WO2022248053A1

WO2022248053A1 - Method of optical performance monitoring in an optical network

Info

Publication number: WO2022248053A1
Application number: PCT/EP2021/064287
Authority: WO
Inventors: Nathalie MORETTE; Ivan FERNANDEZ DE JAUREGUI RUIZ; Yvan Pointurier
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2021-05-27
Filing date: 2021-05-27
Publication date: 2022-12-01
Also published as: EP4335049A1

Abstract

Provided is a computer-implemented method of optical performance monitoring in an optical network (200, 300). The optical network includes at least one optical multiplex section, OMS (202, 308), formed from one or more spans (310). Each span includes a fibre (312) and an optical amplifier (314). The method includes estimating an output power of one or more channels at the output of an optical amplifier of a span in an OMS, based on an estimated lump loss ratio for each span in the OMS. The method includes calculating a power distance between the estimated output power and a monitored output power for each channel through the OMS. The method includes updating estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span in the OMS using an optimisation function to minimise a total power distance for all channels.

Description

METHOD OF OPTICAL PERFORMANCE MONITORING IN AN OPTICAL

NETWORK

TECHNICAL FIELD

The disclosure relates generally to a computer-implemented method of optical perfor- mance monitoring in an optical network, and more particularly, the disclosure relates to an optical performance monitoring device for the optical network. Moreover, the disclo sure also relates to an optical network including at least one optical multiplex section, OMS, and the optical performance monitoring device for monitoring the performance of the optical network. BACKGROUND

An optical network is a communication system that uses light signals, instead of elec tronic ones, to send information between two or more points. The points could be com puters in an office, large urban centers, or even nations in the global telecommunications system. Depending on the optical network size, a 2 decibel, dB, signal-to-noise ratio, SNR, margin may translate into 10-25% of additional capacity of the optical network. The SNR is used to measure a line quality and defines a minimum limit at which a signal level can be decoded with a bit error rate, BER, below a given target. The SNR may be predicted using a quality of transmission, QoT, estimator tool, which leverages all avail able information. The QoT tool is a formula depending on the following inputs. First, at network initial design, the inputs are received from specifications, datasheets, information provided by operators. The inputs provided by the operators may be unreliable. Second, during network operation (i.e. upgrades), the above information may be entered manually in the software-defined networking, SDN, control plane and the other information may be available through monitoring. However, the information that is available through mon- itoring may be inaccurate. The accuracy of the QoT tool mainly depends on the accuracy of its inputs. Additional physical monitoring may help, however, it also has inaccuracies and is expensive. A better QoT inputs accuracy may lead to (a) lower margin (i.e. more capacity, reach, robustness), (b) more information to optimize an optical line (e.g. set/equalize channels launch power). The inaccurate and unreliable inputs to the QoT tool may lead to waste of its capacity. Typically, an optical multiplex section, OMS, designates the optical line between 2 nodes. A node may be a reconfigurable optical add/drop multiplexer, ROADM, or a Dynamic Gain Equalizer, DGE. The lumped losses before/after fibres within the OMS are not very well known. An uncertainty on those losses, which are due to several possibly dirty con nectors and patch cords, can reach several dB. The total span loss for each span (e.g. a loss between 2 amplifiers) is known by power monitoring at Erbium-doped fibre ampli fiers, EDFAs, and the fibre loss is known through fibre length and attenuation coefficient. Hence, the sum of lumped losses on each span is known, but the distribution of the loss before/after the fibre is unknown. Further, these losses strongly affect the power, hence, signal-to-noise ratio (SNR) of signals goes through links (i.e. OMS) via the Raman and Kerr effects. Even, if other existing techniques compute SNR for services, inaccurate knowledge of those lumped losses may lead to inaccurate SNR estimation and power allocation etc. which in turn may lead to unwanted margins, hence the capacity of the optical network is wasted. So, the problem is to estimate with good accuracy the lumped losses at the start and end of each fibre span in the optical network.

A known solution uses an optical time-domain reflectometer, OTDR, which is an optoe lectronic instrument used to characterize an optical fibre. The OTDR measures the longi tudinal loss profile between 2 amplifiers. However, this requires dedicated hardware and signal analysis to locate the losses and estimate their magnitude. Further, the OTDR is expensive, and as the OTDR devices are located next to the amplifiers, the lumped losses to estimate are located either too close to the OTDR, (i.e. in its “blind area”), or too far, where losses estimation is inaccurate.

Another known solution estimates the longitudinal loss using a software-based method such as “digital backpropagation”. The software-based method is complicated, as it may additionally need acceleration hardware and modifications to a digital signal processing chip on optical transponders. Further, the method is not accurate and is only relevant for qualitative (e.g. failure detection) rather than quantitative analysis of an optical line.

Another known solution removes uncertainties on many parameters of the QoT tool by associating machine-learning and analytical models for QoT estimation. Another known solution estimates the QoT based on power monitoring at wavelength selective switches, WSS and SNR monitoring of many services compute “OSNR penalty” of an optical mul tiplex section, OMS, to be used for bit error rate, BER, evaluation of a new service. An other known solution estimates the QoT based on SNR from many services, and computes a shape of gain for each EDFA, then uses them for the BER evaluation of a new service. In these known solutions, the knowledge of amplifiers, power profile etc. are refined based on software techniques only. However, the problem of lumped losses estimation is completely ignored in these known solutions.

Therefore, there arises a need to address the aforementioned technical problem/draw- backs in monitoring a performance of the optical network.

SUMMARY

It is an object of the disclosure to provide a computer-implemented method of optical performance monitoring in an optical network, an optical performance monitoring device for the optical network and an optical network including at least one optical multiplex section, OMS, and the optical performance monitoring device for monitoring the perfor mance of the optical network while avoiding one or more disadvantages of prior art ap proaches.

This object is achieved by the features of the independent claims. Further, implementation forms are apparent from the dependent claims, the description, and the figures.

The disclosure provides a computer-implemented method of optical performance moni toring in an optical network, an optical performance monitoring device for the optical network, and an optical network including at least one optical multiplex section, OMS, and the optical performance monitoring device for monitoring the optical network per formance.

According to a first aspect, there is provided a computer-implemented method of optical performance monitoring in an optical network. The optical network includes at least one optical multiplex section, OMS, formed from one or more spans. Each span includes a fibre and an optical amplifier. The method includes estimating an output power of one or more channels at the output of an optical amplifier of a span in an OMS, based on an estimated lump loss ratio for each span in the OMS. The lumped loss ratio is a ratio of lumped losses before the fibre and lumped losses after the fibre in the span. The method includes calculating a power distance between the estimated output power and a moni tored output power for each channel through the OMS. The method includes updating estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span in the OMS using an optimisation function to minimise a total power dis tance for all channels.

The method accurately estimates the optical performance in the optical network. The method accurately estimates the lumped losses before the fibre and lumped losses after the fibre in the span in the optical network, thereby improving the optical performance estimation of the optical network. The method uses the optimisation function to minimise the total power distance for all channels in the OMS. The method does not require any additional new hardware for monitoring the optical performance of the optical network. The method relies on metrics that are already monitored in a typical optical network for monitoring the optical performance of the optical network. For example, the method mon itors the optical performance of the optical network based on metrics such as per-channel k power within reconfigurable optical add/drop multiplexer, ROADMs or Dynamic Gain Equalizers, DGE, that are Pl(k) (i.e. a start of the OMS) and P2(k) (i.e. an end of the OMS), a signal-to-noise ratio, SNR of established services that can be extracted from pre forward error correction, pre-FEC and bit error rate, BER. The method provides improved accuracy in estimating the signal-to-noise ratio, SNR within 0.2 decibel, dB, down from more than ldB.

The method enables the improved estimation of the lumped losses even if the power var ies in the optical network within a reasonable range for network operation. The method estimates the lumped losses more accurately even in the presence of filtering or other uncertainty on other key physical parameters. The method enables separation of the opti cal amplifier and nonlinear noises, which can be used for health analytics, especially for determining where a loss of optical performance occurs in the optical network. Hence, the method can be used as an optical performance monitor, OPM, or a software OPM. The method can run on a standard computer, thereby eliminating a need for dedicated computing hardware for monitoring the optical performance of the optical network. The outputs of the method may be leveraged by algorithms (e.g. QoT estimation algorithm, etc.)·

Optionally, the method further includes (i) estimating a signal-to-noise ratio, SNR, for each service carried by a channel through the OMS, based on the estimated lump loss ratio for each span in the OMS, (ii) calculating an SNR distance between the estimated SNR and a monitored SNR for each service carried by a channel through the OMS, and (iii) outputting estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span, if a value of the total SNR distance for all channels satisfies a predefined condition. The predefined condition may include one or more of the value falling below a threshold value, and the value increasing after previously decreasing, or the value increasing for a predefined number of iterations.

Optionally, the method further includes requesting a set of monitored inputs from a phys ical layer database, including the monitored output power for each channel through the OMS, and the monitored SNR for each channel through the OMS. Optionally, the method further includes storing the output estimated values in the physical layer database.

Optionally, the method further includes requesting a set of parameters from the physical layer database, including one or more of an average gain, tilt or noise figure value for an amplifier and/or a linear attenuation, chromatic dispersion, nonlinear parameter or length of a fibre and/or a type, bandwidth or filtering response or attenuation of a filter and/or a routing or a spectrum allocation of services and/or a back-to-back noise response of a transponder and/or a polarisation dependent loss parameter of any element.

An initial value of the estimated lump loss ratio for at least one span may be based on arbitrary initialisation values for the lumped losses before the fibre and lumped losses after the fibre for each span. Optionally, the optimisation function is a gradient descent function.

According to a second aspect, there is provided an optical performance monitoring device for an optical network. The optical network includes at least one optical multiplex section, OMS, formed from one or more spans. Each span includes a fibre and an optical ampli fier. The device includes one or more processors that is configured to estimate an output power of one or more channels at an output of an optical amplifier of a span in an OMS, based on an estimated lump loss ratio for each span in the OMS. The lumped loss ratio is a ratio of lumped losses before the fibre and lumped losses after the fibre in the span. The one or more processor is configured to calculate a power distance between the estimated output power and a monitored output power for each channel. The one or more processor is configured to update estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span in the OMS using an optimisation function to minimise a total power distance for all channels.

The optical performance monitoring device accurately estimates the optical performance in the optical network. The optical performance monitoring device accurately estimates the lumped losses before the fibre and lumped losses after the fibre in the span in the optical network, thereby improving the optical performance estimation of the optical net work. The optical performance monitoring device uses the optimisation function to min imise the total power distance for all channels in the OMS. The optical performance mon itoring device does not require any additional new hardware for monitoring the optical performance of the optical network. The optical performance monitoring device relies on metrics that are already monitored in a typical optical network for monitoring the optical performance of the optical network. The optical performance monitoring device provides improved accuracy in estimating the signal-to-noise ratio, SNR within 0.2 decibel, dB, down from more than ldB.

The optical performance monitoring device enables the improved estimation of the lumped losses even if the power varies in the optical network within a reasonable range for network operation. The optical performance monitoring device estimates the lumped losses more accurately even in the presence of filtering or other uncertainty on other key physical parameters. The optical performance monitoring device enables separation of the optical amplifier and nonlinear noises, which can be used for health analytics, espe cially for determining where a loss of optical performance occurs in the optical network.

Optionally, the optical performance monitoring device is further configured to: (i) esti mate a signal-to-noise ratio, SNR, for each service carried by a channel through the OMS, based on the estimated lump loss ratio for each span in the OMS, (ii) calculate an SNR distance between the estimated SNR and a monitored SNR for each service carried by a channel through the OMS, and (iii) output estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span, if a value of the total SNR dis tance for all channels satisfies a predefined condition. The predefined condition may in clude one or more of the value falling below a threshold value, and the value increasing after previously decreasing, or the value increasing for a predefined number of iterations.

Optionally, the optical performance monitoring device is further configured to request a set of monitored inputs from a physical layer database, including the monitored output power for each channel through the OMS, and the monitored SNR for each channel through the OMS.

Optionally, the optical performance monitoring device is further configured to store the output estimated values in the physical layer database. Optionally, the optical perfor mance monitoring device is further configured to request a set of parameters from the physical layer database, including one or more of an average gain, tilt or noise figure value for an amplifier and/or a linear attenuation, chromatic dispersion, nonlinear param eter or length of a fibre and/or a type, bandwidth or filtering response or attenuation of a filter and/or a routing or a spectrum allocation of services and/or a back-to-back noise response of a transponder and/or a polarisation dependent loss parameter of any element.

According to a third aspect, there is provided an optical network. The optical network includes (i) at least one optical multiplex section, OMS, formed from one or more spans, and (ii) the optical performance monitoring device as described above. Each span in cludes a fibre and an optical amplifier.

A technical problem in the prior art is resolved, where the technical problem is that mon itoring the optical performance in the optical network. Therefore, in contradistinction to the prior art, according to a computer-implemented method of optical performance monitoring in an optical network and an optical perfor mance monitoring device for the optical network, the optical network, the optical perfor mance in the optical network is accurately estimated. The method accurately estimates the lumped losses before the fibre and lumped losses after the fibre in the span in the optical network, thereby improving the optical performance estimation of the optical net work. The method uses the optimisation function to minimise the total power distance for all channels in the OMS. The method does not require any additional new hardware for monitoring the optical performance of the optical network. The method relies on metrics that are already monitored in a typical optical network for monitoring the optical perfor mance of the optical network. The method enables the improved estimation of the lumped losses even if the power varies in the optical network within a reasonable range for net work operation. The method estimates the lumped losses more accurately even in the presence of filtering or other uncertainty on other key physical parameters.

These and other aspects of the disclosure will be apparent from and the implementation(s) described below.

BRIEF DESCRIPTION OF DRAWINGS

Implementations of the disclosure will now be described, by way of example only, with reference to the accompanying drawings, in which:

FIG. 1 is a block diagram of an optical performance monitoring device for an optical network in accordance with an implementation of the disclosure;

FIG. 2 is a block diagram of an optical network in accordance with an implementation of the disclosure;

FIG. 3 is an exploded view of an optical network including an optical performance mon itoring device in accordance with an implementation of the disclosure;

FIG. 4 is a flow chart that illustrates a workflow of computing optimisation functions for monitoring an optical performance in an optical network in accordance with an imple mentation of the disclosure; FIG. 5A is a schematic illustration of separation of linear and non-linear effects for all channels for a 5-span OMS using an optical performance monitoring device in accord ance with an implementation of the disclosure;

FIG. 5B illustrates a graph of a sample Amplified spontaneous emission, ASE/ nonlinear interference, NLI, effect separation in accordance with an implementation of the disclo sure;

FIG. 6A illustrates a graph of sample SNR prediction in single 5-spans OMS in accord ance with an implementation of the disclosure;

FIG. 6B illustrates SNR prediction for a multi-OMS network in accordance with an im plementation of the disclosure;

FIG. 6C illustrates a graph that illustrates the SNR prediction for a multi-OMS in accord ance with an implementation of the disclosure;

FIG. 7 is a flow diagram that illustrates a method of optical performance monitoring in an optical network in accordance with an implementation of the disclosure; and

FIG. 8 is an illustration of a computer system (e.g. an optical performance monitoring device) in which the various architectures and functionalities of the various previous im plementations may be implemented.

DETAILED DESCRIPTION OF THE DRAWINGS

Implementations of the disclosure provide a computer-implemented method of optical performance monitoring in an optical network. The disclosure also provides an optical performance monitoring device for the optical network and an optical network including at least one optical multiplex section, OMS, and the optical performance monitoring de vice for monitoring the optical network performance.

To make solutions of the disclosure more comprehensible for a person skilled in the art, the following implementations of the disclosure are described with reference to the ac companying drawings. Terms such as "a first", "a second", "a third", and "a fourth" (if any) in the summary, claims, and foregoing accompanying drawings of the disclosure are used to distinguish between similar objects and are not necessarily used to describe a specific sequence or order. It should be understood that the terms so used are interchangeable under appropri ate circumstances, so that the implementations of the disclosure described herein are, for example, capable of being implemented in sequences other than the sequences illustrated or described herein. Furthermore, the terms "include" and "have" and any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, a method, a system, a product, or a device that includes a series of steps or units, is not necessarily limited to expressly listed steps or units but may include other steps or units that are not expressly listed or that are inherent to such process, method, product, or de vice.

FIG. 1 is a block diagram of an optical performance monitoring device 100 for an optical network in accordance with an implementation of the disclosure. The optical network includes at least one optical multiplex section, OMS, formed from one or more spans. Each span includes a fibre and an optical amplifier. The optical performance monitoring device 100 includes one or more processors 102A-N. The one or more processors 102A- N are configured to estimate an output power of one or more channels at an output of an optical amplifier of a span in an OMS, based on an estimated lump loss ratio for each span in the OMS. The lumped loss ratio is a ratio of lumped losses before the fibre and lumped losses after the fibre in the span. The one or more processors 102A-N are config ured to calculate a power distance between the estimated output power and a monitored output power for each channel. The one or more processors 102A-N are configured to update estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span in the OMS using an optimisation function to minimise a total power distance for all channels.

The optical performance monitoring device 100 accurately estimates the optical perfor mance in the optical network. The optical performance monitoring device 100 accurately estimates the lumped losses before the fibre and lumped losses after the fibre in the span in the optical network, thereby improving the optical performance estimation of the opti- cal network. The optical performance monitoring device 100 uses the optimisation func tion to minimise the total power distance for all channels in the OMS. The optical perfor mance monitoring device 100 does not require any additional new hardware for monitor ing the optical performance of the optical network. The optical performance monitoring device 100 relies on metrics that are already monitored in a typical optical network for monitoring the optical performance of the optical network. The optical performance mon itoring device 100 provides improved accuracy in estimating the signal-to-noise ratio, SNR within 0.2 decibel, dB, down from more than ldB.

The optical performance monitoring device 100 enables the improved estimation of the lumped losses even if the power varies in the optical network within a reasonable range for network operation. The optical performance monitoring device 100 estimates the lumped losses more accurately even in the presence of filtering or other uncertainty on other key physical parameters. The optical performance monitoring device 100 enables separation of the optical amplifier and nonlinear noises, which can be used for health analytics, especially for determining where a loss of optical performance occurs in the optical network.

Optionally, the optical performance monitoring device 100 is further configured to: (i) estimate a signal-to-noise ratio, SNR, for each service carried by a channel through the OMS, based on the estimated lump loss ratio for each span in the OMS, (ii) calculate a SNR distance between the estimated SNR and a monitored SNR for each service carried by a channel through the OMS, and (iii) output estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span, if a value of the total SNR distance for all channels satisfies a predefined condition. The predefined condition may include one or more of the value falling below a threshold value, and the value increasing after previously decreasing, or the value increasing for a predefined number of iterations. Optionally, the optical performance monitoring device 100 is further configured to re quest a set of monitored inputs from a physical layer database, including the monitored output power for each channel through the OMS, and the monitored SNR for each channel through the OMS. Optionally, the optical performance monitoring device 100 is further configured to store the output estimated values in the physical layer database. Optionally, the optical perfor mance monitoring device 100 is further configured to request a set of parameters from the physical layer database, including one or more of an average gain, tilt or noise figure value for an amplifier and/or a linear attenuation, chromatic dispersion, nonlinear param eter or length of a fibre and/or a type, bandwidth or filtering response or attenuation of a filter and/or a routing or a spectrum allocation of services and/or a back-to-back noise response of a transponder and/or a polarisation dependent loss parameter of any element.

FIG. 2 is a block diagram of an optical network 200 in accordance with an implementation of the disclosure. The optical network 200 includes (i) at least one optical multiplex sec tion, OMS, 202 formed from one or more spans. Each span includes a fibre and an optical amplifier. The optical network 200 includes an optical performance monitoring device 204 as described above.

FIG. 3 is an exploded view of an optical network 300 including an optical performance monitoring device in accordance with an implementation of the disclosure. Optionally, the optical network 300 includes a control plane 302 and a data plane 304. The control plane 302 may be the optical performance monitoring device. The control plane 302 may be one or more processors in the optical performance monitoring device or a controller or a management plane. Optionally, the data plane 304 includes optical nodes 306 (e.g. a Reconfigurable Optical Add-drop Multiplexer, ROADM, or a Dynamic Gain Equalizer, DGE), optical transponders, TRX, located at the optical nodes 306, and at least one opti cal multiplex section, OMS 308 between the optical nodes 306. The OMS 308 is formed from one or more spans 310. Each span 310 includes a fibre 312 and an optical amplifier 314. Optionally, the optical transponders include transmitters, TX and receivers, RX (i.e. TRX = transmitter, TX + receiver, RX). Optionally, the optical transponders are used to send or receive optical channels. Optionally, a service 316 is an optical channel transiting from a transmitter to a receiver with no electrical regeneration. The Quality of Transmis sion, QoT, of the service 316 may be measured through its signal-to-noise ratio, SNR, or a bit-error-rate, BER. The BER is optionally measured at each RX before forward-error correction, FEC, and may be directly mapped to the SNR via a mathematical formula.

Optionally, the control plane 302 is used to control/configure/manage the network ele ments (e.g. the ROAD Ms, optical amplifiers, etc.) and the services 316 (e.g. anew service establishment, an implementation of a restoration strategy, etc.). Optionally, the control plane 302 follows the “software-defined networking”, SDN architecture, and the control plane 302 is centralized (i.e. the control plane 302 runs from a dedicated server managed by a network operator) and exchanges messages with each network element of the data plane 304. The control plane 302 may include various software programs and databases. The databases may be a physical layer database 318 that holds data describing a set of parameters of all network elements and services in the optical network 300. The infor mation in the physical layer database 318 may be leveraged to compute the QoT of any service (i.e. an existing or a new service, using a QoT estimator 322). Optionally, the physical layer database 318 is populated by monitored information received directly from the network elements, or datasheet information entered manually by the network operator. Optionally, the optical performance monitoring device (i.e. the control plane 302) is con figured to request a set of monitored inputs from the physical layer database 318, includ ing the monitored output power for each channel through the OMS 308, and the moni tored SNR for each channel through the OMS 308.

Optionally, the control plane 302 includes a lumped loss estimator 320. The lumped loss estimator 320 may be leveraged by other control plane functions (i.e. the QoT estimator 322 functions). The lumped loss estimator 320 may request its inputs (i.e. power and SNR) to the physical layer database 318. Optionally, the set of monitored inputs are a total output optical power of each optical amplifier 314 of each OMS 308, per-channel optical power at the beginning and end of each OMS 308, and the SNR of all services 316 in the optical network 300. The physical layer database 318 may store all the moni tored information and their values. Optionally, the optical performance monitoring device (i.e. the control plane 302) is further configured to store the output estimated values in the physical layer database 318. Optionally, the optical performance monitoring device is further configured to request a set of parameters from the physical layer database 318, including one or more of an average gain, tilt or noise figure value for an amplifier and/or a linear attenuation, chromatic dispersion, nonlinear parameter or length of the fibre 312 and/or a type, bandwidth or filtering response or attenuation of a filter and/or a routing or a spectrum allocation of the services and/or a back-to-back noise response of a tran sponder and/or a polarisation dependent loss parameter of any element. The lumped loss estimator 320 may return the value of the lumped losses before and after each span 310 of each OMS 308 of the optical network 300 to the physical layer database 318.

Optionally, with some values for the lumped losses, and based on the data in the physical layer database 318, the optical performance monitoring device (i.e. the control plane 302) is configured to estimate (a) the SNR of all services 316 using the QoT estimator 322 and (b) the power at the output of each OMS 308. The SNRs and output power may depend on the physics of the propagation of signals in the various OMS 308. The optical perfor mance monitoring device is configured to estimate an output power of one or more chan nels at an output of the optical amplifier 314 of the span 310 in the OMS 308, based on an estimated lump loss ratio for each span 310 in the OMS 308. The optical performance monitoring device is configured to calculate a power distance between the estimated out put power and a monitored output power for each channel. The optical performance mon itoring device is configured to update estimated values for the lumped losses before the fibre 312 and lumped losses after the fibre 312 for each span 310 in the OMS 308 using an optimisation function to minimise a total power distance for all channels. The lumped loss ratio is a ratio of lumped losses before the fibre 312 and lumped losses after the fibre 312 in the span 310. Accurate estimates of the lumped losses may minimize the error (i.e. a first cost function, Cl) between the estimated SNR and a monitored SNR, and between estimated output power and monitored output power (i.e. a second cost function, C2). Optionally, the lumped loss estimator 320 uses an optimisation function (e.g. an iterative algorithm) to minimize the error in two cost functions (i.e. Cl and C2). Optionally, the optimisation function is a gradient descent function. Optionally, one of the cost functions is used as a stopping criterion, while the other cost function is minimized using some optimization framework such as a gradient descent. The optimisation function is assumed to have converged when the cost is small enough. Optionally, the optimisation function starts from possibly random values for the lumped losses, and outputs accurate values for those lumped losses. Those estimates of the cost functions may be stored in the physical layer database 318 for further use by the QoT estimator 322 or any other network function.

Optionally, the optical performance monitoring device reduces the uncertainty on the set of parameters of the QoT estimator 322 for channels carrying the service 316 (e.g. an optical service). The optical performance monitoring device combines the per-channel optical power monitoring per network optical multiplex section, OMS, the total power monitoring per optical amplifier 314, and the SNR of the services 316 deployed in the optical network 300 to estimate the losses located (a) between the optical node 306 and the start of the span 310 immediately following the optical node 306, (b) between the end of the span 310 and the optical node 306 immediately following the span 310. The QoT estimator 322 may be implemented using a known model that includes Kerr and inter channel stimulated Raman scattering (ISRS) effects.

FIG. 4 is a flow chart that illustrates a workflow of computing optimisation functions for monitoring an optical performance in an optical network in accordance with an imple mentation of the disclosure. At a step 402, an arbitrary value between 0 and 1 is set for the lumped losses (i.e. R(i, s) = arbitrary value between 0 and 1) and other QoT inputs are set based on a physical layer database, where, i is an index of an OMS, s is an index of a span on the OMS and R(i, s) is a ratio of the lumped loss before a fibre and after the fibre in the span s on the OMS i. At a step 404, SNR_e(i, k) is computed for each k, where, k is an index of a channel, SNR_e(i, k) is an estimated signal-to-noise ratio, SNR, of a service carried by the channel k going through the OMS i. At a step 406, a total SNR distance Cl (i) is computed between the estimated SNR, SNRe(i,-) and a monitored SNR, SNRm(i,-) for each service carried by the channel k through the OMS i, as follows: Cl (i) = distance (SNRe(i,-), SNR_m(i,-)). At a step 408, it is checked whether C 1 (i) < e, where, Cl is a cost function to determine the total SNR distance, and is used for accurate mod eling of the Kerr effects. If NO, it goes to the step 410. If YES, the estimated lumped losses before the fibre and after the fibre for each span s is outputted as R(i, s), if a value of the total SNR distance, Cl(i) for all channels satisfies a predefined condition e. At a step 410, P2_e(i, k) is computed for each wavelength k, where P2_e(i, k) is an estimated output power of a channel k at the output of OMS i. At a step 412, a power distance C2(i) is computed between the estimated output power, P2_e(i, -) and a monitored output power, P2_m(i, -) for a channel through the OMS, i as follows: C2(i) = distance(P2_e(i, -), P2_m(i, - ), where, C2 is a cost function to determine the power distance, and is used for accurate modeling of the Raman effect. At a step 414, R(i, s) is updated for each span s. Repeat the steps 404 to 414 to estimate an accurate optical performance of the optical network.

Optionally, some of the steps may be re-ordered without a loss of generality. For example, the comparison step on Cl (i.e. the step 406) may be performed after the computation of C2, (i.e. the step 412) etc. The cost function Cl may be used for accurate modeling of the Kerr effects while the cost function C2 may be used for accurate modeling of the Raman effect. Optionally, both the cost functions (Cl and C2) are needed to estimate an accurate performance of the optical network.

Optionally, for example, the initial value of the lumped loss ratio R(i, s) is set as 0.5. The QoT tool may be used, based on the estimated lumped loss ratio R(i, s) and a monitored output power Pl_m(i, k) (power of a signal on the channel k at the start of the OMS i). The other QoT tool inputs may be found in a physical layer database. Optionally, the total SNR distance Cl is computed as follows: Cl (i) = åk(SNRe(i, k) - SNR_m(i, k))². Option ally, the power distance C2 and the lumped loss ratio computed as follows: C2(i)= åk(P2_e(i, k) -P2m(i, k))² and R(i, s) - R(i, s) - A <3C2(i)/ SR(i, s), where, i is an index of the OMS, s is an index of the span on the OMS, k is an index of the channel, PI (i, k) is a power of a signal on the channel k at the start of the OMS i, P2(i, k) is a power of a signal on the channel k at the end of the OMS i, SNR(i, k) is a SNR of a service on the channel k going through the OMS i, R(i, s) is a ratio of the lumped loss before and after the fibre in the span s for each OMS i, X_m is a metric X that is monitored, X_e is a metric X that is estimated, X may be Pl(i, k), P2(i, k) or SNR(i, k).

Optionally, in simple words, the general workflow of the optimisation function starts from arbitrary lumped loss ratio R, then SNR of all services going through the OMS i based on a current value of R is estimated. After the SNR value is estimated, a distance between the estimated and the monitored SNR is computed. The workflow stops if that difference is small enough. Otherwise, output power for the OMS is computed based on the Raman effect only. Then, the distance between the estimated and the monitored output power is computed and the lumped losses ratio R is updated, so that, the estimated and the monitored output power match (e.g. using a gradient descent approach).

FIG. 5A is a schematic illustration of separation of linear and non-linear effects for all channels for a 5-span OMS using an optical performance monitoring device in accord ance with an implementation of the disclosure. The schematic illustration shows an ex perimental 5 -span line OMS and the experimental data on a monitored output power for each channel, a monitored lump loss ratio for each span in the OMS, and a monitored SNR for each service carried by a channel through the OMS. The schematic illustration also shows the experimental data on refined information. Optionally, the optical perfor mance monitoring device estimates linear and nonlinear effects separately in an optical network. Optionally, for example, the data is taken from an experimental line, and a sig- nal-to-noise ratio, SNR, is estimated using the optical performance monitoring device which has already been validated experimentally and may be used for ground truth. Op tionally, the optical performance monitoring device estimated the lumped losses, starting from baseline 50%/50% before/after a fibre in each span in the OMS. Optionally, the estimated lumped losses and the SNR are further refined using other QoT tool inputs using the optical performance monitoring device.

FIG. 5B illustrates a graph of a sample Amplified spontaneous emission, ASE/ nonlinear interference, NLI, effect separation in accordance with an implementation of the disclo sure. The X-axis represents the channel number and the Y-axis represents the ratio PNLI/ASE. The graph shows that, before refinement, the SNR is greater than 5 dB on a ratio of the nonlinear/linear effect, and after refinement, the SNR is less than 0.2 dB on the ratio of the nonlinear/linear effect while comparing to the ground-truth value. The optical performance monitoring device may verify both linear and nonlinear effects that are predicted individually, not just their ratio. The optical performance monitoring device may verify the estimated lumped losses that are important for separating the linear and nonlinear effects for all channels for a 5-span OMS.

The optical performance monitoring device enables separation of linear and nonlinear noises in the optical network, which can be used for health analytics, especially to deter mine where a loss of optical performance occurs in the optical network. Hence, the optical performance monitoring device may be used as an optical performance monitor, OPM, or a software OPM.

FIG. 6A illustrates a graph of sample SNR prediction in single 5-spans OMS in accord ance with an implementation of the disclosure. The X-axis represents the channel number and the Y-axis represents the SNR in decibel, dB. Optionally, the SNR is estimated using an optical performance monitoring device, as described above, for all channels for the 5- span OMS. Optionally, the estimated SNR that is less than 0.2 dB is accurate. The esti mated SNR may be compared with inputs that are not refined (e.g. from datasheet). The SNR has a margin gain up to 1.2 dB, and up to 2 dB in (i.e. in total) optical signal-to- noise ratio, OSNR. The following table illustrates a single OMS network for the SNR prediction.

FIG. 6B illustrates SNR prediction for a multi-OMS (i.e. 3 OMS) network in accordance with an implementation of the disclosure. Optionally, the multi-OMS network shows same accuracy as the single OMS network. FIG. 6C illustrates a graph that illustrates the SNR prediction for a multi-OMS in accordance with an implementation of the disclosure. The graph shows that the SNR margin may be decreased from 1.2 dB to a fraction of a dB. FIG. 7 is a flow diagram that illustrates a method of optical performance monitoring in an optical network in accordance with an implementation of the disclosure. The optical network includes at least one optical multiplex section, OMS, formed from one or more spans. Each span includes a fibre and an optical amplifier. At a step 702, an output power of one or more channels is estimated at the output of an optical amplifier of a span in an OMS, based on an estimated lump loss ratio for each span in the OMS. The lumped loss ratio is a ratio of lumped losses before the fibre and lumped losses after the fibre in the span. At a step 704, a power distance between the estimated output power and a monitored output power is calculated for each channel through the OMS. At a step 706, estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span in the OMS is updated using an optimisation function to minimise a total power distance for all channels.

The method accurately estimates the optical performance in the optical network. The method accurately estimates the lumped losses before the fibre and lumped losses after the fibre in the span in the optical network, thereby improving the optical performance estimation of the optical network. The method uses the optimisation function to minimise the total power distance for all channels in the OMS. The method does not require any additional new hardware for monitoring the optical performance of the optical network. The method relies on metrics that are already monitored in a typical optical network for monitoring the optical performance of the optical network. For example, the method mon itors the optical performance of the optical network based on metrics such as per-channel k power within reconfigurable optical add/drop multiplexer, ROADMs or Dynamic Gain Equalizers, DGE, that are Pl(k) (i.e. a start of the OMS) and P2(k) (i.e. an end of the OMS), a signal-to-noise ratio, SNR of established services that can be extracted from pre forward error correction, forward error correction (FEC) and bit error rate, and bit error rate, BER. The method provides improved accuracy in estimating the signal-to-noise ra tio, SNR within 0.2 decibel, dB, down from more than ldB.

The method enables the improved estimation of the lumped losses even if the power var ies in the optical network within a reasonable range for network operation. The method estimates the lumped losses more accurately even in the presence of filtering or other uncertainty on other key physical parameters. The method enables separation of the opti cal amplifier and nonlinear noises, which can be used for health analytics, especially for determining where a loss of optical performance occurs in the optical network. Hence, the method can be used as an optical performance monitor, OPM, or a software OPM. The method can run on a standard computer, thereby eliminating a need for dedicated computing hardware for monitoring the optical performance of the optical network. The outputs of the method may be leveraged by algorithms (e.g. QoT estimation algorithm, etc.)

Optionally, the method further includes estimating a signal-to-noise ratio, SNR, for each service carried by a channel through the OMS, based on the estimated lump loss ratio for each span in the OMS. The method further includes calculating an SNR distance between the estimated SNR and a monitored SNR for each service carried by a channel through the OMS. The method further includes outputting estimated values for the lumped losses before the fibre and lumped losses after the fibre for each span, if a value of the total SNR distance for all channels satisfies a predefined condition. The predefined condition may include one or more of the value falling below a threshold value, and the value increasing after previously decreasing, or the value increasing for a predefined number of iterations.

Optionally, the method further includes requesting a set of parameters the physical layer database, including one or more of an average gain, tilt or noise figure value for an am plifier and/or a linear attenuation, chromatic dispersion, nonlinear parameter or length of a fibre and/or a type, bandwidth or filtering response or attenuation of a filter and/or a routing or a spectrum allocation of services and/or a back-to-back noise response of a transponder and/or a polarisation dependent loss parameter of any element. An initial value of the estimated lump loss ratio for at least one span may be based on arbitrary initialisation values for the lumped losses before the fibre and lumped losses after the fibre for each span. Optionally, the optimisation function is a gradient descent function. FIG. 8 is an illustration of a computer system (e.g. an optical performance monitoring device) in which the various architectures and functionalities of the various previous im plementations may be implemented. As shown, the computer system 800 includes at least one processor 804 that is connected to a bus 802, wherein the computer system 800 may be implemented using any suitable protocol, such as PCI (Peripheral Component Inter connect), PCI-Express, AGP (Accelerated Graphics Port), Hyper Transport, or any other bus or point-to-point communication protocol (s). The computer system 800 also includes a memory 806.

Control logic (software) and data are stored in the memory 806 which may take a form of random-access memory (RAM). In the disclosure, a single semiconductor platform may refer to a sole unitary semiconductor-based integrated circuit or chip. It should be noted that the term single semiconductor platform may also refer to multi-chip modules with increased connectivity which simulate on-chip modules with increased connectivity which simulate on-chip operation, and make substantial improvements over utilizing a conventional central processing unit (CPU) and bus implementation. Of course, the vari ous modules may also be situated separately or in various combinations of semiconductor platforms per the desires of the user.

The computer system 800 may also include a secondary storage 810. The secondary stor age 810 includes, for example, a hard disk drive and a removable storage drive, repre senting a floppy disk drive, a magnetic tape drive, a compact disk drive, digital versatile disk (DVD) drive, recording device, universal serial bus (USB) flash memory. The re movable storage drive at least one of reads from and writes to a removable storage unit in a well-known manner.

Computer programs, or computer control logic algorithms, may be stored in at least one of the memory 806 and the secondary storage 810. Such computer programs, when exe cuted, enable the computer system 800 to perform various functions as described in the foregoing. The memory 806, the secondary storage 810, and any other storage are possi ble examples of computer-readable media.

In an implementation, the architectures and functionalities depicted in the various previ ous figures may be implemented in the context of the processor 804, a graphics processor coupled to a communication interface 812, an integrated circuit (not shown) that is capa ble of at least a portion of the capabilities of both the processor 804 and a graphics pro cessor, a chipset (namely, a group of integrated circuits designed to work and sold as a unit for performing related functions, and so forth).

Furthermore, the architectures and functionalities depicted in the various previous-de scribed figures may be implemented in a context of a general computer system, a circuit board system, a game console system dedicated for entertainment purposes, an applica tion-specific system. For example, the computer system 800 may take the form of a desk top computer, a laptop computer, a server, a workstation, a game console, an embedded system.

Furthermore, the computer system 800 may take the form of various other devices in cluding, but not limited to a personal digital assistant (PDA) device, a mobile phone de vice, a smart phone, a television, and so forth. Additionally, although not shown, the computer system 800 may be coupled to a network (for example, a telecommunications network, a local area network (LAN), a wireless network, a wide area network (WAN) such as the Internet, a peer-to-peer network, a cable network, or the like) for communi cation purposes through an I/O interface 808.

It should be understood that the arrangement of components illustrated in the figures de scribed are exemplary and that other arrangement may be possible. It should also be un derstood that the various system components (and means) defined by the claims, de scribed below, and illustrated in the various block diagrams represent components in some systems configured according to the subject matter disclosed herein. For example, one or more of these system components (and means) may be realized, in whole or in part, by at least some of the components illustrated in the arrangements illustrated in the described figures.

In addition, while at least one of these components are implemented at least partially as an electronic hardware component, and therefore constitutes a machine, the other com ponents may be implemented in software that when included in an execution environment constitutes a machine, hardware, or a combination of software and hardware. Although the disclosure and its advantages have been described in detail, it should be understood that various changes, substitutions, and alterations can be made herein with out departing from the spirit and scope of the disclosure as defined by the appended claims.

Claims

1. A computer-implemented method of optical performance monitoring in an optical network (200, 300), wherein the optical network (200, 300) comprises at least one optical multiplex section, OMS (202, 308), formed from one or more spans (310), wherein each span (310) includes a fibre (312) and an optical amplifier (314), the method comprising: estimating an output power of one or more channels at the output of an optical amplifier (314) of a span (310) in an OMS (202, 308), based on an estimated lump loss ratio for each span (310) in the OMS (202, 308), wherein the lumped loss ratio is a ratio of lumped losses before the fibre (312) and lumped losses after the fibre (312) in the span (310); calculating a power distance between the estimated output power and a monitored output power for each channel through the OMS (202, 308); and updating estimated values for the lumped losses before the fibre (312) and lumped losses after the fibre (312) for each span (310) in the OMS (202, 308) using an optimisa tion function to minimise a total power distance for all channels.

2. The method of claim 1 further comprising: estimating a signal -to-noise ratio, SNR, for each service (316) carried by a chan nel through the OMS (202, 308), based on the estimated lump loss ratio for each span (310) in the OMS (202, 308); calculating an SNR distance between the estimated SNR and a monitored SNR for each service (316) carried by a channel through the OMS (202, 308); and outputting estimated values for the lumped losses before the fibre (312) and lumped losses after the fibre (312) for each span (310), if a value of the total SNR distance for all channels satisfies a predefined condition.

3. The method of claim 2, wherein the predefined condition is one or more of the value falling below a threshold value, the value increasing after previously decreasing, or the value increasing for a predefined number of iterations.

4. The method of claim 2 or claim 3, further comprising requesting a set of moni tored inputs from a physical layer database (318), including the monitored output power for each channel through the OMS (202, 308), and the monitored SNR for each channel through the OMS (202, 308).

5. The method of claim 4, further comprising storing the output estimated values in the physical layer database (318).

6. The method of claim 4 or claim 5, further comprising requesting, from the physi cal layer database (318), a set of parameters including one or more of an average gain, tilt or noise figure value for an amplifier and/or a linear attenuation, chromatic dispersion, nonlinear parameter or length of a fibre (312) and/or a type, bandwidth or filtering re sponse or attenuation of a filter and/or a routing or a spectrum allocation of services and/or a back-to-back noise response of a transponder and/or a polarisation dependent loss parameter of any element.

7. The method of any preceding claim, wherein an initial value of the estimated lump loss ratio for at least one span (310) is based on arbitrary initialisation values for the lumped losses before the fibre (312) and lumped losses after the fibre (312) for each span (310).

8. The method of any preceding claim, wherein the optimisation function is a gradi ent descent function.

9. An optical performance monitoring device (100, 204) for an optical network (200, 300), wherein the optical network (200, 300) comprises at least one optical multiplex section, OMS (202, 308), formed from one or more spans (310), wherein each span (310) includes a fibre (312) and an optical amplifier (314), the optical performance monitoring device (100, 204) comprising one or more processors (102A-N) configured to: estimate an output power of one or more channels at an output of an optical am plifier (314) of a span (310) in an OMS (202, 308), based on an estimated lump loss ratio for each span (310) in the OMS (202, 308), wherein the lumped loss ratio is a ratio of lumped losses before the fibre (312) and lumped losses after the fibre (312) in the span (310); calculate a power distance between the estimated output power and a monitored output power for each channel; and update estimated values for the lumped losses before the fibre (312) and lumped losses after the fibre (312) for each span (310) in the OMS (202, 308) using an optimisa- tion function to minimise a total power distance for all channels.

10. The optical performance monitoring device (100, 204) of claim 9, further config ured to: estimate a signal -to-noise ratio, SNR, for each service (316) carried by a channel through the OMS (202, 308), based on the estimated lump loss ratio for each span (310) in the OMS (202, 308); calculate an SNR distance between the estimated SNR and a monitored SNR for each service (316) carried by a channel through the OMS (202, 308); and output estimated values for the lumped losses before the fibre (312) and lumped losses after the fibre (312) for each span (310), if a value of the total SNR distance for all channels satisfies a predefined condition.

11. The optical performance monitoring device (100, 204) of claim 10, wherein the predefined condition is one or more of the value falling below a threshold value, the value increasing after previously decreasing, or the value increasing for a predefined number of iterations.

12. The optical performance monitoring device (100, 204) of claim 10 or claim 11, further configured to request a set of monitored inputs from a physical layer database (318), including the monitored output power for each channel through the OMS (202, 308), and the monitored SNR for each channel through the OMS (202, 308).

13. The optical performance monitoring device (100, 204) of claim 12, further con- figured to store the output estimated values in the physical layer database (318).

14. The optical performance monitoring device (100, 204) of claim 12 or claim 13, further configured to request, from the physical layer database (318), a set of parameters including one or more of an average gain, tilt or noise figure value for an amplifier and/or a linear attenuation, chromatic dispersion, nonlinear parameter or length of a fibre (312) and/or a type, bandwidth or filtering response or attenuation of a filter and/or a routing or a spectrum allocation of services and/or a back-to-back noise response of a transponder and/or a polarisation dependent loss parameter of any element.

15. The optical performance monitoring device (100, 204) of any one of claims 9 to

14, wherein an initial value of the estimated lump loss ratio for at least one span (310) is based on arbitrary initialisation values for the lumped losses before the fibre (312) and lumped losses after the fibre (312) for each span (310).

16. The optical performance monitoring device (100, 204) of any one of claims 9 to

15, wherein the optimisation function is a gradient descent function.

17. An optical network (200, 300) comprising: at least one optical multiplex section, OMS (202, 308), formed from one or more spans (310), wherein each span (310) includes a fibre (312) and an optical amplifier (314), and the optical performance monitoring device (100, 204) of any one of claims 9 to

16.