Dense IQ coding method and device for WDM communication systems over optical fibers.

Abstract

The invention relates to a method and a device for WDM transmission in a dual polarization optical fiber. The transmission method uses a specific I / Q coding that allows to suppress the effects of PDL. The modulation symbols to be transmitted on 2N polarization states on N wavelengths are decomposed into real and imaginary numbers. A real vector is constructed by concatenating these real and imaginary numbers. A first reversible linear transformation represented by a dense real matrix is ​​applied to the real vector to provide a transformed real vector. Complex transmission symbols are formed by I / Q combination (240) of the components of the transformed vector, the transmission symbols modulating the different polarization states of the WDM channel.

Description

[Technical field]

[0001] The present invention relates to the field of optical fiber communications, and more particularly to wavelength division multiplexed (WDM) communications.

[0002] (State of the Prior Art) WDM communication systems, commonly used on optical fibres, achieve transmission rates of several Tb / s. Different types of WDM systems are known in the state of the art, those defined by wavelength (Coarse Wavelength Division Multiplexing, CWDM) and, more recently, those defined by frequency (Dense WDM, DWDM). The difference between CWDM and DWDM systems is essentially the spacing between the transmission channels. When the transmission channels are adjacent or overlapping, they are called WDM superchannels and Nyquist superchannels respectively. The term "WDM" is used in the following in a generic sense to encompass the different types of systems mentioned above.

[0003] The use of higher modulation orders and multiplexing with orthogonal polarizations has made it possible to further increase the capacity of these communication systems, but these advances currently face various limitations.

[0004] Firstly, the increasing density of WDM transmission channels and the relative proximity of the subcarriers leads to increasing levels of inter-channel interference (ICI). This interference can be countered by applying an ideal rectangular shaping of the channel in the frequency domain, in other words a waveform that follows a synchronous function in the time domain (so-called Nyquist shaping). Of course, in practice, shaping is imperfect and residual inter-channel interference remains.

[0005] Secondly, different dispersion phenomena such as Chromatic Dispersion (CD), Polarization Mode Dispersion (PMD) and Polarization Dependent Loss (PDL) increase the Bit Error Rate (BER) in different channels. However, if the first two effects can be digitally corrected at reception, the latter cannot due to its non-uniform nature, degrading the performance of WDM transmission systems in terms of BER as a function of flow rate and thus the transmission capacity.

[0006] A paper by Elie Awad, titled "Emerging space-time coding techniques for optical fiber transmission systems," published in 2015, proposes using space-time coding techniques to mitigate the loss of transmission capacity due to PDL. However, these coding techniques add complexity to the transmitter and receiver, since the block of information symbols to be transmitted is coded over multiple consecutive transmission intervals, or Time Transmission Intervals (TTIs), or more commonly, over multiple channel utilizations (CUs).

[0007] The orthogonal polarization precoding technique to mitigate capacity loss due to PDL is described in a paper by C. Zhu et al., entitled “Improved polarization dependent loss tolerance for polarization multilexed coherent optical systems by polarization pairwise coding,” published in J. Lightwave Technolog, vol. 34 no. 8, pp. 1746-1753, 2016.

[0008] This method of precoding in orthogonal polarizations is illustrated diagrammatically in FIG.

[0009] The information symbols (binary words) to be transmitted are converted to symbols of a modulation constellation in q-ary symbol modulators 110-1 and 110-2. The resulting modulation symbols x1, x2 are then rotated by an angle θ in the complex plane using respective rotation modules 120-1 and 120-2, and the rotated symbols The real part of the first rotated symbol and the real part of the second rotated symbol are combined at 130-1 to obtain a first emitted symbol carried by a first polarization component (e.g., a horizontal polarization state). Provides TIFF2025511283000003.tif13150.

[0010] Similarly, the imaginary part of the first rotated symbol and the imaginary part of the second rotated symbol are combined at 130-2 to generate a second emitted symbol carried by a second polarization component (e.g., a vertical polarization state) that is orthogonal to the first polarization component. Provides TIFF2025511283000004.tif13150.

[0011] Then, the optical signals in which the orthogonal polarization components are modulated by the emission symbols x1 and x2, respectively, are transmitted onto an optical fiber.

[0012] However, the precoding method described herein only applies to single carrier transmission systems and not to WDM transmission systems.

[0013] The object of the invention is therefore to propose a method of WDM transmission over optical fibre, and an associated transmission device, which allows the realisation of a high transmission capacity despite PDL and interference between adjacent channels, whilst only requiring a one-time use of a transmission channel to transmit a block of information symbols. Summary of the Invention

[0014] The invention is defined by a method of WDM transmission over an optical fiber with polarization duality, intended to transmit, during the use of a channel, 2N symbols belonging to a modulation constellation on a complex plane, N>1 being the number of WDM channels used for the transmission, the WDM transmission method comprising: The symbols are separated into real and imaginary parts, and a real vector of size 4N (X R ) and A reversible linear transformation represented by a 4N by 4N dense real matrix is ​​applied to the real vector to provide a transformed real vector; 2N complex emission symbols are obtained by performing IQ combining of the 2N components of a first set of transformed real vector components with the 2N components of a second set of transformed real vector components, respectively, the first and second sets being independent, and each complex emission symbol modulates a first state and a second state of polarization of a WDM channel. It is unique in that respect.

[0015] A real vector is typically formed by concatenating a first vector made up of the real parts of the modulation symbols with a second vector made up of the imaginary parts of those same symbols.

[0016] Preferably, the first set of components of the transformed real vector consists of the first 2N components of this vector, and the second set of components of the transformed real vector consists of the last 2N components of this vector.

[0017] According to a preferred embodiment, the characteristic polynomial of a dense real matrix has no real roots. For example, a dense real matrix may be TIFF2025511283000005.tif7150 is a rotation matrix in space.

[0018] The invention also relates to a WDM transmission device over an optical fiber with dual polarization intended to transmit, during the use of a channel, 2N symbols belonging to a modulation constellation in the complex plane, where N>1 is the number of WDM channels used for transmission, the transmission device comprising: A real vector of size 4N formed by the 2N real parts of a symbol and the 2N imaginary parts of the same symbol (X R a first module configured to separate each of the symbols into a real part and an imaginary part to provide a second linear combination module configured to apply a reversible linear transformation represented by a dense real matrix of size 4N×4N to the real vector to provide a transformed real vector; a third IQ combining module configured to combine the 2N components of the first set of transformed real vectors with the 2N components of the second set of transformed real vector components, respectively, the first set and the second set being independent to generate 2N complex emission symbols, each complex emission symbol modulating a first state and a second polarization state of a WDM channel; It is unique in that it includes

[0019] The first module is typically configured to form a real vector by concatenating a first vector made up of the real parts of modulation symbols and a second vector made up of the imaginary parts of like symbols.

[0020] Preferably, the third module is configured such that the first set of components of the transformed real vector consists of the first 2N components of this vector, and the second set of components of the transformed real vector consists of the last 2N components of this vector.

[0021] Advantageously, the characteristic equation of a dense real matrix has no real roots.

[0022] According to a preferred embodiment, the dense real matrix is TIFF2025511283000006.tif7150 is a rotation matrix in space.

[0023] Other characteristics and advantages of the present invention will become apparent on reading the description of preferred embodiments of the invention which follows with reference to the accompanying drawings, in which: [Brief description of the drawings]

[0024] [Figure 1] 1 shows a schematic representation of a fibre optic transmission device previously described which uses precoding on two orthogonal polarisations; [Diagram 2] 1 illustrates a schematic representation of a WDM transmission device over optical fiber with dense IQ coding according to a general embodiment of the present invention. [Diagram 3] WDM TRANSMISSION DEVICE OVER OPTICAL FIBER USING DENSE IQ CODING ACCORDING TO A PREFERRED EMBODIMENT OF THE PRESENT EMBODIMENT [Figure 4A] 4 shows the gain achieved by a WDM transmitting device according to the present invention for different WDM channel courses. [Figure 4B] 4 shows the gain achieved by a WDM transmitting device according to the present invention for different WDM channel courses. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] In the following we consider a WDM transmission system and assume that the fiber is classically affected by PDL attenuation, i.e. the different states of polarization (SOP) of the fiber are not affected by the same attenuation. It is recalled that PDL attenuation is generally introduced by optical elements between fiber sections, in particular doped fiber optical amplifiers (EDFAs), which create energy losses and variations in the optical signal-to-noise ratio (OSNR). However, dispersion effects in the fiber, such as chromatic dispersion (CD) and polarization-mode dispersion (PMD), are neglected because these effects are effectively compensated by the channel equalization in the receiver's DSP.

[0026] The effect of PDL attenuation on the WDM channel (and the signal spatial mode) is shown in column H PDL It can be expressed as follows: (Formula 1) TIFF2025511283000007.tif13150, where TIFF2025511283000008.tif13150 is the gain matrix, TIFF2025511283000009.tif13150 is the polarization rotation matrix, TIFF2025511283000010.tif13150 is the birefringence matrix, γ ∈ [0,1] defines the PDL value, The image is JPEG2025511283000011.jpg13150.

[0027] A WDM transmission system uses a number N of WDM channels (wavelengths or subcarriers), each WDM channel being associated with two polarization states. Thus, at each transmission instant, in other words each use of a channel, the transmission system can transmit 2N modulation symbols, one symbol per polarization state and per WDM channel. The number N is typically chosen to be high, on the order of tens to hundreds. In any case, N>1.

[0028] The idea behind the present invention is to separate the real and imaginary parts of the different modulation symbols and subject all the real and imaginary parts of the different symbols to a reversible linear transformation, which results in averaging the PDL attenuation across different polarization states and different WDM channels.

[0029] FIG. 2 illustrates a schematic representation of a WDM transmission device over optical fiber according to a general embodiment of the present invention.

[0030] The data transmitted in each transmission interval may consist of 2N information symbols, e.g. This is the format of 2N q-ary words in TIFF2025511283000012.tif7150, where Q is the base of the modulation alphabet. The modulation alphabet is specifically the Q-QAM alphabet.

[0031] The information symbols themselves may result, in known manner, from source coding and / or channel coding.

[0032] In all cases, the 2N information symbols are converted to q-ary2N modulation symbols in a q-ary modulator using symbols 210-1, ..., 210-2N, respectively, whose odd indices correspond to a first polarization state and whose even indices correspond to a second polarization state orthogonal to the first polarization state. Thus, x1, ..., x 2N Each of these modulation symbols, denoted as , is subjected to a decomposition in a separation module I / Q 220 into a real part and an imaginary part.

[0033] JPEG2025511283000013.jpg7150 real vector X R which is fed to the linear combination module 230. R are the modulation symbols x1, …x 2N This is obtained by separating and grouping the real and imaginary parts of JPEG2025511283000014.jpg7150. But normally, vector X R can be obtained by concatenating the real and imaginary parts of these symbols in any way, i.e. JPEG2025511283000015.jpg7150, where σ represents any permutation of the 2N components.

[0034] This first module 230 is JPEG2025511283000016.jpg6150 is a lossless linear transform represented by F by X R Combine the elements of The linear transformation is chosen such that the matrix F (representing F in the standard) is dense (or filled), i.e. does not contain zeros. Advantageously, the matrix F has a characteristic polynomial It is chosen so that it has no root in TIFF2025511283000018.tif7150, in other words, no eigenspace. This property is R It ensures efficient mixing of the components of the vectors and, as a result, averaging of the PDL.

[0035] The transformed vector TIFF2025511283000019.tif7150 is (Formula 2) JPEG2025511283000020.jpg26150, where f1, f2, ..., f 4N teeth, TIFF2025511283000021.tif7150 is a linear format.

[0036] The first 2N elements and the last 2N elements of TIFF2025511283000022.tif7150 are then combined two by two in the I / Q combiner module 240 to produce a 2N-dimensional complex vector gives TIFF2025511283000023.tif7150, (Formula 3) The image is JPEG2025511283000024.jpg20150.

[0037] More generally, TIFF2025511283000025.tif7150 Partial transformation vector of the first part of size 2N by selecting 2N components of the vector A second partially transformed vector of size 2N is also created by selecting the western part of the remaining 2N, forming TIFF2025511283000026.tif7150 TIFF2025511283000027.tif7150, so the complex vector is The file is TIFF2025511283000028.tif7150.

[0038] in any case, JPEG2025511283000029.jpg7150 is used to modulate each of the 2N polarization states of N WDM subcarriers / wavelengths.

[0039] FIG. 3 illustrates a WDM transmission device over optical fiber according to a preferred embodiment of the present invention.

[0040] Modules 310-1, ..., 310-N, 320, 330, 340 each perform functions similar to modules 210, 210-N, 220, 230, 240 of FIG.

[0041] This embodiment is a special case of the case shown in FIG. 2, and the linear transformation in the case shown in FIG. TIFF2025511283000030.tif7150 is a rotation in space.

[0042] The fact that the matrix must be complete means that I 4N is the trivial rotation matrix I, which is an identity matrix of size 4N. 4N OR -I 4N Note that we immediately exclude

[0043] Moreover, the rotation matrix does not have a unique (invariant) space since the dimensionality of the space is even.

[0044] Furthermore, the vector X R are the modulation symbols X1, …, X 2N can be obtained by concatenating the real and imaginary parts of TIFF2025511283000031.tif7150 is constructed by selecting a set of 2N components of the transformed vector, each of which is a partially transformed vector such that the set of components is independently related to the next two vectors. The resulting image can be seen as JPEG2025511283000032.jpg7150.

[0045] lastly, JPEG2025511283000033.jpg7150 is used to modulate 2N polarization states of N WDM subcarriers / wavelengths, respectively.

[0046] In the embodiment shown in Figures 2 and 3, dense IQ coding is applied to all WDM component channels, but alternatively dense IQ coding may be applied to blocks of channels, and the lossy linear transformation, e.g., rotation, may be selected differentially for each channel.

[0047] Finally, although the present invention is presented in the context of a dual polarization state optical fiber, those skilled in the art will appreciate that the dense IQ coding methods described above are also applicable to a single polarization state.

[0048] According to a first variant, the channel estimation and the corresponding equalization can be performed per subband. According to a second variant, the channel estimation and the corresponding equalization can be performed globally for the 2N×2N MIMO channel. In both cases, the channel estimation can be based on pilot symbols. For this purpose, CAZAC (Constant Amplitude Zero Auto Correlation) sequences can be used, for example Zadoff-Chu sequences.

[0049] In the case of 2N×2N MIMO channel equalization, the symbols transmitted by the transmitting device can be estimated using a MIMO decoder using ML (Maximum Likelihood) estimation, or more simply, ZF (Zero Forcing) estimation, which aims to multiply the received signal by the pseudo-inverse of the channel matrix, i.e. JPEG2025511283000034.jpg7150 is the estimated matrix of the MIMO channel. Alternatively, in per-subband equalization, the estimate of the transmitted symbols is given by N matrices TIFF2025511283000035.tif7150, where each of these matrices corresponds to a subband. It should be noted that this operation does not include the inverse of the linear transformation represented by the matrix F.

[0050] After separating the real and imaginary parts of each component of TIFF2025511283000036.tif7150, we extract a real vector of size 4N from those components. For example, when the embodiment shown in FIG. 2 or FIG. 3 is used for broadcasting, a first vector consisting of 2N real parts and a second vector consisting of 2N imaginary parts can be formed, and the first vector and the second vector are then concatenated to form a real vector TIFF2025511283000038.tif7150 can be obtained.

[0051] Take JPEG2025511283000039.jpg7150 and apply to its components the inverse of the permutation transformation σ applied to the emission.

[0052] For example, when a real vector is obtained by grouping the real and imaginary parts of a modulation symbol, the vector The first 2N components of TIFF2025511283000040.tif7150 are the real parts of the transmitted modulation symbols. gives an estimate of TIFF2025511283000041.tif13150, the last 2N components are the imaginary parts of the transmitted modulation symbols. gives an estimate of TIFF2025511283000042.tif13150.

[0053] 4A-4B show the gain provided by a WDM transmitting device according to the invention for different WDM channel assumptions.

[0054] PDL Value TIFF2025511283000043.tif7150 is the same for all WDM channels and is equal to 3 dB. TIFF2025511283000044.tif7150 assumed to be equal to π / 2.

[0055] The chromatic dispersion coefficient is equal to 17ps / nm / km, and the polarization dispersion coefficient is Equivalent to TIFF2025511283000045.tif13150.

[0056] The optical fiber consists of 10 sections of 100 km each, with optical amplifiers with constant wavelength gain between successive sections. The symbol rate is 28 Gbaud and the modulation constellation is 16QAM.

[0057] Equalization on receive is achieved by subband ZF equalization.

[0058] Figure 4A gives the bit error rate as a function of the optical signal-to-noise ratio (OSNR) in the fiber for N = 3 subbands. The subbands corresponding to different WDM channels do not overlap, and the signal transmitted in each subband is shaped by a raised cosine root filter with a roll-off factor of 0.1, which ensures the absence of inter-channel interference (ICI).

[0059] Curve 410 gives the bit error rate as a function of the signal-to-noise ratio for PDL values ​​wide by 3 dB in the case of classical WDM transmission without coding.

[0060] Curve 420 gives the BER for B2B (Back to Back) transmission, ie direct transmission without any fiber between transmitter and receiver.

[0061] It should be noted that the 410 curve shows a loss of 2.5 dB of performance in terms of signal-to-noise ratio (OSNR) compared to the direct transmission 420. On the other hand, when the dense I / Q coding method according to the invention is applied, a gain of about 1.2 dB is observed at 430 compared to the uncoded case. Thus, the loss is only about 1 dB compared to the B2B transmission.

[0062] Figure 4B shows again the bit error rate (BER) as a function of the optical signal-to-noise ratio (OSNR) in the fiber, but this time with overlapping sub-bands (Super-Nyquist WDM). Due to inter-channel interference, the performance is degraded compared to the situation shown in Figure 4A. However, dense I / Q coding achieves an SNR gain of about 1 dB over the uncoded case.

Claims

1. A WDM transmission method on a dual-polarized optical fiber intended to transmit 2N symbols belonging to a modulation constellation in the complex plane while a channel is in use, wherein N > 1 is the number of WDM channels used for transmission. The symbols are separated into real and imaginary parts (220), and a real vector of size 4N is formed by the 2N real parts of these symbols and the 2N imaginary parts of these similar symbols (X R ) to provide, A reversible linear transformation (230) represented by a dense real sequence of size 4N × 4N is applied to the real vector to provide the transformed real vector. 2N complex emission symbols are obtained by performing IQ coupling (240) of 2N components of a first set of components of the transformed real vector with 2N components of a second set of components of the transformed real vector, wherein the first set and the second set are independent, and each complex emission symbol modulates the first and second states of polarization of the WDM channel. A WDM transmission method over dual-polarized optical fiber, characterized by the following:

2. The WDM transmission method according to claim 1 on a dual-polarized optical fiber, characterized in that the real vector is formed by concatenating a first vector composed of the real part of the modulation symbol and a second vector composed of the imaginary part of the modulation symbol.

3. The WDM transmission method on a dual-polarized optical fiber according to claim 1, characterized in that the first set of components of the transformed real vector consists of the first 2N components of the vector, and the second set of components of the transformed real vector consists of the last 2N components of the vector.

4. The WDM transmission method on a dual-polarized optical fiber according to claim 1, characterized in that the characteristic polynomial of the dense real sequence does not have real roots.

5. The aforementioned dense real sequence is The WDM transmission method on a dual-polarized optical fiber according to claim 4, characterized in that it is a rotation matrix in space.

6. During the use of the channel, it is intended to transmit 2N symbols belonging to a modulation constellation in the complex plane, where N > 1 is the number of WDM channels used for the transmission. Each of the symbols is separated into a real part and an imaginary part, and a real vector of size 4N (X) is formed by the 2N real parts of these symbols and the 2N imaginary parts of these similar symbols. R A first module (220) configured to provide, A second linear combination module (230) is configured to apply a reversible linear transformation to the real vector by a dense real sequence of size 4N × 4N in order to provide the transformed real vector, The first set of components of the transformed real vector is configured to combine 2N components with the second set of components of the transformed real vector, the first set and the second set are independent, and each complex emission symbol is modulated by a third IQ coupling module (240) that modulates the first and second polarization states of the WDM channel in order to generate 2N complex emission symbols. A WDM transmission device on an optical fiber having dual polarization, characterized by including the following:

7. The WDM transmission device on an optical fiber having polarization duplexity according to claim 6, characterized in that the first module is configured to reshape the real vector by concatenating a first vector composed of the real part of a modulation symbol with a second vector composed of the imaginary part of similar vectors.

8. The WDM transmission device on an optical fiber having polarization duplexity according to claim 6, characterized in that the third IQ coupling module is configured such that the first set of components of the transformed real vector consists of the first 2N components of the vector, and the second set of components of the transformed real vector consists of the last 2N components of the vector.

9. The WDM transmission device on an optical fiber having polarization duplexity according to claim 6, characterized in that the characteristic polynomial of the dense real sequence does not have real roots.

10. The aforementioned dense real sequence is A WDM transmission device on an optical fiber having polarization duplexity, as described in claim 9, characterized in that it is a rotation matrix in space.

Images