Patent Grant
11646790
This application is a U.S. National Phase of International PCT Application No. PCT/IB2020/060372 filed on Nov. 4, 2020, which claims priority to Italian Patent Application No. 102019000020554 filed Nov. 7, 2019, each of which are incorporated herein by reference in their entirety.
This invention refers to a method of controlling reconfigurable optical devices that can be used, for example, but not limited to, in the telecommunications sector in reconfigurable optical networks.
An example of a reconfigurable optical device is described in U.S. Pat. No. 6,892,021. This document describes an optical gain equaliser filter with a Waveguide Grating Router equipped with Mach-Zehnder adjustable optical attenuators, each associated with a relative wavelength of the optical channels used.
Moreover, document Schiffer, P. M. J., et al. “Smart dynamic wavelength equalizer with on-chip spectrum analyzer, IEEE Photonics Technology Letters 12.8 (2000): 1019-1021, describes a dynamic wavelength equalizer using two WGRs and using feedback control via a spectrum analyzer.
In addition, document Schlipf, T. R., et al. “Design and analysis of a control system for an optical delay-line circuit used as reconfigurable gain equalizer”. Journal of lightwave technology 21.9 (2003): 1944, describes an open loop control system for a reconfigurable gain equalizer formed by a two-port lattice-form optical delay-line circuit.
The Applicant has noted that the closed loop control techniques of the known art are too complex, both in computational terms and in relation to the structure of the control circuit.
The present invention addresses the problem of providing an optical system that shows control techniques of a reconfigurable device of the system itself that are not particularly onerous computationally and complex from a structural point of view.
According to a first aspect, an object of the present invention is an optical system as described by claim 1 and its preferred embodiments as defined by claims 2-15.
Another object of the present invention is also the method of controlling an optical system as defined by claim 16.
This invention is described in detail below, by way of example and without limitation, with reference to the attached drawings:
In this description, similar or identical elements or components will be shown in the figures with the same identifying symbol.
In particular, optical system 100 is such as to operate with electromagnetic radiation at wavelengths between 300 nm and 5000 nm, preferably between 1480 nm and 1620 nm.
For example, optical system 100 is a system that operates in the field of optical telecommunications and, in particular, in reconfigurable optical networks.
The reconfigurable optical device 103 (shorter, reconfigurable device) is such that it operates according to the WDM (Wevelength Division Multiplexing) technique on a plurality of M optical channels (at least two optical channels) i.e. with M optical signals with carriers at different wavelengths.
In particular, the reconfigurable device 103 includes at least one adjustable optical element Gi (such as, for example, an optical delay line, an adjustable optical coupler or an adjustable attenuator) configured to operate in WDM. As an example, a single adjustable optical element Gi with M channels or multiple adjustable optical elements Gi operating on M channels can be used.
The reconfigurable device 103 is also equipped with a plurality of N actuators A1-AN, associated with the adjustable optical elements Gi and such as to modify the optical characteristics (for example, the refractive index and/or the attenuation of the medium from which the adjustable device 103 is made) according to the corresponding S1-SN control signals supplied by the control device 110. The reconfigurable device 103 can assume a discrete number of states depending on the value of its N state variables θ1, . . . , θN which are controlled by the N control signals S1-SN.
Note that the N number of A1-AN actuators defines the number of degrees of freedom of the reconfigurable device 103, i.e. the number of independent variables needed to fully determine the state of the reconfigurable device 103 itself.
Advantageously, the number of degrees of freedom N of the reconfigurable device 103 is lower than the number of channels M on which the reconfigurable device itself operates.
The actuators A1-AN can be such as to induce a change in the optical parameters (e.g. phase or amplitude) of the relevant adjustable optical elements Gi. For example, the following devices can be used as actuators A1-AN: thermo-optical, electro-optical, piezoelectric, electro-absorbent, electro-mechanical, electrochemical or fully optical actuators (based or not on non-linear optical effects).
With regard to the exemplyfying sector of reconfigurable optical networks, the reconfigurable device 103 can be, for example: an optical filter, an equalizer filter, a dispersion compensator filter, a FIR filter, an IIR filter, a lattice filter, a binary tree filter.
For example, the reconfigurable 103 device can be made using integrated waveguide technology on an optical platform (or optical chip). Some examples of optical platforms that can be used include: semiconductor platforms (e.g.
silicon, indium phosphide, gallium arsenide), amorphous glasses (silicon dioxide, silicon nitride, silicon oxyfuride, silicon oxycarbon, silicon carbide), polymers and crystals (lithium niobate), possibly integrated with two-dimensional materials (graphene, silicene), and possible hybrid integrations of the same.
According to the example in
In addition, the reconfigurable optical device 103 is equipped with an optical stimulation port 111, connected to the optical source 106, and an optical monitoring port 112, connected to the optical-electric conversion device 200.
The plurality of the M optical input signals in the input signal I occupies an overall band Δλ, which identifies the working wavelength range for the reconfigurable device 103.
In particular, referring to applications in linear mode, in each of the states identified by state variables θ1, . . . , θN the reconfigurable device 103 behaves, in each of the states it can assume, as a time-invariant linear system.
The transmission of the input signal I from the optical input port 101 to the optical output port 102 can be described through the frequency response H12,i(f) of the reconfigurable device 103 or equivalently by the wavelength response H12,i(λ), where the subscript “i” indicates the generic state assumed by the device itself.
Optical source 106 is configured to generate an optical Sin stimulation signal which is supplied to stimulation port 111. Optical source 106 is configured to emit optical radiation over a wavelength range greater than or equal to the operating wavelength Δλ range of the reconfigurable device 103. Optical monitoring port 112 is configured to provide an optical monitoring signal Sout as the output corresponding to the optical stimulation signal Sin.
In the case of integrated optical devices, the optical source 106 can be integrated on the same optical platform (i.e., an optical chip) as the reconfigurable device 103 or can be external to that platform and connected to the reconfigurable device 103 via an optical fibre.
Preferably the optical source 106 includes a superluminescent diode (SLD), but other broadband sources can be used, such as, for example, the “amplified spontaneous emission” noise (ASE noise) of a fibre amplifier (e.g. erbium doped fibre amplifier, EDFA) or a semiconductor optical semiconductor amplifier (SOA), from a “supercontinuum laser” type source, from a laser array (e.g., distributed laser feedback, DFB), a comb spectral array (comb) generated by a fibre comb generator or integrated on an optical chip.
The optical-to-electrical conversion device 200 is configured to receive the monitoring optical signal Sout and provide (for example, on electrical terminals 108) a set of electrical intensity signals SEL1-SELN, each representative of an intensity of the monitoring signal Sout evaluated at a relative wavelength. Note that the set of intensity electrical signals SEL1-SELK has a cardinality equal to K. Preferably, this cardinality K is equal to N, i.e. the group of intensity electrical signals SEL1-SELK has a cardinality equal to the number of degrees of freedom of the reconfigurable device 103.
According to the particular example shown again in
The spectral slicer 105 is configured to transmit to its generic k-th output port a selected Soutk optical signal, corresponding to a portion of the monitoring signal Sout centred around a k-th wavelength λk.
The spectral slicer 105, used to select the K wavelengths k to be monitored can be realized according to different technological and architectural solutions. For example, the spectral slicer 105 is a passive device that, i.e., does not need an external active control to select the λk wavelengths.
Possible architectures that can be used for the spectral slicer 105 include Array Waveguide Gratings” (AWG), echelle gratings and other types of interferometric filters such as Mach Zehnder interferometers, Bragg gratings, ring resonators and any combinations thereof.
The spectral slicer 105 is preferably made in waveguide and is, for example, integrated on the same optical platform as the reconfigurable optical device 103. The spectral slicer 105 can also be realised with alternative technologies, using for example discrete optical components in free space, optical fibre components and combinations of the same.
According to this example, the 104 optical-electric converter comprises a plurality of photo detectors configured to convert the K sampled optical signals Sout(λk) into the K intensity electrical signals SEL1-SELK. The control device 110 is configured to control the plurality of actuators A1-AN according to said set of intensity electrical signals SEL1-SELK, generating the N control signals S1-SN, according to a pre-established control law.
With regard to the control law, for reconfigurability purposes, control device 110 operates so that the i-th state assumed by the reconfigurable device 103, in operating conditions, is as close as possible to an i-th “desired” state. For example, control device 110 is such as to define the plurality of control signals S1-SN using the method of minimising the mean square error between an actual transfer function of reconfigurable device 103 and a desired transfer function of reconfigurable device 103.
The control device 110 can be realised, as an example, by means of a microcontroller, a CPU (Central Processing Unit), an FPGA (Field Programmable Gate Array) or a DSP (Digital Signal Processor), programmed according to the control methodology described below.
Note that the optical system 100 can also include an optical apparatus 300 (APP) operationally associated with the reconfigurable optical device 103. For example, the optical apparatus 300 can be an optical amplifier (in particular, of the doped fibre type) that allows the long distance transmission of optical signals without optoelectronic conversion and regeneration. Optical amplifiers commonly operate on a large number of optical signals, for example, on more than one hundred signals.
According to this example, the reconfigurable optical device 103 can be an equalizer filter configured to equalize the gain band of the erbium doped fibres of the optical amplifier 300 that do not have a constant gain over the entire frequency range occupied by the signals.
The use of the reconfigurable equalizer filter 103 allows adapting the optical amplifier 300 to the needs of a reconfigurable optical network. For example, for the erbium doped fibre optical amplifier 300, the reconfigurable equaliser filter 103 can have a number of degrees of freedom N=20 and operate on a number of optical channels M=100.
In the following, an example of a control method that can be used by the optical system 100 to reconfigure the reconfigurable device 103 will be described.
Referring to the wavelength domain, the optical output signal O (λ) supplied to optical output port 102 is given by the following expression:
O(λ)=H12,i(λ)|(λ) (1)
where I(λ) is the input signal I, expressed in the wavelength domain and H12,i(λ) is the already defined wavelength response of the reconfigurable device 103; the subscript “i” indicates the generic status assumed by the device itself, in relation to the transmission of the input signal I from input optical port 101 to output optical port 102.
To facilitate the understanding of the following mathematical notations, in
The control method starts when there is a request to reconfigure optical system 100 and in particular reconfigurable device 103.
As already mentioned, the described control method performs monitoring and control of the response in wavelength H12,i(λ) so that an i-th “effective” state, defined by an effective wavelength response {tilde over (H)}12,i(λ), assumed by the reconfigurable device 103 under operating conditions is as close as possible to an i-th “desired” state, defined by a response in H12,i(λ).
The optical stimulation signal Sin, generated by the optical source 106, is supplied at the input of stimulation port 111 of the reconfigurable device 103 in order to measure in real time the actual state Hi,e(λ) and evaluate the deviation (i.e., the distance) with respect to the desired state Hi,d(λ).
Note that the desired Hi,d(λ) states are identified in advance and stored, for example in a lookup table, in a control device memory 110, but can also be updated and modified dynamically during operation of the reconfigurable device 103.
The reconfigurable device 103 receives the stimulation signal Sin and returns the monitoring signal Sout(λ) to the optical monitoring port 112; the optical monitoring signal Sout(λ) is described in the following relation:
Sout(λ)=H34,i(λ)Sin(λ) (2)
The function H34,i(λ) identifies the transfer function of the reconfigurable device 103 from the stimulation port 111 to the monitoring optical port 112, when the device itself is in the i-th state to which the transfer function H12,i(λ), is associated, relative to the transmission from the input optical port 101 to the output optical port 102.
It should be noted that for the purposes of the following discussion the reconfigurable device 103 is considered as having the following properties, defined below according to the transfer functions between the optical ports of the device itself:
In the properties indicated above, the term “absence” is to be understood in the sense that the retroreflexsion, coupling or losses indicated above are null or negligible for the purposes of the following discussion.
As the expert in the field acknowledges, the above mentioned properties apply to the relationships:
|H12,i(λ)|2+|H13,i(λ)|2=1
|H31i(λ)|2+|H34,i(λ)|2=1
from which it can be seen that:
|H12,i(λ)|2=|H34,i(λ)|2 (3)
The relation (3) shows how the optical monitoring signal Sout(λ) monitoring, associated with the transfer function |H34,i(λ)|2, provides the same information as the direct monitoring of the optical output signal O (λ), associated with the transfer function |H12,i(λ)|2.
Given the reciprocity of the reconfigurable device 103, the transfer functions |H34,i(λ)|2e|H43,i(λ)|2 are theoretically identical and both could be monitored. However, in practical cases it is convenient to use a stimulation signal that propagates in the opposite direction (counterpropagant) to the signal of interest. In fact, in case of co-propagant signals, the reconfigurable device 103 could be responsible for crosstalk phenomena and transfer part of the input stimulation signal to port 4 towards port 2. Therefore, a counterpropagant configuration is preferable even if it is not the only one possible.
Since the spectrum of the Sin(λ) stimulation signal is known, the transfer function H34,i(λ) can be derived directly from monitoring signal the Sout(λ) (at monitoring port 112) through the relationship (2).
Note also that system 100 operates on the basis of knowledge of the spectrum of the signal Sout(λ) only for a K number of wavelengths, preferably equally spaced, and preferably equal to the number of degrees of freedom N of device 103 (K=N).
Note that the K number of the wavelengths at which the spectrum of the signal Sout(λ) is considered can also be chosen to be greater than the number of degrees of freedom N: K>N. In this case, system 100 is particularly robust against noise, but is more complex than if K is equal to N.
On the other hand, by choosing the number K less than the number of degrees of freedom (K<N) the system 100 performs worse than K≥N.
The K number can be between a minimum Kmin and a maximum Kmax value. For example, the minimum value can be given by Kmin=N−20% N, or Kmin=N−5% N. For example, regarding the maximum value, Kmax=N+100% N, or Kmax=N+50% N, or Kmax=N+20% N.
With regard to the choice of the number K, please note that in the system 100 it is not required to be equal to the number of optical channels M, but can also be less or much less than the number of optical channels M (K<M). For example, when the reconfigurable device 103 is used in an amplification system with M=130 channels, the number of monitored wavelengths K may be less than 15%, i.e. K<15% M. Other possible example values are, K<50% M and K<30% M.
The number K is chosen, depending on the application, by appropriately combining both the above mentioned relations concerning the number of degrees of freedom N and the above mentioned relations concerning the number of optical channels M, considering a compromise between robustness and complexity.
Moreover, it should be noted that for the control method, it is sufficient to know only the intensity of the optical monitoring signal |Sout(λ)|2 and not its phase at various wavelengths.
Therefore, the information that is used by the control device 110 is the intensity of the optical monitoring signal:
|Sout(λk)|2 (4)
where the subscript k=1.2, . . . K indicates the discrete frequency at which the spectral power density |Sout(λ)|2 is sampled.
The spectral slicer 105 receives the monitoring optical signal Sout(λ) and transmits to each relevant output port 109 a sampled optical signal Sout(λk) corresponding to a portion of the monitoring optical signal Sout(λ) centred around the k-th wavelength λk. In particular, the spectral slicer 105 provides on its output ports 109 a plurality of sampled optical signals SS1-SSK in parallel mode.
Each optical output 109 of the spectral slicer 105 is optically connected to a photo detector of the optical-electric converter 104 which measures the input optical intensity and provides on a related terminal 108 an electrical signal SEL
The plurality of electrical signals SEL1-SELK is sent to the control device 110 which monitors in real time the effective frequency response |{tilde over (H)}32,i(λk)|2=|{tilde over (H)}12,i(λk)|2 assumed by the reconfigurable device 103 under operating conditions at k-th wavelength λk solving the equation:
|{tilde over (H)}34,i(λk)|2=|S34,i(λk)|2/|Sin(λk)|2.
Control device 110 compares the current state {tilde over (H)}12,i(λ) with the desired state H12,i(λ) and identifies the control signals S1-SN to be applied, via control terminals 107, to the actuators A1-AN of the reconfigurable device 103 to bring it and keep it in the desired state.
For example, the determination of the control signals S1-SN can be carried out, on the basis of the current state {tilde over (H)}12,i(λ) and the desired state H12,i(λ), according to the method of minimising the mean square error; note however that other methods can also be used. An example of the method of minimizing the mean squared error is described later with reference to a simulation of the control method.
Please note that during an initialisation phase of the optical system 100, the values of the control signals S1-SN to be applied can be taken from a look up table obtained from numerical simulations of the reconfigurable device 103.
Once these values are applied to the reconfigurable device 103, proceed to apply the method described above to bring it to the desired state indicated by the look up table.
Simulation
Σk=1K=[|{tilde over (H)}12,i(λ)|2−|H12,i(λk)|2] (5)
between the current transfer function {tilde over (H)}12,i(λ) and the desired transfer function H12,i(λ) for the generic state i. For the optimization of the transfer function of the reconfigurable device 103 alternative cost functions to those expressed by the equation (5) can be used, as well as other optimization algorithms, such as non-linear optimization, genetic algorithms, particle swarm optimization, machine learning, neural networks and others known in the literature.
In this simulation we first applied the method of the known technique, according to which the transfer functions |H12,i(λk)|2 in the three different states (301, 302 and 303) measured in a high number of wavelengths (open circles), equal to the number of optical channels used in the wavelength range of interest (K=M=130), were taken into account. By applying this conventional method, the curves shown in
Applying instead the control methodology described with reference to the optical system 100, the transfer functions |H12,i(λk)|2 are measured in a limited number of wavelengths (K=N=15) equal to the number of degrees of freedom of the reconfigurable device 103 (full circles). Applying the method described with reference to system 100 we obtained the curves shown in
According to another form of realization of the optical system 100, schematically shown in
Optical output 209 of the tunable optical filter 205 is optically connected to the photodetector 204 which measures the input optical intensity and provides an electrical signal SELk with a current or voltage proportional to the intensity |Sout(λk)|2.
By sequentially tuning the tunable detector 200 over time it is possible to obtain information on the current transfer function {tilde over (H)}12,i(λ) of the reconfigurable device 10, around all the frequencies of interest.
The electrical SEL1-SELK signals output sequentially from the 204 photodetector are sent to control device 110 and provide monitoring of the actual frequency response {tilde over (H)}12,i(λ) of the reconfigurable device 103, under operating conditions.
Possible architectures that can be used for the tunable optical filter 205 include: optical ring resonators, Mach Zehnder interferometers, Bragg gratings, and possible combinations of these.
The tunable detector 200 is preferably realized in waveguide and is preferably integrated on the same photonic platform as the reconfigurable optical device 103, already described, or it can be realized with discrete optical components in free space, optical fibre components and combinations of the same.
For example, for tuning the tunable optical filter 205, the electrical control signal SCR (generated by the control device 110) can be used, which acts on actuators (not shown) integrated in the tunable optical filter 205. These actuators modify the behaviour of the tunable filter 205 by modifying the optical parameters of the material medium in which the light radiation propagates exploiting, for example, the thermo-optical effect, the electro-optical effect, or the elasto-optical effect; alternatively, micromechanical actuators (MEMS) can be used which modify the path of the light radiation in the device.
The curve 802 in
The further optical system 400 comprises a first circulator 401 and a second optical circulator 402. The first optical circulator 401 is equipped with a first port 403 for the input signal I and a second port 404, connected to the optical input port 101 of the reconfigurable device 103 to which the input signal I can be supplied. The optical input port 101 of the reconfigurable device 103 is also such as to feed the second port 404 of the first optical circulator 401 with the monitoring optical signal Sout. The first 401 circulator is equipped with a third port 405 connected to the spectral slicer 105 to supply the latter with the optical monitoring signal Sout.
The second optical circulator 402 includes a relative first port 406 connected to output port 102 of the reconfigurable device 103. The optical output port 102 is such as to supply the output signal O to the second optical circulator 406 and is also such as to receive the stimulation signal Sin.
The second optical circulator 402 is also equipped with a relative second port 407, configured to supply the output signal O, and a relative third port 408, configured to receive the stimulation signal Sin generated by the optical source 106 and to be transmitted to the corresponding first port 406, then to the reconfigurable device 103 (via output port 102).
If the case that the reconfigurable two-port optical device 103 is reciprocal, the transfer function H21 is equal to the transfer function H12.
It should be noted that the structure that has the reconfigurable two-port optical device 103, equipped with the two optical circulators 401 and 402, is also applicable to the form of construction in
The optical lattice filter 103 comprises a plurality of optical couplers K1-K14 and a plurality of actuators Bal 1, 2 . . . 7 and Unbal 1, 2, . . . 6, suitable for introducing delays or imbalances in the optical channels, for a total of thirteen actuators.
The lettice filter 103 in
The optical couplers K1-K14 are an example of the optical elements Gi described with reference to
Note that each of the actuators Bal 1, 2 . . . 7 and Unbal 1, 2, . . . 6 acts on the optical behaviour of a corresponding Kj optical coupler which operates in WDM mode, i.e. it allows the propagation of several optical channels.
In general, the reconfigurable device 103 can be an optical filter that includes as optical elements Gi: binary tree or lattice interferometers, AWG (Arrayed Waveguide Gratings) or similar structures that use as power dividers, for example, MultiMode Interferometers (MMI), directional couplers, or y-branches or similar.
Please note that the solution described above is mainly, but not exclusively, applied in the telecommunications industry, specifically in the field of reconfigurable optical networks. Examples of other possible applications of the lessons described are:
1) optical devices for both fibre and waveguide sensors, which require the stabilisation of optical circuits to process sensor readings,
2) optical devices for distance measurements such as LIDARs, which require a very wide operating temperature range and can take advantage of the stabilization obtained with the invention,
3) optical circuits for 5G wireless networks, which can exploit photonic circuits to improve the performance of mobile network coverage using, for example, beamforming networks with integrated photonic circuits controlled by the method described in the invention.
4) reconfigurable optical circuits that can introduce an adjustable delay time, used for example in optical interferometry, optical tomography and other applications where it is necessary to synchronise relative delays between two or more optical signals;
The optical systems described above are particularly advantageous in terms of simplicity and performance. In fact, these optical systems allow managing their reconfiguration by monitoring a number of signals (i.e. sampled optical signals SS1-SSK) lower than the number of optical channels on which the system itself operates, maintaining the desired performance.
Moreover, the described optical systems have the advantages offered by closed circuit control without requiring a high complexity of the actuator system controlled by the control device.
The lower number of signals to be monitored implies a reduction in the costs related to the necessary components, the physical dimensions of the control system and facilitates packaging operations.
| Number | Date | Country | Kind |
|---|---|---|---|
| 102019000020554 | Nov 2019 | IT | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2020/060372 | 11/4/2020 | WO |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2021/090205 | 5/14/2021 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 6208780 | Li et al. | Mar 2001 | B1 |
| 6892021 | Doerr | May 2005 | B2 |
| 9577763 | Al Sayeed | Feb 2017 | B2 |
| 9985726 | Al Sayeed | May 2018 | B1 |
| 9986317 | Al Sayeed | May 2018 | B1 |
| 10050738 | Akasaka | Aug 2018 | B1 |
| 10536235 | Al Sayeed | Jan 2020 | B2 |
| 20030053175 | Szczepanek et al. | Mar 2003 | A1 |
| 20170134089 | Mansouri Rad | May 2017 | A1 |
| Number | Date | Country |
|---|---|---|
| 2006128254 | Dec 2006 | WO |
| Entry |
|---|
| International Search Report; International Searching Authority; International Application No. PCT/IB2020/060372; dated Feb. 22, 2021; 5 pages. |
| Written Opinion of the International Searching Authority; International Searching Authority; International Application No. PCT/IB2020/060372; dated Feb. 22, 2021; 8 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20220385358 A1 | Dec 2022 | US |
