FR2808361A1 - Method and device for digital signal coding for transmission by modulation of at least one carrier, for use in protected telecommunication system - Google Patents

Abstract

The method for coding a digital signal composed of consecutive binary elements is by mapping onto a set of points of a predetermined constellation having a whole number CC of points, which can be different from a power of two. The device for coding contained in the transmission block (38) of a peripheral station (SPi) in a telecommunication system with a base station (SB) comprises means in the form of markers insertion and data control circuit for regrouping the binary elements of the digital signal into successive binary words of length L, which is obtained from the equation: L = N+C-2N, where N is an integer such that C is in the range from 2N to 2N+1; and mapping circuit utilized for each binary word for mapping on a point among 2N points of the constellation, representing N binary elements in an absolute manner, so that C-2N other binary elements are deductible from former utilization or non-utilization of C-2N points of the constellation, named special points, each representing in a differential manner a binary element among C-2N binary elements of binary word. Each of C-2N binary elements is represented by the special point of corresponding constellation, as a function of occurrence of a predefined event with respect to the value of the considered binary element in a set of consecutive binary words. The predefined event corresponds to a change of value of the considered binary element with respect to its value in the preceding binary word, or to a conservation of value of the considered binary element. The other binary elements are removed from the digital signal and replaced by markers designed to indicate at least one event relative to the other binary elements, and the markers are coded by the intermediary of special points. A predetermined sequence of markers changes the rule of coding. The digital signal is transmitted by the quadrature amplitude modulation (QAM) of a unique carrier, when each binary word modulates consecutively the carrier according to the predetermined constellation, or of a set of carriers, when a set of binary words modulates simultaneously at least one subset of carriers. The digital signal is transmitted according to the method of orthogonal frequency division multiplexing (OFDM). Each point of the constellation corresponds to an amplitude and to a phase of modulation of carrier by the digital signal, and one of the points is situated at the origin of complex plane. The modulation of carrier is controlled by mapping according to the constellation of points. The number of points of the constellation associated with each carrier is calculated as a function of an information representative of the quality of transmission in that channel. The information is received preliminary to the transmission of the digital signal, following a preceding transmission of a digital signal. The reception of digital signal comprises the steps of demodulation of carriers in a manner to obtain a constellation point for each, and of decoding the constellation points according to the rule of coding utilized in transmission in order to obtain binary words of digital signal. The decoding of each constellation point comprises a step of deducing the decoding of special points, as a function of value of binary elements in binary words preliminary received on the same carrier, and a step of extracting from the reconstituted data signal the control data indicative of the quality of transmission, for modifying the rule of coding. The telecommunication system comprises at least one digital signal transmission device, and at least one digital signal reception device. A telecommunication network comprises at least two telecommunication systems.

Description

The present invention relates generally to systems for transmitting a digital signal by carrier modulation (s).

More particularly, the invention relates to a method of encoding a digital signal to be transmitted and composed of consecutive bits, by mapping over a plurality of points of a predetermined constellation having an integer number of points different from a power of two. The invention also relates to a device capable of implementing such a method.

The invention also relates to methods and devices for transmitting and receiving a digital signal by carrier modulation using a coding method of the aforementioned type.

In a system for transmitting a digital signal by modulation of a single carrier, the carrier is modulated by a broadband signal carrying the digital information to be transmitted. In such a system, the symbols constituting the information are transmitted in series, thus occupying a frequency band which must necessarily be greater than the inverse of the duration of a symbol. When the symbol rate becomes too high, it is impossible to guarantee that the channel has identical amplitudes and phase characteristics over the entire frequency space constituting the bandwidth. To overcome this problem, a technique consists of distributing the signal to be transmitted in parallel over several carriers, individually modulated at low bit rate over a portion of the bandwidth. Thus, for each carrier, it is more likely that the frequency and phase characteristics of the transmission channel will be identical for all frequencies constituting the portion of bandwidth used.

A commonly used multi-carrier modulation technique is that known as "Orthogonal Frequency Division Multiplexing" and hereinafter referred to as OFDM. OFDM is a system that distributes the transmission of digital data from one signal to multiple carriers (approximately 256 to 1024 carriers), which are mutually orthogonal, by modulation of these carriers. Each carrier is modulated using for example a quadrature amplitude modulation (QAM) known in English under the designation "quadrature <I> amplitude modulation" </ I> (QAM).

In a schematic manner, the transmission of data carried by a digital signal according to a multicarrier modulation technique such as OFDM, consists firstly in cutting these data into blocks of data. The signals corresponding to these blocks are then used as baseband signals to modulate (for example in QAM) the carriers. Thus, each block of data constitutes an OFDM symbol. These OFDM symbols are finally transmitted independently of each other.

During modulation, each block of data is divided into binary words. Each of these binary words is the subject of a representation, commonly referred to by the expression "cartographic report" or "mapping" in English, by a point in the complex plane also called "Fresnel space". On the other hand, the discrete set of points available in the complex plane for the modulation of a given carrier is commonly referred to as "constellation". At each of the points of said constellation corresponds an amplitude and a modulation phase. In general, when the properties of the transmission channel used vary according to the frequency, each carrier can use with a given average energy, a different maximum number of constellation points, ensuring an identical minimum error rate.

Furthermore, it is known that the rate of information transmitted by a carrier depends on the number of points in the constellation associated with it. The more points in the constellation, the higher the rate of transmitted information. On the other hand, increasing the number of points in the constellation by keeping the same average signal energy, generally implies a decrease in the minimum Euclidean distance between the points. This therefore induces a weakening of the immunity to noise and thus an increased bit error rate.

Consequently, in systems for transmitting digital signals by modulation of one or more carriers, a major objective is to increase, with a constant transmission quality, the transmitted data rate.

One way of approaching this objective proceeds from the observation that the maximum number of points satisfying a predetermined error rate criterion is often different from a power of two. However, for reasons of ease of coding in binary, one generally codes a number of constellation points equal to a power of two.

Therefore, in cases where the maximum number of points of a constellation satisfying a predetermined error rate criterion is different from a power of two, the transmission rate in the channel can be increased, if is able to encode a number of constellation points different from a power of two.

For example, in US Patent No. 5598435 entitled <I> "Digital Modulation </ I> using QAM with <I> multiple signal point constellations </ I> nof equal to <I> â power of </ I> two ", a technique is described for encoding constellations having a number of points different from a power of two. According to this technique, data is transmitted using multiple carriers by quadrature amplitude modulation (QAM). In general, the carriers carry different numbers of bits (bits), to exploit the fact that the properties of the transmission path are more favorable for some carriers than for others. The process of allocating the number of bits makes it possible to encode constellations having a size (i.e., a number of points) different from a power of two. This is achieved by converting the binary representation to be transmitted into a representation in which the base of each element of the representation corresponds to the number of points in the QAM constellation for a respective carrier (or for that number divided by a power of two) .

The technique described in the aforementioned patent has the particular disadvantage of requiring the prior coding of the signal to be transmitted in a base other than the base two, which implies a relatively large amount of time and computing power.

An aim of the present invention is to provide a technique for encoding a digital signal by cartographic reporting on the points of a constellation, this technique making it possible to increase the bit rate of transmitted information with constant transmission quality. In particular, the present invention aims to allow the increase of the bit rate of transmitted information with constant transmission quality, by coding constellations having a size different from a power of two, without going through a prior coding in a different base. than the base two.

To this end, the present invention relates, according to a first aspect, to a method of coding a digital signal intended to be transmitted and composed of consecutive bits, by mapping over a plurality of points of a predetermined constellation having a whole number. C points different from a power of two.

According to the invention, this coding method is characterized in that it comprises the following steps: - grouping the bits of the digital signal into successive binary words of length L obtained by the following equation <I> L = </ I > N + C-2N where N is the integer such that: 2 '<I> <C <<2N +' <I>; </ I> - for each of the binary words, carry out a map report of N bits of the binary word on a point among 2 "points of the constellation, this point among 2" points representing the N bits absolutely, the value of the C-2N other bits of the binary word being deductible from the use or from the previous non-use of C-2N other points of said constellation, said special points, each of which differentially representing a binary element among the C-2 'other binary elements of the binary word.

Thus the coding method according to the invention makes it possible to code the binary information to be transmitted on constellations having a number of points different from a power of two, without going through a prior coding of the information in a base other than the base two. This is made possible by using C-2 'points of the constellation to differentially represent a portion of the binary information to be transmitted. The use of these C-2 "" "other" points of the constellation makes it possible to increase the information rate transmitted, and with constant transmission quality since the maximum number of constellation points ensuring a given minimum error rate is preserved.

According to a main characteristic of the method of coding a digital signal according to the invention, each of the C-2Binary elements is represented by the corresponding special constellation point, according to the occurrence of a predefined event relating to the value of the binary element considered in a plurality of consecutive binary words of the digital signal.

In this way, each of the C-2N bits is differentially represented by a special point in the constellation. The event considered may be predefined, for example, according to the statistical properties as to the value of the bits constituting the consecutive binary words of the digital signal. According to a preferred embodiment of the invention, the predefined event corresponds to a change of value of the binary element considered among said C-2 "binary elements, with respect to its value in the previous binary word.

According to an alternative embodiment, the predefined event corresponds to a value retention of the considered binary element among said C-2 "binary elements, with respect to its value in the previous binary word.

According to another embodiment of the invention, the C-2 "bits are deleted from the digital signal and replaced by markers intended to indicate at least one event relating to the value of the C-2" bits in the digital signal , the markers being coded via special points.

In this way, after having reported the occurrence of a given event relating to the value of one or more bits in the digital signal, by the use of the corresponding special point, it will not be necessary to use again this special point as long as this event does not reproduce, which makes it possible to reduce the number of constellation points used during a plurality of successive binary words of the digital signal. Thus reducing, on average, the number of constellation points used for the same amount of information transmitted, further improves the quality of the transmission.

According to yet another embodiment of the invention, a predetermined sequence of the markers in the digital signal induces a change in the coding rules.

In this way, it will be possible to automatically change the coding rules as a function, for example, of the statistical properties of the digital signal, thereby making the coding adaptive.

According to a first preferred embodiment of the invention, the digital signal is transmitted by quadrature amplitude modulation of a carrier, each of the binary words of the digital signal subsequently modulating the carrier according to the predetermined constellation. According to a second preferred embodiment of the invention, the digital signal is transmitted by quadrature amplitude modulation of a plurality of carriers, a plurality of binary words of the digital signal simultaneously modulating at least one subset of carriers among the plurality of carriers, each according to the predetermined constellation.

In practice, in this second embodiment, the digital signal is transmitted according to an Orthogonal Frequency Division Multiplexing (OFDM) technique.

According to a preferred embodiment of the invention, each of the points of the constellation corresponding to an amplitude and to a modulation phase of a carrier by the digital signal, one of the points of the constellation is advantageously situated at the origin of the complex plan.

Using the point at the origin of the complex plane makes it possible to increase the number of points of the constellation and thus the information rate transmitted, without reducing the quality of the transmission. Indeed, the minimum Euclidean distance between the other points of the constellation remains unchanged. In particular, in such a constellation having a point situated at the origin, the minimum Euclidean distance between the points is increased (hence the immunity to the noise) compared to a constellation having the same number of points all distributed equidistantly on a circle of constant amplitude of the complex plane.

Correlatively, the present invention relates to a coding device of a digital signal characterized in that it comprises means adapted to implement a method of encoding a digital signal as briefly described above.

According to a second aspect, the present invention relates to a method for transmitting a digital signal by quadrature amplitude modulation of at least one carrier whose modulation is controlled by cartographic transfer according to a constellation of points, characterized in that the digital signal is coded according to a method of coding a digital signal as briefly described above. According to a third aspect, the present invention relates to a method for transmitting a quadrature amplitude modulation digital signal of a plurality of carriers whose modulation is controlled by cartographic transfer according to a constellation of points, at least one carrier. being modulated according to a constellation having a number C of points different from a power of two. In short, in accordance with the invention, this method of transmitting a digital signal is characterized in that it comprises the following steps: grouping the bits of said digital signal into successive binary words, each of which being intended to be transported by a carrier, the length L of each of said binary words being obtained, when the number C of points of the constellation for the carrier considered is different from a power of two, by the following equation L = N + C-2 " where N is the integer such that: 2 '<C <2' + '; and for each of said binary words for which the carrier is modulated according to a constellation having a number C of points different from a power of two - perform a map report of N binary bits of the binary word on a point among 2 "points of said constellation, said point among 2" points representing the N binary elements in an absolute manner, the value of C -2 "other bits of the binary word being deductible from the prior use or non-use of C-2N other points of said constellation, so-called special points, each of which differentially represents a binary element among the C-2 "other bits of the binary word.

According to a fourth aspect, the invention relates to a method for receiving a digital signal transmitted according to a method for transmitting a digital signal as briefly described above, characterized in that it comprises the following steps - demodulating the carriers so as to obtain for each of them a constellation point; decode each of the constellation points obtained, according to the coding rules used during the transmission of the digital signal, so as to obtain the corresponding binary words of the digital signal.

Correlatively, the invention relates to a device for transmitting a digital signal and a device for receiving a digital signal characterized, respectively, in that they comprise means adapted to implement, respectively, a method of transmission of a digital signal and a method of receiving a digital signal, as briefly described above.

The invention also relates to a telecommunications system comprising at least one device for transmitting a digital signal according to the invention, and at least one device for receiving a digital signal according to the invention.

The invention also relates to a telecommunications network comprising at least two telecommunications systems as mentioned above.

The present invention further relates to a computer system, such as a computer, having means adapted to implement a method of coding a digital signal, and / or a method of transmitting a digital signal, and or a method of receiving a digital signal, as briefly described supra.

The present invention also relates to a computer program, comprising one or more instruction sequences, able to implement a method of encoding a digital signal according to the invention, and / or a method of transmitting a signal. digital signal according to the invention, and / or a method of receiving a digital signal according to the invention.

The invention also relates to an information carrier, such as a diskette or compact disc (CD), characterized in that it contains such a computer program.

Other features and advantages of the invention will become apparent in the description, hereinafter, of preferred embodiments of the invention. In the accompanying drawings, given by way of nonlimiting examples - Figure 1 schematically illustrates a telecommunications network in which the invention can be implemented; FIG. 2 is a block diagram illustrating a peripheral station of the network of FIG. 1 comprising a transmitting and receiving device according to the invention; FIG. 3 is a block diagram representing the transmission block of the peripheral station of FIG. 2; FIG. 4 is a block diagram representing the marker and control data insertion unit of the transmission block shown in FIG. 3; FIG. 5 is a block diagram representing the reception block of the peripheral station of FIG. 2; FIG. 6 is a block diagram representing the unit for analyzing the channel and decoding the data of the reception block illustrated in FIG. 5; FIG. 7 represents an example of constellations used according to a preferred embodiment of the invention; FIG. 8 represents a constellation that can be used according to an alternative embodiment of the invention; FIG. 9 illustrates an exemplary data stream according to the invention, corresponding to the transmission of OFDM symbols.

FIG. 10 is a flowchart illustrating the methods of encoding and transmitting a digital signal according to the invention.

Referring to the <B> figure </ B> 1, we will describe a telecommunications network in which the invention can be implemented.

The network 10 consists of a station SB, called base, and R peripheral stations SP 1 to SPR, called peripheral stations. The peripheral stations are remote from the base station SB and are likely to move relative thereto.

Each of the peripheral stations SPi (i between 0 and R) is connected to the base station SB by a radio link. The transmission of digital signals between each of the peripheral stations and the base station is performed according to a multi-carrier modulation technique. In practice, in the embodiment chosen and illustrated by the figures appended to this memo, this modulation technique is the known orthogonal frequency division multiplexing (OFDM) technique.

In this context, each of the carriers undergoes an amplitude-phase modulation, commonly known as quadrature amplitude modulation (QAM) (in English <I> "quadrature amplitude modulation" </ I> QAM). Of course, this type of modulation includes pure phase modulations and pure amplitude modulations.

It will be noted here that although in the preferred embodiment described here, the transmission mode used is of the multicarrier modulation type, the invention also applies to a single carrier modulation transmission mode.

Each of the stations of the network 10, peripheral or basic, constitutes a telecommunications system comprising a transmitting device and a device for receiving a digital signal according to the invention.

A peripheral station SPi of the network 10 will now be described in connection with FIG. 2.

A peripheral station SPi according to the invention comprises a data source 20 connected to a transmitting and receiving device 22.

The data source is, for example, a database, or a data processing computer device such as a digital camera, a fax machine, a scanner, a digital camera.

The device 22 for transmitting and receiving the peripheral station comprises a unit 24 for processing transmitted or received data, a transmission block 38, a reception block 72 and a radio module 40.

The data processing unit 24 comprises temporary data storage means such as a random access memory (RAM) 28, permanent data storage means such as a read-only memory (ROM) 30, input means 32 characters, for example a keyboard, data display means 34, for example a screen, data input / output means 36, and a central processing unit (CPU) 26, for example a microprocessor.

As a matter of course, the operation of the peripheral station is as follows.

<U> For transmission </ U> The data source 20 delivers data 100 to be transmitted to the data processing unit 24. The data 100 may be, for example, digital images represented in bitmap mode. These data are stored for example in the RAM 28 and then undergo appropriate processing, for example, in the case of images, compression, or an operation of enlarging the format of the images by interpolation.

Once processed, the data to be transmitted are delivered in the form of a digital signal 600 to the transmission block 38. The latter also receives as input two control signals 740 and 780 from the reception block 72 (these signals will be described further in the description).

The transmission block 72 outputs a signal 500 to the radio module 40. The signal 500 is an OFDM-type modulated analog signal carrying the so-called "useful" data (as opposed to the control data) coming from the unit. 24 data processing and control data whose nature will be described later in the description.

Finally, the radio module transmits a radio signal (not shown) containing the aforesaid useful data and control data on the transmission channel connecting the peripheral station (SPi) to the base station (SB). <U> For reception </ U> The radio module 40 first receives a radio signal (not shown) from the base station and outputs a signal 800. The signal 800 is an OFDM type modulated analog signal. carrying, on the one hand, useful data from the corresponding data processing unit of the base station SB, and, on the other hand, control data inserted by the transmission block of the base station SB. . The signal 800 is input to the receive block 72 which outputs the two above-mentioned control signals 740 and 780, as well as a signal 790 containing the useful dopants from the base station.

Finally, the signal 790 is supplied to the data processing unit 24 for appropriate processing. The processed data can then be stored, for example in the data source, or in storage means such as RAM 28 or a hard disk (not shown) contained in the unit 24.

According to a preferred embodiment, the transmission block 38 and the reception block 72 are made by assembling hardware elements <I> ("hardware" </ I> in English). These hardware elements are more specifically digital electronics cards.

However, alternatively or combined, one can also use software elements ("software" in English). In the latter case, the different digital processing performed by the transmission and reception blocks when these are made in a material way, can be provided by the data processing unit 24, and the entire telecommunications system. (Peripheral station or base station) may then consist of a computer system such as a computer.

Typically, the radio module 40 comprises a known radio transmitter and receiver, one or more mixers, one or more amplifiers, various filters and a radio antenna.

We will now describe, in connection with FIGS. 3 and 4, the transmission block 38 of the transmitting and receiving device 22 of the peripheral station (SPi). FIG. 3 represents the various elements constituting the transmission block 38. FIG. 4 details the unit 60 for inserting markers and control data of the transmission block.

With reference to FIG. 3, the transmission block 38 of the peripheral station SPi comprises the following elements - a unit 60 for inserting markers and control data, - a unit 42 for cartographic reporting (in English "mapping circuiP"). '), - a serial / parallel register (S / P) 44, - an inverse fast Fourier transform (IFFT) calculation unit 46, - a parallel / serial register (PIS) 48, and - a digital / analogue converter (DAC).

As mentioned above, the transmission block 38 receives as input the digital signal 600 containing the data to be transmitted from the data processing unit 24, as well as the control signals 740 and 780 originating from the reception block 72.

At the output, the transmission block 38 delivers the signal 500 (see FIG 2), that is to say an OFDM type modulated analog signal carrying the data to be transmitted initially contained in the digital signal 600.

The methods for encoding and transmitting a digital signal in accordance with the invention are implemented jointly by the marker and control data insertion unit 60 and by the cartographic transfer unit 42, which are therefore essential elements of the invention in the context of the embodiment chosen and shown.

On the other hand, the other elements of the transmission block 38, that is to say the serial / parallel register (SIP) 44, the inverse fast Fourier transform (IFFT) calculation unit 46, the parallel register / series (PIS) 48 and the digital-to-analog converter (DAC), are conventional elements used to carry out the modulation of multiple carriers by a digital signal according to the OFDM technique. Consequently, these elements will not be described here in detail.

From a functional point of view, the marker and control data insertion unit 60 receives the digital signal 600 containing the data to be transmitted, as well as the control signals 740 and 780.

As will be explained below, the signal 740 is a signal containing information representative of the transmission quality of the transmission channel in the direction: base station (SB) to peripheral station (SPi). Conversely, the signal 780 is a signal containing information representative of the transmission quality of the transmission channel in the direction: peripheral station (SPi) to base station (SB). The unit 60 outputs signal 680 which consists of a sequence of groups of bits (bits), each group of bits being fully transportable by an OFDM symbol.

The signal 680 is then input to the map transfer unit 42, as well as the control signal 780. The unit 42 performs, for each group of bits (bits) of the signal 680, the map report of the group considered according to the constellations determined for the carriers so as to control their modulation.

The unit 42 outputs the signal 420 consisting of a succession of binary words each consisting of several bits (bits). Each of these binary words will be transported by a carrier by cartographic report on a point of the constellation associated with the carrier considered, the constellation point used corresponding to a phase (the argument) and an amplitude (the module) of modulation of the carrier.

The signal 420 is input to the serial / parallel register 44 which outputs as many parallel signals as used carriers. Each of these signals corresponds to an input of the Fast Fourier Transform (IFFT) circuit 46. The IFFT circuit 46 is responsible for generating the modulation signals for the real and imaginary part of each of the complex amplitudes represented by the aforementioned bit words. The parallel outputs of the IFFT circuit 46 are then input to the parallel / serial register (PIS) 48 which outputs a corresponding serial signal. This serial signal is input to the digital-to-analog converter (DAC) 50 which outputs the signal 500. As previously mentioned, the signal 500 is then supplied to the radio module 40 which ensures its transmission.

In connection with <B> Figure 4, the unit 60 for inserting markers and control data of the transmit block 38 will now be described in detail.

As shown in FIG. 4, the unit 60 for inserting markers and control data consists of several elements - a buffer unit <I> 64 </ I> ("buffer" in English), - a unit d analysis 66, and a marker insertion unit 68.

The purpose of the buffer unit 64 is to insert in the data to be transmitted (called "useful data") contained in the digital signal 600 the data transported by the signal 740, that is to say the control data representative of the quality of the transmission channel in the receiving direction. The unit outputs a composite signal 640 in the sense that it comprises useful data and control data to be transmitted. According to the embodiment chosen, the unit 64 acts as a buffer for the data corresponding to one or more OFDM symbols.

The signal 740 contains data representative of the transmission quality in the base station to peripheral station direction, i.e. when the peripheral station is receiving. As will be described later, the signal 740 is generated by the reception block 72 of the station, during a reception prior to the current transmission. The method by which information indicative of the quality of transmission over a channel is known is known to those skilled in the art. It is possible, for example, to use techniques implementing pilot digital sequences or pilot carriers that are known at one and the same time. transmitter and receiver.

The signal data 740 that has been inserted into the wanted signal, i.e. the digital signal 600, to form the signal 640 is intended for the remote transmitter (the base station) to adapt future transmission to receiving conditions of the receiver (the peripheral station).

The analysis unit 66 receives the aforementioned signal 640 and the signal 780 as input. The signal 780 contains data representative of the transmission quality in the peripheral station to base station direction, that is to say when the Peripheral station is transmitting. The signal 780 is also generated by the reception block 72 during a previous reception of a signal, denoted Sb, transmitted by the base station. The data contained in the signal 780 is obtained from the control data inserted by the base station buffer unit during the Sb signal transmission process. According to the embodiment described here, the signal 780 contains data making it possible to establish an association between each of the carriers used by the emispion block 38 and a particular constellation. More specifically, the signal 780 enables the unit 66 to associate with each carrier a constellation such that, starting from a given maximum energy and for a fixed minimum error rate, the number of different points (size) of the constellation is the highest.

According to a preferred embodiment of the invention, once the maximum possible constellation size determined for a given carrier, this carrier is associated with a predetermined constellation whose size corresponds to the determined maximum size.

When this maximum size is determined to be different from a power of two, in accordance with the invention, the carrier in question will be associated with a constellation having the same number (different from a power of two) of points.

By way of example, FIG. 7 represents a set of predetermined constellations according to a preferred embodiment of the invention. As can be seen in Figure 7, a set of nine constellations is predefined, having respectively a size (C) of 2 to 9 points. Among these constellations, five have a number of points different from a power of two: those of size C = 3, 5, 6, 7 and 9. We will come back in more detail about these constellations later in the description.

Back to <B> Figure 4, </ B> and in relation to <B> Figure 10 </ B> which illustrates the main steps of the coding and transmission processes of a digital signal in accordance with the The invention will detail the operation of the analysis unit 66 associated with that of the marker insertion unit 68 and that of the cartographic transfer unit 42.

The unit 66 begins by reading the binary data of the signal 640 to be transmitted as indicated in the step E101 of FIG. 10. <B> It </ B> is the data that can be transmitted by an OFDM symbol.

Then, as previously mentioned, the unit 66, through the signal 780, determines the constellation associated with each of the carriers and obtains the size C (number of points in the constellation) as indicated in the step E103.

Then, in step E105, the unit 66 determines whether the size of the constellation C is a power of two.

If the constellation has a size equal to a power of two, we are in the classical case. There then exists an integer N such that C = 2 ", and one goes to step E109 in which the data to be transmitted are grouped into successive binary words of length L with L = N (E107), that is, that is, each comprising L bits (bits).

Then, in the next step (E111), for each binary word obtained by the preceding grouping, all the bits (L bits) constituting this binary word will be the subject of a cartographic report in a conventional manner on one point among the C = 2 "points of the constellation The cartographic report is made by the unit 42 of cartographic report (Fig. 3).

<U> Note </ U>: In order to simplify the written expression, we will sometimes use in the rest of the presentation, instead of the expression: "... is the subject of a report cartography ... ", the expression" ... is represented ... ". Likewise, the terms "binary element" and "bit" will be used equivalently.

It will be noted here that the invention also applies to the case where the transmission of the digital signal is carried out by quadrature amplitude modulation of a single carrier. In this case, each of the binary words of the digital signal consecutively modulates the carrier according to the constellation which is predetermined.

On the other hand, in the case of a modulation of a plurality of carriers, as is the case in the preferred embodiment, a plurality of binary words simultaneously (in parallel) modulate at least a subset of associated carriers. at the same constellation, among the totality of carriers.

Returning to step E105, if the size C of the constellation is not equal to a power of two, according to the invention, the analysis unit 66 calculates, in step E113, the total number N such that 2'v <C <2 "+ '(1) In the next step, E115, we calculate the length L used to group the binary data of the signal to be transmitted by the following equation L = N + C-2 ", <I> (2) </ I> It then proceeds to step E117 in which the data to be transmitted are grouped into successive binary words of length L calculated according to equation (2).

Finally, in step E119, according to the invention, the cartographic report of each of the binary words on the corresponding constellation is carried out (by the unit 42 of cartographic report) in the following manner.

For each binary word of length <I> L = </ I> N + C-2 ", a cartographic report of N binary bits (bits) of the binary word is carried out on a point among 2" points of the constellation. This cartographic report is of the classical type, that is to say that this point (among 2 "points) represents the N binary elements (bits) of the binary word in an absolute way.These 2" points of the constellation considered are said to be " normal ".

In accordance with the invention, the other bits (bits) of the binary word in question which are of C - 2 'are not represented directly on one of the points of the constellation. However, as will be explained later in the description, the value of these other bits will be deductible from the previous use or non-use of the C-2 "other points of the constellation for the modulation of the carrier considered. These points are called "special points." Indeed, each of these special points of the constellation represent in a differential manner the C - 2 "other bits of the binary word.

According to the invention, each of the C-2 "bits of the binary word is encoded by the corresponding special constellation point, as a function of the occurrence of at least one predefined event relating to the value of this bit in a plurality of consecutive binary words of the digital signal (signal 640).

This predefined event corresponds, according to a preferred embodiment of the invention, to a change of value of the binary element considered (among the C-2 "bits of the binary word) with respect to its value in the binary word which precedes immediately the one in the course of cartographic reporting on the constellation associated with the carrier considered.

According to an alternative embodiment of the invention, the event corresponds to the conservation of the value of the bit considered among the C-2 "bits of the binary word, with respect to its value in the binary word previously coded on the constellation.

We can also combine the various events mentioned above.

In practice, the C-2 'bits of the binary word considered are not transmitted by modulation of the carrier associated with the constellation considered. They are actually removed from the digital signal to be transmitted (signal 640) and are replaced by markers inserted in the signal, each intended to indicate the occurrence of one or more events relating to the value of one of the C-2 bits in the binary word that precedes the one being coded.

The markers are represented by the special points of the constellation considered.

According to the invention, a marker is a data item inserted in the digital signal and which carries information about the state of one of the bits of the following binary word (in time) in the data stream contained in the digital signal. According to a particular embodiment of the invention, a marker is represented by a voltage level that can not be confused with a logic high level ("1" logic) and a logic low ("0" logic). If, for example, the logic levels "1" and "0" are respectively coded by the voltage levels, respectively, +5 volts and 0 volts, a negative voltage level can be used for a marker. Each marker has a level of tension of its own to be distinguished from other markers.

In order to determine the appropriate place of the stream of data to be transmitted in which to insert a marker as well as the nature of the marker, the analysis unit 66 proceeds to an analysis of the statistical properties of the data. This analysis may for example determine that on a large sample of successive binary words intended for the same carrier, a bit of the same rank for all these binary words remains practically unchanged. In this case, it will be possible to delete this bit from the data stream, the appropriate marker being inserted in the data stream, for example before the binary word in which the bit considered passes to the value 1. In this example, this marker is used to signaling the change of state of the considered rank bit in the following binary word in the data stream.

The actual insertion of the markers into the data stream is performed by the marker insertion unit 68. For this purpose, the analysis unit 66 supplies the unit 68 with a signal 650 consisting of bits allowing the unit 68 to know the nature and the position of the markers to be inserted in each group of bits of the signal 640 which will be used for the generation by the unit 68 of a group of data that can be fully transported by an OFDM symbol. Such a group of data is thus constituted at the output of the unit 68, binary words from the signal 640 and markers inserted at certain places interspersed between the words.

According to a preferred embodiment, the markers used are responsible for indicating the occurrence of the event, denoted E0, corresponding to the change in value of the bit coded by the marker; and the occurrence of the event, denoted E1, corresponding to the conservation of the previous value of the bit coded by the marker. In this embodiment, EO is the event considered by default.

In a variant of the invention, the marker corresponding to a bit of position determined in a binary word is inserted by the unit 68 as a function of the signal 650, between two successive OFDM symbols allocated to the same carrier, when there is occurrence of a predetermined event on this bit occurring between these two successive symbols.

In this embodiment, the number N is determined such that 2 "<I> <C <2N + ', where N is the number of bits of a binary word encoded by normal points of a C constellation points, and the C-2 'number of bits represented using the C-2N special points of the constellation, for a number, noted Q, of OFDM symbols both large enough that the marker coding improves the throughput compared to a system without a marker, and at the same time small enough so that the transmitter can adapt its number of constellation points per carrier relatively quickly to the change of the characteristics of the transmission channel.This number of OFDM symbols is either definitively fixed, it is variable as a function of the temporal stability of the characteristics of the channel (the number Q is then determined by the unit 66) At the beginning of each sequence of Q OFDM symbols, a sequence of predetermined constellation points distributed over one or several carriers indicates to the receiver the beginning of said sequence. On each of the carriers, at the beginning of each sequence of Q OFDM symbols, the bits coded by markers all have a default value, for example zero, and the events coded on these bits are predetermined, for example the aforementioned event EO for all bits.

In a preferred embodiment of the invention, the marker associated with the special point on which a position bit determined in a binary word is coded, is inserted by the unit 68 between two successive symbols in the serial stream of data transmitted by the data source, when there is occurrence of a predetermined event on this bit occurring between these two successive symbols. The appearance of an event being indicated by the pilot signal 650.

An OFDM symbol is described by several OFDM sub-symbols adjacent to each other, for which the constellation allocated to each carrier is identical. In this embodiment, at the beginning of the OFDM sub-symbol, the bits coded by markers all have a default value, for example zero, and the events coded on these bits are predetermined, EO for example, for all. The appropriate markers at the beginning of the OFDM sub-symbol act in reference to these default values. If there is an error in the receiver's recognition of a constellation point associated with a marker, the error does not propagate further than a complete OFDM symbol. This system is particularly interesting when the number of carriers is large and a large proportion of carriers have similar transmission characteristics. A marker can not be conventionally placed at the end of the OFDM sub-symbol or of the OFDM symbol or, in the embodiment mentioned above, in a sequence sequence of Q OFDM symbols.

For this coding to be interesting the predetermined event must occur sufficiently infrequently. If this event occurs too often, another event with a satisfactory frequency of occurrence must be taken into account.

To achieve this, as mentioned above, the data analysis unit 66 provides the marker insertion unit 68 with a control signal 650. The unit 66 detects the occurrence of events on the binary elements intended to be encoded by the markers and according to the frequency of appearance of these events makes a decision on the coding. If the coding is sufficiently efficient, it is kept as it is, otherwise the unit 66 changes the coding by inserting one or more appropriate markers. The coding is thus adaptive.

The table presented in the appendix provides an exemplary embodiment in which the coding uses several markers whose meaning depends on their sequence of appearance in the data stream.

According to the invention, it can be provided in the coding that a predetermined sequence of markers induces a change in the coding rules.

In the coding methods of the prior art, the number of bits coded per constellation point and therefore the average number of bits transmitted per second (bauds) is equal to D1, with D1 = 1092 (C) where C is the number of constellation points.

Par conséquent le codage selon l'invention apporte une amélioration du point de vue du nombre moyen de bits transmis par seconde (bauds) si l'on a D2 plus grand que D1, ce qui est réalisé pour

Dans un mode préféré de réalisation de l'invention, lorsque la taille de la constellation C est supérieure ou égale à 6, la modulation utilisée qui est adaptée au rapport signal à bruit sur le canal et au taux d'erreur voulu pour chacune des porteuses, est une modulation de phase à C-1 états dont les points de constellation associés sont répartis de manière équidistante sur un cercle de même énergie dans le plan complexe encore appelé plan de Fresnel. The coding according to the invention makes it possible to increase the number of bauds for a constellation of identical size C. Let D2 be the number of bauds in the case of coding according to the invention. Of course D2 depends on the statistical properties of the message to be transmitted, however the coding always adapts to the statistical properties of the message so as to optimize the transmission. This will be highlighted in the following. Suppose to simplify the discussion that all the carriers used are associated with a constellation of the same size C with N normal points and S = C - 2 "special poipts.On the other hand suppose that the number of bits in the message to be transmitted either <I> L = M x P </ I>("x" denotes the multiplication operator), where M denotes the number of packets (bits) to be transmitted and P denotes the size of each packet (in number of In addition, Tk denotes the number of events for the whole message which concern a predetermined position bit in each data packet P. With these assumptions, the number of bauds, D2, in the coding according to the invention, by the following formula

Therefore the coding according to the invention provides an improvement in terms of the average number of bits transmitted per second (bauds) if D2 is larger than D1, which is done for

In a preferred embodiment of the invention, when the size of the constellation C is greater than or equal to 6, the modulation used which is adapted to the signal-to-noise ratio on the channel and to the desired error rate for each of the carriers , is a phase modulation at C-1 states whose associated constellation points are distributed equidistantly on a circle of the same energy in the complex plane also called Fresnel plane.

According to the invention, there is an additional state of zero amplitude and zero phase which corresponds to the origin point in the Fresnel plane.

When the constellation size C is less than or equal to 5, the modulation used is a phase modulation with C states whose associated constellation points are distributed equidistantly on a circle of the same energy in the Fresnel plane. According to an alternative embodiment of the invention, the modulation used for each carrier is a phase and amplitude modulation (QAM) at C states whose associated constellation points are distributed over several energy circles, with possibly one additional point found at the origin.

Figure 8 gives an example of such a constellation. As can be seen in the figure, the represented constellation has 17 points. Among these points, 8 points are distributed equidistantly on a circle (the outer circle) of the complex plane such that its radius is equal to the square root of the maximum energy, Em, available on the transmission channel. 8 other points are distributed equidistantly on a circle (the inner circle) such that its radius corresponds to 0.63 the square root of the maximum energy Em. Moreover, each of the points of the inner circle is out of phase by n18 modulo 7r / 8 relative to the points of the outer circle. In addition, a point is found at the origin.

In another preferred embodiment of the invention, the maximum constellation size is limited to 11 and the constellations used by the map transfer unit 42 are defined by the following table

<SEP> size of <SEP> The <SEP> Number <SEP> of <SEP> points, <SEP> Nm, <SEP> times <SEP> of <SEP> point <SEP> to <SEP> Distance <SEP> minimum
<tb> constellation <SEP> way <SEP> equidistant <SEP> on <SEP> the <SEP> circle <SEP> the origin <SEP> Euclidean
<tb> of energy <SEP> ie <SEP> maximum <SEP> Em
<tb> 2 <SEP> 2 <SEP><I> 0 <SEP> 2 <SEP> Em </ i>
<tb>. <SEP> 3 <SEP> 0 <SEP> 9.73 <SEP> Em
<tb> 4 <SEP> _ <SEP> _ <SEP> 4 <SEP> 0 <SEP><I> 1. <SEP> 41 <SEP> Em </ i>
<tb> 5 <SEP> 5 <SEP> 0 <SEP> 1.20 <SEP> Em
<tb> 6 <SEP> 5 <SEP> 1 <SEP> Em
<tb> 7 <SEP> 6 <SEP> 1 <SEP> Em
<tb> 8 <SEP> 7 <SEP><I> 1 <SEP> 0.87 <SEP> Em </ i>
<tb> 9 <SEP> 8 <SEP><I> 1 <SEP> 0.77 <SEP> Em </ i>
<tb> 10 <SEP> 9 <SEP> 1 <SEP><I> 0.68 <SEP> Em </ i>
<tb> 11 <SEP> 10 <SEP><I> 1 <SEP> 0.62 <SEP> Em </ I> We will now describe an example of application of the methods of encoding and transmitting a digital signal according to the invention. <U> Example </ U> In the example chosen, we have Z = 64 useful carriers and 8 types of predefined constellations. These predefined constellations are for example those represented in FIG. 7.

As can be seen in Figure 7, among these constellations, only constellations Cn2, Cn4 and Cn8 include a number of points equal to a power of two, respectively: 2 ', 22, 23.

For the aforementioned constellations, the cartographic report will be carried out in a conventional manner, without using special points.

On the other hand, the other constellations Cn3, Cn5, Cn6, Cn7 and Cn9 including a number of points of constellation different from a power of two, they will use special points.

If one applies the formula (1) supra to the constellation Cn5 of size C = 5, one obtains N = 2 since one has: 22 <5 <23.

Thus, in a constellation of size 5 like Cn5, we will have N = 2 normal points, and 5 - 22 = 1 special point.

As previously mentioned, the signal 780 (FIG 4) is a signal representative of the transmission quality in the direction of transmission of a signal (peripheral station to base station).

It is assumed that the signal 780 consists in particular of a 576-bit digital message which can be represented in the form of a table indicating the constellation to be used (among those shown in FIG. 7) for each of the 68 carriers. Of these 576 bits, for each carrier, 6 bits are used to indicate the carrier identification number and 3 bits indicate the type of constellation used. Since there are 64 carriers, this makes a total of 64 x (6 + 3) = 576 bits.

In addition, in signal 780, 32 bits indicate for which OFDM symbols, these indications represented in the table below are to use 6 bits are used to determine the carrier number.

<SEP> binaries <SEP><SEP> hexadecimal notation <SEP><SEP> number of <SEP> _sort
<tb> 000000 <SEP> 00 <SEP> 0
<tb> 000001 <SEP> 01 <SEP> 1
<tb> 001001 <SEP> 09 <SEP> 9
<tb> 001010 <SEP> - <SEP> OA <SEP> 10
<tb> 011011 <SEP><U> 113 </ U><SEP> 27
<tb> 011100 <SEP> 1C- <SEP> $ 2
<tb> 101101 <SEP> 2D <SEP> 45
<tb> 111110 <SEP> 3 <SEP>'<SEP><B><U> 62 </ U></U>
<tb><U> 111111 <SEP> 3F <SEP> 63 </ U> 3 bits are used to determine the constellation used.

<SEP> Binaries <SEP><SEP> Hexadecimal <SEP> Constellation <SEP> notation used <SEP> (Number <SEP> of
<tb> points)
<tb> 000 <SEP> 0 <SEP> 2
<tb> 001 <SEP> 1
<tb> 010 <SEP> 2 <SEP> 4
<tb> 011 <SEP> 3 <SEP> 5
<tb> 100 <SEP> 4 <SEP> 6
<tb> 101 <SEP> 5 <SEP> 7
<tb> 110 <SEP> 6 <SEP> 8
<tb> 111 <SEP> 7 <SEP> _ <SEP> 9 <SEP> _ Thus, if the signal 780 is, expressed in hexadecimal, as follows "1000 100F 003 013 023 033 03 03 073 082 092 030 ... 332 342 <B> 351361371381 </ B> 391 3A1 31311 3C1 3D1 3E1 3F1 "This therefore means that the peripheral station will use in its transmission towards the base station, for the OFDM symbols indexed from 1000 to 100F, a constellation of size 5 (identified by "3" in hexadecimal) for the first 8 carriers (00 to 07 in hexadecimal), a size 4 constellation (identified by "2" in hexadecimal) for the next 45 carriers (08 to 34 in hexadecimal) and a constellation of size 3 (identified by "1" in hexadecimal) for the last 11 carriers (35 to 3F in hexadecimal). Thus in this example, for this OFDM symbol, the constellations used will, with reference to FIG. 7, those identified by the reference signs: Cn5, Cn4 and Cn3.

Now consider the first 8 carriers. In the above example, signal 780 indicates that a 5-point constellation should be used for the first 8 carriers (out of 64). Therefore, in this constellation (Cn5), 4 normal points can represent two bits (bits) absolutely, while the special point can represent 1 binary element in a differential way. The size constellation 5 will therefore represent 3 binary elements.

Suppose now that in the signal 640 (Fig. 4) containing the data to be transmitted, this data starts with the following bit sequence (bits) Signa1640: "001011010010011000011011010000111011100010011011 ..." The analysis unit 66, after signal 780, must use a 5-point constellation for the first 8 carriers. Of these 5 points, 4 are normal, 1 is special. The 4 normal points can represent 2 bits. The special point can represent 1 bit differentially. The constellation is therefore used to represent 3 bits. The unit 66 analyzes the signal 640 Signal 640 by 3-bit binary words: "001 011 010 010 011 000 <B> 011011 </ B> 010 000 111 011 100 010 011 011 ..." The unit 66 then analyzes the statistical properties of the signal 640 by 3-bit binary words, thus detects that the first bit of each three-bit binary word does not change in the message portion that can be carried by the first 8 Signal 640 carriers: "001 011 010 010 011 000 011 011 010 000 111 011 100 010 011 011 ... "This first bit can therefore be represented by the special point and the other two bits by the 4 normal points. Since this first bit is always zero, the special point of the constellation (Cn5), for example the point placed at the origin of the complex plane, will not be used to modulate a carrier. Indeed, in this example, by default on reception, the absence of a carrier modulated by a special point is interpreted as meaning that the bit encoded by this special point is zero. The unit 66 then generates the signal 650 which it supplies to the marker insertion unit 68. The signal 650 conveys two pieces of information - the bits to be removed from the signal 640 as represented by markers; - the nature and position of the markers.

For this purpose, the signal 650 comprises a first information in the form of a signal synchronous with the signal 640 such that at each bit represented by a marker in the signal 640 corresponds a "0" in the signal 650; and that every other bit in the signal 640 corresponds to a "1" in the signal 650. This gives in the example "011 011 011 011 011 011 <B> 011011 </ B> 111 111 111 111 111 111 111 111 In addition, the signal 650 includes a second piece of information that provides the bit number position and the nature of the marker to be inserted into the signal 640 after the bits represented by markers are removed.

It is now assumed that the end of signal 640 is as follows "00010101011010111010011011010000 ..." Unit 66, according to signal 780 above, must use a 3-point constellation (Cn3) for the last 11 carriers. On these 3 points, 2 points are normal, 1 point is special. The 2 normal points can represent 1 bit. The special point can represent 1 bit differentially. The constellation serves to represent 2 bits. The unit 66 analyzes the signal 640 "00010101011010111010011011010000 ..." by means of 2-bit words. The unit 66 detects that the first bit of each 2-bit bitword changes little in the message portion that can be carried by the last 11 carrier "00010101011010111010011011010000 ..." This first bit will be represented by the special point and the second bit by the 2 normal points. The other bits, ie beyond the 10th 2-bit words since a carrier will be used to carry the special point representing the marker, will be transported by the next OFDM symbol.

As previously for the first 8 carriers, as regards the last 11, the signal 650 will consist of a first synchronous signal with the signal 640 such that each bit represented by a marker in the signal 640 corresponds to a "0". in the signal 650 and that each other bit in the signal 640 corresponds to a "1" in the signal 650. This gives 1101 01 01 01 01 01 01 01 01 01 "In addition, the signal 650 comprises a second piece of information which provides the position in number of bits and the nature of the marker to be inserted in the signal 640 after deletion of the bits represented by markers.

Thus, in the example above, for the last 11 carriers, a marker coded on the special point S1 (for example the point at the origin) which will serve to modulate the carrier numbered 3A (58 in decimal). As the first 8 carriers and the 45 intermediate carriers each carry 2 coded bits on normal points, the first 53 carriers carry (8 + 45) x 2 = 106 bits. The carriers 53 to 57 (39 in hexadecimal) each carry a bit. encoded on two normal points, which is 5 bits in total. As a result, the marker must be inserted after the 111th bit of signal 640.

In the signal 650, the position of the marker will therefore be represented by 8 bits.

Thus, in this example, the second information conveyed by the signal 650, is the following "6F FF 1" (in hexadecimal) where <B> "6F" </ B> ("111" in decimal) indicates the position after which must be inserted the marker; "FF" is a separator used to separate from the position data field of the nature data field of the marker; "1" is the number of the marker to insert.

For example, in the case where three different markers can be used, as shown in the table in the appendix, the marker-marker number correspondence is as follows

<SEP> bits <SEP> (bits) <SEP><SEP> hexadecimal notation <SEP><SEP> statement for <SEP><SEP> circuit <SEP> 42 <SEP> (mapping)
<tb> 01 <SEP> 1 <SEP> Marker <SEP> 1
<tb> 01 <SEP> 2 <SEP> Marker <SEP> 2
<tb> 11 <SEP> 3 <SEP> Marker <SEP> 3 <U> Note </ U>: In the example given here, the special points S1, S2, S3 are used in the corresponding constellations consecutively to the appearance in the signal of data to be transmitted markers M1, M2, M3 with the meaning mentioned in the table in the Appendix.

The example that has just been described is illustrated in FIG. 9 which represents a data flow intended to modulate the carriers used for the transmission of five OFDM symbols. In FIG. 9, the data corresponding to five symbols respectively 901, 903, 905, 907, 909 follow one another in time.

The symbol 901 corresponds to the example above in which the first 8 carriers use a constellation of size 5. Only normal points (N) are used to modulate the carriers and transmit the last two bits (xx) of the binary words. The first bit is always set to 0. The data corresponding to these 8 carriers using the same constellation, form a first OFDM sub-symbol. The next 45 carriers use a size 3 constellation and the data carried by these carriers represents a second OFDM sub-symbol. The two bits constituting the data are encoded only by normal points (N). Finally a third sub-symbol is transported by the last 11 carriers that use a constellation of size 3 with 1 special point used (S1) to signal a change of value (transition from "0" to "T") of the first bit of the words binaries represented by the points of the constellation. This special point is modulated by the carrier of number 3A (58 in decimal). The second bit of the binary words (x) is encoded by a normal point (N).

Still in Figure 9, the symbol 903 following the symbol 901 uses the same constellations per carrier as the symbol 901. As can be seen the 1st carrier is modulated by the special point S1 to indicate the transition to "1" the 1st bit of the binary word encoded by the next carrier. The 6th carrier is also modulated by the special point S1 to indicate the transition to "0" of the 1st bit of the coded binary word on the next carrier. As regards the last 11 carriers, as can be seen, the first two carriers (numbered 35 and 36 in hexadecimal) of this group of carriers, are modulated both by the special point S1. The special point S1 corresponds to the appearance in the data stream of the marker M1. As can be seen in the table in the Appendix, the succession of two markers M1 (M1M1) is used to indicate an event (change to "1") on the first bit of the binary word which follows the sequence of the two markers, plus one change of event (change to "0" then change to "1") on this first bit for each of the following binary words.

The symbol 905 following the symbol 903 consists of three sub-symbols corresponding respectively to the first 12 carriers associated with a size 6 constellation, then to the next 46 carriers associated with a size constellation 5, and finally to the last 6 carriers associated with a constellation of size 8. The 2 special points (S1, S2) of the size 6 constellation are used, the only special point of the size constellation 5 is not used, the constellation of size 8 does not have points special. The special points S1 and S2 used simultaneously, respectively by the semester and the 4 "n9 carrier, correspond to the succession of markers M1 and M2 in the data stream to be modulated, with the meaning mentioned in the table in the appendix. to indicate an event on a first and a second bit, coded on special points, of the binary word which follows this sequence of markers Thus, in the binary word which will be coded on the 5th carrier, the first and second 4th bits that are coded on the special points, go from 'V' to 'T'. </ B>

The symbol 907 uses the carriers with the same constellations as the symbol 905 above. The first two carriers respectively use the special point S2 and the special point S1, corresponding to an M2M1 sequence of markers in the data stream to be transmitted. The meaning of the M2M1 sequence is given by the table in Appendix <I>: </ I> "Event <I> on bit 1 and 2 + Change of event on bit 1 </ I> <I> and 2" . </ I> This means that the two bits coded by special points of the concerned bit words start with changing state (event: change from "0" to "1") for the word coded on the 5th carrier, then at each subsequent word, the event concerning these bits changes, thus in the binary word coded on the 4th carrier, these two bits return to zero.

The following symbol 9e9 consists of a single OFDM sub-symbol because the 64 carriers use the same 7-point constellation. Only 2 special points (S1 and S3) among the 3 special points available for a 7-point constellation are used. The first two carriers are modulated by special point S3, which corresponds to the sequence of M3M3 markers in the data stream, with the meaning given in the table of the Annex: <I> "</ I> Event <I > on bit 3 + change of event on </ I> <I> bit 3 ". </ I> This means that the third bit (in the order of the weights, the first bit bit of the highest weight coded by a special point) of the binary words concerned begins to move to "1" (event: change to "1") then changes state for each of the words that follow (change to "0" then change to Other sequences of markers with different meanings are given in the table in the Annex.

With reference to FIGS. 5 and 6, the receiving block of the peripheral station shown in FIG. 2, in which the method of receiving a digital signal according to the invention is implemented, will now be described. The reception block 72 of the peripheral station of FIG. 2 provides for the reception of a coded digital signal on transmission, for example by the base station SB, according to a method of encoding a digital signal in accordance with FIG. the invention.

With reference to FIG. 5, the reception block 72 of the peripheral station SPi comprises an OFDM reception unit 80 and a data analysis and decoding unit 70.

The reception unit 80 receives as input the signal 800 coming from the radio module 40 (see Fig. 2 and description above). The signal 800 is an OFDM analog signal carrying data, called "useful" because they are the object of transmission, and control signals.

The OFDM reception unit 80 is a conventional circuit known from the state of the art and intended to provide the OFDM demodulation. Thus, the unit 80 demodulates the carriers used for the transmission, so as to obtain for each of them a constellation point, that is to say a complex symbol, whose real and imaginary parts correspond to the coordinates in the complex plane of the corresponding constellation point. Each of the constellation points obtained corresponds to one of the binary words coded on transmission, or to one of the markers described above, if the constellation point in question is a special point. There are therefore as many complex symbols received as carriers used for transmission (eg 64).

The OFDM demodulation circuit 80 outputs the signal 700 which is a serial signal consisting of complex symbol groups, each group corresponding to an OFDM symbol. The signal 700 is then input to the unit 70 for analyzing and decoding the data.

As mentioned above in the description with reference to FIG. 2, at the output of the unit 70 and therefore of the reception block 72, three signals are delivered: the two control signals 740 and 780 that have already been described in relation with FIG. the description of the transmission block 38, and the signal 790 containing the payload data that will be processed by the data processing unit 24.

Referring now to FIG. 6, unit 70 for analyzing and decoding the data of the reception block 72 will be described in greater detail. In accordance with the invention, the unit 70 decodes each of the reception points. constellation obtained according to the coding rules used during the transmission of the digital signal by a remote transmitter (for example the base station), to obtain the corresponding binary words of the digital signal.

The data analysis and decoding unit 70 consists of three units - a unit 74 for channel analysis and carrier modulation assignment; a decoding unit 76; a unit 78 for separating data. Each of the units 74 and 76 receives as input the signal 700 which, as mentioned above, is a serial signal consisting of groups of complex symbols (corresponding to the obtained constellation points).

The unit 74 for analyzing the channel and assigning modulation to the carriers on the one hand has the function of "analyzing the transmission channel", that is to say of generating from the signal 700 data indicative of the transmission quality of the channel for the reception of a signal by the peripheral station. On the other hand, the unit 74 has the task of analyzing the signal 700 so as to determine the constellation assigned to each of the carriers used by the remote transmitter (the base station) for transmission. The data concerning the quality of the channel and those containing the information relating to the constellations used by the remote transmitter, thus making it possible to deduce the coding rules used by it, are combined in the signal 740 delivered by the unit 74.

The signal 740 is then supplied on the one hand as input to the buffer unit 64 in the unit 60 of the transmission block 38 to extract data indicative of the quality of the channel and insert them in a signal (600) to issue (see description of unit 60 supra); on the other hand, the signal 740 is input to the decoding unit 76 so that it uses the constellation information used to derive the decoding rules therefrom.

The decoding unit 76 will now be described. The unit 76 receives as input the signal 700 and the signal 740 coming from the unit 74. The function of the unit 76 is to extract from the signal 700 the information making it possible to deduce the bits to be inserted into the signal, c that is, those suppressed at issue by the use of markers. This information is obtained by decoding the complex symbols contained in the signal 700, that is to say by decoding the constellation points used for the modulation of the carriers. For this purpose, the rules concerning the decoding are deduced from the information relating to the constellations used by the remote transmitter and provided by the signal 740.

Thus, at each of the special points used, the unit 76 matches a predefined marker. The unit 76 then inserts in the signal 700 the bits deduced from the interpretation of the markers corresponding to the special points used for the modulation. The modified signal 700 constitutes the output signal 760. , <U> Note </ U>: the signal 760 obtained is, with transmission errors, identical to the signal 640 from the buffer unit 64 in the unit 60 for inserting markers and control data, the transmission block 38 of the remote transmitter (for example the base station SB) which is at the origin of the digital signal (800) being received.

Thus, the unit 76 makes it possible, for each of the carriers used for the transmission, to deduce from the decoding of a special point or the decoding of several consecutive special points, the value of the bits represented differentially by this or these special points, according to the value of these bits (bits) in one or more binary words previously received from the same carrier.

The unit 76 delivers the signal 760 at the input of the data separation unit 78.

The function of the unit 78 is to separate in the signal 760 the useful data from the control data sent by the remote transmitter, these control data being indicative of the quality of transmission in the channel in the peripheral station direction (currently "receiver" ") to base station (currently" remote transmitter ").

The useful data is grouped in the signal 790 output by the unit 78, while the control data is grouped in the output signal 780.

Finally, the signal 790 containing the useful data is transmitted to the data processing unit 24 for appropriate processing. The signal 780 is itself delivered to the transmission block 38 to be used as described above in connection with the description of the process of transmitting a signal. In particular, the signal 780 is input to the analysis unit 66 and to the cartographic transfer unit 42, to the transmission block 38.

In summary, concerning the signal 780, the data separation unit 78 extracts the control data inserted by the remote transmitter (base station) during the process from the reconstituted digital signal (signal 760). transmission of the digital signal being received by the peripheral station. These control data are indicative of the quality of the transmission channel concerning the reception by the base station of a signal previously transmitted by the peripheral station. As previously explained in connection with the description of the transmission block 38, the data of the signal 780 will be used jointly by the analysis unit 66 and the cartographic transfer unit 42 to define the coding rules of a digital signal intended to be issued.

Of course, many modifications can be made to the embodiments described above without departing from the scope of the invention.

<B> ANNEX </ B>
<tb> Number <SEP> of <SEP> markers <SEP><SEP> designation <SEP> Marker <SEP> Bit <SEP> encoded
<tb><B> 1 <SEP> mi </ B>
<tb> 2 <SEP> M1 <SEP> - <SEP> 1
<tb> M2 <SEP> 2__
<tb> 3 <SEP> M1 <SEP> 1
<tb> M2 <SEP> 2
<tb> M3 <SEP> 3

Number <SEP> of <SEP> Sequence <SEP> of <SEP> Meaning
<tb> markers <SEP> markers <SEP> encoding
<tb><SEP> data
<tb> 1 <SEP> M1 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 1
<tb> M1 <SEP> M1 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 1 <SEP> + <SEP> change <SEP> of event <SEP> on <SEP> on <SEP> bit <SEP> 1
<tb> 2 <SEP> M1 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 1
<tb> M1 <SEP> M1 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 1 <SEP> + <SEP> change <SEP> of event <SEP> on <SEP> on <SEP> bit <SEP> 1
<tb> M1 <SEP> M2 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 1 <SEP> and <SEP> 2
<tb> M2 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 2
<tb> M2M2 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 2 <SEP> + <SEP> change <SEP> of event <SEP> on <SEP><SEP> bit <SEP> 2
<tb> M2M1 <SEP> Change <SEP> of event <SEP> on <SEP><SEP> bit <SEP> 1 <SEP> and <SEP> 2
<tb> 3 <SEP> M1 <SEP><SEP> event on <SEP><SEP> bit <SEP> 1
<tb> M1 <SEP> M1 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 1 <SEP> + <SEP> change <SEP> of event <SEP> on <SEP> on <SEP> bit <SEP> 1
<tb> M1 <SEP> M2 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 1 <SEP> and <SEP> 2
<tb> M1 <SEP> M3 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 1 <SEP> and <SEP> 3
<tb> M2 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 2
<tb> M2Ml <SEP> Event <SEP> on <SEP><SEP> Bit <SEP> 1 <SEP> and <SEP> 2 <SEP> + <SEP> Change <SEP> Event
<tb> on <SEP><SEP> bit <SEP> 1 <SEP> and <SEP> 2
<tb> M2M2 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 2 <SEP> + <SEP> change <SEP> of event <SEP> on <SEP><SEP> bit <SEP> 2
<tb> M2M3 <SEP> Event <SEP> on <SEP> the <SEP> bit <SEP> 2 <SEP> and <SEP> 3
<tb> M3 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 3
<tb> M3Ml <SEP> Event <SEP> on <SEP><SEP> Bit <SEP> 1 <SEP> and <SEP> 3 <SEP> + <SEP> Change <SEP> Event
<tb> on <SEP><SEP> bit <SEP> 1 <SEP> and <SEP> 3
<tb> M3M2 <SEP> Event <SEP> on <SEP><SEP> Bit <SEP> 2 <SEP> and <SEP> 3 <SEP> + <SEP> Change <SEP> Event
<tb> on <SEP><SEP> bit <SEP> 2 <SEP> and <SEP> 3
<tb> M3M3 <SEP> Event <SEP> on <SEP><SEP> bit <SEP> 3 <SEP> + <SEP> change <SEP> of event <SEP> on <SEP><SEP> bit <SEP> 3
<tb> Ml <SEP> M2M3 <SEP><SEP> event on <SEP><SEP> bit <SEP> 1, <SEP> 2 <SEP>, and <SEP> 3
<tb> MiM1MI <SEP> Change <SEP> of <SEP> bit <SEP> encoded <SEP> by <SEP><SEP> tag <SEP> 1 <SEP> to <SEP> bit <SEP> no <SEP > coded <SEP> b1
<tb> M1 <SEP> M1 <SEP> M2 <SEP> Change <SEP> of <SEP> bit <SEP> encoded <SEP> by <SEP> the <SEP> marker <SEP> 1 <SEP> to <SEP > bit <SEP> no <SEP> encoded <SEP> b2
<tb> M2M2M2 <SEP> Change <SEP> of <SEP> Bit <SEP> Encoded <SEP> by <SEP><SEP> Marker <SEP> 1 <SEP> to <SEP> Bit <SEP> No <SEP > coded <SEP> b2
<tb> M2M2MI <SEP> Change <SEP> of <SEP> Bit <SEP> Encoded <SEP> by <SEP><SEP> Marker <SEP> 1 <SEP> to <SEP> Bit <SEP> No <SEP > coded <SEP> b1
<tb> M3M3M1 <SEP> change <SEP> of <SEP> bit <SEP> encoded <SEP> by <SEP><SEP> tag <SEP> 1 <SEP> to <SEP> bit <SEP> no <SEP > coded <SEP> b1
<tb> M3M3M2 <SEP> change <SEP> of <SEP> bit <SEP> encoded <SEP> by <SEP><SEP> tag <SEP> 1 <SEP> to <SEP> bit <SEP> no <SEP > coded <SEP> b2

Claims (33)

<B> <U> CLAIMS </ U> </ B> A method of encoding a digital signal to be transmitted and composed of consecutive bits, by mapping over a plurality of points of a predetermined constellation having an integer number C of points different from a power of two, characterized in that it comprises the following steps - grouping (E113, E115, E117) the bits of said digital signal (600) into successive binary words of length L obtained by the following equation: <I> L = </ I > N + C-2N where N is the integer such that <I> 2N </ I> <C <2N + '<I>; </ I> - for each of said binary words, perform (E119) a map report of N binary bits of the binary word on a point among 2 "points of said constellation, said point among 2" points representing the N bits absolutely, the value of the C-2 'other bits of the binary word being deductible from the use or previous non-use of C-2N a other points of said constellation, so-called special points, each of which differentially representing a binary element among the C-2N other bits of the binary word. 2. A method of encoding a digital signal according to claim 1, characterized in that each of said C-2N bits is represented by the corresponding special constellation point, depending on the occurrence of a predefined event relating to the value of the considered bit in a plurality of consecutive binary words of said digital signal. 3. Method for encoding a digital signal according to claim 2, characterized in that said predefined event corresponds to a change in value of the binary element considered among said C-2N binary elements, with respect to its value in the word. previous binary. 4. A method of encoding a digital signal according to claim 2 or 3, characterized in that said predefined event corresponds to a value conservation, of the binary element considered among said C-2N binary elements, with respect to its value. in the previous binary word. 5. A method of coding a digital signal according to any one of the preceding claims, characterized in that said C-2Binary elements are removed from said digital signal and replaced by markers intended to indicate at least one event relating to the value of said C-2N bits in the digital signal, said markers being encoded through said special points. 6. Encoding method according to claim 5, characterized in that a predetermined sequence of said markers in the digital signal induces a change in coding rules. 7. A method of coding a digital signal according to any one of the preceding claims, characterized in that said digital signal is transmitted by quadrature amplitude modulation of a single carrier, each of the binary words of said digital signal modulating consecutively. said carrier according to said predetermined constellation. The method of coding a digital signal according to any one of claims 1 to 6, characterized in that said digital signal is transmitted by quadrature amplitude modulation of a plurality of carriers, a plurality of binary words of the digital signal simultaneously modulating at least a subset of carriers of said plurality of carriers, each according to said predetermined constellation. 9. A method of coding a digital signal according to claim 8, characterized in that said digital signal is transmitted according to an orthogonal frequency division multiplexing technique. A method of encoding a digital signal according to any one of the preceding claims, wherein each of the points of said constellation corresponds to an amplitude and a modulation phase of a carrier by said digital signal, characterized in that one of the points of the constellation is the point at the origin of the complex plane. 11. A method for transmitting a digital signal by quadrature amplitude modulation of at least one carrier whose modulation is controlled by cartographic report according to a constellation of points, characterized in that said digital signal is coded according to a method. encoding a digital signal in accordance with any one of claims 1 to 10. 12. A method for transmitting a digital signal by quadrature amplitude modulation of a plurality of carriers whose modulation is controlled by cartographic report according to a constellation of points, at least one carrier being modulated according to a constellation whose number C points is different from a power of two, characterized in that it comprises the following steps - group (E113, E115, E117) the bits of said digital signal into successive binary words, each of which being intended to be transported by a carrier, the length L of each of said binary words being obtained, when the number C of points of the constellation for the carrier considered is different from a power of two, by the following equation <I> L = </ I> N + C-2N where N is the integer such that: 2 '<I> <C <<I> 2N +' <I>; </ I> and, for each of said binary words for which the carrier is modulated according to a constellation having a no a number of points different from a power of two - performing (E119) a map report of N binary bits of the binary word on a point among 2 "points of said constellation, said point among 2" points representing the N binary elements so absolute, the value of the other C-2N bits of the binary word being deductible from the prior use or non-use of C-2N other points of said constellation, so-called special points, each of which differentially representing a binary element among the C-2N other bits of the binary word. 13. Method for transmitting a digital signal according to claim 12, characterized in that each of said C-2Binary elements is represented by the corresponding special constellation point, as a function of the occurrence of at least one predefined relative event. to the value of the considered bit in a plurality of consecutive binary words of said digital signal. 14. Method for transmitting a digital signal according to claim 13, characterized in that said at least one predefined event corresponds to a change in value of the binary element considered among said C-2 "binary elements, relative to to its value in the previous binary word. 15. A method of transmitting a digital signal according to claim 13 or 14, characterized in that said at least one predefined event corresponds to a value conservation of the binary element considered among said C-2 "binary elements, by relative to its value in the previous binary word. 16. A method of transmitting a digital signal according to any one of claims 12 to 15, characterized in that said C-2 "bits are removed from said digital signal and replaced by markers intended to indicate at least one event relating to the value of said C-2 "bits in said digital signal, said markers being encoded through said special points. 17. A method of transmitting a digital signal according to claim 16, characterized in that a predetermined sequence of said markers in the digital signal induces a change in coding rules. 18. A method of transmitting a digital signal according to any one of claims 12 to 17, characterized in that said digital signal is transmitted according to an orthogonal frequency division multiplexing technique. A method of transmitting a digital signal according to any one of claims 12 to 18, wherein each of the points of said constellation corresponds to an amplitude and a modulation phase of a carrier by said digital signal, characterized in that one of the points of the constellation is the point at the origin of the complex plane. 20. A method of transmitting a digital signal according to any one of claims 12 to 19, characterized in that the number of points of the constellation associated with each of said carriers is calculated according to information representative of the quality. transmission channel in which said digital signal is transmitted. 21. A method of transmitting a digital signal according to claim 20, characterized in that said information representative of the transmission quality of the transmission channel is received prior to the transmission of said digital signal, following a previous transmission of a digital signal. 22. A method of receiving a digital signal emitted by a method of transmitting a digital signal according to any one of claims 12 to 21, characterized in that it comprises the following steps - demodulating said carriers in a manner to obtain for each of them a point of constellation; decoding each of the constellation points obtained according to the coding rules used during the transmission of said digital signal, so as to obtain the corresponding binary words of said digital signal. 23. A method of receiving a digital signal according to claim 22, characterized in that the step of decoding each of the constellation points comprises the following step - for each of said carriers, to deduce from the decoding of a special point or decoding of several consecutive special points, the value of the bits represented differentially by this or these special points, as a function of the value of these bits in one or more binary words previously received from the same carrier. 24. The method of receiving a digital signal according to claim 23, characterized in that it further comprises the following step: extraction of said reconstituted digital signal of control data indicative of the transmission quality of the transmission channel, said data being intended to allow the modification of said coding rules as a function of the quality of transmission in the channel. 25. Device for coding a digital signal to be transmitted and composed of consecutive bits, by cartographic report on a plurality of points Of a predetermined constellation having an integer number C of points different from a power of two, characterized in that it comprises - means (60) for grouping the bits of said digital signal into successive binary words of length L obtained by the following equation: <I> L = </ I> N + C-2N where N is the integer such that: 2 "<C <2" + '; cartographic reporting means (42), used for each of said binary words, for carrying out a cartographic report of N binary bits of the binary word on a point among 2 "points of said constellation, said point among 2" points representing the N elements in absolute terms, the value of the C-2 "other bits of the binary word being deductible from the prior use or non-use of C-2" other points of said constellation, so-called special points, each of which represents Differentially, a binary element among the C-2 "other bits of the binary word. 26. Device for coding a digital signal according to claim 25, characterized in that it comprises means adapted to implement a method of encoding a digital signal according to any one of claims 2 to 10. 27. A device for transmitting a digital signal by quadrature amplitude modulation of a plurality of carriers whose modulation is controlled by cartographic transfer according to a constellation of points, at least one carrier being modulated according to a constellation whose number C points is different from a power of two, characterized in that it comprises - means (60) for grouping the bits of said digital signal into successive binary words, each of which being intended to be transported by a carrier, the length L of each of said binary words being obtained, when the number C of points of the constellation for the carrier considered is different from a power of two, by the following equation <I> L = </ I> N + C- 2N where N is the integer such that: 2 '<C <2' + '; cartographic reporting means (42) used for each of said binary words for which the carrier is modulated according to a constellation having a number C of points different from a power of two, to carry out a cartographic report of N binary elements of the word binary on a point among 2 "points of said constellation, said point among 2" points representing the N binary elements in an absolute way, and carry out a cartographic report of the C-2 "other bits of the binary word on respectively C-2" points, so-called special points, of said constellation, said special points representing the C-2 "other bits differentially. 28. Device for transmitting a digital signal according to claim 27, characterized in that it comprises means adapted to implement a method of transmitting a digital signal according to any one of claims 13 to 21. 29. Device for receiving a digital signal transmitted by a device for transmitting a digital signal according to claim 27 or 28, characterized in that it comprises - demodulation means (80) for demodulating said carriers of in order to obtain for each of them a point of constellation; decoding means (76) for decoding each of the constellation points obtained, according to the coding rules used during the transmission of said digital signal, so as to obtain the corresponding binary words of said digital signal. Digital signal receiving device according to claim 29, characterized in that said decoding means comprise means (76) for deriving from the decoding of a special point or the decoding of several consecutive special points, the value bits represented differentially by this or these special points, depending on the value of these bits in one or more previously received binary words of the same carrier. 31. Device for receiving a digital signal according to claim 30, characterized in that it further comprises: extraction means (78) of said reconstituted digital signal of control data indicative of the transmission quality of the transmission channel. transmission, said data being intended to allow the modification of said coding rules as a function of the quality of transmission in the channel. Telecommunications system comprising at least one digital signal transmission device according to claim 27 or 28, and at least one digital signal receiving device according to claim 29, 30 or 31. 33. Telecommunications network comprising at least two telecommunications systems according to claim 32.