Passive surveillance system

Abstract

A passive surveillance system is provided that comprises a first and second receiving system, each with a corresponding antenna, with at least one of the receiving antennas being moveable in relation

Description

This invention relates to a passive surveillance system operable at radio frequencies. In particular it relates to a surveillance system suitable for location and identification of radio transmitters of unknown position within a region. Locating the position of radio transmitters is a function often required by the military, and also by civil authorities. In a military scenario knowing the whereabouts of enemy transmitters can be important in planning attacks or appreciating the capabilities of the enemy. Civil applications include the location of pirate radio transmitters, in identifying intentional or unintentional interferers to other radio systems, or in search and rescue operations. One system for the location of a transmitter comprises a directional antenna coupled to a receiving system to record a direction of arrival of a signal from a transmitter. This will then give a track somewhere on which the transmitter is located. If the directional antenna is then moved to a different location, and the measurement repeated then a second track can be recorded, and the intersection of the two tracks will be the likely location of the transmitter. As an improvement, the two tracks can be recorded simultaneously using two separate directional antenna systems, and the data combined to produce a result more efficiently. Another system for locating transmitters utilises the Time Distance of Arrival or Frequency Distance of Arrival approach (TDOA / FDOA). These are two separate methods that are closely related, and in practice are often carried out together, to increase the accuracy of the resultant position information. The TDOA approach uses two antennas, which need not be directional, and each is arranged to receive the signal of interest. The time of arrival of the signal at each antenna is compared, and the difference calculated. This difference can be used to plot an elliptical curve on which the transmitter should be located. A repetition of the measurement at different antenna positions generally produces a different curve, and the intersection of the curves locates the transmitter position in two dimensions. The FDOA system is broadly similar, but instead of measuring the time of arrival, the instantaneous frequency of the signal is recorded. Of course, if the transmitter and the two receiving antennas are all stationary in relation to each other then the frequencies measured will be the same. However, if there is relative movement then generally a Doppler shift will be added to one or both antennas. This Doppler can be measured and compared, and the difference used to plot a curve on which the transmitter lies. Again, a repetition of the measurement at different positions will result in additional curves being generated, and the overlap of the curves is indicative of the position of the transmitter. To produce the required movement, FDOA is very often implemented on two or more vehicles, typically aircraft. In practice, a Cross Ambiguity Function (CAF) is often used to calculate the TDOA and FDOA of an input signal generated by an emitter, and so estimate the location of the emitter. More details of this can be found in “Wideband TDOA / FDOA Processing Using Summagion of Short Time CAFs”, Ullman et al, IEEE Trans. On Signal Processing, Vol 47 No 12, December 1999. A disadvantage of the above location systems is that they are all used with prior knowledge of a transmitting signal’s characteristics before location takes place. Also, the sensitivity of the systems tends to be low, leading to lower location accuracy of the system. Such prior art systems are unfocused in the sense that they make only basic assumptions regarding platform motion, and assume a constant time difference of arrival and a constant frequency difference of arrival between the two receive antennas. It is the aim of the current invention to at least ameliorate the stated problems of the prior art. According to the present invention there is provided a passive surveillance system comprising: a first receiving system having a first receive antenna, and a second receiving system having a second antenna, with at least one of the first and second antennas being moveable in relation to a region of interest; position sensing means adapted to measure the relative positions of the first and second antennas; storage and processing means for storing information received by the first and second receive systems and the position sensing means over a predetermined time interval as at least one of the first and second antenna means moves in relation to the region of interest, and for processing the received information; characterised in that: the processing means is adapted to correlate the stored information obtained using the first receiving system with that obtained from the second receiving system in a focused manner, such that emitters producing a signal of strength above a threshold may be detected. The system is preferably arranged to compute the location of any detected emitters. Such computation of the location is carried out by the process of correlation of the signals received by the first and second receivers, coupled with knowledge of the positions of the platform or platforms on which the first and second antennas are mounted. The system is further preferably arranged to measure transmission characteristics of any emitters detected. These characteristics include, but are not limited to, the frequency, bandwidth, duty cycle, agility and modulation present on the detected signal. Such measurement of the signal characteristics may be done in any suitable manner. The system according to the present invention may be arranged to operate on a single moving platform, with both the first and second antenna being positioned on the platform, or may preferably be arranged to operate using a separate platform for each antenna. Note that herein a focused system is taken to mean a system that performs hypothesis tests against measured signals and the variation of signal difference between antennas with time, that accounts for the substantially precise motion, as measured by the position sensing means, of the surveillance platforms over the integration time period. This allows, in principle, an arbitrary integration time period and spatial resolution to be achieved, subject to the persistence of the emitter, the accuracy of the position sensing means, the accuracy of the syntonisation between the receiver systems and the visibility of the emitter. In contrast to the present invention, the prior art assumes a constant time and Doppler frequency difference which limits the effective integration time of the prior art systems, and hence lowers the sensitivity and spatial resolution obtainable. The present invention has utility at radio frequencies extending up to millimetric wavelengths, and references to “radio”, or “radio frequency" should be construed accordingly. The present invention provides additional sensitivity over the prior art due to the process of storage of received data and subsequent correlation of the data. The invention is particularly beneficial over the prior art when Low Probability of Intercept (LPI) Radar, or where spread spectrum communications signals such as frequency hopping or CDMA systems are employed. The additional sensitivity to such signals allows increased system detection ability and improved detection and location accuracy of the emitters of the signals. Such improvements allow the system to better distinguish between emitters located physically closely together. Longer data collection time intervals as compared to the prior art also provide an improved performance in situations where multipath signals exist. This is because the improved spatial resolution over the prior art permits the different paths a signal takes to be individually identified and characterised. The multipath signal will be delayed as compared to the directly received path, and will also be subject to increased dispersion, and these characteristics can be used in to detect such signals. The storage means is preferably arranged such that it allows for synchronisation and syntonisation of the data received from the first and second receivers. Here, synchronisation is knowledge relating to the relative reception time of a signal received at each receiver, and syntonisation is knowledge relating to the instantaneous frequency at a point in time of signals received at each receiver. Having synchronised and syntonised information allows the correlation process to generate a result having much greater accuracy. The synchronisation and syntonisation process may be achieved by incorporating suitable timing means within each receiving system. Such timing means may comprise an off-air clock such as that provided by the Global Positioning System (GPS). Alternatively, a dedicated radio or optical link between the two receivers may be employed which communicates relative timing information. For greater accuracy an atomic clock may be employed at each receiver. This may be a rubidium clock, or a caesium clock, the caesium clock being the most precise. The higher the frequency being located or detected the more accurate the timing means will need to be, due to the shorter wavelengths involved. As a rough guide, GPS is likely to be adequate up to VHF emitter frequencies, whereas a rubidium clock is likely to be adequate up to frequencies around 1GHz. A caesium clock may be required for frequencies of around 10GHz and above. The position sensing means is required so that platform movement during the pre-determined time interval can be corrected for in relation to the signals received by the receiver systems. The position sensing means is also used to give the position of the platforms relative to the region being viewed, and so provide a point of reference for the resultant data generated by the surveillance system. The measurement of platform movement for the purposes of movement compensation of the received signals is the most critical of these in terms of required accuracy. Accuracy requirements are determined by the wavelengths of the signals within the band of interest. GPS systems are suitable for lower frequencies, and inertial navigation systems are preferred for frequencies of the order 1 GHz or above. The system may preferably be arranged to produce an image showing emitter activity detected over the region of interest, as a function of time. Such an image may be used in combination with a geographic map of the region of interest to conveniently locate the absolute positions of the emitters. A rectilinear image may be produced that is substantially aligned to a geographic map of the region, the alignment being by means of a ground reference in the region. Using this approach, the region of interest may be divided into pixels. The correlation process may then be arranged effectively as a hypothesis testing process. The hypothesis being tested is that a transmitter exists at each pixel location in the image, whilst remaining consistent with the limitations of system resolution. The hypothesis is tested by looking for common signal components in the two receiver outputs that have a phase relationship, over the emission bandwidth and respective receiver positions, that is consistent with an emission source located at each pixel position. Such correlation is sensitive to the variation in relative phase between the two signals due to the knowledge of position provided by the position sensing means. The pixel locations are, for the purposes of computational feasibility, arranged as a two dimensional (2D) surface, preferably arranged to correspond to the shape of the earth’s surface over the region of interest. The pixel spacing on the 2D surface should be chosen so as to adequately sample the spatial resolution achievable with the integration period and surveillance bandwidth employed by the system. Hypothesis tests may be performed on the 2D surface, or at points of intersection of projections from the image plane onto the earth’s surface. The surface on which the hypothesis tests are performed is called the focus surface. The correlation process may also be thought of as a synthetic beamforming process, with the beam being formed by the correlation process rather than by the antenna itself. The correlation process uses the data stored over the predetermined time interval, with this time interval effectively being an integration period of the correlator. The larger this integration period the more directional the beam that is formed, which leads to a higher resolution and smaller pixel size in any image produced using the data. The integration step effectively allows the system to “focus” in on different parts of the region of interest. There are limits on how long this integration time can be however, as inaccuracies within the timing and position sensing means and with other parts of the system provide diminishing returns as the integration time is increased. A typical integration time period achievable with the current invention is 100s or more, depending on the frequencies observed and the limitations of the position sensing means, and processing and storage means. The accuracy requirements of the position sensing means of the two antennas are driven by the needs of the desired absolute location accuracy of the transmitters in the region of interest, and also on the need to maintain data coherence between the first and second receiver systems. This second need is the most demanding, with the accuracy required being proportional to the wavelength of the transmitters of interest within the region, and also dependent again on the integration period. An advantage of the present invention is that the surveillance system is passive in the sense that no energy needs to be directed towards the region of interest. The system relies purely on signals received using the first and second receive antennas. According to a second aspect of the current invention there is provided a method of detecting a radio signal source within a region of interest, the method comprising the steps of: arranging a first receiver having a first receive antenna on a first platform and a second receiver having a second receive antenna on one of the first platform or a second platform; moving at least the first platform in relation to the region of interest whilst recording data received by the first and second receiver systems; combining, and correlating using a processing means the stored data received by the first and second receiver systems in a focused manner to detect radio emitters within the region. The invention will now be described in more detail, by way of example only, with reference to the following Figures, of which: Figure 1 diagrammatically illustrates a typical scenario in which the present invention may be employed; Figure 2 shows a block diagram of a typical system setup for implementation of the current invention; Figure 3 diagrammatically illustrates a simplified arrangement of transmitters and platform receive antennas to show the basic correlation process; Figure 4 is a block diagram showing processing stages used in one embodiment of the current invention; Figure 5 is a block diagram showing processing stages used in a second embodiment of the current invention; and Figure 6 is a block diagram showing more detail of one part of the processing stages with the second embodiment. Figure 1 shows a typical scenario in which the present invention may be employed. A first platform 1 and a second platform 2 are shown in proximity to a region of interest 3. The platforms in this case are both aircraft, although the scope of the invention is such that it also encompasses space based platforms such as satellites and also surface based platforms such as vehicles, ships, or even man-portable systems. Platform 1 has a receive antenna able to receive signals from the region of interest 3 as indicated by antenna angle of reception 4, and platform 2 has a receive antenna able to receive signals from the region of interest 3 as indicated by antenna angle of reception 5. Within the region 3 are a plurality of emitters 6 , the quantity and characteristics of which are initially unknown to the platforms 1, 2. As the platforms 1,2 move in relation to the region 3 they record any received signals within a bandwidth of interest over a pre-determined time interval, this time interval setting the maximum integration time of the system. This time interval will depend upon the required resolution of the system, the speed over the ground of the platforms 1,2, and other factors such as available processing, power and data storage facilities. As an example, an emitter at 1GHz observed from receivers on two individual platforms 10km from the emitter, and travelling at 100m / s around the emitter, and separated by an angle of 60° with respect to the emitter will be observed at 3m resolution using a 10s integration time period. A data link 7 between the two platforms is provided to enable the necessary position synchronisation and syntonisation of the platforms 1,2, and also to transfer data received by the platforms 1,2 as necessary. Figure 2 shows a top level block diagram of the system components mounted on the two platforms 1,2. Here, the first platform 1 is acting as a master, and so is carrying out the processing of the data received from both platforms. Platform 1 has mounted thereon a receive antenna 8, connected to receiver 9. An accurate clock system providing the function of a synchronisation (and syntonisation) source 10 also provides an input to the receiver. The output of the receiver 9 is fed to data storage and processing means 12. Navigation sensor 11 also provide an input to the storage and processing means 12. Such a navigation sensor may comprise an off-air navigation system such as GPS, but other systems such as inertial navigation systems may be used, alternatively or, for improved accuracy, in combination. The accuracy requirements of the navigation means is dependent upon the system integration time. For integration time periods of the order of one second or longer an inertial navigation systems will preferably be used. The second platform 2 has mounted thereon a receive antenna 13, connected to receiver 14. A synchronisation (and syntonisation) source 15 similar to that described in relation to that on platform 1 also provides an input to the receiver 14. The output of the receiver 14 is fed to a data link transmitter 17. Navigation sensor means 16 similar to that described in relation to that on platform 1, and measuring the position of platform 2, also provide data to the data link transmitter 17. The data link transmitter 17 transmits data to a data link receiver 18 on platform 1, from where the received data is passed to the storage and processing means 12. The receive antennas 8,13 need not be narrow beamwidth antennas, as the system achieves its spatial resolution within the beam widths of the antennas. Indeed, a narrow antenna beam pattern is sometimes undesirable, as then some form of antenna stabilisation may be necessary, either to maintain the beam on a fixed point, or to maintain the beam in a fixed direction as the platform on which it is mounted moves. To generate more accurate emitter location information and to provide better resolution, knowledge of the positions of the platforms 1,2 is required in relation to the region of interest, along with the topography of the region. This may be provided by, for example, a GPS system and a digital terrain database such as DTED supplied by the US Department of Defense. The receivers 9,14 provide an intermediate frequency (IF) or a complex (l / Q) baseband signal that unambiguously represents the bandwidth of interest. The receivers 9,14 may comprise multiple receivers or individual Digital Down Conversion channels each arranged to provide coverage over different bandwidths of interest, and the resultant output signals may be combined within the processing means 12. Each receiver is arranged to time stamp the signals from its corresponding antenna, and to supply this timing information to the processor means 12 along with the received data. The timing accuracy of this time stamp is typically of the order of 10ps when the platforms are manned aircraft. The processor means 12 is programmed to carry out a correlation of the data stored in the storage means, and to present the result of the correlation in graphical form on display and control means 19. Figure 3 shows a simple representation of two receive antennas, corresponding to antennas 8, 13 in Figure 2. Also shown are three transmitters positioned in unknown locations within the region of interest. The following is a discussion of the correlation process that is carried out within the processing means 12 (shown in Figure 2), and a general correlation function is derived. Consider an onmi-directional emitter E^, stationary at location ei radiating a signal er). Receive antennas 8,13, at positions ra( / )and rb( / ) receive signals of ai(z) and bi(t) respectively. Ignoring propagation effects such as multipath, dispersion and the scaling of signal magnitudes with range, the signals receive at antennas 8, 13 are: at ( / )=si (r-|ra(r)-ei| / c) ^(r)=5i( / -|rb( / )-ei| / c) Equations 1 where c is the speed of light. Thus the signal received at each antenna 8, 13 is a delayed copy of the radiated from the transmitter. The delay varies slowly with time due to the changing receive antenna position. In general the signal at each antenna will be the superposition of such responses from all transmitters within view of the receive antennas: Sa (0 = £ an (0 = JX (Hro (0 - | / C) n n Equations 2 h W IM n n The ideal operation performed by the processor for each pixel is given by the correlation of the two signals sa(f) and sb(t), as given by the following equation Equation 3 where sa(t) represents the data stream from receiver A as a function of time t sb(t) represents the data stream from receiver B as a function of time t ra(r) and rb(r) are as defined above, ri;j is the spatial position at which the hypothesis test for image pixel i,j is performed sitj is the resulting (complex) value of pixel i,j is the integration period over which the integral is performed. The operation defined by the Equation 3 is performed using signal processing techniques, approximating the integration by a summation. Further approximations and factorisations may be made to allow a good approximation of the specified algorithm to be achieved at low computational cost, although a direct implementation is possible. Any suitable factorisation of Equation 3 may be employed. Figure 4 shows the architecture of one such processing algorithm that may be employed to implement the correlation process. This processor algorithm is optimised for use when the platforms are travelling in near-linear trajectories; the algorithm explicitly compensates for deviations of platform trajectories from nominal linear trajectories. Baseband data from a first receiver on a first platform is represented at 20, as a linear sequence of l / Q samples over the integration time period of the system. This data 20 is partitioned at 21 into two time scales, fast time and slow time. The partitioning operation 21 divides the data 20 into a plurality of segments e.g. 22a of equal length, each typically around 1ms long. The segments 22a-22h are stacked together to form a two dimensional data array 22. Here, fast time represents time within a single segment (the x axis of 22), whereas slow time represents time between segments (the y axis of 22) Baseband data 23 from a second receiver on a second platform is partitioned in partition operation 24, to produce data array 25. This partitioning 24 is similar to that of the partition operation 21, with the exception that each segment e.g. 25a is increased in length with respect to segments e.g 22a, with contiguous segments effectively overlapping each other in terms of the data they contain. This overlap facilitates a good approximation to the exact correlation function shown in equation 3, in the correlation process between a segment e.g 22a from the data array 22 with a segment e.g 25a from the data array 25.. The time interval between segments e.g. 25a is thus made exactly the same as that between segments 22a. The two dimensional data arrays 22, 25 are then correlated with each other in the fast time dimension using a fast correlation algorithm at stage 26. This comprises a forward fast Fourier transform (FFT) of each data array in the fast time dimension, a point to point multiplication of the FFTs of the respective data arrays, followed by an inverse FFT in the fast time dimension. The data array 22 from the first receiver is time reversed in the fast time dimension and complex conjugated before applying this correlation process. The next processing stage is fast time motion compensation 27. This corrects the data in the fast time dimension using a phase shift and a time delay, achieved by interpolation, to make it appear that the data was obtained from platforms each moving along a straight line. The line to which the data is corrected is chosen using a least mean squares fit to the position of the platform as reported in navigation data. The correction applied is varied with both fast time and slow time to achieve the most accurate correction for emitters at any position on the focus surface. The next processing stage is slow time motion compensation 28. This resamples the data in the slow time dimension using interpolation. Data is sampled at constant angular increments of the bisector of the two platform positions about the region centre. The final processing stage is a passive bistatic aperture synthesis processing engine 29. This carries out the final stage in the correlation process, effectively carrying out a full set of hypothesis tests as described earlier in a single set of operations, without having to do individual hypothesis tests for each pixel. This exploits the regularity in the tests applied to adjacent pixels arising from the regular structure of pixels in the image, and the linear platform trajectories to which the data has been pre-corrected. The processing is done using any suitable factorisation, such factorisations are analogous to, and based on, those employed in synthetic aperture radar systems, but are modified to accommodate the passive nature of the current invention, and the use of dual platforms. Suitable base algorithms are, for example, the Polar Reformat algorithm, the Range Migration algorithm, and the Tomographic Back Projection algorithm. More details of these can be found in W.G Carrara et al, “Spotlight Synthetic Aperture Radar”, Artech House, 1995. The output of the processing engine 29 is an image 30 showing any detected emitters and their locations in the focus plane. 5 Figure 5 shows the architecture of an alternative processing algorithm to that presented in Figure 4, that may be employed to implement the correlation process. This processor algorithm is adapted for use with arbitrary platform trajectories, employing no explicit motion compensation stage, but instead is intrinsically matched to arbitrary trajectories. The stages of this processing algorithm are identical to that of the previously presented algorithm up to and including the fast time correlation stage, and reference numerals are shared between Figures 4 and 5 to indicate the similar stages. The motion compensation stages are omitted as the passive bistatic 15 processing engine 31 incorporates an algorithm that is able to perform processing for arbitrary platform motion. Figure 6 details the processing that occurs within the processing engine 31. A 20 four stage process is employed. The first stage 32 is an FFT in the fast time dimension to obtain data sampled as a function of slow time ts and fast time angular frequency ojf. 25 The second stage 33 mixes out the region centre phase response by multiplying the 2-D data by a region centre reference function: where ra(rj and rb(rj are the positions of the two surveillance platforms, measured as a function of slow time relative to the region centre. The third stage 34 interpolates the data into a new space (kx,ky) according to a reference mapping: s(ts,a)f)^ s(kx,ky) defined by where and are the projections of ra(rj and rb(rj respectively in the focus plane. This reference mapping optimises the focusing of the system about the scene centre to a first order of precision. This mapping is not 5 limited to a particular system configuration or platform trajectory, and so allows single platforms to be used. Note that a third component, kz\s a component not in the focus plane and is therefore ignored. The final stage 35 is a 2-D inverse FFT to obtain the image ^(x, j), defined in 10 the co-ordinates of the focus plane. This produces the final image 36 shown in Figure 5 The skilled person will be aware that other embodiments within the scope of 15 the invention may be envisaged, and thus the invention should not be limited to the embodiments as herein described.

Claims

1. A passive surveillance system comprising:a first receiving system having a first receive antenna, and a second receiving system having a second antenna, with at least one of the first and second antennas being moveable in relation to a region of interest;position sensing means adapted to measure the relative positions of the first and second antennas;storage and processing means for storing information received by the first and second receive systems and the position sensing means over a predetermined time interval as at least one of the first and second antenna means moves in relation to the region of interest, and for processing the received information;characterised in that:the processing means is adapted to correlate the stored information obtained using the first receiving system with that obtained from the second receiving system in a focused manner, such that emitters producing a signal of strength above a threshold may be detected.

2. A surveillance system as claimed in claim 1 wherein the system is arranged to use information produced by the correlation process to locate the position of detected emitters.

3. A surveillance system as claimed in claiml or claim 2 wherein the system is arranged to extract transmission characteristics of any detected emitters.

4. A surveillance system as claimed in any of claims 1 to 3 wherein the processing means is adapted to produce an image indicating the geographic positions within the region of interest of any transmitters detected.

5. A surveillance system as claimed in claim 4 wherein the image is a rectilinear image aligned to a geographic region using at least one ground reference.

6. A surveillance system as claimed in any of claims 1 to 5 wherein the processing means is adapted to provide, from the information from each receiver, a two dimensional array of data associated with each receiver, by partitioning stored information from each receiver into a plurality of segments that are stacked into the two dimensional array, the array having a fast time axis and a slow time axis, and the processing means being further adapted to correlate the arrays associated with each receiver to produce a fast-time correlated array.

7. A surveillance system as claimed in claim 6 wherein the two dimensional array of data associated with one of the receivers is arranged such that a subset of data within a single row of the array corresponding to the fast axis is common with that of a contiguous row.

8. A surveillance system as claimed in claim 7 wherein the processing means:includes means for generating a fast Fourier transform (FFT) of the fast-time correlated array;includes means for multiplying the FFT by a scene centre reference function comprising the expected phase difference, between the first and second antennas, of an emitter located at the centre of the region of interest;includes means for interpolating the multiplied FFT to apply a reference mapping to correct for emitters not at the centre of the region of interest; and includes means for applying an inverse FFT to the interpolated, multiplied FFT.

9. A surveillance system as claimed in claim 8 wherein the means for interpolating is adapted such that the reference mapping is chosen to optimise the focusing of the system to a first order of precision about the scene centre.

10. A method of detecting a radio signal source within a region of interest, the method comprising the steps of:arranging a first receiver having a first receive antenna on a first platform and a second receiver having a second receive antenna on one of the first platform or a second platform;moving at least the first platform in relation to the region of interest whilst recording data received by the first and second receiver systems;measuring the movement of the at least first moving platform with respect to the region of interest,combining, and correlating using a processing means the stored data received by the first and second receiver systems in a focused manner to detect radio emitters within the region.

11. A method as claimed in claim 10 wherein transmission characteristics of any detected emitters of any detected emitters are extracted by the system.

12. A method as claimed in claim 10 or claim 11 wherein the correlation process locates the position of detected emitters. / 13. A method as claimed in 12 wherein the processing means is arranged to:a) provide, from the information from each receiver, a two dimensional arrayof data associated with each receiver, by partitioning stored information from each receiver into a plurality of segments that are stacked into the two dimensional array, the array having a fast time axis and a slow time axis, wherein an array associated with one of the receivers is arranged such that a subset of data within a single row of the array corresponding to the fast axis is common with that of a contiguous row, and to produce a fast-time correlated array by correlating the two arrays in the fast-time dimension;b) generate a fast Fourier transform (FFT) of the fast-time correlated array;c) multiply the FFT by a scene centre reference function comprising the expected phase difference, between the first and second antennas, of an emitter located at the centre of the region of interest;d) interpolate the multiplied FFT to apply a reference mapping to correct for emitters not at the centre of the region of interest;e) produce an inverse FFT of the interpolated data to generate an image of the locations of detected emitters.

14. A method as claimed in claim 13 wherein the interpolation step c) has a 5 reference mapping chosen to optimise the focusing of the system to a first order of precision about the scene centre.

Images