update using an accumulation associating two by two said SF complex time samples associated with said complex correction data with said SF preconstructed complex temporal samples from said pre-construction module when m = 1, or with said SF complex time samples from said update entity implemented for a previous data stream when 1 <n-KM, delivering SF complex time samples used for the next data stream; a spreading unit implementing a multiplication of each data stream of said modulated CDMA signal by a constellation symbol derived from said preprocessing unit by a spreading factor spreading code SF; a scrambling unit of said CDMA signal originating from said spreading unit; a unit for transmitting and / or storing a CDMA signal originating from said scrambling unit. Such a transmission device is particularly suitable for implementing the transmission method according to the invention as described above.

This transmission device may of course include the various characteristics relating to the transmission method described above, which can be combined or taken separately. Thus, the features and advantages of this device are the same as those of the transmission method. Therefore, they are not detailed further. According to a particular embodiment, said pre-construction module comprises at least one level-weighting entity according to which each Xm constellation symbol modulating a data flow associated with one of said M data streams is multiplied by a factor of Gm gain, and in that it comprises a power adjustment unit whose input is fed by said scrambling unit and whose output feeds said transmission unit and / or storage and said leveling entity level.

According to a particular embodiment, said correction module further comprises: a summation entity of the coordinates of said constellation symbol, called original coordinates of said constellation symbol, with coordinates representative of a complex displacement (dAm, dBm ) of said constellation symbol on the abscissa and ordinate axes of the complex plane of the constellation of said symbol, and - a weighting entity of the real or imaginary part, respectively, of said complex displacement as a function of said original coordinates of said constellation symbol . In yet another embodiment, the invention relates to a computer program comprising instructions for carrying out a method as described above, when this program is executed by a processor. The method according to the invention can therefore be implemented in various ways, in particular in hard-wired form or in software form. 4. List of Figures Other features and advantages of the invention will appear more clearly on reading the following description of a particular embodiment, given as a simple illustrative and non-limiting example, and the accompanying drawings, among which: FIGS. 1A and 1B respectively represent the processing diagram of a CDMA signal and a simplified block diagram of the PAPR reduction system according to the invention; FIG. 2 illustrates the main steps of a transmission method according to the invention; FIG. 3 illustrates a detailed block diagram of the PAPR reduction system according to the invention; FIG. 4 illustrates the operating principle of a weighting entity of the real part, respectively imaginary part, of a complex displacement applied to the constellation symbol modulating a data stream. FIG. 5 illustrates the structure of a transmission device according to the invention; FIG. 6 illustrates two types of corrected constellations obtained according to the invention; FIGS. 7A and 7B respectively illustrate a comparison between an original constellation and a corrected constellation according to the invention for a modulation of the MAQ64 type. 5. DESCRIPTION OF AN EMBODIMENT OF THE INVENTION 5.1 GENERAL PRINCIPLE The invention is therefore based on the use of a constellation correction control modulating a CDMA signal in such a way as to optimally reduce the crest factor of the time signal S (t) CDMA which would be transmitted without correction at the output of the CDMA transmission device. Such a CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M separate data streams forming said CDMA signal, each constellation symbol and each data stream being identified by an index m, m being an integer such that 1.rrKM . Since the time signal CDMA S (t) is not known from the method according to the invention, the latter preconstructs a digital signal representative of the dynamic point of view and peak values of the analog signal at the output of the CDMA transmission device. In other words, this pre-construction consists in obtaining an "image" of the analog signal at the output of the CDMA transmission device and in progressively correcting, data flow by data flow, each constellation symbol modulating a data stream to obtain a preconstructed and corrected global CDMA signal whose peak factor is reduced. Once the pre-construction has been performed, the method according to the invention implements a correction phase for each data stream, which consists of detecting the amplitude peaks within the set of preconstructed complex temporal samples representative of the M is a data stream, each data stream of index m being respectively multiplied by a spreading code Sm distinct from spreading factor SF, SF being an integer. The detected peaks thus correspond to peaks of amplitude greater than a predetermined threshold corresponding to the desired final peak factor level. Detecting these peaks for each data stream makes it possible to advance data streams by data flow the overall decrease in CDMA signal peaks greater than the predetermined threshold corresponding to the desired final peak factor level. In the presence of these peaks, the method according to the invention delivers a complex correction data making it possible to optimize the modification of the constellation associated with the CDMA signal in order to reduce its peak factor. In particular, for each data stream considered, the complex correction data results from a complex correlation between: on the one hand, the temporal samples corresponding to the peaks of amplitude greater than a predetermined threshold detected among the SF temporal complex samples ( thus 2.SF samples in total) preconstructed from said pre-construction phase when m = 1, or among SF complex time samples from an update step implemented for a previous data stream when 1 <n- KM, and - the SF real time samples associated with the spreading code of the data flow considered on the other hand. The invention therefore makes it possible to adapt data streams by data stream to the constellation of the CDMA signal to be transmitted.

New CDMA signal modulation constellations for reducing the peak factor are thus obtained according to the invention. In the following, with reference to FIG. 1A, is described the general scheme for processing a CDMA signal aimed at reducing the crest factor. A CDMA signal is, according to the embodiment as described by the general diagram in relation to FIG. 1A, processed according to a succession of steps: to the transmission 1000: generation 101 of source, encoding and interleaving data 102 of said data supplying interleaved data, modulation of said interleaved data 103, for example in a modulation using a quadrature amplitude modulation (QAM) or phase shift (CDM) type constellation, correction 104 of the constellation symbols each modulating a data stream associated with one of said M data streams for reducing the crest factor according to the method of the invention, SP time spread CDMA data streams by multiplying each of the M data streams of the CDMA signal by a spreading code distinct from one data stream to another and scrambling SC 105, pulse shaping by means of a shaping filter 107, transmission 10 700 of the CDMA signal on a transmission channel 108 in the presence of noise for example a Gaussian white noise 109; and at reception 1010: reception 110 of a received CDMA signal, pulse shaping 111, reception 112 by means of a RAKE receiver, demodulation 113, deinterleaving and decoding 115 of said demodulated CDMA signal, determination 116 of bit error rate (or "bit error rate").

The invention therefore proposes a specific correction technique 104 which makes it possible to effectively reduce the crest factor while being simple to implement. In addition, the correction according to the invention is implemented only for transmission and does not require modifications of existing receivers. The method of reducing peak factor according to the invention is presented according to the simplified block diagram of FIG. 1B. More specifically, the essential steps of the transmission method according to the invention are implemented between the conventional steps of modulation 103 and SP spreading and scrambling SC 105. More precisely, according to FIG. 1B, the method according to the invention corresponds to a servo system retroactive type (or "feed-back" in English) nested with a system of direct type correction (or "Feed-Forward" in English).

This method is non-iterative, in other words a correction relating to a block of M constellation symbols respectively modulating M data stream of the CDMA signal is entirely calculated in a duration of SF "chips", a chip corresponding to an element binary duration Tc of a spreading code length SF used to form the CDMA signal considered. As further detailed in connection with FIG. 2, the operation of the method consists, in a preprocessing phase, of pre-constructing, before spreading, the complex temporal CDMA signal which "could" be obtained at the output of the transmission device at from the different data streams of the CDMA signal which are modulated by constellation symbols Xm = Am + jBm. It should be noted that according to the embodiment shown in relation with FIGS. 1A and 1B, the invention implements a power adjustment A_P 106 after spreading and scrambling making it possible to weight the power allocated to each data stream of the CDMA signal. .

Such a power adjustment is taken into account in the pretreatment phase, and more particularly in the pre-construction stage of the CDMA signal which "would" be output from the CDMA transmission device by means of an A_P weighting entity. 106. Accordingly, the pretreatment unit that constructs the CDMA signal that "would" be output from the uncorrected CDMA transmission device communicates with the spreading and scrambling units 105 and the power adjusting unit 106. in order to implement a weighting 1069 of the constellation symbols modulating the data streams of the CDMA signal so that the power adjustment performed prior to the transmission is taken into account in the preprocessing step. To carry out the pre-construction of the CDMA signal, a first X11 parallel / serial conversion of the symbol block Xm, index k, as well as the block of XM G1 corresponding gain adjustment parameters Gr obtained by the weighting entity Gm allowing a power adjustment A_P 106. This parallel / series conversion is obtained by means of the decoder modules DEC 1058 and 1064 respectively which shift each symbol / gain factor relative to each other and by MUX multiplexing modules 1062 and 1065 respectively.

Then from the symbol sequences and gain factors the products 1069 of the coincident terms are made, and the series of results produced [XMGM, XniGni, X1G1] is obtained. This operation is intended to introduce readjusted level signals, representative of what they would be before summation of the different data streams and transmission in the channel 108. Then for each spreading code Sm which multiplies to origin a data flow index m modulated by a symbol Xm, are then calculated simultaneously all the samples of the temporal response of this code "modulated" by the product Xm.Gm samples that would be obtained after spreading SP and scrambling SC 105 and readjustment of power A_P 106 without further intervention. It should be noted that the scrambling operation is applied identically for all data streams. Since this operation is linear, it will be assumed later that for an interpolation factor of the shaping filter L equal to 1, it can be reduced downstream of the summation and does not intervene in the formation of peaks. amplitude, it is then not useful for the pre-construction to take into account the subsequent operation of scrambling Sc. Then, starting from the first data stream and then gradually, the different temporal responses are accumulated in the ACC1 accumulation entity. At the end of the accumulation when the data stream of index M has been processed, the signal vector EM [] at the output of the module ACC1 thus consists of the set of SF complex samples calculated in parallel, which at the end of the accumulation construction will correspond to the serial samples of the temporal CDMA signal that "would" be obtained at the output of the CDMA transmission device. At the end of the block of series symbols of index k, that is to say for the symbol of index M, the set of time samples consisting entirely of the vector EM [] contained in module ACC1 1055 are then loaded. in a second module ACC2 1051 which, in this way, is initialized by switching a switch to connect the output of the module ACC1 1055 to the input of the module ACC2. The first ACC1 module 1055 is then completely reset and the pre-construction calculations restart on a new block of symbols [XM ... X1] of index k + 1 following.

At the same time, the sequence of the constellation symbols of the block of index k current has transited in a delay line Z-SF 1068 of duration SF, so that the first symbol X1 comes out when the module ACC2 has just been initialized by the module ACC1 and the switch is switched back to the input values from the weighting module 1053. The switch then remains in this state for the entire processing time, namely of correction of the index block k.

In a second step, the module ACC2 has an operation similar to that of the first module ACC1, which consists in pre-constructing the temporal CDMA signal that could be obtained after spreading and summing, but this time taking account of the constellation corrections made to the data streams. by data stream by the method according to the invention for the purpose of reducing the crest factor. Thus, just like the ACC1 module, the ACC2 module thus preconstructs a temporal signal E'm [], with the difference that to generate this signal, the time responses of the codes modulated only by the displacements are no longer accumulated. constellation symbol correction dm since at the initialization of the ACC2 module the accumulation of all the temporal responses of the codes modulated by the original constellation symbols Xm has already been loaded. At the ACC2 update accumulator, the start situation therefore corresponds to a situation where all the samples of the signal are known and in particular the peaks of the original signal that are to be eliminated: E'0 [] = Em [], and these peaks will evolve only according to the corrections of constellation symbol made. The operation of pre-construction of the signal E'm [] carried out by ACC2 is accordingly a reactualization of the signal EM []. From the beginning of the index symbol block k, the vector E'm [] of the corrected temporal samples is therefore progressively modified at each clock cycle and accumulation, and a simple named amplitude peak detection operation 1050 suffices. to identify power peaks greater than a determined threshold value a, this threshold value in fact setting the desired final peak factor level. The amplitude peak detection module 1050 sets all signal values between - a and + a to zero and retains only the peak values.

Next comes a correlation operation 1052 between the binary sequence of the spreading code Sm of index m (multiplying the corresponding data flow of index m) and the peak value samples at the output of the amplitude peak detection module. 1050. The result of this operation makes it possible to elaborate a correction information Fm by displacement dm = dAm + j.dBm of the constellation symbols Xm, which will progress the reactualization of the signal in the direction of the global decrease of amplitude peaks. of the CDMA signal. Thus, when the polarity of the real part, respectively imaginary, of the correlation product is positive, then a correction by a negative displacement of Am, respectively dBm, of the constellation symbol must be chosen, conversely if the polarity is opposite a correction by a displacement of Am , respectively dBm, positive must be selected to obtain A'm, respectively B'm, the new components of the constellation symbol X'm with: A'm = Am + dAm B'm = Bm + dBm In particular, the magnitude of the displacements that will be made may not be fixed but proportional to the result of the correlation product, and thus fully determine the complex displacements of the constellation symbols in order to obtain a reduced peak factor, however a weighting module 1053 is necessary to be able to weight the correction values and in particular: according to the type of constellation and the location of the symbol in the a constellation. For example, for quadrature phase shift keying (QPSK) type modulation, much larger displacements will be allowed than for sixty-four quadrature amplitude modulation type modulation. 64-QAM. In the case of displacements of constellation symbols corresponding to the ACE (Active Constellation Extension) technique based on a displacement carried out in the direction of a distance from the decision axes, for example, external displacements will not be possible. authorized only for symbols located on the periphery of the constellation. At the output of the weighting module 1053, a complex displacement value dm is thus obtained which according to FIG. 1.B is applied both to the symbol Xm and taken into account in the ACC2 module after weighting at level 1070.

Thus, optionally, the update (2024) according to the invention takes into account the optional power adjustment (FIG. 1, 23, FIG. 2) of the data flows forming the CDMA signal carried out before transmission and / or CDMA signal storage. Throughout the correction of the constellation symbols of the block of index k, when m varies from 1 to M, the method according to the invention therefore determines a sequence of correction displacements dm which will be applied to the symbols of both. Xm origin to form a sequence of symbols X'm leading to a lower peak factor after reintroduction (DEMUX 1061 demultiplexing and C-DEC 1059 registration) upstream of spreading SP 105 different data streams, and therefore also counted in the acceleration accumulator ACC2 after weighting in level 1070 by the corresponding gain factor in order to obtain a correctly updated signal before each new detection / correction of the peaks. The sequence of displacements dm will have the effect, at each update operation performed for each data stream, to reduce the amplitude of the detected peaks compared to what it could have been without correction, this update will therefore produce new values. E'm [], of the sequence of time samples preconstructed in parallel, with a regression constraint on the samples of higher level and with the objective that this amplitude becomes lower than a at the end of the correction and update processing of the symbol block [XM ... X1] of current index k. The system is then reset at the ACC1 and ACC2 accumulator blocks in order to restart the pretreatment factor reduction method according to the invention for the following symbol block [XM ... X1] of index k + 1 . 5.2 Detailed Description and Implementation of the Different Stages of the Transmission Method According to the Invention FIG. 2 represents in detail all the steps implemented according to the invention in order to develop a correction of the constellation which modulates each data stream of a CDMA signal and this to reduce the peak factor of the transmitted CDMA signal while Figure 3 represents the physical implementation of these steps. These two aspects are detailed below. 5.2.1 Description of the different steps of the method according to the invention Thus, as illustrated in FIG. 2, the transmission method according to the invention makes it possible to apply a correction of the modulation constellation with a view to reducing the peak factor of the CDMA signal to be transmitted comprises a preprocessing step (20). More specifically, this pretreatment step (20) comprises a preconstructing phase (201) of SF preconstructed complex temporal samples representative of the M data streams respectively multiplied by a spreading factor distinct code SF, SF being an integer. Indeed, the reduction of the peak factor must be applied to the CDMA time signal which would be transmitted without correction at the output of the CDMA transmission device. To do this, and according to the embodiment represented by FIG. 2, the pre-construction phase (201) optionally sets (represented in dotted line) when a power adjustment (23) of the data streams is also optionally (shown in dashed line) implemented before transmission and / or storage (24), at least one level weighting step (20110) according to which each Xm constellation symbol modulating a data flow associated with one of said M streams of data is multiplied by a gain factor Gm. Thus, optionally, the pre-construction (201) according to the invention takes into account the optional power adjustment (FIG. 1, 23, FIG. 2) of the data flows forming the CDMA signal carried out before transmission and / or storage of the CDMA signal. Moreover, as illustrated by FIG. 2, the pre-construction phase (201) further comprises, for a current data flow of index m modulated by a constellation symbol Xm, m being an integer such that 1.rn .M a pre-construction step (2011) of a set of SF complex time samples associated with the current data flow of index m. Then, the pre-construction phase (201) of the transmission method 20 according to the invention comprises a storage step 2012 and accumulation combining two by two SF complex time samples associated with the current data stream of index m with the previously memorized set of SF complex temporal samples associated with the previous m-1 data streams. Once the M data streams have been processed one after the other according to these two stages of pre-construction (2011) and storage (2012), we obtain at the end of the pre-construction phase. (201) a signal vector En, [] consisting of the set of complex temporal samples calculated for the M data streams of the block of index k corresponding to the time signal CDMA Sk (t) which would be transmitted by a device of CDMA transmission without any correction being implemented.

In particular, considering a CDMA system with M data stream for which each spreading code comprises SF bits called "chips" with M SF and including the scrambling operation, the CDMA signal at the output of the CDMA transmission device has for expression: M SF -1 Sk (t) Gk, m Xk, m Sme. Scr, k.SF + n - P (t - n Tc) m = 1 n = o Where: Sr ', is a pseudo random sequence which takes normalized values in the set SF SF Scr' k SF ri is the identical complex scrambling sequence (which NI NI does not interfere with peak amplitude formation when considering a shaping filter interpolation factor equal to 1) for each data stream whose actual components and imaginary take standardized values in the set - + -, -}, P (t) is the function over the time interval Tc, G a is a gain factor V 2 V 2 dedicated to the data flow d index m supposed constant over the symbol duration Ts, which corresponds to the weighting of the block of symbols of index k, Xk, m = Ak, m. J. Bk, rn is the constellation symbol which modulates the data flow of index m on the symbol duration Ts, and which belongs to the block of symbols of index k. If the envelope of the CDMA signal must not exceed a certain power threshold a in order to be able to reduce the peak factor, Sk (t) must check: 15k (01 <a Vn, Vk What is obtained when: M IScrl. Z Gk, m.Xk, m Sm, n <a Vn, Vk m = 1 In other words: [M 2 [yi 2 / G kart, Akan, Sn "+ 1 G k, m Bk mm, SI" < This result therefore shows that the scrambling operation can be neglected in the pre-construction of a CDMA signal in order to serve as a reference for the calculation of a symbol correction. of constellation to reduce amplitude peaks and thus the crest factor.

Then the transmission method 20 according to the invention comprises a phase of correction (202) of this signal vector En, [] transferred by activation of a switch SW (the switch being composed for example of two elementary switches SW1 and SW2 represented on FIG. 3 and associated, respectively, with the purely temporal SF temporal samples and the SF purely temporal temporal samples of the SF complex temporal samples) making it possible to connect the pre-construction module (not represented but composed of the pre-construction entities 302, of accumulation 303 and optional weighting at level 1069 of FIG. 3) and the correction module (not shown but composed of the detection 306, correction 307, construction 304 and update 305 entities of FIG. 3) respectively implementing the pre-construction phase and the correction phase. Then once this transfer (shown in Figure 2 by a dotted line) performed, the correction phase becomes independent of the pre-construction phase by opening the same switch SW. The correction phase (202) makes it possible to correct progressively the signal vector En, [] resulting from the pre-construction phase (201). More precisely, a correction is made as data flows by data flow, by small complex and variable movements from one data flow to another (dAm, dBm), of each original value of the data flows. real and imaginary components (Am, Bm) of a constellation symbol modulating a data stream to obtain a preconstructed and corrected CDMA signal whose peak factor is reduced. Thus, the correction phase (202) implements, for the data flow of index m = 1 modulated by a constellation symbol X1, a step of detection (2021) of P complex time samples, P being an integer, corresponding at power peaks whose amplitude is greater than a predetermined threshold a among the complex temporal samples of the signal vector En, [].

Among the SF complex temporal samples of the signal vector En, [], one detects for example, for a number SF = 256, ten peaks corresponding to: fourth, twentieth, fifty-fifth, seventy-eighth, hundredth, one hundred and twenty- second, one hundred and fifty-first, one hundred and seventy-third, one hundred and seventy-two-and-twenty-seventh, and two hundred and twentieth complex temporal samples of the signal vector En, []. The detection step delivers these P = 10 samples and the other SF-P = 246 temporal samples are set to zero to form a set of M complex temporal samples to be corrected. Then, the correction phase (202) of the transmission method 20 according to the invention comprises a second correction step 2022 of the constellation symbol X1 modulating the current data flow of index m = 1 as a function of the set of SF complex temporal samples to be corrected resulting from the detection step (2021) described above, delivering a complex correction data item F1 of the constellation symbol modulating the current data flow of index m = 1.

More precisely, the complex correction data item F1 results from a complex correlation operation (20220) of the SF complex time samples to be corrected, comprising the P = 10 samples whose amplitude is greater than the predetermined threshold a resulting from the detection step. 2021 and SF real time samples (respectively imaginary, the imaginary part of the spreading code can be zero if necessary) associated with a spreading code SF spreading factor SF multiplying the flow of index data m = 1 constituting the CDMA signal. The correction step 2022 implements a summation of the coordinates of the constellation symbol Xi, called original coordinates of the constellation symbol, with coordinates representative of a complex displacement (dA1, d131) of the constellation symbol on the axes. abscissas and ordinates of the complex plane of the constellation of the symbol, the complex displacement being selected by means of the complex correction data F1, among the complex displacements belonging to at least one of the following categories: real displacement, respectively imaginary displacement complex, negative when the sign of the real part, respectively of the imaginary part, of the complex correction data F1 is positive; real or imaginary displacement of the positive complex displacement when the sign of the real part or of the imaginary part of the complex correction data item F1 is negative.

In addition, according to a feature of the invention, the absolute value of the real or imaginary part of the complex displacement is proportional to the real or imaginary part of the complex correction data F1. Once this complex displacement has been selected from the complex correction data Fi, the summation of the original coordinates (A1, B1) of the constellation symbol X1 with the coordinates (dA1, dB] of the selected complex displacement is performed and delivers the new coordinates (A'1, B'1) of the corresponding corrected constellation symbol X'1. The complex correction displacement is therefore performed in the opposite direction to the formation of the amplitude peaks and in a variable manner from one data stream to another. Indeed, because the method causes from one data stream to the other an overall decrease of the CDMA signal peaks, the set of SF complex time samples to be corrected also varies from one data stream to another. , and consequently the correction data relating to a data flow as well. This results in a series of complex displacements that evolve from one data stream to another in an adaptive manner with respect to the peak factor level of the desired CDMA signal. However, two data streams can also each have a complex displacement close to their respective constellation symbol, or even the same. In the case where m> 1, for example m = 5, the detection step (2021) is carried out among the complex temporal samples of the updated signal vector E'm [] taking into account by accumulation the corrections associated with the five fluxes of previous data, namely the data streams of index 1, 2, 3, 4 and 5. Then, the correction phase (202) of the transmission method 20 according to the invention comprises a step of construction (2023) of a set of SF preconstructed complex temporal samples associated with the correction data of the constellation symbol Xm modulating the data flow of index m, in other words the previously defined complex displacement. Such a construction step notably implements a multiplication by the spreading code Sm associated with the data flow of index m. Then, the correction phase (202) of the transmission method 20 according to the invention comprises a step 2024 of updating the signal vector En, [] by pairing pairs of SF complex time samples of the signal vector En, [] with the set of SF preconstructed complex temporal samples (from the construction step 2023 above) representative of the complex displacement dm correction carried out delivering a signal vector E'm [] used for the next data stream. It should be noted that when a power adjustment (106) is optionally implemented the SF preconstructed complex temporal samples representative of the complex displacement dm of correction are also weighted in level (20230) before the updating step 2024. L the set of detection (2021), correction (2022), construction (2023) and update (2024) steps is applied to each of the M data streams associated with the M data stream forming the CDMA signal. Once these M data streams have been processed, a block of M constellation symbols corrected at the end of the correction phase 202 of the preprocessing step (20) is thus obtained. Then, the method according to the invention implements a spreading step (21) by multiplying each data flow associated with one of the M data streams of said modulated CDMA signal by a corrected constellation symbol resulting from the step pretreatment by a spreading factor spreading code SF, a scrambling step (22) of the CDMA signal from the spreading step, and a step of transmitting and / or storing (24) a signal CDMA from the scrambling step (22). Optionally (represented in dashed line), prior to said transmission and / or storage step (24), said method further comprises a power adjustment step (23) whose input is fed by said scrambling step (22). ) and whose output feeds said transmission and / or storage step (24) and the step of leveling (20110) reciprocally implemented in the pre-construction phase (201). 5.2.2 Physical implementation of the different steps of the method according to the invention In accordance with the diagram of FIG. 3, the implementation of the method according to the invention comprises a set of modules / processing entities, as subsequently described for a factor interpolation of the shaping filter equal to one for which the contribution of the scrambling operation relative to the process that generates the crest factor can be assumed to be completely zero.

With respect to a conventional chain which generates a signal Sk (t) from the sequence of values (Am, Bm) of coordinates in the complex plane which define a constellation symbol Xm (Xm = Am + j.Bm), the method according to the invention generates at the end of the second correction phase (202), corrected values (A'm, B'm) data flow data streams which will give after spreading and scrambling, optionally power adjustment a corrected signal Sk (t) in which the amplitude peaks which affect the signal Sk (t) have been attenuated so as to be less than a predetermined threshold a. To do this, from a block of index k, symbols [XM ... X1] which modulate the data flows, the operating principle is based on pre-construction before spreading a complex signal Sk (t) representative of the temporal CDMA signal that could be obtained after summation of the different data streams if no particular action was taken to reduce the crest factor. It should be noted that in FIG. 3 a black vertical bar 3000 in a module / entity indicates that the (the) module / entity is registered and that at each edge of a cycle clock of duration Tc , a new input is loaded and / or a new output is updated, the corresponding values are then maintained during a clock cycle. The module 311 receives the block [XM ... X1] parallel constellation symbols and assigns an increasing delay of a "chip" period (a "chip" corresponding to a spreading elementary elementary bit) additional to each symbol , starting from a zero delay for X1, a delay of Tc for X2, 2.Tc for X3, etc .... and finally (M-1) .Tc for the last symbol XM. The purpose of this operation is then to be able to serialize the various constellation symbols via a multiplexer or selector switching at the time "chips". The serial train of the symbols is then applied at the input of the method according to the invention according to the real and imaginary components Am and Bm resulting from an imaginary real part decomposition 309. The interest of this embodiment, as represented by FIG. FIG. 3 resides in the possibility of restoring the synchronism of the parallel train of the symbols once corrected with a minimum of delay, thanks to the complementary operations performed by the module 312 at the end of the processing according to the invention. The latency of the entire process inserted between modulation (103) and spreading / scrambling (105) is then only two symbol time periods Ts, that is to say 2.Ts = 2.SF.Tc . The role of the module 312 is therefore to realign all the symbols of the block of subscript k, after the constellation symbol corrections have been applied. These corrections are obtained in the form of a serial train at the output of an imaginary real part decomposition 310 of the method according to the invention and are added 3100 selectively to the symbols by means of a demultiplexer (1061) of this kind. that each correction is in coincidence with the symbol for which this correction has been calculated, then the module 312 assigns a decreasing delay of one "chip" period to each symbol, starting from a delay SF.Tc for X'1 , a delay of (SF-1) .Tc for X'2, (SF-2) .Tc for X'3, etc. and finally (SF-M + 1) .Tc for the last corrected symbol X'M. At the output of the module 312 we finally obtain a block of constellation symbols [X'M ... X'1] corrected, thus delayed by 2.Ts with respect to the original block [XM ... X1] at the input of system, and to which the spreading operation SP (105) can be applied. The module 313 performs the serialization of the gain factors [GM ... G1] subsequently applied by the power adjustment module 106, in a manner analogous to that performed by the module 311, so that the sequences of elements Xm and Gm are in coincidence to form the weighted elements Xm.Gm supplying the pre-construction entity 302. In parallel, the generator module 301 has the role of delivering at each clock stroke, simultaneously, the sequence of the SF binary samples which constitute the index code m associated with the data flow of index m in order to spread each of the symbols Xm by a code Sm respectively. These samples are stored in ROM memory or computed algorithmically by a pseudo-random generator. If the samples Sine of the order n, constitute the elements of the code Sinn it is expressed as follows: Sm [] = [Smo, Sma,, Smn,, SmsF _ The pre-construction entity 302 receives the succession of the constellation symbols Xm from the block [XM ... X1] according to the two real and imaginary components Am and Bm, each weighted by the respective gain factors Gm, and simultaneously calculates the sequence of the SF complex samples (ie 2 .SF temporal samples in total corresponding to the calculations of SF real samples and SF imaginary samples forming said SF complex samples) corresponding to the spreading of these components by the same code Sm. At the output of the pre-construction entity 302, the following two real and imaginary components are obtained: Sm °. Gm-Am Sm, o. Gm. Bm S Gm-Am Sma. Gm. Bm - - - - - - AAm11 = ABm11 = Sm, sF-2. Gm-Am Sm, sF -2. Gm-Sm Bm, SF-1- Gm-Am-Sm, sF-1. Gm. Then, at each clock stroke and constellation symbol Xm, all the samples of the real and imaginary parts of the signal Sk (t) are preconstructed progressively by accumulating and storing in ACC1 accumulation entity 303 (forming when is combined with the pre-construction entity 302 the pre-construction module according to the invention), the current calculation results of the terms AAm and ABm [], with respectively the different results of the operations previously performed to calculate these same terms . Since ACC1 accumulation entity 303 is registered, two output signal vectors of this module are obtained which have the following expression: at initialization, accumulation entity ACC1 303 being set to zero: [] = [ 0] BBil] = [0], and - to order m, for 1 <m M: m-1 m-1 BAm11 = AAill BBm11 = ABill i = ti = t Or in other words: m-1 51, SF-2. Gi. Ai i = t m-1 Si, SF-1. Gi. Ai i = 1 m-1 51, SF-2. Gi. Bi 1 = 1 m-1 Si, SF-1. Gi. Bi -1 = 1 BAm11 = BBm [] = - m-1 Gi. Bi i = t m-1 If G. Bi i = t_ m-1 Gi. Ai i = t m-1 Gi. Ai i = t At the end of the processing block of constellation symbols [XM ... X1] of index k, when m = M, the set of temporal samples of the real and imaginary parts of Sk (t) over the duration a symbol of duration Ts, with the exception of a last pair of values which is neglected (the terms AAM [] and ABm [] can not be taken into account before re-initialization of ACC1 accumulation entity registered), is contained in the vectors respectively BAM [] and BBNI at the output of the accumulation entity ACC1 303. A "V" validation signal is then activated during a clock cycle, and an initialization signal " ends »resets all the accumulators of the accumulating entity ACC1 303. The calculation of the real and imaginary parts of Sk + 1 (t) then restarts on a new block of constellation symbols of index k + 1. Activation of the "V" signal then has the effect of loading the 2.SF temporal samples (SF temporal samples for each real and imaginary component) of the accumulation module ACC1 303 into a second accumulation entity ACC2 305, also called entity discount. The second accumulation entity ACC2 305 is thus reinitialized at this time, by the closing of switches SW1 and SW2 controlled by the signal V in order to connect the respectively real and imaginary outputs of the accumulation module ACC1 303 to the corresponding inputs of the discounting entity 305. Then, the switches SW1 and SW2 are switched back to the remainder of the block of index constellation symbols k, and the updating entity 305 then works in "accumulation only" mode from the time samples provided by the construction 304 allowing the construction of SF complex samples associated with the complex correction data obtained according to the invention. At the same time, the sequence of the constellation symbols Xm of the block of index k as well as the components Am, Bm and Gm have respectively transited in delay lines 314 and 315 of duration SF; in this way the first symbol X1 of the index block k comes out when the updating entity 305 has just been reset and contains the two real and imaginary components of the signal Sk (t) delivered by the first accumulation entity ACC1 303. As shown in FIG. 3, the method according to the invention makes it possible to determine complex displacements dm = dAm + j.dBm at the symbols Xm in order to obtain at the output a modified signal S'k (t) of weaker crest factor. To determine the displacement solutions (dAm, dBm) for each symbol Xm, and in the same way as for the pre-construction implemented by the combination of the pre-construction entity 302 and the accumulation entity ACC1 303, the method according to the invention progressively preconstructs the sequence of the SF samples of each real and imaginary component of the temporal response of the corrected signal S'k (t) that could be obtained after summation of the different data streams. The real and imaginary temporal responses of the summation of the set of original symbols Xm, spread by the Sm codes and weighted in level by the factors Gm, having already been loaded into the updating entity 305 at its initialization, and the operation being linear, it is only necessary to compute in the building entity 304 and to accumulate in the updating entity 305 the real and imaginary temporal components of the level-weighted correction movements (dAm, dBm) 1070 by the Gm gain factors at the output of the line 315, then spread by means of the construction entity 304 by the Sm codes at the output of the generator module 301.

From the first symbol X1 of the block of index k of constellation symbols, the construction entity 304 then receives, successively at the rhythm of the clock, the SF temporal samples corresponding to the displacement values (dAm, dBm) for each symbol Xm following, as well as the sequence of SF time samples of the corresponding Sm code generated by the generator module 301.

These SF samples are delivered by the generator module 301 identically to the pre-construction entity 302 belonging to the pre-construction module and to the building entity 304 belonging to the correction module, because the delay between the operations making to intervene the components of the same index corresponds to the period of a symbol of duration Ts exactly because of the application of the delay line 315.

At the output of the construction entity 304 two signal vectors are thus obtained: CdAm [] and CdBm [] which have the expression: Sm, o. Gm. dAm Sm, o. Gm. dBm - Sma. Gm. dAm Sma. Gm. dBm - - - - - - Sm, SF-2- Gm. dAm Sm, SF-1. Gm. dAm Sm, SF-2 Gm. dBm, Sm, SF-1. Gm. dBm_ CdAmll = CdBm [] = Then at each clock stroke and starting from the response of Sk (t) loaded at the initialization, all the real and imaginary samples of the signal S'k (t) are updated progressively by accumulating in the discounting entity 305, the current results of the transactions carried out in the building entity 304. The discounting entity 305, also called the second accumulation entity ACC2, is registered identically to the entity of the entity. accumulation ACC1 303, we thus obtain two signal vectors at the output of the updating entity 305 which have the following expression: for m = 1, the accumulators of the updating entity 305 are initialized by the accumulation entity ACC1 303 and: Dili = BAm11 = EM AAill DB1 [] = BB m [] = r = 11 AB [- for 1 <m M: M-1 m-1 DAm [] = AAill + CdAill i = tj = 1 M -1 m-1 dBm [] = ABill + C dB / i = tj = 1 From the output of the updating entity 305, a detection entity 306 then allows discriminating power peaks greater than a determined threshold value, by setting all signal values between - a and + a to zero and keeping only the peak values. This threshold value a in fact sets the desired final crest factor level.

At the output of the detection entity 306, two signal vectors defined as follows are obtained: EAm11 = Discr (DAm []) EBm [] = Discr (DBm []) With: Discr (DAm, n) = IDA ' m, n if [(DAm, n) 2 U0Bm, n) 21 0 otherwise Discr (DBm, n) Bn if [(DAm, n) 2 + (DBm, n) 2]> a2 0 otherwise An example of vectors for which samples 0, (SF-4), and (SF-2) constitute peaks of CDMA signal of amplitude greater than a obtained at the output of the detection entity 306 is given below: DAm, o-DBmo - 0 o DAm, 3 DBm, 3 E Amll = E Bm [] = DAm, SF-4 D Bm, sr-4 DAm, SF-2 D Bm, sr-2 - Detection entity 306 then provides both signal vectors defined above at a correction entity 307 which will perform a correlation product between the Sm code samples of index m delivered by the generator module 301, and the real and imaginary components, respectively EAm [] and EBm [ ], at the output of the detection entity 306 for discriminating peaks. If Emll = EAm11 + j.EBm [], the correlation product carried out has the expression: Fm = CES m (0) = r SF-1 E L.12 = 0 m, n-Fm Slr 9 (e) - EnF [ EAME. Sine], 5: 5'm (Fm) = vsF1) pp (, i [p 1 L.12 = The result of this operation makes it possible to elaborate a displacement correction information dm = dAm + j.dBm of the symbols of constellation Xm, which will advance the reactualization of the signal in the direction of the overall decrease of the amplitude peaks of the CDMA signal Thus when the polarity of the real (imaginary) part of the correlation product is positive then a displacement correction dAm ( dBm) negative of the constellation symbol must be chosen, conversely if the polarity is opposite a displacement correction dAm (dBm) positive must be selected to obtain A'm (B'm), which will be the new coordinates of the constellation symbol X m with: A'm = Am + dAm B'm = Bm + dBm As for the techniques Tl-CES, CD, ACE and TR of the prior art applied to the OFDM signals and not to the CDMA signals, the method according to US Pat. invention results in a modification of the data flow modulation constellation are CDMA signal before with: each data stream forming the CDMA signal is spread and then summed, and then obtain a CDMA signal with reduced crest factor. The advantage of the method according to the invention is the flexibility of constellation correction. Indeed, any type of correction can be applied according to the invention, as long as it can result in a controlled complex movement of the real and / or imaginary components of the constellation symbol associated with a data stream. Thus, the complex displacements of positive or negative direction on the abscissa axis and the ordinate axis of the complex plane, can lead to keep the constellation points in their decision area or in the constellation of origin or in the move outside. In relation to FIG. 6, the method according to the invention makes it possible to apply both complex displacements inside (61) as well as outside (62) of the original constellation. As a result, the implementation of the method according to the invention delivers two new, more general classes of constellation correction: the class 61 ICS (of the "Inside Constellation Shift") for which, when the modification applied remains moderate, the moved constellation points remain inscribed in the constellation of origin, and the OCS (Outside Constellation Shift) class 62 for which the points are moved outside the constellation of origin. Thus, the method according to the invention makes it possible to obtain two new types of constellation called "ICS" or "OCS" depending on whether the constellation points are moved respectively inwardly or outwardly of the original constellation. The constellation correction implemented according to the invention can therefore be substituted for any of the constellation corrections of the prior art PAPR reduction techniques applied essentially to the OFDM signals and not to the CDMA signals, by adopting them, for each of them taken separately, all the advantages. Since the proposed system makes it possible to combine several techniques of the prior art, another advantage is then to be able to regain efficiency because the respective distinct defects of these different techniques of the prior art can to a certain extent be compensated for. According to an advantageous variant of the invention, the amplitude of the displacements that are imposed may not be fixed from one data stream to another, but advantageously be proportional to the result of the correlation product. Thus, when dm = dAm + j. dBm = -K. Fm with 0 <K, the expression Cdml = + j. CdBm [] of the output signal of the building entity 304 becomes: Cdm [] = dm. Gm. Sm = -K. Fm. Gm. Sm Considering, for n = p, a particular complex sample constituting an isolated maximum signal peak, we obtain: Fm, p = Emp. Sm, p, and Cdm, p = -K. Em, p.Sm, p.Gm.Sm, p, or: S, np. S - V p m, P SF We thus obtain: Cdm, p = -K. Gm. Dm, p / SF And in the order m + 1 (for m <M-1), the pre-built sample of the peak of maximal amplitude in n = p becomes: Dm + i, p = Er = -11- Ai , p + Er = 11C - K.Gm. Dm, p / SF Having posited beforehand that: Dm, n [] = DAm, n [] + j. DBmell Am ,, [] = AAme [] + j. ABm ,, [] Which amounts to: Dm_Ftp = (1 - K .GmISF) .Dm, p And finally, considering that this signal peak was present in the original signal, we deduce the expression to it. order m + 1, for 0 <m <M: m M-1 Dm + 1, p = [ni_ K. GmISF) 1. [I Ai, p1 = 1 i = t The term rjr_1 (1 - K.GIISF ) only decreasing as the processing of the k-constellation symbol block is progressed according to the method of the invention, this result demonstrates that from the initial time response of the CDMA signal Sk ( t) having an isolated peak at the sample n = p, the method according to the invention therefore reduces the amplitude of this peak to each new symbol Xm and correction displacement dm, until it has become amplitude less than a. For a set of signal peaks, the operation being linear and according to the result of the correlation product, the method according to the invention reduces, as a priority, the combination of the peaks of greatest contribution, then finally all the peak levels of the CDMA signal will be equalized, so that these become less than a. It should be noted that unlike other methods of crest factor reduction, no "new generation" of peak can occur during treatment since in this eventual case, it would be immediately taken into account in the progression of the algorithm.

The results, namely real and imaginary parts, of the correlation product can therefore fully determine the complex displacements of the constellation symbol symbols in order to obtain a reduced peak factor. A weighting entity 308 is advantageously implemented according to one embodiment of the invention represented by FIG. 3, in order to be able to weight the correction values as a function of the constellation type as well as of the location of the symbol in the constellation. . For example, for a given noise equivalent degradation, much larger displacements may thus be allowed for a QPSK-type modulation constellation than for a MAQ64-type modulation constellation. In the case of displacements of type ACE or of several types of constellation symbol correction combined: CD and ACE for example, all the symbols will not be modified in the same way; in the second case ACE for example, small displacements will be allowed for the symbols inside the constellation, and more important outside for the symbols located at the periphery of the constellation. The advantage of this weighting 308 is also to make the method fully compatible with the use of several different constellations in the same block of constellation symbols [XM ... X1] modulating each of the M data streams, in particular of rates to be transmitted or of different propagation conditions between the different data streams that are transmitted within the CDMA signal. The operating principle of the weighting entity 308 is illustrated in FIG. 4. The weighting entity 308 receives as input the real and imaginary parts of the correlation product and applies to each of them a gain respectively GHAn, and GHBm, followed by a saturation of the positive and negative values, distinctly, according to threshold values respectively VSAm +, VSAn, - and VSBm +, VS13 ,, -. From a Vern input signal, the saturation function of the real (imaginary) part restores a signal VSn, which verifies the following conditions: VSAm + if Vem> VSAm + VSm = VSAm - Vem if Vem <VSAmautrement A "Param Symb" block identifies the current Xm symbol in the constellation, and according to its position determines the gain and saturation threshold values to be applied to the real and imaginary components of the correlation product result in order to deliver displacement values dAm and dBm conform to the type of correction sought.

For example, a block of constellation symbols [XM ... X1], whose constellation symbols Xm forming a MAQ64 constellation, as shown in FIG. 7A, is considered. Xm = Am + j.Bm with: Am, Bm E (± 7 ± 7, ± 7, ± 7%). The different gains and saturation threshold values, in the processing branches of the real and imaginary parts of the weighting entity 308, can then be defined as follows: 1/2 if lAmi = 7/4 I GHB m = 1/2 if 1/41 = 7/4 GHAm = 1/32 if otherwise VSBm + = 1/32 if otherwise VSAm + = 11/4 Am = +7/4 Am = -7 / 4 otherwise 11/4 Bm = +7/4 Bm = -7/4 otherwise 1/8 1/8 1/32 1/32 -11/4 if Am = -7/4 -11/4 if Bm = -7/4 VSAm- = - 1 / 8 if Am = +7/4 VSBm- = - 1/8 if Bm = +7/4 -1/32 otherwise -1/32 otherwise With these values we obtain the corrected constellation of Figure 7B. Thus, according to the invention, despite relatively large imposed modifications to reduce the peak factor, the ACE (peripheral point) deviations completely compensate the deterioration of the BER brought by CD (internal points). 5.3 Description of the transmission device according to the invention Finally, in connection with FIG. 5, the simplified structure of a device for transmitting a code division multiple access signal, said CDMA signal, said CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M distinct data streams forming said CDMA signal, each constellation symbol being identified by an index m, m being an integer such that 1.rrKM. Such a transmission device comprises a storage module 50 comprising a buffer Mem, a processing unit 51, equipped for example with a microprocessor itP, and driven by the computer program 52, implementing the transmission method according to the invention. At initialization, the code instructions of the computer program 52 are for example loaded into a RAM before being executed by the processor of the processing unit 51. The processing unit 51 receives as input a signal CDMA x. The microprocessor of the processing unit 51 implements the steps of the transmission method described above, according to the instructions of the computer program 52, to perform a correction of the modulation constellation aimed at reducing the peak factor of the CDMA signal. x. For this, the transmission device comprises, in addition to the buffer memory Mem, a preprocessing unit, comprising: a pre-construction module (302, 303, 1069) of complex temporal samples, delivering SF complex preconstructed temporal samples representative of the M data streams respectively multiplied by a spreading factor distinct factor SF, SF being an integer, and a constellation symbol correction module which modulate the M data streams of said CDMA signal, said correction module comprising the following entities, implemented for each data stream: a detection entity 306 of P samples, P being an integer, of said SF preconstructed complex temporal samples from said pre-construction module when m = 1, or among SF temporal samples from a refresh entity implemented for a previous data stream when 1 <n-KM, having a higher than a predetermined threshold a, said detecting step delivering said P samples and SF-P null samples constituting SF complex time samples to be corrected, a correction entity 307 of the constellation symbol modulating said data flow according to said SF samples complex timepieces to be corrected, delivering a complex correction data Fm of said constellation symbol, a construction entity 304 of SF complex time samples associated with said complex correction data item, a refresh entity 305 of said SF preconstructed complex time samples, said update implementing an accumulation associating two by two said SF complex temporal samples associated with said complex correction data with said SF preconstructed complex temporal samples resulting from said pre-construction phase when m = 1, or with said SF temporal samples complexes resulting from an update step implemented for a previous data stream when 1 <n-KM, delivering SF complex time samples used for the next data stream, a spreading unit 105 implementing a multiplication of each data stream of said CDMA signal modulated by a constellation symbol from said preprocessing step by a spreading factor spreading code SF; a scrambling unit 105 of said CDMA signal from said spreading unit, a unit for transmitting and / or storing a CDMA signal from said scrambling unit. These means are controlled by the microprocessor of the processing unit 51.

Claims (15)

REVENDICATIONS1. A method of transmitting a code division multiple access signal, said CDMA signal, said CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M distinct data streams forming said CDMA signal, each constellation symbol Xm being identified by an index m, m being an integer such that 1.rrKM, characterized in that it comprises the following steps, for at least one block, of M constellation symbols of said CDMA signal: a preprocessing step (20), comprising: a pre-construction phase (201) of SF preconstructed complex temporal samples representative of said M data streams respectively multiplied by a spreading factor distinct code SF, SF being an integer, and o a phase of correction (202) of the constellation symbols modulating the M data streams of said CDMA signal, said correction phase comprising the following steps, repeated for each flu x of data: detection (2021) of P samples, P being an integer, among said SF preconstructed complex temporal samples resulting from said pre-construction phase when m = 1, or among SF complex temporal samples resulting from a step of update implemented for a previous data stream when 1 <n-KM, having a power greater than a predetermined threshold (a), said detection step delivering said P samples and SF-P null samples, constituting SF time samples complex to correcting, correcting (2022) the constellation symbol modulating said data stream according to said SF complex time samples to be corrected, delivering a complex correction data (Fm) of said constellation symbol Xm, building (2023) SF associated complex temporal samples to said complex correction data, updating (2024) said SF preconstructed complex time samples, ladi the update using an accumulation associating two by two said SF complex temporal samples associated with said complex correction data with said SF preconstructed complex temporal samples from said pre-construction phase when m = 1, or with said SF complex temporal samples from an update step implemented for a previous data stream when 1 <n-KM, delivering SF complex time samples used for the next data stream; - a spreading step (21) implementing a multiplying each m-index data stream of said modulated CDMA signal with a constellation symbol X'm from said preprocessing step by spreading code Sm of spreading factor SE; a scrambling step (22) of said CDMA signal resulting from said spreading step; a step of transmission and / or storage (23) of a CDMA signal originating from said scrambling step. 2. A transmission method according to claim 1, characterized in that said preprocessing step (20) (201) implements at least one step of leveling (20110, 20230) according to which each constellation symbol Xm modulating one of said M data stream is multiplied by a gain factor Gm, and in that prior to said step of transmission and / or storage (24) said method further comprises a power adjustment step (23) whose input is powered by said scrambling step and whose output feeds said transmitting and / or storing step (23) and said step of leveling (20110). 3. A transmission method according to claim 1, characterized in that said complex correction data (Fm) results from the complex correlation (20220) of said SF complex temporal samples to be corrected with SF real time samples associated with an Sm spreading code. spreading factor SF multiplying said data flow of index m. 4. Transmission method according to claim 1, characterized in that said transmission method uses a shaping filter (107) whose interpolation factor L is equal to one. Transmission method according to claim 1, characterized in that said preprocessing step applies at least one delay line (314, 315) of duration equal to said spreading factor SF to said block, of M constellation symbols of said CDMA signal. . 6. Transmission method according to claim 1, characterized in that said pre-construction phase comprises the following steps, repeated for each data stream of index m: pre-construction (2011) of SF complex temporal samples associated with said stream of index data m multiplied by a spreading code Sm, storing (2012) said SF complex temporal samples associated with said data stream, by accumulating associating two by two said SF complex temporal samples associated with said data flow with SF temporal samples Complexes associated with previous data flows. 7. A transmission method according to claim 1, characterized in that said correction step implements a summation of the coordinates of said constellation symbol, called original coordinates of said constellation symbol, with coordinates representative of a complex displacement ( dAm, dBm) of said constellation symbol on the abscissa and ordinate axes of the complex plane of the constellation of said symbol, said complex displacement being selected by means of said complex correction data (Fm), among the complex displacements belonging to at least one one of the following categories: real or imaginary displacement of said complex displacement, negative when the sign of the real part, respectively of the imaginary part of said complex correction data item (Fm) is positive; real or imaginary displacement of the positive complex displacement when the sign of the real part, respectively of the imaginary part, of said complex correction data item (Fm) is negative. 8. A transmission method according to claim 5, characterized in that the absolute value of the real part, respectively imaginary, said complex displacement is proportional to the real part, respectively imaginary, said complex correction data (Fm). 9. Transmission method according to claim 5, characterized in that said correction step also implements a weighting (308) of the real part, respectively imaginary, of said complex displacement as a function of said original coordinates of said constellation symbol. 10. Transmission method according to claim 1, characterized in that it further comprises a step of weighting said complex correction data (Fm) by a weighting factor K. 11. A transmission method according to claim 1, characterized in that it further comprises a switching step for transferring said SF preconstructed complex temporal samples from said pre-construction phase input of said step of detecting said phase. correction. 12. A device for transmitting a code division multiple access signal, said CDMA signal, said CDMA signal comprising blocks consisting of M constellation symbols respectively modulating M separate data streams forming said CDMA signal, each constellation symbol being identified by an index m, m being an integer such that 1.rrKM, characterized in that it comprises, for at least one block, M constellation symbols of said CDMA signal: - a preprocessing unit, comprising: o a module pre-construction complex (302, 303, 1069) of SF preconstructed complex temporal samples representative of said M data streams respectively multiplied by a spreading factor distinct code SF, SF being an integer, and o a correction module constellation symbols modulating the M data streams of said CDMA signal, said correction module comprising the following entities implemented for each data stream: u detecting entity (306) of P samples, P being an integer, among said SF preconstructed complex temporal samples from said pre-construction module when m = 1, or among SF complex time samples from a updating entity for a previous data stream when 1 <n-KM, having a power greater than a predetermined threshold (a), said detection entity delivering said P samples and SF-P null samples, constituting SF complex time samples to be corrected, a correction entity (307) of the constellation symbol modulating said data flow according to said SF complex time samples to be corrected, delivering a complex correction data (Fm) of said constellation symbol, a construction entity (304) of SF samples complex temporal data associated with said complex correction data, an entity for updating (305) said SF temporal samples preconstructed complex orels, said update using an accumulation associating two by two said SF complex temporal samples associated with said complex correction data with said SF preconstructed complex temporal samples from said pre-construction module when m = 1, or with said SF complex time samples from said update entity implemented for a previous data stream when 1 <n-KM, delivering SF complex time samples used for the next data stream; a spreading unit (105) implementing a multiplication of each index data stream m of said modulated CDMA signal by a constellation symbol X'm originating from said preprocessing unit by a factor spreading code Sm spreading SF; a scrambling unit (105) of said CDMA signal originating from said spreading unit; a unit for transmitting and / or storing a CDMA signal originating from said scrambling unit. 13. Transmission device according to claim 12, characterized in that said pretreatment unit comprises at least one level weighting entity (1069, 1070) according to which each Xm constellation symbol modulating a data flow associated with one of said M data stream is multiplied by a gain factor Gm, and in that it comprises a power adjustment unit (106) whose input is fed by said scrambling unit and whose output feeds said transmission unit and and / or storing and said level-weighting entity (1069). Transmission device according to claim 12, characterized in that said correction module further comprises: a summation entity (3100) of the coordinates of said constellation symbol, called original coordinates of said constellation symbol, with representative coordinates a complex displacement (dAm, dBm) of said constellation symbol on the abscissa and ordinate axes of the complex plane of the constellation of said symbol, and a weighting entity (308) of the real or imaginary part of said complex displacement according to said original coordinates of said constellation symbol. A computer program comprising instructions for implementing a transmission method according to claim 1 when the program is executed by a processor.