Methods, devices and programs for demodulating a data frame

Abstract

The invention relates to a method for demodulating a data frame comprising N information words [P0.. PN-1], in which a word p i occupying a position i ϵ [0, N — 1] in the frame is modulated by a symbol C pi from a dictionary of K symbols, wherein the method comprises the steps of determining (301), for a plurality of symbols C Pi and C Pj of the frame, where i,j ϵ [0, N — 1], a vector of K probabilities P i / j , referred to as the relative distribution, comprising probabilities associated with K possible values of an estimated difference between the symbols C Pi and C Pj }; combining (303) a relative distribution Pi / j with a particular probability distribution mask i / j determined (302) according to a characteristic of the channel; and demodulating (304) the frame on the basis of at least one portion of the combined distributions.

Description

[0001] DESCRIPTION Title: Methods, devices and programs for demodulating a data frame Technical field The invention generally belongs to the field of telecommunications and more particularly relates to a modulation technique for improving the demodulation performance of a data frame, in particular in communication systems which rely on a relative representation of the information for the purpose of its transmission and / or its processing. Prior art Communication systems are known which rely on a relative representation of the information for the purpose of its transmission and / or its processing. A common example is the use of differential modulations. Consider for example a frame of 5 symbols [^^0, ^^1, ^^2, ^^3, ^^4], each encoding ^^ bits of information, and thus being able to take 2 ^^values. A possible differential representation of the frame can be [^^0⁄ 1 , ^^1⁄ 2 , ^^2⁄ 3 , ^^3⁄ 4 ] = [^^0 − − ^^2, ^^2 − ^^3, ^^4 − ^^3]. In this example, it is therefore the difference between the values ​​that is coded, and a single absolute reference is sufficient to decode the complete frame. Of course, in such a configuration, the subtraction, or any other operation, is done in the finite field considered, here F ^^(i.e. modulo ^^). Such a representation can be used for transmission or signal processing purposes within the communication chain. For example, when variations between consecutive samples are small, transmitting differences between samples can be particularly effective. Such a differential representation of the frame can also be useful in certain receivers, for example when receiving a signal modulated by the cyclic rotation of a root sequence of complex symbols, such as CCSK modulation (for Cyclic Code-Shift Keying in English). CCSK modulation proposes to modulate the data to be transmitted by the cyclic rotation / shift of a sequence of complex symbols called the root sequence. The root sequence is such that its shifted versions are orthogonal to each other, that is to say, it offers a good autocorrelation function.Thus, each binary word to be transmitted is associated with a particular cyclic shift of the same root sequence. For example, by choosing a root sequence composed of the 4 complex symbols [a; b; c; d], CCSK modulation allows 2 bits of information to be modulated orthogonally via the rotations of this sequence. The binary word '00' could correspond to the transmission of the sequence [a; b; c; d]; the binary word '01' could correspond to the transmission of the sequence [d; a; b; c]; the binary word '10' could correspond to the transmission of the sequence [c; d; a; b], etc. CCSK sequences can advantageously be transmitted in a CP-OFDM frame (for Cylic Prefix Orthogonal Frequency Division Multiplex in English), for example in a time-frequency frame in which each of the N time steps contains a complete CCSK sequence of size K distributed over all the subcarriers in frequency as shown in Figure 1.The receiver's knowledge of the absolute offset of a particular sequence in the frame allows the demodulation of the entire frame. For example, when the offset of the first sequence is known, for example when a particular sequence is present at the beginning of the frame by convention, the receiver can demodulate the frame by estimating the offsets of the other sequences relative to the known sequence. It is therefore necessary to have a good knowledge of the differences between the sequences in the same frame to correctly demodulate the data. However, it sometimes happens that disturbances do not affect the symbols transmitted in the same frame in the same way, which can cause problems during its demodulation. Indeed, in certain circumstances, the communication system can undergo variations in the communication channel over time.This is for example the case in the presence of Doppler effect due to a relative displacement between the receiver and the transmitter. This evolution of the channel is all the faster as the relative transmitter-receiver speed is high. A typical example is the case of satellite communications, in which the relative speeds considered can be particularly high. There is therefore a need for a method which makes it possible to improve the demodulation performances when the channel undergoes disturbances affecting differently the symbols of the same frame, for example when these disturbances vary over time, in particular when the communication is based on a relative representation of the information. Summary of the invention To this end, a method is proposed for demodulating a data frame comprising N information words [^^0 .. ^^^^−1] in which a word ^^^^ occupying a position ^^ ∈ [0, ^^ − 1] in the frame is modulated by a symbol ^^.^^^^ of a dictionary of ^^ symbols, the method comprising the following steps: -Determination, for a plurality of symbols ^^^^^^ and ^^^^^^ of the frame, with ^^, ^^ ∈[0, ^^ − 1], of a vector of K probabilities ^^^^ / ^^, called relative distribution, comprising probabilities respectively associated with ^^ possible values ​​of an estimated difference between the symbols ^^ ^^^^ and ^^ ^^^^ , - Combination of a relative distribution ^^ ^^ / ^^ with a particular probability distribution mask ^^⁄ ^^determined according to a characteristic of the communication channel, - Demodulation of the frame from at least part of the combined distributions. The operation of difference between symbols discussed in this document depends on the modulation technique used. In the case of CCSK modulation, such a difference corresponds for example to the value of a cyclic shift of the sequence considered with respect to another sequence. In the case of QPSK modulation, the difference operator is adapted to estimate phase or angle shifts. The difference between symbols can also correspond to a difference between index values ​​associated with the symbols in a symbol dictionary, following an indexing system equal to the set {0, 1, …, K-1} up to an isomorphism, and their respective probabilities.The difference is thus estimated in a finite group of size K, that is to say modulo the cardinality of the dictionary of symbols when we assume an indexing in the form {0, 1, …, K-1}. We recall here that a group, in the mathematical sense of the term, is a set equipped with a unique internal composition law which is associative, possessing a neutral element. A finite group is a group whose number of elements is finite. A relative distribution. ^^ represents a probability of the relative value of the symbol ^^ ^^^^ compared to the symbol ^^ ^^^^ , knowing the channel estimated using the symbol ^^ ^^^^ (and vice versa). Thus, when the channel varies over time, for example in the presence of Doppler, and the distributions ^^ ^^⁄ ^^ are affected by the correlation between the channels undergone by ^^ ^^^^ and ^^ ^^^^, then for values ​​of ^^ and ^^ far apart in the OFDM frame, these distributions are not necessarily reliable. According to another example, when considering time-constant channels and several spatial layers, for example in a MIMO (Multiple Input Multiple Output) system using spatial multiplexing, the level of channel correlation between different layers can affect the estimation of a difference between symbols. For example, in a system using 4 spatial layers, we can have layer 1 and 4 slightly correlated and layer 1 and 2 and 1 and 3 completely decorrelated. The determination and application of a suitable mask makes it possible to limit the influence of the ^^ distributions ^^ / ^^unreliable. The method thus proposes to combine probability distributions of a difference between symbols of the frame with masking distributions which are determined according to a characteristic of the channel on which a level of confidence in the estimation of the symbols or the differences of the symbols depends, so as to adjust the effect of certain probability distributions on the demodulation process. The relative distributions thus filtered, representative of probabilities of difference between symbols, are used to demodulate the information words of the frame. For example, when a symbol of the frame is known to the receiver, like a pilot symbol, the knowledge of the differences between the symbols allows the demodulation of the complete frame. The method makes it possible to give a variable importance to the distributions representative of these differences according to their reliability. The demodulation performances are thus improved.According to a particular embodiment, the demodulation of the frame is carried out from at least one partial relative distribution ^^ (^^) ^. ^ / ^^ determined by relating a first combined relative distribution determined from symbols ^^ ^^^^ and ^^ ^^^^ with a second combined relative distribution determined from a symbol ^^ ^^^^ and the symbol ^^ ^^^^ , with^^, ^^ ≠ ^^, and ^^ ∈ [0; ^^ − 1].It is thus proposed to determine relative distributions called partial ^^ (^^) ^ ^ / ^^ from masked relative distributions determined for symbols ^^ ^^^^ and ^^ ^^^^ on the one hand, and on the other hand from combined relative distributions determined for the symbols ^^ ^^^^ and ^^ ^^^^. In other words, we determine a difference between a first symbol and a second symbol by transitivity, that is to say from a difference between the first symbol and a third symbol and from an index difference between the second symbol and the same third symbol. We thus obtain a plurality of estimates of the index difference between the first and the second symbol based on different symbols. The partial relative distributions ^^ (^^) ^ ^ / ^^ determined for a particular ^^ value are then combined by product to obtain a consolidated version of a relative ^^ distribution ^^⁄ ^^ in which distributions deemed unreliable have little influence. In a particular realization, a particular probability distribution mask ^^⁄ ^^ is determined according to the reliability level determined for the symbols ^^ ^^^^ and ^^ ^^^^and / or a determined reliability level for the difference between these symbols. The masking distributions are for example determined according to the reliability of the estimation of the symbols and / or the reliability in the estimation of a difference between symbols. In this way, it is possible to determine a particular masking distribution which is configured to limit the influence of unreliable estimations and / or to give more importance in a combined distribution to the most reliable symbols and / or differences between symbols. According to a particular embodiment, a relative distribution ^^ ^^ / ^^ is convolved with a Kronecker distribution with a peak at 0 (denoted Kronecker[0]) when the number of symbols that separate symbols ^^ ^^^^ and ^^ ^^^^in the frame is less than a threshold, and with a uniform distribution when this number is greater than or equal to said threshold, the threshold being determined according to a characteristic of the communication channel. In this way, the differences between symbols which are distant in the frame are not taken into account for the demodulation process. In the presence of a Doppler effect for example, the symbols at the start of the frame and at the end of the frame do not experience the same channel because the latter varies over time. In this way, the method avoids introducing these unreliable estimates into the demodulation process. The threshold from which the relative distributions are suppressed is defined according to a characteristic of the communication channel. According to a particular embodiment, a relative distribution ^^ ^^ / ^^ is convolved with a probability distribution mask ^^⁄ ^^ intermediate between a Kronecker distribution [ 0 ]and a uniform distribution ^^ such that mask^^⁄ ^^ = ^^^^ / ^^^^[0] + being determined according to a characteristic of the communication channel. Such an arrangement makes it possible to finely adjust the level of consideration of a distribution ^^ ^^ / ^^ particular during the demodulation step, depending on the degree of confidence that we wish to grant it. According to a particular embodiment, a probability distribution mask ^^⁄ ^^ is determined by the number of symbols separating the symbols ^^ ^^^^ and ^^ ^^^^in the frame and a level of temporal variability of the channel. When the channel is subject to temporal variations, for example in the case of a Doppler effect, the symbols at the beginning and end of the frame are not affected in the same way, so that an estimated difference between these symbols may prove to be unreliable. By determining a mask which takes into account both the distance separating pairs of symbols in the frame and a level of temporal variability of the channel, for example a Doppler level, the least relevant probability distributions can be discarded and thus improve the demodulation performance. According to a particular embodiment, the method is such that a symbol ^^ ^^^^ is a sequence of complex values ​​obtained by applying a particular cyclic shift to a root sequence of size ^^, a sequence ^^ ^^^^being received on ^^ subcarriers of an OFDM symbol. The information to be transmitted is modulated by a cyclic shift of a root sequence known to the receiver. This is for example a CCSK type modulation using a Zadoff-Chu type root sequence. The CCSK sequences (or symbols) are integrated into a CP-OFDM frame so that each element composing a CCSK sequence, called a chip, is transmitted on a separate OFDM subcarrier. In its simplest form, we then obtain a time-frequency frame in which each of the N time steps contains a complete CCSK symbol of size K distributed over all the subcarriers in frequency, as shown in Figure 1. The root sequence (and therefore all its cyclic shifts) being known to the receiver, the transmitted CCSK sequences can then serve as pilot sequences for synchronization and channel estimation, while carrying information.The calculated differences can thus correspond to differences between indexes associated with the symbols or differences between the offsets of the sequences composing the symbols. According to another aspect, the invention relates to a device for demodulating a data frame comprising a sequence of information words [^^0 .. ^^^^−1] in which a word ^^^^ occupying a position ^^ ∈ [0, ^^ − 1] in the frame is modulated by a symbol ^^^^^^ from a dictionary of ^^ symbols, the device comprising a processor coupled to a memory in which program instructions are recorded configured to implement the following steps: - Determination, for a plurality of symbols ^^^^^^ and ^^^^^ of the frame, with ^^, ^^ ∈[0, ^^ − 1], of a vector of ^^ probabilities ^^^^ / ^^, called relative distribution, comprising probabilities respectively associated with K possible values ​​of an estimated difference between the symbols ^^. ^^^^ and ^^ ^^^^, - Combination of a relative distribution ^^ ^^ / ^^ with a particular probability distribution mask ^^⁄ ^^determined according to a characteristic of the communication channel - Demodulation of the frame from at least part of the combined distributions. The invention also relates to a communication terminal comprising a demodulation device as described above. In a particular embodiment, the different steps of the demodulation method are determined by computer program instructions. Consequently, the invention also relates to a computer program comprising instructions adapted to the implementation of the steps of a demodulation method as described above, when the program is executed by a processor. This program can use any programming language, and be in the form of source code, object code, or intermediate code between source code and object code, such as in a partially compiled form,or in any other desirable form. The invention also relates to a computer-readable information medium on which is recorded a computer program comprising instructions for executing the steps of a demodulation method as described above. The information medium may be any entity or device capable of storing the program. For example, the medium may comprise a storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, a flash memory, or a magnetic recording means, such as a hard disk. Furthermore, the information medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means. The program according to the invention may in particular be downloaded from a network such as the Internet. Alternatively,the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the method in question. The various embodiments or characteristics mentioned above may be added independently or in combination with each other, to the steps of the demodulation method. The devices, terminals, programs and information carriers have advantages similar to those conferred by the method to which they correspond. Brief description of the figures Other characteristics and advantages will appear on reading a preferred embodiment described with reference to the appended drawings among which: - Figure 1 represents an environment adapted to implement the demodulation method according to a particular embodiment,- Figure 2 illustrates a time-frequency frame in which each of the N time steps contains a complete symbol of size K distributed over all the subcarriers in frequency, - Figure 3 is a flowchart representing the main steps of a demodulation method according to a particular embodiment, - Figure 4a is a graph showing an example of relative distribution, - Figure 4b represents a graph showing a Kronecker distribution[0], - Figure 4c is a graph illustrating a relative distribution resulting from the convolution of the distributions represented in Figures 4a and 4b, according to a particular embodiment, - Figure 5a is a graph showing an example of relative distribution, - Figure 5b represents a graph showing a uniform distribution, - Figure 5c is a graph illustrating a relative distribution resulting from the convolution of the distributions represented in Figures 5a and 5b,according to a particular embodiment, - Figure 6a is a graph showing an example of a relative distribution, - Figure 6b represents a graph showing a masking distribution intermediate between a Kronecker[0] distribution and a uniform distribution, - Figure 6c is a graph illustrating a relative distribution resulting from the convolution of the distributions represented in Figures 6a and 6b, according to a particular embodiment, and - Figure 7 is a block diagram representing the architecture of a device suitable for implementing the demodulation method in a particular embodiment. In the following description, embodiments are described on the basis of non-limiting examples making it possible to explain the concepts on which the invention is based. In particular, although the examples and the terminology used may refer to certain well-known technologies or standards,These references are not limiting and other technologies may be adapted to implement the concepts of the invention. For example, the CCSK modulation technique referred to may be replaced by other modulation techniques, such as QPSK (Quadrature Phase Shift Keying), or QAM (Quadrature Amplitude Modulation), without it being necessary to modify the invention. Figure 1 represents an environment suitable for implementing the demodulation method according to a particular embodiment. The environment comprises a communication system 100, comprising a transmission device 101 and a reception device 102. The transmission device 101 is for example a communication terminal, a user equipment (UE), a base station, a connected object,a telecommunications satellite, etc. The receiving device 101, for example user equipment, a terminal, a vehicle or a connected object, or any other device suitable for receiving wireless data. The devices 101 and 102 comprise radiofrequency communication means, for example a transceiver suitable for enabling the devices 101 and 102 to communicate by exchanging time-frequency data frames, for example CP-OFDM frames. In a particular embodiment, such frames comprise a plurality of information words, for example binary words of size ^^ bits, which are modulated by a CCSK symbol, i.e. by a particular sequence of complex values ​​obtained by cyclic shift of a root sequence of size 2, ^^which has a good autocorrelation function. Such a sequence, for example a Zadof-Chu sequence. Thus, the communication system 100 implements a CCSK-CP-OFDM modulation. The CCSK-CP-OFDM approach combines the CP-OFDM technique with CCSK modulation with the aim of achieving small packet communications for the IoT, with high energy efficiency. It can however be noted that other modulation techniques can be envisaged without modifying the invention, as indicated above. Figure 2 represents an example of a CCSK-CP-OFDM frame in which each of the N time steps contains a complete CCSK sequence of size ^^ = 2^^ distributed over all the^^ subcarriers in frequency, so that each symbol can code a word of size ^^ bits. The set of ^^ CCSK sequences constitutes a dictionary of symbols of cardinality ^^.In a particular embodiment, each symbol of the dictionary is associated with a distinct index so that the ^^ symbols of the dictionary are respectively associated with ^^ index values. In the case of a CCSK modulation, each CCSK sequence of the dictionary can thus be associated with a unique numerical value representing the offset of the sequence with respect to the root sequence. According to a particular embodiment, the ^^ index values ​​with which the symbols of the dictionary are associated constitute a finite group ^^ of size ^^, provided with an addition law admitting 0 as a neutral element, so that each symbol of the dictionary is associated with a distinct value belonging to the finite group ^^.Operations on indexes, in particular sums and differences, are thus performed modulo the cardinality ^^ of the symbol dictionary when assuming an indexing in the form {0, 1, …, K-1}, so that any difference (or sum) modulo ^^ between elements of the set is also a value belonging to the set. In a particular realization, index values ​​are defined by the group ^^ℤ / ^^^^ℤ, considering modular arithmetic on the remainders of division by ^^^^ and ^^ a strictly positive integer, ℤ being the set of relative integers. For example, with ^^ = 2 and ^^ = 4, we obtain a set of indexes [0, 2, 4, 6] such that any difference (or sum) modulo ^^^^ between elements of the set is also a value belonging to the set.Thus, in this example, we actually observe that 0 – 2 = -2 [8] = 6, 0 - 4 = -4 [8] = 4, 2 – 4 = -2 [8] = 6, etc… A particular implementation of the demodulation method will now be described with reference to Figure 3. The method is for example implemented by the device 102 of Figure 1. During a first step 300, the device 102 receives a CCSK-CP-OFDM data frame comprising N information words [^^0 .. ^^^^−1] in which a word ^^^^ occupying a position^^ ∈ [0, ^^ − 1] in the frame is modulated by a CCSK sequence ^^^^^^ from a dictionary of ^^sequences. Thus, for each symbol C. Pi at a position i in the frame, and for each symbol Ck of the dictionary (C1, C2, ... , CK), we determine a probability vector P(CPi = Ck), in which P(CPi = Ck) is the probability that the symbol at position i is the symbol C k. Such a distribution is for example determined in a conventional manner by cross-correlation operations of the symbol at position i with the different symbols of the dictionary. In step 301, the device 102 determines so-called relative symbol distributions from the symbol distributions determined in step 300. Each symbol ^^ ^^^^ of the frame being represented by a symbol probability distribution which can be assimilated to a random variable Xi, the difference between two symbols ^^ ^^^^ and ^^ ^^^^ of the frame is, in the sense of the present invention, a differential random variable Y^^,^^ = ^^^^ − ^^^^ ∈ F^^ whose distribution, called relative distribution, is noted^^ ^^⁄ ^^ . For example, consider a frame of symbols as a sequence of discrete random variables ^^ = [^^1 ^^2 … ^^^^−1] ∈ Ϝ ^ ^ ^ ^ of distribution law ^^ ^^=[^^^^1 … ^^^^^^−1]. It is possible to express the matrix containing the set of differential variables Y ∈ Ϝ ^^×^^^^ such that Y^^,^^ = ^^^^ − ^^^^ ∈ F^^ and the matrix of corresponding distributions ^^Y such that ^^Y^^,^^ = ^^^^^^ ⊛ ^^^^^^ where ⊛ is the operator of cross-correlation. Given the similarity of sum and difference in the context of modular arithmetic in a finite field, we can also define the matrix containing the set of random variables Y ∈ Ϝ ^^×^^^^ such that Y^^,^^ = ^^^^ + ^^^^ ∈ F^^ and the matrix of corresponding distributions ^^Y such that ^^Y^^,^^ where ∗ is this times the convolution operator. In this document we note "Relative distribution ^^ ^^⁄ ^^ » the distribution of the random variable resulting from the difference (or sum) of the two random variables ^^ ^^ and ^^ ^^. A relative distribution represents probabilities of differences between pairs of symbols received in the frame. These relative distributions are determined for all pairs of symbols of the frame, ^^, ^^ ∈ [0, ^^ − 1], by cross-decorrelation or convolution operations. More precisely, a difference between a symbol ^^ ^^^^ and a symbol ^^ ^^^^ is determined by relating these symbols by cross-correlation or convolution. This difference corresponds, for example, to the value of an offset between two CCSK sequences of the frame. Thus, for each pair of symbols (^^^^^^, ^^^^^^), we determine a vector ^^ ^^ / ^^ whose size is equal to the cardinality K of the symbol dictionary, and whose element ^^ ^^ / ^^[^^] includes a probability that the relative shift between symbols of value ^^. These relative distributions allow to obtain a matrix of relative distributions in which an element of coordinates (i, j) is a relative distribution ^^ ^^ / ^^ : ] ^^0⁄ ^^−1 ⋯ ^^^^−1⁄ ^^−1 According to a particular embodiment, the K symbols of the dictionary are associated with K distinct index values ​​of a finite group of order K, so that the differences between the symbols are expressed by differences between the indices associated with them. The method comprises a step 302 of determining a masking matrix of dimensions ^^ ∗ ^^ in which an element (^^; ^^) ^^, ^^ ∈ [0, ^^ − 1] is a Kronecker distribution [0], a Kronecker uniform distribution ^^ or an intermediate distribution between a Kronecker distribution and a uniform distribution, a particular distribution being selected according to a reliability level granted to the difference between the symbols ^^ ^^^^ and ^^ ^^^^ in the frame. For example, the device 102 may obtain a transmission error rate associated with a symbol ^^ ^^^^ and a transmission error rate associated with a symbol ^^ ^^^^, and determine a mask distribution ^^⁄ ^^ according to the error rates obtained, so that the lower the error rate, the more the distribution resembles a Kronecker distribution [ 0 ] , and the higher the error rate, the more the distribution resembles a uniform distribution. According to a particular embodiment, each element (^^; ^^) ^^, ^^ ∈ [0, ^^ − 1] of the masking matrix is ​​determined according to the relative position of the symbols ^^ ^^^^ and ^^ ^^^^ in the frame and according to a level of temporal variability of the channel. Indeed, under certain conditions, notably when the channel undergoes temporal variations (such as a Doppler effect), the reliability of the relative distribution can be inversely proportional to the symbol spacing ^^ ^^^^ and ^^ ^^^^in the frame. According to a particular embodiment, an element (^^; ^^) ^^, ^^ ∈ [0, ^^ − 1] of the masking matrix is ​​a mask distribution ^^⁄ ^^ which is a Kronecker distribution [0] when the number of symbols that separate symbols ^^ ^^^^ and ^^ ^^^^ in the frame is less than a threshold, and a uniform distribution ^^ when the difference is greater than or equal to said threshold, the threshold being determined according to a characteristic of the communication channel. According to a particular implementation, an element (^^; ^^) ^^, ^^ ∈ [0, ^^ − 1] of the masking matrix is ​​a mask distribution ^^⁄ ^^ intermediate between the Kronecker distribution [0] and the uniform distribution ^^ such that mask^^⁄ ^^ = ^^^^ / ^^^^[0] + (1 − ^^^^⁄ ^^ )^^, the parameter ^^^^ / ^^being determined according to the number of and ^^ ^^^^in the frame, and / or according to a characteristic of the communication channel, such as a level of temporal variability of the channel. Different characteristics of the communication channel can be taken into account to determine one or more masking distributions. This may be a level of temporal variability of the channel, but not only. For example, in the case of a MIMO system, some layers may be particularly correlated and others not, so that the symbols of a frame may be influenced differently by the channel. The masking distributions can thus be determined according to a level of correlation between the layers of a MIMO system. The method comprises a step 303 during which the matrix of relative distributions determined in step 301 is combined by convolution with a masking matrix determined in step 302. Thus, each relative distribution ^^ ^^ / ^^ is convolved with a masking distribution mask^^⁄ ^^ so as to give more or less importance to the distribution depending on the reliability of the symbols ^^ ^^^^ and ^^ ^^^^, or a reliability associated with their difference. The method finally comprises a demodulation step 305 during which at least part of the relative distributions convolved in step 304 is used to estimate the value of a symbol from another taken as reference. For example, the demodulation can be carried out from the probabilities of differences between a symbol whose value and location in the frame are known, for example a pilot symbol, and the other symbols of the frame. Knowledge of a reference symbol and the probabilities of differences between this symbol and the other symbols of the frame makes it possible to estimate the value of these other symbols. According to another example, the demodulation can consist of a shifted demodulation from the consolidated relative distributions.Offset demodulation consists of demodulating the frame by taking an arbitrary symbol value as a reference and performing an integrity check on the information words thus obtained. If the integrity check does not validate the frame, a new demodulation is performed by taking another symbol value as a reference and a new integrity check is performed. These steps are repeated until the frame is validated by the integrity check. In a particular embodiment, the demodulation is performed on the basis of recombined relative distributions, obtained by relating at least a first relative distribution to which a mask has been applied, determined from ^^ symbols. ^^^^ and ^^ ^^^^with a second masked relative distribution, determined from a symbol^^^^^^ and the symbol ^^^^^^ , with ^^, ^^ ≠ ^^, and ^^ ∈ [0; ^^ − 1]. In other words, we determine a difference between a first symbol and a second symbol by transitivity, that is to say from a difference between the first symbol and a third symbol and from an index difference between the second symbol and the same third symbol. We can thus obtain a plurality of relative distributions ^^ (^^) ^ ^ / ^^ so-called partial, representing the difference between the first and second symbol, but based on different symbols. Partial relative distributions ^^ (^^) ^ ^ / ^^ determined for a particular ^^ value are then combined by product to obtain a consolidated version of a relative ^^ distribution ^^⁄ ^^ on the basis of which the demodulation step takes place. Figure 4a is a graph showing an example of relative distribution ^^^^ / ^^ =[0.23, 0.1, 0.34, 0.1, 0.23] representing the probabilities of difference between symbols^^ ^^^^ and ^^ ^^^^ of a data frame for a symbol dictionary of cardinality K=5. Figure 4b represents a graph showing a masking distribution mask ^^⁄ ^^ = [1, 0, 0, 0, 0], which is a Kronecker distribution [0]. Figure 4c is a graph illustrating the result of the convolution of the distribution ^^ ^^ / ^^ of Figure 4a with the masking distribution of Figure 4b. The resulting distribution is thus identical to the original distribution of Figure 4a. In other words, convolution by the Kronecker distribution [0] does not affect the distribution. Figure 5a is a graph showing an example of a relative distribution ^^ ^^ / ^^ =[0.23, 0.1, 0.34, 0.1, 0.23] representing the probabilities of difference between symbols^^ ^^^^ and ^^ ^^^^of a data frame for a symbol dictionary of cardinality K=5. Figure 5b represents a graph showing a masking distribution mask ^^⁄ ^^ = [1 / 5, 1 / 5, 1 / 5, 1 / 5, 1 / 5], or a uniform distribution. Such a distribution is, for example, selected to mask relative distributions that are considered unreliable. Figure 5c is a graph illustrating the result of the convolution of the distribution ^^ ^^ / ^^ of Figure 5a with the masking distribution of Figure 5b. The resulting distribution is a uniform distribution whose effect on demodulation will be neutral. Figure 6a is a graph showing an example of a relative distribution ^^ ^^ / ^^ =[0.23, 0.1, 0.34, 0.1, 0.23] representing the probabilities of difference between symbols^^ ^^^^ and ^^ ^^^^ of a data frame for a symbol dictionary of cardinality K=5. Figure 6b represents a graph showing a masking distribution mask ^^⁄ ^^= [0.6, 0.1, 0.1, 0.1, 0.1], which is an intermediate distribution between a Kronecker distribution [0] and a uniform distribution. Such a distribution is, for example, selected to mask relative distributions whose reliability is average. Figure 6c is a graph illustrating the result of the convolution of the distribution ^^ ^^ / ^^of Figure 6a with the intermediate masking distribution of Figure 6b. The resulting distribution is one that reduces the gap between low and high probabilities, so that the effect on demodulation is mitigated. Although the method is described here with reference to discrete distributions, the present proposal can be generalized to continuous distributions, for example by approximating the continuous distribution by a discretized distribution, or by modeling the distributions by their parameters (for example the mean ^^ and the variance ^^ as parameters of a normal distribution). It is then appropriate to express the relationship between the parameters of the relative distributions. Figure 7 represents a simplified architecture of a device 700 adapted to implement the demodulation method according to a particular embodiment.The device 700 comprises a data processing module comprising a storage space 701, for example a memory (MEM), a processing unit 702, equipped for example with a microprocessor (PROC), and controlled by a computer program (PGR) 703 whose instructions are configured to implement the demodulation method as described previously in relation to FIG. 3. At initialization, the code instructions of the computer program 703 are for example loaded into the memory 701 before being executed by the processor of the processing unit 702. The microprocessor of the processing unit 702 implements, according to the instructions of the computer program 703, the steps of the demodulation method described above with reference to FIG. 3.For this, in addition to the memory 701 and the processor 702, the device comprises communication means 704, for example an OFDM transducer adapted to receive a signal on a plurality of orthogonal carriers. The communication means 704 are for example configured by computer program instructions to allow the reception of at least one time-frequency frame in which each of the N time steps contains a complete symbol of size K, for example a CCSK sequence, distributed over all the subcarriers in frequency, and to demodulate the N OFDM symbols received. The device 700 also comprises a module 705 for determining a matrix comprising a set of probability distributions, called relative distributions, each element of the matrix being a vector of K probabilities representative of at least one difference between symbols ^^. ^^^^ and ^^ ^^^^ of a plurality of pairs of symbols (^^ ^^^^ ; ^^ ^^^^) formed from the symbols composing a frame received by the communication means 704, with ^^ ∈ [0; ^^ − 1] and ^^ ∈ [0; ^^ − 1], a difference being calculated according to modular arithmetic. The module 707 is for example implemented by computer program instructions configured to calculate a difference between index values ​​associated with two symbols of the frame by a cross-correlation and / or convolution operation and to store the result of the difference between the elements of a pair (^^ ^^^^ ; ^^ ^^^^ ) of sequences at the coordinates (^^; ^^) in the matrix. The device 700 comprises a module 706 for determining a masking matrix comprising a set of mask distributions ^^⁄ ^^ determined according to a level of reliability of the symbols ^^ ^^^^ and ^^ ^^^^and / or a reliability level of the difference between these symbols. Module 706 may be implemented by program instructions configured to assign a Kronecker distribution [0] to the element at coordinates (i; j) of the matrix when the number of symbols that separate symbols ^^ ^^^^ and ^^ ^^^^ in the frame is less than a threshold, and a uniform distribution ^^ when the difference is greater than or equal to said threshold, the threshold being determined as a level of temporal variability of the channel. According to a particular embodiment, the program instructions are configured to assign to the element (i; j) of the matrix an intermediate distribution between the Kronecker distribution [0] and the uniform distribution ^^ such that mask^^⁄ ^^ + (1 − ^^^^⁄ ^^ )^^, the parameter ^^ ^^ / ^^ being determined according to the number of symbols separating the symbols ^^ ^^^^ and ^^ ^^^^in the frame and a level of temporal variability of the channel. The device 700 also comprises a module 707 for calculating a convolution of the matrix of relative distributions determined by the module 705 with the masking matrix determined by the module 706. The module can be implemented by program instructions configured to calculate such a convolution and produce a convolved relative distribution matrix in which the importance of the distributions is affected by the corresponding masking distributions. The device 700 finally comprises a demodulation module 708 adapted to demodulate the symbols transmitted in the received frame from the relative distributions convolved by the module 707.For example, the module 708 is implemented by program instructions adapted to carry out an offset demodulation of the data frame based on the convolved relative offsets, for example from the offsets of one or more particular rows of the convolved matrix. According to another example, the instructions are configured to demodulate the symbols of the frame from the differences between these symbols and a reference symbol whose location and value are known. In a particular embodiment, the device 700 is integrated into a communication terminal, a connected object, a vehicle, a gateway, an access point, a communication satellite or even a base station.

Claims

CLAIMS 1. Method for demodulating a data frame comprising ^^ information words [ ^^0 .. ^^^^−1] dans lequel un mot ^^^^ occupant une position ^^ ∈ [0, ^^ − 1] dans la trame est modulated by a symbol ^^ ^^^^ of a dictionary of ^^ symbols, the method comprising the following steps: - Détermination (301), pour une pluralité de symboles ^^^^^^ et ^^^^^^ de la trame, avec ^^, ^^ ∈ [0, ^^ − 1], d’un vecteur de ^^ probabilités ^^^^ / ^^, dit distribution relative, comprenant probabilities associated with ^^ possible values ​​of an estimated difference between the symbols ^^ ^^^^ and ^^ ^^^^ , - Combination (303) of a relative distribution ^^ ^^ / ^^ with a particular probability distribution mask ^^⁄ ^^ determined (302) according to a characteristic of the communication channel, - Demodulation (304) of the frame from at least a part of the combined distributions.

2. Method according to claim 1 in which the demodulation of the frame is carried out from at least one partial relative distribution ^^ (^^) ^ ^ / ^^ determined by relating a first combined relative distribution determined from symbols ^^ ^^^^ and ^^ ^^^^with a second combined relative distribution determined from a ^ symbol ^^^^^ et du symbole ^^^^^^ , avec ^^, ^^ ≠ ^^, et ^^ ∈ [0; ^^ − 1].

3. Method according to any one of the preceding claims in which a particular probability distribution mask ^^⁄ ^^ is determined according to the reliability level determined for the symbols ^^ ^^^^ and ^^ ^^^^ and / or a determined reliability level for the difference between these symbols.

4. Method according to any one of the preceding claims in which a relative distribution ^^ ^^ / ^^ is convolved with a Kronecker distribution with a peak at 0 when the number of symbols that separate symbols ^^ ^^^^ and ^^ ^^^^ in the frame is less than a threshold, and with a uniform distribution when this number is greater or equal to said threshold, the threshold being determined according to a characteristic of the communication channel.

5. Method according to any one of claims 1 to 3 in which a relative distribution ^^ ^^ / ^^ is convolved with a mask probability distribution ^^⁄ ^^ intermediate between a Kronecker distribution [0] and a uniform distribution ^^ such that m ask^^⁄ ^^ = ^^^^ / ^^^^[0] + (1 − ^^^^⁄ ^^ )^^, le paramètre ^^^^ / ^^ étant déterminé selon une characteristic of the communication channel.

6. Method according to any one of the preceding claims in which a probability distribution mask ^^⁄ ^^ is determined by the number of symbols separating the symbols ^^ ^^^^ and ^^ ^^^^ in the frame and a level of temporal variability of the channel.

7. A method according to any one of the preceding claims in which the method is such that a symbol ^^ ^^^^ is a sequence of complex values ​​obtained by applying a particular cyclic shift to a root sequence of size ^^, a sequence ^^^^^^ being received on ^^ subcarriers of an OFDM symbol.

8. Device for demodulating a data frame comprising N information words [ ^^0 .. ^^^^−1] dans lequel un mot ^^^^ occupant une position ^^ ∈ [0, ^^ − 1] dans la trame est modulated by a symbol ^^ ^^^^ of a dictionary of ^^ symbols, the device comprising a processor (702) coupled to a memory (701) in which are recorded program instructions (703) configured to implement the following steps: - Détermination, pour une pluralité de symboles ^^^^^^ et ^^^^^^ de la trame, avec ^^, ^^ ∈ [0, ^^ − 1], d’un vecteur de ^^ probabilités ^^^^ / ^^, dit distribution relative, comprenant probabilities associated with ^^ possible values ​​of an estimated difference between the symbols ^^ ^^^^ and ^^ ^^^^ , - Combination of a relative distribution ^^ ^^ / ^^ with a particular probability distribution mask ^^⁄ ^^ determined according to a characteristic of the communication channel, - Demodulation of the frame from at least part of the combined distributions.

9. Communication terminal comprising a demodulation device according to claim 8.

10. Computer program comprising instructions configured to implement a demodulation method according to any one of claims 1 to 7, when the instructions are executed by a processor.

11. Information medium readable by a computer on which instructions configured to implement a demodulation method according to any one of claims 1 to 7 are recorded.

Images