Abstract
A system for transmission of an optical signal, the system including an optical coupler for splitting said signal into a first copy and a second copy. The optical coupler has an input for receiving the optical signal, a first output for the first copy and a second output for the second copy. The system also includes a first optical guide connected to the first output, a second optical guide connected to the second output and a superposition module for coherently superimposing the first copy and the second copy of the signal.
Claims
1. A system for transmission of an optical signal, the system including: an optical coupler configured to split the optical signal into a first copy and a second copy, the optical coupler having: an input configured to receive the optical signal; a first output for the first copy; and a second output for the second copy; a first optical guide connected to the first output; a second optical guide connected to the second output; and a superposition module configured to coherently superimpose the first copy and the second copy.
2. The system of claim 1, wherein the optical coupler is a 50:50 coupler.
3. The system of claim 1, wherein the superposition module includes a Mach-Zehnder (MZ) module having a first 22 Mach-Zehnder interferometer (MZI) and a second 22 MZI, the outputs of the first 22 MZI being coupled to the inputs of the second 22 MZI.
4. The system of claim 3, wherein the MZ module further includes a third 22 MZI, wherein the outputs of the second 22 MZI are coupled to the inputs of the third 22 MZI.
5. The system of claim 4, wherein a phase controller is located on an arm of one of the first 22 MZI, the second 22 MZI, or the third 22 MZI.
6. The system of claim 5, wherein the phase controller comprises a piezoelectric transducer.
7. The system of claim 5, wherein the MZ module further includes a feedback controller including a detector that is connected to a control element of the phase controller.
8. The system of claim 1, wherein the first optical guide is connected to the first output of the optical coupler via a first spectral inverter and wherein an input of the superposition module is connected to either the first optical guide or the second optical guide via a second spectral inverter.
9. The system of claim 1, further comprising a plurality of optical couplers, arranged into M layers, configured to split the optical signal into 2.sup.M copies, wherein the superposition module is one of M layers of superposition modules configured to coherently superimpose the 2.sup.M copies into a single output signal.
10. A method of transmitting a signal, the method comprising: splitting the signal into: a first copy which propagates along a first optical guide; and a second copy which propagates along a second optical guide; and coherently superimposing the first copy and the second copy onto one another.
11. The method of claim 10, wherein the step of superimposing is carried out by a Mach-Zehnder (MZ) module having a first Mach-Zehnder interferometer 22 (MZI) and a second 22 MZI, the outputs of the first 22 MZI being coupled to the inputs of the second 22 MZI.
12. The method of claim 11, wherein the MZ module further includes a third 22 MZI, wherein the outputs of the second 22 MZI are coupled to the inputs of the third 22 MZI.
13. The method of claim 12, wherein a phase controller is located on an arm of one of the first 22 MZI, the second 22 MZI, or the third 22 MZI and configured to control a phase of the copy of the optical signal that propagates through that arm.
14. The method of claim 13, wherein the phase controller is a piezoelectric transducer.
15. The method of claim 13, wherein the MZ module further includes a feedback controller including a detector that is connected to a control element of the phase controller.
16. The method of claim 10, wherein the first optical guide is connected to a first output of a optical coupler via a first spectral inverter configured to perform a first step of spectral inversion and wherein an input of the superposition module is connected to either the first optical guide or the second optical guide via a second spectral inverter configured to perform a second step of spectral inversion.
17. The method of claim 10, further comprising: using a plurality of optical couplers, arranged into M layers, to split the optical signal into 2.sup.M copies; and using M layers of superposition modules to coherently superimpose the 2.sup.M copies into a single output signal.
18. A system for transmission of a first signal from a first transmitter and a second signal from a second transmitter, the system including: an optical coupler having: a first input configured to receive the first signal and a second input configured to receive the second signal; and a first multiplexed output and a second multiplexed output connected such that, in use, the coupler multiplexes the first signal with the second signal to provide a resulting signal; and splits the resulting signal into a first multiplexed signal at the first multiplexed output and a second multiplexed signal at the second multiplexed output; a first optical guide connected to the first multiplexed output via a first spectral inverter; a second optical guide connected to the second multiplexed output; and a superposition module configured to coherently superimpose the first multiplexed signal and the second multiplexed signal onto one another, wherein the superposition module has a first input connected to the first optical guide and a second input connected to the second optical guide, wherein one of the inputs its connected to its respective optical guide via a second spectral inverter.
19. The system of claim 18, wherein the superposition module includes a Mach-Zehnder (MZ) module.
20. The system of claim 18, wherein the MZ module includes a first 22 Mach-Zehnder interferometer (MZI) and a second 22 MZI, the outputs of the first 22 MZI being coupled to the inputs of the second 22 MZI.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Embodiments of the present invention are now described with respect to the drawings, in which:
[0042] FIG. 1 shows a schematic of a known transmission link with no nonlinear compensation;
[0043] FIG. 2 shows a schematic of known transmission links with nonlinear compensation, in particular (a) mid-span spectral inversion (MSSI) and (b) a pair of phase-conjugated twin-waves co-propagating on two orthogonal polarizations of the same optical guide, and superimposed electronically at the receiver;
[0044] FIG. 3 shows a schematic of transmission links employing the approach of the first two aspects of the present invention, for reduction in nonlinear distortion, and in particular (a) with one pair of fibres in use and (b) M pairs of fibres in use;
[0045] FIG. 4 shows a schematic of transmission links employing the approach of the first two aspects of the present invention, including ESSI, in which (a) one pair of fibres is in use, (b) M pairs of fibres are in use and (c) M pairs of fibres are in use but with a different arrangement of spectral inverters;
[0046] FIG. 5 shows a schematic of transmission links employing the approach of the third aspect of the invention, combining ESSI and SDM, in which (a) one pair of fibres is in use, and (b) M pairs of fibres are in use, but still with only two transmitters;
[0047] FIG. 6 shows a schematic of the phase-controlled MZ modules used in all aspects of the present invention, to coherently superimpose two signal copies after propagation over independent fibres, in which (a) two 22 couplers are used, and (b) three 22 couplers are used; and
[0048] FIG. 7 shows a graph of the experimental Q-factor measured using a 10G optical transceiver operating at 9.953 Gbps. The Q-factor was calculated from the bit error count measured by the transceiver.
DETAILED DESCRIPTION OF THE INVENTION
[0049] FIG. 1 shows a known transmission link which employs no nonlinear compensation techniques. The gross structure of the transmission links shown in FIGS. 1 to 5 is the same, and details occurring in more than one drawing will not be described more than once, for conciseness. The following description will focus on those features which are most important to aspects of the invention. FIG. 1 shows a transmission link between transmitter 2 and receiver 8. The transmission link includes an optical amplifier 4, and N network modules 5. Each optical network module 5 includes a length of optical fibre (denoted by the coils depicted) and an additional optical amplifier 6. It is noted that the transmission link may include other features, such as network elements.
[0050] FIG. 2(a) shows a transmission link which includes two network modules 5. For the reasons set out above, this pairing is equivalent to a single network module 5 of e.g. FIG. 1. However, since in FIG. 2(a), the MSSI technique is employed to compensate for nonlinear distortion, an MSSI module 9 is required in the mid-span of the transmission link. As a result there are only N/2 network modules 5 in this particular configuration. In FIG. 2(b), rather than an MSSI module 9, the PCTW method is used for compensation, and as a result a transmitter 12 and receiver 18 are specially adapted for this method, at increased complexity and cost.
[0051] FIG. 3(a) shows a transmission link according to the present invention. In this example, the signal is split into two copies, each of which propagate along one of the two optical fibres 15a, 15b. Then, inside network module 5, the first and second copies are coherently superimposed using MZ module 10, into a single signal which then enters optical amplifier 6 for transmission to the receiver 8. FIG. 3(b) has an identical structure, except the signal is split into 2M copies by M layers of couplers (not shown). Then, a subsequent M layers of MZ modules are required to recombine the copies into the single signal which is output to the amplifier 6.
[0052] In order to quantify the maximum capacity improvement provided by the twin-fibre method of the present invention, the concept of nonlinear signal-to-noise ratio (SNR) must be introduced. Approximating the nonlinear Kerr distortion as an additive Gaussian noise means that the SNR.sub.NL can be defined as:
[00001]
[0053] In the above: I is the launch power density, n.sub.0 is white optical amplified spontaneous noise (ASE), I(I/I.sub.0).sup.2 is the nonlinear noise due to Kerr distortion, and I.sub.0 is the nonlinear characteristic power density. For the optimal launch power density, SNR.sub.NL is maximal, and given by:
[00002]
[0054] By modifying these expressions to account for the twin-fibre method of the present invention, it can be shown that the method improves SNR.sub.NL by 6 dB in the nonlinear regime and by 2 dB for the optimal launch power, as compared to the system as shown e.g. in FIG. 1.
[0055] In the nonlinear regime, for the system shown in FIG. 3(a), SNR.sub.NLI/[.sup.2.Math.I(I/I.sub.0).sup.2]. For the system shown in FIG. 1, SNR.sub.NLI/[I(I/I.sub.0).sup.2].
[0056] In the optimal launch power density regime, for the system as shown in FIG. 3(a), (SNR.sub.NL).sub.opt (2I.sub.0/n.sub.0).sup.2/3.
[0057] For the system shown in FIG. 3(b), in which there are M pairs of fibres, rather than only two, in the nonlinear regime, in the nonlinear regime, SNR.sub.NLI/[1/(2M).sup.2.Math.I(I/I.sub.0).sup.2] and in the optimal launch power density regime, (SNR.sub.NL).sub.opt (2M.Math.I.sub.0/n.sub.0).sup.2/3, where M is the number of parallel pairs of fibres.
[0058] This is, in the nonlinear regime, SNR.sub.NL increases by 6 dB each time the number of fibres is doubled, and in the optimal launch power regime, SNR.sub.NL increases by 2 dB each time the number of fibres is doubled. Finally, an increase of 6 dB on the SNR.sub.NL allows to increase the modulation constellation from 4-QAM to 16-QAM, for the same transmission distance, thereby doubling the system spectral efficiency from 2 bits/s/Hz/pol (current systems) to 4 bits/s/Hz/pol. Note that optical links operating at 4-QAM with additional SNR.sub.NL margin require a lower SNR.sub.NL increase.
[0059] FIG. 4(a) shows a similar configuration to FIG. 3(a) except, first and second spectral inverters 11 are located between the splitting point and the MZ module. The first spectral inverter 11 inverts one of the copies of the signal from the transmitter 2, and the second spectral inverter 11 reinverts the signal into its original form. Then when the two copies are coherently superimposed at the MZ module 10, the noise cancels out, which supplements the effect of the twin-fibre method as described in an earlier section. FIGS. 4(b) and 4(c) show similar arrangements to FIG. 4(a), with more layers of couplers (not shown) and MZ modules. The increased degree of splitting of the original signal results in an improved effect from the twin-fibre approach. The numbers of required spectral inverters 11 differs depending on the points in the tree of couplers (not shown) and MZ modules at which they are located.
[0060] The method shown in FIG. 4 presents an additional SNR.sub.NL improvement coming from the phase conjugation on the top of the improvement coming from the twin-fibres method. The typical SNR.sub.NL improvement provided by phase conjugation is around 2 dB. Therefore, the usage of the schematic shown in FIG. 4a, employing only two fibres, allows the increase of the SNR.sub.NL by 8 dB in the nonlinear regime and by 4 dB in the optimal launch power regime, without any modification of the transmitter, receiver, intermediate nodes or amplifiers. Moreover, the usage of phase conjugation at the edge of the span in this case does not impose any routing constrains or link power symmetry present in the MSSI method, the link is inherently symmetric. Besides the nonlinear distortion reduction given by the overlaid twin-fibre method, the ESSI method provides the same theoretical gain than the conventional MSSI but overcomes its main disadvantages: the phase conjugators are located at the edge of the spans (instead of a fixed mid-span location), the power evolutions are inherently similar between the two parallel fibres (that can be located in the same cable or being cores of a same multi-core fibre).
[0061] FIG. 5(a) shows an example of an embodiment of the third aspect of the invention, in which signals from two transmitters 22, 32 are first multiplexed at the coupler (not shown). The multiplexed signal is then split into two signals, one of which passes through spectral inverter 11. After this point in the system, the process is the same as that in FIG. 4(a). In the same way, FIG. 5(b) is analogous to FIG. 4(b).
[0062] In FIG. 5(b), since the number of fibres (2M) is bigger than the number of optical signals/transmitters (2 signals), a reduction of the nonlinear distortion does take place besides the nonlinear compensation provided by the ESSI technique. In the nonlinear regime, SNR.sub.NL I/[1/M.sup.2.Math.I(I/I.sub.0).sup.2] and in the optimal launch power density regime, (SNR.sub.NL).sub.opt) (M.Math.I.sub.0/n.sub.0).sup.2/3 where M is the number of parallel pairs of fibres. So, for one pair of fibres the gain is zero but for more than one pair, in the nonlinear regime, SNR.sub.NL increases by 6 dB each time the number of fibres is doubled, and in the optimal launch power regime, SNR.sub.NL increases by 2 dB each time the number of fibres is doubled.
[0063] FIGS. 6(a) and 6(b) show two examples of MZ modules for use in all aspects of the present invention. In FIG. 6(a) there are two MZIs making up the MZ module, and in FIG. 6(b) there are three MZIs making up the MZ module. In both cases, one of the rightmost outputs of the MZ module is connected to a network element and the other is connected to a detector for a feedback control system.
[0064] In order to validate the present invention, the schematic in FIG. 3(a) has been assembled experimentally where the phase controlled MZI design in FIG. 6(a) is used. The experimentally assembled phase controlled MZI have shown an excess loss as low as 0.35 dB, demonstrating that this method can be implemented with negligible excess loss. The system experiment considered 10 Gbps intensity modulated (IM) signals, direct detected (DD) after coherent superimposition of two copies transmitted over two spools of 20 km of conventional single-mode fibre (SMF-28). The 10 Gbps signal was generated and detected using a commercial transceiver from JDSU part number JXP01TMAC1CX5GE2. Using the transceiver, the bit-to-error rate (BER) was measured and the Q-factor (Q) calculated using Q(BER)=erfc(BER/{square root over (2)}). Note that Q.sup.2 is directly proportional to SNR.sub.NL. Finally, the results for the twin-fibre scheme and for the reference conventional (single-fibre) scheme were obtained under the same noise loading conditions. The results are shown in FIG. 7.
[0065] FIG. 7 shows that the twin-fibre invention improves the Q.sup.2 (thereby SNR.sub.NL) by 2.33 dB when operating at the respective optimal launch power. This improvement is slightly higher than the value predicted of 2 dB, which can be related with a reduction of the stimulated Brillouin scattering not accounted in section 4. Stimulated Brillouin scattering is responsible for the reflection of a fraction of the power of an incident beam. Above a certain threshold power stimulated Brillouin scattering can reflect most of the power of an incident beam. Moreover, in FIG. 7, it can be seen that in the linear regime the twin-fibre system presents a small penalty which is due to the 0.35 dB excess loss of the MZI. In the nonlinear regime, the SNR.sub.NL improvement approaches the 6 dB predicted in section 4.
