GB2451615A - Radar reflection simulator for generating synthetic radar reflections - Google Patents

Abstract

The invention relates to method of simulating the reflection of radar pulse from a three dimensional object. A model 40 represents a three-dimensional object and the model is sliced a plurality of slices A,B,C. Surfaces 70,72,74 are created for each slice to represent the reflection of a radar pulse. The cumulated reflections of a pulse from each slice are represented by a discrete reflection from a surface. In this way, the reflection from an object is represented as a plurality of reflections, each reflection being from a respective surface. False colour may also be added to the model and each surface in order to represent characteristics of the reflected radiation, such as the strength of the reflection.

Description

RADAR REFLECTION SIMULATOR

The present invention relates to the generation of synthetic radar reflections. These synthetic reflections can be used for testing radar equipment and providing training aids by removing the need for a set of real objects.

Radar has been used extensively in both civilian and defence applications for determining the distance to targets as well as some of the targets' physical attributes. As with similar technologies such as sonar and lidar (Laser Imaging, Detecting and Ranging), radar operates on the basis of the detection of reflected energy.

Several techniques exist for testing radar equipment and radar imaging algorithms without the need for using real targets. A first technique involves creating a representation of a target that is a gross simplification; for example, representing a target as one or a few single-point reflectors. Simulated radar reflections from the simplified representation can be fed into radar equipment for testing. While being useful for testing that equipment works at a low level of complexity, this technique is severely limited in the number of aspects of the equipment that can be tested.

A second approach is to use a mathematical model of a real object. This model can be used to generate pseudo-radar reflection data for testing equipment and is particularly suitable for testing software representations of the final system; it often makes real-time testing impractical due to the time taken to run the models.

While this technique can test equipment at a high level of complexity it is computationaHy intensive. This makes the technique inflexible for testing alternative scenarios as the time taken to carry out a number of tests would be prohibitive.

A third approach uses actual measurements from a real target, or a scale model.

Although this approach uses data from real measurements, it does not require actual targets because historic measurements can be used. These measurements can be fed into real equipment to test the effectiveness of its processes. This approach is much less computationally intensive than using a mathematical model and can test more aspects of the equipment than a gross simplification of the target can. However, the technique also suffers from being inflexible. In particular, this approach does not allow the testing of radar equipment with radar reflections from alternative types of objects, alternative aspects of the same object, or variants of the object outside the scope of pre-recorded data.

According to one aspect of the present invention there is provided a method of simulating a reflection of a pulse of energy from a three dimensional object, the method comprising the steps of: providing a model that represents the three dimensional object; and representing the reflection of a pulse of energy from the model as a reflection from a surface, wherein continuous reflections from the model are represented by a discrete reflection from the surface.

The present method thus provides a reflecting surface by using a mathematical approach to provide target reflection data that can be generated to test a receiver such as a radar receiver. The surface may be generated in real time during the testing of a receiver, or else may be generated before the commencement of any testing. The advantages or each of the previously described prior art approaches are retained and their limitations are reduced.

In this way it is possible to represent the overall reflection from a three dimensional object by a discrete reflection from a surface. This method significantly simplifies the generation of synthetic pulse reflections from a complex shape without serious degradation to a sampling receiving system.

The present method may be particularly useful for testing radar imaging algorithms and radar equipment. The simplified representation of the reflection from the model of a complex object may then be used to test the effectiveness of detection, identification, imaging and ranging algorithms, either in software, firmware or hardware, appropriate to the stage of the radar development to be supported. In particular, the present method may allow the testing of imaging and ranging algorithms in a wide variety of different scenarios. For example, a model of a ship may be adapted to introduce an extra life boat, or a model of a car may be adapted to introduce a roof-rack. It may be desirable to test whether the various algorithms are able to detect, or are sensitive to, the adaptations to a model. Equally it may be desirable to test whether certain components of a model are detectable from different directions of view. The simplification of the representation of the reflection from a model enables this to be done without imposing restrictive computational demands.

As the present method is a method of simulating the reflection of a pulse of energy, it is clear that the pulse, the object, the direction of view and the direction of irradiation may also be simulated. The direction of view of an object may be modified as desired so that the object can be viewed in different orientations.

Similarly the direction of irradiation may be modified so that simulated pulses of energy are incident on the object from a particular aspect.

Where the direction of view and the direction of irradiation are equal, the analogous real world situation may be where a pulse transmitter and a reflection receiver are collocated. It is desirable that this situation can be simulated but it is also desirable to simulate reflections where a reflection receiver is located remotely from a pulse transmitter.

Preferably the method further comprises the steps of: slicing the model into a plurality of slices; and representing the reflection of a pulse of energy from the model as a plurality of reflections, each reflection being from a surface corresponding to a respective slice, wherein continuous reflections from a slice are represented by a discrete reflection from a respective surface.

In this way, reflections from an object may be represented at a high level of detail.

There may be several slices taken through an object giving a three dimensional depth to the representation of the reflection. In many situations it may be desirable to represent this three-dimensional depth so that receiver's depth resolution can be tested in relation to the algorithms it is required to process.

Preferably, the or each surface comprises points of common reflection. That is to say that points on the same surface reflect a pulse from the direction of irradiation towards the direction of view such that all reflections from that surface would be received at a theoretical detection point at substantially the same time.

The or each surface may be substantially flat, or a substantial plane. Generally speaking points of common reflection lie on a curved surface. However, it may be preferable to use flat surfaces as a first order approximation to curved surfaces.

The pulse of energy may be reflected from a direction of irradiation towards a direction of view. The direction of irradiation is the direction in which the pulse travels before it is reflected by the model and the direction of view is the direction from which the reflected energy is viewed.

The or each surface may have an area for reflecting the pulse, which area is the cross-sectional area that would be encountered by the pulse prior to reflection.

Where there is more than one surface, each surface may have a reflecting area that is substantially equal to the cross-sectional area of the corresponding slice that would be encountered by the pulse. In this way the reflection from a surface may represent the reflection from those points in the corresponding slice that are not in the shadow of another slice.

It may be comparatively straightforward to determine the cross-sectional area of a slice, as viewed from the perspective of a pulse prior to reflection, when the slice is considered in isolation to the other slices. This may be done, for example, by examining a two-dimensional perspective image of the isolated slice from the viewing direction. However, it may be more complex to determine the reflective cross-sectional area of a slice when the slice is partially in the shadow of other slices. Generally the reflection from slices which are partially in shadow will only be from areas of the slice that are not invisible to the pulse due to the masking effect of preceding slices.

Each surface may comprise a plurality of pixels. In this way a surface may be digitally represented, and each pixel may be treated as a single point reflector.

The number of pixels on a surface may be a function of the image resolution that is required. When determining the reflective area of a surface that corresponds to a slice that is partially in the shadow of other slices, it may be more convenient to determine whether individual surface pixels are in the shadow of pixels in preceding slices.

Each reflecting surface for each slice may be determined in a successive manner.

The first surface may be determined by examining a two-dimensional perspective view of the first slice. The second surface may be determined by examining those areas of the second slice that are not in the shadow of the first slice. This may be easily determined by subtracting the cross-sectional reflecting areas of the first slice from the second slice. This technique may further help to minimise computational demands.

The plurality of surfaces may be arranged so that there is a series of successive reflections, each reflection being from a different surface, and each reflection occurring at a different time. In this way, the surfaces may be arranged in a stack with one behind the other in the direction of view. Thus, as the pulse of energy sweeps past the model it may encounter successive surfaces at predetermined time steps. Preferably the time delay between successive reflections from different surfaces is substantially the same, or otherwise suited to the receiving equipment..

Preferably the pulse of energy is a radio frequency pulse. In this way, the present method may be used in radar simulations. It will be appreciated that a pulse may be a radio frequency pulse at a changed frequency, as is used in continuous wave radar. Radar is a particularly notable application due to the popuiariy of its use in both civilian and military applications.

As well as radar, the present method is also applicable to simulating the reflection of other types of energy pulse. Lidar (Laser Imaging Detection and Ranging) uses pulsed lasers for ranging and to determine other properties of target objects.

Sonar uses pulsed sound waves for the same purpose. It is apparent that the present method is also applicable to sonar and lidar so that the pulse of energy may be an acoustic pulse or a pulse of light (whether the light be optical, infra-red, ultraviolet or from some other part of the electromagnetic spectrum).

The representation of the reflection from the or each surface may comprise a representation of at least one property of the reflected energy. In this way it is possible to represent further details of the reflection of energy. For example, part of an object may have a high reflectivity; this may be a represented property in a the model, or in one of the surfaces comprising points of common reflection.

Other aspects of the reflected energy that may be represented include amplitude, phase, polarisation, direction of reflection, and frequency (or Doppler) shifting.

Each surface may comprises a plurality of pixels and each pixel may comprise a representation of at least one property of the reflected energy from that surface.

Thus, each pixel can have a unique personality in terms of its response to the incident energy.

The relative strength may be a represented property of the reflected energy. The strength of the reflected energy may be dependent on the reflectivity of parts of the object. As such, the strength of the reflected energy may vary across the object. The angle of incidence may also be an important factor in determining the strength of the reflected energy. Thus, even areas of low reflectivity may provide a strong reflection at grazing incidences. Multipath reflections may also add to the overzl1 strength of rcficctions.

The reflected energy may have a frequency. The relative change in frequency of the reflected energy may be a represented property. In this way, both the Doppler shift of a whole object and any micro-Doppler shifts due to moving elements within an object may be represented. Optionally, only variations in the Doppler shift within the model may be represented, an overall Doppler shift may be applied outside the model.

Preferably the representation of the at least one property of the reflected energy is according to a colour scheme. The representation may be a "false" colour in the sense that the choice of colour is purely arbitrary. A strong reflection could be represented by a dark blue, while a weaker reflection could be represented by a lighter shade of blue, but red could equally be chosen. Where the surface comprises a number of pixels, each pixel has a false colour attached to it.

Of course, it would be possible to represent the at least one aspect of the reflected energy in other "invisible" ways; for example, using a hidden numerical representation which may be attached to individual pixels. However, the use of colour allows an individual to examine the model and/or the plurality of surfaces and to readily ascertain certain reflection characteristics.

Many properties of the reflected energy may be represented by false colours.

However, as more aspects are represented, immediate human comprehension of the reflection may become more difficult. It may be that only up to three properties can be readily presented using false colours. A processing machine, by contrast, may have no difticuUy in rendering the way in which the reflected energy is represented and handled. Optionally, there may be more parameters attached to each pixel than are presented for human viewing, the rest being used for machine processing.

At least one property of the reflected energy may be represented according to a colour scheme in at least one of the model and the plurality of surfaces. Thus, false colour may be applied either before or after the model is sliced, as appropriate.

The method may comprise the further step of simulating the detection of the reflections from the or each surface. In this way it may be possible to create a representation of the surfaces, as they would be perceived by a real world receiver such as a radar receiver.

In addition, the method may comprise the step of inputting the simulated detection of the reflections into a receiver for processing. Thus the surfaces produced from the sliced model can be imaged as if they were elements of a real object as detected by a radar receiver, for example. The receiver may then be used to create an image of the model in order to test the functionality and sensitivity of the receiver. The receiver, or any parts of it, may be in hardware, or alternatively in software.

The input point in the receiver for the simulated detection may be selectable and the detection of the reflections from the or each surface may be simulated according to the input point. Thus, different aspects of the receiver may be tested depending on the choice of input point. The receiver may comprise aspects such as an antenna, a demodulator and a processor and the input point may be before any of these elements. The detection of the reflections may be simulated accordingly to provide a suitable input to the selected input point.

The method may comprise the step of selecting the direction of view of the model and selecting the direction of irradiation of the model. This selection may be manual by a user, or may be automatic.

According to another aspect of the present invention there is provided a reflection simulator for simulating the reflection of a pulse of energy comprising: storage means adapted to store a model representing a three dimensional object; and a surface calculator, arranged to determine a surface, so that the reflection of a pulse of energy from the model is represented as a reflection from the surface, wherein continuous reflections from the model are represented by a discrete reflection from the surface.

Preferably the reflection simulator also comprises a slicer for slicing the model into a plurality of slices, wherein the surface calculator is also arranged to determine a surface for each slice, so that the reflection of a pulse of energy from the model is represented as a plurality of reflections, each reflection being from a respective surface, and wherein continuous reflections from a slice are represented by a discrete reflection from a respective surface.

The reflection simulator may be for simulating radar reflections and may also comprise a detection simulator for simulating the detection of the reflections from the or each surface. In this way, it may be possible to test the sensitivity of real According to another aspect of the present invention there is provided a method of simulating a reflection of a pulse of energy from a three dimensional object, the method comprising the steps of: providing a model that represents the three dimensional object; and representing at least one property of the energy reflected from the model according to a false colour scheme, so that a false colour is applied to surfaces of the model, wherein the false colour is representative of a property of the reflected energy other than colour.

In this way the model can be visually inspected and certain reflection characteristics may be readily ascertained. For example, the strength of the reflected energy and any Doppler shifting, where the pulse of energy is a pulse of radiation, may be represented using false colour. This may enable a human to immediately understand which elements of a model provide strong reflections and/or Doppler shifted reflections.

Any of the method features may be provided with any of the apparatus features and vice-versa.

Preferred features of the present invention will now be described, purely by way of example, with reference to the accompany drawings, in which: Figures iA-iC show a radar transceiver transmitting and detecting radar pulses at different epochs, as background to the present invention; Figure 2 shows a flow diagram of a procedure for simulating the generation of radar reflections in an embodiment of the present invention; Figures 3A-3G show a representation of a three-dimensional model of a vehicle that is sliced into a plurality of surfaces in an embodiment of the present invention; Figure 4 shows an example of a typical radar system architecture, the functionality of which may be tested in an embodiment of the present invention; Figure 5 shows a schematic diagram representing the internal operations of a synthetic radar simulator in an embodiment of the present invention; and Figures 6A and 6B show another schematic diagram representing a synthetic radar simulator used to test equipment in an embodiment of the present invention.

By way of background, Figures 1A -IC show a simple representation of a radar transceiver transmitting and detecting radar pulses at three separate epochs.

In Figure IA, a radar transceiver 2 transmits a pulse of radio frequency energy 8A which advances at the speed of light. In the first epoch, the pulse SA travels towards spaced reflective objects 4 and 6. The pulse 8A has a spherical wavefront as it emanates outwards with the centre of the sphere being coincident with the transceiver 2.

Figure 1 B shows the pulse 8B in a second epoch arriving at a first reflective object 4. The pulse 8B has dispersed from the first epoch to the second epoch such that the energy of the pulse is distributed over a wider area. Upon reaching the first reflective object 4, part of the pulse continues, and part is reflected back towards the radio transceiver 2. Part of the pulse is also absorbed by the first reflective object 4. In reality, reflected energy may be scattered in all directions, with a portion of it being returned to the radar transceiver 2.

Figure 1C shows the pulse 8C arriving at a second reflective object 6 in a third epoch. Upon reaching the second reflective object 6, part of the pulse continues and part of the pulse is reflected back towards the radar transceiver 2. The pulse 8C disperses from the second epoch to the third but nevertheless there is a central gap in the pulse, as returned from the second reflective object 6, owing to reflection and absorption from the first reflective object 4. At the same time as the pulse arrives at the second reflective object 6, the reflection from the first reflective object 4 is approaching the radar transceiver 2. The transceiver 2 may be arranged to sample the received reflections at a sampling rate for analogue to digital conversion.

Ii. will be appreciated that the receiver will be subject to, in this case, two return pulses separated in time, with energy proportional to the properties of the first and second reflective objects 4 and 6. In a more complex situation there would be many reflections from surrounding material, referred to in the art as "clutter".

Reflecting objects in the real world have complex shapes that produce many reflections in dependence on a large number of variables.

Figure 2 shows a flow diagram of a method of simulating a radar reflection and testing radar ranging and imaging equipment. Aspects of the simulation process are encoded as software algorithms and may be run on a computer.

The first step in the simulation process is a modelling step 20. At the modelling step 20, a three-dimensional model is created or obtained to represent a real world object. Any kind of object can be modelled, depending on the application, from commercial ships to farm vehicles. The model may be based on measurements, drawings, or scale models and is typically prepared using a computer-aided drawing (CAD) package.

At the orientation and irradiation step 22 the model is developed for radar specific applications. A user is prompted to select a direction of view for the model and at least one direction of irradiation, this may be most appropriate if a specific scenario is being prepared prior to the testing of radar equipment. Alternatively the direction of view and/or the direction of irradiation are selected automatically, this may be more appropriate in a real-time simulation.

The direction of view of the model determines the orientation of the model from the viewing perspective. In real-world terms, this would equate to the position of a radar receiver with respect to the position and orientation of a target. The direction of irradiation is the direction from which radar pulses would be incident on the object. In real-world terms this would equate to the position of a radar transmitter with respect to the position and orientation of a target. There may be more than one direction of irradiation to represent the real world situation of multiple radar transmitters.

In this method of simulating the reflections from an object it is important to observe that there are no actual pulses of energy directed at the model, and thus references to the "direction of irradiation" are purely hypothetical.

Upon selection of a direction of view and a direction of irradiation, the three-dimensional model is sliced in the slicing step 24. The model is generally sliced in such a way that the surfaces separating adjacent slices comprise points of common reflection. These points of common reflection on a surface are "common" in the sense that each common reflection point reflects energy from the direction of irradiation towards the direction of view such that all of the reflected energy from a surface would be received at a theoretical detection point at substantially the same time. Put in a different way, reflected energy from the common points of reflection has substantially the same "time of travel" between the theoretical points of transmission and detection.

The present method is applicable to radar, which uses electromagnetic radiation travelling at the speed of light. Thus, radar reflections from a real-world target such as a vehicle or building would all be received at a detector at "substantially" the same time, when using macroscopic time references. However, there will be a microscopic spread in the time of received reflections; the length of the spread being the length of time it takes for a pulse to transit the object in question. The resolution in range of these time-spaced reflections varies from system to system; it may be on the order of a few centimetres or a few metres depending on the application. Therefore, in this context, "substantially" the same time means within around the resolution of the equipment. For a resolution in range of 1 metre, the resolution in thne would be in the region of 3ns.

In situations where the direction of view and the direction of irradiation are the same, as is the case in the simple diagram of Figure 1, the points of common reflection would tie on a spherical surface which may be between adjacent slices.

In this situation the model should be sliced so that the surfaces between adjacent slices are surfaces on a sphere whose centre is the location of the radar transceiver 2. Where the distance between the radar transceiver and a reflecting object is much greater than the size of the reflecting object, the wavefront of a pulse will be approximately flat at the location of the reflecting object. Thus, the spherical surfaces may be approximated to flat reflecting surfaces, or planes. This will be especially appropriate where the curvature of the spherical su face across the object is smaller than the resolution in range of the relevant radar equipment.

Where the direction of view is not equal to the direction of irradiation, the points of common reflection will also be on a curved surface. This surface can be calculated, and approximated to a plane, if appropriate, in the same way as if the two directions were the same. In embodiments where there is more than one direction of irradiation or direction of view, simulating multiple radar transmitters or receivers, the model must be sliced an appropriate number of times to create a different set of surfaces comprising points of common reflection for each transmitter-receiver pair.

Figure 3A shows a two-dimensional perspective view of a model of a three-dimensional vehicle 40 as viewed from the direction of view, and with the direction of irradiation being the same as the direction of view. The model comprises wheels 42a,b, and a cabin 44 comprising a windshield 46, a roof 45 and a side door 50 with a window 48. The model also comprises a bonnet 52, radiator grill 54, front and rear bumpers 56a,b, and a rear storage compartment 58.

Figure 3B shows a top view of the model of Figure 3A as oricntcd with its main axis F at an angle G to the direction of view E. The model is sliced in the slicing step 24, shown in Figure 2, so that points of common reflection lie on surfaces between adjacent slices. In this example, it is a reasonable approximation for the surfaces to be flat planes. Figure 3B shows three surfaces 70, 72, and 74 defining three slices A,B,C taken through the model of the vehicle U) in a direction orthogonal to the direction of view E. Each slice represents a time-step sweep of a pulse of energy through the model in the direction of irradiation. The three surfaces 70, 72, 74 are coincident with the divisions between the slices A,B,C.

Howev er, this need not be the case. The surfaces 70, 72, 74 may be equally representative of the reflections from a slice if they are positioned in the middle, at the front, or at some arbitrary point in the appropriate slice.

Only three slices A,B,C are shown in this example. However, for applications (for example military applications) where a resolution in the range of a few centimetres is required, several hundred slices may be needed. Typically the width of each slice is a function of the resolution in range of the radar application.

The number of slices may be roughly equal to a dimension of the model divided by the resolution in range. Alternatively, and at the lowest level of approximation, there may only be one slice that corresponds to the entire model.

In this situation there would be a single corresponding surface and a discrete reflection from that surface would represent the reflections from the entire object.

Referring back to Figure 2, having sliced the model, the cumulated reflections from each slice, due to a pulse of energy travelling in the direction of irradiation are then simulated by the reflection simulation step 26. This process may be understood more readily with reference to the model shown in Figure 3.

Figure 3C shows the portions of the vehicle which may be responsible for radar reflections from the first slice A. Figures 3D and 3E show the portions of the vehicle 40 responsible for cumulated reflections of a pulse due i.e slices A and B, and A, B and C respectively. This may also be understood by analogy to using a white screen, inserted at the slicing point, such that parts of the object behind the screen cannot be seen.

Figure 3F shows the portions of the vehicle 40 responsible for reflections from the second slice B. This image is obtained by subtracting the image of Figure 3C from that of Figure 3D. In a similar way, Figure 3G shows the portions of the vehicle 40 responsible for reflections from the third slice C; this image is obtained by subtracting the image shown in Figure 3D from that of Figure 3E.

Thus, Figures 3C, 3F and 30 represent the portions of the vehicle 40 responsible for reflections from slices A, B and C respectively. These portions of the model may be represented as images on two-dimensional surfaces for representing aspects of the reflection of a pulse. The surfaces 70, 72, 74 are located backwardly of the slice A,B,C that they represent, with respect to the direction of view.

Again using the analogy of the white screen, obscuration may be dealt with by "blacking-out" those areas in a current slice that would be obscured by previous slices. Blacked-out areas would then represent areas of no reflection in a current slice.

Of course, in a digital system a surface may actually be a web of separate pixels, rather than a continuous surface. In this way, each pixel may be like a single point reflector that has its own reflection characteristics. The number of pixels in a surface is selected based on the desired resolution of the object; it is apparent that a higher number of pixels places higher computational demands on the system.

In practice, this technique is used to dispense with the full model of the vehicle 40 and replace this with two-dimensional images at the location of the three surfaces 70, 72 and 74. Each two-dimensional image represents the cumulated reflections that would be experienced by a pulse of energy advancing past a slice of the model. Thus together, three two-dimensional images can represent the overall reflection from the three-dimensional model in a particular direction of view and with a particular direction of inadiation. Should a new direction of view or direction of irradiation be required, the model may be re-sliced to determine new surfaces comprising new points of common reflection. In this way, and referring back to Figure 2, the orientation and iiradiation step 22, slicing step 24, and reflection simulation step 26 can be repeated as required, either in real-time or "offline" prior to any tests. Generally, it is particularly advantageous, in terms of saving computational effort, if these steps are undertaken offline.

Having created surfaces that represent the reflection of a pulse from a slice, the reflection of a pulse of energy from a model can be simulated. The detection simulation step 28 is operable to simulate the detection of reflections from the or each surface. This creates an output equivalent to a digitiser output that would be produced by a radar receiver. General noise, "clutter" and/or reflections from other objects may be combined with the simulated reflection at the detection simulation step 28.

The output from the detection simulation step 28 is fed into the testing step 30.

The testing step 30 involves passing the simulated detection through a radar reflection processing system, whether it be in software, firmware or hardware. It may be desirable to test radar processing software in development, for fine-tuning before the software is implemented in radar receiving equipment. In addition, it may be desirable to test the radar processing equipment to ensure that the final hardware is performing as expected from previous software tests. The output from the testing step 30 is the standard output from the radar processing equipment; this may include a range together with a three-dimensional image of the target object. This output is sent to the comparison step 32 which compares the output from the testing step 30 with the three-dimensional model obtained in the modelling step 20, or conceivably the reflection surfaces in the case where steps 20-24 are pre-prepared offline. The results of this comparison may then be output in an outputting step 34 which may be arranged to generate an analysis report detailing the performance of the radar processing equipment.

Figure 4 shows an example of a typical radar system architecture, the operation of which may be tested. A transmitter 80 transmits radar pulses via the "real world" 81 to be received by a receiver 82. The transmitted radio pulses have a carrier frequency and a pulse frequency, as determined by a carrier frequency tuner 84 and a modulation frequency tuner 83 respectively; these may typically be provided on the transmitter 80.

The transmitter 80 comprises a modulator 85 and an antenna 86. Radio waves transmitted from the transmitter 80 then follow a typical path through the "real-world" 81 on the way to the receiver 82, this typically involves: a first pass through the atmosphere 87, reflection from a target 88, further reflections and added noise from clutter 89, and a second pass through the atmosphere 90.

The receiver 82 comprises an antenna 91, demodulator 92, de-compressor 93, processor 94 and display 95 for receiving and processing radar reflections. The antenna 91 receives reflected radar from the real world 81, the demodulator 92 mixes the carrier down to a baseband, the dc-compressor 93 extracts the signal from the modulation, and the processor 94 processes the remaining signal to produce an image of a target, finally this image is displayed on a display 95.

Mixing at the demodulator 92 requires a knowledge of the carrier frequency and mixing at the dc-compressor 93 requires a knowledge of the modulation frequency, to this end the demodulator 92 and the dc- compressor 93 have respective connections to the carrier frequency tuner 84 and the modulation frequency tuner 83. The processor 94 has a feedback connection to control either or both of the modulation frequency tuner 83 and the carrier frequency tuner 84, if required.

The real-world 81 may be simulated and the simulated reflection may be input to the receiver 82 to test its functionality and sensitivity. However, depending on the application different aspects of the receiver may be tested. Four possible injection points I-IV are shown for inputting a test signal to the receiver 82.

The first injection point I is before the receiver antenna 91. The test signal must be input to the first injection point I via a dummy transmitter (not shown). The real-world 81 is simulated and modified by the transmitter pulse format to modulate a radio frequency transmission. All of the components of the receiver 82 are then tested.

The second injection point II is before the demodulator 92. Thus, the test signal is the same as the test signal to the first injection point I, but it may be input by direct connection, rather than via a dummy transmitter.

The third and fourth injection points III, N are before the de-compressor 93 and processor 94 respectively. A test signal may be injected directly to the third injection point Ill using simulated real world data modified by transmitter pulse format, independently of the carrier frequency. Finally, a test signal may be input directly to the fourth injection point IV using simulated real world data only.

In this way, various aspects of a hardware radar system may be tested. The same concept may be applied to testing software models of the hardware system pre-build.

As well as improving the efficiency of radar reflection simulations, the accuracy of the simulated reflections from an object is enhanced. In one embodiment this is achieved through the use of false colour attached to some of the attributes of the model.

The properties of the reflected energy from a surface depend, in part, on the properties of the surface. For example, if a surface has a high reflectivity then reflections from that surface may be strong. However, the strength of reflections may also be affected by other factors such as the angle of incidence and potential multipath reflections. It will be understood that determining the strength of a reflection from a surface may be a complex process which is dependent at least on the properties of the surface, the direction of view and the direction of irradiation.

The strength of the reflection may be stored as an associated parameter to the model, but preferably this is displayed visually on the model through the use of false colour.

It will be understood that multipath reflections can increase the expected strength of a reflection from a surface. This will arise where the geometry of the object is such that the "time of travel" of a multipath reflection is substantially the same time as for reflections from common reflection points.

In Figure 3A, it may be that the radiator grill 54 provides a particularly strong radar reflection. This could be represented in accordance with a colour scheme.

For example, the radiator grill 54 could be shown in red with the intensity of red dependent on the strength of reflection. It is advantageous to show the strength of reflections using false colour so that a model can be visually inspected, and so that certain reflection characteristics can be readily ascertained by eye. Thus, the mode' and/cr any rcfiection surfaces 70,72,74 may be provided with false colour.

As well as the strength of reflection, other properties of the reflected energy that may be represented with false colour include the phase of the reflected signal and any frequency (or Doppler) shifting of the signal. Each parameter may be represented as a different false colour, and as many aspects may be represented as are required. As the surfaces 70,72,74 may comprise pixels, each pixel may have its own false colour. In this way the properties of the reflected energy, or other properties of the object, may be represented by a pixel with its own unique character.

False colour may be used to create a detailed model of a real world target. This allows immediate human comprehension of some of the reflection characteristics of a model and is a more convenient way of increasing the information that can be gleaned from the model. For example, and again referring to Figure 3A, if the vehicle 40 were moving forward then there may be an overall Doppler shift that is "blue-shifted" in the viewing direction. However, the wheels 42a,b, rotate relative to the rest of the vehicle 40 so that the upper half of the wheels 42a,b, is comparatively "blue-shifted" and the lower half of the wheels 42a,b, is comparatively "red-shifted". This effect, known as micro-Doppler, may be an important factor in testing whether apparatus can identify objects that have moving parts. For example, tank tracks provide a different effect to that of wheeled vehicles. Also aircraft jet engine parts and their position on aircraft wings are useful identification features if the direction of view is from the front or rear of the aircraft, for example. Another variant in the general use of false colour is to create an inter-relationship between different parts of a model. For example, a helicopter viewed from an oblique angle may necessitate a model such that part of the helicopter body would be periodically obscured as a rotating blade passes between the viewing position and the body. In this case, an embedded parameter in the model, representing the speed of rotation of the blades, could be made to vary the response of the body section that is periodically obscured. An associated mathematical model may be used to modify the stored parameters dynamically.

The reflection characteristics of a model may be frequency dependent. Thus, different false colours may be appropriate depending on the frequency of the radiation that is used.

Figure 5 shows a schematic diagram representing the internal operations of a synthetic radar simulator 100. The simulator 100 comprises a three-dimensional model 110, stored as series of data structures in storage means. A slicer 112 is arranged in communication with the three-dimensional model and is arranged to slice the model as required in dependence on the direction of view and direction of irradiation. The slicer 112 passes information about the slices to a surface calculator 114 which uses infonnation about the three-dimensional model to calculate surfaces, so that the continuous reflections from a slice can be represented by a discrete reflection from a respective surface. A detection simulator 116 is arranged to create an output, equivalent to a digitiser output that would be produced by a radar receiver, and the simulated detection is sent to external imaging equipment 118 in order to test the equipment. The imaging equipment 118 creates an output that comprises ranging information and an image of the model; this is sent to a comparator 120. The comparator 120 is arranged in communication with both the three-dimensional model 110 and the imaging equipment 118 and is arranged to compare the image and the model and to generate an output report for analysis by an external analysis unit 122.

Figure 6A shows another schematic diagram showing the internal operations of a synthetic radar simulator, as connected to equipment for real-time testing. A 3D model 130 is connected to a false colour applicator 132 which applies false colour to the model according to the test parameters: the direction of view, the direction of irradiation, and the frequency of the radio pulse. The model is then sliced by a slicer 134 nd the surface calculator 136 calculates appropriate surfaces for representing the reflections from respective slices. The reflections from the various surfaces are then combined with the reflections from other targets and clutter at a sununing station 138 which in turn sends data to a detection simulator 140. The detection simulator 140 then generates an appropriate test signal for injection at an appropriate point into the test equipment 142, by the process explained with reference to Figure 4. Finally the test equipment 142 provides an output to an operator display 144 for analysis.

The process described above is intended to be undertaken in real-time. That is to say, the model may be continually re-sliced, with surface continually re-calculated, every modulation cycle. A real-time process may only be appropriate where the computational power is sufficient for the application in question.

Figure 68 shows a modification to Figure 6A, whereby slicing and surface calculations may be undertaken off-line (i.e. not in real-time). A feedback is provided from the surface calculator 136 to the false colour applicator 132 so that the 3D model 130 can be sliced according to as many permutations of the test parameters (the direction of view, the direction of irradiation and the frequency of the radio pulse) as is appropriate. For every permutation of test parameters required, a different set of surfaces from the surface calculator 136 is stored in a storage unit 146. Calculating surfaces for every permutation of test parameters (within a certain range) would be coinputationally intensive. However, since this calculation may be undertaker off-line where there are no time constraints, this is not a problem.

In a real-time test, test parameters 148 are input to a view determination unit 150 which selects the appropriate set of surfaces from the storage unit 146. These surfaces are then input to the sununing station 138 and the test equipment 142 can be tested as previously described.

Claims (29)

1. Claims 1. A method of simulating a reflection of a pulse of energy from a three dimensional object, the method comprising the steps of providing a model that represents the three dimensional object; and representing the reflection of a pulse of energy from the model as a reflection from a surface, wherein continuous reflections from the model are represented by a discrete reflection from the surface.
2. 2. The method of claim I further comprising the steps of: slicing the model into a plurality of slices; and representing the reflection of a pulse of energy from the model as a plurality of reflections, each reflection being from a surface corresponding to a respective slice, wherein Continuous reflections from a slice are represented by a discrete reflection from a respective surface.
3. 3. The method of any of the preceding claims wherein the or each surface comprises points of common reflection.
4. 4. The method of any of the preceding claims wherein the or each surface is substantially flat.
5. 5. The method of any of the preceding claims wherein the or each surface has an area for reflecting the pulse, which area is the cross-sectional area that would be encountered by the pulse prior to reflection.
6. 6. The method of any of the preceding claims wherein the or each surface comprises a plurality of pixels.
7. 7. The method of any of claims 2 to 6 wherein the plurality of surfaces are arranged so that there is a series of successive reflections, each reflection being from a different surface, and each reflection occurring at a different time.
8. 8. The method of claim 7 wherein the time delay between successive reflections from different surfaces is substantially the same.
9. 9. The method of any of the preceding claims wherein the pulse of energy is a radio frequency pulse.
10. 10. The method of any of claims 1 to 8 wherein the pulse of energy is an acoustic pulse.
11. 11. The method of any of claims 1 to 8 wherein the pulse of energy is a pulse of light.
12. 12. The method of any of the preceding claims wherein the or each surface comprises a representation of at least one property of the reflected energy from that surface.
13. 13. The method of any of the preceding claims wherein the or each surface comprises a plurality of pixels and each pixel comprises a representation of at least one property of the reflected energy from that surface.
14. 14. The method of claim 12 or claim 13 wherein the relative strength of the reflected energy is a represented property.
15. 15. The method of any of claims 12 to 14 wherein the reflected energy has a frequency and the relative change in frequency of the reflected energy is a represented property.
16. 16. The method of claim 15 wherein the phase of the reflected energy is a represented property.
17. 17. The method of any of claims 12 to 16 wherein the representation of the at least one property of the reflected energy is according to a colour scheme.
18. 18. The method of any of the preceding claims wherein at least one property of the reflected energy is represented according to a colour scheme in at least one of the model and the or each surface.
19. 19. The method of any of the preceding claims further comprising the step of simulating the detection of the reflections from the or each surface.
20. 20. The method of claim 19 further comprising the step of inputting the simulated detection of the reflections into a receiver for processing.
21. 21. The method of claim 20 wherein the input point in the receiver for the simulated detection is selectable and the detection of the reflections from the or each surface is simulated according to the input point.
22. 22. The method of any of the preceding claims further comprising the step of selecting the direction of view of the model and selecting the direction of irradiation.
23. 23. A reflection simulator for simulating the reflection of a pulse of energy comprising: storage means adapted to store a model representing a three dimensional object; and a surface calculator, arranged to determine a surface, so that the reflection of a pulse of energy from the model is represented as a reflection from the surface, wherein Continuous reflections from the model are represented by a discrete reflection from the surface.
24. 24. The reflection simulator of claim 23 further comprising: a slicer for slicing the model into a plurality of slices; and wherein the surface calculator is arranged to determine a surface for each slice, so that the reflection of a pulse of energy from the model is represented as a plurality of reflections, each reflection being from a respective surface, and wherein continuous reflections from a slice are represented by a discrete reflection from a respective surface.
25. 25, The reflection simulator of claim 23 or claim 24 wherein the pulse of energy is a radio frequency pulse.
26. 26. The reflection simulator of any of claims 23 to 25 further comprising a detection simulator for simulating the detection of the reflections from the or each surface.
27. 27. A method of simulating a reflection of a pulse of energy from a three dimensional object, the method comprising the steps of: providing a model that represents the three dimensional object; and representing at least one property of the energy reflected from the model according to a false colour scheme, so that a false colour is applied to surfaces of the model, wherein the false colour is representative of a property of the reflected energy other than colour.
28. 28. A method substantially as described herein with reference to and as illustrated in the accompanying drawings.
29. 29. A reflection simulator substantially as described herein with reference to and as illustrated in the accompanying drawings.