WO2020239551A1

WO2020239551A1 - Method for determining at least one precoder for a transmitting device of a wireless communication system

Info

Publication number: WO2020239551A1
Application number: PCT/EP2020/063990
Authority: WO
Inventors: Dinh Thuy Phan Huy; Raphaël Visoz
Original assignee: Orange
Priority date: 2019-05-24
Filing date: 2020-05-19
Publication date: 2020-12-03
Also published as: EP3977629A1; FR3096528A1

Abstract

The invention relates to a method for determining at least one precoder for a transmitting device (2) of a wireless communication system (1) equipped with a plurality of transmitting antennas, the transmitting device (2) being associated with: - a first precoder P_1 for optimizing a transmission of data in the form of beams, - a regulatory constraint relating to an electromagnetic quantity outside a defined regulatory zone. Moreover, said method comprises: - a step (E10) of obtaining a set G comprising a plurality of precoders respectively associated with mutually distinct directions, - a step (E20) of determining a precoder P_2 corresponding to a representation of the precoder P_1 as a function of the set G, - a step (E30) of identifying, among the directions associated with the precoders of the set G, directions referred to as "overshooting directions", under which the regulatory constraint is not respected when the transmitting device is associated with said precoder P_2, - a step (E40) of determining a precoder P_3 as a function of the precoder P_2, so that the regulatory constraint is respected according to said overshooting directions when the transmitting device is associated with the precoder P_3.

Description

Title of the invention: Method for determining at least one precoder for a transmitter device of a wireless communication system

Prior art

The present invention belongs to the general field of

telecommunications and in particular in the field of wireless communications implemented on radio-type networks such as mobile networks (eg 3G, 4G, 5G, etc.), WiFI (Wireless Fidelity), etc.

[0002] It relates more particularly to a method for determining at least one precoder for a transmitter device of a wireless communication system fitted with a plurality of antennas. It also relates to a method for transmitting at least one data stream by such a sending device, the data stream being intended to be received by a receiving device comprising at least one receiving antenna. The invention finds an application

particularly advantageous, although in no way limiting, in the case of a communication system comprising a transmitter device and a receiver device each equipped with a plurality of antennas, also called MIMO (Multiple Input Multiple Output) communication system.

For a transmitter device belonging to a wireless communication system and comprising a plurality of transmitting antennas, it is known to use a precoder making it possible to transmit simultaneously, to a receiver device belonging to the same system , one or more data streams via the various transmitting antennas of the transmitting device. Such a precoder is based on the knowledge by the transmitting device of the propagation channel (or transmission channel) which separates it from the receiving device. It allows the transmitter device to deliver data to the receiver device with high spectral efficiency thanks to the formation of beams (also called “beamforming” in the English literature) transporting the data streams.

[0004] Each precoder is further designed so that the output of

data in the form of beams is optimized with regard to a criterion determined operating mode, such as for example a data transmission quality of service criterion (maximization of the throughput or of the power received at the level of the receiving device, absence of interference between time symbols at the level of the receiving device, etc.) , or a network spectral efficiency criterion taking into account interference generated on other users, or a network energy efficiency criterion, etc. Known examples of precoders are for example those of the zero forcing type (or ZF for “Zero Forcing” in English), those of the maximum ratio transmission type (or MRT for “Maximum Ratio Transmission” in English), those carrying out training. clean beams (or “eigenbeamforming” in English), those carrying out a formation of beams by decomposition into eigenvalues (or “SVD” for “Singular Value Decomposition” in English), and so on.

[0005] Furthermore, the current design of a precoder, although aimed at optimizing a determined criterion, nevertheless remains carried out independently of any regulations relating to the limitation of the exposure of persons (public or workers ) electromagnetic fields, and capable of being applied where the emitting device is intended to operate.

[0006] Such a regulation initially fixes field levels (electric, magnetic) which must not be exceeded depending on the

transmission frequency of the sending device. These field levels constitute exposure limits which can be formulated in an equivalent manner in terms of the power radiated by the emitting device and received by a person.

[0007] For example in France, and with regard to workers, such

regulations are included in the Labor Code in Articles R. 4453-1 to R. 4453-34 (as of April 2019). Regarding the public, the relevant regulations are provided by Decree No. 2002-775 following the transposition of European recommendation 1999/519 / CE.

[0008] Secondly, an exposure limit is associated with a regulatory distance from the transmitting device, typically imposed by geographical implementation constraints of the transmitting device and defined by an entity in charge of the management of the transmission system. communication (for example a telephone operator). In this way, it is possible to define a regulatory zone around the emitting device beyond which an exposure limit must not be exceeded.

[0009] The fact that a precoder is designed without taking into account the regulations to which the transmitter device will be subject is problematic. Indeed, due to the optimization of said operating criterion, it may happen that a limit

exposure is not respected in one or more beam directions outside the regulatory zone associated with the emitting device.

[0010] FIG. 1A represents schematically, in a sectional view

horizontal and in accordance with the state of the art, a data stream transmission mode between a transmitter device 2 and a receiver device 3, this transmission being optimized with regard to an operating criterion determined using a precoder.

[0011] FIG. 1A corresponds more particularly to a superposition of two representations respectively associated with distinct physical quantities, namely:

- a first representation relating to the geographical implementation of the transmitter device 2 and of the receiver device 3. According to this first

representation, FIG. 1A also illustrates diffuser elements D_i, for i varying from 2 to 6, of a type known per se, and positioned in the environment of the transmitter 2 and receiver 3 devices in order to ensure the convergence of the data transmitted towards the receiver device 3. The relative positions of the transmitter 2 and receiver 3 devices, as well as of the diffusers D_i, according to this first representation, can therefore be identified by geographical coordinates, for example in a Cartesian coordinate system;

a second representation relating to a radiation diagram of the emitting device 2. According to this second representation, FIG. 1A illustrates the formation of six beams F_i, for i varying from 1 to 6, generated by the emitting device 2 by means of the precoder. A circle C_P is also represented, this circle C_P surrounding the emitting device 2. As explained in more detail later, this circle C_P is here representative of a maximum transmission power of the “EIRP” type (acronym of the expression “Power. Isotropic Radiated Equivalent ”, in English“ EIRP ”for“ Equivalent Isotropie Radiated Power ”) which should not be exceeded in order to comply with a regulatory constraint determined as a function of said regulatory distance (and therefore ultimately of said regulatory zone) as well as an electric field exposure limit. More specifically, said circle C_P corresponds to a maximum transmission power EIRP not to be exceeded as a function of an angular direction considered with respect to the emitting device 2. It should be noted that in practice, the shape taken by this circle C_P defined by said maximum transmission power EIRP is arbitrary. It is nevertheless represented here in the form of a circle to simplify FIG. 1 A. In addition, it is important to note that said circle C_P is not an effective representation of said regulatory zone, the latter also being able to take any form whatsoever. , but is nevertheless linked to it via in particular said exposure limit

(in other words, the regulatory area is not shown in Figure 1A).

[0012] Moreover, according to this second representation of FIG. 1A, each

beam takes a substantially oblong shape delimited by a border also representative of an EIRP type emission power according to the direction associated with this beam (direction also called azimuth).

[0013] Therefore, and as can be seen in FIG. 1A, the implementation of the precoder implies that the regulatory constraint is not respected along the direction of the beam F_1. Indeed, a portion of the beam F_1 extends outside the circle C_P, which means that the maximum transmission power associated with the regulatory constraint is exceeded outside the regulatory zone.

In order to get around this problem, it has been proposed to reduce the

electrical power applied to the transmitting device, so that the effective transmit power of the latter is less than the maximum transmit power for which it was designed.

[0015] The fact of thus reducing the transmission power makes it possible to make the operation of the transmitter device compliant with respect to the exposure limits imposed outside the associated regulatory zone. However, such a procedure cannot be considered satisfactory insofar as it affects the power emitted in all directions downwards. beams. This results in a strong degradation of the operating conditions of the communication system (reduction of the bit rate or of the power received at the level of the receiving device, appearance of interference between time symbols received, etc.).

[0016] This disadvantageous situation is illustrated, by way of example in no way

limitative, in Figure 1B which corresponds to Figure 1A after a power reduction in accordance with the prior art has been applied to the transmitter device 2.

As illustrated in Figure 1B, the beam F_1 is now contained in the circle C_P, so that the regulatory constraint is now respected by the transmitter device 2. It can nevertheless be observed that the other beams F_2 to F_6 saw their respective sizes significantly decrease, so as to be now very far from the maximum transmission power associated with the circle C_P. These reductions greatly degrade the transmission of data from the sending device 2 to the receiving device 3.

Disclosure of the invention

[0018] The present invention aims to remedy all or part of the

disadvantages of the prior art, in particular those exposed above, by proposing a solution which makes it possible to obtain a precoder for a transmitter device of a wireless communication system, so that, on the one hand, this transmitter device complies with regulations relating to electromagnetic exposure limits outside a defined zone, and on the other hand, the degradation of an operating criterion of said system is much more limited than that of the solutions of the prior art.

[0019] To this end, and according to a first aspect, the invention relates to a method for determining at least one precoder for a transmitter device equipped with a plurality of transmitting antennas, the transmitter device being associated with:

- a first precoder P_1 making it possible to optimize, with regard to a determined operating criterion, a transmission of data streams in the form of beams,

- a regulatory constraint corresponding to non-exceeding, outside a regulatory zone defined around said emitting device, a threshold value relating to an electromagnetic quantity.

In addition, said method of determination comprises:

- a step of obtaining a set G comprising a plurality of precoders respectively associated with directions distinct from each other,

- a step of determining a precoder P_2 corresponding to a

representation of the precoder P_1 as a function of the set G,

- a step of identifying, among the directions associated with the precoders of the set G, of directions known as "overshoot directions" according to which the regulatory constraint is not respected when the sending device is associated with said precoder P_2,

- a step of determining a P_3 precoder as a function of the P_2 precoder, so that the regulatory constraint is met along said overrun directions when the sending device is associated with the P_3 precoder.

[0020] Thus, the method is based on obtaining such a set G with the help of which the precoder P_1 is then decomposed. It should in fact be understood that simply consulting the P_1 precoder does not, in general, easily determine the directions associated with the emission beams resulting from the use of said P_1 precoder. In other words, the precoder P_1 is structured above all so that its application to data stream symbols allows the generation of beams achieving an optimization of the operating criterion. However, it is not possible, a priori, to identify directly from the single precoder P_1, which directions of space are involved in this optimization. Thus, said set G is obtained with the objective of facilitating the identification of these directions of space used by the precoder P_1.

The set G is then used to identify the directions from among those of G which are requested by the precoder P_1, thus making it possible to overcome the difficulties of reading said precoder P_1. The precoder P_2 thus obtained consists of a representation of the precoder P_1 as a function of the set G.

[0022] The invention therefore makes it possible, on the basis of the precoder P_2 and of the set G, to precisely identify the directions of the overrun, so as to obtain a very precise description of the behavior on transmission of the transmitter device at the with regard to regulatory constraint. Thus, unlike what is done in State of the art, here we avoid assuming that all the directions contributing to the formation of the beams participate in exceeding the threshold value. In other words, one distinguishes, among said directions of the set G, those for which it is useful to envisage a reduction in the transmission power of the emitting device.

[0023] The P_3 precoder has the effect of correcting the effects of the P_2 precoder, and therefore also of the P_1 precoder, in that it targets a reduction in radiated power specifically in the direction of the excess that has been identified. The invention therefore makes it possible not only to comply with the regulatory constraint, but also to very effectively limit the degradation of the operating criterion in comparison with the solutions of the prior art. It is in fact understood that by targeting a reduction in power along only the direction of the overrun, it is thus possible to avoid abruptly reducing the overall transmission power of the emitting device. In other words, we avoid reducing the transmit power in directions that are not overshoot directions.

[0024] In particular embodiments, the determination method may further include one or more of the following characteristics, taken in isolation or in any technically possible combination.

[0025] In particular modes of implementation, the identification step

comprises, for each direction considered of the set G:

- an estimate of the power radiated by the emitting device and emitted beyond said zone in said direction,

a comparison of said power with a threshold value, so as to determine whether said direction corresponds to a direction of overshoot.

In particular embodiments, when the set G admits a matrix representation, the column vectors of this matrix are respectively associated with said distinct directions between them, and are also mutually orthogonal for the inducing Hermitian scalar product the Frobenius matrix norm.

In particular embodiments, the step of determining the precoder P_3 comprises, for each direction of overshoot considered: an estimate of the power radiated by the transmitting device associated with the precoder P_2 and received beyond said regulatory zone along said direction of overshoot,

- an update of the precoder P_2, so that the power radiated by the transmitter device associated with the updated precoder P_2 and received beyond said regulatory zone in said direction of overshoot is less than the transmitted power received before said update up to date,

the estimation and the update being iterated for the direction of overshoot considered until the regulatory constraint is met along said direction of overshoot,

the P_3 precoder corresponding to the last update of the P_2 precoder, once all the overshoot directions have been considered.

In a first preferred embodiment:,

- a direction associated with a precoder of the set G and distinct from the overshoot directions is called "non-overshoot direction",

the method comprising, for at least one direction of non-overshoot, a step of updating the precoder P_3, so that:

- the power radiated by the transmitter device associated with the updated precoder P_3 and received in said regulatory zone in said non-exceeding direction is greater than the corresponding received power before said update,

- the regulatory constraint is respected according to said direction of non-exceeding.

The fact of updating the precoder P_3 according to this first preferred mode

advantageously allows to redistribute, in a direction of no

overshoot, a portion of the antenna's operating power that was released during the step of determining the P_3 precoder to meet the regulatory constraint following the overshoot directions. By proceeding in this way, the operating criterion associated with the transmitter device is therefore improved, while respecting the regulatory constraint.

[0030] In other words, the transmission capacities are advantageously exploited by

power of the emitting device in the direction of non-overshoot considered. We therefore benefit from the advantages of updating the precoder P_2 (compliance with the regulatory constraint), while ingeniously compensating for the loss of quality of the transmission of the data streams inherent in the reduction in transmission power necessary to satisfy the regulatory constraint.

In particular modes of implementation, the step of updating the precoder P_3 is iterated for each of the directions considered of non-overshoot, an update step associated with an iteration being carried out from the precoder P_3 obtained during the previous iteration.

[0032] In a second preferred embodiment:

- each beam obtained thanks to the precoder P_1 is associated with a direction, called "beam direction",

the method comprising steps of updating the precoder P_3

mutually independent and associated respectively with the beam directions, so as to obtain a plurality of updated P_3 precoders and that, for a given beam direction:

- the power radiated by the emitting device associated with the precoder P_3 updated for said beam direction and received in said zone

regulatory in said beam direction is greater than the corresponding received power before said update.

[0033] Thus, the invention according to this second preferred embodiment makes it possible to obtain a plurality of updated P_3 precoders. Each of these updates retains most of the advantages of the pre-update P_3 precoder, namely that the regulatory constraint remains respected according to the beam directions distinct from the beam direction considered during the update. However, each of these updates also offers the possibility of exploiting the transmitting capabilities of the transmitting device to obtain an emission beam whereby the regulatory constraint is exceeded. In other words, during an update of the precoder P_3, the invention here offers the possibility of exceeding the threshold value associated with the regulatory constraint for a beam direction.

Such an implementation is advantageous because it allows, when it is planned to transmit by means of the plurality of updated P_3 precoders, in turn and during determined respective times, to improve the criterion d 'exploitation associated with the sending device, while respecting the regulatory constraint on average over time.

In particular embodiments, said method is intended to be implemented by a device, called an "ancillary device", distinct from said transmitter device, said method further comprising a step of transmitting the ancillary device to the device. sending device:

- of the P_3 precoder determined according to the invention; or

- the updated precoder P_3 determined according to said first preferred mode of implementation, or

- updated P_3 precoders determined according to said second preferred embodiment.

According to a second aspect, the invention relates to a method of transmitting at least one data stream by a transmitter device equipped with a plurality of transmitting antennas, the data stream being intended to be received. by a receiving device equipped with at least one receiving antenna, the transmitting device being associated with:

- a precoder P_3 determined beforehand according to the invention; or

an updated precoder P_3 determined beforehand according to said first preferred embodiment.

According to a third aspect, the invention relates to a method of transmitting at least one data stream by a transmitter device equipped with a plurality of transmission antennas, the data stream being transmitted in directions beams and intended to be received by a receiving device equipped with at least one receiving antenna, said transmitting device being associated with:

- a regulatory constraint corresponding to not exceeding, beyond a regulatory zone defined around said emitting device, a threshold value relating to an electromagnetic quantity,

- updated P_3 precoders previously determined according to said second preferred embodiment,

- a regulatory issuance period,

said method comprising successive transmission steps associated respectively with said updated precoders P_3, each of said transmission steps being carried out for a determined fraction of the duration regulatory emission, so that said regulatory constraint is observed on average during said regulatory duration in any beam direction.

The fact of thus successively using the various updated precoders P_3 makes it possible to limit the violation of the regulatory constraint to a determined period of time, so that the regulatory constraint is satisfied on average over time, with regard to the regulatory duration associated with the issuing device. In other words, the formation of beams by the emitting device is thus statistically limited in time.

[0039] In addition, the use of said updated P_3 precoders is advantageous because it makes it possible to exploit the maximum transmission power of the transmitter device for each of the beams. Thus, the power given up during the

determination of the precoder P_3 is here redistributed in turn for the emission of each of the beams for a determined duration. In the end, the operating criterion is improved on average over time, compared to the sole use of the non-updated P_3 precoder, while respecting the regulatory constraint on average over time.

[0040] According to a fourth aspect, the invention relates to a computer program comprising instructions for implementing a method of

determination or emission according to the invention when said program is executed by a processor.

[0041] According to a fifth aspect, the invention relates to a recording medium readable by a computer on which is recorded a computer program according to the invention.

[0042] According to a sixth aspect, the invention relates to a device for determining at least one precoder for a transmitter device equipped with a plurality of transmitting antennas, the transmitter device being associated with:

- a first precoder P_1 making it possible to optimize, with regard to a determined operating criterion, a transmission of data streams in the form of beams to a receiving device, said first precoder P_1 being known to the determination device,

- a regulatory constraint corresponding to non-exceeding, beyond a regulatory zone defined around said emitting device, a threshold value relating to an electromagnetic quantity.

In addition, said determination device comprises:

- an obtaining module, configured to obtain an assembly G comprising a plurality of precoders respectively associated with distinct directions between them,

- a first determination module, configured to determine a precoder P_2 corresponding to a projection of the precoder P_1 on the set G,

an identification module, configured to identify, from among the directions associated with the precoders of the set G, directions called “overshoot directions” according to which the regulatory constraint is not respected when the sending device is associated with said precoder P_2 ,

a second determination module, configured to determine a precoder P_3 as a function of the precoder P_2, so that the regulatory constraint is complied with along said overrun directions when the sending device is associated with the precoder P_3.

[0043] In particular embodiments, the determination device may further include one or more of the following characteristics, taken in isolation or in any technically possible combination.

In particular embodiments, said determining device corresponds to:

- to the sending device; or

to a device, called an “ancillary device”, distinct from said transmitter device, said ancillary device further comprising communication means configured to transmit at least one precoder to the transmitter device.

In particular embodiments, said annex device corresponds to the receiving device equipped with at least one receiving antenna.

[0046] According to a seventh aspect, the invention relates to a system for

wireless communication comprising a sending device equipped with a plurality of transmitting antenna and a receiving device equipped with at least one receiving antenna, said transmitting device being associated with:

a precoder P_3 determined beforehand according to the invention; or an updated precoder P_3 determined beforehand according to said first preferred embodiment; or

- P_3 updated precoders determined beforehand according to said second preferred mode of implementation.

Brief description of the drawings

Other characteristics and advantages of the present invention will emerge from the description given below, with reference to the accompanying drawings which illustrate an exemplary embodiment thereof without any limiting nature. In the figures:

[Fig. 1A] FIG. 1A schematically represents, in a horizontal sectional view and in accordance with the state of the art, a mode of transmission of data streams between a transmitter device and a receiver device, this transmission being optimized with regard to an operating criterion determined by means of a precoder;

[Fig. 1 B] Figure 1 B corresponds to Figure 1 A after a power reduction in accordance with the prior art has been applied to the transmitter device;

[Fig. 2] FIG. 2 schematically represents an exemplary embodiment, according to the invention, of a wireless communication system comprising a transmitter device and a receiver device;

[Fig. 3] FIG. 3 shows schematically in the form of a flowchart an embodiment, according to the invention, of a method for determining at least one precoder for the sending device;

[Fig. 4] FIG. 4 schematically represents the effects of the application to the emitting device of FIG. 1A of a precoder P_3 according to the invention;

[Fig. 5] FIG. 5 schematically represents a first preferred mode of implementation of the determination method of FIG. 3;

[Fig. 6] FIG. 6 schematically represents the effects of the application to the emitting device of FIG. 1A of an updated P_3 precoder obtained according to said first preferred mode of implementation;

[Fig. 7] FIG. 7 schematically represents a second preferred embodiment of the determination method of FIG. 3; [Fig. 8A] FIG. 8A schematically represents the effects of the application to the emitting device of FIG. 1A of a first updated precoder P_3 obtained according to said second preferred embodiment, said first precoder P3 being associated with a beam F_1;

[Fig. 8B] FIG. 8B diagrammatically represents the effects of the application to the emitting device of FIG. 8A of a second updated precoder P_3 obtained according to said second preferred embodiment, said second precoder P3 being associated with a beam F_2 ;

[Fig. 8C] FIG. 8C schematically represents the effects of the application to the emitting device of FIG. 8B of a third updated precoder P_3 obtained according to said second preferred embodiment, said third precoder P3 being associated with a beam F_3 ;

[Fig. 8D] FIG. 8D schematically represents the effects of the application to the emitting device of FIG. 8C of an updated fourth precoder P_3 obtained according to said second preferred mode of implementation, said fourth precoder P3 being associated with a beam F_4 ;

[Fig. 8E] FIG. 8E schematically represents the effects of the application to the emitting device of FIG. 8D of an updated fifth precoder P_3 obtained according to said second preferred mode of implementation, said fifth precoder P3 being associated with a beam F_5 ;

[Fig. 8F] FIG. 8F schematically represents the effects of the application to the emitting device of FIG. 8E of an updated sixth precoder P_3 obtained according to said second preferred mode of implementation, said sixth precoder P3 being associated with a beam F_6 ;

Description of embodiments

[0048] The present invention finds its place in the field of data flow exchanges in a wireless telecommunications network.

[0049] Figure 2 shows schematically an embodiment of a

wireless communication system 1 according to the invention. As illustrated by Figure 1, the communication system 1 is a system comprising:

- a transmitter device 2 equipped with M transmit antennas TX1, TX2, ..., TXM, M designating an integer greater than 1; and

- a receiving device 3 equipped with a plurality of receiving antennas RX1,,

RXN, N denoting an integer greater than 1.

In this way, the communication system 1 forms a multi-antenna system or MIMO (acronym for the English expression "Multiple Input Multiple Output).

However, it should be noted that the fact of considering a plurality of receiving antennas at the level of the receiving device 3 in no way constitutes a limitation of the invention. Thus, nothing excludes considering a receiving device 3 equipped with a single receiving antenna.

The transmitter device 2 and the receiver device 3 are configured here to communicate with each other via the wireless telecommunications network. However, no limitation is attached to the form taken by the transmitter 2 and receiver devices 3. For example, whatever device is considered, it may be a base station or even a terminal. Preferably, the transmitter device 2 is a base station, and the receiver device 3 is a terminal.

The remainder of the description relates more specifically to a network of

mobile telecommunications offering a mode of communication according to a TDD mode (acronym for the English expression "Time Division Duplex").

[0054] Nothing, however, excludes considering other types of

wireless telecommunications, fixed or mobile, operating in TDD mode, but also in FDD mode (acronym of the English expression "Frequency Division Duplex"), those skilled in the art being able to make the necessary adjustments to adapt the invention as described below.

The transmitter device 2 and the receiver device 3 are separated by a propagation channel 4. It is assumed here that the communication system 1 uses during communications between the transmitter device 2 and the receiver device 3, a waveform multi-carrier of the OFDM type (for “Orthogonal Frequency Division Multiplexing”). The use of such a waveform has for consequence that for a given (sub) carrier, the propagation channel 4 is flat in frequency (ie all the frequencies are attenuated in a similar way by the propagation channel 4) and is written in the form of a complex matrix noted H, of dimension equal to the product of the number of receiving antennas considered (N in this example) by the number of transmitting antennas considered (M in this exemplary embodiment). Because of this property, in the remainder of the description, the effect of the invention is described only with reference to a single carrier, the invention being applied in the same way to the other carriers on which it is based. the waveform used by the communication system 1.

The transmitter device 2 is configured to apply in transmission, on the

data that it sends to the receiving device 3, a precoding which is based on knowledge, at all times, by the sending device 2 of the propagation channel 4 which separates it from the receiving device 3 (ie of the coefficients of the matrix H) .

As indicated above, in the example considered here, we are in the context of communications in TDD mode. In such a context, it is generally accepted that there is reciprocity between the channel going in the sender-to-receiver direction and the channel going in the receiver-to-sender direction. In other words, if we denote He and Hr the matrices of the channels estimated respectively by the sending device 2 and by the receiving device 3, it is customary to consider in a TDD context that these two estimates are substantially equal, the matrices He and Hr then being substantially equivalent. If, moreover, the estimate made is of good quality, we also obtain that the H-channel matrix is substantially equivalent to the He and Hr matrices.

[0058] The estimate of a propagation channel in a wireless network is a

conventional operation known to those skilled in the art, and therefore not detailed here further. It can in particular be based on the sending of sequences comprising pilot symbols (also called "pilot sequences" hereinafter) on the propagation channel that it is sought to estimate.

In order to precode the data intended to be transmitted, the transmitter device 2 is associated with a precoder P_1, hereinafter also referred to as "first precoder P_1". This first precoder P_1, for example calculated by the sending device 2 on the basis of the knowledge of the Hr channel obtained by TDD reciprocity, has been designed so as to optimize, with regard to a determined operating criterion, a transmission of data streams in the form of beams to the receiver device 3. The number of independent data streams that can be sent simultaneously by the sending device 2 to the receiving device 3 is denoted K, and is conventionally less than or equal to the smallest number among the integers N and M.

Said operating criterion corresponds for example to a criterion of quality of service of the data transmission (maximization of the throughput or of the power received at the level of the transmitter device 3, absence of interference between temporal symbols at the level of the device. transmitter 3, etc.). Known examples of precoders capable of optimizing a quality of service criterion are for example those of the zero forcing type (or ZF for “Zero Forcing”), those of the maximum ratio transmission type (or MRT for “Maximum Ratio Transmission”). "In English), those performing clean beam formation (or"

eigenbeamforming "in English), those carrying out a formation of beams by decomposition in eigenvalues (or" SVD "for" Singular Value

Decomposition ”in English), etc.

For the remainder of the description, it is considered that the precoder P_1 is of the MRT type, so as to optimize a quality of service criterion corresponding to a maximization of the throughput received at the level of the receiver device 3. The choice of a Such a precoder of course only constitutes a variant implementation of the invention, any other precoder based on knowledge of the transmission channel and making it possible to simultaneously transport several data streams being able to be considered.

As the communication system 1 uses an OFDM type multiplexing technique, the first precoder P_1 admits a representation in the form of a matrix with complex coefficients and of dimension equal to M x K (ie M multiplied by K). The precoder P_1 also being of the MRT type here, the matrix which is associated with it corresponds to ^l H, that is to say the conjugate transpose of the channel matrix H (the index "t" applied to a matrix corresponds thus to the operation of transposition-conjugation). The representations of the precoding matrices in the case where the precoder differs from an MRT precoder are well known to those skilled in the art. These details are therefore not repeated here because they go outside the scope of the invention.

Furthermore, nothing excludes either considering an operating criterion which differs from a quality of service criterion, such as, for example, a criterion of spectral efficiency of the network taking into account the interference generated on d other users, or a criterion of energy efficiency of the network, or even an operating criterion combining several criteria between them, these criteria possibly being of the quality of service type or not.

It should therefore be understood that the operating criterion is a criterion set by an entity that owns the communication system, such as for example a company wishing to offer communication services capable of satisfying customers as part of an optimization of the quality of service. Therefore, said operating criterion differs from any regulatory framework with which the communication system 1 is likely to face due to national legislation applicable to it.

In the present embodiment, the transmitter device 2 is associated with a regulatory constraint corresponding to the non-exceeding, outside a zone defined around said transmitter device, called "regulatory zone", of a threshold value relating to an electromagnetic quantity. As mentioned previously, such a constraint results from a regulatory framework linked to the legislation in force where the communication system 1 is established.

For the remainder of the description, it is considered, in no way limiting, that the communication system 1 is located in France. Therefore, this system 1 is subject to a regulatory framework aimed at defining limits for the exposure of the public to electromagnetic field emissions, as specified in Decree No. 2002-775. More particularly, it is considered here that the radio signals generated by the transmitter device 2 have a frequency between 2 GHz and 300 GHz. Consequently, the regulation indicates in its appendix 2.2 (table A) that the threshold value of electric field E not to be exceeded, for such a frequency range, is equal to 61 V / m (volt per meter). Furthermore, for the remainder of the description, an emission scenario corresponding to a rural environment (Rma scenario, acronym of the English expression "Rural Macrocell") is also considered for which the sending device is implemented in height, namely here a height of 35 meters. In addition, a regulatory distance D is imposed here by geographical implementation constraints of the transmitter device 2 (height of the transmitter device 2, as well as typically an access zone prohibited to the public around said transmitter device 2), and defined for example by an entity in charge of the management of the communication system 1 (for example a telephone operator). This distance D corresponds to a distance counted from the sending device 2.

According to these assumptions, it is possible to estimate a power received at the distance D from the transmitter device 2. Correlatively, it is possible to estimate a maximum transmission power of EIRP type associated with the transmitter device 2 corresponding to said said device. power received when the threshold value of 61 V / m is reached at distance D. This estimate, under the assumption of propagation in free space, is obtained using the following formula:

E = sqrt (30 * EIRP) / D

where the unit of E is the volt per meter, the unit of EIRP is here the Watt (the conversion between Watts and dBm being known to those skilled in the art), and where the unit of D is the meter. Thus, in this exemplary embodiment, the maximum transmission power is calculated to be substantially equal to 63 kW (ie 78 dBm).

Said regulatory zone is then in correspondence with said maximum transmission power EIRP, as has already been described above in the case of Figures 1 A and 1 B (in Figures 1A and 1 B, said maximum power EIRP is represented as a circle C_P).

[0071] It should be noted that the previous example concerning France was

given for illustrative purposes only. Thus, no limitation is attached to the country that can be considered, the skilled person being able to access the ad hoc regulatory framework.

Nothing excludes either considering other frequency ranges for the signals transmitted by the transmitter device 2, so as to obtain a threshold value for the electric field which is different from 61 V / m. In general, no limitation is attached to the type of communication network used, as was mentioned above, or to the frequency range considered. In addition, other emission scenarios can also be considered, such as for example Umi or Uma scenarios (respective acronyms of the English expressions “Urban Microcell”, “Urban Macrocell”).

In addition, the electromagnetic quantity considered in the remainder of the

description for the regulatory constraint concerns the maximum transmission power EIRP, the calculation of which is presented above according to the threshold value of the electric field and the regulatory distance. However, nothing excludes considering another electromagnetic quantity, such as for example said threshold value of electric field, even possibly yet another quantity, such as for example a magnetic field intensity (expressed in amperes per meter), a magnetic induction ( expressed in teslas), etc. In general, those skilled in the art know how to translate a threshold value associated with a given electromagnetic quantity into an equivalent threshold value associated with another electromagnetic quantity.

Finally, nothing excludes the regulatory area from being determined other than by means of a calculation formula. For example, such an area can correspond to a delimited area following an on-site measurement campaign, around a transmitter device of the same type and by means of dedicated tools, such as for example a broadband isotropic probe, a measurement analyzer. spectrum, etc. According to yet another example, the regulatory zone is obtained by numerical simulations from a modeling of the emitting device 2, taking into account the environment around the latter. It is also recalled that no limitation is attached to the shape of the regulatory zone defined around the transmitter device 2.

For the remainder of the description, we consider the situation whereby the

threshold value defined by the maximum transmission power EIRP is exceeded outside the regulatory zone along at least one of the beams when the transmitter device 2 transmits by means of the first precoder P_1. Such a situation is for example illustrated in FIG. 1A. According to the invention, the first precoder P_1 and the regulatory constraint associated with the transmitting device 2 (namely therefore said regulatory zone, as well as the threshold value relating to an electromagnetic quantity) are both used by a device configured for carry out processing aimed at determining at least one precoder:

- intended for sending device 2, and

- capable, in particular, of reducing the power radiated by the emitting device 2, so that the regulatory constraint is respected in the direction of a beam for which the threshold value was previously exceeded.

The processing carried out by the device in question is carried out by implementing a method for determining said at least one precoder. To this end, the device intended to carry out said processing comprises for example one or more processors and storage means (magnetic hard disk, electronic memory, optical disc, etc.) in which data and a computer program are stored, as a set

of program code instructions to be executed in order to implement all or part of the steps of the method for determining at least one precoder.

Alternatively or in addition, said device also comprises one or more programmable logic circuits, of FPGA, PLD, etc. type, and / or specialized integrated circuits (ASIC), and / or a set of discrete electronic components, etc. suitable for implementing all or part of the steps of the method for determining at least one precoder.

In other words, the device intended to carry out said treatments

comprises a set of means configured in software (specific computer program) and / or hardware (FPGA, PLD, ASIC, etc.) to implement all or part of the steps of the method for determining at least one precoder.

The remainder of the description relates more specifically to the case where the device

implementing said determination method is the transmitter device 2 belonging to the communication system 1. However, nothing excludes considering an implementation of this method by a device, called an "ancillary device", separate from said transmitter device 2, as explained.

later. Furthermore, for the sake of simplification of the notations, the convention is now adopted according to which any precoder P is identified with its matrix representation. Thus, in the present exemplary embodiment, the term “P_1” designates both the first precoder P_1 as such, but also the matrix associated with the latter.

[0081] Figure 3 shows schematically in the form of a flowchart a

embodiment, according to the invention, of the method for determining at least one precoder for the transmitter device 2.

As illustrated in Figure 3, the method for determining at least one precoder comprises several steps. In its general principle, said method consists first of all in determining a set of precoders respectively associated with directions distinct from each other. A representation of the precoder P_1 is then sought as a function of the precoders of this set, so as to be able to precisely identify the transmission directions along which the regulatory constraint is not respected outside said zone is considered. Another precoder is finally determined in order to guarantee a targeted reduction in the radiated power of the emitting device 2, more particularly in the direction of the beams for which the threshold value was previously exceeded, so that the regulatory constraint is respected.

To this end, the determination method firstly comprises a step E10 of obtaining a set G comprising a plurality of precoders respectively associated with distinct directions between them.

As mentioned above, the matrix P_1 encodes, via its coefficients, the information necessary for carrying out the MRT type precoding. However, its simple reading does not make it possible, except in particular cases, to easily determine the directions associated with the emission beams resulting from the use of said precoder P_1. Thus, said set G is obtained with the objective of making it easier to read the matrix P_1, more particularly to make it possible to identify which directions in space are used by the precoder P_1 to satisfy the operating criterion. Each precoder of the set G corresponds to a precoding vector denoted gj, the index i being included in the interval [1, L], where L is an integer strictly greater than 1 corresponding to the number of directions considered for said set G. In addition, such a precoding vector g_i is a vector with complex coefficients of size equal to M (number of transmitting antennas).

For the rest of the description, the direction associated with a vector g_i is

also identified by said index "i", without causing confusion. To denote such a direction, the expression "direction i" is used.

Each precoder gj being associated with a defined direction i of space, its application to the transmitter device 2, for the transmission of a data stream, generates a beam in said direction i. In other words, the application of a precoder gj to the transmitter device 2 generates an antenna diagram for which the radiated power EIRP is maximized in the direction i.

It should be noted that the beam directions i associated with the various precoders g_i differ a priori from the beam directions resulting from the application of the precoder P_1, except in the case where at least one column vector of said precoder P_1 is equal to one of said vectors g_i.

It is thus understood that the set G corresponds to a collection of vectors gj (or even a collection of beams), so that it also admits a representation in the form of a matrix of dimension equal to M x L. Each vector column of this matrix (also noted G according to the convention mentioned above) therefore corresponds to one of the vectors gj.

In a particular mode of implementation, the directions i associated with the precoders g_i of the set G are mutually orthogonal. For example, in the case of a linear array of transmitting antennas (so-called “Uniform Linear Array” array in English literature) whose spacing e between antennas is equal to l / 2, the following vector g_i :

corresponds to a precoder pointing in the direction i equal to Q, with respect to an axis perpendicular to said network of transmission antennas (j represents for its part a complex number whose square is equal to -1). If the angle Q, considered satisfies the following relation:

,,,

there exists k integer such that

it follows from the properties of the sine function that, on the one hand,

2 i L

sin Q _Ϊ = - for i = 0, (fe = 0)

L 2

so that the angles traversing the interval [0 °, 90 °] are described, and on the other hand

so that the angles traversing the interval [-90 °, 0 °] are described. It is therefore understood that the vector g_i constructed in this way are associated with distinct directions capable of being distributed uniformly in the plane as a function of the number L. Moreover, when the angle Q satisfies the above relations, the vector gj corresponds, in a manner known to those skilled in the art, to a column of the so-called “Discrete Fourier Transform matrix” whose element placed in line m, with m = 1, ..., M, and in column i, with therefore i = 1, ..., L, is written:

The vectors g_i thus defined are mutually orthogonal for the Hermitian scalar product inducing the Frobenius matrix norm.

However, nothing excludes considering precoders gj which are not orthogonal to each other for said Hermitian scalar product, since their directions respective remain distinct from each other. For example, in a manner substantially similar to the previous example in which the matrix G is a Fourier transform matrix, it is possible to envisage precoding vectors gj so that the element placed at the line m and at the column i of the matrix G is written:

the numbers b, being between 0 and 1, and all distinct from each other. A vector gj thus formed corresponds, in a manner known to those skilled in the art, to a vector called an “oversampled DFT vector” (“DFT” being the acronym for “Discrete Fourier transform” in the English literature. ).

The invention is not limited to the hypothesis of a linear configuration of the network of transmitting antennas. Thus, nothing excludes considering, for example, a network of antennas distributed uniformly in vertical and horizontal directions (network called "Uniform Plane Array" in the English literature). In this case, a precoding vector can for example correspond to the kronecker product of column vectors belonging to a Fourier transform matrix, for example a kronecker product between a vector associated with an elevation angle and a vector associated with a azimuth angle.

In general, no limitation is attached to the set G, and therefore to the shape of its matrix, since the directions considered are distinct. Note however that the precoding vectors g_i are

preferably designed so that the overlap of the antenna patterns respectively associated with these precoding vectors is minimized. The term “minimized overlap” refers here to a configuration in which, whatever the direction envisaged, the product of the powers radiated by the transmitting antennas, by application of the various precoders gj, corresponds to an EIRP power less than one. threshold value in

correspondence with the maximum transmission power imposed by the regulatory constraint outside the area around the emitting device 2. The fact of designing the assembly G in this way advantageously makes it possible to facilitate the implementation of the invention, such as described in detail below, to ensure compliance with the regulatory constraint by the sending device 2. Without loss of generality, it is considered for the remainder of the description, this essentially for the purpose of simplifying the notations, that the columns of the matrix G are orthogonal and normalized to 1. This amounts to considering that the product G ^l G is equal to the identity matrix (of dimension L).

In a particular mode of implementation, the set G corresponds to a determined set. For example, the matrix corresponding to the set G is stored in annex storage means of the system 1 of

communication, such as for example a database stored on a server. These ancillary storage means are distinct from the storage means of the transmitter device 2, so that obtaining the set G corresponds to a transmission of said matrix to the transmitter device 2 by communication means of the communication system 1. Once said matrix has been transmitted, the sending device 2 stores it in its storage means, so as to be able to implement the determination method according to the invention.

No limitation is attached to the configuration of said means of

communication, which can be wired or wireless, as well as using any type of known transport protocol.

Alternatively, the set G is determined directly by the transmitter device 2. For example, obtaining the set G corresponds to a calculation of the coefficients of the matrix associated with said set G by the transmitter device 2.

The determination method then comprises a step E20 of

determination of a P_2 precoder corresponding to a representation of the P_1 precoder as a function of the set G.

As has been mentioned previously, the simple reading of the matrix P_1 does not allow, except in particular cases, to easily determine the directions associated with the transmission beams making it possible to maximize the throughput received at the level of the receiver device 3. As soon as then, the fact of representing the precoder P_1 as a function of the set G makes it possible to overcome this defect.

[0100] Insofar as the matrix G is formed of column vectors

orthogonal and unitary, the projection v_proj of a vector v of dimension M on the column space of the matrix G corresponds to an orthogonal projection and verifies the following relation:

v_proj = G x ( ^l G xv).

[0101] Also, in the present mode of implementation, if we denote pi _< _1 and pi _< _2 the k-th respective columns of the matrices P_1 and P_2, said vector p ^ _2 satisfies the following relation for all k included between 1 and K (the number of data streams to be sent):

P _k _2 = G x ('G xp _k _1).

This formulation of the matrix P_2 implies that each column vector p ^ _2 is expressed as a linear combination of the column vectors of the matrix G. Thus the fact of projecting the matrix P_1 on the set G makes it possible to obtain a detailed representation, in terms of direction i of space, of the way in which a precoder associated with a column vector rk_1 of said matrix P_1 acts. This detailed representation is encoded in the matrix P_2 which results from this projection. It is then understood that the matrix P_2 is also of size M x K.

For the remainder of the description, a compact notation is adopted by designating by V | <_2 the vector ( ^l G rk_1), for any index k between 1 and K. The coefficients of such a vector vi _< _2 therefore correspond to a linear combination of the vectors g_i to obtain the vector pi _< _2. We also denote by V_2 the matrix of size L x K whose k-th column vector is equal to vi _< _2. This matrix V_2 therefore comprises all the information relating to the linear combinations expressing the precoder P_2 as a function of the precoders g_i of the set G, and therefore ultimately as a function of the directions associated with said precoders g_i. The relation between the precoder P_2 and the set G therefore results in the following formula:

P_2 = G V_2.

The fact of determining the matrix V_2 is particularly advantageous insofar as it is possible to obtain the matrix P_2 directly from the latter, the emitting device 2 already having knowledge of the matrix G. In addition, it It is also advantageous for the sending device 2 to store the matrix V_2 rather than the matrix P_2 given that the dimension L x K of V_2 is less than the dimension M x K of P 2. It should also be noted that the choice consisting in determining P_2 by carrying out an orthogonal projection of P_1 on the space generated by the column vectors of G only constitutes a variant implementation of the invention. Indeed, the expression of P_1 as a function of the set G can be seen, more generally, as the search for a solution to an optimization problem consisting in minimizing a distance between the precoder P_1 and the space generated by the column vectors of G. The formulation of this optimization problem can vary, in particular, as a function of the standard considered for evaluating said distance. Even more generally, any optimization method known to those skilled in the art can be implemented.

[0105] Subsequently, the determination method comprises a step E30

of identification, among the directions associated with the precoders of the set G, of directions known as "DP overshoot directions" according to which the regulatory constraint is not respected when the sending device is associated with said precoder P_2.

The term “direction of overshoot” therefore refers here to a direction i along which the radiated power EIRP is exceeded outside the regulatory zone.

[0107] Symmetrically, and for the remainder of the description, a direction

associated with a precoder of the set G and distinct from said overshoot directions is called "non-overshoot direction N DP".

[0108] Said step E30 therefore aims to identify the directions of

overshoot DP among the directions respectively associated with precoders 9_i

[0109] In a particular mode of implementation, the identification step E30

comprises, for each direction i of the set G, and initially, an estimate of the power F radiated in free space by the emitting device 2 and received outside said regulatory zone in said direction i. For example, said radiated power F is estimated at the edge of said zone, that is to say at the level of the geographical limit corresponding to the regulatory distance used to calculate the maximum transmission power associated with the regulatory constraint. The evaluation of said power F is carried out using the following formula:

F = | | Hj x G x V_2 || ² ,

or :

- Hj denotes the channel matrix for a receiving antenna under the assumption of a free space data stream transmission. This matrix Hj, insofar as it is associated with a single direction i, in fact corresponds to a row vector comprising M columns, so that the quantity Hj x G x V_2 forms a vector comprising K columns. This vector Hj is for example estimated by calculations, by numerical simulations or even by measurements

previously carried out, for example by taking into account the emission diagram of the antennas of the emitting device 2. The obtaining of such a vector Hj being known to those skilled in the art, it is not detailed further here .

- 1 |. || ² denotes the Euclidean norm of a vector (ie the sum of the squares of the absolute values of the components of a vector).

[0110] Secondly, once the radiated power has been calculated, step E30 comprises for the direction i considered a comparison of said power F with a threshold value, so as to determine whether said direction i corresponds to a overshoot direction DP. The threshold value considered here corresponds to the maximum transmission power associated with the regulatory constraint.

[0111] Consequently, by taking the previous notations, we deduce that if the

quantity || Hj x G x V_2 || ² exceeds the power threshold value, the direction i considered is identified as being an overshoot direction DP.

Thus, once step E30 has been completed, that is to say once all the directions i associated with the precoders g_i have been tested, a subset of the set G is available for which the associated directions are DP overflow directions. It is important to note that the number of directions of

DP overrun is greater than or equal to 1, since it has been assumed that the sending device 2 does not comply with the regulatory constraint when it transmits with the precoder P_1 (and therefore P_2), but not necessarily strictly greater than 1. [01 13] Once said DP overshoot directions have been identified, the determination method comprises a step E40 of determining a precoder P_3 as a function of the precoder P_2, so that the regulatory constraint is met along said DP overshoot directions when the transmitter device is associated with the precoder P_3.

[01 14] Thus, the precoder P_3 targets a reduction in radiated power

specifically following the DP overshoot directions that have been identified. Said precoder P_3 therefore not only makes it possible to comply with the regulatory constraint, but also to very effectively limit the degradation of the operating criterion in comparison with the solutions of the prior art.

[01 15] In a particular mode of implementation, step E40 of determining the precoder P_3 is executed iteratively. To this end, step E40 comprises, for each direction of overshoot DP, and during a first iteration associated with this direction of overshoot DP, an estimate of the power radiated in free space by the transmitter device 2 associated with the precoder P_2 and received outside said regulatory zone in said direction of overshoot DP.

[01 16] Thus, as explained previously, it is a question here of evaluating the quantity F equal to || Hj x G x V_2 || ² , where Hj corresponds here to the free space transmission channel vector for the overshoot direction DP considered during iteration.

Of course, when this quantity has already been estimated during the identification step E30, it can be stored by the storage means of the transmitter device 2, which can therefore reuse it without having to perform the corresponding calculations again.

[01 17] Still during the same iteration, and for the direction of overshoot DP considered, step E40 then includes an update of the precoder P_2, so that the power radiated in free space by the associated transmitter device 2 at the updated precoder P_2 and received outside of said regulatory zone in said direction of overshoot DP is less than the received power estimated before said update.

[01 18] For example, for the index i belonging to the interval [1, L] and associated with the direction of overshoot DP considered, the coefficients located on the line of index i of the matrix V_2 are reduced by a determined step, for example equal to 1 Watt. It will be understood that the fact of modifying the coefficients of V_2 in this way implies modifying the coefficients of the matrix P_2, the latter being equal to G x V_2. Thus, by updating the matrix V_2, the matrix P_2 is updated. In addition, by reducing the coefficients placed on line i of the matrix V_2, the radiated power is reduced in the direction of overshoot DP

considered.

[0119] Once the matrix V_2 (and therefore ultimately the matrix P_2) has been updated, the quantity F equal to || Hj x G x V_2 || ² is evaluated again (therefore during a following iteration) and compared with the power threshold value. If this quantity is less than the threshold value, this means that the radiated power along the direction of the initial overshoot DP has decreased sufficiently for the regulatory constraint to be respected from now on in this direction, which therefore no longer constitutes a direction of overshoot DP.

[0120] Conversely, if the quantity F is greater than the threshold power value, the coefficients located on the line of index i of the matrix V_2 (already updated previously) are again reduced by said determined step. This process is repeated until the quantity F is less than the power threshold value, in other words until the regulatory constraint is respected according to the said direction of excess DP considered.

[0121] This update of the precoder P_2 is carried out for all the directions of overshoot DP identified during step E30. The finally obtained precoder P_3 corresponds to the last update of the precoder P_2, once all the overflow directions DP have been considered.

It should be noted that this particular mode of implementation has been described by considering that each direction of overshoot DP was processed one by one and separately. This choice of implementation is only a variant

implementation of the invention. Thus, it is possible to envisage iterations during which several directions of overshoot DP are considered. For example, if several overshoot directions DP are considered simultaneously, the coefficients of the lines corresponding to these directions in the matrix V_2 are reduced by said step determined in parallel. Furthermore, nothing excludes either considering a step (for the reduction of the coefficients) which differs for one or more directions of overshoot, or even also for one or more coefficients of the same row (which amounts to perform iterations not line by line, but coefficient by coefficient). As an alternative, or else in addition, it is possible to envisage an adaptive reduction step, as the quantity F approaches the power threshold, for example by evaluating the variation of the radiated and received power in an overshoot direction. depending on the variation of the pitch. In this way, the updating of the precoder P_2, and therefore the obtaining of the precoder P_3, can be done in an informed manner, by finely controlling the reduction in transmission power of the emitting device 2.

[0124] Figure 4 shows schematically the effects of the application to

transmitter device 2 of FIG. 1A of the precoder P_3. It is considered here that the beams shown in Figure 1A are formed by means of the first precoder P_1.

[0125] As illustrated in FIG. 4, the respective sizes of the beams F_2 to F_6 are substantially identical to the corresponding sizes represented in FIG. 1 A. In other words, in comparison with FIG. 1 B, no drop in power of transmission takes place according to the beams F_2 to F_6, which effectively limits the degradation of the operating criterion (the rate received at the level of the receiver device 3 in the case of FIG. 4 is greater than the rate received in the case of figure 1 B). In addition, in Figure 4, the size of the beam F_1 has decreased, in comparison with the size shown in Figure 1 A, so that this beam F_1 is now contained, in an adjusted manner, in the circle C_P translating the maximum power emission to comply with the regulatory constraint.

For the rest of the description, we denote by V_3 the matrix of size L x K

corresponding to the last update of the matrix V_2 during step E40. According to such a notation, we therefore have that the matrix P_3 is equal to G x V_3.

[0127] FIG. 5 schematically represents a first preferred mode of implementation of the determination method of FIG. 3, in which the precoder P_3 obtained is updated to exploit in an improved manner the operation of the transmitter device 2 while complying with the regulatory constraint.

[0128] As illustrated in FIG. 5, the determination method comprises, for at least one direction of non-exceeding NDP, a step E50 of updating the precoder P_3, so that:

- the power radiated in free space by the transmitter device 2 associated with the updated precoder P_3 and received in said regulatory zone in said direction of non-exceeding NDP is greater than the corresponding received power before said update,

- the regulatory constraint is respected according to said direction of non-exceeding NDP.

[0129] For example, in a manner analogous to what has been described above for the implementation of step E40, and for the index i belonging to the interval [1, L] and associated with the direction of no exceeding NDP considered, the coefficients located on the line of index i of the matrix V_3 obtained at the end of step E40 are increased by a determined step.

It is understood that the fact of thus modifying the coefficients of V_3 implies a modification of the coefficients of the matrix P_3. Thus, by updating the matrix V_3, the matrix P_3 is updated. In addition, by increasing the coefficients placed on line i of matrix V_3, the radiated power is increased in the direction of non-overshoot NDP considered.

[0131] The increase in the coefficients of line i of the V_3 matrix is carried out under compliance with the regulatory constraint. In this way, the updated precoder P_3 makes it possible to guarantee that the radiated power remains below the threshold power value (therefore in fine electric field) in the direction of non-exceeding NDP considered. Part of the operating power of the antenna which had been abandoned during step E40 to satisfy the regulatory constraint according to the directions of excess DP is therefore used here again to improve the operating criterion, while respecting regulatory constraint.

It should be noted that the mode illustrated by FIG. 5 has been described so far by considering an update of the precoder P_3 for a single direction of no. NDP overrun. However, and preferably, when there is a plurality of non-exceeding NDP directions, all said non-exceeding NDP directions are considered, for example in an iterative manner. In other words, the coefficients of the matrix V_3, for the lines corresponding to all these non-overshoot directions NDP, are increased by a determined step.

[0133] Alternatively, when there is a plurality of directions of no

NDP overrun, only part of these directions are considered. By proceeding in this way, it is possible to favor NDP non-exceeding directions according to which the radiated power, before updating, was

particularly weak.

[0134] In general, no limitation is attached to the way in which the NDP non-exceedance directions are considered among themselves. In addition, it should also be noted that all the technical considerations relating to the determination of the pitch during step E40 still apply here (not identical for all directions, not different between one or more directions, adaptive pitch, etc. .).

[0135] Figure 6 shows schematically the effects of the application to

transmitter device 2 of FIG. 1 A of the precoder P_3 updated according to said first preferred mode of implementation. As illustrated in Figure 6, the respective sizes of the beams F_2 to F_6 are now greater than the corresponding sizes shown in Figure 4, without however these beams leaving the circle C_P. In other words, the regulatory constraint is respected by all the beams, the rate received at the level of the receiver device 3 in the case of FIG. 6 being however greater than the rate received in the case of FIG. 4. Consequently, the criterion of operation has been improved in the case of FIG. 6. This improvement results from a redistribution towards the beams F_2 to F_6 of the power abandoned to transmit in the direction of the beam F_1. Moreover, it is noted that the size of the beam F_1 remains identical between FIG. 4 and FIG. 6.

[0136] FIG. 7 diagrammatically represents a second preferred embodiment of the method for determining in FIG. 2, in which the precoder P_3 obtained is updated a plurality of times so as to obtain a plurality of precoders intended to be used in turn to exploit in an improved manner the operation of the transmitter device 2, while respecting the regulatory constraint.

[0137] As illustrated in FIG. 7, the determination method comprises a plurality of updating steps E50 I. These update steps are independent of each other. By "independent", we refer here to the fact that the precoder obtained at the end of one of the update steps is not used for the execution of the other update steps. In addition, said updating steps are respectively associated with the directions of the beams obtained by means of the precoder P_1 with which the emitting device 2 is initially configured.

[0138] Thus, in this preferred mode, and for such a given beam direction, the precoder P_3 is updated so that:

the power radiated in free space by the emitting device 2 associated with the precoder P_3 updated for said beam direction and received in said regulatory zone along said beam direction is greater than the corresponding received power before said update.

In an exemplary implementation, using the previous notations, the method firstly comprises an identification of the lines V_ {3, 1} of the matrix V_3 which are not zero (ie the lines which include at minus a non-zero coefficient). The indices I of these non-zero lines are stored. Then, for each of said stored indices I, considered independently of one another, the matrix P_3 is updated, so as to obtain a plurality of updated matrices P_3. More particularly, for such a stored index I, the coefficients of V_3 located on the line associated with said index I are iteratively increased by a determined step. The process is then repeated for each of said stored indices I, so as to obtain I updated matrices P_3.

[0140] By thus increasing the coefficients of the matrix V_3 for an index I

stored, said matrix V_3 is updated, and therefore also matrix P_3. Moreover, such an increase has the effect that the power radiated in only one of the beam directions associated with the precoder P_1 increases, up to potentially exceeding the regulatory constraint. On the other hand, the regulatory constraint remains respected according to the other beam directions.

[0141] It is noted here again that no limitation is attached to the way in which the step determined during each step E50_l is defined.

[0142] Moreover, and as illustrated in FIG. 7, the updating steps E50_l are carried out one after the other, considering the different indices I for which there are non-zero lines in the matrix V_3. However, nothing excludes considering that said steps E50_1 are executed in parallel.

[0143] It is therefore understood that in this second preferred embodiment, each updated precoder P_3 is capable of generating a violation of the regulatory constraint in only one of the beam directions.

[0144] The fact of determining such updated P_3 precoders is advantageous in the context of a specific method of transmitting at least one data stream according to the invention. For such an issue process, it is considered that the transmitter device 2 is, in addition to the regulatory constraint, also associated with a regulatory issue duration. Such a regulatory issuance period corresponds for example to a period during which the constraint

regulatory must be observed on average along a beam direction. By "respected on average", we refer here to the fact that:

- the threshold value of the regulatory constraint may be exceeded during the regulatory period depending on the beam direction considered,

- the time average (calculated over a time interval corresponding to the regulatory duration) of the considered electromagnetic quantity generated does not exceed the corresponding threshold value outside the regulatory zone. By way of non-limiting example, the regulatory duration is equal to 6 minutes.

[0145] Said transmission method further comprises transmission steps

successive respectively associated with said updated precoders P_3, each of said transmission steps being carried out during a determined fraction of the regulatory transmission time, so that said constraint regulatory compliance is observed on average during said regulatory period in any beam direction.

[0146] For example, if the regulatory duration is equal to 6 minutes, and if we

considers six transmission beams, the fraction allocated to the transmission following each updated P_3 precoder is equal to 1 minute.

It should be noted that these examples of regulatory duration (6 minutes) and of fraction allocated to each updated precoder P_3 (1 minute) are given purely by way of illustration. Thus nothing excludes having other values of regulatory duration and corresponding fractions (the said fractions may for example differ from each other), as soon as the regulatory constraint is satisfied on average along each beam direction during the course. of the said regulatory period. There is also nothing to exclude that the sum of the fractions allocated respectively to the updated P_3 precoders is less than the regulatory duration. In this case, it is possible to envisage using during the remaining time (that is to say the regulatory time minus the sum of the allocated fractions) another precoder, such as for example the first precoder P_1, as soon as the constraint regulatory is satisfied on average in each beam direction during said regulatory duration.

[0148] The fact of successively using the various updated P_3 precoders in this way makes it possible to limit the violation of the regulatory constraint to a short period of time. The regulatory constraint is thus satisfied on average. To this is added the fact that, in a manner analogous to the implementation described with reference to FIG. 5, the power abandoned during step E40 to comply with the regulatory constraint is here redistributed in turn for the transmission of each of the beams for a determined period. In the end, the operating criterion is improved on average over time.

[0149] FIGS. 8A, 8B, 8C, 8D, 8E, 8F diagrammatically represent the effects of the application to the emitting device 2 of FIG. 1A of a plurality of precoders P_3 updated according to said second preferred updating mode implemented. More precisely, FIGS. 8A to 8F correspond to successive applications of updated precoders P_3 and respectively associated with beams F_1 to F_6. In addition, in these Figures 8A to 8F, the transmitter device 2 is associated with a regulatory duration of 6 minutes, and each updated precoder P_3 is applied to device 2 for a duration of 1 minute.

[0150] As illustrated in Figure 8A, the respective sizes of the beams F_2 to F_6 are substantially identical to the corresponding sizes shown in Figure 4 (and therefore also shown in Figure 1A). The size of the beam F_1 is for its part substantially identical to the corresponding size in FIG. 1A. In other words, the beam F_1 is in excess, and the constraint

regulation is not respected according to its direction. It should also be noted that the updated precoder P_3 so that the beam F_1 is exceeded can correspond to the precoder P_2 here.

[0151] FIGS. 8B to 8F differ from one another, as well as from FIG. 8A, in that the beams F_2 to F_6 are in turn in excess for 1 minute.

[0152] Although the beams F_1 to F_6 are therefore in turn exceeded for 1 minute, the regulatory constraint is however observed on average for the regulatory period of 6 minutes and in any direction of the beams. In addition, the fact of using the transmission capacities of the transmitter device 2 in a fractional way over time ultimately improves the operating criterion compared to the case of Figure 4.

[0153] It therefore clearly emerges from the present description that the modes described with reference to FIGS. 5 and 7 correspond to alternatives for implementing the invention in order not only to comply with the regulatory constraint, but also to limit the degradation of the criterion d. 'operation even more efficiently than that obtained from the single precoder P_3.

It should be noted that the invention has been described up to now in the context where the device implementing the determination method is the transmitter device 2. Such a choice constitutes only one variant of implementation of invention. Thus, nothing excludes considering that the determination method is implemented by an annex device, which may be the receiver device 3 of the communication system 1 or even a third-party device. When this is the case, the determination method then comprises a step of transmitting, from the ancillary device to the sending device 2, at least one of the determined precoders (precoder P_3 and / or precoder P_3 updated according to said first preferred mode implementation and / or precoders P_3 updated according to said second preferred mode of implementation), so that said transmitter device 2 can encode the data streams to be transmitted. Among the implementation variants, it is understood that the determination device can be included in the transmitter device or in the receiver device.

It is also understood that the device implementing the determination method must have knowledge of the first precoder P_1. For example in the case where the annex device corresponds to the receiver device 3, the precoder P_1 is determined by said receiver device 3 (the receiver device 3 having previously acquired, in a manner known per se, the

knowledge of the He channel from the sender device 2 to the receiver device 3), before being transmitted to the sender device 2 via the H channel of which they both know by reciprocity. According to another example, if the ancillary device is a third-party device, it is equipped with communication means (wired or wireless, and capable of implementing any communication protocol.

known communication) to perform said transmission.

Finally, the invention has also been described up to now by considering a hypothesis of emission in free space with regard to the emitting device 2. This hypothesis therefore relates to the modeling of the environment of propagation of waves from the emitting device 2. The choice of such a hypothesis is however only an implementation variant of the invention that it may be realistic to consider in the context of an assessment of compliance with the regulatory constraint. . This being the case, other modeling hypotheses can be envisaged, such as for example by taking into account prerecorded data from the propagation environment in the vicinity of the emitting device 2. In other words, and in general, no limitation is possible. is attached to the modeling hypotheses of the propagation environment. A person skilled in the art will know how to adapt the calculations made according to the invention, in particular the calculations of power received in and outside the regulatory zone, according to a given modeling assumption.

Claims

[Claim 1] Method for determining at least one precoder for a transmitting device (2) equipped with a plurality of transmitting antennas, the transmitting device (2) being associated with:

- a precoder P_1 making it possible to optimize, with regard to a criterion

determined operating mode, a transmission of data streams in the form of beams,

- a regulatory constraint corresponding to not exceeding, outside a regulatory zone defined around said emitting device (2), a threshold value relating to an electromagnetic quantity,

said determination method comprising:

- a step (E10) of obtaining a set G comprising a plurality of precoders respectively associated with directions distinct from each other,

- a step (E20) of determining a precoder P_2 corresponding to a representation of the precoder P_1 as a function of the set G,

- a step (E30) of identifying, among the directions associated with the precoders of the set G, of directions called “overshoot directions” according to which the regulatory constraint is not complied with when the sending device (2) is associated with said precoder P_2,

a step (E40) of determining a precoder P_3 as a function of the precoder P_2, so that the regulatory constraint is respected along said overrun directions when the sending device (2) is associated with the precoder P_3.

[Claim 2] The method of claim 1, wherein the step

(E30) identification comprises, for each direction considered of the set G:

- an estimate of a power radiated by the emitting device (2) and received beyond said regulatory zone in said direction,

a comparison of said power with a threshold value, so as to determine whether said direction corresponds to a direction of overshoot. [Claim 3] Method according to any one of claims 1 to 2, in which the step (E40) of determining the precoder P_3 comprises, for each direction of overshoot considered:

- an estimate of a power radiated in free space by the emitting device (2) associated with the precoder P_2 and received beyond said regulatory zone along said direction of overrun,

- an update of the precoder P_2, so that the power radiated by the emitting device (2) associated with the updated precoder P_2 and received beyond said regulatory zone in said direction of overshoot is less than the estimated received power before said update,

the estimation and the update being iterated for the direction of overshoot considered until the regulatory constraint is respected according to said direction of overshoot,

[Claim 4] A method according to any one of claims 1 to 3, wherein:

a direction associated with a precoder of the set G and distinct from the overshoot directions is called “non-overshoot direction”, the method comprising, for at least one non-overshoot direction considered, an update step (E50) the precoder P_3, so that:

- the power radiated by the emitting device (2) associated with the updated precoder P_3 and received in said regulatory zone in said non-exceeding direction is greater than the corresponding received power before said update,

[Claim 5] A method according to claim 4, in which the step (E50) of updating the precoder P_3 is iterated for each of the directions of non-overshoot considered, an updating step associated with an iteration being carried out from of the P_3 precoder obtained during the previous iteration. [Claim 6] A method according to any one of claims 1 to 3, wherein:

the method comprising steps (E50J) of updating the precoder P_3 which are independent from one another and respectively associated with the beam directions, so as to obtain a plurality of updated precoders P_3 and that, for a given beam direction:

the power radiated by the emitting device (2) associated with the precoder P_3 updated for said beam direction and received in said regulatory zone along said beam direction is greater than the corresponding received power before said update.

[Claim 7] A method according to any one of claims 1 to 6, said method being intended to be implemented by a device, called an "ancillary device", distinct from said transmitter device, said method

further comprising a step of transmitting from the ancillary device to the transmitting device (2):

- of the precoder P_3 determined according to any one of claims 1 to 3; or

- the updated P_3 precoder determined according to any one of claims 4 to 5; or

- updated P_3 precoders determined according to claim 6.

[Claim 8] A method of transmitting at least one data stream by a transmitting device (2) equipped with a plurality of transmitting antennas, the transmitting device (2) being associated with:

- a precoder P_3 determined beforehand according to a method according to any one of claims 1 to 3; or

- an updated precoder P_3 determined beforehand according to a method in accordance with any one of claims 4 to 5.

[Claim 9] A method of transmitting at least one data stream by a transmitting device (2) equipped with a plurality of transmitting antennas, the data stream being transmitted in beam directions, said device transmitter (2) being associated with:

- updated P_3 precoders previously determined according to the

claim 6,

- a regulatory issuance period,

said method comprising successive transmission steps associated respectively with said updated precoders P_3, each of said transmission steps being carried out for a determined fraction of the regulatory transmission time, so that said regulatory constraint is observed on average during of said regulatory duration in any beam direction.

[Claim 10] A computer program comprising instructions for implementing a determination method according to any one of claims 1 to 7 or a transmission method according to any one of claims 8 to 9 when said program is executed by a processor.

[Claim 11] A computer readable recording medium on which is recorded a computer program according to claim 10. [Claim 12] Device for determining at least one precoder for a transmitter device (2) equipped with a plurality of transmitting antennas, the transmitting device (2) being associated with:

- a precoder P_1 making it possible to optimize, with regard to a criterion

operating mode, a transmission of data streams in the form of beams, said precoder P_1 being known to the determination device,

- a regulatory constraint corresponding to the failure to exceed, beyond a regulatory zone defined around said emitting device (2), a threshold value relating to an electromagnetic quantity,

said determination device comprising:

- an obtaining module, configured to obtain a set G comprising a plurality of precoders respectively associated with directions distinct from each other,

a first determination module, configured to determine a precoder P_2 corresponding to a representation of the precoder P_1 as a function of the set G,

- an identification module, configured to identify, from among the directions associated with the precoders of the set G, directions called “overshoot directions” according to which the regulatory constraint is not respected when the sending device (2) is associated said P_2 precoder,

a second determination module, configured to determine a precoder P_3 as a function of the precoder P_2, so that the regulatory constraint is respected along said directions of overrun when the sending device (2) is associated with the precoder P_3.

[Claim 13] A device according to claim 12, said corresponding determining device:

- to the sending device (2); or

- to a device, called "ancillary device", distinct from said transmitter device (2), said ancillary device further comprising means for

communication configured to transmit at least one precoder to the sending device (2).

[Claim 14] Device according to claim 13, said annex device is a receiving device (3) of the data stream equipped with at least one receiving antenna.

[Claim 15] Wireless communication system (1) comprising a transmitting device (2) equipped with a plurality of transmitting antennas and a receiving device (3) equipped with at least one receiving antenna, said transmitting device (2) being associated with:

- an updated precoder P_3 determined beforehand according to a method in accordance with any one of claims 4 to 5; or

- updated P_3 precoders previously determined according to the

claim 6.