AU1786100A

AU1786100A - Method for controlling the power of a transmitter by the signals received

Info

Publication number: AU1786100A
Application number: AU17861/00A
Authority: AU
Inventors: Guillaume Calot; Cedric Lapaille
Original assignee: Alcatel CIT SA; Alcatel SA
Current assignee: Alcatel Lucent SAS
Priority date: 1998-12-30
Filing date: 1999-12-23
Publication date: 2000-07-24
Also published as: ATE403280T1; JP2003521833A; JP4536259B2; FR2788181B1; DE69939205D1; TW443053B; WO2000041336A1; EP1017186A1; EP1017186B1; ES2312202T3; FR2788181A1

Abstract

The telecommunications receiver power control method has a transmitter (40) transmitting towards a receiver (52), and the receiver sending back a signal to the transmitter to adjust the power level. The packet of received signals is filtered and compared with a reference value. The output is sent back to the transmitter, using output from the last received signal (48) and providing a stabilized output more resistant to sharp transmitter gain changes.

Description

METHOD FOR CONTROLLING A TRANSMITTER POWER BY MEANS OF RECEIVED SIGNALS The present invention relates to a method for controlling a transmitter power so that the signal received by a receiver has a substantially constant quality. 5 The invention relates generally to a telecommunications system in which the transmitter to receiver signal propagation channel has a variable gain. It relates in particular, but not exclusively, to a telecommunications system such a satellite-based telecommunications system, in which the propagation of signals between the transmitter and the receiver may vary, for example due to a 10 variation in the weather. In such a system, the transmission channel gain variations are mainly caused by rain, scintillation and masking; such variations induce a sharp attenuation of the received signal compared with transmission under favourable conditions. Scintillation results from multiple paths of signals which cause additive and substractive 15 combinaisons. Masking occurs when an antenna tracks a moving source such as a satellite and when obstacles are located on the path of the transmitted signal. Moreover, the noise of the signal received may vary either owing to the variation of the propagation conditions or due to the fact that the noise source (whether it comes from the transmitter or from an outer interference) is variable. 20 A telecommunications system must ensure a minimum quality of service. For example, in the transmission of digital signals, this condition requires that the bit error rate is always less than a required rate. To meet this requirement, the signal to noise ratio of the received signal must at all times be greater than a pre-determined value. To solve this problem, the most common solution is to assign the 25 transmission power a sufficient value so that, regardless of the connection between the transmitter and the receiver, the signal to noise ratio obtained is always at least equal to a minimum value. However, this solution is not generally satisfactory since it entails supplying an excessive power to the transmitter, which may lead to limiting the transmission capacity of a system of which the transmitter forms a part. Indeed, when 30 transmission occurs via a satellite which relays transmissions from several transmitters, the available power in the satellite, is limited and a high power supplied to a transmitter reduces the power available for the other. In other terms, the number 2 of transmitters which may be relayed by the satellite is therefore reduced. The transmission power is therefore excessive. Another solution to the problem of variations in the transmitter power or variations in noise levels is regulation. The transmitter power is controlled by the 5 receiver. For this purpose, the signal to noise ratio of the received signal is determined; this ratio is compared with a reference value, then, from the result of the comparaison, a set point is generated from the receiver to the transmitter. The purpose of said set point is to correct the power of the transmitter so that the transmitted signal supplies the receiver, after propagation in the transmission 10 channel, with a signal whose signal to noise ratio is equal to the reference signal. To compare, in the receiver, the signal to noise ratio to the reference, in general, before comparaison, a low pass filtering type smoothing of the measurement of the signal to noise ratio is performed in order to limit the noise from the transmission channel. 15 The implementation of such regulation raises problems that are difficult to solve in the event of a substantial delay in the transmission and/or processing of signals from the transmitter to the receiver and from the receiver to the transmitter. In fact, the set point received by the transmitter is intended to correct a situation which may have shifted between the transmission of the set point by the receiver and its 20 reception by the transmitter and this situation may further evolve between the transmission of a data signal (applying the set power value) by the transmitter and its reception by the receiver. This lag between the moment of requested correction and the moment it is received by the receiver makes it impossible to regulate the power when the transmission channel gain - i.e. the propagation conditions in the present 25 example - varies substantially during the transmission time. This offset problem is exacerbated in the case of a telecommunications system with sporadic traffic (i.e. with a variable throughput) and/or in the case where the set points are transmitted from the receiver to the transmitter, only with actual information to transmit. 30 We disclose herein an example of an asynchronous transmission system such as an ATM ("Asynchronous Transfer Mode") system in which the receiver is a transmitter which sends data to the transmitter which then performs as a receiver. In this case, the set points are only transmitted with data packets or cells (of. information 3 or signalling) transmitted from the receiver to the transmitter. Since the traffic is sporadic, the time that has elapsed between, on the one hand, the application of set power value to the transmitter, and, on the other, the event generating this set point in the receiver, is variable and may reach high values. Moreover, the sporadic nature 5 of the traffic between the transmitter and the receiver entails a variable time between the moment of transmission and reception in the receiver. In particular, in a "multimedia" type transmission system, the data traffic presents highly variable throughputs. For example, the transmission of E-mail requires a much smaller throughput than the transmission of video data. 10 The invention relates to a method for controlling the power of a transmitter and offers very simple solutions to the problems above mentioned. The method according to the invention is characterised in that, whenever the receiver receives information from the transmitter, it determines a set power rating that the transmitter is required to provide according to, on the one hand, the 15 comparison between a characteristic of the received signal with a reference, and on the other hand, the power at which the received information has been transmitted and, finally, on the other hand on the set points previously transmitted but not recorded in the information received owing to transmission delays. The transmitter power is known by the receiver since the latter determines it 20 via the set point values. Once the set transmitted power value is registered by the transmitter, the stipulated power will take into account the previous received set points. The servo control may therefore operate with zero instability. The difficulty that may result from the sporadic nature of the traffic from the 25 receiver to the transmitter is solved in that the time of transmission of the set points is known in the receiver; the latter may take this into account to generate the subsequent set point. The sporadic nature of the traffic from the transmitter to the receiver also has no adverse effect since the power of the transmitter is determined by the receiver; the 30 latter knows at all times the transmission power of the signal it receives. This method therefore overcomes instability or malfunctions which could result from transmission and processing delays.

4 In one embodiment, the signal representing the power of the transmitter is stored for a time t' equal to the sum of the propagation time tP from the receiver to the transmitter, the time te taken to process or acknowledge the set point by the transmitter, and the propagation time tP from the transmitter to the receiver. This time 5 t' is known and is generally constant. The signal representing the power rating is then updated each time the set point is transmitted from the receiver to the transmitter. If the previously generated set point has not been transmitted to the transmitter whereas new information is received by the receiver, the new set point, generated from the new information received; replaces the set point which has not been transmitted. 10 In a preferred embodiment, the characteristic of the received signal is determined as follows: the instantaneous signal to noise ratio of this received signal is measured, the latter signal to noise ratio is divided by the transmission power of the received signal, the result of the division is smoothed (par example, by applying a low pass filter) and this smoothed division signal is multiplied by the'transmission power 15 of the received signal. The effect of this smoothing is to intrinsically reduce the noise of a signal. Indeed, it makes it possible to substantially reduce the noise spread across the entire spectrum, the filtering only retaining the low part of the energy spectrum. The smoothing performed on the division of the signal to noise ratio by said 20 transmission power of the received signal makes it possible to substantially improve the quality of regulation. Indeed, the instantaneous signal to noise ratio numerator which is measured is the product of the signal issued by the attenuation (or the gain) of the transmission channel (micro-wave propagation, in the example) and if the smoothing is performed on this product, it would cause power variations to be 25 integrated, which would cause oscillations of this smoothed signal to noise ratio during sudden changes of the transmission channel gain. Such oscillations of the measured signal would then supply inadequate values of the signal to noise ratio measured versus its real value, inducing an excessive set point value. Consequently, when the signal to noise ratio of the received signal is divided 30 by the transmission power, these oscillations are largely overcome, which improve the 5 precision of the measurement of the characteristic of the received signal and hence confers a not excessive value on the transmission power. This facility may be used independently from the facility involving generating the set point from previously generated and stored set points. 5 In other words, the invention consists in using the transmission power of the received signal to generate the power set point of the transmitter. The signal representing the transmission power of the received signal may be used with the previously generated set points to determine the new set point; this signal, which represents the transmission power of the received signal, may, independently or in 10 combination, be used to limit said oscillations of the measured signal to noise ratio. In particular, to generate the set point, it is not always essential to use previously generated set points; this applies in particular when for example, the propagation time is fairly low or when the set points are issued from the receiver to the transmitter at a fairly low frequency, or when each cell contains information on the power with 15 which it has been issued. In a telecommunications system for which information is transmitted from the receiver to the transmitter, since the set points are transmitted with this information, it is preferable to provide an analog system for controlling the transmission power from the receiver to the transmitter. In this case, the transmission power of the received will 20 be controlled at the transmitter. In other words, in this situation, the receiver and the transmitter each perform the dual function of transmission and reception. The invention provides for a telecommunications method which a transmitter transmits information to a receiver with a power that varies according to a set point supplied by the receiver, this set point being established by a comparison between a 25 characteristic of the received signal and a reference characteristic, the purpose of the set point being to maintain the power of the transmitter at a level so that the characteristic of the received signal is constantly equal to or similar to the reference characteristic. The method is characterised in that, as a delay is provided in the transmission of the signals between the transmitter and the receiver, the set point is 30 generated, in the receiver, whenever information is received, from, on the one hand, said comparison between the characteristic of the received signal and the reference characteristic, and, on the other hand, from a signal representing the transmission power of the received signal.

6 According to an embodiment, the set point is generated from the previously generated set points and transmitted to the transmitter, although the latter has been unable to register them owing to transmission delays. According to an embodiment, since the characteristic is a smoothed signal to 5 noise ratio in the receiver, the instantaneous signal to noise ratio of the received signal is determined, it is divided by a signal representing the transmission power of the received signal, this ratio is smoothed and the smoothed ratio is multiplied by the signal representing the transmission power of the received signal, the result of this multiplication being the characteristic which is compared to the reference. 10 According to an embodiment, the update at time t, in the receiver, of the signal representing the transmission power of the received signal occurs after a time t' has elapsed following transmission of a set point from the receiver to the transmitter, this time t' being equal to the time of the transmission delay tP from the receiver to the transmitter, of the processing time te, or acknowledgement of the set point in the 15 transmitter and the transmission delay time tP from the transmitter to the receiver; this update consists in multiplying the previously stored power by the set point transmitted at time t. According to an embodiment, the set point C(t) is generated from the following formula: T (t) 20 C(t) = ' x ?(t) T x (t + t'V wherein y, is the value of the reference characteristic, ?(t) the value of the characteristic measured at time t in the receiver, and Tx(t) and Tx(t+t') signals representing the transmission power of the signal received at times t and t+t' respectively. 25 According to an embodiment, the set point C(t) is generated from the following formula: C(t) f 1 7(t) r c 7 Wherein Tref is the value of the reference characteristic, i(t) the value of the characteristic measured at time t in the receiver, and [IC the set point or product of the set points previously transmitted but not yet registered by the transmitter. According to an embodiment, the set point transmitted by the receiver to the 5 transmitter is transmitted simultaneously with information or signalling data. According to an embodiment, the data transmitted from the receiver to the transmitter is in the form of digital data in cells or packets, each set point being transmitted at the top of the cell or packet. According to an embodiment, the information is transmitted from the 10 transmitter to the receiver being digital data transmitted by cells or packets, the characteristic of the received signal is determined at each cell. According to an embodiment, the flow of information from the transmitter to the receiver or from the receiver to the transmitter is of the sporadic type. According to an embodiment, since the purpose of the transmitter is also to 15 receive information from the receiver and since the purpose of the receiver is to transmit information to the transmitter, the transmission power control of the receiver is performed from a set point supplied by the transmitter. In addition, the present invention provides for a receiver to implement the telecommunications method of the invention. As this receiver also transmits signals to 20 the transmitter, it includes means to generate set points and memory means to store signals representing the transmission power of the received signal, the value of the signal representing the transmission power of the stored and received signal being updated whenever signals are transmitted from the receiver to the transmitter. According to an embodiment, the receiver includes a circular memory with a 25 capacity t, wherein t' is the sum of the transmission delay tP from the receiver to the transmitter, the processing time te registered in the transmitter and the transmission delay time tP from the transmitter to the receiver. Finally, the present invention provides for an application of the telecommunications method according, pursuant to the invention, to a satellite-based 30 telecommunications system, in which is provided comprising a control station and a plurality of terminals, the terminal and control station communicating via the satellite.

8 Other features and advantages of the invention will become more apparent in light of the description of certain embodiments there of with reference to the accompanying drawings, in which: figure 1 is a diagram showing a satellite-based telecommunication system 5 by satellite, figure 2 is a diagram showing known devices regulating the transmission power from a receiver, figure 3 is a diagram explaining the defects of the system of figure 2, figure is a diagram showing a control system according to the invention, and 10 figures 5 and 6 are diagrams explaining some operating features of the system illustrated in figure 4. The example that will be described in relation to the figures pertains to a telecommunications system in which the surface of the terrestrial globe is divided into zones 10 (figure 1) of which only one has been illustrated on the figure In each zone 15 there is, on the one hand, a central control or connection station 20, and, on the other, terminals or subscriber stations 16, 18, etc. Terminals 16, 18, etc. communicate via a low or medium orbit satellite 14. In the example, the altitude of the satellite is about 1,500 km. This satellite 14 moves through an orbit 12 in which other satellites are located. To cover the terrestrial globe 20 or a large part of the latter, several orbits 12 are provided. As soon as satellite 14 looses slight of zone 10, the subsequent satellite (not shown), par example on the same orbit 12, takes over the communication. The control and connection station 20 controls communications between terminals 16, 18, etc. In particular, it assigns frequency, power and code resources 25 for each terminal. For this purpose, each station 20 communicates with each terminal, also via satellite 14. The communications between terminals are performed via station 20. In other words, when terminal 16 communicates with terminal 18, terminal 16 sends data to station 20 via the satellite and station 20 re-transmits this data to terminal 18, 30 also via the satellite. Station 20 is connected to a terrestrial network 22, of the ATM type in the example. This station 20 is this connected via a ATM switch 34 to a wide band network 36, a narrow band network 38, and to servers 28. The narrow band network 9 38 enables the connection of users 30 and servers 24. Moreover, the large band network 36 enables the connection of users 32 and servers 26. Such a telecommunications system of the ATM type enables a high throughput of data with a high capacity and short delay due to transmission. 5 In an asynchronous network, especially of the ATM type, the data is in digital form and organised in packets or cells including, for the ATM standard, 384 bits (or symbols) of data and 40 bits (or symbols) of header. The problem addressed by the invention is to assign to each communication from the control station 20 to a terminal 16, 18 and from a terminal 16, 18 to the 10 control station, a power resource just necessary so that the signal to noise ratio meets the specifications. Consequently, the communications will have the required service quality guaranteeing a bit error rate less than a pre-determined limit without excessive power consumption. Indeed, the power issued must be just the amount required since the power available in the satellite is limited and if a communication 15 requires more power, the excess power is deducted from the power attributed to the other communications. Moreover, since the terminals are wide distribution devices whose price must be as low as possible, they should preferably have a low power. To control the transmission power, the signal to noise ratio is determined on 20 reception and the transmission power is regulated so that the signal to noise ratio is equal to a reference value. In the case of a satellite-based transmission system, the power received by the receivers may vary to a large extent, in particular owing to variations in propagation conditions due to random changes in the weather. In particular, the 25 propagation deteriorates substantially when it rains as opposed to clear weather. The propagation also deteriorates as a result of scintillation and masking. Propagation may also introduce noise and the noise source may feature variable characteristics. Besides thermal noise, the causes of noise may be interference due to the use of the same transmission frequencies for adjacent zones 30 or jamming by other transmission systems. The slaving of the transmission power to the signal to noise ratio at reception raises a difficult problem to solve in the case of the satellite-based transmission system owing to the propagation times of the set point transmitted by the receiver to 10 the transmitter and the time taken for transmission of signals from the transmitter to the receiver. In the example, the delay due to signal propagation from a terminal to the control station or from the control station to a terminal is about 25 ms. This time 5 frame is maintained constant by means of buffers. The actual propagation time varies constantly owing to the displacement of the satellite with respect to the zone, which causes a variation in the propagation distance between a terminal and the control station. However, in order to facilitate system management, the delay is maintained at a constant level by means of the above-mentioned buffers. 10 Moreover, since the signal to noise ratio is measured on received cells, the measurement frequency depends on the traffic flow which intrinsically varies within a multimedia communication system of the type described. Par example, the transmission frequency of the cells is lower in the case of E-mail than in the case of image or programme transmissions. 15 Moreover, it is preferable that the power set point transmission from the receiver to the done be done at the same time as the transmission of a cell since, as this set point only requires a small number of bits, it is preferable not to use a full cell only for transmission of the set point. It is therefore necessary to wait until (data or signalling) information has to be transmitted from the receiver to the transmitter to 20 transmit this set pint. The sporadic nature of the traffic further increases the regulation difficulty since it entails non-deterministic, i.e. unforeseeable delays. Consequently, a standard servo control system, where used as it, cannot operate correctly. Such a system is represented in figure 2. This figure shows that the transmitter 40 comprises an input 42 receiving a set point from a receiver 44. The 25 transmission time between the control output 46 of the receiver 44 and the set point input 42 of the transmitter 40 corresponds to a delay symbolised by block 48. Moreover, the transmission of cells from the transmitter 40 to the receiver 44 is performed, as in the reverse transmission, by micro-wave link via satellite. This transmission provides for a channel 50 that also causes a delay. In the receiver 44, 30 cells are received by a receiving device 52 and the signal to noise ratio y - is No constantly calculated (stack 54) on each cell received.

11 To limit the noise introduced by the channel 50, smoothing is performed (block 56'), i.e. in the example, through a low pass filter, on the signal supplied by block 54. This smoothed signal to noise ratio ? is compared at block 58, called a "decision block", to a reference value Yref applied at its input 60. Block 58 delivers, at 5 the output 46, the set point to the transmitter 40 so that the latter adjusts its transmission power according to the comparison between the smoothed signal to noise ratio and the reference value Yref The smoothing 56' introduces a constraint on the transmission power. Indeed, this smoothing causes oscillations on the measured value during sudden 10 variations of the input signal. These oscillations, which do not reflect actual oscillations of the received signal to noise ratio, are applied at block 58 and, therefore during one oscillation, this block 58 receives values i which my be less than the reference Yref, whereas the real signal to noise ratio does not fall below the reference value. Consequently, the transmission power must usually be chosen with 15 an excess value to take these measurement oscillations into account. As mentioned below, the invention, in one of its aspects, enables these unwanted oscillations to be minimized. Moreover, as will be described in relation to figure 3, the research conducted under the present invention reveal that the delays in the servo control loop 20 illustrated in figure 2 prevent correct operation of this loop and this research has helped to understand the malfunction of the servo control and to propose a solution for this. For the present project, a simplified example has been considered in which it has been assumed that the power Pe has, at the beginning of the communication, a 25 value of 1 and channel 50 presents an attenuation which increases continuously over time, this attenuation increasing by a factor a at each time unit. In other words, at the end of a time unit, the power is attenuated by a factor a, at the end of two time units the power received is attenuated by the factor a 2 and, at the end of n time units, the power is attenuated by the factor an. Moreover, in this example, the propagation time 30 from the transmitter to the receiver and from the receiver to the transmitter is one time 12 unit, the set point is issued by the receiver 44 one time unit after it has been received and, in analog mode, the transmission power is updated one time unit after reception of the set point by the transmitter. Finally, the value of the noise No is assumed to be that of the unit. 5 The upper line 62 (figure 3) symbolises the variation of the power Pe of the transmitter 40 versus time, the time units being marked from 0 to 15 on the lower horizontal line 64 which symbolises the receiver. Under line are 64 symbolised the values y of he signal to noise ratio measured at each time unit and the last line symbolises the set points transmitted by 10 the receiver to the transmitter. At time 0, the transmitter transmits at power 1. At time 1, the receiver receives the power a (owing to the attenuation a by time unit, transmission from the transmitter to the receiver being symbolised by an oblique arrow from top to bottom with mention of the attenuation). In these conditions, the receiver issues an increase 15 set point to offset the attenuation. This set point is therefore an increase request in a reverse ratio, i.e. al. However, this set point is only registered by the transmitter at time 4 owing to the above-mentioned delays. In this conditions at time 1, the transmission power is still 1, and the powers received (here, equal to y) at times 2, 3 and 4 are a 2 , a 3 and a 4 respectively, which causes, at these moments, the 20 transmission points a-2, a- 3 and a- 4 respectively. At time 5, the receiver receives a power a 5 .a -1 = a 4 , hence a set point a-4. At this time 5, the transmission power is a- 3 , which causes a receiving power of a 3 at time 6. This operating mode reveals clear instability of the transmission power. At time 14, for instance, the transmission power is a 3 , whereas it should be a~ 25 (considering the above-mentioned delays). This dysfunction is due to the fact that the set points are generated each time according to the received signal without taking into account the previous set points which were not acknowledged by the transmitter owing to transmission and processing delays. 30 For example, figure 3 shows that the set point sent at time 6 by the receiver has a value of a- 3 . However, this set point, which will be executed at time 9 by the 13 transmitter does not take into account the fact that the signal received at time 6 corresponds to the signal issued at time 5 and that at times 6, 7 and 8, the transmitter will have changed power versus time 5. Figure 4 illustrates a servo control design to solve the above two problems 5 i.e. to minimize oscillations of the measured parameter on the one hand, and, on the other, to exercise proper, reliable and strong servo control of the transmission power according to the measurement performed in the receiver. In this figure 4, the elements corresponding to those of figure 3 bear the same reference numbers. The fixtures illustrated in figure 4 differs from the one 10 illustrated in figure 2 in that, according to the invention, in receiver 44, a block 70 is provided for controlling the filtering block 56' and the set point transmission block 58'. It is possible, by means of block 70, to determine in relative value the power at which was issued the cell on which the signal to noise ratio is instantaneously 15 calculated at block 54. The signal representing the power with which the cell received, measured and processed at block 54, 56' and 58' was issued, may be determined in the receiver 44 since the latter determines the power of the transmitter 40. In order to determine the power with which the cell has been issued, account 20 is taken of the set points previously sent by the receiver 44 to the transmitter 40. In other words, contrary to the situation illustrated in figures 2 and 3, the set point sent by the receiver to the transmitter takes into account both the received signal and the transmission power of the cell. Moreover, the power set point requested at a given moment takes into 25 account the set points previously sent by the receiver but which have not yet been acknowledged by the transmitter. The invention therefore stabilises control despite the inevitable delays between the generation of the set point and its acknowledgement by the transmitter. Block 70 (figure 4) presents an output 72 connected to the input of block 56' 30 and delivers to the latter information denoting the value of the power T" at which the cell to be filtered at block 56' has been issued.

14 Eb At this block 56', prior to filtering, the signal i delivered by block

N

0 54 is divided by the transmission power value T. Given that fact that Eb = GT., the signal filtered by block 56' is the signal

G/N

0 , where in G is the gain (or attenuation) of the channel 50, i.e. it only 5 corresponds to the transmission channel attenuation signal to noise ratio. This system is designed to dampen oscillations of the filtered signal during changes in the gain gradient G. In this way, the power of the transmission signal may be reduced in comparison with cases where such oscillations occur. Finally, block 56' multiplies the signal G/NO, again smoothed by the value T, 10 of the transmission power to be able to compare this signal to the reference Yref at block 58'. Moreover, block 70 comprises an input/output 74 connected to an input/output 76 of decision block 58. The latter generates the power set points to be issued on the output 46 and supplies on the input 74 of block 70 information on 15 when the set point is transmitted, i.e. on the time when the output 46 transmits the set point to the transmitter. To generate the set point, block 58' registers all the set points previously issued and which are not taken into account by the received cell owing to the propagation and processing times. More specifically, the received cell which is 20 processed at blocks 54, 56' and 58', takes into account the set points issued until time t-t' and not the set points issued between the moments t-t' and t. t is the present moment and t' the sum of the signal propagation time tP from the receiver 44 to the transmitter 40, the signal processing time te in the transmitter 40 and, finally, the propagation time tP from the transmitter 40 to the receiver 44. 25 The time t has a constant value since, on the one hand, as previously mentioned, a constant value is attributed to the propagation time tP via the buffers and, on the other hand, the processing time t e in the transmitter 40 is also constant.

15 Even if no buffer is provided to impose a constant transmission time between transmission by the receiver (or the transmitter) and reception by the transmitter (or the receiver), the time tP is known at each moment since the distances separating the transmitter and the receiver from the satellite are also known at each moment. 5 To determine, at moment t, the power at which a received cell has been issued, the power set point requested power (by the signal on output 46) at moment t-t must be known. In the current example, the signal issued on the output 46 is a multiplying factor. For example, if, at time t, the power of the transmitter is PO, the set point will 10 be a number C 1 so that the cell received at moment t+t' is the power P 0 C. This example only applies if between moments t-t' and t no set point has been issue. If during this time interval other set points have been issued, for example C 2 and C 3 , the set point issued will always be C 1 , although in this case the transmission power of the cell received at time t+t' will be P = POC 1

C

2

C

3 15 Consequently, the transmission power of each cell received may be known at any time provided the product of the set point issued until moment t-t' is known. It is not essential to know the transmission power of the transmitter at initialisation; only the variation of this power, i.e. the variation of the set points, needs to be known. In the embodiment described, to limit the size of the memory of block 70, 20 only the requested powers are kept in the memory between times t-t' and t which correspond to the transmission powers of the cells received between times t and t+t'. In the current example, the set points are issued only when a (information or signalling) communication is established from the receiver 44 to the transmitter 40. "Communication" here denotes any type of transmission of information, which 25 encompasses both the transmission of actual data and the transmission of signalling data such as data indicating a change of satellite, the position of the satellite or verification data. In this case, the set point is issued at the top of one of the communication cells. Consequently, transmission of the set point is not necessarily immediate. It may 30 happen, therefore, that a cell is received by the receiver 44, whereas the set point 16 corresponding to the previously analysed cell has not been issued to the transmitter 40 (since in the interim no traffic has occurred from the receiver to the transmitter). In this situation, the new set point, generated from the last cell received, replaces the previous set point. 5 In the light of the all the above explanation, the value of the set point C(t) formed by block 58' after the reception of each cell is: y(e T, (t) y (e 1 (1) C(t) - Yrf T e y(t) T, (t + t' ) (t) 1-IC in the above formula, C(t) is the set point, i.e. a multiplication factor, yref is the reference value applied to the input 60 of block 58', y (t) is the output signal of 10 the filtering block 56' applied to an input of block 58', T.(t) is the image at time t, at block 70, of the transmitter power at time t-tP and T,(t+t') is the image at time t+t' of the transmitter power at time t+t'-tp. The latter image is known at moment t since it depends on the set points issued by the receiver up until this moment. The ratio TX corresponds to the reverse of the product of the set Tx (t + t') 15 points of power issued by the receiver and that have not yet been taken into account by the transmitter. Indeed, it is not necessary to take into account several times these previously generated set points. HC is this product in the above formula (1). Moreover, the value T, (image of the transmission power) is updated in the memory of block 70 whenever a set point is issued; this block 70 being informed of 20 this transmission by the signal at its input 76. This update involves, after a time t' has elapsed after transmission of the set point, multiplying the previous value T. by the set point issued. In other words, at time t, the set point is generated on the basis of the power of the cell issued at time t =t-tp-tr (tr is the duration of the measurement and generation of the set point in the receiver), and this set point issued at time t will be 25 applied by the transmitter at time t 1 +t'+ tr and will only be received by the receiver before time t 1 +t'+tp+tr=t+t'. Under such conditions, to update the power Tx, the following procedure is followed: when a set point is transmitted a time t, 'at the output 17 46, by a cell, the stored power T, of block 70 is updated at time t+t'. This update at time t+t' involves multiplying the stored value T. by the set point issued at time t. As previously mentioned, when a new set point is generated on the basis of a received cell, while the previous set point has not be issued, the new set point 5 replaces the one that could not be sent. An example of an operating mode is illustrated in figure 5 which is a figure similar to figure 3. This figure, as in figure 3, considers a transmitter whose initial power has a value of 1 and a transmission channel which weakens by a coefficient a every time 10 unit. The transmission time, one way or the other, between the transmitter and the receiver is one time unit. The processing time in the receiver is one time unit. On the other hand, with regard to figure 3, for the sake of simplification, it has been assumed that the processing time in the transmitter is slight. The upper line 100 shows the variation in the power transmitted by the transmitter over time. The lower 15 line 102 represents the receiver. The area beneath this line shows the variations in the value ? over time. Also shown is the variation over time: of the set points pertaining to the power C(t), of the image T. of the power of the transmitter 40 formed at block 70, and, finally, of the set points of power sent by the receiver 44 to the transmitter 40 but which have not yet been registered by the transmitter 40. 20 At a given time t,, the image T, of the power Pe is the image of the power of a cell transmitted at time t-t, (t-1 in the current example). In the present operating example, the transmission of cells from the transmitter to the receiver (oblique lines, from line 100 to line 102) is irregular. Consequently, between moments 0 and 2, between moments 4 and 7 and between 25 moments 10 and 14, transmission is performed at the rate of one cell per time unit, whereas between the moments 2 and 4, two time units elapse; the same applies between moments 8 and 10. Similarly, the transmission of cells from the receiver to the transmitter (horizontal dotted lines and oblique lines from line 102 to line 100). Transmission occurs at each time unit between times 2 and 6, and between times 10 18 and 14. However, four time units separate the transmission of one cell between moments 6 and 10. For figure 5, the above formula (1) has been used to determine the set points C(t) and the value T, updated has also be used and consists in multiplying the 5 value Tx present at time t by the set point at time t-t'. In this example, the value at time t' is two time units (twice the propagation time, assuming that the processing time te is zero). For example, at time 7, the set point is: 1 1 2 C(7) a -__ C a4 a- 2 10 In the above calculation, is the reverse of the set point still not "- 2 received by the transmitter which is indicated by the last line at time 7. The values T (t) and T,(t+t') may also be used. Par example: 1 Tx(7) 1 a-3 2 C(7) = T,(9 = a

GC

4 T, (9) 11 4 a- 5 To generate the value T. at time 7, the value present in memory 70, 15 indicated at time 6, is considered, and this value is multiplied by the set point present at time 6-t', i.e. at time 4. At time 4, the set point was 1. The value T, then remains X- 3 at time 7. At times 9, 10 and 11 the value T, is maintained constant since at times 7, 8 and 9 (i.e. at times t-t), no set point has been issued by the receiver. It is noteworthy that despite the sporadic nature of the traffic, the power Pe 20 correctly follows the attenuation due to propagation (notwithstanding propagation and processing delays), despite the irregularity of the traffic. The latter also causes irregularities in the transmission power variations, although such irregularities (e.g. from time 10 to time 11) do not impact the stability of the servo control.

19 It is also noteworthy that the process is resistant to disturbances such as signal losses from the receiver to the transmitter, i.e. disturbances such as signals issued by the receiver 40 are not received by the transmitter 44. Figure 6 illustrates the effect of a loss of cells from the receiver to the 5 transmitter. This example is similar to the one represented in figure 5. The assumptions are the same. This example shows that the cells issued by the receiver at times 6, 7, 8 and 9 have not reached the transmitter, whereas the latter should have received set points at times 7, 8, 9 and 10 respectively. 10 In this situation, the power Pe reaches the correct value at time 13, i.e. only two time units after the end of the interruption of the transmission from the receiver to the transmitter. An embodiment of block 70 involves providing means for storing the image of the transmission power of the received cells and updating this power via a clock or 15 timer which updates this value to the time which is triggered upon transmission of a set point and which updates this value to the time t' = t + te + tP, the new value representing the previous one multiplied by the set point during transmission of the set point. If, at the arrival of a new cell, the previous set point has not been transmitted because no cell has been transmitted from the receiver to the transmitter, 20 the new set point established from the last received cell replaces the previous set point. Block 70 may use, for example, a circular buffer with a capacity of t. Although only the power control at the receiver has been described, it is easy to understand that there is another power control loop for which the receiver 25 transmission power is controlled at the transmitter.

Claims

1. A telecommunication method, wherein: - a transmitter (40) transmits information to a receiver (44) with a power that varies according to a set point supplied by the receiver, 5 - this set point is established from a comparison between a characteristic (?) of the received signal and a reference characteristic (7ref), - the purpose of the set point is to maintain the power of the transmitter at such a level that the characteristic of the received signal is constantly equal to or similar to the reference characteristic, 10 - since a delay occurs in the transmission of signals between the transmitter and the receiver, the set point is generated in the receiver whenever information is received, from, on the one hand, said comparison between the characteristic of the received signal and the reference characteristic and, on the other hand, from a signal representing the transmission power of the received signal 15 characterised in that the set point is generated from the set points previously generated and transmitted to the transmitter but which the latter could not registered owing to the transmission delays.

2. A method according to claim 1, characterised in that since the characteristic is a smoothed signal to noise ratio, in the receiver, the instantaneous signal to noise 20 ratio (i) of the received signal is determined, it is divided by a signal (Tx) representing the transmission power of the received signal, this ratio is smoothed and the smoothed ratio is multiplied by the signal (T.) representing the transmission power of the received signal, the result (?) of this multiplication being the characteristic which is compared to the reference. 25

3. A method according to claim 1 or 2, characterised in that the update, at time t, in the receiver, of the signal (Tx) representing the transmission power of the received signal is made after a time t' has elapsed following transmission of a set point from the receiver to the transmitter, this time t' being equal to the sum of the transmission delay tP from the receiver to the transmitter, the processing time 21 te or acknowledgement of the set point in the transmitter and the transmission delay time tP from the transmitter to the receiver, and in that this update involves multiplying the previously stored power by the set point issued at time t.

4. A method according to any one of the previous claims, characterised in that the 5 set point C(t) is generated from the following formula: C(t) = yf -T (t) i(t) Tx (t + V ) wherein Tref is the value of the reference characteristic, y (t) the value of the characteristic measured at time t in the receiver, Tx(t) and T.(t+t') are the signals representing the transmission power of the signal received at times t and t+t' 10 respectively.

5. A method according to any one of claims 1 to 3, characterised in that the set point C(t) is generated from the following formula: C(t) ref 1 i(t) IC in which Yref is the reference characteristic value, y (t) the value of the 15 characteristic measured at time t in the receiver and IC the set point or product of the set points previously issued but not yet registered by the transmitter.

6. A method according to any one of the previous claims, characterised in that the set point transmitted by the receiver to the transmitter is transmitted simultaneously with information or signalling data. 20

7. A method according to claim 6, characterised in that the data transmitted from the receiver to the transmitter is in the form of cells or packets of digital data, in that each set point is transmitted into the header of the cell or packet.

8. A method according to any one of the previous claims, characterised in that the information transmitted from the transmitter (40) to the receiver (44) being digital 25 information transmitted by cells or packets, the characteristic of the received signal is determined at each cell. 22

9. A method according to claim 6, 7 or 8, characterised in that the traffic flow of information from the transmitter to the receiver or from the receiver to the transmitter is of the sporadic type.

10. A method according to any one of the previous claims, characterised in that since 5 the transmitter (40) is also intended to receive information from the receiver (44) and since the receiver is intended to transmit information to the transmitter, the transmission power of the receiver is controlled from a set point supplied by the transmitter.

11. A receiver design to implement the telecommunications method according to any 10 one of the previous claims, characterised in that, since this receiver (44), also transmits signals to the transmitter (40), it has also a means (58') to generate set points and memory means (70) to store signals representing the transmission power of the received signal, the signal representing the transmission power of the stored and received signal being updated each time signals are sent from the 15 receiver to the transmitter.

12. A receiver according to claim 11, characterised in that whereby it comprises a circular memory (70) with a capacity of t', wherein t' is the sum of the transmission delay tP from the receiver to the transmitter, the processing time te calculated in the transmitter and the transmission delay time t from the 20 transmitter to the receiver.

13.Application of the method according to any one of the claims 1 a 10 to a satellite-based telecommunications system in which a control station (20) and a plurality of terminals (16, 18) are provided, the terminals and control station communicating via the satellite.