Determining the location of optical fibres

Abstract

This application relates to methods and apparatus for determining a location of at least part of an optical fibre (104). The method involves taking a data set comprising sensor output data acquired by performing fibre optic sensing on the optical fibre (105), where the data set comprises sensor output data from each of a plurality of sensing portions of the first optical fibre. The method comprises analysing the data set to detect at least one characteristic signal (CS) which corresponds to the detected response to an event which is not caused to occur for the purpose of locating the first optical fibre. The method involves determining a first known location associated with the characteristic signal and determining a location of at least one sensing portion of the first optical fibre based on the characteristic signal and the first known location.

Description

[0001] DETERMINING THE LOCATION OF OPTICAL FIBRES

[0002] This application relates to methods and apparatus for determining the location of at least part of an optical fibre, in particular to determining the location of at least part of an optical fibre that forms part of, or is co-located with, some asset or infrastructure so as to provide an indication of the location of that asset or infrastructure.

[0003] There are a number of applications in which it may be desirable to be know the location of some asset or infrastructure, but the location of that asset or infrastructure may not be known, at least to a desired degree of accuracy.

[0004] For example, it may be desirable to know the location of underwater or subsea cables, such as power or communication cables. Inspection and / or repair of such subsea cables may be difficult and costly, involving sending equipment and / or personnel to the seabed to identify and inspect the cable and / or raise it to the surface for repair. If the location of the cable is not known, or the position can only be estimated to a relatively large area, this may also involve a relatively significant underwater search to try to locate the cable, which can greatly add to the time and costs involved, particularly if the cable is buried in the seabed. In addition, there may be a desire to know the location of the subsea cable in order to define zones in which certain activities, such as ships dropping anchors or trawling the seabed, is prohibited to avoid damage to the cable. In some cases, the location of the cable at the time of deployment may only be known to a certain accuracy, e.g. the cable may have been deployed from a ship on the sea (or other waterbody) surface, where the surface position of the ship at the time of deployment is only estimated. Even if the surface position of the ship at the time of deployment is known to a relatively high accuracy, the actual location of the cable on the seabed may vary from the surface position and may vary over time due to underwater currents, tidal forces, shifting sands and the like.

[0005] Likewise, for onshore applications, there may be assets or infrastructure such as data or power cables or pipework or the like, that may be located underground or within structures, such as walls or conduits or the like. In some cases, such as in a large industrial site or urban environment, there may be a relatively complex network of such cables / pipes and the exact location of a given cable / pipe may not be known over its entire length. Again, this can be problematic in terms of locating the relevant asset for maintenance and / or repair. Likewise, it may be desirable to know the location of theasset to avoid accidental damage, e.g. to know the location of underground pipes when planning some ground works.

[0006] Such assets or structures may, in at least some cases, be considered to be hidden infrastructure, in the sense that they are structures that are not easily visible and / or accessible.

[0007] Embodiments of the present disclosure relate to methods and apparatus for locating optical fibres that form part of, or are co-located with, some asset or infrastructure, e.g. some hidden infrastructure, so as to provide geolocation of that asset or infrastructure.

[0008] Thus, according to an aspect of the invention there is provided a method of determining a location of at least part of a first optical fibre, comprising taking a data set comprising sensor output data acquired by performing fibre optic sensing on the first optical fibre, wherein the data set comprises sensor output data from each of a plurality of sensing portions of the first optical fibre. The method comprises analysing the data set to detect at least one characteristic signal, wherein the characteristic signal corresponds to the detected response to an event which is not caused to occur for the purpose of locating the first optical fibre; determining a first known location associated with the characteristic signal; and determining a location of at least one sensing portion of the first optical fibre based on the characteristic signal and the first known location.

[0009] In some examples, the characteristic signal may comprise a transient characteristic signal. In which case, determining the first known location associated with the characteristic signal may comprise determining a source location of the transient characteristic signal. The method may further comprise determining a generation time of the transient characteristic signal at the source location and determining the location of the at least one sensing portion of the first optical fibre may comprise determining a time difference between the generation time of the transient characteristic signal at the source location and the time of detection of the characteristic signal at the at least one sensing portion. In some implementations, the transient characteristic signal may be generated by a lightning strike. Determining the first known location associated with the characteristic signal and determining the generation time of the transient characteristic signal at the source location may comprise determining the time and location of the lightning strike from a lightning tracker data set.In some implementations, the transient characteristic signal may be generated by an earthquake or earth tremor. Determining the first known location associated with the characteristic signal may comprise determining at least one of a source location of the earthquake or earth tremor and / or a position of at least one seismic wave generated by the earthquake or earth tremor.

[0010] In some implementations, the transient characteristic signal may comprise a characteristic signal generated by a moving source. The moving source may comprise a motor power vehicle. The first optical fibre may be part of an underwater cable and the moving source may comprise a water going vessel. The characteristic signal may comprise at least one of a characteristic acoustic signal generated by the water going vessel and a characteristic pressure variation caused by movement of the water going vessel through the water. Determining the first known location associated with the characteristic signal may comprise determining the position of the water going vessel over time. The water going vessel may be a ship and determining the position of the ship over time may comprise interrogating a data set of automatic identification system data. Determining the location of the at least one sensing portion of the first optical fibre may comprise determining at least a first sensing portion as a sensing portion under a crossing point of the water going vessel over the first optical fibre and determining the location of the at least one sensing portion of the first optical fibre may comprise determining a crossing time of the water going vessel over the first optical fibre and identifying the location of the at least a first sensing portion based on the position of the water going vessel at the crossing time. Determining the at least a first sensing portion as a sensing portion under a crossing point of the water going vessel over the first optical fibre may comprise analysing sensor output data using a model trained by machine learning.

[0011] In some examples, the characteristic signal may comprise an acoustic signal generated by operation of some fixed infrastructure. The fixed infrastructure may comprise a traffic control apparatus and the characteristic signal may comprise an acoustic signal generated by traffic flow in response to operation of traffic control apparatus.

[0012] In some examples, the characteristic signal may comprise a thermal stimulus.

[0013] In another aspect there is provided an apparatus for determining a location of at least part of a first optical fibre comprising a processor configured to: take a data setcomprising sensor output data acquired by performing fibre optic sensing on the first optical fibre, wherein the data set comprises sensor output data from each of a plurality of sensing portions of the first optical fibre and process the data set to: detect at least one characteristic signal, wherein the characteristic signal corresponds to the detected response to an event which is not caused to occur for the purpose of locating the first optical fibre; determine a first known location associated with the characteristic signal; and determine a location of at least one sensing portion of the first optical fibre based on the characteristic signal and the first known location.

[0014] The apparatus may operate according to any of the embodiments of the method described herein.

[0015] Also provided is a computer readable medium comprising computer readable code that, when run on a suitable processor, cause the processor to perform the method of any of the embodiments described herein.

[0016] The discussion above considers various different features or options of the embodiments of the disclosure. It will be understood that, unless clearly incompatible, each of the features described above may be implemented in combination with any one or more of the other features described, depending on the particular use case.Embodiments of the present disclosure will be now be discussed with reference to the accompanying drawings, of which:

[0017] Figures 1a- 1c illustrate one example of a method of locating an optical fibre with generation of a transient propagating signal;

[0018] Figures 2a and 2b illustrate another example of a method of locating an optical fibre with generation of a transient propagating signal; and

[0019] Figures 3a and 3c illustrate the generation of multiple different transient propagating signals in different known locations;

[0020] Figure 4 illustrates a method of locating an optical fibre based on movement of a constant source;

[0021] Figures 5a and 5b illustrate acoustic data acquired from fibre optic sensing on an underwater cable as a motor-powered ship passed overhead;

[0022] Figures 6a - 6e illustrate pressure data acquired from fibre optic sensing on an underwater cable as a ship passed overhead; and

[0023] Figure 7 illustrates a method of locating an optical fibre based on control traffic flow.

[0024] Embodiments of the present disclosure relate to methods and apparatus for determining the location of at least part of an optical fibre, in particular to determining the location of part of an optical fibre that form part of, or is co-located with, some asset or infrastructure, so as to provide an indication of the location of that asset or infrastructure. Determine the location of at least part of the optical fibre may comprise geolocating that part of optical fibre, i.e. determining the geographical position of the optical fibre. Locating the optical fibre may comprise determining the location of the optical fibre in two-dimensions, e.g. determining two-dimensional co-ordinates in a suitable coordinate system, such as longitude and latitude or similar. In some examples, locating the optical fibre may comprise determining a three-dimensional location, e.g. to also provide an indication of elevation or depth with regard to some defined surface, such as the local ground surface or sea level. In some examples,determining the location may comprise determining a relative location with respect to some other optical fibre and / or some other asset.

[0025] As noted in the discussion above, there may be a number of applications where it is desirable to know the location, e.g. the geographical location, of some asset, which may, for example, comprise a pipeline or cable, for instance a power cable or communication cable or the like. Such an asset may, in some cases be considered to be some hidden infrastructure.

[0026] By way of example, as noted previously, it may be desirable to know the location of an underwater cable, which could, for instance, be a power cable or communications cable or the like. It may be desirable to know the location for the purposes of maintenance, e.g. inspection and / or repair, and / or to avoid performing activities in the vicinity of the underwater cable that could result in damage to the underwater cable. As also noted above, in some cases the position of the cable when deployed may only be known relatively roughly, e.g. based on an estimate of the surface position of a ship deploying the cable and, even if the position of the cable at initial deployment (or subsequently) was known relatively accurately, the position of at least part of the cable may have changed significantly over time. In some cases, it may additionally or alternatively be desirable to know the relative location of one cable with respect to another cable. For instance, two underwater cables may follow generally the same path as one another. If it is known that the cable with a first optical fibre is the cable that needs repair, it may be desirable to be able to determine which of the two underwater cables is the one with the first optical fibre, e.g. as seen from a first direction which runs generally along the path of the underwater cables, is the relevant cable with the first optical fibre on the right or the left.

[0027] In embodiments of the present disclosure, fibre optic sensing may be performed on a sensing optical fibre which forms part of, or is otherwise co-located with, the asset it is wished to locate, at least in one location. The fibre optic sensing may be performed on the sensing optical fibre to detect stimuli acting on one or more sensing portions of the sensing optical fibre and, in at least some embodiments, the fibre optic sensing may detect stimuli arising from incident mechanical waves or similar acting on the sensing optical fibre. The stimuli acting on the sensing portions of the sensing optical fibre can be analysed to determine at least some characteristic signals of interest. The characteristic signals of interest, as will be described in more detail below, are signalsthat have at least one characteristic that allow the signal to be identified as a signal of interest in the sensor output of the fibre optic sensor and for which a location associated with the signal of interest, e.g. the location of the source of the signal of interest, is known or can be separately determined (by some other means). Based on the detection of the characteristic signals of interest and the relevant locations associated with such signals, information regarding the location of the relevant sensing portion(s) of the optical fibre, and hence the location of the co-located asset, can be determined, as will be discussed in more detail below.

[0028] In some examples, the sensing optical fibre may an optical fibre that has previously been co-located with the relevant asset, e.g. the hidden infrastructure, for some other purpose or reason. As will be described in more detail below, the sensing optical fibre can, in some examples, be a conventional optical fibre such as would typically be used for optical communications, e.g. a conventional single-mode optical fibre as would typically be used for telecommunications. There are a number of use cases where such an optical fibre may be formed as part of, or co-located with, some asset for the purpose of allowing data communication.

[0029] For example, where the asset is a communication cable, such as an underwater telecommunication cable or some onshore communication cable as may be deployed in an industrial or urban environment, the communication cable itself will typically comprise a plurality of optical fibres for optical data communication and any of these optical fibres could be used as the sensing optical fibre. In some cases, there may be some spare or unused optical fibres, e.g. optical fibres provided for redundancy, which could be used for the fibre optic sensing, otherwise one of the optical fibres which is generally used for data communication could, when required, be repurposed for fibre optic sensing.

[0030] At least some other underwater or subsea cables, such as underwater power cables, may also typically be provided with one or more optical fibres for the purpose of communication. For instance, an underwater power cable for power transfer, e.g. for transferring power from an offshore wind farm to an onshore grid, may often also include at least one optical fibre for allowing communication with the offshore windfarm and such an optical fibre could be used as a sensing fibre.In onshore applications, some assets such as a buried power cable or oil pipeline or the like may also be provided with one or more optical fibres formed as part of, or arranged to run along the path of, the asset to allow for data communication, e.g. for monitoring and / or control of the pipeline.

[0031] The ability to perform (at least some types of) fibre optic sensing on conventional optical fibres, as would typically be used for telecommunication, and which have already been deployed together with an asset whose location it is desired to determine, allows for the techniques described herein to be used for the location of assets that have previously been deployed, without the need for any retrofitting of a sensing optical fibre to the relevant asset.

[0032] However, in some use cases, an optical fibre could be deliberately included as part of, or co-located with, the relevant asset for the purposes of allowing fibre optic sensing to be performed. E.g. if an underwater cable were to be deployed that would not otherwise include an optical fibre, one could be formed as part of that cable, or attached to the cable prior to or during deployment, so as to provide a sensing optical fibre that runs along the length of the underwater cable to allow for location detection after deployment, e.g. to account for any movement of the cable after deployment. In this case, the sensing optical fibre may be a conventional optical fibre as could be used for telecommunications, as such optical fibres are readily commercially available and the sensing optical fibre could, when not used for fibre optical sensing, be separately used for data communication. In some cases, however, if a sensing optical fibre is deliberately included to allow for fibre optic sensing, then the sensing optical fibre may be selected to have some properties which are particularly suited for a particular type of fibre optic sensing, e.g. an optical fibre in a cable structure which is particularly sensitive to incident strain / vibration and / or which includes features such as fibre Bragg gratings (FBGs) to act as discrete sensors.

[0033] Various different fibre optic sensing techniques are known and may be employed in embodiments. For instance, the fibre optic sensing may involve distributed acoustic sensing (DAS) based on Rayleigh backscatter.

[0034] As will be understood by one skilled in the art, one form of DAS sensor, that could be used in embodiment of the disclosure, repeatedly interrogates the sensing optical fibre by launching one or more pulses of coherent optical radiation into the sensing opticalfibre and uses the techniques of OTDR (optical time domain reflectometry) to associate the backscatter received at different times after launch with different sensing portions of the sensing optical fibre at different distances along the sensing optical fibre. As the interrogating radiation is coherent, the backscatter from any given sensing portion is an interference signal that depends on the distribution of scattering sites within that sensing portion, which is generally random. In the absence of any environmental stimuli acting on the sensing portion, the backscatter from a given sensing portion should be the same for each interrogation (assuming the properties of the optical radiation launched into the sensing fibre are the substantially same for each interrogation). An environmental stimulus, such as a strain arising from an incident acoustic wave acting on a sensing portion, can cause a change in the distribution of the scattering sites within that sensing portion, with a change in the Raleigh backscatter from that sensing portion, as the sensing portion effectively acts as a virtual interferometer which is affected by strains causing a change in the optical path length between different scattering sites of the sensing portion. This change in the backscatter from a given sensing portion can be detected and used to provide a signal indicative of the stimulus acting on that sensing portion. In this way, the sensing optical fibre can effectively be divided into a plurality of longitudinal sensing portions along its length, which may or may not be contiguous depending on the processing, and where the size and location of the sensing portions are defined by the form of interrogating radiation used and / or the processing applied to the backscatter.

[0035] It will be understood by one skilled in the art that the discussion above is just one example of a suitable DAS sensor and various other varieties of DAS sensor could be used. For instance, some DAS sensors use interrogating radiation with pulses at different optical frequencies to one another and / or some other time varying frequency modulation and / or use a local oscillator with some frequency difference to the interrogating radiation which is launched into the sensing optical fibre to allow for processing of the detected signals with regard to a carrier frequency defined by the relevant frequency difference. Additionally or alternatively, some DAS sensors may repeatedly interrogate the sensing optical fibre where each interrogation involves launching interrogating radiation which is encoded by some modulation pattern, e.g. a modulation of amplitude and / or phase, with the backscatter from a given interrogation being identified by correlation of the backscatter with the relevant modulation code. In some examples, polarisation-based sensing and / or frequency domain techniques may be used to provide additional or alternative functionality.It will also be understood that whilst DAS is suitable for detecting incident acoustic waves acting on the sensing fibre, DAS is also sensitive to other stimuli that can act on the sensing fibre to cause a change in optical path length of a sensing portion, and this will include strains on the sensing fibre causes by vibration or other types of mechanical waves, e.g. other pressure or seismic waves, as well as temperature variations. Depending on the stimulus of interest, filtering may be applied to the sensor output to provide suitable sensor output signals.

[0036] Whilst the fibre optic sensing may, conveniently, be performed using DAS, in particular when the sensing optical fibre is a conventional single mode optical fibre suitable for telecommunications, other fibre optic sensing techniques could additionally or alternatively be used. For instance, the fibre optic sensing could comprise distributed strain sensing (DSS) or distributed temperature and strain sensing system (DTSS) based on Brillouin scattering. Additionally or alternatively, the fibre optic sensing may make use of discrete sensors formed in, or coupled to, the sensing optical fibre at different positions, for instance using a fibre Bragg grating (FBG) sensing system or the like. In some applications, the fibre optic sensing could additionally or alternatively monitor for other environmental stimuli other than strain / vibration and in some cases the fibre optic sensing could comprise distributed temperature sensing (DTS), for instance based on Raman scattering.

[0037] In general, the fibre optic sensing provides a means for monitoring for characteristic signals acting on the sensing optical fibre. The characteristic signals may comprise any type of signal that causes some mechanical displacement, and shall include acoustic or other pressure waves or pressure variations in fluid such as water or air, and shall also include seismic or other waves that may propagate through a solid medium such as the ground. Note, as used herein, the term acoustic does not limit to sound that may be audible to a human.

[0038] As noted above, the characteristic signals are signals which have characteristics that allow the characteristic signals to be identified as such in the sensor output generated by the fibre optic sensing. That is, the sensor output can be analysed to detect the presence of a characteristic signal. As such, the characteristic signal has one or more properties that allow it is be identified and discriminated from ambient noise that maybe detected by the fibre optic sensing. The characteristic signals are also signals that can be identified as originating from a defined source location.

[0039] In some examples, the characteristic signals may be signals which are specifically generated for the purposes of allowing the optical fibre to be located. This may form part of what may be termed an active method of location, as the characteristic signals are actively generated. Actively generating the characteristic signals can allow at least some properties of the characteristic signals to be specifically controlled, which can ease the detection of the characteristic signals. For example, the characteristic signals could be generated at one or more specific acoustic frequencies and / or some characteristic modulation, e.g. in frequency and / or amplitude, could be applied to the characteristic signal provide ease of detection. The active generation of the characteristic signals can also be controlled to occur at a defined generation time which allows the sensor output in a defined time window based on the defined generation time to be analysed to detect the characteristic signal, which can aid in identifying the characteristic signals.

[0040] In some examples, however, the characteristic signals may be signal of opportunity which are not specifically generated for the purpose of locating the sensing optical fibre. Such characteristic signals of opportunity may be generated by natural events or may be man-made but generated as part of some other activity. As will be described in more details below, there are some events or activities that may provide a characteristic signal that can be identified in the sensor output of the fibre optic sensor and for which a location can be identified. The use of characteristic signals of opportunity can be referred to as a passive method of location.

[0041] The active method of location controls the time and location of the generation of the characteristic signal and also the properties of the characteristic signal, which can greatly ease the identification of the characteristic signals in the sensor output compared to passive sensing. However, the active method of location does involve having to have some equipment and / or personnel in the vicinity of the sensing optical fibre in order to generate the characteristic signal, and there may be a need to generate a plurality of different characteristic signals in different location, which could cover a large area, to be able to determine the location of the sensing fibre to a desired accuracy across a certain length of the sensing fibre - which could be tens or hundreds of kilometres in length. This may be time consuming and costly.Whilst embodiments of the present disclosure may involve active methods of location, embodiments may particularly additionally or alternatively use passive methods of location and the ability to perform passive methods of location represents a particularly novel aspect of this disclosure.

[0042] Active methods of location will be described first, to outline some of the principles of determining location using fibre optic sensing.

[0043] Active Methods of Location

[0044] As noted above, active methods of location involve controlled generation of at least one characteristic signal at a known generation location and at a known generation time.

[0045] For active methods of location, the characteristic signal may, for example, comprise some relatively high amplitude transient signal which will propagate from a source location to the sensing fibre, e.g. some kind of acoustic or mechanical impulse such as a gunshot or similar. Given the known generation time of the characteristic signal, a certain time window can be defined for analysis of the sensor output from the fibre optic sensing to detect the characteristic signal in the sensor output. For instance, the start of the time window may be based on the generation time, and the end of the time window may be based on a maximum expected propagation delay between the source location and one or more sensing portions of the sensing optical fibre. The characteristic signal may be detected by looking for a relatively high amplitude signal in the sensor output that affects the relevant sensing portion(s) in the time window.

[0046] In some environments, the high amplitude transient signal may be generated to have a significantly higher amplitude than the expected ambient noise and thus may comprise the only high amplitude transient in the sensor output in the relevant time window. In some cases, the characteristic signal may comprise some distinctive modulation, e.g. may comprises a series of two or more transients at defined intervals, so as to aid detection of the relevant characteristic signal in the sensor output. In some examples, there may be different characteristic signals generated at different times at the same location to provide different corresponding time windows for analysis, i.e. there may be repeated generation of characteristic transient signals, which each propagate to thesensing fibre, to provide more date for analysis to hopefully reduce error and improve accuracy.

[0047] The analysis of the sensor output from the fibre optic sensing may comprise identifying the characteristic signal in the sensor output and the using time-of-arrival analysis to determine information about the location of the sensing fibre.

[0048] In some examples, the time of arrival of the characteristic signal at different sensing portions of the sensing optical fibre (which may be individual sensing portions or channels of a DAS sensor and / or individual sensors such as FBGs, depending on the type of fibre optic sensing applied) may be analysed.

[0049] Figures 1a - 1c illustrate one basic example of how the characteristic signals may be detected and analysed to determined location information.

[0050] Figure 1a illustrates generally that, for an active method of location, signal CS may be generated as a characteristic signal by a suitable signal generator 101. The characteristic signal CS is generated at a known source location 102, which may, for instance, by determined by a positioning system 103 such as a GPS (global positioning system) device or the like, which may be co-located with the signal generator 101. As noted above, in some examples, the characteristic signal CS may comprise one or more relatively high amplitude transient signal, e.g. such an acoustic signal as could be generated by a gunshot, or a short, loud acoustic signal such a siren or foghorn or the like, or (for onshore application) some percussive stimulus creating a ground shock.

[0051] The characteristic signal CS is generated at the source location 102 where it is expected that the characteristic signal CS will be able to propagate to a sensing optical fibre. In the example of figure 1a, the sensing optical fibre forms part of a fibre optic cable 104 deployed in an area and which may not be readily visible, e.g. may be underwater or buried in the ground or within a structure. The exact position of the fibre optic cable 104 may not be known, at least over part of the length of the fibre optical cable 104. In some examples, the fibre optic cable 104 may be formed as part of, or otherwise co-located with, some other linear asset such as a power cable or pipeline or the like (not illustrated in figure 1a).To allow the location of the sensing optical fibre of the fibre optic cable 104 to be determined, fibre optic sensing is performed on the sensing optical fibre of the fibre optic cable 104. One end of the sensing optical fibre of the fibre optic cable 104 may thus be connected to a suitable interrogator unit 105 for performing fibre optic sensing on the sensing optical fibre. The interrogator unit 105 may, for example be a DAS interrogator unit. The interrogator unit 105 may be coupled to an end of the sensing optical fibre of the fibre optic cable 104 which is remote from the location of the acoustic generator. As will be understood by one skilled in the art, some forms of fibre optic sensing can operate on lengths of optical fibre of tens or even hundreds of kilometres in length. The position of a first and / or second end of the fibre optic cable 104 may be known, but the path of at least part of the fibre optic cable 104 between the first and second ends may be unknown, at least to a desired degree of accuracy. For instance, for an underwater cable that runs from a first onshore location to a second onshore location via a body of water, the first and second onshore locations may be known, but the underwater path of the cable may not be known. The first and / or second onshore location may be accessible, and the interrogator unit may thus be installed at the first and / or second onshore location as appropriate. The sensor output from the interrogator unit 105 may be provided to a processor 106, which could be colocated with the interrogator or which could be remote, for processing to determined the location of the sensing fibre. It will be noted that the fibre optic sensing should be performed during the period in which the characteristic signal CS is being generated and propagates to the sensing optical fibre of the fibre optic cable 104, but the sensor output could simply be recorded for subsequent analysis - although in some cases some contemporaneous processing to determine location may be performed to inform subsequent measurement, i.e. to inform whether another characteristic signal should be generated, if the properties of the characteristic signal should be changed and / or whether the generation location should be changed.

[0052] As discussed above, performing fibre optic sensing on the sensing optical fibre of the fibre optic cable 104 can provide sensing from a plurality of different sensing portions of the sensing optical fibre, at different distances along the length of the sensing optical fibre. The sensor output from each sensing portion, or channel, of the fibre optic sensor can then analysed to look for a detected signal component corresponding to the characteristic acoustic signal. Where the characteristic signal is simply a high amplitude transient acoustic or similar signal, the analysis of the sensor output from each sensing portion may simply comprise looking for a corresponding relatively hightransient in the detected signal, within a time window based around the time of generation of the characteristic signal.

[0053] Whilst the output from each sensing portion could be analysed individual, in some examples the output from multiple adjacent sensing portions may be analysed together to aid in identifying the characteristic signal, as it may be expected that sensing portions which are relatively close to one another will be affected by the characteristic signal relatively closely in time.

[0054] In general, the characteristic signal CS, which is generated at a generation time TO, may propagate through any relevant medium, such as water and / ground, to the sensing optic fibre of the fibre optic cable 104 (unless otherwise attenuated). When the characteristic signal CS is incident on any sensing portion, the resultant strain or vibration can cause a detectable signal component in the sensor output of that sensing portion. The time at which such signal component appears in the output of the relevant sensing portion depends on the propagation time for the characteristic signal CS to propagate from its source 102 to the relevant sensing portion, which in turn depends on the distance of the source 102 of the characteristic signal to the sensing portion and the propagation speed in the relevant medium.

[0055] Figure 1b illustrates a plan view of the path of the fibre optic cable 104 with respect to the source 102 of the characteristic signal CS and illustrates that, in this example, the path of the fibre optic cable 104 is substantially linear, at least within the area of the acoustic signal generator 101. As such, individual sensing portions SP can be seen as being spaced linearly from one another. Figure 1b illustrates that the characteristic signal may radiate outwards from the source location 102 with time, and may thus interact with the different sensing portions of the sensing optical fibre of the fibre optic cable 104.

[0056] In this example, the characteristic signal CS is generated at a time TO and propagates outwards and reaches a first sensing portion, labelled SP1, at a time T1 later, resulting in a detectable high intensity signal component. In general, where the propagation speed of the characteristic signal CS (e.g. the speed of sound in the relevant medium for an acoustic signal) is substantially the same in each direction (or exhibits substantially the same variation in each direction), then the first sensing portion SP1 to detect the characteristic signal will the sensing portion which is closest to the sourcelocation 102 of the acoustic signal. The characteristic signal then continues to propagate outwards and is detected next, in this example, by the sensing portions on either side of the first sensing portion SP1, and then the next adjacent sensing portions and so on. Figure 1c illustrates an example of the sensor output from the plurality of sensing portions, illustrating detected signal intensity as a function of time and sensor channel, i.e. distance along the sensing fibre. As noted, the sensing portion SP1 is the first to register the characteristic signal, at the time T 1 and then the characteristic signal is detected by successive sensing portions on either side of this first sensing portion SP1 at successively later times. This results in, in this example, a characteristic chevron or V-shape in the sensor output. This characteristic V-shape is indicative of a characteristic acoustic or similar signal propagating in the area and affecting sensing channels in turn, and the analysis may, in some cases, involve looking for such a characteristic response in the sensor output from the plurality of different sensing portions as a way of, at least partly, identifying the characteristic signal in the sensor output.

[0057] Having identified the characteristic signal in the sensor output of the plurality of sensing portions, the sensor output corresponding to the characteristic signal can be analysed to provide location information about the location of the fibre optic cable 104.

[0058] It will be understood that the time of arrival of the characteristic acoustic signal at each sensing portion will be a function of the path length that the characteristic signal travels to reach that sensing portion and the propagation speed along that path. If the medium through which the characteristic signal travels is relatively homogeneous, the path between the source location 102 and the relevant sensing portion SP could be a substantially straight path and thus the path length may simply be the straight-line distance between the source location 102 and the relevant sensing portion, and the propagation speed along each path may also be substantially constant. Thus, the travel time between the source location 102 of the characteristic signal CS and a given sensing portion will be equal to d*v, where d is equal to the distance between the source location 102 and the relevant sensing portion and v is the propagation speed, which may (in this simplified example) be taken to be constant and the same for each sensing portion. Thus, the time difference between the time TO at which the characteristic signal is generated and the relevant time that the characteristic signal can be seen in the output from the sensing portion is indicative of the distance between that sensing portion and the source location 102 of the characteristic signal CS. Figure1c illustrates a time difference 5T between the time TO at which the characteristic signal AS is generated, which can be measured accurately, e.g. by the positioning apparatus 103, and the time T1 at which the characteristic signal is detected in the sensing portion SP1, which can be recorded by an accurate clock, e.g. a GPS based clock, which could, for example, be formed as part of the interrogator unit 105. The sensor output from the interrogator unit 105 may thus be accurately time stamped. In this example, this time difference ST, which may be seen as a time offset, is indicative of the shortest distance d between the source 102 of the characteristic signal CS and the sensing optical fibre of the fibre optic cable 104. If this offset time period ST were substantially equal to zero, i.e. there was no substantially no time difference between the time TO at which the characteristic signal is generated and the time T1 at which the characteristic signal is detected by the sensing portion SP1 , then this would indicate that the source 102 of characteristic signal was effectively at substantially the same location as the sensing portion SP1. This could be the case, for instance, if the fibre optic cable 104 were buried in the ground just below the surface of the ground and the signal generator 101 were located substantially above the relevant sensing portion of sensing optical fibre of fibre optic cable 104. If the offset time ST is not zero, then as mentioned the offset time may correspond to d*v and if the speed of propagation v were known, or could be estimated, then the actual distance d between the source 102 of the characteristic signal CS and the sensing portion SP1 could be determined or estimated. For a fibre optic cable buried in the ground near to the surface, the determined distance d may effectively be the lateral distance between the source 102 of the characteristic signal CS and the sensing portion SP1. However, for a fibre optic cable 104 buried in the ground at some significant depth, or for an underwater cable at depth (where the source 102 of the characteristic signal is at, or near, the surface) then the distance d would depend on the lateral distance and also the depth - but if the depth could be estimated, the lateral distance could be calculated.

[0059] Having determined the distance between the source 102 of the characteristic signal CS and the closest sensing portion SP1 of the fibre optic cable 104, a similar analysis could be applied for each of the other sensing portions SP (provided the characteristic signal can be identified in the sensor output for that sensing portion), so as provides a distance to each of the sensing portion. Given a known spacing between the sensing portions, a plot of the position of the sensing portions relative to the source 102 of the acoustic signal can be determined.It will be understood that figures 1a-1c illustrate an example where the path of the sensing optical fibre, in the vicinity of the generator 101 of the characteristic signal, is substantially straight or linear, but in some cases the path of the sensing optical fibre may not linear. Figure 2a illustrates a plan view of a different example, where the sensing optical fibre of the fibre optic cable 104 has a path that is not straight (at least in the vicinity of the source location 102 of the characteristic signal CS). Again, the characteristic signal CS, which is generated at a time TO, will radiate outwards and reach the closest sensing portion SP1 at a time T1, before going on to reach the next closest sensing portions and so on (assuming the propagation speed in different directions is substantially homogeneous). Figure 2b illustrates an example of the detected response to the characteristic signal CS. Again, the offset time between the time TO at which the characteristic signal is generated and the time at which is the characteristic signal is present in the sensor output for a give sensing portion is indicative of the distance between that sensing portion and the source location 102. If the (average) propagation speed is known or can be estimated, the distance between the source location 102 and the relevant sensing portion SP can be determined or estimated and, given the known spacing between the sensing portions, a plot of the path of the fibre optic cable determined.

[0060] It should be noted that in the examples of figures 1a-1c and 2a-b, the analysis to determine location of the sensing optical fibre of the fibre optical cable 104, based on a characteristic signal generated at a single location will have some ambiguity, in particular some symmetric ambiguity, in that any symmetrical arrangement of the sensing optic fibre with regard to the source location 102 (which could be rotationally symmetric around the source location 102 or symmetric with regard to any axis through the source location 102) would result in the same sensor output.

[0061] For instance, for the example with a linear path for the fibre optical cable 104, a source which is a lateral distance d1 from the closest sensing portion SP1 in one direction would produce the same sensor output as if the source had instead been the lateral distance d1 in the opposite direction. Thus, whilst it may be possible to determine a lateral distance of each of a plurality of sensing portions of the sensing optical fibre of the fibre optic cable 104 from the source 102 of the characteristic signal CS, it may not be possible, based on this analysis alone, to determine the direction in which the cable lies, e.g. for a fibre optic cable 104 that runs generally east-west, the cable could be laterally north or laterally south of the source location 102. In some cases, thisambiguity may be resolved by some other prior knowledge, e.g. if the signal generator 101 is positioned such that it is known the fibre optic cable must be to the south, this ambiguity can be reduced or potentially eliminated. Otherwise, the ambiguity can be at least partly resolved by also analysing the response to characteristic signals generated in different locations.

[0062] Figure 3a illustrates this principle. Figure 3a illustrates that a first characteristic signal CSa is generated at a first generation location 102a and fibre optic sensing applied, as discussed above, to detect the time of the arrival of the first characteristic signal at one more sensing portions of the sensing optical fibre. A second characteristic signal CSb is generated at a second generation location 102b, which is different to the first generation location 102a. The second characteristic signal CSb could be generated at the same time as the first time characteristic signal CSa, in which case at least one property of the first and second characteristic signals CSa and CSb should differ from one another to allow for each characteristic signals CSa and CSb to be correctly identified in the sensor output. Conveniently, however, the first characteristic signal may be generated at the first generation location 102a at a first time and then the signal generator 101 moved to the second generation location 102b and used to generate the second characteristic signal CSb.

[0063] In any event, the respective time offsets <5Ta and 5Tb between the respective first and second characteristic signal being generated and the respective characteristic signal being detected in the senor output for a given sensor portion may be determined. Based on the known first and second generation locations, the respective distances da and db from the first and second generation locations 102a and 102b can be estimated as discussed above and the position of the sensing portion that is a distance da from the first generation location 101a and distance db from the second generation location 101b can be determined. As illustrated in figure 3a, in two dimensions this may limit the position of the sensing portion to one of two possible locations. Thus, if the elevation or depth of the sensing optical fibre is known or can be neglected, the two-dimensional position can be narrowed to one of two locations using the sensor output from just one sensing portion. In practice, the results from the plurality of different sensing portions, where the distance between the adjacent sensing portions is known, may be used to resolve any ambiguity. Additionally or alternatively, as illustrated in figure 3b, generating at least a third characteristic signal CSc in a third generation location 102c may improve the positional estimate and remove the ambiguity in twodimensions. It will be understood by one skilled in the art that the active method of location of the present disclosure may thus use multilateration type techniques to determine the position of one or more sensing portions of the sensing optical fibre.

[0064] The discussion above has assumed that the propagation speed of the characteristic signal between the source location 102 is known or can be estimated and also assumes that the propagation speed does not significant differ in different directions. For many applications, this assumption may be valid. For instance, where the sensing optical fibre forms part of an underwater cable and the characteristic signal is an acoustic signal which is generated near the surface, e.g. by a sonar projector or the like, it will be understood that the speed of propagation may vary throughout the waterbody with depth, e.g. due to changes in temperature and / or salinity etc. with depth. However, the cable may be expected to lie on the waterbody bottom at substantially the same depth and thus the same variation in propagation speed may be experience for the characteristic acoustic signals travelling to each sensing portion and thus an average propagation speed may be used. In some cases, the propagation speed could be determined as part of the process of generating the characteristic signal, e.g. by detecting the round-trip time for echoes to return from a known depth or by using hydrophones at different depths.

[0065] As discussed above whilst active methods of location can usefully be used to determine location information regarding the geographical location of sensing optical fibre, this can involve a relatively timely and costly surveying process and at least some embodiments make use of a passive method of location.

[0066] Passive method of location

[0067] In the passive method of location, the sensor output from performing fibre optic sensing may be analysed to detect any characteristic signals of opportunity, which are signals that occur in the environment in the which the sensing optical fibre is deployed that are not specifically generated for the purposes of determining the location of the sensing optical fibre, but which, nevertheless can be identified as being of a certain type and associated with a given location.

[0068] Transient SignalsIn some examples, the characteristic signal of opportunity could be a transient signal of a relatively high amplitude, e.g. a relatively high amplitude acoustic or similar signal, that can be detected above the ambient background noise. If the generation time and the generation location of the characteristic signal of opportunity is known or can be determined, then processing similar to that described above for the active method of location can be used to determine location information for the sensing optical fibre.

[0069] Such transient high-amplitude signals could, in some examples, be generated by natural causes. For instance, lightning strikes on the ground generate a relatively intense mechanical shock wave in the ground, which can propagate though the ground as an characteristic signal CS generated at a source location (which is the location of the lighting strike). In many parts of the world, the time and location of lightning strikes are tracked, to relatively high precisions, for instance using radio detection (at suitable frequencies) to detect the electromagnetic waves generated by a bolt of lightning. The time of arrival of the radio waves at different detection stations can be used to triangulate the position of the lightning strike with a relatively high accuracy. The data regarding the time and position of lightning strikes is available and can be used in a passive method of location according to an embodiment.

[0070] Lightning strikes may also occur on the sea or ocean or other waterbody surface. Lightning may, in particular, tend to strike wind turbines or other offshore installations, which can also be generate shock / acoustic wave which could propagate to an underwater cable.

[0071] If a thunderstorm occurs in area where a sensing optical fibre is located, fibre optic sensing may be performed on the sensing optical fibre during the storm. In some cases, the fibre optic sensing may be performed specifically if it is known that a thunderstorm is occurring, or is likely, but in some applications fibre optic sensing may be routinely performed on a given sensing optical fibre, e.g. in a generally continuous or periodic manner as part of some other monitoring purposes, such as health or other monitoring, and the monitoring may happen to coincide with the occurrence of a thunderstorm. In which case, the sensor output generated during the storm may be analysed.

[0072] The sensor output data may be analysed based on an event driven approach and / or a data driven approach. In the event driven approach, the data set regarding the timeand position of detected lighting strikes could be processed to detect one of more lightning strikes of interest. In this case, each lightning strike could be seen as generation of a characteristic signal CS at a source location 102 at a generation time in a similar manner as discussed with respect to figures 1 b-1 c, 2a-b and 3a-b. The sensor output data in a relevant time window based on the generation time could be analysed as discussed above to determine whether the characteristic signal can be detected in the sensor output from one or more sensing portions and, is, the data could be processed to determine location information in a similar manner as discussed above. If there are multiple lightning strikes at different locations at different times, the sensor output data may be processed based on each of the lightning strikes, in a similar manner discussed with respect to figures 3a and 3b.

[0073] In the data driven approach, the sensor output from sensing portions may be analysed to detect any unusual signals, which may include some high amplitude signal component. In the event such an unusual signal component is detected, it may be analysed to determine some properties and to try to classify the signal. If, for example, the signal is classified as being possible detection of a lightning strike, then data regarding any lightning strikes in the general area of the sensing optical fibre may be obtained and analysed to determine if the detected signal does correspond to an characteristic signal generated by a lightning strike.

[0074] It should be noted that as well as generating a mechanical shock wave, a lightning strike on the ground will also cause some localised heating and fibre optic sensing that provides some indication of temperature changes may also provide some additional information for characterising the sensor output in response to a lightning strike.

[0075] Other natural phenomena may also result in large detectable transient signals. For instance, earthquakes or earth tremors may general seismic waves that can travel large distances. Many parts of the world are provided with seismic monitoring equipment that can detect such seismic waves and information regard the propagation of such seismic waves is thus available. In the event of an earthquake or tremor of sufficient magnitude, the data from the sensing optic fibre can be analysed to detect the arrival of the various seismic waves, e.g. the p and s waves. Time of arrival analysis, similar to that discussed with respect to the figures 1a-c, 2a-b and 3a-b (but on larger scale) may be applied and / or the position of the different seismic wavecomponents at different times may be known or can be interpolated based on the seismic monitoring.

[0076] In some cases, the high amplitude transient characteristic signals of interest could be man-made signals for which a time and location of generation is known. For instance, in some urban or industrial applications there may be emergency alarms or the like which may generate a loud noise. Such alarms may be located in known positions and may be tested at defined testing times. If it is known that a particular alarm at a given location is tested at a specific time, this could be taken as an event of interest and the sensor output from the fibre optic sensing in the relevant time window could be analysed as discussed above. Similarly, foghorns or the like may be located in known locations and tested at defined times. Clock towers with bells or similar may also generate louds signals at defined times. Institutions such as schools or colleges may have bells or other similar audio alerts that generate relatively loud sound at defined times in defined locations, e.g. to signify the beginning or end of certain periods.

[0077] Religious buildings may also be a source of acoustic signals, e.g. the call to prayer, church bells etc at defined times.

[0078] These active methods could be used to provide an estimated geographical location of at least one sensing portion. Additionally or alternatively, the method could be used to provide a relative location of one sensing optical fibre with respect to another sensing optical fibre. In some use cases, there may be two sensing optical fibres, each possibly formed as part of, or co-located with, a different asset. For example, power cables may often be laid in pairs a few tens of meters apart from one another. In some cases, the general physical location of the pairs of cables may be known, or could be determined by any of the methods described herein. It may still, however, be desirable to locate a particular sensing fibre with a particular cable. For example, consider there are two power cables that run largely parallel to one another, with a separation of about 10m. Each of the power cables may comprise a sensing optical fibre which may be used for fibre optic sensing, e.g. for health monitoring. If the sensing performed on one of a first one of two sensing optical fibres indicates a potential fault, e.g. acoustic signals or thermal signals that could be indicative of partial discharge, then it may be desirable to inspect and repair as necessary the cable that contains the first sensing optical fibre, at a point that corresponds to the sensing portions of the first sensing optical fibre where the potential fault was detected - but this requires each sensing fibre to be associated correctly with the relevant power cable. A relative location of thesensing fibre can be determined by detecting a characteristic signal which is known to be propagating in a certain direction, e.g. a seismic wave or propagating acoustic wave, and determining which of the two sensing optical fibres detects the characteristic signal first.

[0079] Moving constant sources (e.g. ship crossing)

[0080] In some examples, the characteristic signal of opportunity may correspond to a relatively constant source which moves and where the movement of the constant source is known. For instance, the moving source could correspond to an acoustic signal generated by a vehicle in motion and / or a pressure variation caused by the movement of the vehicle.

[0081] For example, a motor-powered ship or other vessel travelling on the surface of a waterbody may generate a substantially constant sound due to the operation of the engines. Most commercially ships are fitted with locating devices. For instance, the Automatic Identification System (AIS) is an automated tracking system that tracks the position of ships through transponders on the ships and which transmits data regarding ship position. AIS transponders are required to be fitted on a large number of ships and current and / or historic AIS data regarding ship position at a given time is readily available.

[0082] Thus, data regarding the position a ship and how that position changes over time is available. As such, it is possible to track the movement of the ship, which as noted above can thus be seen as a moving acoustic source. The acoustic signals generated by the ship may propagate to, and be detected by, a sensing optical fibre used for fibre optic sensing, e.g. a sensing optical fibre formed as part of an underwater cable. In this case, as the acoustic source is substantially constant, it is not generally possible to identify a specific generation time as such and thus time-of-travel type techniques, based on the travel time of the acoustic signal from the source to the sensing fibre, are not generally applicable. However, the movement of the acoustic source, i.e. ship, with respect to the sensing optical fibre may nevertheless result in a detectable characteristic signal in the sensor output that can be used to determine some location information.For example, figure 4 illustrates the example of a ship 201 travelling in a constant direction with a path 202 that crosses over an underlying underwater fibre optic cable 104. Fibre optic sensing, such as DAS, is performed on a sensing optical fibre of the fibre optic cable 104 as the ship crosses overhead and the data processed to detect, in this example, the acoustic signals generated by the ship, i.e. low frequency content below a certain threshold may be filtered out. Figure 4 illustrates three different positions of the ship along the path 202 and shows an example of the senor output of the fibre optic sensing, illustrating acoustic intensity for each channel, i.e. sensing portion along the length of the sensing optical fibre. Initially, at a first position 203a which is relatively far from the fibre optic cable, acoustic signals generated by the ship may be largely attenuated before reaching the fibre optic cable and thus the sensor output may correspond to any ambient noise. As the ship gets closer to the fibre optic cable, the channels or sensing portions closest to the ship may start to detect the sounds generated by the ship and thus may start to show increase acoustic intensity in the sensor output. At a second position 203b the output from a number of channels may thus show an increased acoustic intensity. As the ship 201 gets closer still the detected intensity will generally increase to a point when the ship is directly overhead, illustrated by the third position 203c.

[0083] The intensity as the ship approaches the cable and then crosses the cable may exhibit a pattern which increases to a maximum at a first point before the ship crosses the cable and which then drops as the ship gets closer to the cable and crosses directly overhead. Once past the crossing point, the intensity may rise again to a maximum at a second (which is a similar distance away from the cable as the first point) and then may drop away as the ship gets further away. Thus, as the ship passes over the fibre optic cable and then starts to move away from the fibre optic cable, the sensor output from the fibre optic sensing would then exhibit a similar evolution but in reverse, and as the ship gets relatively fare away, the detected intensity will drop back to just ambient noise.

[0084] Such a pattern of intensity (for a given channel) can arise due to the sensitivity of the sensing fibre for fibre optic sensing. As noted above, the measurement signal for DAS (and some other types of fibre optic sensing) results from longitudinal strain on the sensing optical fibre to result in a change in optical path length within a given sensing portion. As such, the fibre optic sensing is more sensitive to incident stimuli that result in a longitudinal strain on a sensing portion. For acoustic signals which are longitudinalpressure waves, this results in a variation in sensitivity with direction of incidence of the acoustic wave. Acoustic waves which propagate along the longitudinal axis of the sensing optical fibre (at the relevant sensing portion) may thus directly cause longitudinal strain. However, acoustic waves which propagate in a direction orthogonal to the sensing optical fibre may result in transverse strain. This may result in sideways movement of the sensing optical fibre, but relatively little change in longitudinal length. Thus, the fibre optic sensor may be relatively insensitive the acoustic waves travelling perpendicular to the fibre optic cable.

[0085] The pattern of intensity at a given sensing portion thus depends on relatively attenuation of the acoustic signals reaching that sensing portion and the angle of incidence for such signals, which evolves over time as the ship moves relative to the sensing optical fibre.

[0086] If the ship is travelling in a constant direction and at a constant speed with respect to a sensing optical fibre that is relatively straight (in that area), the evolution of the sensor output from the fibre optic sensing would generally be symmetrical in time, as the ships draws closer, passes over the fibre optic cable, and then gets further away.

[0087] This evolution of the sensor output from the multiple sensing portions of the fibre optic sensor thus is a characteristic signal indicative of a ship travelling above the sensing optic fibre. As noted above, where the ship crosses directly over the sensing optical fibre, there may be a pattern of increased intensity to a local maximum, then a drop in intensity and then another increase to the similar local maximum before dropping back to ambient noise level. In general, all sensing channels in the vicinity of the crossing pint will exhibit a similar profile, but the maximum detected intensity would generally correspond to the sensing portion(s) that lie directly underneath the ship at the crossing point. In addition, multipath reflections etc. may also contribute to the detected intensity for the sensing portion(s) that directly underlie the crossing point.

[0088] Figures 5a and 5b illustrate two examples of data acquired by performing fibre optic sensing on an underwater sensing optical fibre as a ship passed overhead. In the examples of figures 5a and 5b, the data was acquired by performing DAS on a suitable sensing optical fibre formed as part of an underwater cable and filtering out any low-frequency component, so the measurement signal is an indication of acoustic activity. Figures 5a and 5b illustrate waterfall plots of the acoustic intensity detected by each ofa plurality of channels, i.e. sensing portions, of the fibre optic sensor over time. In practice, the acoustic intensity may be represented by colour, but in the black-and-white versions of figure 5a and 5b, a darker shade represents increased intensity.

[0089] The example of figure 5a illustrates an example where the ship travelled along a track that has a shallow angle to the path of the sensing optic fibre, i.e. just off parallel. The general path of the ship and how it tracked along the sensing optic fibre with time can be seen in the sensor output - and this could be correlated with the AIS data to help identify the characteristic signal in the sensor output data and also confirm that the data corresponds to the correct ship. The greatest acoustic intensity in the sensor output, indicated as 501, corresponds to the ship crossing over the sensing optical fibre. The relevant channel or sensing portion SPc which corresponds to the crossing point can be determined, as can the time Tc at which the crossing occurred.

[0090] The example of figure 5b illustrates an example where the ship travelled along a track that was substantially perpendicular to the path of the sensing optical fibre. This results in a different pattern in the sensor output data and the acoustic signal from ship is detected by a smaller number of channels of the fibre optic sensor (as the ship heads directly towards those channels, crosses and then heads away in a generally orthogonal path) but again relevant channel or sensing portion SPc which corresponds to the crossing point can be determined, as can the time Tc at which the crossing occurred.

[0091] In both of the examples of figures 5a and 5b the symmetrical nature of the sensor output with time can be seen. For instance, for figures 5a the symmetric response is clear in the channels around channel 350 or channel 650.

[0092] The discussion with respect to respect to figure 4 and figure 5a and 5b has focussed on detecting acoustic signals generated by a ship. Additionally or alternatively, it is possible to detect a signal due to variations in pressure caused by the movement of the ship, i.e. variations from the normal hydrostatic pressure. Movement of a ship, or other water going vessels including submersible craft, causes displacement of water and thus can lead to a variation in pressure. This variation in pressure can result in pressure variations acting on an underwater cable comprising a sensing optical fibre which can be detected by fibre optic sensing. Such variations in water pressure extendfor some distance into the seabed and thus even underwater cables buried in the seabed may experience such a pressure variation.

[0093] For sensing such as DAS, such a pressure variation may appear in the sensor output as a relatively low-frequency component (such as may have been excluded by filtering from the example data of figures 5a and 5b). Some embodiments may, however, specifically look for a measurement signal corresponding to a pressure variation as a characteristic signal.

[0094] The pressure variation caused by a ship approaching and crossing a sensing optical fibre may, again, exhibit a particular pattern. As the ships approaches the sensing optical fibre, the movement of the ship is effectively pushing water out of its way, with the result of creating a pressure bow wave extending from the ship. Thus, as the ship approaches the sensing optical fibre, it effectively pushes more water towards the sensing optical fibre increasing the pressure on the sensing optical fibre. As the ship reaches a position overhead the sensing optical fibre, the pressure bow wave will have passed and the pressure acting on the sensing optical fibre will decrease - which can result in an effective suction force acting on the seabed. Once the ship passed the sensing optical fibre, water will rush to fill the gap left behind the ship and the pressure may thus rise again, until the pressure dissipates and normal hydrostatic pressure (subject to wind / waves etc.) returns. This again creates a characteristic pattern in the measurement signal from an affected sensing portion, and the time of the ship crossing can be determined from the time of the local minimum between the two local maxima.

[0095] Figures 6a -6d shows example data collected from four different vessel crossing above an underwater sensing optical fibre. Each of the plots 6a-6d shows, for a single sensing portion, an indication of determined pressure - based on the low pressure component from fibre optical sensing - against time as the vessel passed overhead. In each case the rise in pressure as the vessel approaches, the drop in pressure when the vessel crosses, and the subsequent rise in pressure can be clearly seen. From this, the time of crossing can be clearly determined. This pressure variation will be greatest for the sensing portions that the ships passes directly over.

[0096] Figure 6e illustrates a two-dimensional map / waterwall plot of the data acquired from a plurality of sensing portions extending over several kilometres of the sensing optical fibre and illustrates the pressure detected with time. It can be seen that each sensingportion experiences a relatively steady cyclic variation in pressure, which is due to surface waves varying the pressure experienced by the sensing optical fibre. However, there is a very pressure variation, which a characteristic increase, decrease and increase due to a ship crossing (in the centre of the map / waterfall plot), which is clearly distinguishable from the normal background variation and which can be clearly identified as being a ship (or other vessel) crossing.

[0097] Thus, using acoustic and / or pressure date, the sensor output data from performing fibre optic sensing can be analysed to detect a characteristic signal corresponding to a ship moving above the underwater sensing optical fibre. As noted above, the position of a ship over time may be available due to the AIS data or similar. Thus, the position of a ship over time can be determined from AIS data, or similar, and correlated with the sensor output from performing fibre optic sensing. In particular, where the ship crosses over the sensing optical fibre, sensor output from the fibre optic sensing can be analysed to detect the relevant sensing portion(s) or channel(s) of the sensing optical fibre that are at the crossing point and the relevant time at which the ship crosses the fibre. The AIS or other ship positioning data can then be used to determine the ship position at the relevant time (possibly taking the expected propagation time for the acoustic signal to propagate from the ship to the sensing optical fibre into account) to thus determine the location of the relevant sensing portion(s).

[0098] Over time, with multiple crossings by different ships at different times and in different locations, the geographical location of a plurality of different sensing portions of the sensing optical fibre can be determined by locating the crossing points and the path of the fibre optic cable 104 can be interpolated between the location of these known sensing portions.

[0099] In some implementations, the data may be analysed to detect relevant patterns associated with a vessel passing overhead or near a sensing optical fibre (depending on the whether the data comprise acoustic and / or pressure data). In some implementations, the data may be analysed using a classifier or similar to determine whether or not it corresponds to movement of a vessel. The classifier can be trained on historic data which is known to correspond to vessel movement. New data which is identified as resulting from vessel movement data may be analysed to determine some location information.In some cases, the data may be processed or analysed using at least one model which is determined by machine learning.

[0100] For instance, in one example, a data set which corresponds to sensor output data acquired from performing fibre optic sensing fibre in-situ may be performed. In some cases, the fibre optic sensing may be performed on the relevant sensing fibre which it is wished to locate, when at least some positions of the sensing optical fibre are known to a reasonable accuracy. The data set may, for instances, correspond to sensor data acquired over a period of several days or months. The known positions of the sensing optical fibre may be determined using the pressure or acoustic data to identify a point of crossing as discussed above.

[0101] For the application to using ships movement, AIS data regarding movement of ships in the general location of the sensing optical fibre may be obtained. This data can be taken as a ground truth and used to label each location along the sensing cable at each timestep in the dataset to say whether there was a vessel close to it or not at that point. In some cases, information such as the distance of the vessel to the position along the sensing optical fibre and / or the size of the vessel may be included.

[0102] Data snippets from the data set corresponding to times and locations when a vessel was close to the cable (say within 1 km in one example) may then be extracted, along with corresponding data snippets from times when there were no vessels close to the cable. The number of data snippets may be, in some examples, in the thousands.

[0103] The data snippets may, in some examples, may processed to provide some indication or summary of at least some key data content, for instance by generating statistical summaries and / or performing gradient mapping.

[0104] The data snippets and / or the key data content may then be fed into a model builder which then produces a model. The model builder may be configured to test a range of model types, including, for instance, random forests and neural networks and then selects the one or more models that best match the training data. Hyperparameter tuning may be conducted where relevant. The (or each) resultant model is then able to take data it hasn’t seen and determine whether it contains content generated by a vessel not. If enough data is provided, the model should also be able to determine information such as the distance of the vessel that made the noise from the cable.Applying this model on new data can provide an estimate of the distance of the vessel from multiple points along the sensing optical fibre which allows triangulation of its position. Doing this at multiple distinct timesteps allows plotting of the vessel path and, therefore, identify when and where the lateral offset between the vessel and the cable is zero - which thus locates the position of the sensing optical fibre.

[0105] It should be noted that machine learning, e.g. training a classifier or model to detect the relevant characteristic signal in the sensor output and / or determine location could be applied to any of the embodiments of the passive method of location.

[0106] The discussion above has focussed on the acoustic signals or pressure variation from a single ship. In at least some use cases there may well be times when only a single ship may be travelling the in the vicinity of the sensing optical fibre at a given time. Thus identify the relevant AIS or other positional data from the ship may be straightforward and it can be assumed that any activity in the sensor output data at the relevant time corresponds to that ship. In some cases, two or more ships could be travelling relatively closely to each other and there may be a need to discriminate between the signals from the two or more ships. In some cases, it may be able to discriminate the signals from the two ships based on the movement of the ships themselves, i.e. if the two ships are heading in different directions and the response can be tracked along the sensing fibre, it may be possible to determine which ships the relevant detected signals correspond to. In some cases, some analysis of the detected signals may be used to discriminate between the ships, for instance the properties of the acoustic signals detected from a tanker or cargo ship may be sufficiently different to those generated by a passenger ferry or fishing trawler or the like.

[0107] The ability to detect vessels passing by or crossing over an optical fibre by analysing signals detected by performing fibre optical sensing on the optical fibre represents a novel aspect of this disclosure, as does using positional information known about the vessel, such as AIS data, to locate the geographical position of the optical fibre.

[0108] Whilst the discussion above has focussed on detecting the signals generated by a ship travelling above an underwater cable, which may be the acoustic signals, the same general principles would apply to any moving constant acoustic source in which the position of the acoustic source over time is known. For instance, thus could apply toother types of vehicles, such as aircraft flying at a relatively low altitude over onshore fibre optic cables, e.g. such a cable buried in the ground, or ground-based vehicles in which the position of the vehicle is known or can be tracked by some other means.

[0109] Operation of fixed infrastructure

[0110] In some examples, the characteristic signal of interest may arise due to the operation of some fixed infrastructure. In these examples, the fixed infrastructure may not, itself, directly generate a detectable signal of interest, but operation of the infrastructure may lead to some detectable signal of interest.

[0111] For instance, the fixed infrastructure could comprise some traffic flow control apparatus, such a set of traffic lights. Operation of the traffic lights controls traffic flow and the traffic flow will generate traffic noise. The resulting traffic noise may be detected and, in particular, the pattern of the traffic flow noise may depend on the operation of the lights and the timing or phasing of the operation of the lights. The position of the traffic lights is fixed and can be known and, in at least some instances, the timing of operation of the lights may be set or controlled by some central control. Information about the timing of the operation of the lights may be available and can be correlated with sensor output data collected from performing fibre optic sensing on a sensing optic fibre.

[0112] Figure 7 illustrates one example of this principle. Figure 7 illustrates a road or highway 701 with a first intersection 702a, where traffic flow over the first intersection is controlled by a first set of traffic lights 703a.

[0113] It may be known that a fibre optic cable 104 is deployed somewhere in the area, but the path of the fibre optic cable 104 may not be known, e.g. because the fibre optic cable is buried or in some underground conduit. If, however, one end of the fibre optic cable 104 is accessible, so that fibre optic sensing can be performed on a sensing optical fibre within the fibre optic cable, then fibre optic sensing may be performed and the sensor output could be analysed with respect to a known timing of operation of the first set of traffic lights. For instance, the first set of traffic lights may be controlled to permit traffic to pass in one direction, say the left-right directions as illustrated, in first time periods and to permit traffic permit traffic to pass in a different direction, say the up down direction as illustrated in second time periods. The sensor output data from thefibre optic sensing may be analysed with knowledge of the timing and relative durations of the first and second periods.

[0114] For instance, consider the example illustrated in figure 7 where the fibre optic cable 104 runs generally along the same path as the road in the first direction. During a period when the traffic lights are allowing traffic to flow in the first direction, there may be relatively continuous traffic noise detected at all of the sensing portions along the path of the road. When the light change to amber and red, to stop traffic flow in this direction the traffic noise will cease as the traffic flow comes to a halt, and this will generally stop first at the location of the intersection as the vehicles closest to the intersection stop, and then this cessation of traffic noise will spread along the sensing portions of the sensing fibre as vehicles join the queue of traffic at the lights.

[0115] The lights will then go green at the start of the second period, to allow traffic to flow in the other direction. In this case traffic noise will start to be detected by the sensing portions in the vicinity of the intersection, and some quite intense signals may be detected by the sensing portions that pass under part of the intersection road.

[0116] However, if the predominant or only traffic flow allowed is in the second direction (e.g. up-down as illustrated), then sensing portions that are spaced relatively away from the intersection in the first direction will not detect much significant traffic noise and the traffic in that direction will be standing still due to the red light in the that direction. At the end of the second period, when the lights turn back to amber and red, the traffic noise will again stop as the traffic flow stops. At the start of the next first period, when the lights go green to allow traffic flow in the first direction, the traffic will start to flow, from the front of the queue, and thus traffic noise will again start to be detected along from multiple sensing portions along the first direction, starting at the sensing portions closest to the intersection.

[0117] In this way, by analysing the sensor output from the fibre optic sensing, information about the position of the sensing portion of the sensing optical fibre with regard to the road can be determined due to the known operation of the firsts et of traffic lights.

[0118] Figure 7 illustrates that there may be a second intersection 702b which is controlled by a second set of traffic lights 703b. The relative timing or phasing of the operation of the second set of traffic lights may be different to that of the first set of traffic lights, and this difference in timing or phasing can be used to identify and distinguish acoustic signalsarising from the flow control of the first set of traffic lights from acoustic signals arising from the flow control of the first set of traffic lights.

[0119] It will be understood that the discussion of figure 7 illustrates just one example of how traffic flow noise, as controlled by one or more set of traffic lights, could be analysed to determined location information and there will be other ways that the data could be analysed depending on a particular raid layout. Also, whilst the operation of traffic lights has been described, other similar analysis could be based on pedestrian crossings (if timing data of the operation of the crossing is available) or rail crossings or the like.

[0120] Other Stimuli

[0121] The discussion above has focussed mainly on fibre optic sensing to detect any propagating mechanical pressure or similar wave that may induce strain or vibration of the sensing optical fibre. As noted previously, fibre optic sensing may additionally or alternatively respond to some other environmental stimuli such as temperature changes. In some cases, the signals of interest comprise some other stimuli acting on the sensing optical fibre, either in addition to or instead of an acoustic signal of interest.

[0122] For instance, as discussed above, lightning strikes may provide some localised heating. In some cases, operation of fixed infrastructure could result in some pattern of heating or cooling, there may cause localised temperature changes of one or mor sensing portions of a sensing optical fibre in the vicinity of the fixed infrastructure.

[0123] Solar heating may cause temperature changes and, in some applications the time at which solar heating occurs (or is blocked by some intervening object) may be function of the position of the sun in the sky and the location of the sensing optical fibre. In some cases, given the times at which a sensing portion of sensing optical fibre exhibits temperature changes that are attributable to the location of the sensing fibre being in direct sunlight or in shadow, it may be possible to determine the location of the relevant sensing portion.

[0124] A thermal stimulus may also propagate through a medium and the time of arrival of a thermal stimulus at a given sensing portion of optical fibre may give an indication of the amount of medium the stimulus has to travel through. For instance, solar heating ofthe ground will warm the surface of the ground first and this thermal stimulus will gradually progress deeper into the ground. If it is known that a particular area is directly exposed to the sun at a particular time of day, the time at which the heating is registered by a buried sensing optical fibre may given an indication of the depth of the sensing optical fibre. In some cases, this could be correlated with some weather data, or data from solar power generation, to estimate the radiant heating imparted to the ground. The propagation of a thermal stimulus could, in particular, be used to provide an indication of relative position of two sensing optical fibres. For instance, if two sensing fibres are buried at different depths, the sensing optical fibre nearest the surface would be expected to register a response to solar heating first.

[0125] Similar principles could be applied where a pipeline or similar is used at certain times for the transport of fluids which may be significantly above or below ambient temperature and thus which have a heating or cooling effect on their surrounding environment.

[0126] In general, therefore, at least some embodiments relates to methods and apparatus for passive location of at least part of sensing optical fibre, on which fibre optic sensing is or was performed. The passive location may make use of characteristic signals of opportunity which are not specifically generated for the purposes of locating the sensing optical fibre. The analysis may be event driven, based on determining that an event of interest occurred in the general vicinity of the sensing optical fibre and analysing the sensor output data to try to detect the characteristic signals resulting from the event, and / or analysis may be data driven, in which signals of possible interest may be detected in the sensor output data and analysed to categorise the type of signal, to allow for correlation with a possible event of interest.

[0127] It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims. Features from various embodiments may be combined and used together except where expressly indicated otherwise. The word “comprising” does not exclude the presence of elements or steps other than those listed in a claim, “a” or “an” does not exclude a plurality, and a single feature or other unit may fulfil the functions of several units recited in the claims. Any reference numerals or labels in the claims shall not be construed so as to limit their scope.

Claims

CLAIMS1. A method of determining a location of at least part of a first optical fibre comprising:taking a data set comprising sensor output data acquired by performing fibre optic sensing on the first optical fibre, wherein the data set comprises sensor output data from each of a plurality of sensing portions of the first optical fibreanalysing the data set to detect at least one characteristic signal, wherein the characteristic signal corresponds to the detected response to an event which is not caused to occur for the purpose of locating the first optical fibre;determining a first known location associated with the characteristic signal; and determining a location of at least one sensing portion of the first optical fibre based on the characteristic signal and the first known location.

2. The method of claim 1 wherein the characteristic signal comprises a transient characteristic signal.

3. The method of claim 2 wherein:determining the first known location associated with the characteristic signal comprises determining a source location of the transient characteristic signal;the method further comprises determining a generation time of the transient characteristic signal at the source location; anddetermining the location of the at least one sensing portion of the first optical fibre comprises determining a time difference between the generation time of the transient characteristic signal at the source location and the time of detection of the characteristic signal at the at least one sensing portion.

4. The method of claim 2 or claim 3 wherein the transient characteristic signal is generated by a lightning strike.

5. The method of claim 4 wherein the determining the first known location associated with the characteristic signal and determining the generation time ofthe transient characteristic signal at the source location comprises determining the time and location of the lightning strike from a lightning tracker data set.

6. The method of claim 2 wherein the transient characteristic signal is generated by an earthquake or earth tremor.

7. The method of claim 6 wherein determining the first known location associated with the characteristic signal comprises determining at least one of a source location of the earthquake or earth tremor and / or a position of at least one seismic wave generated by the earthquake or earth tremor.

8. The method of claim 1 wherein the characteristic signal comprises a characteristic signal generated by a moving source.

9. The method of claim 8 wherein the moving source comprises a motor power vehicle.

10. The method of claim 8 or claim 9 where the first optical fibre is part of an underwater cable and the moving source comprises a water going vessel.

11. The method of claim 10 wherein the characteristic signal comprises at least one of a characteristic acoustic signal generated by the water going vessel and a characteristic pressure variation caused by movement of the water going vessel through the water.

12. The method of claim 11 wherein determining the first known location associated with the characteristic signal comprises determining the position of the water going vessel over time.

13. The method of claim 12 wherein the water going vessel is a ship and determining the position of the ship over time comprises interrogating a data set of automatic identification system data.

14. The method of any of claims 10 to 12 wherein the determining the location of the at least one sensing portion of the first optical fibre comprises determining at least a first sensing portion as a sensing portion under a crossing point of thewater going vessel over the first optical fibre and wherein determining the location of the at least one sensing portion of the first optical fibre comprises determining a crossing time of the water going vessel over the first optical fibre and identifying the location of the at least a first sensing portion based on the position of the water going vessel at the crossing time.

15. The method of claim 10 wherein determining the at least a first sensing portion as a sensing portion under a crossing point of the water going vessel over the first optical fibre comprises analysing sensor output data using a model trained by machine learning.

16. The method of claim 1 wherein the characteristic signal comprises an acoustic signal generated by operation of some fixed infrastructure.

17. The method of claim 14 wherein the fixed infrastructure comprise a traffic control apparatus and the characteristic signal comprises an acoustic signal generated by traffic flow in response to operation of traffic control apparatus.

18. The method of claim 1 wherein the characteristic signal comprises a thermal stimulus.

19. An apparatus for determining a location of at least part of a first optical fibre comprising a processor configured to:take a data set comprising sensor output data acquired by performing fibre optic sensing on the first optical fibre, wherein the data set comprises sensor output data from each of a plurality of sensing portions of the first optical fibreprocess the data set to:detect at least one characteristic signal, wherein the characteristic signal corresponds to the detected response to an event which is not caused to occur for the purpose of locating the first optical fibre;determine a first known location associated with the characteristic signal; and determine a location of at least one sensing portion of the first optical fibre based on the characteristic signal and the first known location.A computer readable medium comprising computer readable code that, when run on a suitable processor, cause the processor to perform the method of any of claims 1 to 18.

Images