WO2011070404A1

WO2011070404A1 - Optical system and method for monitoring the physical structure of optical networks, based on otdr with remote detectors

Info

Publication number: WO2011070404A1
Application number: PCT/IB2009/055763
Authority: WO
Inventors: Liliana Rocha Nicolau Lopes Da Costa; António Luís JESUS TEIXEIRA
Original assignee: Universidade De Aveiro
Priority date: 2009-12-11
Filing date: 2009-12-15
Publication date: 2011-06-16
Also published as: PT104880A

Abstract

The present invention relates to an optical system and method for monitoring the physical structure of optical networks, based on an optical time domain reflectometer (OTDR (2)) with remote detectors (600), (610), (611), (612), (613), (620), (621) and (622). The current use of the OTDR technique with remote detectors for detecting, identifying and locating physical faults in optical networks can be complex and expensive, such as when applied in networks having a tree topology with multiple hierarchy levels. The proposed method makes it possible to implement a simple and centralized monitoring system, and at the same time saves spectral bandwidth, is easy to integrate in any optical access network topology, eliminating the limitations usually associated with other conventional reflectometric techniques, reduces measurement and fault detection times and the expenditures associated with the implementation of an OTDR monitoring system in the physical layer of fiber optical networks.

Description

Title of Invention: OPTICAL SYSTEM AND METHOD FOR MONITORING PHYSICAL NETWORK STRUCTURE

OPTICS, RODT-BASED WITH REMOTE SENSORS

Technical field of the invention

[1] This paper presents a time domain optical reflectometer (RODT) developed for application in mapping and characterization of the physical structure of optical access networks. This monitoring system makes it possible to discover physical faults, verify physical connectivity and characterize the state of any optical path by placing remote optical sensors at critical points in the network. Therefore, the optical system allows to monitor and characterize an optical path in a sectioned manner and / or from its starting point to its ending point.

[2] Remote sensors create reference points, measurable as univocal events in the reflectometer trace, distinct in both the time domain and the frequency domain (or wavelength domain). These remote sensors also create a unique signature for each optical path of the network when illuminated by different wavelength monitoring channels, enabling remote identification and characterization of individual fiber sections of that path. In order to read this signature, the RODT light source emits more than one wavelength or consists of a set of lasers operating at different frequencies. These remote sensors are preferably based on Bragg optical fiber diffraction grids because they have low insertion losses and are low cost optical components. The Bragg networks of each remote sensor react to each monitoring channel and may reflect it (fully or partially) or transmit it. The characteristics of the RODT and remote optics sensors proposed herein are an innovation in the state of the art. In addition, remote sensors can prevent receivers on networked devices from receiving light from the monitoring system, preventing malfunctions or damage to these devices due to the presence of the monitoring signal.

[3] Because remote sensors are passive optical components, they do not require power or maintenance, making them ideal for application to optical access networks. As the only active element of the optical system is the RODT and since it operates from a single central point of the network, the optical system can be operated manually or automatically and allows integration into the upper tier management system. network for the purposes of protection and maintenance of long-term network infrastructure.

[4] In addition, the insertion of these wavelength-sensitive remote sensors into key network points also allows to delimit and differentiate network areas, which makes it possible to implement a physical network infrastructure shared by multiple operators. , each of which is assigned a particular range of wavelengths to monitor their part of the network.

[5] The monitoring strategy based on the optical system and the method of

The monitoring presented here, allows to map any fiber optic network topology (tree, multilevel tree, ring or hybrid), as long as there is access to the physical paths to integrate the remote sensors.

[6] The advantages of the proposed optical monitoring method are associated with its innovation, notably as regards spectral bandwidth savings for monitoring purposes, easy integration into any optical access network topology, the elimination of limitations normally associated with other conventional reflectometry techniques, the reduction in measurement time and failure detection time and, in essence, the significant reduction in costs associated with the implementation of an optical RODT monitoring system in the infrastructure. physical structure of fiber optic networks.

[7] Although the optical system and its monitoring method which is the invention have been developed for the supervision, certification and maintenance of fiber optic access communication networks, it is applicable in other areas where remote monitoring of the physical state of infrastructure is required. - structures such as buildings and bridges, provided that the means used to propagate the monitoring signal is a fiber optic waveguide.

Description of the prior art

[8] References cited:

US 5771250 - 06/1998 -Shigehara et al.

US 5907417 - 05/1999 - Darcie et al.

US 6009220 - 12/1999 - Chan et al.

US 6512610 - 01/2003 - Minami et al.

EP 1578038 - 09/2005 - Yuksel et al.

US 7206478 - 04/2007 - Yeh

EP 1772979 - 04/2007 - Vandewege et al.

US 0062408 - 03/2008 - Lai et al.

US 6771358 - 08/2008 - Shigehara et al.

With the emergence of passive access optical networks as a solution to address the limitations associated with reduced bandwidth and the growing number of users resulting from the provision of new voice and data services by operators, There was also a need to ensure quality of service, safety, robustness and agility in preventing and repairing failures. In the current context, and for future generation access networks, it is imperative that the operator be able to see in real time the physical state of the network and the conditions of data transmission. Implementing a monitoring system that can attest to the good state of the physical network as a whole and timely detect and locate a critical failure that could cause service disruption is an important factor in the success of such networks. Currently, the most commonly used technique for performing this function is temporal domain optical reflectometry (RODT). Because it can be used in a non-invasive manner, occupying a wavelength, or spectral band, different from that used for traffic transmission, because it is a simple technique for information processing, it is a solution. RODT has been the most popular choice in the literature to investigate the implementation of physical fault monitoring in optical access networks.

] Although technology has already reached maturity and commercial success, its

The application for real-time monitoring in the current context of optical access networks is still a field to explore. In addition, an optical reflectometry monitoring system also finds application in other life stages of an optical network, beyond supervision, such as certification of the initial installation of the fiber optic infrastructure and the optical components that make up the network (such as power splitters, for example), and checking for repair or replacement operations of fiber sections or network components. Currently, a team of specialist technicians use an RODT manually to certify network installation and for maintenance and / or repair functions. The latter usually entails high operating costs in terms of time, equipment, human resources and eventually partial or total interruption of service. This situation is aggravated by the large number of optical paths to be verified due to the high number of users and / or the long range or geographical coverage which makes monitoring and certifying the physical structure of the network with an RODT a very complex task and subject to misinterpretation.

Of the many possible methods of optical reflectometry, the conventional direct detection RODT technique requires the least initial investment. A conventional direct detection RODT acts as an 'optical radar', ie it sends a short pulse from one end of the fiber optic and detects the reflected pulse power evolution over the travel time to the other end. of the fiber and back to the RODT. The reflected energy is produced by the Rayleigh scattering phenomenon and, after being detected directly with a photodiode, presents an exponential decrease in relation to the travel time of the pulse. The travel time of the pulse in the optical fiber is proportional to its length, and therefore it is possible to obtain an RODT trace showing the reduction of the reflected pulse power relative to the fiber length. Normally multiple identical pulses are sent at a maximum repetition rate equal to twice the pulse travel time through the fiber to avoid pulse overlap and lead to errors in RODT measurements. Since the reflected optical power level is usually low, it is necessary to increase the signal-to-noise ratio of the monitoring signal received by the RODT in order to achieve a greater dynamic range of measurement. This is why multiple pulses are sent: averaging the reflected power values with multiple pulses eliminates much of the noise from the measurement. However, processing information acquired with multiple pulses implies high measurement times, typically in the order of tens of seconds to tens of minutes.

[12] The presence of fiber discontinuities is also detected by RODT. These are usually junctions between fiber and / or optical components, where the refractive index varies, creating so-called Fresnel reflections. These discontinuities, visible in the trace obtained, can be of two types: reflective events and non-reflective events.

[13] Reflective events are identified as peaks located along the Rayleigh scatter trace and are usually caused by fiber cuts, connectors, mechanical splices or unfinished fiber. Non-reflective events appear as abrupt power drops located in the rayleigh scatter trace and are usually caused by fusion splicing or bending in the fiber.

[14] The use of a conventional direct detection RODT technique is a compromise between the initial cost of the monitoring system and the time cost for acquiring, processing and interpreting the measurements obtained with the reflectometer. Its major limitation is the compromise between the spatial resolution of the RODT, which is inversely proportional to the temporal width of the pulse, and the maximum distance.

measurable, which is proportional to the dynamic range of measurement of the RODT. The lower the resolution, the higher the energy of the transmitted pulses, so we can monitor longer optical paths, or paths subject to higher levels of insertion losses. However, using a lower resolution implies detecting fewer events, and consequently, losing detail about the optical path characterization information. Potentially for example failures in detecting an event as important as a cut of a fiber. For example, imagining a situation where two reflective events are located at a distance of the same order or less than the RODT resolution makes it impossible to distinguish one event of the other, since the power reflected by one of them masks the power reflected by the other. When considering point-to-multipoint networks, this problem becomes even more difficult to solve as monitoring signals originating in parallel optical paths overlap in the RODT trace.

[15] A time domain optical reflectometer can be implemented in a number of ways: in terms of the light source used, the pulse generation method, the type of detection used and the fiber interrogation method. Modifications to the conventional direct detection RODT method are intended to improve one or more of the following: increasing dynamic range, increasing maximum measurable fiber distance, resolving the compromise between resolution and dynamic range, using resources / network components simultaneously for traffic monitoring and transmission and facilitate the adoption of the monitoring strategy regardless of the type of fiber optic network. In this sense, some methods provide solutions that adapt to the characteristics of the fiber optic network itself and the limitations imposed by data transmission simultaneously with RODT monitoring, the presence of active optical components along the optical paths and the topological complexity of the optical fiber network itself. network. The following documents describe some of these methods.

[16] US Patent 5771250 describes a conventional RODT monitoring system for peer-to-peer networks, where Bragg networks are placed at optical path terminations with users. These networks serve two purposes: first, to filter the monitoring signal so that it does not hinder the reception of information by users and, secondly, to create reflective events to demarcate the end of each optical path. In the present invention, the insertion of remote sensors aims to create distinct signatures by optical path, allowing them to be differentiated from each other, if the network has a point-to-point or point-multipoint structure, and produce traces of RODT overlaid. In addition each reflective element referred to in the document is comprised of a single Bragg diffraction grating which reacts to a single wavelength or channel.

monitoring, has the highest reflectivity possible, and is placed only at the end of each optical path. The optical system presented herein also differs from the present invention in that the reflective elements are placed only at the network termination points.

[17] US 5907417 describes an RODT based monitoring system used in conjunction with an optical spectrum analyzer, while the optical system of the present invention is based on an RODT and reflective elements. The use of the technique described in the document implies, besides the reflective elements, the presence Optical multiplexer / demultiplexer on remote nodes of the network contrary to the proposed invention which only provides for the insertion of remote sensors and does not require the presence of other optical components that influence or determine the technology used for data signal transmission. The document also differs from the present invention in that instead of using remote sensors as reference points at intermediate points and at network terminations with users, a set of couplers and a waveguide are used to make a mesh. feedback (additional fiber optic connection) so that the monitoring signal is reflected back to the RODT. The optical system proposed in this invention does not provide for the use of a return mesh.

[18] US 6009220 describes an RODT method that avoids the use of a dedicated light source solely for monitoring purposes. High bandwidth noise from amplifiers placed at midpoints of the optical network is the light source of the RODT and the receiver is a spectral analyzer. Reflective elements with different wavelengths (Bragg grids) and reduced spectral width are used when compared to the RODT light source. While the present invention provides for the installation of remote sensors only, the reflective elements herein are used as an integral part of an optical system further comprising an optical amplifier, a circulator, an optical receiver and an additional fiber optic path for directing, detect and carry the monitoring signal reflected by the reflective elements (Bragg networks) back to where the monitor itself is located. The technique proposed in the document also differs from the invention in that the monitor used herein is not an RODT, but an optical spectrum analyzer; the reflective elements are comprised of a single Bragg network and the need to allocate a wavelength or monitoring channel for each reflective element used in the network; instead of using a dedicated light source located at a central point in the network, it takes advantage of the noise generated by erbium doped fiber amplifiers to act on the reflective elements; and further by detecting the monitoring signal generated in the amplifier and reflected by the Bragg networks is detected locally next to each amplifier, unlike the present invention where the signal processed by the remote sensors is generated, detected and processed remotely in the RODT

[19] US 6512610 describes an RODT-based monitoring system and reflective elements (Bragg networks) for application to a tree-structure multipoint point network (the topology most commonly used in passive optical access networks). This document differs from the present invention in that it is based on calculating the attenuation produced by the fiber distance between two comparisons with data obtained before failures for a single wavelength or monitoring channel, the reflective elements being made up of a single Bragg network. It therefore differs from the present invention, whose method is based on analyzing and comparing RODT traces for more than one distinct wavelength or monitoring channel, and whose remote sensors may consist of more than one Bragg network. and may transmit or reflect (wholly or partially) different monitoring channels. Furthermore, in the proposed invention the analysis of monitoring signals that are processed by remote sensors for the purpose of detecting and identifying physical faults is performed in both time and frequency domains, unlike the technique described in the document where it is detected. analysis is performed only in the time domain.

[20] EP 1578038 describes a method and device for performing RODT in real-time optical networks. The reflectometer is placed in the central office of the optical access network. This method was developed for tree topology networks. While the optical system of the present invention is comprised of fixed and passive remote sensors located at network midpoints and terminals, switchable reflective elements (active, requiring an electronic control / actuation system) are placed at each network termination. with users. Each switchable reflective element reflects or not light at a single wavelength (depending on the control signal it receives and processes), unlike the remote sensors considered in the present invention.

[21] US 7206478 describes a method of monitoring physical failures in an optical network based on frequency spectrum analysis. As such, it differs from the present invention in that monitoring signal reception is based on an optical spectrum analyzer rather than an RODT. The monitoring band signals are allocated in a different band than that used for traffic transmission, as in most similar methods. The light source used is local and is a semiconductor optical amplifier, unlike the optical system proposed here, where the generation, reception and processing of the monitoring signal is done by the RODT (located at a central point of the network) and by inserting remote sensors. As in some of the previous methods, Bragg networks with different wavelengths are placed at the terminations with the users and act as reflective elements of the monitoring signal. The presence or absence of the reflections produced by these reflective elements allows the detection of physical faults. In the proposed invention the analysis of monitoring signals that are processed by remote sensors is performed in both time and frequency domains, unlike the technique described in the document where such analysis is performed only in the frequency domain. As such, the latter can only detect the occurrence of failures, unlike the invention which allows to detect, identify and locate critical faults.

[22] EP 1772979 describes an RODT-based monitoring system embedded in optical termination hardware with users. Its main advantage is the reuse of optical components that already exist in the network (transmitters and receivers), unlike the present invention where RODT hardware is totally independent of the hardware used for transmitting / receiving data signals. In this document, the wavelengths or channels of

The monitoring systems are the same as those allocated for traffic transmission, unlike the optical system of the invention which operates on a completely different spectral band than that allocated for data transmission. Thus, the technique described in the document entails a significant decrease in the temporal freedom to perform monitoring operations, a restriction not existing in the present invention. Another aspect that differentiates the invention is that the former does not aim at the insertion of remote sensors as an integral part of the optical monitoring system.

[23] US 0062408 describes an RODT method for tree structure passive access optical networks based on the use of an RODT in conjunction with an optical spectrum analyzer (or alternatively a demultiplexer, a switch and a photodetector). The optical system of the invention described herein does not include an optical spectrum analyzer. Reflective elements with different wavelengths, together with bypass elements, are placed at the optical terminations with the users. As with other previously described methods, the presence of reflected signals indicates the presence of network failures. RODT is used to determine the exact location of the fault. While in the invention the optical system consisting of the RODT and remote sensors is used to simultaneously detect, identify and locate physical faults, in the document the RODT is used for fault detection and the optical spectrum analyzer (or similar equipment). ) is used to identify and locate in which branch of the tree-structured network the fault occurred.

[24] US 6771358 describes an RODT method for tree-structured optical networks. Reflective elements based on optical filters are again used as identifiers of the different fiber optic branches. In this method, the reflective elements are placed just after the star coupler at the beginning of each branch. It is by calculating the longitudinal distribution of the monitoring signal attenuation in each fiber optic branch that fault detection is performed. The technique described in the document provides for the use of reflective elements, each of which is allocated a distinct wavelength or monitoring channel. Thus, the invention differs from the document in that each remote sensor may react to more than one wavelength or monitoring channel, and that sets of two or more remote sensors may react to the same wavelength as the light source of the RODT

Detailed Description of the Invention

[25] Following an in-depth study of the state of the art in relation to techniques of

Monitoring for optical access networks, based on RODT measurements and the addition of reflective elements to the network structure, some limitations were found regarding their concrete implementation in the current context of optical access networks. The prior art of RODT is considered to be based on the addition of reflective elements at certain points of the fiber optic network. Each of these reflective elements has a single central reflection wavelength and reacts to a single wavelength or monitoring channel. The set of reflective elements may be operated using a single monitoring channel or multiple monitoring channels. These reflective elements are, in most cases, fiber-optically engraved Bragg diffraction grids. The RODT can use either a wavelength tunable light source or a light source with a very wide spectral width. By allocating one and only one distinct wavelength to each Bragg network (and each reflective element), a high spectral band must be allocated for monitoring, and therefore the first solution involves the use of a high cost light source. . The second solution involves the existence of wavelength demultiplexing / multiplexing elements or optical amplifiers at midpoints of the network to be monitored, which restricts the field of application of the prior art RODT.

[26] In the prior art, the monitoring strategy is based on the identification of

reflective events, caused by Bragg networks, on the RODT trace obtained or as measured with a spectrum analyzer for the various wavelengths (simultaneously or successively) as a means of verifying connectivity and good physical state between two diametrically opposed points of the network that define an optical path. Simply put, if a given reflective event is not present in the Rayleigh scatter reflected optical spectrum and originated by the Fresnel reflections, or if the RODT trace shows attenuation variations compared to previous dashes, then a physical fault exists. on the optical path between the RODT and the Bragg network placed at or at the end of that path. It is assumed that there is only one reflective element (one Bragg network) per optical path. Physical failure can then be localized, as the technique analyzes the power reflected by Rayleigh scattering in the time domain, and the pulse travel time is converted to distance. [27] When considering a high number of end users and / or long geographical reach, the high number of network branches is an important limiting factor for the application of this type of RODT techniques. It is therefore imperative to achieve a spectrally efficient monitoring method occupying the smallest possible spectral band. For example, its application in a tree structure network with multiple levels of hierarchy becomes complex, making it difficult to uniquely identify the fiber section where the fault occurred and its location. The complexity of a network topology translates into a high number of fiber sections, terminations and / or hierarchy levels. The greater this complexity, the greater the costs inherent in using the prior art RODT: it implies a high spectral band occupied for monitoring purposes, which could be used for traffic transmission; consequently, it also implies a wider wavelength tuning range for the RODT light source, which makes it even more expensive. In addition, a method that can be applied to both tree networks, ring networks, and hybrid topologies is desirable.

[28] For reasons of compatibility between current and future networks, or even to facilitate network upgrade operations, an RODT method that does not imply any change in network architecture and does not depend on the components that make it up is the most appropriate. For example, in some methods a spectral analyzer is also used to detect the monitoring signal (in conjunction or not with an RODT) by differentiating it into wavelength in order to identify the reflections produced by different reflecting elements. The RODT and / or spectral analyzer is typically located in the network central office, acting as a centralized monitoring subsystem. Although in some of the prior art documents presented, the detection and further conversion of the optical to electrical domain of the monitoring signal is performed locally (near the reflective elements) and not remotely from the central office of the network. Even in some of these documents, even using reflective elements, it is necessary to install return loops or fiber optic paths dedicated only to transport the monitoring signal to the place where it is processed (through an RODT, a or both). The invention avoids the costly installation of these optical paths dedicated solely to network monitoring as it operates directly over the fiber infrastructure used for the transport of paid cargo traffic. In addition, the application of the invention does not imply the presence of network components such as optical amplifiers or wavelength multiplexers / demultiplexers, nor does it influence traffic transmission.

[29] However, the most important criterion for the implementation of a Monitoring in access networks, with technological and commercial viability, is undoubtedly the cost, which translates into several levels: initial investment in the installation and cost of the optical monitoring system itself, busy spectrum band, time required and ability to monitor network-wide, consequences for simultaneous traffic transmission, and number / type of undetectable physical failures.

[30] The RODT method that is the subject of this invention provides a solution to these problems. The optical system is based on placing an RODT at a network center point for interrogation of the optical network, and for processing the information embedded in the reflected monitoring signal, translating it into optical power attenuation versus fiber distance values . This RODT has the ability to send light in more than one wavelength, each wavelength corresponding to a monitoring channel. Monitoring channel spacing is typically considered to be 1 nm or less for a tunable wavelength laser. Although this is the preferred approach, it is also possible and economically viable to use a set of fixed lasers (for medium range networks but not too many optical paths) or tunable lasers (for short or medium range networks with a very large number of optical paths). The choice of the central wavelength of the monitoring channels in the context of optical access networks typically falls on long wavelengths, the attenuation of which due to their spread on optical fiber is reduced. Typically, in the case of optical access networks, the monitoring channels are allocated in the U band (1625nm to 1675nm), as it is expected that the L bands (1565nm to 1625nm), the C band (1535nm to 1565nm) and others of shorter wavelengths are already allocated for traffic transmission or other network functions.

[31] Remote sensors are added to all critical points in the network,

namely, all endpoints of a network hierarchy level, which is equivalent to adding one remote sensor per fiber section. Thus, the invention proposed herein offers a solution for optimizing the number of wavelengths or channels required to monitor all individual fiber sections of a network, as well as resolving ambiguity in the detection, identification and location of physical faults arising from physical faults. of using a conventional RODT. The remote sensors used, and their arrangement in the network to be monitored, differ from those of the prior art. The remote sensors used in the invention react to more than one wavelength which we will hereinafter refer to as the monitoring channel. This is achieved, for example, by using phase-shifted or multiple phase-shift Bragg networks or concatenation. of uniform Bragg nets. The use of this type of remote sensors allows significant savings of the spectral band for monitoring. Since networks with equal spectral characteristics can be reused to reference more than one key point of the network, if their location meets a given set of conditions, this method has even greater spectral efficiency. In addition, when fewer monitoring channels are used, less time is required to monitor all optical paths that make up the network.

[32] In addition to minimizing cost associated with measurement time and spectral band

Occupied for monitoring, this RODT method allows the use of a light source, namely a laser, of low cost, tunable wavelength (in temperature or current, for example) in a reduced range. A concrete example is a temperature tunable coarse DFB laser in a typical range of 2 to 10 nm. It is to be noted that expensive tunable lasers typically have tuning ranges typically in the order of 20nm to 100nm. The cost reduction of the RODT relative to its light source is further supported by the fact that a low power laser (typical average power with continuous signal up to 10 mW) can be used, and preferably a semiconductor laser, but not limited to it. technology, as the field of application of this RODT method is restricted to medium or short range fiber optic networks (typically hundreds of meters to many tens of kilometers). In addition, because remote sensors based on Bragg diffraction grids are used as spatial demarcation points, which react to more than one monitoring channel, the problem resulting from the compromise between RODT resolution and dynamic range is eliminated. , as it becomes possible to identify physical failures even when the RODT resolution is too high, causing reflective events to overlap. This means that high RODT resolution values can be used without impairing their fault detection performance, allowing you to enjoy a wider dynamic range and / or relax the requirements of the RODT photodetector sensitivity. The substantial reduction in the cost of the RODT system is a significant advance of the method proposed here compared to the methods described in the documents presented above and also present in the literature.

[33] Another distinguishing factor of the invention is that remote sensors do not operate solely as power reflection elements for a given input channel.

but also as transmission elements. Spectral information is used to create an amplitude code: each Bragg network can reflect (wholly or partially, depending on its location) or transmit a given monitoring channel. Each remote sensor has one or more Bragg networks. When making remote sensors with networks with multiple phase shifts, for example, it is possible to print an amplitude code for multiple monitoring channels with a single Bragg network. If uniform Bragg networks are used, the various networks with a spectral profile allocated to different monitoring channels must be concatenated, which can be achieved in one single component (a practically achievable recording process, slightly spaced from the diffraction grids on the same optical fiber).

[34] Looking at the RODT trace obtained with a given monitoring channel, we observed a signature, which can be analyzed in the time domain and the frequency domain, which is unique for each network path and each fiber sections that make it up. Failure is detected by observing this signature in the temporal domain using RODT, and then comparing at least one signature on a given monitoring channel with the table, which specifies the spectral profile of the remote sensors and which defines the type of events that remote sensors generate for each monitoring channel. This monitoring strategy makes it possible to detect physical failures faster as the remote sensors used reuse the same monitoring channels under certain conditions.

[35] In addition, the proposed method allows for example to implement a simple and fast system for identifying fiber sections or optical paths with multiple fiber sections (fiber sections may be common to more than one optical path), very useful in repair or replacement operations, especially when there are multiple physically bundled fiber sections on the same cable. Eventually, the RODT light source can operate in continuous mode, ie without pulse generation, to perform network monitoring even faster.

[36] The proposed RODT method can be applied to any type of optical access network topology and architecture regardless of its constituent components. Detecting and locating physical faults with the proposed RODT technique does not require the prior existence of RODT information obtained in the network installation phase: knowledge of the base network topology is sufficient. The measurements made are related only to this topological information, i.e. number and length of fiber sections and location of hierarchy levels.

[37] Since the spectral band allocated for the monitoring signal is different from

allocated for traffic transmission, when necessary, the coupling of the RODT to the network can be performed very simply on any type of network through an optical element that allows the monitoring signal generated by the RODT light source to be coupled and the signal to be decoupled. network and remote sensors. The coupling / uncoupling element is typically placed in the first fiber section of the network, right after the connections of all equipment and transmission / reception of traffic signals. The optical component capable of coupling / uncoupling the monitoring signal to the network may be an optical circulator, a WDM coupler or a power divider coupler, with a coupling ratio that is typically between 10% to 50%. If there is simultaneous transmission of traffic or other optical signals on the network with the monitoring signals, an optical component must also be added next to the RODT receiver that filters the monitoring signal so that it is not contaminated or masked. by the presence of other signals that propagate in the network, which also generate Rayleigh dispersion and Fresnel reflections. The filter should typically have a bandpass feature, which rejects only traffic signals, letting the monitoring signals pass, and can be performed on various technologies such as Bragg networks, 'add / drop', 'arrayed-' filters. waveguide gratings', among others. The RODT mating / uncoupling element and the optional optical filtering element next to the RODT receiver may eventually be a single optical component (such as a WDM coupler operating on at least distinct spectral bands).

[38] The limitations of the proposed method are mainly related to: (i) insertion losses of remote sensors; (ii) the spectral band available for monitoring purposes; (iii) the number and maximum length of fiber sections, which is mainly dependent on the characteristics of the RODT, namely its dynamic range of measurement, which in turn depends on the spatial resolution of the RODT and the energy of the transmitted pulses.

[39] It is understood that not all types of physical failures that may occur in a fiber optic network are covered by the proposed RODT monitoring method. Only the most critical faults that impair physical connectivity in optical paths such as fiber cuts, unconnected connectors or severe fiber bends. Minor increases in insertion losses in optical components can only be detected when the RODT trace quality allows for the interpretation of small variations in Rayleigh scatter reflected power, which may not occur in all cases. In order to carry out this type of measurements it is necessary to compare current RODT measurements with the RODT information obtained at the network installation stage. Although the optical system of the present invention allows the complete high resolution characterization of a given optical path (thus it can also be used in network certification when installing or repairing), the detection of these events, usually caused by scanning processes. aging of components or replacement of new fiber sections or components is performed as a secondary operation. The priority function of the optical system of the The invention is to supervise the physical state of a complex network to detect, identify and locate critical physical faults as quickly and accurately as possible. Another consideration is that fiber cuts usually result in non-reflective events (at most they may result in low reflectance reflective events) and are therefore difficult to detect. The maximum reflectance value caused by a critical failure determines the minimum reflectivity value for Bragg grids that make up the remote sensors of the optical system when they are placed at midpoints of the grid and / or are in a series configuration. When remote sensors are placed at extreme points (terminals) of the network optical paths and / or when remote sensors are in a parallel configuration, their reflectivity can and should be higher as long as it does not cause saturation of the RODT photodetector. .

[40] The monitoring strategy proposed here, and the advantages arising from the

implementation of the optical monitoring system become clearer and more noticeable in the following description of the configuration, where the monitoring method of the present invention is also presented.

Description of Drawings

[41] Figure 1 illustrates a tree-structured network with three levels of hierarchy to which the optical monitoring system of the invention is applied.

[42] Figure 2 illustrates the table with the spectral characteristics of Bragg networks constituting the codewords assigned to each of the eight remote sensors used in the optical system of the example in Figure 1.

[43] Figures 3 (a) and 3 (b) illustrate the time domain signature of the

monitoring for each monitoring channel, corresponding to the spectral characteristics of the Bragg networks constituting each remote sensor, used in the example of Figure 1 and described in Figure 2.

[44] Figure 4 illustrates the occurrence of physical failures in the network of Figure 1.

[45] Figures 5 (a) and 5 (b) illustrate the RODT traces obtained with the method of

monitoring proposed in the invention for each monitoring channel of the network illustrated in Figure 1 and used as an example. The dashed RODT traces correspond to the case of Figure 1, where there are no physical failures. For the case of Figure 4, where physical failures occur, the RODT traces are indicated with a continuous line.

[46] Proposed Configuration

[47] An example of an optical fiber network with three hierarchy levels is shown in figure 1, indicated in the figure as' Level Γ, 'Level 2' and 'Level 3'. The place of division between different levels of hierarchy is called a node. RODT (2) is coupled to the network through an optical coupler (3), which also filters the allocated band for the monitoring signal, isolating it from the band allocated for traffic transmission, preventing Rayleigh dispersion caused by the traffic signals to mask the

monitoring. Examples of optical couplers used for this purpose are: distinct spectral band couplers / decouplers called two wavelength division multiplexing (WDM) couplers, an optical circulator or a power splitter coupler. In this application, the RODT (2) is placed next to the central office (1) of the network.

[48] The RODT laser-generated monitoring signal (2) follows the same path as optical traffic signals. The following describes how the RODT (2) laser is operated if it is a tunable laser with a reduced tuning range. First, its emission wavelength is tuned to coincide with one of the channels or wavelengths allocated for monitoring. The laser is modulated directly with the short pulses and they are transmitted through the fiber section (500) at hierarchy level 1. Being operating in fault detection, identification and location mode (main mode of operation), the width of the Pulses typically range from 1 μ s to 20 μ s, although the RODT light source can also operate in continuous mode, ie without pulse generation. If you want to characterize an optical path or section of an optical path, you may need to use a higher RODT resolution, which means typically use a smaller pulse width than that used for detection, identification and troubleshooting. If the transmitted monitoring channel corresponds to one of the wavelengths reflected by the remote sensor (600), then part of the pulse power is reflected back to the RODT (2) and the other part goes to the power divider (40). ). If the transmitted monitoring channel corresponds to one of the wavelengths transmitted by the remote sensor (600), it acts as a transparent element, appearing in the acquired RODT trace as a non-reflective event and with a small associated power loss. At hierarchy level 2, the monitoring signal is divided and follows through four fiber sections (510), (511), (512) and (513) and reaches remote sensors (610), (611), (612) ) and (613), where a process similar to that of remote sensor (600) occurs, depending on the monitoring channel used at that time. It is assumed that in the network terminal equipment 700, 701, 702, 702, 710, 711, 712 there is optical filtering and therefore the monitoring signal is discarded so as not to hinder the transmission and reception of traffic from these network points. At hierarchy level 3, the signal from fiber section 513 is processed by remote sensor 613 and, depending on its wavelength, goes wholly or partially to power divider 41. Then the monitoring signal is transmitted by the monitoring sections. fiber (520), (521) and (522). Remote sensors 620, 621, and 622 process the monitoring signal in the same manner as at previous hierarchy levels.

[49] The temporal evolution of all reflected power propagating from right to left is detected due to Rayleigh scattering, Fresnel reflections and the presence of remote sensors, and is detected and processed in the RODT (2). This information is first analyzed in the time domain and, to facilitate the interpretation of the trace obtained, the time scale is converted to a spatial scale, with the prior knowledge of the value of the speed of light propagation in the fiber. In this way it is possible to quantify the distance to the RODT (2) at which reflective and non-reflective events caused by remote sensors occur. Upon completion of the measurement information processing process, the RDOT (2) laser is tuned to another monitoring channel and so on. After obtaining each trace of optical power versus distance for each monitoring channel, the measurements are interpreted in order to verify the good physical state of the network:

[50] (i) when one or more reflective events, caused by the passage of monitoring channels through remote sensors, are not present in one or more traces obtained, it means that one or more physical faults have occurred;

(ii) when one or more non-reflective events arising from the passage of monitoring channels through remote sensors present as reflective events in one or more of the traces obtained, one or more physical faults have occurred;

(iii) fault identification and location results from comparing the obtained RODT trace with a given monitoring channel, a table with the network optical path signatures for each wavelength, or with at least one trace trace. reference (trace obtained in the absence of faults), anterior to the same monitoring channel;

(iv) the optical path signature table has information about the distances of the remote sensors, the hierarchy levels and the spectral characteristics of the Bragg networks that make up each remote sensor, for each wavelength or channel. allocated for monitoring;

(v) Optical path signatures not only enable the detection, identification and location of critical physical faults, but also identify the presence of fiber optic sections and optical components.

[51] Figure 2 illustrates a table showing the spectral characteristics of

Bragg which constitute the remote sensors used in the example of Figure 1. Each Bragg network signals information on 2 monitoring channels. Each of them is associated with a central wavelength, λΐ and λ2, and for each remote sensor, is assigned a codeword. This code consists of two elements: T means that the monitoring signal is transmitted and R means that the monitoring signal is reflected. In the case of remote sensors 600 and 613, the monitoring channels are partially reflected as the sensors are located at intermediate points of the network. The reflectivity value of Bragg networks, remote sensors located at intermediate points of the network, depends on their distance to the RODT (2) and the insertion losses in this optical path. The Bragg networks constituting the remaining remote sensors 610, 611, 612, 620, 621 and 622 have higher reflectivities as they are at network termination points. Mathematically, the maximum reflectivity of the Bragg networks, situated at intermediate points, is given by the following mathematical expressions for series configuration and parallel configuration cases, respectively:

[52] [Chem. l]

= _s IT ^

[53] [Chem.2]

[54] where α is the attenuation of the fiber, L _; is the length of the ith fiber section, R is the reflectivity of Bragg networks, β _ΚΝ is the value of the insertion losses of the optical components present in each node, N is the number of fiber sections and P _D is the minimum optical power detectable by the RODT receiver. P _R is the optical power reflected by Rayleigh scattering and is expressed by:

[55] [Chem.3]

P = a: _s .Sv _g WP _iK

[56] where W is the time width of the pulses transmitted by the RODT, a _s is the Rayleigh scattering attenuation coefficient, S is the Rayleigh recapture coefficient, v _g is the group velocity of light and P _in is a peak optical power of the pulses transmitted by the RODT.

[57] While the monitoring signal is processed by remote sensor Bragg networks, Rayleigh dispersion, caused by the fiber sections of the network, produces a residual signal from the monitoring signal, which propagates in all the directions. Part of this residual signal propagates from right to left and is also detected at the RODT receiver (2). The presence of optical components and transitions between fiber sections of the network may also give rise to another phenomenon of monitoring signal reflection: the so-called Fresnel reflections. In order that Fresnel reflections do not mask the reflections produced by Bragg networks of remote sensors, the latter must have sufficient reflectivity to exceed the maximum reflectance value produced by Fresnel reflections. Therefore, the minimum reflectivity of the Bragg (600) and (613) networks must be greater than the maximum power value caused by a reflective event, originated by Fresnel reflection, situated at the same location as a given Bragg network. .

[58] The two-dimensional code (time and frequency), assigned to the Bragg networks of each remote sensor, makes it possible to uniquely identify a given optical path and the fiber sections that make it up, as different signatures are assigned to each channel. monitoring This code is assigned regardless of the network topology, ie the same code can be applied to ring networks (where remote sensors are in series), tree networks (where Bragg networks are in parallel) or to hybrid or multi-level networks (where some remote sensors are in series and others in parallel). The case of the network illustrated in Figure 1 corresponds to a multi-level network.

[59] However, there are some rules for constructing the code to assign to each remote sensor:

(i) at the termination of each light path a remote sensor is placed;

(ii) an optical path is defined as a fiber optic link between the first hierarchy level and the remaining hierarchy levels and consists of one or more fiber sections in series;

(iii) For each pair of optical paths whose length (in case of serial optical paths) or the length difference (in case of parallel optical paths) is less than the spatial resolution of the RODT, it is necessary to allocate at least one length. distinct wave Thus, the total number of monitoring channels M is equal to the total number of optical path pairs that meet this condition;

(iv) the value of RODT Res spatial resolution is given by:

[Chem.4]

2 ^and

where vg is the group velocity of light, W is the time width of the pulses transmitted by the RODT, and t,. is the response time of the RODT photoreceptor. The temporal width of the pulses depends on the maximum distance to be measured with the RODT;

(v) in the extreme case, where all optical paths have the same length (serial configuration) and / or if the length difference between any pair of optical paths (parallel configuration) is less than the RODT spatial resolution, so M equals the total number of optical paths;

(vi) with M monitoring channels, we can get a total of Q = 2M - 1 codewords, composed of 'transmitted' and 'reflected'; The only word that cannot be used is that which implies the transmission of all M monitoring channels;

(vii) Of these Q combinations, there are M combinations that have a single monitoring channel that is reflected. These specific M combinations must be chosen for the N optical paths that satisfy condition (iii).

[60] In the example illustrated in Figure 1, we have a total of 8 optical paths, of which there are 2 pairs of fiber sections that meet condition (iii) - the pair of fiber sections (610) and (611) and the pair of fiber sections 620 and 621. Therefore we can allocate a minimum of M = 2 channels for monitoring.

[61] The complementary codewords, R 'and' RT, must be assigned to the remote sensor pair (600) and (610), as with the remote sensor pair (620) and (621), according to condition (vii). The remaining remote sensors 600, 612, 613 and 622 may be assigned any of the codewords: 'R T, Ύ R' and 'R R'.

[62] Figures 3 (a) and 3 (b) illustrate the working principle of the method of

proposed monitoring for each of the monitoring channels, λΐ and λ2, respectively. For example, when the RODT laser is tuned to the center wavelength monitoring channel λΐ, the time signature shown in Figure 3 (a) is given by passing the monitoring signal through the various optical paths of the network and by each of the remote sensors (which are represented on the abscissa axis of the graph). It can be observed that the signatures for the remote sensors 610 and 611 are complementary, allowing to differentiate the occurrence of physical failures in the fiber sections 510 and 511. The same is true for remote sensors 620 and 621. Note that repeating the code words 'TR' for remote sensors 612 and 621 and RT for remote sensors 620 and 622 is possible since condition (iii) does not apply. to the optical paths where the referenced remote sensors are located.

[63] Figure 4 gives an example of physical failures that can occur in the network shown in Figure 1. The following situation is considered: there is a section (80) in the fiber section (510) and a section (81) in the fiber section (521). Recall that the fiber sections 510 and 511 have the same length. The same is true for fiber sections 520 and 521.

[64] Figures 5 (a) and 5 (b) show the RODT traces obtained for the monitoring channels with central wavelengths λΐ and λ2, respectively. To facilitate understanding of the working principle of the method of proposed monitoring, the RODT traces for the case of Figure 1 (no physical faults) are plotted, while the RODT traces for the case of Figure 4 (physical faults) are plotted with continuous lines.

[65] Figure 5 (a) shows the RODT traces for central wavelength monitoring channel λΐ. In the possession of the table of Figure 2, it can be guaranteed that there are no physical failures in the fiber sections 500, 513 and 522, given the presence of the reflective events caused by the remote sensors 600, 613. and (622). To check for faults in fiber sections 510, 511, 620, and 621, both traces for monitoring channels with the central wavelengths λΐ and λ2 must be obtained. In Figure 5 (b), where the RODT trace for the center wavelength monitoring channel λ2 is shown, the reflective event caused by the remote sensor (610) is not present. As the reflective event of remote sensor 611 is present in the trace of Figure 5 (a), it means that a physical fault 80 has occurred and that such fault is undoubtedly in the fiber section 511. Physical failure (81) is similarly detected: the presence of the reflective event caused by the remote sensor (620) in the trace of Figure 5 (a) and the absence of the reflective event caused by the remote sensor (621) in the trace of Figure 5 ( b) attest that a physical failure (81) has occurred in the fiber section (521). The remaining information present in the trace of Figure 5 (b), namely the presence of the reflected event caused by the remote sensor (612), confirms the absence of physical faults in the fiber section (512).

[66] The example application of the RODT monitoring system presented here can easily be extended to other types of network topologies provided that it is possible to integrate RODT equipment and remote sensors into the physical layer of the fiber optic network. .

Claims

Optical system for identifying physical failures in a fiber optic network characterized by:

- by a time domain reflectometer (RODT) located at a central point of the optical fiber network capable of emitting pulses of light, preferably of short time duration and at more than one frequency, having a light source capable of generate two or more monitoring channels with different central wavelengths;

- by remote sensors associated with a two-dimensional code (amplitude and frequency) located at fiber optic network key points diametrically opposed to the location of the RODT and such that at least one remote sensor per fiber section exists in a given path. optical;

- and by at least one frequency sensitive element of at least two monitoring channels sent by the RODT.

Optical system for identifying physical failures in a fiber optic network according to Claim 1, characterized in that the light source is comprised of at least one wavelength tunable semiconductor laser.

Optical system for identifying physical failures in a fiber optic network according to claim 2, characterized in that the semiconductor laser is tunable to wavelengths in the range 2nm to 10nm.

Optical system for identifying physical failures in a fiber optic network according to claim 1, characterized in that the light source comprises at least one fixed wavelength semiconductor laser.

Optical system for identifying physical failures in a fiber optic network according to claim 1, characterized in that the remote sensors are Bragg networks.

Optical system for identifying physical failures in a fiber optic network according to claim 5, characterized in that the Bragg networks are uniform.

Optical system for identifying physical failures in a fiber optic network according to claim 5, characterized in that the Bragg networks are phase-shifted. Fiber optic network physical fault identification system according to Claim 1, characterized in that the coupling of the RODT to the fiber network is via an optical component capable of coupling the monitoring signal generated by the RODT light source. and uncoupling the monitoring signal reflected by the network and remote sensors, preferably located in the first fiber section of the network, right after all equipment connections and transmission / reception of traffic signals.

Optical system for identifying physical failures in a fiber optic network according to claim 1, characterized in that it comprises an optical filter element next to the RODT receiver capable of filtering the monitoring signal.

Optical system for identifying physical failures in a fiber optic network according to claims 8 and 9, characterized in that the optical component and the optical filtering element are a single component, which component is preferably a WDM coupler or a power splitter coupler. ability to couple / uncouple at least two different spectral bands.

Optical system for identifying physical failures in a fiber optic network according to Claim 1, characterized in that the RODT light source operates continuously without pulse generation in order to quickly perform the priority detection function, identification and location of physical faults.

Optical system for identifying physical failures in a fiber optic network according to claim 1, characterized in that it operates in a spectral band distinct from that allocated for traffic transmission.

Method for monitoring the physical structure of optical networks using the optical system described in the preceding claims, characterized in that:

- obtain the individual signatures of the various optical paths of the network simultaneously in the time domain and the frequency domain for each monitoring channel generated by the RODT in a successive and cyclical manner;

- tune successively and cyclically for each monitoring channel, the RODT light source taking into account each of the allocated wavelengths for monitoring;

- send the optical monitoring signal to the network;

- detect the evolution of the power reflected by the monitoring signal as it travels the various optical paths of the network and illuminates remote sensors in the time domain;

- record the evolution of the monitoring signal strength in the time domain by the RODT, and analyze the power attenuation versus fiber distance for each of the monitoring channels;

- Detect faults by comparing the results with a predefined code (in amplitude and frequency) for each remote sensor.

Method for monitoring the physical structure of optical networks according to claim 13, characterized by analyzing the reflective events and non-reflective events generated by remote sensors and detected by the RODT, detecting the occurrence of faults, identifying the source of such faults and proceeding to correct them. location.

Method for monitoring the physical structure of optical networks according to claim 13, characterized in that:

- allocate at least one wavelength for monitoring for each pair of optical paths whose length (for optical paths in a serial configuration) or whose difference in length (for optical paths in parallel configuration) is less than or equal to the resolution RODT space;

- determine the total number of monitoring channels (M) which shall be equal to the total number of optical path pairs that satisfy the above condition.

Method for monitoring the physical structure of optical networks according to claim 13, characterized in that it presets the spectral characteristics of the Bragg networks constituting the remote sensors.

Method for monitoring the physical structure of optical networks according to claims 15 and 16, characterized in that

preset the spectral characteristics of the Bragg grids so that each remote sensor has a number of uniform Bragg grids equal to or greater than the minimum number of wavelengths allocated for monitoring.