FR2821702A1 - OFDM signal modulations optimized reception having two part header section synchronizing frame/correcting error second header part and carrying out second synchronization using inter correlation then demodulating. - Google Patents

Abstract

The signal reception frame processing has digital words with a lead section. The header section has two parts. A synchronization operation (500) is carried out detecting a new frame separately from the header section. Correction (502) is carried out by correcting information in the second part of the header section. A second synchronization attempt (504) is carried out by intercorrelation between information previously corrected and a predetermined signal. Following the results of the second synchronization a demodulation operation (506) is carried out demodulating the frame digital word assembly.

Description

<Desc / Clms Page number 1>

The present invention relates to a method and to an optimized reception device.

 The invention is described here, by way of nonlimiting example, in its application to signals modulated according to an OFDM type modulation (orthogonal frequency division multiplexing, in English "Orthogonal Frequency Division Multiplex"), and more particularly, to signals transmitted in accordance with the Hiperlan 2 standard.

 When transmitting a message using an orthogonal frequency division modulation, the binary data of the message to be transmitted are divided into data blocks. Each of these data blocks is transmitted independently and constitutes, after baseband modulation, an OFDM symbol.

 In each of these data blocks, the elements are also grouped by subsets, each subset then undergoing a cartographic report on a discrete set of points in the Fresnel space, each of these points representing a phase and an amplitude possible. This application is bijective.

 Thus, for example, in a message consisting of the following sequence {000011100100011110001010100100100 ...}, we can extract, in the order of presentation, a block of 16 binary data 0000111001000111 with which we associate the map report 1 + j, 1+ j, -1-j, 1-j, -1 + j, 1 + j, -1 + j, -1-j. So we have a set of 8 complex elements of a vector V.

 These vectors are then multiplied by a fast inverse discrete Fourier transform matrix M to obtain an OFDM symbol made up of a series of complex amplitudes. This symbol is transmitted after other processing.

 The OFDM symbol is received, after its passage in the transmission channel, in a demodulator, from which a vector with complex elements V 'will be extracted by multiplying the amplitudes constituting this symbol by a discrete Fourier transform matrix M', such that M. M '= Id (matrix

<Desc / Clms Page number 2>

identité). Des décisions au maximum de vraisemblance sur la partie réelle et imaginaire permettent de retrouver la séquence de symboles initiale, puis de restituer les éléments binaires associés.

identity). Maximum likelihood decisions on the real and imaginary part make it possible to find the initial sequence of symbols, then to restore the associated binary elements.

 For more details on this modulation technique, one can consult, for example, the work entitled "OFDM for Wireless Multimedia Communications" by Richard VAN NEE and Ranjee PRASAD published at Artech House.

 There are, in particular in the various publications used to develop the Hiperlan 2 standard, various solutions for performing the synchronization steps prior to the demodulation of the OFDM symbols received.

 For example, document HL12ERi6A entitled "Proposa / br a common preamble for Hiperlan 2 and IEEE802. 1" by Nokia and Ericsson (available from ETSI ("European Telecommunications Standards Institute")) provides the basic structure of a preamble adapted to the desired characteristics of the transmission system.

 This structure consists of three parts A, B, C respectively proposed as means of carrying out the following operations: - part A: automatic gain control and rough time synchronization; - part B: automatic gain control, fine time synchronization and estimation of the frequency offset; - part C: estimation of the channel.

Furthermore, the document HL 13S0N1A entitled "HL2 Physical Layer Synchronization - structure of the preamble" from Sony (also available from ETSI) proposes to use part C of the preamble as a means of estimating the channel in the case of 'a burst of the "Broadcast" type and to use it as a means of channel estimation and as a means of fine time synchronization and correction of the frequency offset in the case of a "Downiink" type burst (the "Broadcast" bursts "and" Downlink "are illustrated in Figure 1 described below).

The work entitled "OFDM for Wireless Multimedia Communications" cited above also describes the various problems that must be solved by a

<Desc / Clms Page number 3>

récepteur OFDM et propose plusieurs méthodes pour effectuer les opérations de synchronisation en temps et en fréquence.

OFDM receiver and offers several methods for performing time and frequency synchronization operations.

 When the data received is too damaged to be recovered correctly, the known receivers perform unnecessary calculations, which represents a considerable energy consumption; in particular, when the receiver is included in a mobile telephone powered by battery, this unnecessary consumption of energy is detrimental to the autonomy of operation of the telephone.

 The present invention aims to remedy the aforementioned drawbacks.

 To this end, the present invention provides a method for receiving signals in the form of frames comprising a preamble and data, this preamble comprising at least two parts, remarkable in that it comprises steps according to which: - a first synchronization operation, consisting in detecting the arrival of a new frame from the first part of the preamble, - a correction operation is carried out, consisting in correcting information contained in the second part of the preamble of the new frame, - a second synchronization attempt is made, by intercorrelation between the previously corrected information and a predetermined signal, and - depending on the result of the second synchronization, a demodulation operation is carried out or not, consisting in demodulating all of the data contained in the frame.

 Thus, the invention makes it possible to provide a signal intended to authorize or not the complete demodulation of a received burst. By inhibiting the end of demodulation in the case of a signal too disturbed to provide valid data, the invention makes it possible to save the energy of the receiver, which is particularly useful when, for example, it is located in a mobile terminal powered by a battery.

<Desc / Clms Page number 4>

According to a particular characteristic, during the first synchronization operation, the error in coarse frequency is also determined.

 Thus, as will be seen later in the detailed description, it is the same autocorrelation function which makes it possible both to detect the arrival of a new frame and to determine the error in coarse frequency.

 According to a particular characteristic, during the correction operation, the error in fine frequency is determined.

 According to a particular characteristic, during the second synchronization attempt, an attempt is made to perform fine time synchronization.

 According to a particular characteristic, the predetermined signal is a validation signal which controls the demodulation operation.

 According to a particular characteristic, the demodulation operation implements demodulation of the OFDM type.

 According to a particular characteristic, the signals are in the form of frames conforming to the Hiperlan 2 standard, the first part of the preamble consisting of sequences A and B and the second part of the preamble consisting of sequence C.

 For the same purpose as that indicated above, the present invention also provides a device for receiving signals in the form of frames comprising a preamble and data, this preamble comprising at least two parts, remarkable in that it comprises: a first synchronization unit, for detecting the arrival of a new frame from the first part of the preamble, - a correction unit, for correcting information contained in the second part of the preamble of the new frame, - a unit for synchronization, to carry out a second synchronization attempt by intercorrelation between the previously corrected information and a predetermined signal, and - a demodulation unit, for, depending on the result of the second synchronization, demodulating or not all the data contained in the frame .

<Desc / Clms Page number 5>

The present invention also relates to an apparatus for processing digital signals, comprising means suitable for implementing a reception method as above.

 The present invention also relates to an apparatus for processing digital signals, comprising a reception device as above.

 The present invention also relates to a telecommunications network, comprising means suitable for implementing a reception method as above.

 The present invention also relates to a telecommunications network, comprising a reception device as above.

 The present invention also relates to a mobile station in a telecommunications network, comprising means suitable for implementing a reception method as above.

 The present invention also relates to a mobile station in a telecommunications network, comprising a reception device as above.

 The present invention also relates to a base station in a telecommunications network, comprising means suitable for implementing a reception method as above.

 The present invention also relates to a base station in a telecommunications network, comprising a reception device as above.

 The invention also relates to: - a means of storing information readable by a computer or a microprocessor retaining instructions of a computer program, allowing the implementation of a reception method as above, and - a removable information storage means, partially or totally, readable by a computer or a microprocessor retaining instructions of a computer program, allowing the implementation of a reception method as above.

 The invention also relates to a computer program product comprising sequences of instructions for implementing a reception method as above.

<Desc / Clms Page number 6>

The special features and advantages of the receiving device, the different digital signal processing devices, the different telecommunications networks, the different mobile stations, the different base stations, the different storage means and the computer program product being similar to those of the reception method according to the invention, they are not repeated here.

 Other aspects and advantages of the invention will appear on reading the detailed description which follows of a particular embodiment, given by way of nonlimiting example. The description refers to the accompanying drawings, in which: - Figure 1 schematically illustrates the structure of a frame in accordance with the Hiperlan 2 standard; - Figure 2 schematically illustrates the structure of the preambles of the different phases of a frame according to Figure 1; - Figure 3 is a flowchart illustrating the main steps of a reception method according to the present invention, in a particular embodiment; - Figure 4 shows schematically the structure of a receiving device according to the present invention, in a particular embodiment; - Figure 5 shows schematically the structure of the first synchronization device included in a reception device according to the present invention, in a particular embodiment; - Figures 6a and 6b illustrate the operation of the first synchronization device of Figure 5; - Figure 7 shows schematically an exemplary embodiment of the frequency error correction units included in the receiving device of Figure 4; - Figure 8 shows schematically the structure of the second synchronization device included in a reception device according to the present invention, in a particular embodiment;

<Desc / Clms Page number 7>

- les figures 9a, 9b et 9c illustrent le fonctionnement du second dispositif de synchronisation de la figure 8 ; - la figure 10 illustre schématiquement la constitution d'une station de réseau ou station de réception informatique adaptée à mettre en oeuvre un procédé de réception conforme à la présente invention ; et - la figure 11 représente sous une forme schématique simplifiée un réseau de télécommunications conforme à la présente invention, dans un mode particulier de réalisation.

- Figures 9a, 9b and 9c illustrate the operation of the second synchronization device of Figure 8; - Figure 10 schematically illustrates the constitution of a network station or computer reception station adapted to implement a reception method according to the present invention; and - Figure 11 shows in simplified schematic form a telecommunications network according to the present invention, in a particular embodiment.

 By way of illustration, the preferred embodiment is described in the case of a data transmission according to the Hiperlan 2 standard. The invention is therefore applied to the various parts of the preamble as defined in the document ETSI 101 475 V1 . 2.1 entitled "Broadband Radio Access Network; HIPERLAN type 2; Physical layer '.

 Figure 1 shows schematically the structure of a Hiperlan 2 frame (MAC frame). It summarizes the information contained in the ETSI TR 101 683 V1 documents. 1.1 and ETSI 101 475 V1. 2.1.

 The frame is made up of several phases, commonly called bursts (in English "bursf"): - the general broadcast phase or "Broadcast" burst, located at the start of the frame, which contains information intended for all receivers (transmission from the base station to the mobiles); - the "Downlink" burst, which carries information intended for particular receivers (transmission from the base station to the mobiles); - the "Direct link" burst, which allows receivers to exchange information directly, without going through a base station (transmission from mobile to mobile); - the "Uplink" burst, which carries information intended for the base station (transmission from mobile to base station) - the "Random access" burst, which allows mobiles which have no channels assigned in the "Uplink" burst, to communicate with the base station.

<Desc / Clms Page number 8>

Each of these bursts has a header called a preamble. Figure 2 shows the constitution of these different headers, which are all based on the use of particular data sequences called A, RA, B, IB, C.

The content of these sequences has been determined so that they have particular properties with regard to certain mathematical operations. Reference is usefully made on this subject to the Hiperlan 2 standard and to the contributions which enabled the development of this standard, these documents being available from ETSI.

 According to a classic scheme in OFDM, the useful symbols (here called "data") are preceded by a prefix, composed of the repetition of a certain number of samples of the symbol. In the context of the Hiperlan 2 standard, this prefix, designated by CP (in English "Cyclic Prefix") is the copy of the last 16 samples of the following symbol.

 The "Broadcast" phase being located at the head of the frame, its preamble will have the task of waking up and synchronizing the receiver, which leads to a longer preamble than for the other phases.

 As can be seen in FIG. 2, the C sequences are present in all types of bursts and therefore it is possible to apply the invention, not only to the "Broadcast" burst, but also to the other parts of the message.

 The flowchart of Figure 3 illustrates a particular embodiment of the reception method according to the present invention.

 It is assumed that signals are received in the form of frames. These frames include a preamble and data. The preamble has at least two parts. For example, in the case of a frame conforming to the Hiperlan 2 standard, a “Broadcast” type preamble such as that illustrated in FIG. 2 is considered. The first part of the preamble consists of the sequences A and B and the second part of the preamble consists of the sequence C.

 During a first step 500, a first synchronization operation is carried out, consisting in detecting the arrival of a new

<Desc / Clms Page number 9>

trame à partir des séquences A et B, par étude de l'amplitude du signal d'autocorrélation calculé sur les séquences A et B.

frame from sequences A and B, by studying the amplitude of the autocorrelation signal calculated on sequences A and B.

o fréquence grossière. During step 500, the error in coarse frequency (in English "coarse") is also estimated over the rest of the frame, from the value of the phase of the autocorrelation signal, and the error is corrected. error in

o rough frequency.

 The sequence used to carry out this first synchronization operation is designated by the reference 1 on the "Broadcast" burst illustrated in FIG. 2.

 A possible embodiment corresponding to step 500 is described in detail below in relation to FIG. 5.

 Then, during a step 502, a second correction operation is carried out, consisting in estimating the residual fine-frequency error from the phase of the autocorrelation signal calculated on the sequences C of the "Broadcast" preamble, and to correct the error in fine frequency (in English "fine") throughout the remainder of the frame.

Then, during a step 504, an attempt is made to second synchronization, by intercorrelation between the corrected symbols C and their expected theoretical value. Depending on the result of this operation, it generates or not a validation signal indicating that the data are of sufficiently good quality to be demodulated. If the final demodulation of the data is validated, there is also generated, in step 504, in a manner known per se, a sub-sampling command and a windowing signal (in English "windowing") for the fast Fourier transformation (FFT), allowing correct demodulation of data
A possible embodiment corresponding to step 504 is described in detail below in relation to FIG. 8.

 If the demodulation operation is carried out, in the next step 506, conventionally, the signal is sub-sampled and it is cut into blocks adapted to the size of the FFT.

 It is assumed that signals modulated according to OFDM type modulation are processed. In this case, we know that the useful symbols are

<Desc / Clms Page number 10>

précédés d'un préfixe composé de la répétition d'un certain nombre d'échantillons du symbole. Dans le cadre de la norme Hiperlan 2, on a vu plus haut que ce préfixe CP est la recopie des 16 derniers échantillons du symbole suivant.

preceded by a prefix composed of the repetition of a certain number of samples of the symbol. As part of the Hiperlan 2 standard, we saw above that this prefix CP is the copy of the last 16 samples of the following symbol.

 During the demodulation step 506, these prefixes CP are deleted and the OFDM demodulation is carried out.

 Then, it remains to: - estimate the channel by demodulation of the symbol C and compare the signal obtained with the theoretical signal emitted, - carry out a channel correction, - correct the common phase error due to the non-enslavement of the clock d sampling of the analog digital 1 converter located at the input of the receiver, - extract the symbols carried by the subcarriers, and - extract binary data from these symbols.

 These operations are described in more detail below.

 FIG. 4 schematically represents the architecture of a reception device according to the present invention.

 In the following description, we have deliberately omitted to describe in detail certain parts of the receiver according to the invention, given that these are known per se and are not essential to the implementation of the invention. This particularly concerns the so-called final demodulation part.

 However, for greater clarity in the operation of the receiver, the elements illustrating the fast Fourier transformation (FFT), the channel estimation, the channel equalization, the common phase correction, the reverse map transfer on the carrier ( in English "carrier demapping") and the reverse QPSK cartographic report (in English "QPSK demapping") are nevertheless represented on the drawings.

 At the input of the reception device illustrated in FIG. 4, the analog signal received by the radio frequency (RF) interface is sent to an analog / digital conversion unit 10 which samples the received signal and the

<Desc / Clms Page number 11>

convertit en un signal numérique. A titre d'exemple nullement limitatif, on peut choisir une fréquence intermédiaire égale à 25 MHz et un facteur de sur- échantillonnage égal à 4, ce qui conduit à une fréquence d'échantillonnage de 100 MHz.

converts to a digital signal. By way of nonlimiting example, one can choose an intermediate frequency equal to 25 MHz and an oversampling factor equal to 4, which leads to a sampling frequency of 100 MHz.

 The digital signal is then transmitted to an IF demodulation unit 20 which brings the modulated OFDM signal around 25 MHz in baseband, in a manner known per se.

 The baseband signal is then transmitted simultaneously to a coarse time and frequency synchronization unit 40 and to a coarse frequency error correction unit 30. The coarse synchronization unit 40 deduces from this signal, on the one hand, information 41 for the arrival of a signal to be demodulated and, on the other hand, information 42 representative of the coarse frequency error.

 These two pieces of information are transmitted to the correction unit 30, which starts the demodulation process, corrects the received baseband signal by subtracting the coarse frequency error 42 and supplies the corrected signal to an estimation unit 50 and fine frequency correction.

 The unit 50 corrects the residual frequency error, as described below, and transmits the fully frequency corrected signal to a fine time synchronization unit 60 and to a subsampling and windowing unit 70.

 The fine synchronization unit 60, the operation of which is described below, provides, on the one hand, a sub-sampling and windowing command 61 for the FFT allowing correct demodulation, to the sub-sampling unit 70 and windowing and, on the other hand, a validation signal 65 authorizing the final demodulation of the data.

 The validation signal 65 controls all of the units located downstream, namely: - the unit 70 for sub-sampling and windowing, - a unit 80 for fast Fourier transformation FFT, - a unit 90 for estimating channel, - a channel equalization unit 100,

<Desc / Clms Page number 12>

- une unité 110 de correction de phase commune, - une unité 120 de report cartographique inverse sur la porteuse, et - une unité 130 de report cartographique inverse QPSK.

- A unit 110 for common phase correction, - a unit 120 for reverse map transfer on the carrier, and - a unit 130 for reverse map transfer QPSK.

 The subsampling and windowing unit 70 performs a subsampling operation on the signal received from the frequency estimation and fine correction unit 50 according to the supplied subsampling and windowing command 61 by the fine synchronization unit 60.

 The sub-sampled signal is also cut into data blocks adapted to the size of the FFT (windowing operation) in the unit 70.

 These data blocks are then transmitted to the fast Fourier transformation unit 80, which performs the demodulation proper. After demodulation, the part of these data blocks which corresponds to the transmission of perfectly known fixed sequences (sequences C of the preamble) is simultaneously sent to the channel estimation unit 90, which compares it with the theoretical sequences transmitted and in deduces the disturbance due to the channel.

 A disturbance signal 91 is sent by the channel estimation unit 90 to the channel equalization unit 100, which applies the necessary correction to the demodulated signals from the fast Fourier transformation unit 80, which unit 100 receives on another input.

 Once the demodulated signals have received this so-called "channel" correction, they are transmitted to the common phase correction unit 110, which performs a rotation of the points received so as to compensate for the sampling error at receiver input.

 The corrected signals are then transmitted to the unit 120 for reverse cartographic transfer on the carrier, which performs a deinterlacing operation of the subcarriers and outputs the complex signals as they were ordered at the input of the OFDM modulator.

 These signals are then transmitted to the reverse mapping report unit 130 QPSK, which performs a reporting operation.

<Desc / Clms Page number 13>

cartographique inverse de celle utilisée à l'émetteur de façon à restituer les signaux binaires tels qu'ils étaient ordonnés à l'entrée de l'émetteur OFDM.

reverse map to that used at the transmitter so as to reproduce the binary signals as they were ordered at the input of the OFDM transmitter.

 Thus, the invention provides for double synchronization, the coarse synchronization unit 40 in time and frequency constituting a first synchronization device and the fine synchronization unit 60 constituting a second synchronization device.

 FIG. 5 illustrates an exemplary embodiment of the first synchronization device, using an autocorrelation function to determine the arrival of a signal of the desired type and therefore, trigger the "awakening" of the receiver.

In this example, the autocorrelation function is performed as follows:
The complex input signal 200 is simultaneously sent to a delay input unit 210, which delays the signal 200 by a number of samples, denoted D, and to a multiplier 230. The output of the unit 210 delay input is then transmitted to a conjugation unit 220 which transforms the complex numbers received into their conjugate complexes.

 The output of the conjugation unit 220 is transmitted to the second input of the multiplier 230. The output of the multiplier 230 is sent to a first averaging unit 240 which calculates the average of the signal received on the last MD points.

 Furthermore, the complex input signal 200 is also sent to a module calculation unit 260 which calculates the module of the complex number received. This module is sent to a second averaging unit 270 which performs the same operation as the first averaging unit 240. The second averaging unit 270 outputs a normalization signal 271.

 The output signal of the first averaging unit 240, denoted x (i) in FIG. 5, is sent to a first input of a divider 250, the second input y (i) of the divider receiving the normalization signal 271 from of the second averaging unit 270.

 The output of the divider 250 is a complex number which constitutes the output of the autocorrelation operator.

<Desc / Clms Page number 14>

The amplitude of this number is transmitted to a threshold unit 300, which performs a thresholding operation and transmits the data above a predetermined threshold to a maximum detection unit 310, which applies a maximum detection algorithm to these data. .

 The maximum detection unit 310 produces the signal 41 to start the demodulation process.

 The phase of the complex number from the autocorrelation operator is transmitted to a unit 320 which extracts the value of the coarse frequency error therefrom and delivers it in the form of signal 42.

 FIGS. 6a and 6b illustrate the operation of the first synchronization device in FIG. 5.

 The graph at the top of figure 6a represents the amplitude of the calculated autocorrelation signal and the correspondence of the peaks obtained with the sequence of data received (bottom of figure 6a).

 When receiving a Hiperlan 2 type signal, two peaks appear clearly when the correlation window, defined by the value of the delay D and the size of the averaging window MD (the parameters D and MD having been defined above in relation to FIG. 5), is located on the sequences received "RA-A-RA" or "BBB". The correlation window is shown on the drawing by hatching.

 The autocorrelation signal being normalized (that is to say that its maximum amplitude is 1), a thresholding is applied to this signal (see thresholding unit 300 in FIG. 5), for example to the threshold value of 0.8 then a maximum detection algorithm (see unit 310 for maximum detection in FIG. 5) which makes it possible to detect peaks.

 In the particular embodiment described here, the detection signal of the first peak is used to determine the arrival of a new frame to be demodulated as well as to indicate its approximate time position (signal 41 in FIGS. 4 and 5).

 In this embodiment, the length of the sequences A, RA, B and IB is 64 points, the delay D is 64 points and the size of the averaging window MD is 192 points.

<Desc / Clms Page number 15>

FIG. 6b illustrates the correspondence between the values of the amplitude (English term "magnitude" plotted on the ordinate of the graph at the top of FIG. 6b, the abscissa representing the samples, or "samples" in English) and of the phase of the autocorrelation signal (illustrated in the bottom graph of Figure 6b).

 One of the important features of the sequences A, RA, B and IB chosen is that, when the autocorrelation signal described above is calculated, the value of its phase at the time of the appearance of the peaks on its amplitude is significant by l frequency error resulting from demodulation FI (unit 20 in FIG. 4).

 We see in Figure 6b that the error is centered around 180 degrees opposite the first amplitude peak and is centered around 0 opposite the second amplitude peak. By using the signal generated previously to memorize this error value, we therefore have the information necessary to correct the coarse error in frequency, thanks to the unit 30 of FIG. 4.

 In the particular embodiment described here, the first peak is used and, to improve the operation of the system and its sensitivity to a possible detection error, the value of the frequency error is averaged over 20 points before being transmitted to the correction unit 30 (signal 42 in FIGS. 4 and 5).

 FIG. 7 shows an exemplary embodiment of the frequency error correction units 30 and 50 of FIG. 4.

 The unit 30 applies the correction of the coarse frequency error calculated by the first synchronization device. The unit 50 determines the residual frequency error and corrects it.

 The input signal (from the demodulator FI 20) is transmitted to the first input of a multiplier 410.

 The frequency error signal 42 supplied by the coarse synchronization unit 40 in time and in frequency feeds a unit 420 which transforms it into a complex number by the operation cor = exp (-jk <p), (p being the error in normalized frequency and k the index of the sample to be corrected.

<Desc / Clms Page number 16>

signal cor est ensuite envoyé sur la deuxième entrée du multiplieur 410, qui effectue une première correction.

signal cor is then sent to the second input of multiplier 410, which performs a first correction.

 The elements 42, 410 and 420 form the unit 30 for correcting the error in coarse frequency.

 The signal from the multiplier 410 simultaneously feeds a unit 430 for estimating the error in fine frequency and the first input of a second multiplier 450. The output a of the unit 430 feeds a unit 440 which transforms it into a number complex by the operation cor = exp (-jka), a being the normalized frequency error and k the index of the sample to be corrected. This number is then sent to the second input of the multiplier 450, which performs the fine phase correction.

 The unit 430 performs the function of fine estimation of the frequency error and the units 440 and 450 perform the function of fine correction of the frequency error. The elements 430, 440 and 450 therefore form the unit 50 for fine estimation and frequency correction.

 The signal from the multiplier 450 is a frequency corrected signal.

 The operation of the units 30 and 50 is as follows. In both cases, once the error has been determined, the correction is made by multiplying the signal by exp (-jkp), p being the error in normalized phase and k the index of the sample to be corrected.

 The computation of the fine frequency error (unit 430) is carried out in the same way as that of the coarse frequency error (unit 350 of FIG. 5), but by applying the autocorrelation function described above to the C sequences. of the preamble. The longer these sequences, the more precision can be improved. The parameters used for this autocorrelation function are for example a delay D of 256 points and an averaging window MD of 384 points.

 One can, in the same way as previously, extract the value of the error by measuring the phase at the instant of the amplitude peak of the autocorrelation signal calculated on the C sequences, or else measure the phase at an instant t1 taking as a time reference the peak detected on the signal

<Desc / Clms Page number 17>

d'autocorrélation calculé sur les séquences A et B. L'écart théorique entre ces deux pics étant connu, il suffit d'appliquer un retard au signal 41.

autocorrelation calculated on sequences A and B. The theoretical difference between these two peaks being known, it suffices to apply a delay to signal 41.

 As before, to improve the functioning of the system and its sensitivity to a possible positioning error, the value of the frequency error is averaged over 20 points before being transmitted to the correction part.

 FIG. 8 represents an exemplary embodiment of the second synchronization device, that is to say the fine time synchronization unit 60, using an intercorrelation function to precisely determine the instant of arrival of a known signal and therefore, do a fine time synchronization of the system.

 In the particular embodiment described here, the cross-correlation function, performed by the unit 501 illustrated in FIG. 8, is applied to the sequences C of the Hiperlan 2 preamble. The theoretical value of these sequences is sent to an input of l unit 501, for example by reading a table 520 containing the value of the samples of the sequences C in the time domain.

 The input signal of the second synchronization device, that is to say the received signal corrected in frequency, passes through a sub-sampling unit 510 before attacking the second input of the intercorrelation unit 501, which is the entrance to a 540 multiplier.

 The output of table 520 passes into a conjugation unit 530 which transforms the complex numbers into their conjugates before transferring them to the second input of the multiplier 540. The output of the multiplier 540 is sent simultaneously to a first averaging unit 560 which averages the signal received on the last 64 points and to a first module calculation unit 550 which calculates the module of the complex number received and transfers it to a second averaging unit 570, which also averages the signal received on the last 64 points.

 The output of the first averaging unit 560 is sent to a first input x (i) of a divider 580, the second input y (i) of the divider

<Desc / Clms Page number 18>

580 recevant un signal de normalisation 571 issu de la seconde unité de moyennage 570.

580 receiving a normalization signal 571 from the second averaging unit 570.

 The output of the divider 580 is a complex number which constitutes the output of the cross-correlation unit 501.

 A second module calculation unit 590 then extracts the module from this signal and transfers it to a threshold unit 600, which performs a thresholding operation before transferring the signal to a maximum detection unit 610, which performs the search for the maximum of this signal and supplies as an output the signal 61 for controlling the sub-sampling and windowing unit 70 (see FIG. 4) and the validation signal 65 intended for the units involved in the final demodulation of the useful symbols.

 The operation of the second synchronization device is as follows.

 In the Hiperlan 2 standard, two symbols C of 64 samples each are emitted following the sequences A and B.

 The second synchronization device uses these symbols C to perform fine time synchronization and confirm the validity of the data received and therefore, control their demodulation.

où cross-sorrel désigne la valeur de la fonction d'intercorré ! ation, receivedC désigne la séquence C reçue et theoretical désigne la séquence C théorique. At the input of this device, the signal to be processed is still oversampled; it is therefore necessary to carry out a sub-sampling operation before being able to carry out the cross-correlation. In practice, this intercorrelation function is coupled to a subsampling function by a factor of 4 and the mathematical function performed is as follows:

where cross-sorrel denotes the value of the crossover function! ation, receivedC designates the received C sequence and theoretical designates the theoretical C sequence.

 The signal 591, result of this operation, is illustrated in FIG. 9a: two peaks appear at the end of the sequences C received.

<Desc / Clms Page number 19>

The intercorrelation function being normalized, it is possible to easily detect these peaks, for example, as described above, by applying a thresholding then a maximum detection algorithm, to determine the phase of undersampling. optimal as well as the precise position of the first useful symbol which will have to be demodulated.

 FIGS. 9a, 9b and 9c illustrate the operation of the second synchronization device in FIG. 8.

 The graph at the top of FIG. 9a represents the amplitude of the cross-correlation signal when receiving a signal including the C sequences; the vertical arrows illustrate the correspondence of the peaks obtained with the sequence of data received.

 When receiving a Hiperlan 2 type signal, two peaks appear clearly when the correlation window (of size equal to 64 points in the example described here) is located on the received sequences C.

 Thanks to the choice of the size and content of the sequences to which this operation relates, the peaks generated are clearly more marked than those generated in the first synchronization device and, consequently, the temporal precision as well as the reliability of the decision to demodulate. the data is better.

 FIG. 9b shows an example of generation of the validation signal 65 based on the detection of the first peak. In the nonlimiting example described here, a maximum detection algorithm is applied to the amplitude of the signal resulting from the cross-correlation as soon as it exceeds a threshold of 0.6. The detected peak being the first, a delay of 256 points (length of the oversampled symbol C) is then applied before activating the validation signal, which is thus in phase with the start of the first useful symbol received.

 FIG. 9c illustrates the generation of the sub-sampling and windowing command 61 for the FFT.

 The curve at the top of FIG. 9c represents the intercorrelation of the preamble C.

 The curve below the cross-correlation represents the frequency corrected samples.

<Desc / Clms Page number 20>

In the nonlimiting example described here, the command 61 for sub-sampling and windowing consists of a series of pulses of width 1 point separated by 3 points so as to perform the sub-sampling by 4 of the signal to be demodulated. This series of pulses is shown at the bottom of FIG. 9c. This sub-sampling thus reduces the size of the OFDM symbols from 256 points to 64 points.

 Thus, it follows from the above description that, before the arrival of a new frame, the receiver according to the present invention, illustrated in FIG. 4, is in the standby position: only, the conversion unit 10 analog / digital, the IF demodulation unit 20 and the coarse time and frequency synchronization unit 40 are activated. The first synchronization device is then active.

 Then, as soon as the unit 40 detects the arrival of a new frame, as described in relation to FIGS. 5 and 6, the receiver goes into a second phase, in which, by validating the signal 41 to start the process of demodulation, the coarse frequency error correction unit 30, the fine frequency estimation and correction unit 50 and the fine time synchronization unit 60 are further activated. The second synchronization device is then active.

 Finally, as soon as this second synchronization device succeeds in fine time synchronization, as described in relation to FIGS. 7 and 8, the validity of the data received is confirmed and all of the constituent units of the receiver are validated by the validation signal 65 .

 FIG. 10 schematically illustrates the constitution of a network station or computer reception station, in the form of a block diagram.

 This station includes a keyboard 1011, a screen 1009, an external information recipient 1010, a wireless receiver 1006, jointly connected to an input / output port 1003 of a processing card 1001.

 The processing card 1001 comprises, linked together by an address and data bus 1002: - a central processing unit 1000;

<Desc / Clms Page number 21>

- une mémoire vive RAM 1004 ; - une mémoire morte ROM 1005 ; et - le port d'entrées/sorties 1003.

- a random access memory RAM 1004; - ROM 1005 read only memory; and - the input / output port 1003.

 Each of the elements illustrated in FIG. 10 is well known to those skilled in the art of microcomputers and transmission systems and, more generally, information processing systems. These common elements are therefore not described here.

 It is further observed that the word "register" used in the description designates, in each of the memories 1004 and 1005, both a low-capacity memory area (some binary data) and a high-capacity memory area (allowing store an entire program).

dans la description, les mêmes noms que les données dont ils conservent les valeurs. La mémoire vive 1004 comporte notamment : - un registre"données reçues", dans lequel sont conservées les données binaires reçues, dans leur ordre d'arrivée sur le bus 1002 en provenance du canal de transmission. The random access memory 1004 stores data, variables and intermediate processing results, in memory registers carrying,

in the description, the same names as the data whose values they keep. The random access memory 1004 includes in particular: a “received data” register, in which the binary data received are stored, in their order of arrival on the bus 1002 coming from the transmission channel.

 The ROM 1005 is adapted to keep the operating program of the central processing unit 1000, in a "Program" register.

 The central processing unit 1000 is suitable for implementing a reception method in accordance with the invention, as illustrated by the flow diagram of FIG. 3.

 As shown in FIG. 11, a network according to the invention consists of at least one station called base station SB designated by the reference 64, and several peripheral stations called mobile terminals SPi, i = 1, ..., M, where M is an integer greater than or equal to 1, respectively designated by the references 661, 662, ..., 66M. The peripheral stations 661, 662, .... 66M are distant from the base station SB, each linked by a radio link with the base station SB and capable of moving relative to the latter.

<Desc / Clms Page number 22>

 The base station 64 may include means suitable for implementing a reception method according to the invention or may comprise a reception device according to the invention and at least one of the mobile terminals 661 may include means suitable for putting implement a reception method according to the invention or include a reception device according to the invention.

Claims (22)

 1. Method for receiving signals in the form of frames comprising a preamble and data, said preamble comprising at least two parts, characterized in that it comprises steps according to which: - a first synchronization operation (500) is carried out, consisting in detecting the arrival of a new frame from the first part of the preamble, - a correction operation (502) is carried out, consisting in correcting information contained in the second part of the preamble of said new frame, - we performs a second synchronization attempt (504), by intercorrelation between the previously corrected information and a predetermined signal, and - depending on the result of the second synchronization, a demodulation operation (506) is carried out or not, consisting in demodulating the set data contained in said frame.  2. Reception method according to claim 1, characterized in that, during the first synchronization operation (500), the error in coarse frequency is also determined.  3. Reception method according to claim 1 or 2, characterized in that, during the correction operation (502), the error in fine frequency is determined.  4. Method according to claim 1,2 or 3, characterized in that, during the second synchronization attempt (504), an attempt is made to perform fine time synchronization.  5. Method according to any one of the preceding claims, characterized in that the predetermined signal is a validation signal (65) which controls the demodulation operation (506). <Desc / Clms Page number 24>

6. Method according to any one of the preceding claims, characterized in that the demodulation operation (506) implements demodulation of the OFDM type.  7. Method according to any one of the preceding claims, characterized in that the signals are in the form of frames in accordance with the Hiperlan 2 standard, the first part of the preamble consisting of sequences A and B and the second part of the preamble consisting of sequence C.  8. Device for receiving signals in the form of frames comprising a preamble and data, said preamble comprising at least two parts, characterized in that it comprises: - first synchronization means (40) for detecting the arrival of 'a new frame from the first part of the preamble, - correction means (30, 50), for correcting information contained in the second part of the preamble of said new frame, - second synchronization means (60), to make a second synchronization attempt by intercorrelation between the previously corrected information and a predetermined signal, and - demodulation means (70,80, 90,100, 110,120, 130) for, depending on the result of the second synchronization, demodulating or not all the data contained in said frame.  9. Device according to claim 8, characterized in that the first synchronization means (40) further determine the error in coarse frequency.  10. Device according to claim 8 or 9, characterized in that the correction means (30.50) determine the error in fine frequency.  11. Device according to claim 8,9 or 10, characterized in that the second synchronization means (60) attempt to effect fine time synchronization.  12. Device according to any one of claims 8 to 11, characterized in that the predetermined signal is a validation signal (65) which controls the means (70,80, 90,100, 110,120, 130) for demodulation. <Desc / Clms Page number 25>

13. Device according to any one of claims 8 to 12, characterized in that the demodulation means (70, 80, 90, 100, 110, 120, 130) implement demodulation of the OFDM type.  14. Device according to any one of claims 8 to 13, characterized in that the signals are in the form of frames in accordance with the Hiperlan 2 standard, the first part of the preamble consisting of sequences A and B and the second part of the preamble being made up of the sequence C.  15. Apparatus for processing digital signals, characterized in that it comprises means suitable for implementing a reception method according to any one of claims 1 to 7.  16. Apparatus for processing digital signals, characterized in that it comprises a reception device according to any one of claims 8 to 14.  17. Telecommunications network, characterized in that it comprises means adapted to implement a reception method according to any one of claims 1 to 7.  18. Telecommunications network, characterized in that it comprises a reception device according to any one of claims 8 to 14.  19. Mobile station in a telecommunications network, characterized in that it comprises means suitable for implementing a reception method according to any one of claims 1 to 7.  20. Mobile station in a telecommunications network, characterized in that it comprises a reception device according to any one of claims 8 to 14.  21. Base station in a telecommunications network, characterized in that it comprises means suitable for implementing a reception method according to any one of claims 1 to 7.  22. Base station in a telecommunications network, characterized in that it comprises a reception device according to any one of claims 8 to 14.