Pre-distortion of an optical signal with frequency division multiplexed sub-carriers

Abstract

A method is provided for pre-distorting an optical signal, intended to be transmitted over an optical transmission line and including a plurality of frequency division multiplexed optical sub-carriers. The optical signal is obtained from an optical conversion of an electrical signal including a plurality of frequency division multiplexed sub-carriers. This method includes obtaining a plurality of corrective factors, corresponding to the electrical sub-carriers, as a function of a plurality of values, determined for the optical sub-carriers, of a first parameter dependent on the power of the sub-carriers, and application of the corrective factors to the amplitudes of the corresponding electrical sub-carriers, so as to equalize, at the output of the optical transmission line, the values of the first parameter for the sub-carriers of the optical signal.

Claims

1. A method for pre-distorting an optical signal, intended to be transmitted over an optical transmission line having a configuration and comprising a plurality of frequency division multiplexed optical sub-carriers, said optical signal being obtained from optical conversion of an electrical signal comprising a plurality of frequency division multiplexed electrical sub-carriers, the method comprising: storing several sets of corrective factors in association with several possible optical transmission line configurations; recovering, from among the stored sets, a set of corrective factors for the respective electrical sub-carriers, which are associated with a configuration corresponding to the configuration of the optical transmission line, each of the corrective factors of the recovered set being a function of a first parameter that is dependent on the power of the respective sub-carrier; and applying the recovered set of corrective factors to amplitudes of the respective electrical sub-carriers, so as to equalize, at an output of said optical transmission line, values of the first parameter for the sub-carriers of the optical signal.

2. The method according to claim 1, further comprising a preliminary act of measuring the values of the first parameter at the output of the optical transmission line and calculating the corrective factors as a function of the measured values of the first parameter before being stored.

3. The method according to claim 1, in which the first parameter is the signal-to-noise ratio (SNR.sub.i) of the respective optical sub-carrier.

4. The method according to claim 3, in which the corrective factor .sub.A,i corresponding to the ith sub-carrier of the electrical signal is calculated as follows: ${}_{A,i}=\sqrt{\frac{1}{{\mathrm{SNR}}_i}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{1}{{\mathrm{SNR}}_i}\right)}}$ in which SNR.sub.i is the signal-to-noise ratio of the ith optical sub-carrier at the output of the optical transmission line and N is the number of sub-carriers of the optical signal.

5. The method according to claim 1, in which the first parameter is the power of the respective optical sub-carrier.

6. The method according to claim 5, in which the corrective factor .sub.A,i corresponding to the ith sub-carrier of the electrical signal is calculated as follows: ${}_{A,i}=\sqrt{\frac{P_{\mathrm{SC},0}}{P_{\mathrm{SC},i}}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{P_{\mathrm{SC},0}}{P_{\mathrm{SC},i}}\right)}}$ in which P.sub.SC,0 is the power at the output of the optical transmission line of a reference sub-carrier from among the sub-carriers of the optical signal, P.sub.SC,i is the power at the output of the optical transmission line of the ith sub-carrier of the optical signal and N is the number of sub-carriers of the optical signal.

7. A device for pre-distorting an electrical signal comprising a plurality of frequency division multiplexed sub-carriers, intended to be converted into an optical signal transmitted over an optical transmission line having a configuration and comprising a plurality of frequency division multiplexed optical sub-carriers, said device comprising a processor configured to: store several sets of corrective factors in association with several possible optical transmission line configurations; recovering, from among the stored sets, a set of corrective factors for the respective electrical sub-carriers, which are associated with a configuration corresponding to the configuration of the optical transmission line, each of the corrective factors being a function of a first parameter dependent on the power of the respective sub-carrier; and apply the recovered set of corrective factors to the amplitudes of the respective electrical sub-carriers, so as to equalize, at an output of said optical transmission line, values of the first parameter for the sub-carriers of the optical signal.

8. The device according to claim 7, further comprising storage means suitable for storing the several sets of corrective factors, the processor being arranged to access the storage means in order to recover the corrective factors to be applied to the respective amplitudes of the sub-carriers of the electrical signal.

9. A method for pre-distorting an optical signal, intended to be transmitted over an optical transmission line and comprising a plurality of frequency division multiplexed optical sub-carriers, said optical signal being obtained from optical conversion of an electrical signal comprising a plurality of frequency division multiplexed electrical sub-carriers, the method comprising: obtaining a set of corrective factors, each corrective factor corresponding to one of the electrical sub-carriers and being a function the power or the signal-to-noise ratio of the respective optical sub-carrier; and applying the set of corrective factors to amplitudes of the respective electrical sub-carriers, so as to equalize, at an output of said optical transmission line, the powers or the signal-to-noise ratios of the sub-carriers of the optical signal.

10. The method according to claim 9, in which the set of corrective factors are functions of the signal-to noise ratios of the respective optical sub-carriers, and the corrective factor .sub.A,i corresponding to the ith sub-carrier of the electrical signal is calculated as follows: ${}_{A,i}=\sqrt{\frac{1}{{\mathrm{SNR}}_i}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{1}{{\mathrm{SNR}}_i}\right)}}$ in which SNR.sub.i is the signal-to-noise ratio of the ith optical sub-carrier at the output of the optical transmission line and N is the number of sub-carriers of the optical signal.

11. The method according to claim 9, in which the set of corrective factors are functions of the powers of the respective optical sub-carriers, and the corrective factor .sub.A,i corresponding to the ith sub-carrier of the electrical signal is calculated as follows: ${}_{A,i}=\sqrt{\frac{P_{\mathrm{SC},0}}{P_{\mathrm{SC},i}}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{P_{\mathrm{SC},0}}{P_{\mathrm{SC},i}}\right)}}$ in which P.sub.SC,0 is the power at the output of the optical transmission line of a reference sub-carrier from among the sub-carriers of the optical signal, P.sub.SC,i is the power at the output of the optical transmission line of the ith sub-carrier of the optical signal and N is the number of sub-carriers of the optical signal.

12. The method according to claim 9, further comprising a preliminary act of measuring the values of the powers or the signal-to-noise ratios of the sub-carriers of the optical signal at the output of the optical transmission line and calculating the corrective factors as a function of the measured values before being stored.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Other features and advantages will become apparent on reading, in the following detailed description, about particular embodiments, given as nonlimiting examples, and from the attached figures in which, in addition to FIGS. 1A and 1B that have already been discussed:

(2) FIG. 2 illustrates an optical transmission system in which an embodiment of the present disclosure can be implemented;

(3) FIG. 3 illustrates a pre-distortion device according to an embodiment of the present disclosure; and

(4) FIG. 4 illustrates a pre-distortion method according to an embodiment of the present disclosure.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

(5) Reference is first of all made to FIG. 2 which illustrates an optical transmission system in which an aspect of the present disclosure can be implemented.

(6) This transmission system comprises optical transmitters (a single transmitter 10 being illustrated here, in a nonlimiting manner) connected, via an optical wavelength division multiplexer 20, to the input of an optical transmission line 30. The output of this optical transmission line 30 is connected to a plurality of optical receivers (just one receiver 50 being illustrated here, in a nonlimiting manner), via an optical wavelength division demultiplexer 40.

(7) In such a system, each optical transmitter transmits an optical sub-band comprising N frequency division multiplexed sub-carriers (N being equal to 2.sup.k, k being an integer, for example N=128, 256, 512, . . . ), and the different optical sub-bands are wavelength division multiplexed in order to construct an optical signal of super-channel type and are thus transmitted over the line 30 before being wavelength division demultiplexed at the output to be directed to their respective optical receivers.

(8) The optical transmission line 30 typically consists of an alternation of an optical fibre with various optical components having spectral transfer functions which are specific to them. FIG. 2 illustrates a particular case in which this line 30 comprises the repetition of n (n being an integer such that n1) similar sets (or spans) consisting of an optical fibre (for example of SSMF type) followed by an optical amplifier (for example of EDFA type) which is in turn followed by a reconfigurable wavelength optical add/drop multiplexer ROADM. In other words, the transmission line 30 comprises n cascaded ROADM multiplexers, which exhibit a cumulative spectral transfer function likely to greatly affect the performance of an OFDM optical signal passing over this line 30, because of the imperfection of the rectangular nature of the spectral transfer functions of these multiplexers.

(9) The optical transmitter 10 is, for its part, configured to transmit an OFDM optical signal, that is to say an optical signal (denoted S.sub.OPT) made up of a plurality N of frequency division multiplexed optical sub-carriers, from an electrical signal S.sub.E which is itself made up of the same number N of frequency division multiplexed electrical sub-carriers. As already indicated, such an OFDM optical signal can notably correspond to a constituent optical sub-band of an optical signal of super-channel type as described previously.

(10) More specifically, the transmitter 10 comprises electro-optical conversion means 15 capable of converting the electrical signal S.sub.E into an optical signal S.sub.OPT, these means typically being implemented in the form of an electrical amplifier (RF driver), powered by the electrical signal S.sub.E so as to control an optical modulator (for example of C-MZM type) in order for the latter to modulate an optical signal generated by an optical source (not illustrated). The N optical sub-carriers of the optical signal S.sub.OPT are thus obtained from the electro-optical conversion of the N electrical sub-carriers of the electrical signal S.sub.E.

(11) In order to take account of the impact of the optical devices present on the line 30, notably of the multiplexers of ROADM type, the transmitter 10 further comprises a pre-distortion device 11, in the form of a module inserted upstream of the electro-optical conversion means 15, which acts on the electrical signal S.sub.E before the latter is converted into an optical signal S.sub.opt. In other words, the signal processing that makes it possible to neutralize the effects of the multiplexers of ROADM type of the line 30 on the optical signal is performed in the electrical domain.

(12) Reference is now made to FIG. 3 which more particularly illustrates a pre-distortion device according to an embodiment of the present disclosure.

(13) This pre-distortion device 11 comprises processing means 12, which typically consist of a processor, implemented for example in the form of a DSP module. These processing means 12 generate, from useful data received as input, a plurality of N sub-carriers in digital form and supplies these sub-carriers, in parallel, to a digital-analog converter 13, which converts them into an analog signal intended to be supplied to the electro-optical conversion means 15 of the optical transmitter 10.

(14) These processing means 12 are arranged in particular to apply, to the amplitude of each ith electrical sub-carrier generated, a corrective factor .sub.A,i, specific to this ith sub-carrier since the amplitude of the ith electrical sub-carrier is expressed in digital form, the corrective factor .sub.A,i, is also expressed in digital form.

(15) In other words, where an identical nominal transmission power P.sub.S,SC is normally used for the different electrical sub-carriers generated from an OFDM signal, the amplitude of each sub-carrier is multiplied here by the corrective factor .sub.A,i specifically corresponding to this sub-carrier in the processing means 12, so that the signal S.sub.E exhibits a distorted spectrum (i.e. not flat) at the output of the processing means 12.

(16) On account of this, in the device 11, the processing means 12 can be associated with a read-only memory (not illustrated) forming a storage medium, that can be read by these processing means on which is stored a computer program according to an aspect of the present disclosure, comprising instructions for the execution of the steps of a pre-distortion method according to an aspect of the present disclosure, these steps being described later with reference to FIG. 4.

(17) In addition to the abovementioned processing means 12, the device 11 advantageously comprises storage means 14, typically implemented in the form of a storage table (or look-up table, LUT), capable of storing a certain number of corrective factors .sub.A,i, that can be applied to the amplitude of the electrical sub-carriers by the processing means 12.

(18) In particular, a single series of corrective factors can be stored therein, when these corrective factors are calculated from measurements performed specifically for the line 30 concerned.

(19) Alternatively, several series of corrective factors corresponding to different possible transmission line configurations can be stored therein beforehand, in which case the processing means 12 access the storage means 14 to check therein as to whether there is a series of corrective factors .sub.A,i, associated with a line configuration corresponding (exactly or largely) to the specific configuration of the line 30 and, if there is, recover these corrective factors .sub.A,i, in order to apply them to the respective amplitudes of the electrical sub-carriers to which these factors correspond.

(20) Reference is now made to FIG. 4 which illustrates a pre-distortion method according to an embodiment of the present disclosure.

(21) In this method, the pre-distortion device 11 is advantageously configured beforehand (step E1), in order to allow for a subsequent implementation of the actual pre-distortion when useful data are to be transmitted.

(22) During this preliminary configuration step E1, a certain number of sets of N corrective factors .sub.A,i, intended to be applied to the amplitude of the N electrical sub-carriers before optical conversion, are calculated so as to allow for the equalization, at the output of the optical transmission line 30, of the values Q.sub.1 taken by a first parameter Q, dependent on the power of a signal, for the different multiplexed optical sub-carriers.

(23) This first parameter Q is typically the signal-to-noise ratio (SNR), which reflects the impact of the spectral transfer functions of the optical devices both on the power of the optical signal and on that of the optical noise, in which case efforts are made to equalize the N signal-to-noise ratio values SNR.sub.i specific to the different sub-carriers of the signal S.sub.OPT at the line output 30, the equalization of these parameters making it possible to obtain a better bit error ratio.

(24) In a simplified variant, if it is considered that the noise power is relatively uniform over all of the spectrum of the optical signal S.sub.OPT, this first parameter Q is the optical power P, in which case efforts are made to equalize the N power values P.sub.i specific to the different sub-carriers of the signal S.sub.OPT at the line output 30.

(25) As already indicated, a first variant preliminary configuration consists, for a given transmission line 30 at the input of which there is a pre-distortion device 11 already installed, in measuring (step E11) the N values of this first parameter Q for the different sub-carriers from an OFDM optical signal transmitted in this line 30, this signal being able to be a specific test signal, even a signal carrying useful data transmitted previously, and in calculating (step E12) each of the N corrective factors .sub.A,i, from the N measured values of this first parameter Q. Once calculated, the corrective factors .sub.A,i, can then be stored (step E13), so as to be able to be then directly applied by the processing means 12 in the generation in the electrical domain of the OFDM signal carrying useful data.

(26) A single series of corrective factors can thus be stored, notably when the optical signals transmitted by the transmitter 10 follow the transmission line 30 to a single destination, and therefore passes through the same cascade of optical devices.

(27) Alternatively, when the line 30, at the input of which the device 11 is installed, comprises multiple branches, and therefore the optical signals transmitted by the transmitter 10 can follow optical paths having different cascades of optical devices depending on the planned destination, it is possible to store several series of N corrective factors, evaluated for each possible destination of an optical signal transmitted by the transmitter 10 and following the line 30. In the latter case, when they have to generate an optical signal passing through the line 30 to a particular destination, the processing means 12 search the storage means for the corrective factors .sub.A,i, associated with this specific destination.

(28) A second variant preliminary configuration consists in determining several sets of corrective factors .sub.A,i, corresponding to a certain number of possible transmission line configurations (these configurations being able to be identified by a value of a parameter R which is associated with them), and in storing, in the storage means 14, for each of these transmission line configurations, the duly determined sets of corrective factors, advantageously in association with the value of the parameter R characterizing the line configuration which is associated with them.

(29) In this case, the preliminary configuration step E1 comprises, for a specific line configuration: the determination (step E11), at the output of the optical transmission line for which the configuration is considered, values of the first parameter Q for each of the optical sub-carriers of the signal S.sub.OPT, typically following the transmission of a test OFDM optical signal in this line, by measuring the value Q.sub.i of this first parameter Q for each ith optical sub-carrier by means of an optical spectrum analyser; the calculation (step E12) of the corrective factors .sub.A,i, respectively associated with each of these optical sub-carriers from the values Q.sub.i determined for the first parameter Q; and the storage (step E13), in the storage means 14 of the pre-distortion device 11, of these corrective factors .sub.A,i, advantageously in association with a specific parameter R value identifying this line configuration.

(30) These operations are repeated a number M of times (M1) for different optical transmission line configurations.

(31) This last variant makes it possible to preconfigure the pre-distortion device even before its installation in the transmitter 10 at the input of the transmission line 30 (e.g. in the factory), by storing therein a certain number of series of N corrective factors applicable for possible optical transmission line configurations.

(32) The processing means 12 then simply have to select, from the stored sets of N corrective factors, the set associated with an optical transmission line configuration corresponding, either strictly identically, or in the closest possible approximation, to the specific configuration of the transmission line 30.

(33) To illustrate this last case, it can be supposed that, in a first approximation, out of the different optical devices present on the transmission line considered, only the ROADM multiplexers present in this line are taken into account, these being considered to most affect the spectra of the optical signal S.sub.OPT. In a second approximation, it can be considered that all these ROADM multiplexers have similar spectral transfer functions.

(34) In such a case, different series of corrective factors can be calculated for optical lines having different numbers of ROADM multiplexers, and the parameter R characterizing the configuration of these optical lines can simply be the number of ROADM multiplexers present in these transmission lines.

(35) The preliminary configuration step E1 then consists in proceeding to measure the values Q.sub.i of the first parameter Q on a first test transmission line containing only a single ROADM multiplexer (i.e. R=1) in order to obtain a first set of corrective factors .sub.A,i,R=1, then to measure the values Q.sub.i of the first parameter Q with a test transmission line comprising two ROADM multiplexers (i.e. R=2) in order to obtain a second set of values of corrective factors .sub.A,i,R=2, and so on, up to a sufficient number M of ROADM multiplexers to be able to cover all the possible optical transmission line configurations used in real useful data transmission conditions.

(36) The different sets of corrective factor values .sub.A,i, are thus stored, possibly in association with the parameter R value identifying the line configuration employed to calculate these values (in this case the number of ROADM multiplexers in the optical line), in the storage means 14, in order to be able to be recovered subsequently by the processing means 12, as in the exemplary table below:

(37) TABLE-US-00001 R .sub.A,1 . . . .sub.A,i . . . .sub.A,N 1 1.05 . . . 1 . . . 1 2 1.1 . . . 1.05 . . . 1.02 . . . . . . . . . . . . . . . . . .

(38) More generally, an optical transmission line configuration can be characterized by the different quantities of L optical devices of distinct types present in this line (L1), in which each distinct type of device has a distinct spectral transfer function likely to affect the spectrum of the optical signal S.sub.OPT. In this case, the storage means 14 first stores the series of corrective factors .sub.A,i calculated for line configurations corresponding to the different possible combinations of these different quantities of optical devices (the parameter R can then be a doublet, a triplet, etc., depending on the number L of distinct types of devices). This enables the processing means 12 to retrieve the corrective factors to be applied, in the transmission of the signal over the transmission line 30, after having obtained the information on the quantities of optical devices of different types that this line 30 contains.

(39) Alternatively, these series of corrective factors .sub.A,i can be calculated for different line configurations corresponding to a specific type of optical device (e.g. ROADM) associated with a variable distance separating this optical device from the device preceding it in the line concerned.

(40) The storage means 14 can be fed with series of corrective factors corresponding to all or some of the abovementioned variants, the processing means 12 then being able to recover several series of corrective factors that can be used for the line 30, according to different criteria, and to calculate the corrective factors .sub.A,i to be applied from these corrective factors (e.g. by averaging them).

(41) Once the preliminary configuration of the pre-distortion device 11 has been performed, the pre-distortion can then be applied in a nominal operating mode of the optical transmitter 10, i.e. when the latter has to transmit OFDM signals carrying useful data.

(42) In such a nominal operating mode, the processing means 12 access (step E2) the storage means 14 to recover the corrective factors .sub.A,i to be applied respectively to the electrical sub-carriers of the signal S.sub.E to be transmitted.

(43) In the case where several sets of N corrective factors .sub.A,i are stored in the storage means 14, the processing means 12 select the set to be recovered that best corresponds to the configuration of the line 30 in which the signal S.sub.E is to be transmitted.

(44) This can be done, for example, by retrieving, if it has been entered, a parameter R value identifying this configuration, for example when R identifies the destination of the data in a line 30 with branches or when R identifies the number of ROADM multiplexers in this line 30 and by recovering the N corrective factors .sub.A,i associated with this value in the storage means 14. In this case, the parameter R is obtained beforehand by the processing means 12, for example in the form of a piece of information supplied to the device 11 by the operator of the transmission line 30, or by automatic detection of this parameter.

(45) Once provided with these N corrective factors .sub.A,i, the processing means 12 apply (step E3) these factors to the respective amplitudes of the N corresponding electrical sub-carriers, in the generation of the sub-carriers in the electrical domain, before the thus distorted electrical signal S.sub.E is supplied to the digital-analog converter 13, the latter supplying the electrical signal, in analog form this time, to the electro-optical conversion means 15 for the latter to convert (step E4) the electrical signal into an optical signal S.sub.OPT transmitted to the line 30.

(46) In one embodiment, the first parameter Q, the values of which are to be equalized for the different optical sub-carriers at the line output 30, is the signal-to-noise ratio SNR at the output of the optical transmission line, such that the corrective factors .sub.A,i serve to pre-distort the spectrum of the OFDM signal in transmission in order to make the particular signal-to-noise ratio values SNR.sub.1 of the different OFDM sub-carriers uniform in reception.

(47) The corrective factors .sub.A,i are then calculated as follows: For the ith optical sub-carrier of the signal S.sub.OPT, the signal-to-noise ratio SNR.sub.i is expressed as follows:

(48) ${\mathrm{SNR}}_i=\frac{P_{\mathrm{SC},i}}{P_{n,i}}$

(49) in which: P.sub.SC,i is the power of the ith optical sub-carrier of the signal S.sub.OPT at the output of the line 30. P.sub.n,i is the power of the noise (mainly caused by the amplified spontaneous emission, called ASE, from the optical amplifiers) affecting this ith optical sub-carrier, at the output of this line 30.

(50) It can be seen in this equation that both the power P.sub.SC,i of the ith optical sub-carrier and the noise power P.sub.n,i affecting it are likely to be impacted by the spectral transfer functions of the optical devices present on the line 30. It is then possible to express (as power) the corrective factor .sub.P,i to be applied for the ith sub-carrier as a function of the signal-to-noise ratio SNR.sub.i for this sub-carrier:

(51) ${}_{P,i}=\frac{}{{\mathrm{SNR}}_i}$

(52) in which is a standardizing constant which can be defined as follows.

(53) For the total power of the signal to be conserved in the application of the pre-distortion (to avoid having to modify the parameters of the RF driver electrical amplifier on each pre-distortion), a first condition to be met by the corrective factors is as follows:

(54) $\underset{i=1}{\overset{N}{.Math.}}\left({}_{P,i}.Math.P_{\mathrm{SCC},i}\right)=P_{\mathrm{SCC},\mathrm{total}}$

(55) in which: P.sub.SSC,i is the transmission power of the ith sub-carrier if no pre-distortion is applied; P.sub.SSC,total is the total transmission power of all the sub-carriers if no pre-distortion is applied.

(56) In as much as the transmission power P.sub.SSC,1 is identical for all the sub-carriers in the absence of pre-distortion, this equation becomes:

(57) $P_{\mathrm{SCC},1}.Math.\underset{i=1}{\overset{N}{.Math.}}{}_{P,i}=P_{\mathrm{SSC},\mathrm{total}}$

(58) $\underset{i=1}{\overset{N}{.Math.}}{}_{P,i}=\underset{i=1}{\overset{N}{.Math.}}\left(\frac{}{{\mathrm{SNR}}_i}\right)=N$

(59) hence the equation:

(60) $=\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{1}{{\mathrm{SNR}}_i}\right)}$

(61) and:

(62) which makes it possible to obtain the corrective factor (in power) for the ith sub-carrier:

(63) ${}_{P,i}=\frac{1}{{\mathrm{SNR}}_i}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{1}{{\mathrm{SNR}}_i}\right)}$

(64) The corrective factor .sub.A,i to be applied, expressed this time in amplitude, is then obtained as a function of the signal-to-noise ratios of the N sub-carriers of the optical signal S.sub.OPT at the receiver, according to the following equation (1):

(65) 0$\begin{array}{cc}{}_{A,i}=\sqrt{{}_{P,i}}=\sqrt{\frac{1}{{\mathrm{SNR}}_i}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{1}{{\mathrm{SNR}}_i}\right)}}& \left(1\right)\end{array}$

(66) To obtain these corrective factors .sub.A,i to be applied, it is therefore sufficient to first estimate the signal-to-noise ratio values SNR.sub.i of the N sub-carriers of the optical signal S.sub.OPT at the output of the transmission line 30 or of a transmission line of similar configuration.

(67) In another embodiment in which the first parameter Q is the power of the signal, the corrective factors are calculated in order to substantially equalize the values of the optical powers P.sub.SC,i of the sub-carriers of the signal S.sub.OPT at the output of the optical transmission line 30. In particular, in the case where the spectral noise density of ASE, generated by the chain of optical amplifiers present on the equalization of this type of parameter gives good performance levels in terms of bit error ratio, while being simpler to implement than the equalization of the signal-to-noise ratios.

(68) Given this last hypothesis, in order to conserve the total power of the signal in the application of the distortion, the parameter SNR.sub.i can be advantageously replaced, in the formula (1), by the ratio of the powers

(69) $\frac{P_{\mathrm{SC},i}}{P_{\mathrm{SC},0}}$

(70) in which: P.sub.SC,0 is the power at the output of the optical transmission line 30 of a reference sub-carrier from among the sub-carriers of the optical signal; P.sub.SC,i is the power at the output of the optical transmission line of the ith sub-carrier of the optical signal.

(71) The reference sub-carrier is advantageously a sub-carrier situated at the centre of the spectrum of the optical signal S.sub.OPT, in order to minimize the amplitude variations due to the pre-distortion for the sub-carriers situated at the edge of this spectrum.

(72) Thus, for the ith sub-carrier, the corrective factor to be applied is then obtained by replacing, in the preceding equation (1), the parameters SNR.sub.i in the manner indicated above, in order to obtain the following equation (2):

(73) $\begin{array}{cc}\sqrt{{}_{P,i}}=\sqrt{\frac{P_{\mathrm{SC},0}}{P_{\mathrm{SC},i}}.Math.\frac{N}{\underset{i=1}{\overset{N}{.Math.}}\left(\frac{P_{\mathrm{SC},0}}{P_{\mathrm{SC},i}}\right)}}& \left(2\right)\end{array}$

(74) Obviously, the invention is not limited to the exemplary embodiments described above and represented, from which it will be possible to provide other embodiments and forms, without in any way departing from the framework of the invention.