EP2243227A1

EP2243227A1 - Communication by return pathway from a terminal to a transmitter for reducing in particular interference between beams from the transmitter

Info

Publication number: EP2243227A1
Application number: EP08867695A
Authority: EP
Inventors: Thomas Salzer; Marios Kountouris
Original assignee: France Telecom SA
Current assignee: Orange SA
Priority date: 2008-01-03
Filing date: 2008-12-12
Publication date: 2010-10-27
Also published as: CN101971516A; WO2009083680A1; US20100284294A1

Abstract

The invention relates to a communication by return pathway from a terminal to a transmitter for reducing in particular interference between beams from the transmitter. A telecommunication system comprises a transmitter designed to render simultaneously active a first number of beams (Fl, F2, F3), in the guise of the resources for a plurality of user terminals. Within the sense of the invention, each terminal: calculates, while taking account of interference noise, at least one second preferred, number, of beams to be rendered simultaneously active by the transmitter, and sends to the transmitter, through the return pathway, an indication of this second preferred number, among information about the quality of reception. On its side, the transmitter (BS) can then adjust the first number of beams to be rendered active, as a function of the returns from the terminals, for a next burst.

Description

Communication by way of a return of a terminal to a transmitter, in particular to reduce interference between beams emanating from the transmitter

The present invention relates to a communication on a return path from a terminal to a transmitter associated with a telecommunication network, with a view to informing this transmitter about the quality of reception of data by the terminal.

In particular, it concerns a communication (or "metric") that would make it possible to adapt the transmission mode, for example, in a space division multiple access or "SDMA" system (for "Space Division Multiple Access"). Such a system is characterized by the use of multiple antennas (so-called "MIMO" technique for "Multiple Input Multiple Output") on transmission to generate beams that can be allocated to different user terminals. Thus, a transmitter (for example a base station in a cellular network) may comprise a plurality of antennas generating a set of beams that can be allocated to one or more terminals. Thus, a transmitter comprising M antennas can generate at most M beams simultaneously. However, in practice, as will be seen below, a number K, less than or equal to M, beams is generated, in particular to avoid interference between beams on the same channel from a user terminal.

The use of multiple antennas to serve several users (SDMA system), thanks to a set of beams, is part of the recent standards for mobile radio networks, such as the E-LJTRAN standard described in particular in: 3GPP TS 36.212, Version 8.0.0 - "Physical Channels and Modulation (Release 8)".

The invention is therefore presented later in a pure SDMA system, but it can also be applied in a hybrid system combining the SDMA technique with any other multiple access technique. In fact, to serve a plurality of terminals (for example more than M terminals), it can in particular be provided to serve K terminals at a time t, then K other terminals at a time t + 1, etc. Alternatively, it can be expected to serve K terminals to a frequency band F, then K other terminals to a frequency band F + 1, etc. Thus, the invention can be applied in a hybrid system combining the SDMA technique with, in particular TDMA (for "Time Division Multiple Access") if it is decided to allocate resources by time slots, or the OFDM or FDMA (for Frequency Division Multiple Access) if it is decided to allocate resources in separate frequency bands, or CDMA (for "Code Division Multiple Access") or others, as long as one channel return is used to refine the choice of transmission parameters.

Indeed, the invention relates to the allocation of beams to users "downstream" where the configuration of the transmission mode (number of beams, modulation, coding, beam orientation, etc.) is generally determined using information carried by a rising path (terminals to the transmitter), this upstream channel being called "return path" (or "feedback").

However, this way of return is very expensive in terms of flow. It is therefore preferable to find a compromise between the quality of the information obtained by this feedback and the amount of information sent. In an SDMA system where the choice of resource allocation depends on the expected reception quality information for a large number of user terminals and for a large number of beams, the amount of information sent on the return path can become quickly prohibitive. Reducing this amount would imply the implementation of a less efficient resource allocation technique.

State-of-the-art techniques therefore aim to reduce the amount of information on the return path while allowing the use of a connection allocation and configuration algorithm (called "link configuration"). which is the most efficient possible. The choice of the number of active beams simultaneously according to the state of the art is generally done at the base station, only, which represents the disadvantage that the base station does not know the impact of the interference. between beams on the quality of the link to the user terminal. There are many feedback methods to inform the base station of this impact of the interference, but this information is given a posteriori and requires, in any case, a quantity often prohibitive feedback. Without additional feedback on the impact of the interference, the base station is generally forced to transmit a fixed number of beams in parallel to allow a terminal to evaluate the impact of the interference and to estimate the interference. link quality reliably. Such a fixed configuration of the number of beams, however, is not optimal for some user terminals, depending on their channel state.

In fact, reference is made to FIG. 1, on which is represented a space division multiple access ("SDMA") system (for "Space Division Multiple Access"). The multiple antennas at the transmitter (for example a base station BS) are used to generate separate beams F1, F2, F3 which represent the resources that can be allocated to different terminals T1, T2, T3, T4. The number of distinct beams that can be generated simultaneously is generally equal to the number of antennas included in the base station, this number being denoted by M hereinafter (with M = 3 in the example of FIG. 1). In principle, the maximum number of terminals that can be served simultaneously is also M.

A group of M antennas can generate a multiplicity of sets of M distinct beams. The optimum choice of one of these sets depends on the relative position of the terminals to be served simultaneously and the state of their radio channels.

A particularity of an SDMA system is the fact that the orthogonality of the resources is not ensured, which creates interference between the signals transmitted on different beams at the receiver of a terminal. To limit this interference, it is recommended to make a choice of a group of beams (in transmission mode) which will be adapted to a choice of a group of terminals to be used simultaneously.

By way of example, to illustrate this point, the base station BS of FIG. 1 has three antennas and could therefore serve three of the four terminals present simultaneously. If we compare all the beams created by the antenna network of the base station with the state of the channels of the user terminals (these channels being here defined only by their position), one understands the compromise that must be found. First, the base station BS must choose between the user terminals T3 and T4 which are both covered by the beam F3 but can not be served simultaneously. A possible choice would therefore be to serve the user terminal T1 with the beam F1, the terminal T2 with the beam F2, and the terminal T3 with the beam F3.

However, the user terminal T2 is between the beams F2 and F3, which means that it receives the two beams with a similar quality. As a result, signals transmitted with the beams F2 and F3 arrive at the receiver of the terminal T2 with a similar power, which generates strong interference. The quality of the signal, if the latter was transmitted on the beam F2, is not ensured.

Knowing that a signal transmitted on the beam F2 also creates a little interference at the user terminals Tl and T3, the use of the beam F2 is not optimal compared to the configurations of the terminals represented here by their respective positions. Thus, the best choice for the system would be to save the power necessary to serve the user T2 and to serve, on the other hand, only users T1 and T3. Then, users T2 and T4 could be served with another set of beams, for example F2 and F3 beams oriented otherwise.

It will then be understood from this example that the optimal number of active beams for a transmission, which is K-2 in this example of FIG. 1, is not necessarily the maximum number M of beams that the transmitter is capable of producing. simultaneously (with M = 3 in the example of FIG. 1), because of the interference that can be generated by a maximum number M of active beams simultaneously.

However, the transmitter BS has no prior knowledge on the exact state of the channels of the user terminals at the time of transmission and more particularly on the interference generated on such channels by the beam allocation and the mode chosen transmission. It is therefore difficult for him to choose: the transmission mode, i.e. the beam set and the number K of active beams simultaneously, and the beam allocation to each user terminal.

The present invention improves the situation.

It proposes for this purpose a telecommunication method in a system comprising at least one transmitter arranged to simultaneously activate a first number of beams, as resources for a plurality of user terminals, in which the user terminals receive, by said beams, telecommunication data.

At least one of the terminals transmits to the transmitter, by return, an indication of at least a second preferred number of beams to be active simultaneously by the transmitter.

Note that the second preferred number is specific to the terminal, in that it results for example from the calculation of the terminal, alone, or that it is for example determined by the terminal, alone, on the basis of information prerecorded in this latest. It is transmitted to the transmitter for possible consideration by it in particular to adjust the number of beams to be active simultaneously. However, this number of beams made active will not necessarily correspond to the preferred number of beams by a terminal.

This second preferred number is calculated by the terminal, preferably taking into account an interference noise.

Thus, it will be taken into account in this calculation the interference that may result from the simultaneous reception of several beams at the same terminal, for example via the estimation of a signal to interference plus noise ratio. Moreover, in a particular embodiment, it will be sought to maximize an estimate of the overall bit rate that the transmitter can provide to all the terminals, which estimation of the overall bit rate is a function of the ratio signal on interference plus noise mentioned above. A terminal can then determine its preferred number of beams as being the number of beams maximizing this estimate of the overall rate.

It is therefore proposed to require the terminal an indication of its preferred mode of transmission (typically the number of beams tolerated in parallel). This indication of the terminal as to its preferred mode of transmission can then define a particular metric, used on the return path from the terminal to the base station, to inform the base station of this preferred mode.

Thus, the invention proposes to rely on a terminal for the latter to determine its preferred transmission mode, which can be the subject of a new metric definition used on the return path to inform the transmitter. This metric may, for example, be part of a set of feedback metrics (or "feedback" hereinafter) allowing the base station to operate in order to optimize the overall throughput served to the user. set of user terminals while ensuring a satisfactory quality of service for each user terminal. A typical set of metrics may consist of: an index of the preferred beam by the terminal (an integer for example), - a reception quality value in this beam, a transmission mode indicator (for example an integer or a binary number indicating the preferred number of beams, according to possible alternative embodiments of the invention).

Nevertheless, the invention is not limited to the application of such metrics. For example, it is possible to provide feedback on the transmission mode indicator by a priori fixing, at the base station level, the beam used for a given user or by determining this preferred beam by other means. as the return channel (eg an estimate made by the uplink base station). It is therefore possible to provide a metric in which one of the information conveyed by the other metrics above could be readily available at the base station without requiring return terminals.

Other characteristics and advantages of the invention will appear on examining the detailed description below, and the appended drawings in which, in addition to FIG. 1 illustrating the allocation of resources in an SDMA system described above: FIG. 2 schematically illustrates the processing by the transmitter of the information communicated by the terminals on the return path, FIG. 3 schematically illustrates the steps of a method in the sense of the invention.

FIG. 2 represents an example of a functional diagram implemented, within the meaning of the invention, by a base station. The base station is equipped with M antennas and can thus transmit at most M flow of information to user terminals. The M antennas, however, generate K distinct beams simultaneously, with K ≤ M to avoid interferences between beams. It is assumed here that a separate beam immediately corresponds to a served terminal. It can alternatively be provided that several beams can be used by the same terminal, which can be done in practice for example depending on the state of the radio channel and if the terminal, for its part, has a sufficient number of antennas For the reception.

The set of beams used simultaneously for a transmission is denoted Ω = {w _| , ..., w _A. }. Although the implementation of the invention is not, of course, limited to such an embodiment, it is therefore hypothesized that: K ≤ M.

The set Ω is then composed of K vectors of size M, denoted W _n = {w _v ..., w _M ].

Each vector w ,, thus represents a distinct beam and has M complex coefficients (as components), these coefficients corresponding in practice to the weights applied to each antenna branch to generate a beam w ,,. In the example shown in FIG. 2, the base station receives from a terminal information on the quality of the link Q to optimize the processing S21 of modulation and coding for this terminal, as well as information on a preferred beam w _pre f to finally optimize the control S22 antenna beams ANTl ... ANTM. For the purposes of the invention, the base station also receives from the terminal a preferred number of K _prej beams to optimize the number K of beams to be transmitted.

More particularly, to generate the set of K beams, allocate them to the user terminals and perform "link adaptation" (which consists in choosing an optimal modulation and coding processing in step S21), the base station receives and interpreting such information transmitted to it on the return path by a user terminal. It can be, for example, two types of information, as indicated above: a first information w _pre f on the beam or beams preferred by the terminal, which can therefore depend on the position of the terminal relative to the base station and / or the state of its radio channel, a second information Q on the quality of the radio link reachable in the beam or beams preferred by the terminal, this second information then making it possible to decide on the effectively allocating a beam to a user and performing the link adaptation.

For example, the first information may be the index of a preferred vector in a set of vectors known by the terminal and the base station. This set can be defined by a dictionary (or "code-book"), common between the base station and the terminal. It can be alternatively provided that the base station can select the beams independently and transmit pilots on each beam to allow each terminal to identify them.

The size of the overall set of beams that it is possible to generate by the base station (taking into account in particular the different possible beam shapes) is, in general, greater than or equal to the number M defined previously and this set will be noted in the following Ω '= {w,, ... ^^} with N ≥ M (and in particular with TV = M in the aforementioned case of a transmission of pilots in each beam as indicated above) . It will be noted that the set of beams actually transmitted Ω can then be a subset of the set Ω '. It will then be understood that the base station can freely orient or refine the K beams of the set Ω from the feedback received from the terminals, by making its choice in the set Ω 'of the TV possibilities of beams.

The second information received can generally be based on an estimate, at the terminal, of the Signal to Interference plus Noise Ratio (or SINR) ratio. This estimate will be described later for a particular embodiment of the invention. The feedback, per se, can be done for example according to the method set forth in document FR-2 893 468.

Within the meaning of the invention, it is proposed to associate a third type of information. This additional information indicates to the base station how many other streams (beams) the user terminal can tolerate in parallel with its current flow of information. This information is advantageously quantified and its quantification is based on an estimate of the impact of the interference of the other beams on the quality of its current radio link. The value assigned in feedback relative to this third piece of information may be an integer or even a single bit, according to alternative embodiments described below.

In what follows, we consider that each terminal explicitly knows the vectors of the set Ω '. The base station can transmit for example a pilot signal on each antenna enabling the terminal to estimate a complex coefficient h _m representing the effect of the mobile radio channel between each antenna of the base station and the receiving antenna of the terminal. The radio channel between the base station and the terminal can thus be represented by a vector h = {/ ?,, ..., h _m , ..., h _Vf ). As indicated above, the terminal may alternatively have several receiving antennas. In addition, alternatively or in addition, the radio channel can be described by several complex coefficients at once, as for example in the case of a frequency selective channel. In these variants, the channel is then described by a matrix (rather than a single vector h) and the following expressions, given in the embodiment where a single vector h is assigned to a channel, can be adapted from appropriate way. It will therefore be understood that the invention is in no way limited to the assignment of a single vector to a channel.

From the estimation of the vector channel h and the knowledge of the set Ω 'of the beams, the terminal can estimate its quality of radio link in each of the beams under the assumption of a transmission mode using a set Ω which is a subset of the set Ω '. Here, we also assume that the same power is allocated to all beams. However, it may be provided to adapt the following expression to the case of a variant according to which the power allocation is not homogeneous.

According to the above hypothesis, the signal to interference plus noise (SINR) ratio, for a given beam w, can be written as follows:

SINR (Vf _n , Ω), n = 1, ..., N, where σ represents the inverse of the signal-to-noise ratio (therefore the ratio between a reception noise and the power of the useful signal received by the terminal) and K the cardinality of the set Ω. In addition, the notation h.vv ₍ denotes the dot product between the column vector h and the column vector W ₁ , ie hw, = h ^? W ₍ .

It should be noted that, according to the definition of this metric, the terminal must make an assumption on the set Ω which will actually be chosen by the base station, the set Ω being unknown to the terminal at the time when it considers its quality of reception. Therefore, the terminal can not calculate exactly the interference part (first term of the denominator).

It is shown, however, that it is possible to establish an estimate of the ratio SINR in a beam w, ₍ given with knowledge only the size K of the set Ω, the vector h and the global set Ω '. this effect is usefully referred to the document:

"Efficient Metrics for MIMO Broadcast Channels with Limited

Feedback ", M. Kountouris, Francisco R., D. Gesbert, D. Slock, T. Salzer, in Proceedings IEEE ICASSP, Hawaii, USA (April 2007).

We can then choose here an estimate of the form:

SINR (W _k , K) =, k = 1, ..., K, (1)

This estimate is satisfactory under the assumption that the base station chooses a set Ω of orthogonal (or approximately orthogonal) vectors. This precaution is in any case desirable to reduce interference between beams transmitted simultaneously. The beam selection algorithm of the base station therefore generally seeks to achieve such a configuration.

The function C (K, M, σ ² ) in the relation (1) above serves to estimate the impact of the interference with the other beams, with advantage as a single hypothesis the number K of transmitted beams simultaneously. This is a generally non-linear and configurable function. It can be optimized according to the knowledge available to the terminal, for example on the mode of adaptation and selection of the beams at the base station. It is therefore possible for the terminal to choose, among all the vectors of the set Ω ', a preferred beam, represented by the vector w _pιef given by an expression of the type: w _pref = argmax _Ω , | hw _(/ | that is, the vector w _n of the global set Ω 'which has the largest scalar product in absolute value with the channel vector h) In other words, the preferred vector \ v _p , _ef is therefore the one that, advantageously, maximizes the projection on the channel vector h.

The index k of this vector (with W ^ = vtprej) in the global set Ω 'therefore represents one of the elements of the feedback to the base station. It should be noted that the terminal could also determine several beams with an order of preference and inform the base station according to a succession of values given in this order of preference.

In a second step, the terminal determines its preferred mode of transmission. It is considered here that the K beams generated simultaneously are likely to create more or less interference with the beam served to a given user. As it is desired to minimize interference for all users served

(As mentioned previously with reference to FIG. 1), the number K is therefore defined as the optimal number of beams tolerated in parallel for each user.

Of course, a terminal will prefer to be served without interference from other beams served simultaneously with his. In this case (ideal case where one can consider simply K-I for the calculation of the SINR report), its SINR report is simply written as follows: lh.w,

SINR (Vf _pιet , I) = σ

From this SINR report value, the terminal can estimate the bit rate that the base station is able to transmit to it in this configuration, this bit rate being noted R (w _pιet 1). Generally, the terminal has for this purpose a correspondence table allowing it to associate a flow SINR report. He may also, by approximation from the Shannon limit, calculate the flow as follows:

* (w _PM , D = Glog (l + SINR (w _pref , l)), where G is a constant of the system which depends in particular on the available frequency band and other parameters known by the terminal. then the maximum rate that can be provided to this terminal if chosen by the base station.

However, the terminal can estimate its bit rate for transmitting K beams in parallel in the same way, with:

R (Yf ^, K) = Glog (1 + SINR (Yf ^ _n K))

hw _A where w, = w and <9 = arccos hw _A '

For the base station, it is best to serve multiple users simultaneously to maximize the overall system throughput. However, the fact of serving in addition to a user terminal increases the interference and therefore decreases the rate per user. Usually, the base station can not estimate a priori the impact of this interference created on the quality of the links of the users because it knows only their preferred beam w _ptej but not their channel vector h.

An embodiment of the invention then proposes to calculate the optimal configuration at the terminal through an approximation of the system flow. For this purpose, an approximation of a homogeneous network is made, where each user is served with the same rate. The total throughput therefore simply represents the rate per user multiplied by the number K. This assumption is used only for the calculation of the optimal mode of transmission but does not limit the scope of application of the invention.

The terminal can then find its optimal configuration K _pιef as follows:

K _pιe , = argmax (*: χ R (yv _pιef , K)) (2) λ '= 1, 2, .M

This number will then be transmitted on the return channel and allow the base station to serve the user terminal in its preferred mode, which maximizes the throughput taking into account the overall system including also the other terminals.

In practice, it will often be useful not to report the K _pre number, per se, but simply to report in binary mode whether the terminal may or may not tolerate other users in parallel with it. The terminal can therefore choose between a configuration K = I and K = M and signal its preference by a single bit of feedback.

It should be noted that, in the case where K = M, the function C (K, M, σ ² ) becomes trivial and equation (1) is written:

SINR (W _k , K)

FIG. 3 shows schematically the main steps of the above method in one embodiment.

As indicated above, the transmitter BS transmits to the terminal pilot signals enabling the terminal to estimate the coefficients h ₁ ,..., H _1, each representing a radio channel between an antenna of the transmitter and this terminal (or terminal antenna). From these coefficients h \, ..., h, y, the terminal is able to constitute in step S31 the vector h representing the global channel between the transmitter and the terminal. The beam that this terminal prefers is determined according to the vector h representing the overall channel. It is recalled that the preferred beam is represented by a beam vector w _pre f having, among the overall set of possible beams Ω ', the largest scalar product, in absolute value, with the vector h representing the global channel. In fact, according to the equation (1) given above, this preferred beam represented by the vector w _pre f maximizes, among the set of possible beams Ω ', the signal to interference plus noise ratio, estimated as a function of the channel vector h at step S32 of FIG. 3 and denoted SINR (h) in this FIG. 3. The search for maximum of the SINR ratio (h), knowing the channel vector h, gives, at the end of step S32 of FIG. 3, the vector of the preferred beam w _pre f.

In generic terms, it will be understood that the terminal determines the preferred beam w _pre / which maximizes, among a set of possible beams Ω ', an estimate of the signal to interference plus noise ratio SINR (wy _t /, K). In particular, the beam preferred by the terminal w _pre f is determined as a function of a global channel (vector h) between the beams from the transmitter and this terminal, the global channel h being estimated from information transmitted by the transmitter to the terminal on values of coefficients h _\ , ..., h _\ ι each representing a channel between a beam from the transmitter and the terminal.

In the following step S33, the terminal deduces, from equation (2) given above, the preferred number of beams K _pre f defined, according to this equation (2), as being the number of beams at make active to maximize the overall rate, noted KxR (w _preβ K), that the transmitter is able to transmit to all terminals. It is recalled that this overall rate KxR (w _pre f, K) is estimated as a function of the signal to interference plus noise ratio K).

It is also recalled that the signal to interference plus noise ratio SINR (My ₆ _/., K) is estimated by calculating a function C (K, M, σ ^2), inverse variation of the ratio SlNR (w _pre f, K) , and depending at least on: the number of active beams K, a maximum number of beams M that the transmitter can make active simultaneously and which is generally equal to the number of antennas to the transmitter, and the ratio σ ² between a reception noise and the power of the useful signal received by the terminal, this noise and power can be measured by the terminal.

Once the value K _pιej determined, it can for example be coded, in step S34, on a single bit signifying that the terminal: only tolerates a single active beam, or can tolerate a maximum number M beams assets. In the next step S35, the value of the preferred number K _pιe f, thus coded on a bit in the example described, is transmitted to the transmitter by the return channel, possibly with an indication of the preferred beam represented by the vector. w _{p / ef} and a Q value representing the quality of the radio link.

It will be noted that the steps illustrated in FIG. 3 are implemented by the same communicating entity, namely the user terminal. As such, the present invention also aims at such a terminal comprising means for implementing the above method (for example a storage and / or work memory, and a processor). The present invention also relates to a computer program intended to be executed by such a processor.

However, the invention is not limited, of course, to the embodiment illustrated in FIG. 3. Moreover, on its side, the transmitter BS can adjust the number K of beams made active at least according to the indications of FIG. preferred numbers of K _pιe f beams, communicated by the user terminals. As a purely illustrative example, a base station may not immediately serve a terminal indicating (especially in the embodiment where the preferred number K _pte f is coded on a single bit) that it will not tolerate the transmission of too many beams. This terminal can be served in a next burst, for example in a TDMA transmission mode, combined with an SDMA mode. As such, the present invention also aims at such a transmitter BS, then comprising means (for example still a storage and / or working memory, as well as a processor) for adjusting the number K of beams made active at least by according to the indications by the terminals of their preferred number of beams K _pre f. The present invention also relates to a computer program intended to be executed by such a processor.

The present invention also relates to a telecommunication system including at least one terminal within the meaning of the invention and an issuer within the meaning of the invention. In an advantageous embodiment, such a system may be a space division multiple access or "SDMA" system.

The present invention also relates to the metric itself, making it possible to transmit to the base station the indication of the preferred number of K _pre / beams. As such, it then targets a signal transmitted by return of a terminal to a transmitter comprising the information on the quality of reception of telecommunication data, and in particular the preferred number of K _pre f beams.

Thus, it is proposed to have a user terminal indicate its preferred mode of transmission. In particular, the terminal transmits to the transmitter on the return channel an indication of the number K _pre f he prefers, to make beams simultaneously active by the transmitter. It is then proposed an associated feedback metric. The implementation of the invention has the following advantages in particular: the choice of the transmission mode is made at the terminal and does not require additional feedback to inform the transmitter of the impact of the interference between beams on the quality of the link, the only knowledge necessary for the terminal is the overall set of beams that can be generated by the base station, - no knowledge of the channels of other user terminals and allocation decisions is necessary, the feedback metric for choosing the mode is not very complex. It can be an integer or just a single bit.

Claims

claims

A telecommunication method in a system comprising at least one transmitter (BS) arranged to simultaneously activate a first number (K) of beams, as resources for a plurality of user terminals, in which the user terminals (T1, T2 , T3, T4) receive, by said beams, telecommunication data, characterized in that at least one of the terminals transmits to the transmitter, by way of return, an indication of at least a second preferred number (K _pre f) beams (K _pre f) to be active simultaneously by the transmitter.

2. Method according to claim 1, characterized in that said second preferred number (K _pre j) is calculated by said terminal taking into account an interference noise.

3. Method according to claim 1, characterized in that said terminal:

estimates a global bit rate (KxR (w _pre f, K)) that the transmitter is capable of transmitting to all the terminals,

and determines said second preferred number of beams (K _pre f) as the number of beams to be active to maximize the estimate of said overall bit rate.

4. Method according to claim 3, characterized in that said terminal estimates said overall bit rate to be maximized (KxR (w _pre f, K)) as a function of a signal to interference plus noise ratio. K)).

5. Method according to claim 4, characterized in that the signal to interference ratio plus noise K)) is estimated by calculating a function (C [KM, (y ² ), of inverse variation of said ratio, and depending at least on:

the number of active beams (K), a maximum number of beams (M) that the transmitter can make active simultaneously,

and a ratio (σ ² ) between a reception noise and the power of the useful signal received by the terminal, measured by the terminal.

6. Method according to claim 1, characterized in that said indication of the second preferred number (K _pιe j), transmitted to the transmitter by the return channel, is coded on a single bit signifying that a terminal: only tolerates a single active beam, or - can tolerate a maximum number (M) of active beams.

A method of telecommunication in a system comprising at least one transmitter (BS) arranged to simultaneously activate a first number (K) of beams, as resources for a plurality of user terminals, in which the transmitter (BS) transmits to the user terminals, by said beams, telecommunication data, characterized in that the transmitter adjusts said first number (K) of beams made active at least according to an indication of a second preferred number (K _pιej ) of beams to be active simultaneously, transmitted by at least one of said terminals on said return path.

Terminal intended for a telecommunication system comprising at least one transmitter arranged to make simultaneously active a first number (K) of beams, as resources for a plurality of user terminals, characterized in that it comprises means for implementation of the method according to one of claims 1 to 6.

9. Computer program comprising instructions for implementing the method according to claim 1 when the program is executed by a processor.

Transmitter for a telecommunication system in which said transmitter is arranged to make simultaneously active a first number (K) of beams, as resources for a plurality of user terminals, characterized in that it comprises means for implementation of the method according to claim 7.

A computer program comprising instructions for implementing the method of claim 7 when the program is executed by a processor.

Telecommunication system comprising at least one terminal according to claim 8 and at least one transmitter according to claim 10.

13. Signal transmitted by way of return by at least one terminal to a transmitter in a telecommunication system in which the transmitter is arranged to make simultaneously active a first number (K) of beams, as resources for a plurality of terminals users, the transmitter (BS) transmitting to the user terminals (T1, T2, T3, T4), by said beams, telecommunication data, characterized in that the signal includes an indication of a second preferred number of beams (K _pre j) to be active simultaneously by the issuer.