Back...

Synthetic aperture sonar
H.D. Griffiths, S. Dugelay, S. Chapman*, R. Voles, C.F.N. Cowan+ and T.A. Rafik**

Synthetic aperture techniques can also be applied to sonar systems, where they have applications in surveying of oil pipelines, wellheads and wrecks, and in minehunting. There are many similarities to synthetic aperture radar, but some differences, mostly associated with the much lower velocity of propagation.

A current study is examining motion compensation techniques, using both IN information and autofocus processing, and developments such as squint-mode and spotlight-mode processing. There is also a conflict between maximum unambiguous range and proper sampling of the synthetic aperture, which may be tackled by a number of approaches, and the relative merits of these are being investigated. An experimental system for sea trials is currently being designed.

A previous study showed how interferometric (bathymetric) techniques could be used, in the same way as with interferometric SAR, to obtain three-dimensional images of the target scene. A practical system has been designed and built to operate in the tank at Loughborough, and has been used to obtain 3-D images of a number of test targets. This is able to introduce controlled platform motion errors and has been used to provide experimental verification of the motion compensation algorithms.

*DERA Portsdown
+Queen's University, Belfast
**Loughborough University
Project support: DERA (Bincleaves), EPSRC

Texture characterisation in high-resolution sonar images
H.D. Griffiths, J. Dunlop, R. Voles, M. Di Bisceglie*, T. Stathaki+ and C. Naldi*

In analysis of high-resolution sonar imagery of the seabed, it is important to properly model the image texture in order to be able to detect objects which do not conform to this model.

This work attempts to apply statistical models such as the compound K-distribution to sonar imagery. This is represented as the product of a Rayleigh speckle component and the square root of a gamma distributed modulation process. Such models have been successfully applied in the past to high-resolution maritime radar clutter and to airborne SAR imagery of land target scenes. Previous workers have modelled a single sonar image of sand ripples by the K-distribution, but it does not appear to have been applied to a full range of sediment types.

Images from a number of different sidescan sonar systems have speckle that is normally distributed with non-zero mean. The figure below shows that the Rayleigh model provides a poor fit to these data. Modelling the speckle with the Ricean distribution gives the generalised K-distribution, which provides an excellent fit to all of the sonar data tested. A number of possible physical explanations for the Ricean speckle have been proposed.

The K-distribution also allows the spatial correlation properties to be modelled, and a procedure for the estimation of the parameters of the bidimensional generalised K-distribution has been devised. Detection algorithms based on autoregressive (AR) models whose parameters are estimated through joint second and third order statistical techniques have been developed and demonstrated.

*University of Naples
+Imperial College
Project support: DERA (Bincleaves), EPSRC (CASE), CNR (Italy)



First-order statistics of speckle from a sidescan sonar image of a muddy seabed with Rayleigh distribution (dashed line) and Ricean distribution (solid line).

An IR spectral signature detector for naval surveillance
H.D. Griffiths, D.R. Selviah and R.C. Coutinho

In this project we investigate a technique for detection of low-radiance targets embedded in a brighter maritime clutter, exploring their temporal coherence properties. The approach uses an interferometer to measure the coherence profile (interferogram) of a scene in the field of view. As the interferogram is the Fourier Transform of the spectrum, sharp bandpass filtering of this radiation gives it a sinc envelope. The aim is to optimise the system to have a maximum change in the interferogram, when a narrow spectral feature from a target comes into the field of view. Clutter rejection ratios of up to 115 dB were reported in previous work by Sutton.



Coherence profile of a 20 W tungsten halogen lamp.

AIR TRAFFIC CONTROL SYSTEMS


Aeronautical telecommunications network development
K. Woodbridge, P. Lane and F. O'Connell

The International Civil Aviation Organisation (ICAO) is currently standardising a data communications environment for the aeronautical community. This environment, known as the Aeronautical Telecommunications Network (ATN), is being designed to allow fixed and mobile, ground to ground and ground to air links. This involves a terrestrial backbone network as well as air to ground subnetworks based on VHF datalink, secondary radar and satellite communications systems. The availability of the ATN in the communications infrastructure is essential to realise the full benefits of datalink applications which will form the backbone of next generation integrated air traffic control systems. This project aims to develop a set of evolutionary paths for future aeronautical networks. Consideration is being given to physical topologies and techniques such as modulation schemes, multiple access methods, provision of network capacity, required communications protocols and future upgradability. Capacity, coverage and integrity are also critical issues in a safety critical function such as ATC. The project is closely guided by the current developments in air traffic management systems and will link with other UCL research on mobile communications systems, air traffic control radar and satellite communications.

Project support: National Air Traffic Services Ltd

Azimuth errors in air traffic control secondary surveillance radar
K. Woodbridge and L. Vinagre

NATS currently operates some 20 secondary surveillance radar (SSR) stations throughout the UK for monitoring air traffic in UK airspace. Recorded data from these sites has shown significant deviations in positional accuracy from some stations. These target azimuth errors have been traced to multipath reflections from obstacles close to the radar head. The current methods used to estimate SSR target azimuth errors due to obstacle shadowing are based on work carried out for the FAA more than twenty years ago. The object of the study is to review and update these procedures and to produce an easily usable manual system of calculating azimuth errors for field use by NATS engineers. This will be important when considering the effects of any new developments or obstructions on current or planned SSR sites. In addition it is proposed to study the possibility of developing fully automated methods of obtaining azimuth error data caused by any type of obstacle. The research will involve investigation of electromagnetic field profiles at the antenna using fast Fourier Transform Techniques and the effect of various obstacle types on both secondary and some primary radar returns. Consideration of the possible effects on SSR datalinks will also be considered.

Project support: National Air Traffic Services Ltd



National Air Traffic Services' air traffic surveillance radar at Clee Hill (photo courtesy of the Civil Aviation Authority).

Satellite and radar datalinks for air traffic surveillance and control
K. Woodbridge and A. Toema

The automatic dependent surveillance (ADS) system is a new air traffic control (ATC) technique that uses datalink for aircraft position reporting. ADS forms part of the overall aeronautical telecommunications network environment being standardised for the future global air traffic management system. Position data may be derived from on board aircraft flight management systems or from the Global Positioning System (GPS). This position data is downlinked and then routed via ground networks to air traffic control centres enabling efficient air traffic management even in remote and oceanic areas out of range of ground based radar surveillance. The datalink can be made via VHF, Mode S secondary radar or using a satellite communications (SATCOM) system. This project is examining new modulation and access techniques such as code division multiple access (CDMA) and their ability to meet the high levels of integrity and reliability needed for ATC datalinks. In addition we are exploring the benefits of using the newer low earth orbit (LEO) satellite constellations for ADS SATCOM applications in addition to the established INMARSAT links. The Mode S secondary radar datalink modulation and access techniques will also be considered at a later stage in this work.

Project support: Eurocontrol Brussels

Phased array multifunction air traffic control radar
H.D. Griffiths and K. Woodbridge

Current air traffic control (ATC) radars are largely based on technology developed 20-30 years ago using rotating antenna, single frequency systems with several separate units required at a major airport. Such systems will eventually be unable to cope with fast update multi-tasking which will be required by the continued rapid growth in air traffic. This project is therefore a feasibility study into the potential of more advanced solid state active array radar for multi-function ATC applications. This work will also cover radar data integration into the local airport data system and the global aspects of the connection with the overall ADS and CNS systems including the Aeronautical Telecommunications Network. An initial study has resulted in 2 and 4 faced antenna designs combining airfield approach, secondary surveillance and ground movement radars with beam patterns optimised for each individual task. Advanced work on array thinning has resulted in a significant reduction in the number of elements required. An estimate of costings based on the latest information from manufacturers suggests that this system may be comparable in cost to a typical new conventional ASR. Further studies on the overall radar system design including power budget is underway.


ANTENNAS AND MICROWAVES

Accuracy and resolution of superresolution methods
H.D. Griffiths, D.H. Brandwood* and D. Salter*

Superresolution is a term applied to a group of parameter estimation techniques using multiple signal channels with digital signal processing. In general the parameters of more than one point source of data can be estimated from a single set of data and the resolution achievable between sources is better than that given by standard methods. In particular these methods are applicable in the radar, radio or sonar fields for signal, or target, direction finding (DF), where the parameters required are the azimuth and elevation of each signal. In this application the processing is applied to the outputs of an array of antenna elements or other sensors.

Superresolution techniques may be divided into two main groups. MUSIC is a well known example of the spatial spectrum methods and IMP is an example of maximum likelihood methods. The performance of MUSIC, in particular, has been analyzed when limited by receiver noise, assuming perfect knowledge of the array and its response as a function of direction. In practice the limitation in performance is often the quality of the array description, or calibration, and the aim of this work is to determine the performance, in terms of both the signal resolution and parameter estimation accuracy, that is achievable, statistically, with given array description errors. Both MUSIC and IMP will be considered.

*Roke Manor Research
Project support: Roke Manor Research

AC properties of composite material systems
H.D. Griffiths, I.J. Youngs*, A.L. Cullen and A.C. Lynch

The AC properties of composite materials and the physical factors which influence them are being studied to enable the optimised design of materials with controlled electromagnetic response at RF and microwave frequencies. In the first instance, the composite materials to be studied are based on dispersions of a conducting material (metal, carbon and conducting polymers) in an insulating host matrix. It is probable that both 2- and 3-dimensional systems will be of interest. An improved theoretical representation will be sought, particularly for strongly interacting (i.e. near percolation) systems which are anticipated to exhibit more novel properties. Observation of plasma-like phenomena at frequencies substantially below optical frequencies would be a considerable achievement.

*Structural Materials Centre, DERA Farnborough
Project support: DERA Farnborough

Microwave forceps
H.D. Griffiths and A.L. Cullen

A fundamental procedure in surgery is the joining or closure of blood vessels. The present procedure involves the use of staples, which hold the tissues together until a natural healing process occurs and the tissues form a permanent weld. The use of staples, although an advance on the use of nylon thread, is not entirely satisfactory, since the staple is eventually rejected. A technique which avoids the use of physical means such as staples would obviously be preferable. It is known that if the temperature of the tissue can be raised to about 55 degrees C, natural sealing will take place. Heated forceps have been tried, but adhesion of the tissue to the tips of the forceps is then a problem.

We have therefore conceived the idea of microwave forceps, which apply a controlled amount of microwave power to the tissue that is to be joined or sealed. The frequency is 2.45 GHz. The power is fed via miniature coaxial cable to the tips of the forceps.

An analysis has been developed to predict the temperature distribution within the tissue as a function of spatial coordinates, time, and power level, and this is being used to determine the optimum applied power as a function of time. An early experimental system (using power from a microwave oven source) has given encouraging results, and trials are about to begin with a new system with better control of the power.

Project support: Royal Society

Modulated scatterer techniques in antenna measurement
P.V. Brennan, H.D. Griffiths and R. Benjamin

The modulated scatterer antenna measurement technique was originally devised at UCL in the 1950s in the form of spinning dipoles. The technique measures the field radiated from an antenna or scattered from a target using a scatterer which may be either mechanically or electrically modulated with synchronous detection of the modulated scattered signal.

The current work aims to extend the scope of the original method by considering the use of an array of modulated probes, each of which may be modulated at a different frequency. The probe outputs may be summed and then Fourier analysed to deduce the field present at each probe position. This has the principal advantage of improved measurement speed. A separate development is the investigation of sparsely-sampled arrays of probes. The near-field or aperture field regions of most antennas are very well behaved and contain surprisingly little information. This fact may be exploited to enable a reduction in the sample spacing of any near-field probing technique. We have already found that a considerable reduction in probe spacing is possible with little impact on measurement accuracy. This method can be extended to the sparse, possibly quasi-random sampling of an antenna in two dimensions which would allow the measurement of antenna patterns in both planes without any assumption of separability.

Project support: British Aerospace

Non-linear dynamics approach to characterising circuit disruption
P.V. Brennan and R.J. Bullock

The phenomenon of chaos in circuits has received much attention over recent years, from a largely analytical viewpoint. This work, however, aims to investigate the optimum signals required to actually cause chaos and general disruptive behaviour in circuits. Our role has been to develop test procedures and a number of circuit elements, particularly PLLs and AGC loops, with which to produce chaotic behaviour and observe and record its effects. The work programme involves two other groups of researchers. The University of Dundee are investigating optimum waveforms for causing chaos in PLL and AGC circuits whilst UMIST analysing signals recorded in various parts of actual generic circuits (developed by UCL) which have been subjected to the specified waveforms. The results may subsequently be used to test commercially available equipment. The overall aim of the programme is to determine whether insight gained from individual sub-circuits can be used to generate optimal waveforms to disrupt complete systems. Our results have shown good correspondence between measured and modelled behaviour and have confirmed the ability of an optimum waveform to most readily cause non-linear and chaotic behaviour.

Project support: DERA (Fort Halstead)

Agile synthesiser studies for mobile telecomms
P.V. Brennan, R. Walkington

Mobile phone services are moving towards spread-spectrum techniques involving fast frequency hopping. This relies on highly agile synthesised local oscillators which may need to change channel within a few micro-seconds. There are several well known methods with which this may be accomplished, such as fractional-N synthesis, but these currently lack the potential for simple, reliable and economical implementation.

This work, performed in close association with industry, is a study of the presently available frequency synthesis techniques and the possible development and integration of new ones to fulfil the demanding technical and economic requirements of future mobile communications. A particularly promising candidate is the sigma-delta fractional-N technique which we are planning to develop into an optimum waveform which may be stored in memory and clocked-out at a very fast rate, allowing loop operation at very high reference frequencies.

Project support: EPSRC (studentship), Nokia Telecommunications Ltd