I/Q CODING METHOD FOR SDM COMMUNICATION SYSTEM OVER OPTICAL FIBRE

Abstract

The present invention relates to a method for dual-polarisation SDM transmission over optical fibre. The transmission method uses specific I/Q coding for combating the effects of PDL. The modulation symbols to be transmitted on the 2N polarisation states of the N wavelengths are broken down into real values and imaginary values (220). A first orthogonal linear transformation (230-1) is applied to the vector of the real values thus obtained and a second orthogonal linear transformation (230-2), separate from the first, is applied to the vector of the imaginary values thus obtained. A complex scalar solving an irreducible polynome of [X] in is multiplied with the first or second transformed vector before the two transformed vectors are summed (240) in order to provide a vector of transmission symbols for modulating the different states of polarisation of the spatial elementary channels.

Claims

1. SDM transmission method over optical fiber with polarization duality, intended to transmit, during one channel use, 2N symbols belonging to a modulation constellation in the complex plane, N>1 being the number of spatial elementary channels used for transmission, characterized in that: said symbols undergo a separation into real part and imaginary part (220-520) to provide a first vector consisting of the real parts of these symbols and a second vector consisting of the imaginary parts of these same symbols; a first orthogonal linear transformation (230-1, . . . , 430-1) is applied to the first vector to provide a first transformed vector; a second orthogonal linear transformation (230-2, . . . , 530-2), distinct from the first, is applied to the second vector to provide a second transformed vector; a complex scalar, solution of an irreducible polynome from [X] in is multiplied to the first or to the second transformed vector, before the two transformed vectors are summed to provide a vector consisting of 2N complex emission symbols, each complex transmission symbol modulating a first state and a second polarization state of a spatial elementary channel.

2. SDM transmission method over optical fiber with polarization duality according to claim 1, characterized in that the first linear transformation is the composition of a first rotation with a first non-trivial permutation and/or a first non-trivial reflection in .sup.2N and that the second linear transformation is the composition of a second rotation with a second non-trivial permutation and/or a second non-trivial reflection in .sup.2N.

3. SDM transmission method over optical fiber with polarization duality according to claim 2, characterized in that the first permutation is composed of an even plurality of transpositions and that the second permutation is composed of an odd plurality of transpositions, or vice versa.

4. SDM transmission method over optical fiber with polarization duality according to claim 3, characterized in that the first rotation and the second rotation are identical.

5. SDM transmission method over optical fiber with polarization duality according to claim 2, characterized in that the first orthogonal linear transformation is identity.

6. SDM transmission method over optical fiber with polarization duality according to claim 1, characterized in that the complex scalar, ? is chosen such that ?.sup.2N is not a positive real.

7. SDM transmission method over optical fiber with polarization duality according to claim 5, characterized in that the complex scalar is equal to j with j.sup.2=?1.

8. SDM transmission method over optical fiber with polarization duality according to claim 7, characterized in that the number N is odd with N?3.

9. SDM transmission method over optical fiber with polarization duality according to claim 1, characterized in that the spatial elementary channels are propagation modes in the optical fiber.

10. SDM transmission method over optical fiber with polarization duality according to claim 1, characterized in that the optical fiber is of the multi-core type and that the elementary spatial channels are different cores of said fiber.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Other characteristics and advantages of the invention will appear on reading a preferred embodiment of the invention, described with reference to the attached figures including:

[0026] FIG. 1, already described, schematically represents an optical fiber transmission device using pre-coding on two orthogonal polarizations;

[0027] FIG. 2 schematically represents a SDM transmission device over optical fiber with IQ coding according to a general embodiment of the invention;

[0028] FIG. 3 schematically represents a SDM transmission device over optical fiber with IQ coding according to a preferred embodiment of the invention;

[0029] FIG. 4 schematically represents a SDM transmission device over optical fiber with IQ coding according to a first embodiment of the invention;

[0030] FIG. 5 schematically represents a SDM transmission device over optical fiber with IQ coding according to a second embodiment of the invention;

[0031] FIG. 6 shows an example of the gain provided by an SDM transmission device according to the invention.

DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS

[0032] We will consider in the following a transmission system with spatial diversity (SDM) over optical fiber. Spatial diversity may be due to the plurality of modes and/or cores in the fiber. In the case of a classic multimode fiber, the diameter of the core is large enough to allow the propagation of several modes at the wavelength considered. In the case of a multi-core fiber, propagation takes place in a plurality of elementary cores of the fiber. The case of a bundle of single-mode fibers with reduced cladding thickness is compared below to a multi-core fiber.

[0033] The SDM transmission systems considered below can be of one and/or the other type, it being understood that the spatial elementary channels are then propagation modes and/or cores of an optical fiber.

[0034] We further assume that the optical fiber is classically affected by PDL attenuation, in other words that the different states of polarization in the fiber do not undergo the same attenuation. It is recalled that PDL attenuation is generally introduced by optical elements between fiber sections, in particular doped fiber optical amplifiers (EDFA) which create energy losses and fluctuations in optical signal to noise ratio or OSNR (Optical Signal to Noise Ratio). Abstraction will be made however of the polarization dispersion (PMD) as this effect can be effectively corrected by channel equalization in the DSP of the receiver.

[0035] The effect of PDL attenuation for a spatial elementary channel can be expressed by the H.sub.PDL matrix applied to the two polarization states:

H.sub.PDL=D.sub.?R.sub.?B.sub.?

where

[00001]$D_{?}=\left(\begin{array}{cc}\sqrt{1+?}& 0\\ 0& \sqrt{1-?}\end{array}\right)$

is the gain matrix,

[00002]$R_{?}=\left(\begin{array}{cc}\cos ?& -s\mathrm{in}?\\ \sin ?& \cos ?\end{array}\right)$

is the polarisation rotation matrix and

[00003]$B_{?}=\left(\begin{array}{cc}\exp \left(i?\right)& 0\\ 0& \exp \left(i?\right)\end{array}\right)$

is the birefringence matrix with ??[0,1] defining the value of PDL PDL, ?.sub.dB=log.sub.10(?), with

[00004]$?=\frac{1+?}{1-?}$

and ?,??[??,?].

[0036] The SDM transmission system uses a plurality N of spatial elementary channels, each spatial elementary channel being associated with two polarization states. Thus, at each transmission instant, in other words at each use of the channel, the transmission system can transmit 2N modulation symbols, one symbol being transmitted per polarization state and per spatial elementary channel. The number N is generally chosen high, of the order of several tens or even more. In any case N>1 and, preferably, N>2.

[0037] The idea underlying the present invention is to separate the real parts and the imaginary parts of the different modulation symbols and to subject them to distinct orthogonal linear transformations before recombining them in the complex plane to then modulate the light signal with the various spatial elementary channels/polarizations. We thus carry out an averaging of the PDL attenuation over the different polarization states and the different spatial elementary channels.

[0038] FIG. 2 schematically represents a SDM_transmission device over optical fiber according to a general embodiment of the invention.

[0039] The data to be transmitted at each transmission interval is in the form of 2N information symbols, for example 2N q-ary words with q?log.sub.2 Q where Q is the cardinal of the modulation alphabet. The modulation alphabet may in particular be a Q-QAM alphabet.

[0040] The information symbols may themselves result from source coding and/or channel coding, in a manner known per se.

[0041] In all cases, the 2N information symbols are respectively converted into 2N modulation symbols in the q-ary symbol modulators 210-1, . . . , 210-2N. The odd indices of these symbols correspond to a first polarization state and the even indices to a second polarization state, orthogonal to the first. Each of these modulation symbols, denoted in the following x.sub.1, . . . , x.sub.2N, is then subjected to a decomposition into a real part and an imaginary part in the separation module I/Q, 220.

[0042] The respective real parts of these modulation symbols (x.sub.1), . . . , (x.sub.2N) form a vector X.sub.R in .sup.2N which is supplied to a first linear combination module 230-1. This first module combines these real parts by means of a first orthogonal linear transformation, F, represented by a matrix F?O(2N,), to provide a first transformed vector {tilde over (X)}.sub.R, in .sup.2N.

[0043] Similarly, the imaginary parts of the modulation symbols form a vector X.sub.1 in .sup.2N which is supplied to a second linear combination module, 230-2. This second module combines these imaginary parts by means of a second orthogonal linear transformation, G, represented by a matrix G?O(2N,), to provide a second transformed vector, {tilde over (X)}.sub.1, in .sup.2N.

[0044] The orthogonal linear transformations F and G are advantageously chosen distinct. For example, one of them could be a direct orthogonal linear transformation, in other words the corresponding matrix will be an element of the special orthogonal group SO(2N,), and the other will be an indirect orthogonal linear transformation.

[0045] The second transformed vector is then multiplied in 240 by a complex scalar value ?, solution of a polynome of [X], irreducible in . Preferably, ? will be chosen not to be a norm of an element in the complex plane, in other words ?.sup.2N should not be a positive real.

[0046] The first transformed vector and the second transformed vector thus multiplied are finally summed in the adder 250 to provide a vector in .sup.2N, {tilde over (X)} whose complex elements {tilde over (x)}.sub.1, . . . , {tilde over (x)}.sub.2N are transmission symbols respectively used to modulate the 2N polarization states of the N spatial elementary channels. More precisely the component of a first polarization state (for example a horizontal polarization component) of a spatial elementary channel of index n will be given by ({tilde over (x)}.sub.n) and that of a second polarization state (for example a vertical polarization component) of this spatial elementary channel will be given by ({tilde over (x)}.sub.n), or vice versa.

[0047] Thus, the vector {tilde over (X)} can be expressed, up to a multiplicative coefficient, in the form:

[00005]$\begin{array}{cc}\stackrel{?}{X}=F{X}_R+?{\mathrm{GX}}_I& \left(1\right)\end{array}$

[0048] According to a variant not shown, the first transformed vector is multiplied by the complex scalar value ? in place of the second transformed vector, the first transformed vector thus multiplied being then summed with the second transformed vector to provide the vector {tilde over (X)}.

[0049] FIG. 3 schematically represents a SDM transmission device on optical fiber according to a preferred embodiment of the invention.

[0050] Modules 310-1, . . . , 310-2N, 320, 330-1 and 330-2 respectively fulfill the same functions here as modules 210-1, . . . , 210-2N, 220, 230-1 and 230-2 in FIG. 2.

[0051] Unlike the embodiment illustrated in FIG. 2, the first transformed vector and the second transformed vector are combined by the I/Q combination module, 340, to form the complex vector {tilde over (X)}={tilde over (X)}.sub.R+j{tilde over (X)}.sub.I of .sup.2N. In other words, this embodiment is deduced as a special case from the general embodiment with ?=j, the I/Q combination module here replacing the multiplier 240 and the adder 250.

[0052] Advantageously, the complex scalar a is not a norm, in other words N is chosen odd with N?3.

[0053] The complex elements {tilde over (x)}.sub.1, . . . , {tilde over (x)}.sub.2N of the vector {tilde over (X)} are respectively used to modulate the 2N polarization states of the N spatial elementary channels.

[0054] FIG. 4 schematically represents a SDM transmission device over optical fiber with IQ coding according to a first embodiment of the invention.

[0055] Modules 410-1, . . . , 410-2N, 420, 430-1, 430-2, 440 respectively perform the same functions as modules 310-1, . . . , 310-2N, 320, 330-1, 330-2 and 340 of FIG. 3.

[0056] This embodiment is a particular case of the preferred embodiment of FIG. 3 in that the first linear transformation is direct, that is to say a rotation R in space in .sup.2N.

[0057] The second linear transformation results from the composition of this rotation R with a non-trivial permutation P in .sup.2N and/or a non-trivial reflection S in .sup.2N. By non-trivial permutation, we mean a permutation distinct of identity I, By non-trivial reflection, we mean a reflection distinct from ?I.

[0058] The permutation can be composed of an even number of transpositions in which case the second linear transformation is still a rotation, or it can be composed of an odd number of such transpositions.

[0059] The permutation can be cyclic, the second linear transformation then being represented by the matrix PR where P?{?, ?.sup.2, . . . , ?.sup.2N-1} set of possible permutations (except the trivial permutation) and where ? is the cyclic permutation matrix defined by:

[00006]$?=\left(\begin{array}{ccccc}0& 1& 0& .Math.& 0\\ 0& 0& 1& .Math.& 0\\ .Math.& & ?& ?& .Math.\\ 0& 0& .Math.& 0& 1\\ 1& 0& .Math.& 0& 0\end{array}\right)$

[0060] As in the general case, the roles first and second linear transformations can be interchanged. In other words, the rotation R can be applied to the vector of imaginary parts X.sub.I and the compound of rotation and permutation and/or reflection (S)PR/S(P)R can be applied to the vector of real parts X.sub.R.

[0061] FIG. 5 schematically represents a SDM transmission device over optical fiber with IQ coding according to a second embodiment of the invention.

[0062] Modules 510-1, . . . , 510-2N, 520, 530, 540 respectively perform the same functions as modules 310-1, . . . , 310-2N, 320, 530-2 and 540 in FIG. 3.

[0063] This exemplary embodiment is a particular case of the preferred embodiment of FIG. 3 in that the first linear transformation is trivial and equal to identity I, and that the second linear transformation results from the composition of this rotation R with a trivial or non-trivial permutation P in .sup.2N. The first vector and the second transformed vector are here combined to form the complex vector {tilde over (X)} of symbols intended to modulate the 2N polarization states as above.

[0064] In all cases, the received optical signal is spatially (per propagation mode and/or core) demultiplexed and per polarization state. The 2N?2N MIMO channel can be estimated, for example using an LS (Least Squares) algorithm from pilot symbols. The symbols transmitted by the transmission device can then be estimated using a MIMO decoder using an ML (Maximum Likelihood) estimate or more simply a ZF (Zero Forcing) estimate aimed at multiplying the signal received by the pseudo-inverse of the channel matrix, namely {tilde over ({circumflex over (X)})}=(H.sup.HH).sup.?1H.sup.HY where ? of size 2N?2N is the estimated matrix of the MIMO channel.

[0065] After separation of the real and imaginary parts of each of the components of {tilde over ({circumflex over (X)})} and formation of a first vector X.sub.R consisting of the 2N real parts and of a second vector X.sub.I consisting of the 2N imaginary parts, a first inverse orthogonal transformation F.sup.?1 is applied to the first vector X.sub.R and a second inverse orthogonal transformation G.sup.?1 is applied to the second vector, multiplied by ?.sup.?1, ?.sup.?1X.sub.I. We can then estimate the real and imaginary parts of the modulation symbols from the components of the same rank of the vectors thus obtained.

[0066] FIG. 6 shows in an example the gain provided by a SDM transmission device according to the invention for N=4 spatial elementary channels, here elementary cores of a multi-core fiber (MCF).

[0067] The value of PDL, ?.sub.dB was assumed to be the same for all spatial elementary channels and equal to 4.5 dB, the polarization rotation, ? was equal to ?/2.

[0068] The optical fiber was made up of 10 sections of 100 km each, an optical amplifier with constant gain on the various modes being provided between consecutive sections. The symbol rate was 12 Gbauds and the modulation constellation was 16-QAM.

[0069] The chosen embodiment was that of FIG. 5 with P=Id.sub.2N.

[0070] The estimate upon receipt was carried out using an ML estimator.

[0071] FIG. 6 gives the bit error rate (BER) as a function of the optical signal-to-noise ratio (OSNR) in the multicore fiber. In this case, we have ?.sup.2N=1 but we nevertheless observe a gain (of OSNR) of more than 1 dB compared to a non-IQ coded SDM system.