GB2598941A - Detection system - Google Patents

Abstract

A detection system for monitoring perimeters and railway tracks, the system comprising a first surface waveguide structure having first 5a and second 5b transceivers with elongate surface waveguide 4, such as a wire, therebetween. Signal generator 31a, 31b injects a signal into a transceiver to produce a surface wave along the surface wave guide. Signal processor 32a, 32b obtain a signal received at the transceiver. Signal analyser 33 produces a detected signature corresponding to objects in proximity to the surface waveguide based on the received signal and the detected signature is compared to one or more reference signatures. The reference signature may include information corresponding to amplitude, time of flight, or phase. The received signal may result from transmission or reflection of the produced signal from objects in proximity to the surface waveguide. The detected signature may be compared to a plurality of reference signatures stored in a library to provide information on environmental conditions.

Description

DETECTION SYSTEM

This invention relates to a detection system for monitoring objects in proximity to a detector and in particular for monitoring long sections of land such as perimeters, and railway tracks.

Background of the Invention

There are many ways in which security of restricted areas can be managed. Challenges arise where the area to be secured is to be unmanned or it is impractical to provide permanent security on site. One particular area where this challenge exists is along the edges of large areas with long perimeters such as airports, international borders and in particular railway tracks. Railways tend to include regions which are extremely long but relatively narrow and which are often remote. The nature of railways is that they are particularly hazardous areas for trespassers. Trespassers may have different motivations such as theft of valuable materials such as copper cables but also children may enter railway properties out of curiosity. It is therefore important to be able to try to minimise and deal with any trespassers that may access a railway line or the immediate surroundings, to both avoid injury to the trespassers but also to prevent theft or damage to railway infrastructure.

Traditional security was generally provided by having physical barriers to try to prevent access to the railway. Simply maintaining fences along a track is time consuming and expensive and vulnerable to persistent intruders, who may simply break through or climb over a fence.

Other methods include providing CCTV cameras for monitoring sections of the track. This may also include other techniques such as having motion detection devices along with the physical security such as fences, to limit or deter trespassers. Where the area to be secured is potentially several miles long, as with a railway, one camera can only observe a small section in detail and so many cameras may be needed to cover even relatively short lengths of track. Therefore additional measures are often needed particularly at sites which are more vulnerable.

CCTV cameras can provide good coverage of relatively small areas of track but to cover long sections will require many cameras and each camera will need to be monitored either by a human security person or using image analysis techniques to identify trespassers.

Infra-red, lidar or radar detection can also be utilised but these have drawbacks in that they tend to have limited range and are not able to see around corners and so may need repeated stations for monitoring sections of track.

Other options include using acoustic sensing which can be basic microphones detecting sound to more sophisticated acoustic sensing techniques such as using optical fibres to detect the vibrations from sound impinging on the optical fibre. However, this is limited to relatively close range and the noise of trespassers may be lost in the background sounds from wind, rain and other environmental noise. It is also restricted to detecting active noise, so a stationary object placed on a railway line will make no sound and will not be detected. Similarly a trespasser who remains silent may be overlooked.

Many of these techniques require expensive equipment which is often expensive to install and maintain.

In addition to trespassers on the line, there may be other objects which may be moving in the vicinity of a railway tracks such as animals and foliage. Some of these movements maybe of little concern, such as small rodents or foliage swaying in the wind. This can lead to false alarms from simple motion detection systems. However, larger animals may represent a problem, for example life stock that has escaped onto a railway track, which clearly does represent an issue. Other issues that may represent problems for operating a railway might include trees or branches that have fallen onto the line, landslips or rock falls may also obstruct the railway as well as other objects that may inadvertently end up on a railway line posing a threat to themselves or passing trains.

There is therefore a need to provide security along long stretches of railway which is less expensive than installing and maintaining the infrastructure for multiple cameras or other motion sensing devices without the need for additional security personnel and or additional complex and more expensive infrastructure.

The present invention therefore aims to improve on or ameliorate the problems described above to provide a system for monitoring perimeters, particularly long perimeters, as simply and efficiently as possible.

Summary of the Invention

Therefore, according to the present invention there is provided a detection system comprising: a first surface wave guide structure having a first transceiver, a second transceiver, and an elongate surface wave guide arranged between said first transceiver and said second transceiver; a signal generator for injecting a signal into said second transceiver, to produce a surface wave along said surface wave guide; a signal processor for obtaining a first received signal received at the first transceiver, and a signal analyser for producing a detected signature corresponding to objects in proximity to the conductor, based on the first received signal from the signal processor; and a comparator arranged to compare the detected signature to one or more reference signatures.

This arrangement allows changes in the environment around the surface wave guide structure to be detected and analysed by comparing the detected signature to one or more reference signatures to identify a match and obtain an indication of the current environment which may include elements such as persons in it.

The reference signature may include information corresponding to at least one of: amplitude; time of flight; phase, and the profile of these parameters over time. By utilising some or all of these elements of the detected signals, different aspects of the monitored environment can be determined allowing detailed information about location, nature and proximity of objects in proximity to the line.

The signal processor is preferably further arranged to obtain a second received signal, received at the second transceiver, wherein said detected signature is also based on the second received signal from the signal processor. By utilising the second signal, additional information can be obtained about the environment around the detector. This allows more detailed information to be obtained which would include both transmission and reflected signals affected by interactions with objects that may be present.

The signal generator may further be arranged for injecting a signal into said first transceiver, to produce a surface wave along said conductor, said signal processor is arranged for obtaining a third received signal received at the second transceiver and a fourth received signal received at the first transceiver; and said detected signature is also based on the third received signal and the fourth received signal from the signal processor.

In this way, yet further information is obtained about the environment around the detector system. By combining the four signals together, a more detailed analysis of the objects in the proximity of the wire can be obtained.

The detection system may include one or more additional surface wave guide structures similar to said first surface wave guide structure, wherein the signal analyser also analyses the signal from each of the additional surface wave guide structures to produce said detected signature. By providing additional surface wave guide structures, using different wires which are spaced apart from the first wire yet more information can be obtained about the environment which will be different to that obtained from the first wire due to the spatial separation of the wires. By arranging the wires at spaced apart intervals, each wire will collect information from a slightly different perspective allowing it to better differentiate objects compared including obtaining improved directional information as moving objects approaching the detection system will impact each wire differently.

The signal generator is preferably arranged to inject signals of more than one frequency. By using different frequencies, the different information available at each frequency can be utilised to further improve the quality of the signature obtained and thereby obtain a better interpretation of the environment around the detector. Signals of different frequencies can be injected onto the wire and the respective detected signals can then be combined to provide an improved detected signature. The frequency may be varied by repeatedly injecting signals of different frequencies sequentially or the signal frequency may be scanned across a range of frequencies. Alternatively a multiband or broadband signal maybe injected onto the line which includes multiple frequencies simultaneously.

Preferably at least one of said reference signatures is produced based on a known environment around the surface wave guide. By producing reference signatures based on a known environment, the known environment can be correlated to the signatures to help to produce a library of signatures which once matched to the detected signature can be used to provide an indication of whatever environment that signature corresponds to.

At least one of the reference signatures may be determined by obtaining a detected signature and storing it along with information corresponding to the environment around the detection system when the detected signature is obtained.

A plurality of said reference signatures can be stored in a library and the detected signature compared to the reference signatures stored in a library to identify the best matching reference signature. By matching the detected signature to one of a plurality of reference signatures in a library, the closest matching signature can be identified to provide the best suggestion of the current environment around the detector which may correspond to a person in proximity to the line or some other variation such as a train passing or a tree fallen on the line.

Preferably, a plurality of said reference signatures correspond to the same physical environment under different variable environmental conditions and wherein said comparator compares the detected signature to said plurality of said reference signatures to identify the closest matching reference signature to provide information on the environmental conditions. This allows the system to assess the current environmental conditions such as the current weather. By obtaining an indication of the current weather, the information can be used to provide weather information but also to provide improved accuracy for determinations of the best matching signature and future. For example by determining that the weather is raining, preference can be given to those reference signatures which relate to rain and exclude reference signatures which relate to dry conditions. In that way improved accuracy of characterising the environment around the detector is provided.

The comparator preferably monitors how the closest reference signature to the detected signature changes over time for tracking the movement of objects around the surface wave guide structure. By monitoring the environment and identifying objects in the proximity of the detector and then seeing how the detected objects change over time, allows the system to track the movement of objects further aiding the detection of certain types of events such as persons alongside the track or other parameters being monitored.

The elongate surface wave guide may be formed in a number of ways such as a conductive wire with an insulator or a plurality of wound wires. The waveguide preferably has a

S

hydrophobic coating. The hydrophobic coating helps to prevent water collecting on the guide which may have a substantial and potentially overwhelming effect on the detection of other objects. For example, a drop of water hanging on the guide may have a significantly greater effect on the detected signal than a person a few metres away from the guide. By providing a hydrophobic coating the accumulation of water on the surface wave guide can be minimised.

The signal generator may include one or more separate signal generators for injecting a signal into one or more of the transceivers in the system. Although a single signal processor may be used to provide all the signals to be injected onto the guide or guides, it may be more convenient or effective to have separate signal generators for each guide and/or each transceiver. For example, where the transceivers are arranged at each end of the wire, it may be better to have a signal generator local to the transceiver to avoid having a single signal generator at the far end of the wire that requires the signal to be somehow communicated to the transceiver at the far end.

The signal processor may include one or more separate signal processors for obtaining received signals from one or more or said first, second, third, fourth and any additional transceivers. In a similar way to the signal generators, the signals obtained from the transceivers may be processed by a single processor or a plurality of processors, for example with a signal processor for each transceiver. Where a single processor is used, the signals may be sent separately at different time intervals or the signals maybe detected using other techniques such as multiplexers to repeatedly analyse the signals from each of the transceivers effectively simultaneously. However, it may be more convenient to have separate signal processors at each end of the wave guide, so that signals can be received and processed at the respective ends of the wave guide, rather than having to communicate the unprocessed signals back along the length of the wave guide.

Signals received by at least one of said transceivers may be communicated from one end of the wire to the other along the wire using a surface wave on the wire. Surface waves are capable of carrying information and so by injecting a signal into one end of the surface wave guide to be received at the other end, the signal can be arranged to carry communication information from one end of the guide to the other. In this way the signals detected by the transceiver at one end of the guide can be sent along the guide back to the other end so that the signals can be processed together in a single unit at one end of the guide. Furthermore, the guide may be used for carrying other communications, either in relation to the operation of the detection system or as a way of carrying other communication signals unrelated to the detection system.

The present invention also provides a method for monitoring a detection zone around a first surface wave guide structure having: a first transceiver; a second transceiver; and an elongate surface wave guide arranged between said first transceiver and said second transceiver, the method comprising: injecting a signal into said second transceiver, to produce a surface wave along said conductor; obtaining a first received signal received at the first transceiver, and analysing the first received signal to produce a detected signature corresponding to objects in proximity to the surface wave guide; and comparing the detected signature to one or more reference signatures.

Brief Description of the Drawings

The present invention will now be described in more detail by reference to the attached drawings in which: Figure 1 shows a schematic layout of the system of the present invention; Figure 2 shows a typical arrangement of a detector system according to the present invention; Figure 3 shows a schematic layout of the detector system; Figure 4 shows a modified version of the detector system provided with a second line; and Figure 5 shows the schematic layout of the detector system of Figure 4.

Detailed Description

Figure 1 shows a typical layout of a detector system according to the present invention. In this example, the system is being used along a railway line but as noted above, it may be used in other applications. The detector system 1 is configured as an elongate line 4 extending longitudinally along a section of railway comprising one or more tracks 2a, 2b arranged generally in parallel with each other. The detector line 4 generally follows the path of the railway to form a detection region along a perimeter. Although Figure 1 only

B

shows one detector system 1, a similar detector system may be arranged on the other side of the railway to monitor the other perimeter. In this way any incursion onto the railway from either side can be detected.

Other configurations of the detector may also be used by including additional detector systems for improving sensitivity across the full width of the railway corridor.

Figure 2 shows a more detailed view of the detective system of the present invention. The detector line is formed by a wire 4 which is arranged in an elevated position extending along the length of the region to be monitored, to provide a detection area surrounding the wire 4. The wire is preferably maintained in a tensioned state to minimise any sag in the wire between the supports 22. This also helps to avoid any movement of the wire due to air flow caused by wind or passing trains etc. For shorter lengths or where the tension in the wire can be sufficient to prevent excessive sagging, the intermediate supports 22 may not be required.

At each end of the wire there is termination where the signal is fed or launched onto the line and recovered from it. In this embodiment that is provided by horns 20a, 20b in which the wires are terminated. The wire and horns (acting as a launcher and catcher respectively) form an electromagnetic surface wave line, sometimes referred to as a Goubau line. The horns 20a, 20b are connected to an input source for generating an electromagnetic field which propagates along the line in the form of a surface wave. The wire 4 is provided with a dielectric coating which causes a slight reduction in the phase velocity to a speed slightly less than the speed of light, to constrain the electromagnetic radiation around the surface of the wire. This means the wave only propagates in the direction of the transmission line's axis. By constraining the electromagnetic wave such that the energy primarily propagates along the axis of the wire 4, allowing it to act as an electromagnetic surface wave guide. The attenuation of the signal along the line is very low so that the signal can be transmitted over considerable distances without great loss of signal strength.

The containment of the electromagnetic wave means that there is very little transmission of energy away from the line which is also helpful in avoiding interference to other equipment which may be sensitive to high frequency signals.

The electromagnetic surface wave propagates along the line from the launcher, horn 20a, at one end and is terminated in catcher, horn 20b at the other end. The field from the surface wave formed on the line extends radially around the wire to define a region which forms the detection zone. Objects within the detection zone will influence the wave which allows such objects to be detected and monitored. The proximity of such objects to the wire will affect the surface wave allowing identification, location and tracking of such objects in proximity to the line.

The detection zone around line is sensitive to objects in a region radially around the line 4. This includes anything below the line as well as above or to the side. As a result, it is desirable to raise the wire above the ground, to prevent the ground undesirably influencing the detection by the wire or attenuating signals detected as a result of objects above or to the side of the line. The wire is therefore preferably suspended a distance above the ground such that the ground is not in the part of the detection region closest to the wire, although it may still be within the normal detection range. The wire 4 is typically raised between 50 centimetres and 150 centimetres above the ground. The height is preferably around the magnitude of one wavelength. However, different wavelengths may be used and so the height is preferably selected based on the longest wavelength used in the system to avoid higher attenuation of those longer wavelength signals. However, other limitations may restrict the height or require it to be higher.

In the environment of a railway line, it is convenient to have a small raised wire along the railway corridor which would not otherwise hinder operation of the railway. Furthermore, fences are often placed along the edges of the corridor to physically prevent trespassers entering the railway area. Such fences can be and often are formed of a series of multiple strands of wire arranged one above the other and supported on posts arranged periodically along the direction of the fence. One of these wires may be utilised for the detector wire 4 or a detector wire 4 maybe mounted on the fence posts in addition to the existing wire. This means the system can be easily implemented in a railway environment without unnecessary infrastructure works and with relatively low cost, as a simple insulated wire is all that is required.

Whilst surface waves tend to propagate better along straight wires, they can be directed to travel round corners where the wire 4 is bent, as shown in Figure 1. Very sharp bends in the wire will tend to cause significant losses, but more gentle bends such as those that might be found on a railway track where the pitch is very low, tend to have relatively low loss of signal along the surface wave line. As such, the lines can be arranged to follow the track as it traverses corners with little modification other than to modify the direction of the wire. Alternatively, where larger changes in direction are required, surface waves can be reflected and deflected using known techniques. However, in general, simply changing the direction of the wire is sufficient to follow the track without significant impact on the attenuation of the signal passing along the wire 4.

In use, the horns 20a, 20b are attached to a respective transceiver unit 5a, 5b at each end of the wire 4. Depending on the configuration, the transceiver units are arranged for producing a high frequency signal to be fed onto the wire 4 through the horns 20a, 20b. The transceiver may also include a receiver for detecting a surface wave propagated along the line from a horn at the other end. The transceiver unit preferably includes a diplexer unit for separating the outgoing transmitted signal and any incoming signal for detection.

The detector 1 may be configured in a number of different ways. In its simplest form, one of the horns 20a acts as a transmitter feeding the high frequency signal onto the wire so that it propagates along the length of the wire and is received by the horn 20b at the other end. When the signal is fed onto the line, it propagates along the line and is affected by objects which are proximal to the line. For example, a person 8 standing close to the wire 4 will influence the surface wave passing along the line. The presence of the person will affect the transmitted signal received by the horn at the other end of the wire 4 and will also cause a signal to be reflected back towards the transmitting horn 20a which can detect the reflected signal. By having a receiver with the transmitter in the horn 20a, this reflected signal can also be detected. Similarly, the horn 20b may include a transmitter to allow as signal to be transmitted from it to the first horn 20a. Again, objects such as a person 8 alongside the detector will affect the signal transmitted to horn 20a and will also reflect a signal back to the transmitting horn 20b.

Figure 3 shows the basic lay out of a detection system according to the invention. A control system 30 is connected to the transceivers 5a, 5b for generating and receiving the signals passed along the wire 4. The control system 30 includes a signal generator 31a for providing the high frequency signal to the transceiver 5a. Similarly, a separate signal generator 31b generates the signal provided to the transceiver 5b. The signals received at the transceivers are also fed to the control system 30 and passed to respective target signal processing units 32a, 32b. The transceivers may include diplexers to allow signals to be put onto the line and signals extracted to be from the line without interfering with each other. Other devices may be used such as a circulator.

The target signal processing units 32a, 32b process the received signals and pass them back to a main processing unit 33 which analyses the signals received from each of the transceivers 5a, 5b. The main processing unit 33 therefore receives signals received at transceivers 5a, 5b resulting from the signal put on to the wire 4 from transceiver 5a and also receives signals received at transceivers 5a, 5b resulting from the signal put onto the wire 4 from transceiver 5b as a result of either transmission or reflection of the signal from any objects in proximity to the wire 4.

In this embodiment, the four different signals collected from the transceivers allow a more complete picture of the environment around the wire to be developed. The signals are analysed to determine the time of flight, the difference in time of arrival as well as the amplitude and phase profile of the received signals, to determine the type and location of objects along the line as well as their relative size and proximity to the wire 4. Observation of the amplitude profile over time allows the effect of different objects at different positions to be assessed. The phase profile helps to quantify objects in proximity to the wire. The phase profile is obtained by comparing the received signals to a reference signal, such as the input signal to obtain phase information.

Simple analysis of these signals alone can provide an indication of objects moving around the wire and provide an indication of their relative position and nature although, as described in more detail below, more detailed information can be derived by comparing the signals to reference from previous measurements. Thus by comparing the received signals with previous and subsequent information, changes in the environment around the wire 4 can be detected.

Analysis of the signals received over time, both in terms of the amplitude and profile of the signals received and the timing of the signals, allows determination of the location of objects along the length of the line as well as information about their impact on the signal. Larger objects will tend to have a larger impact on the signal and objects that are closer to the wire 4 will also affect the signal more than objects further away. However, other factors will affect the signal such as how reflective of electromagnetic waves the object is and how much it absorbs. A small object that is very reflective may have a similar signal response to a larger less reflective object. In addition, the geometry and orientation of the object will affect the signal response. A person standing side-on to the wire will tend to have a weaker response than one facing the wire.

By measuring reflected signals as well as the transmitted signal, further information can be gleaned about the nature and proximity of any objects. By using the different signals available, it is possible to develop models for differentiating and categorising objects to determine their position, proximity and features.

The main processing unit 33 collects the information produced by the target signal processing units 32a, 32b. It may then utilise machine learning to produce an expected response from a particular environment. For example, when the system is first installed and activated, an initial set of data can be obtained by the main processing unit 33 which reflects the normal environment surrounding the detector 1 at that time and in the current conditions. In this way the system can identify what it should expect to observe under typical circumstances. In this way any changes to the received signals will be indicative of a change in the environment around the wire 4. These changes may be caused by a number of different effects along the length of the wire 4, including: people or animals approaching the wire 4; structural changes such as trees or rocks falling onto the line; and effects due to weather and other environmental conditions such as rain, humidity and temperature.

Using machine learning, the main processing unit 33 can be provided with information around some of these parameters so that it can produce a set of standard models reflecting the range of responses that might be expected under certain conditions and so derive an acceptable set of changes in the environment. For example, the system may adapt to learn how the received signals are impacted by different kinds of weather. Wind may have an impact on the wire itself causing it to move slightly as a result, which may be a significant impact on the signals received. Again, the machine learning algorithm used in the main processing unit 33 can analyse these changes so that it recognises the impact that, for example, high wind may have on the received signals. In this way, it can not only identify when there is high wind but also adjust its sensitivity to avoid false detections as a result. Further learning may be used to then recognise the difference between the impact of wind only on the system and how that compares to the impact of wind when there is also an intruder in the proximity of the wire. This allows the system to continue to identify intruders even when it might appear to be overwhelmed by the effect of signals caused by rain, wind or other effects.

The main processing unit 33 determines a likely scenario for the current environment which it provides to the data egress manager 34 which can provide data to an appropriate recipient. This could be sending an alarm to a signalman if a trespasser is detected or there is an object on the line. It may provide other information to other locations for example if there is high wind or rain. It may also be configured to redirect CCTV cameras or send audio messages to be played to trespassers using loudspeakers, warning them to move away from the area.

In the above examples, intruders approaching a line, or rain or other effects such as wind will have a short term and potentially quite substantial effect on the signals detected, allowing the system to easily identify potential incidents. With suitable learning algorithms, the detector can be developed to differentiate between the different types of impact on the transmitted and received signals. Environmental changes such as wind, rain and temperature will affect the detected signals. Rain drops will be detected, wind may cause the line to move, temperature may cause the line to contract and tighten or expand and relax and so sag more. These effects can be detected and effectively provide a method of determining the current environmental conditions. The effects may also be accounted for by using external information such as a temperature information or weather information. Using temperature information, the system would know the expected effect on the system and can compare that to the actual effect to update its model and/or detect other factors such as an object near the wire which may have otherwise been overlooked as an environmental change.

As noted above, the signals are analysed to identify parameters of the received signals such as the time of flight of the signals between transmission and reception, the difference in time of arrival at each transceiver, as well as the amplitude, profile and phase profile of the received signals. The measurements may be repeated at different frequencies and using different wires, where multiple wires are employed, to develop additional reference parameters. These parameters can be combined as a signature corresponding to the current environment around the wire 4. Such signatures can be stored as reference signatures corresponding to different sets of circumstances. Sets of reference signatures can be developed which reflect the expected signals under certain conditions such as the monitored environment under standard conditions but with a person at a certain distance along the wire and away for it.

By producing a set of such reference signatures, a library can be developed reflecting a host of different variations in the environment from changes in the physical environment, such as rock falls, landslides, to transient changes caused by passing trains, animals, people etc. Each signature will typically include reference information or metadata which reflects the environment of that signature. For example, once a signature is recorded, additional information can be associated with it to identify what the environment includes.

For example, the information may prescribe "Normal environment, with object of human size and nature, located 27 meters from one end of the wire at a distance of 1.5 meters, no rain, wind low, temperature 24C", or "Normal environment, no additional proximal objects, located 27 meter, heavy rain, wind low, temperature 10C" etc. The metadata may be simple text such as the above to allow a human operator to review the current status of the data may be structured into specific parameters, with flags (e.g. object present/not present, rain/no rain, temperature high/low, passing train in detection zone, etc.) or specific values or ranges (e.g. distance of detected object along the wire, proximity of object to wire, size of object, temperature, wind speed, etc.).

By comparing the current signature to examples from the library, an estimate of the current status of the monitored area can be derived. For example, if a trespasser enters the monitored area at a certain point in rainy conditions a particular current signature would be obtained. By matching the current signature to the closest reference signature, corresponding to a person within 1 meter of the wire 75 meters along it in rainy conditions, the system can offer a suggestion of what is happening.

However, the generation of these reference signatures can also be used to identify changes which happen slowly over a longer period of time. For example, if the system is initially set up to detect its environment, any foliage around the railway will be accounted for. However, if substantial growth occurs such that it may represent a hazard to railway traffic, then it will slowly cause a change in the signals detected which will progressively vary from the reference signature under a given set of circumstances. In this way, even slow changes can be detected by comparing current conditions with historical or reference signatures. The system may compare the current signature to a recent signature to monitor short term changes, as well as comparing the signal to the library to look at absolute matches for specific conditions.

Where a sudden change occurs, a significant change in the received data will allow such changes to be detected. However, in a real environment there are often changes which do not represent a hazard that needs to be considered. For example, passing trains are a perfectly normal and acceptable occurrence. Other events such as authorised railway workers may be transiently present on the line. If this is known in advance or happens at specific times, then the system can take account of that.

Other events may occur, such as birds flying in close proximity to or landing on the wire 4, which may cause a transient change in the detected signals. To avoid false alarms, certain events may be ignored unless they continue for a period of time. However, this can lead to significant issues being overlooked. For example, if a tree falls onto the railway line there will clearly be a significant period of movement as the tree is falling which can be identified by comparing the current signature with a recently obtained signature. However, once the tree has hit the ground, it will become stationary and the detected signals will no longer be changing and so it might be assumed that it was just a transient event such as a passing train or a gust of wind. It might appear that everything is back to normal after such a transitory event.

However, by comparing the current signature with a historical signature from the library of a signature recorded before the event, it will be apparent that there has been a significant change in the environment around the detector wire 4. Similarly, other events such as rock falls or landslides might have a transitory effect on the signals but then settle down with no movement whilst still leaving a significant hazard to passing trains. Again, by comparing a historical or even a relatively recent reference signature to the current environment, a sudden change involving movement which suddenly stops can still trigger an alarm if the environment at the end of the process is not close to the initial environment or the reference signature. The change from the initial environment will manifest itself primarily as different base-line system attenuation/gain. The library of signatures may even include signatures reflecting fallen trees at specific positions and so matching the current signature to that can be used to identify the specific nature of the change.

The system may use the learning process to identify a large set of reference signatures corresponding to different environmental conditions. So, as noted above, when the weather is windy or rainy or hot or cold, then the environment may be significantly different such that the reference signature for each of these conditions and combinations of them cover might be stored. These selection of reference signatures may also be aided by using other information from other detectors such as temperature or rain sensors, which the main processing unit obtains information from, to identify the best reference signature to use. It may then compare the best match reference signature to the current conditions to identify any variation which might indicate a long term change or a short term effect such as a trespasser on the line. Where variation between the current received signals and the reference signature is significant, the signals may be used as a further reference signature or to update the existing reference signature.

The set of reference signatures may be integrated into a neural network or stored as sets of signatures in a lookup table to be compared with received data. By comparing a current signature with the stored signature in a lookup table, an indication of the meaning of the current signature can be obtained. For example, when the system obtains a signature of the current environment around the wire 4, this can be referred to the lookup table to identify the best match to the stored signatures, which may be one of those that represents a person on the line, that may identify a trespasser. The stored data can be refined by reference to the environmental conditions. So a signature that may look like a person on the line when it is raining may not be considered to represent one when the weather is fine. By modifying the way in which the signature is compared to reference signatures in the lookup table based upon the determined weather conditions, an improved output can be obtained to avoid false alarms whilst still identifying issues that need to be addressed, such as trespassers on the line.

In addition to a range of reference signatures for different environmental conditions, the system may also obtain reference signatures corresponding to certain types of event which may be significant. For example, the system may be trained by having a person approach the wire 4 at different locations and from different directions and also potentially under different environmental conditions to generate a reference signature reflective of these kinds of events. In this way, when the system is monitoring a live environment, it can compare the detected signature with one of its reference signatures to identify the most likely scenario for the signature it is detecting. Therefore, if a person approaches the line at a particular location along the wire 4 then that will be compared to potential reference signatures to identify the best match which might correlate to a particular type of object at a particular location, e.g. a human being at a distance of 21 metres from transceiver 5a. Other references may be developed for other potential objects such as animals, trees, rock falls etc., including different sizes, shapes, profiles of people and objects.

In addition, when the system is being initiated, any bends or curves in the wire 4 will have an impact on the signature that is detected. As the system is initiated to determine the signature under different circumstances, the effect of any non-uniformities or bends which may cause reflections due to impedance mismatch can also be taken into account.

However, these can be taken into account when determining the initial signature and also when identifying signatures corresponding to certain specific events. For example, detecting a person approaching the portion of the wire 4 on a bend may have a different effect on the detected signals than if that person were to approach one of the straight sections of the wire 4. However, as part of identifying the signatures of people approaching the line at different positions, the effect of the bend can be taken into account.

The system may initially be provided with a starter set of signatures which may be quite generic in nature. The system can then be trained and additional signature developed which more closely reflect the environment in which the system is located.

By using the multiple measurements obtained from using the two transceivers 5a, 5b and detecting both reflected and transmitted signals, multiple sets of data can be developed. This can be further improved by using different frequencies to collect a different set of data to form part of the signature. Different frequencies will provide different responses to different objects and so a different perspective can be obtained to help to further differentiate between different types of objects etc. These can then be compared to corresponding reference signatures prepared using similar multi-frequency analysis to identify potential matching signatures in each scenario.

When the current signature is compared to reference signatures in the library, a best match reference signature may be selected but other techniques may be used. For example, a majority vote can be taken to identify the most likely scenario. For example, if a person has approached the wire, several scenarios and their respective reference signatures may closely match the measured signature. By considering several of the closest matches, better target classification and location can be obtained. This may be done by means of a simple majority of matching signatures or using a more probabilistic or soft approach.

For example, in a simple majority approach, three out of the four best matching signatures may suggest there is no intruder, whereas one signal does but, in this case, the majority would overrule the probable false alarm from the dissenting one. Where a probabilistic approach is used, each reference signature might be assessed to suggest a probability of it being reliable and then that probability applied to each of the best matches to assess an overall probability of a certain eventuality. Again, considered the issue of a trespasser on a line, the five best matching reference signatures might be given respective probabilities that they reflect a person in the detection zone of 100%, 85%, 75%, 25% and 65%. This reflects an overall averaged likelihood of 70% of there being a person on the line. Then only if the overall probability is greater than a threshold amount, perhaps 50%, is a consensus decision given as an intruder detection. Similar analysis might be done to identify other events such as Is it raining?". Each of the best matching signatures might be similarly ranked according to the likelihood of the signature being predictive of rain to determine an overall probability.

The above examples are only one of many techniques which may be used to match signatures to identify events and environments around the detection area.

By preparing multiple signatures, when a target approaches the detector, it can be matched to one of the signatures suggesting a particular target at a particular location. If the target moves then the signature matching may suggest a different signature representing a similar target at a slightly different location. By repeatedly carrying out this analysis, the system can identify that the sequence of signatures reflects the events that would be expected when an object of a certain size moves in a certain direction. This allows targets to be tracked and monitored as they move relative to the detector.

The learning process may be entirely automated or aided by human input. For example, the system may be trained by deliberately introducing objects into the environment to provide training material for the system to develop a model and reference signatures with information being provided as to the severity and significance of these models. The system may employ artificial intelligence (Al) to allow it to learn and develop the model, for example using a neural network to recognise the different environments and their impact on the signals received and then acting on that information in an intelligent way. The learning system may be given a generic set of basic signatures which it develops to accommodate local variations and environment or it may start completely from scratch with no initial information. The Al may then be used to interact adaptively with the human world by informing a controller, or actively responding such as sounding sirens, providing messages over loudspeakers to potential trespassers, stopping trains etc. In one example, the system may initially be configured to operate in conjunction with a CCTV system so that objects which arrive in the proximity of the wire 4 can be observed by the CCTV system to allow correlation between objects which are problematic and objects or events which are not. Such a CCTV system, possibly in conjunction with other detection systems such as motion sensors, can be used to supplement the decision-making processes of the machine learning system to improve the accuracy of the models being developed.

In the exemplary arrangement described above, the surface wave signal is propagated along the wire 4 using a signal provided by the signal generators 31a, 31b. The signal generators may produce a single frequency signal to be propagated along the line but they may also be operated using different frequencies, as suggested above. For example, signal generator 31a and signal generator 31b may use different frequencies to allow them to operate simultaneously without interfering with each other. Using a single frequency can be advantageous as the transceivers can have a narrowband receiver which will provide improved noise rejection and also improved range for the signals for a given power. This will allow for detectors covering a greater distance or provide detectors which operate with lower power.

Alternatively, each transceiver may utilize multiple frequencies to allow it to obtain a set of signatures for each frequency or signatures that include multiple sets of data provided by different frequencies. This may be done by using a broadband signal including several different frequencies or it may sequentially cycle through a set of different frequencies, or simply scan across a range of frequencies over a period of time. By utilising different frequencies, the response from objects in the proximity of the wire 4 differs, allowing greater differentiation of the objects in the environment of the wire 4. Objects which have a size similar to or greater than the order of magnitude of the wavelength of the signal propagating on the wire 4 tend to produce a stronger response, in terms of absorption and reflection, than objects with sizes that are significantly smaller than the wavelength in magnitude. In particular, objects bigger than half a wavelength of the signal will tend to provide a much stronger response than much smaller objects. For example, a signal with a wavelength of around 0.5m will be much more strongly affected (e.g. reflected, absorbed) by a person-sized object than by a raindrop. In contrast, a signal with a wavelength of 1mm will be affected significantly by a rain drop.

This can be useful for differentiating between objects such as people compared to objects such as raindrops which are clearly very different in size. Also, by comparing the environment across different frequencies and looking at the amplitude and phase profile of the received signals, it is possible to determine whether a change is caused by a large object more distant from the wire or a smaller object in close proximity to a wire etc. With a single frequency, it is difficult to differentiate between say a small bird landing on the wire and a person standing a metre away from the wire. However by analysing the environment using different frequencies, the respective responses will vary, allowing both the nature (dependent on size, reflectivity, absorbance, orientation, geometry etc.) and proximity to be assessed. This can be particularly important as a bird landing on the wire is clearly not a significant hazard whereas a person trespassing on the line or a tree which is fallen is clearly more significant.

The system may use a wide range of frequencies from a few tens of MHz to several tens of GHz. Different frequencies may be more effective at identifying objects of different sizes and at different proximities. Utilising multiple frequencies allows the benefits of the different frequencies to be combined. Frequencies as low as 10Mhz may be useful in the same way that frequencies as high 100GHz might be. Preferred ranges would be between 100Mhz and 10 GHz. However, many countries have limitations on what frequencies may be used and so to avoid the requirement for special licencing, the system may be operated using available or open frequencies such as those in the ISM bands which are generally freely available to be used or usable under certain restrictions. However, because of the nature of the surface wave and the fact that very little energy is radiated away from the wire 4, the system can be arranged to operate over a large range of frequencies without causing interference by radiating energy away from the system and so may be used with frequencies that would not be allowed if they were being radiated into the environment.

Breaks in the wire 4 may result in the surface wave not being contained and causing some radiation of energy away from the system. However this can be easily detected by the transceivers 5a, 5b and the system shut down to prevent energy being radiated. In addition, breaks may be indicative of trespassers on the line and so any detected breaks can be used to raise an alarm. Significant deformation of the wire, which may result from damage such as one of the supports being broken or perhaps a branch falling onto the wire 4, can also be detected due to the impact on the signal transmitted along the line. This again may cause an alarm to be raised.

In the example shown in Figure 2, as described above, the detector includes a single wire 4 arranged between the horns 20a, 20b of the transceivers 5a, 5b. However, the detector may include two or more wires working together to provide additional information about the environment around them. Figure 4 shows a modified version of the detector of Figure 2, including two wires 4a, 4b which work together to provide additional information. In the arrangement shown in Figure 4, the wires are arranged one above the other and spaced apart from each other. In the example shown, the second wire 4b is mounted on the same posts as those shown in Figure 2. Additional wires may also be added to further improve the information collected to aid in identifying objects.

The arrangement in Figure 4 is simply an example of how two wires may be arranged but this is not restrictive and the wires may be arranged in different configurations. For example, the wires may be arranged at the same height but displaced horizontally from each other. Similarly, multiple wires may be arranged in different configurations rather being simply one above the other. For example, they may be arranged in a square configuration, in the case of a four wire system, or a triangle for three.

Depending on the frequency of the signal used to form the surface wave, the range of detection around the wire will vary. The signal strength of the surface wave developed around the wires tends to decline with distance radially away from and along the wire and so the sensitivity and the ability to detect objects decreases the further away from or along the wire the object may be. In view of this, for the wires to work synergistically, they are preferably arranged close enough together so that the detection region around them overlaps to some extent so that objects approaching the line are detected by both wires. However, this is not essential and even if only one of the wires can detect an object close by, that may be sufficient to identify targets.

Figure 5 shows a modified control system for a two wire system such as that shown in Figure 4. The control system is similar to that Figure 3 but includes additional signal generators 31c, 31d and target signal processing units 32c, 32d to produce signals and process received signals from the extra two transceivers Sc, 5d.

In the arrangements shown in figures 3 and 5, it is assumed that the transceivers 5a-5d our connected to the control system 30 which may be located close to one or the other transceiver or possibly even somewhere else. Furthermore, the signal generators 31a- 31d and target signal processing units 32a-32d may be located separately to the other elements of the control system 30. For example, the signal generator 31 and/or the target signal processing unit 32 for a transceiver maybe co-located with the transceiver and the connections between them and the main processing unit 33 extending from the transceiver to the control system 30.

Although the examples above suggest using separate signal generators and target signal processing units for each transceiver, a single signal generator may be used with the generated signal being multiplexed to each of the transceivers. Similarly, as single target signal processing unit may have the signals from two or more transceivers demultiplexed so that it analyses the signals from multiple transceivers sequentially.

The wire 4 used to carry the surface wave can also be used as a means of communicating between the ends of the wire 4. In this way, the signal received by a transceiver can be processed by a target signal processing unit 32 which is co-located with the transceiver.

The data from the target signal processing unit 32 can then be transmitted back over the wire 4 to be received at the other end. For example, if transceiver 5b is currently providing a surface wave on wire 4, the surface wave will produce reflected signals received back at transceiver 5b and transmitted signals received at transceiver 5a. The target signal processing unit 32c co-located with transceiver 5a receives the transmitted signal and processes it to provide output data representing the received signal. The data can then be transmitted back along line 4 to transceiver 5b. The signal received at transceiver 5b is then extracted and passed to the main processing unit 33.

This arrangement allows the main processing unit 33 to be located close to one end of the wire 4 without being connected to transceiver 5a other than via the wire 4. The control unit may also send signals along the wire 4 to be received at the transceiver 5a to instruct a signal generator co-located with transceiver 5a to produce signals to initiate a surface wave from horn 20a. The surface wave generated will then be carried along wire 4 to be received by transceiver 5b whilst transceiver 5a will receive any reflected signals from objects adjacent to the wire 4. Again, the received signals at transceiver 5a can be processed by the target signal processing unit 32a which are then communicated back along wire 4 to transceiver 5b to be passed to the main processing unit 33. It will be appreciated that other forms of communication separate to the wire 4 may alternatively be used.

In the examples above, the wire 4 is formed from a conducting central core with an insulating dielectric around it. This configuration allows the wire to operate as a medium for carrying a surface wave. However, surface waves can be contained with other arrangements such as a plurality of helically wound wires. Other options that might be used to carry a surface wave are cables which have an armour wire sheath around them with an insulator surrounding the armour. These might be, for example, power cables with conducting cores or optical fibres for communication. The slightly larger dimensions of the armour sheath can provide an improved surface wave carrying capability but this is offset by increased size and dimensions of the cable used. However, it is quite common to have trackside cabling used for communications and power, next to railway lines. These are typically laid on the ground or in troughing and so the proximity to the ground can significantly affect the generation of surface waves and reduce the detection capability. However by suspending the cables in the air similar to the arrangement shown in figures 2 and 4, the detection range is considerably improved. In this way, existing cables might be utilised by simply relocating them onto an elevated support.

One issue that many security and monitoring systems have to address is avoiding false alarms due to an object or event appearing to be more significant than it actually is. One example is where an object that is not a significant issue is detected but appears to be more significant than it actually is. For example, a bird landing on the wire 4 would clearly provide a very strong signal which might be comparable to a larger object further away. As described above, this kind of issue can be addressed by making use of different frequencies or multiple wires which have different sensitivity to objects of different natures and also comparing the signature detected as the object approaches. The signature of a person approaching the wire 4 will be different to a small object approaching the wire 4 even though it may appear similar ultimately. In this way, before the bird arrives on the wire, the system can already have identified that an object approaching is a bird sized object and so when it ultimately lands on the wire 4, the system can differentiate it from a person approaching the wire.

Using multiple wires also provides a slightly different aspect on the detection area which can allow the angle of approach to be determined, e.g. from above like a bird or from the side like a person or animal.

Similarly, raindrops can have a substantial effect on the detector. A raindrop passing close to the wire 4 or even landing on it might have an effect comparable to a larger object further away. Again, by monitoring the signature of the received signals over a period of time, the system is able to differentiate between a raindrop approaching and a larger object approaching the wire. Furthermore, having identified that a raindrop sized object is approaching the wire, the system can then monitor the situation and if repeated small objects approach the wire then the system might conclude that there is rain falling. Having established that it is now raining, the system can modify its operation and perhaps use a different reference signature for ongoing monitoring.

The detector may use other sources of information to establish environmental conditions and to adjust the detection signatures accordingly. For example, the detector may include an alternative method for detecting rain or wind, or may obtain information from a weather service.

Furthermore, the system may switch to a different mode of operation such as by modifying the frequency or frequencies being used. By avoiding frequencies which have a wavelength which is close to the dimensions of a raindrop, the sensitivity to raindrops can be reduced. Therefore under normal good weather conditions, the range of frequencies may include frequencies which have wavelengths in the order of magnitude of raindrops.

When rain is detected, those frequencies may be excluded from being used. Whilst this may have some effect on the ability of the detector to identify objects, it has the considerable advantage of being able to supress the detection of raindrops and allow the system to continue to operate in different weather scenarios. It also allows the system to determine that it is raining, snowing etc. and also the intensity of the rain, all of which may be important for identifying flooding, potential land slips, or even risk to safe operation of the trains.

To minimise the possibility of water drops collecting on the wire 4 the wire may be coated with a hydrophobic coating to repel water landing on it. For example, the wire may be coated with enamel to provide an insulator but which also has a water repelling property. The wire may alternatively be coated with other water repellent materials such as Teflon®.

The ongoing assessment and comparison of the current signature with historical references can also be used to look at other environmental changes such as ground water content or ground movement.

The suspended wire 4 is able to move to a limited extent in response to sound or vibration. These small movements will affect the signature detected and by comparison with the reference signal can produce a signal corresponding to the vibration allowing the system to detect vibration due to objects in proximity to the wire such as trains. However, as sound and vibration can travel further than the range of the surface wave detection system, the sound or vibration can potentially detect those objects further away than the normal detection range. This can provide an additional pre-warning of the approach of an object.

This could be potential trespassers allowing the system to perhaps focus attention on a section of track or identify how trespassers are approaching the track. It might also be used to identify approaching trains to desensitise or completely stand down any warning that might be activated due to the train being detected.

The horns 20a-20d used in the detector are 3D shaped structures which are adapted to produce a surface wave of appropriate shape and diameter on the wire 4. The precise shape of the horn is generally related to the frequencies being used for the surface wave. The horn shape will be optimised for the frequencies being used in a particular application. The horns are relatively easy to produce either as plastic mouldings or 3D printed structures which are can be simply provided with a conductive coating on their inner surface. The horns may have other configurations and may be flat arranged parallel to the wire or may be formed as discs perpendicular to the wire.

This relatively simple structure for the horns and wire 4 means that the system can be constructed relatively easily and at much lower cost than installing regularly spaced CCTV systems to monitor an extended length of track. The wires can be mounted on simple posts or uprights, which may already be in existence, and so the installation is also relatively simple.

Although the embodiment above utilises horns, the use of horns is not essential and other structures can be used as launchers to pass the signals from the feed onto the line and similarly catchers for collecting the surface wave signal from the line.

In the examples above, the signal generators 31a-d are shown as separate elements.

However, it will be appreciated that a single generator may be used to distribute signals to each of the transceivers, this could be simultaneously or sequentially. Alternatively, a single signal generator may be provided at each end of the wire 4 to feed the transceivers at that end. Similarly, the target signal processing units 32a-d are shown as separate elements, but these may also be processed by a single unit by analysing the signals transmitted sequentially or using a multiplexor to analyse the signals essentially simultaneously. One or more of the target signal processing units 32a-d may be located at the respective ends of the wire or co-located at a common location.

Claims (27)

1. CLAIMS1. A detection system comprising: a first surface wave guide structure having a first transceiver, a second transceiver, and an elongate surface wave guide arranged between said first transceiver and said second transceiver; a signal generator for injecting a signal into said second transceiver, to produce a surface wave along said surface wave guide; a signal processor for obtaining a first received signal received at the first transceiver, and a signal analyser for producing a detected signature corresponding to objects in proximity to the conductor, based on the first received signal from the signal processor; and a comparator arranged to compare the detected signature to one or more reference signatures.
2. 2. A detection system according to claim 1, wherein said reference signature includes information corresponding to at least one of: amplitude; time of flight; phase, and the profile of these parameters over time.
3. 3. A detection system according to claim 1 or 2, wherein the signal processor is further arranged to obtain a second received signal, received at the second transceiver, wherein said detected signature is also based on the second received signal from the signal processor.
4. 4. A detection system according to claim 1, 2 or 3, wherein: the signal generator is further arranged for injecting a signal into said first transceiver, to produce a surface wave along said conductor, said signal processor is arranged for obtaining a third received signal received at the second transceiver and a fourth received signal received at the first transceiver; and said detected signature is also based on the third received signal and the fourth received signal from the signal processor. 5. 6. 7. 8. 9. 10.
5. A detection system according to any one of the preceding claims, further comprising one or more additional surface wave guide structures similar to said first surface wave guide structure, wherein the signal analyser also analyses the signal from each of the additional surface wave guide structures to produce said detected signature.
6. A detection system according to any one of the preceding claims, wherein the signal generator is arranged to inject signals of more than one frequency.
7. A detection system according to any one of the preceding claims wherein at least one of said reference signatures is produced based on a known environment around the surface wave guide.
8. A detection system according to any one of the preceding claims wherein at least one of said reference signatures is determined by obtaining a detected signature and storing it along with information corresponding to the environment around the detection system when the detected signature is obtained.
9. A detection system according to any one of the preceding claims wherein a plurality of said reference signatures are stored in a library and the detected signature is compared to the reference signatures stored in a library to identify the best matching reference signature.
10. A detection system according to any one of the preceding claims wherein a plurality of said reference signatures correspond to the same physical environment under different variable environmental conditions and wherein said comparator compares the detected signature to said plurality of said reference signatures to identify the closest matching reference signature to provide information on the environmental conditions.
11. 11. A detection system according to any one of the preceding claims wherein said comparator monitors how the closest reference signature to the detected signature changes over time for tracking the movement of objects around the surface wave guide structure.
12. 12. A detection system according to any one of the preceding claims wherein said elongate surface wave guide has a hydrophobic coating.
13. 13. A detection system according to any one of the preceding claims wherein said signal generator includes one or more separate signal generators for injecting a signal into one or more or said first, second, third, fourth and any additional transceivers.
14. 14. A detection system according to any one of the preceding claims wherein said signal processor includes one or more separate signal processors for obtaining received signals from one or more or said first, second, third, fourth and any additional transceivers.
15. 15. A detection system according to any one of the preceding claims wherein signals received by at least one of said transceivers are communicated from one end of the wire to the other along the wire using a surface wave on the wire.
16. 16. A method for monitoring a detection zone around a first surface wave guide structure having: a first transceiver; a second transceiver; and an elongate surface wave guide arranged between said first transceiver and said second transceiver, the method comprising: injecting a signal into said second transceiver, to produce a surface wave along said conductor; obtaining a first received signal received at the first transceiver, and analysing the first received signal to produce a detected signature corresponding to objects in proximity to the surface wave guide; and comparing the detected signature to one or more reference signatures. 17. 18. 19. 20. 21. 22. 23.
17. A method for monitoring a detection zone according to claim 16, wherein said reference signature includes information corresponding to at least one of: amplitude; time of flight; phase, and the profile of these parameters over time.
18. A method for monitoring a detection zone according to claim 16 or 17, further comprising obtaining a second received signal, received at the second transceiver, wherein said detected signature is also based on the second received signal.
19. A method for monitoring a detection zone according to claim 16, 17 or 18, further comprising: injecting a signal into said first transceiver, to produce a surface wave along said conductor, obtaining a third received signal received at the second transceiver and a fourth received signal received at the first transceiver; and analysing the third received signal and the fourth received signal to produce said detected signature.
20. A method for monitoring a detection zone according to any one of claims 16 to 19, further comprising one or more additional surface wave guide structures similar to said first surface wave guide structure, further comprising analysing the signal from each of the additional surface wave guide structures to produce said detected signature.
21. A method for monitoring a detection zone according to any one of claims 16 to 20, wherein said injecting said signals includes injecting signals of more than one frequency.
22. A method for monitoring a detection zone according to any one of claims 16 to 21, wherein at least one of said reference signatures is produced based on a known environment around the surface wave guide.
23. A method for monitoring a detection zone according to any one of claims 16 to 22, wherein at least one of said reference signatures is determined by obtaining a detected signature and storing it along with information corresponding to the environment around the detection zone when the detected signature is obtained.
24. 24. A method for monitoring a detection zone according to any one of claims 16 to 23, further comprising storing a plurality of said reference signatures as a library and comparing the detected signature to the stored reference signatures to identify the best matching reference signature.
25. 25. A method for monitoring a detection zone according to any one of claims 16 to 24, wherein a plurality of said reference signatures correspond to the same physical environment under different variable environmental conditions and further comprising comparing the detected signature to said plurality of said reference signatures to identify the closest matching reference signature to provide information on the environmental conditions.
26. 26. A method for monitoring a detection zone according to any one of claims 16 to 25, further comprising comparing how the closest reference signature to the detected signature changes over time for tracking the movement of objects around the surface wave guide structure.
27. 27. A method for monitoring a detection zone according to any one of claims 16 to 26, further comprising signals communicating signals received by at least one of said transceivers, from one end of the wire to the other along the wire using a surface wave on the wire.