See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225662626

A review of GPR for landmine detection

Article  in  Sensing and Imaging · September 2006

DOI: 10.1007/s11220-006-0024-5

CITATIONS READS

108 4,858

1 author:

David Daniels
The University of Manchester
74 PUBLICATIONS   4,621 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Minehound View project

2014 IEEE International Conference on Antenna Measurements & Applications View project

All content following this page was uploaded by David Daniels on 04 January 2019.

The user has requested enhancement of the downloaded file.

 Page 1 of 11

UWB RADAR FOR THE DETECTION OF BURIED ORDNANCE

David J Daniels,
ERA Technology Ltd,
Cleeve Road,
Leatherhead,
Surrey KT22 7SA,
United Kingdom
Tel ++ 44 (0) 1372 36 7084 Fax ++ 44 (0) 1372 367138,david.daniels@era.co.uk

Abstract— the detection of buried ordnance, IED and landmines has proved to be a successful application of
UWB radar. UWB radar techniques are being routinely researched for these applications and systems are in
production. This paper will review the technology that has been implemented in production systems and
assess current research into hand-held, stand-off vehicle-based radar systems and airborne systems. These
three distinct modes of operation pose fundamentally different challenges in terms of the physics of
propagation and the radar system design. The paper will also highlight the future engineering challenges to
achieve not only detection using UWB radar but also recognition and identification.

Background

Landmine detection using UWB radar has made considerable advances since the first work carried out in the
1970’s and early 1980’s by researchers in the UK and US. The early work revealed many of challenges
posed by the basic physics of propagation and set the foundation of understanding. The impetus for further
development was given by the Falklands conflict and the UK pushed the physics and technology further in
work carried out at ERA Technology for the UK MoD.

However, the end of that conflict and the resources needed to develop the technology were key factors in
reducing the priority given to further research in the UK. The aftermath of the Gulf War in the early 1990’s
caused the military to renew its interest in the mid 1990’s. A growing awareness on the part of the military of
the vulnerability of soldiers and vehicles to what is now termed asymmetric warfare resulted in resources
being channelled into military research programmes in the US, UK, Germany, France and Russia etc,
particularly for vehicle based sensors to detect landmines and IEDs.

UWB radar challenges for landmine detection

The severe propagation path losses encountered by electromagnetic waves in soil result in operation where
the wavelengths radiated are greater than, or of the same order of magnitude as, the dimensions of the
landmine. Conventional radar systems operate at wavelengths where the target dimensions are generally
much larger than the wavelength of the incident radiation, i.e. the optical region, whereas UWB radar
operates between the Rayleigh and Mie (or resonance) region. The total path losses within a few wavelengths
may be as much as 100dB depending on the material and, as UWB radar systems do not have a total loop
gain much in excess of 120dB, the designer has a major challenge to detect landmines signatures within very
short time ranges of typically 20ns.

Scattering of electromagnetic energy from a landmine results from the impedance differences of the
landmine compared with the host material. The ground contains significant levels of radar scatterers such as
stones, animal burrows as well as man made detritus and, consequently radar encounters extremely high
levels of clutter at short ranges. Overcoming this as well as achieving an adequate signal/noise ratio is a
major technical challenge. Daniels [1] provides further details of the design challenges associated with
landmine detection.

© Copyright ERA Technology 2008

 Page 2 of 11

Canonical targets such as cylinders, which are similar to landmines, have well understood free space
scattering characteristics that will be modified by the dielectric properties of the soil. The mine may have a
number of scattering centres, each with their own angular radiation pattern and, in the case of plastic
landmines; the internal structure of the mine may generate additional scatterers.

Most minimum metal landmines may be considered as multiple layered dielectric cylinders, each interface
causing a reflection, the impact of the small internal metallic fuse being minimal. The landmine is
surrounded by soil, which is a lossy dielectric whose relative dielectric constant depends mainly on the water
content. Typically relative dielectric constant of the soil varies from 3 in dry sand to greater than 16 in wet
and waterlogged soils.

Table 1 Relative dielectric constants of explosives

Substance Name Relative

Dielectric
Constant

TNT 2,4,6-Trinitrotoluene 2.70

Detasheet PETN 2.72
PETN Pentaerythritol tetranitrate 2.72
Comp B RDX TNT 2.90
Octol HMX TNT 2.90
Tetryl 2,4,6-Trinitrophenyl-N-methylnitramine 2.90
Semtex-H RDX-PETN 3.00
HMX Cyclotetramethylene-tetranitramine 3.08
Comp C-4 RDX 3.14
RDX RDX Hexahydro-1,3,5-trinitro-1,3,5-triazine 3.14
AN Ammonium nitrate 7.10
NG Nitroglycerin 19.00

The explosive used in landmines is typically nitrogen based with a relative dielectric constant of between 2.7
and 3.5, ammonium nitrate being a notable exception as shown in Table 1.

Landmines can also be found in fresh water, which has a relative dielectric constant of approximately 80 but
a very low loss tangent; hence, it is quite feasible to detect landmines in fresh water or soils saturated in fresh
water, which also has the benefit of increasing impedance contrast. Salt water on the other hand completely
attenuates radar signals. It should be noted that the ground and surface are quite likely to be inhomogeneous
and contain inclusions of other rocks of various size as well as man-made debris. Thus the signal to clutter
performance of the radar is likely to be an important performance factor. Clutter may be regarded as any
radar return that is not associated with the wanted landmine and needs to be defined with respect to a
particular application.

Hand-held UWB radar systems use separate, man-portable, transmit and receive antennas, which are placed
just above the surface of the ground and moved in a known pattern over the surface of the ground under
investigation. This generates, in real time, data or an image. By systematically surveying the area in a regular
pattern, a radar image of the ground can be built up. Alternatively, the equipment may be designed to provide
an audible warning of target presence while the antenna is swept over the ground. Vehicle based or airborne
systems use much larger linear arrays of antennas to illuminate a swathe of the ground surface ahead of the
platform and rely on the movement of the vehicle to create the data, which may be processed using SAR
techniques.

Close-in or proximal UWB radar can be operated so that the antenna is very close to the ground surface and
target such that the energy transfer is predominantly either induction or quasi-stationary (i.e., in the near
field). Some workers have reported detection by means of evanescent wave propagation. Stand-off UWB

© Copyright ERA Technology 2008

 Page 3 of 11

radar systems can be operated such that the energy transfer is in the far field region and this, in turn, brings
challenges of energy transfer and above-ground clutter rejection.

GPR technology

Introduction

Although alternative techniques (such as acoustics, NQR, radiometry and EMIS, etc) have been proposed,
UWB radar has emerged as the technology of choice for improved landmine detection. Recent developments
using hand-held dual sensor technology combining electromagnetic induction EMI and ground penetrating
radar (GPR) have enabled improved discrimination against metal fragments to be demonstrated in live
minefields and reductions of up to 7:1 compared with the standard metal detector have been achieved in the
field by hand held systems. Vehicle-based systems have been developed that use arrays of antennas and
generate 3-D data, which is then processed to provide a rolling map of detections.

The signal and image processing options for vehicle-based landmine detection are more extensive because
the radar and its platform generate 3-D data. In general, vehicle-based systems concentrate on anti-tank
landmines because it is difficult to achieve adequate cross range resolution at realistic budgets. Options for
signal and image processing include image inversion and synthetic aperture techniques for image
enhancement.

Modulation techniques

Amplitude modulation

The majority of UWB radar systems use impulses of radio frequency energy variously described as baseband,
video, carrier less, impulse, monocycle or polycycle. A sequence of pulses, typically of amplitude within the
range between 20 V to 200 V and a pulse width within the range 200 ps to 50 ns at a pulse repetition interval
of between several hundred microseconds to one microsecond, depending on the system design, is applied to
the transmit antenna. The output from the receive antenna is applied to a flash A/D converter or a sequential
sampling receiver.

The principle of the sampling receiver is therefore a down conversion of the radio frequency signal in the
nanosecond time region to an equivalent version in the micro or millisecond time region. The
incrementation of the sampling interval is terminated at a stage when, for example, 256, 512 or 1024
sequential samples has been gathered. The process is then repeated. The challenge with most of these
sample and hold circuits is the limited dynamic range of the sampling diodes which is in the order of 50-
60dB at best. The poor noise figure of the sampling gate can be improved by using a wideband low noise RF
amplifier prior to the gate The typical noise figure of a 1 GHz amplifier is 2.4 dB hence an immediate
improvement in system noise figure is achieved. However, the sampling gate may now be vulnerable to
saturation by high level signals caused by targets at very short ranges.

The stability of the timing increment is very important and generally this should be 10% of the sampling
increment; however, practically, a stability in the order of 10 ps to 50 ps is achieved. The effect of timing
instability is to cause a distortion, which is related to the rate of change of the RF waveform. Evidently,
where the RF waveform is changing rapidly, jitter in the sampling circuits results in a very noisy
reconstructed waveform. Where the rate of change of signal is slow, jitter is less noticeable. Normally
control of the sampling converter is derived from a sample of the output from the pulse generator to ensure
that variations in the timing of the latter are compensated automatically.

Frequency domain radar

An FMCW radar system transmits a swept carrier frequency by means of a voltage-controlled oscillator
(VCO) over a chosen frequency range on a repetitive basis. The received signal is mixed with a sample of
the transmitted waveform and results in a difference frequency which is related to the phase of the received

© Copyright ERA Technology 2008

 Page 4 of 11

signal, hence, its time delay and, hence, the range of the target. The difference or intermediate frequency (IF)
must be derived from an I/Q mixer pair if the information equivalent to a time domain representation is
required as a single ended mixer only provides the modulus of the time domain waveform. An inverse
complex frequency-time transform can reproduce a time domain equivalent to the impulse radar but most
frequency domain radars produce amplitude only. The effect of windowing the IF waveform is significant
and the unwindowed case gives rise to the well-known sinc (sin (x)/x) function. This limits the dynamic
range of the receiver, whereas a windowed case can potentially achieve a better dynamic range, albeit at the
disadvantage of reduced resolution. An FMCW radar system is particularly sensitive to certain parameters.
It requires a high degree of linearity of frequency sweep with time to avoid spectral widening of the IF and,
hence, degradation of system resolution. Practically a useful system should aim to keep all non-linearities
less than 0.1%.

A stepped frequency continuous wave radar radiates a sequence of frequencies over a defined bandwidth in
regular increments. As any repetitive pulsed signal can be transformed to a frequency domain representation,
which consists of line spectra whose frequency spacing is related to the pulse repetition rate and whose
envelope is related to the pulse shape. Hence, a repetitive impulsive waveform can be synthesised by
transmitting a sequential series of individual frequencies whose amplitude and phase is accurately known.
The main advantages of the stepped frequency continuous wave radar are its high dynamic range (>150dB)
and low noise floor as well as the ability to avoid certain frequencies when transmitting, thus making
compliance with licensing requirements much easier than most other types of modulation. In the case of a
synthesised radar, simple anti-aliasing filters cannot remove contributions to the received signal outside the
ambiguous range. It is therefore important to choose operating parameters that minimise the aliasing effect.
One method is to iteratively determine the range gate of the target. The initial measurement is taken with
low resolution and large ambiguous range. The range to the target is determined and the resolution is then
increased.

Normally the radar is calibrated both to establish a reference plane for measurement as well to reduce the
effect of variation the frequency characteristics in components and antennas. The radar system will
introduce additional phase shifts on the transmitted and received system as a result of the electrical lengths of
the signal paths to the antennas and the effective radiation phase centre of the antennas. This means that the
phases of the transmitted and received signals will be different at each integer frequency and will require
compensation.

Noise modulation

Noise modulated radar offers some very attractive possibilities to the designer of UWB radar systems. The
radiated power is evenly spread throughout the spectrum and the receiver is less susceptible to interference.
However, until recently, such systems were relatively rare. Developments over the last few years are
changing that situation and more efforts are being put into their development. The basic principle of
operation of noise radar is that of a correlator. The radar transmits a noise signal and the received signal is a
time delayed version τs of the transmitted signal. In the receiver, the transmitted signal is used via a variable
delay τv to cross correlate the received signal. The key issues with noise modulated radar concern the
integration time and the fundamental noise floor of the system as well as the sidelobes of the correlation
function.

Handheld Landmine Detection

Recent developments using hand held dual sensor technology have enabled improved discrimination against
metal fragments to be demonstrated in live minefields and reductions of up to 7:1 compared with the
standard metal detector have been achieved in the field by hand held systems such as MINEHOUND,
Daniels [2] and AN/PSS-14, Doheny [3].

Handheld landmine systems are more limited in the signal processing algorithms that can be applied because
they usually only have a single transmit-receive antenna pair and with only a few exceptions, do not form an
image. research into landmine discrimination based on the analysis of A-scans by means of complex
© Copyright ERA Technology 2008
 Page 5 of 11

resonances, wavelets, time- frequency characteristics, neural networks, fuzzy sets, Gaussian mixture models,
order statistics and template matching, has been carried out and methods based on time-frequency
characteristics are reported by Wong [4], Lopera [5], as well as Daniels [6] who showed the feasibility of
discriminating between AP landmines and typical false landmines on a small data set. An example of an
UWB radar system is provided by MINEHOUND and is described in more detail in the following section.

The UWB radar module is comprised of two main sections; these are the RF and the digital sections as
shown in Figure 1.

Figure 1 Minehound radar module system diagram

Sampling is carried out by sequentially incrementing the sample time position each pulse repetition interval
up to 512 samples and then repeating the process. For example, a sampling increment of 50 ps is added to
the previous pulse repetition sampling interval to enable sampling of the received signal at regular intervals.
Using a digitally generated slow ramp and analogue generated fast ramp to create the sequence generates the
incrementally timed samples.

The radar software runs on an Analog Devices Blackfin™ DSP and is responsible for data acquisition at the
full data rate of 16Mbs-1, the updating of timing pulses to the digital to analogue converters every 32µs and
general housekeeping tasks. In addition to this, the DSP runs the signal processing algorithms which convert
the radar data to audio every 16 ms. Further tasks include communication with other devices which are
connected to the radar board. These can include devices such as metal detectors (in the case of
MINEHOUND™) and PDA devices. The radar consumes 2.3W and contains all the processing needed to
provide audible warning of a landmine target as shown in Figure 2.

© Copyright ERA Technology 2008

 Page 6 of 11

Figure 2 MINEHOUND Radar transmitter receiver and processor

VEHICLE BASED UWB RADAR

Vehicle based systems have been developed that use arrays of antennas and generate 3-D data, which is then
processed to provide a rolling map of detections. The signal and image processing options for vehicle based
landmine detection are more extensive. In general, vehicle based systems concentrate on anti-tank landmines
because it is difficult to achieve adequate cross range resolution at realistic budgets. Options for signal and
image processing include image inversion and synthetic aperture techniques for image enhancement,
principal component analysis (PCA), independent component analysis (ICA) techniques and hidden Markov
models.

Although a number of vehicle based GPR systems have been trialled and reported on over the last 5 years the
most extensively reported system has been the NIITEK radar shown in Figure 3.

Figure 3 NIITEK radar system

An extensive evaluation of four different landmine discrimination algorithms applied to data collected with a
NIITEK vehicle-mounted radar system over approximately 42,000 m2 of ground is reported by Wilson [7].
The four algorithms evaluated comprised: a hidden Markov model (HMM), GEOM a feed-forward-order-

© Copyright ERA Technology 2008

 Page 7 of 11

weighted average (FOWA) network to discriminate between landmines and clutter, SCF a frequency analysis
based on comparison of similarity features and EDS a fuzzy K-nearest neighbour (K-NN) algorithm to
generate a mine confidence level.

Work by ERA Technology resulted in the MINDER CAP technical demonstrator. The architecture of the
MINDER CAP system is based around 16 receivers each of which sequentially sample the signal incident on
receive antenna elements. The transmitters are synchronised by adjacent receivers and a central master clock.
The overall system was developed initially as part of an ERA private venture development programme and
subsequently for the UK MoD MINDER CAP programme. An example of the image output of the prototype
array radar system is shown in Figure 4. The system was successfully trialled at DRES in Canada and
subsequently in the UK.

Figure 4 Typical radar map output from the MINDER radar system
The complete MINDER system is shown in Figure 5.

Figure 5 MINDER CAP Technology demonstrator vehicle

A recent development by ERA Technology is an 8-channel radar system as shown in Figure 6. Each radar
channel operates as a self-contained module and is triggered by an interface board, which ensures that each
pulse is transmitted in its own pre-assigned time slot. The data from each radar module is concatenated and
fed by a USB interface to a laptop computer. The pulse repetition interval of one microsecond and the
sampling window of 25ns of each module define the timing of the radar boards. This enables a theoretical
maximum of 40 time slots but practically a maximum 32 time slots is available, allowing for a guard band
around each.

© Copyright ERA Technology 2008

 Page 8 of 11

Figure 6 Complete 8-channel radar system

The complete radar system includes the antenna array, radar system, laptop interconnecting cables and well
as a shaft encoder to provide position information. The rack electronics is contained within a standard format
19” 4U height rack mount enclosure. The electronics requires approximately 25 watts at an operating voltage
of 10 – 30 VDC. The rack comprises eight independent radar channels which are synchronised to minimise
interference, drift compensated to remove the effects of environmental temperature and controlled through
an LCD interface on the front of the panel. The radar sampling is triggered through the use of a shaft encoder
for vehicle applications or may be set to free run at a user-defined speed. Eight independent transmit and
receive antenna pairs are mounted in a secure, ruggedised antenna enclosure. The lateral antenna spacing is
approximately 18cm which allows for detection of AT and IED sized targets. The antenna cabling length is
2.5m, which is sufficient to allow for installation of the system on a typical four-wheel drive vehicle.

Figure 7 Laptop screen display showing test target in free space

The radar software provides a rolling map display of the radar data. The software allows the user to change
colour palette and gain and also allows for recording and playback of radar data. The software will run on a

© Copyright ERA Technology 2008

 Page 9 of 11

32-bit windows platform, installation on a laptop running 32 bit windows XP would be recommended. A
screen shot of the software is shown in Figure 7.

Forward looking radar systems have been investigated by a number of organisations and the most advanced
has been developed by Planning Systems Incorporated (PSI) (now a QinetiQ company) of Long Beach,
Mississippi, who developed a test bed forward-looking ground penetrating synthetic aperture radar
(FLGPSAR) system. This is a circularly polarized, stepped frequency radar that can operate over the 400 to
4000 MHz frequency band. The antenna system is composed of specially designed wideband dual spiral
antenna elements (instead of conventional horns). PSI’s Maverick Ground Penetrating Radar (patent #
6,445,334) is used to integrate forward-looking and downward looking detection systems. On smooth, dry
terrain and against buried plastic mines, a PD of 0.9 with a PFA of 0.002 have been reported. PSI notes that
increased roughness of terrain can be expected to increase the PFA and wet soil conditions will decrease the
PD. An example of the difference between single-look and multi-look modes for a flush buried TM62M
mines with a radar tilt angle of 22 degrees is shown in Figure 8.

Figure 8 Single look (lhs) and multi-look (rhs) processing for a buried mine (20dB dynamic range) (courtesy
Planning Systems)
Vehicle-based GPR must achieve orders of magnitude performance improvement to enable route clearing
military operations to proceed at speed. A total system performance of a probability of detection better than
0.99, with a probability of false alarm less than 10-4, is called for if route clearance at convoy speeds is to be
achieved. Humanitarian clearance may tolerate speed reduction but still requires high detection rates. This
applies to both stand-off and close-in GPR systems. Simonsen [8] and Voles [9] give detailed analyses of
radar performance.

Airborne systems

Several organisations have attempted to develop airborne UWB radar systems for landmine detection and
these are described by Daniels [10] but airborne UWB for landmine detection is an enormous technical
challenge. However a new generation of unmanned airborne vehicle may provide suitable platforms for the
close in GPR systems if low level flight paths can be achieved. This would allow UAV reconnaissance
ahead of convoys and would reduce the need to mine protect vehicles.

Summary

System
• UWB radar systems for landmine detection have a loop gain in the order of 120dB, which sets their
order of magnitude performance.
• The radiated power is limited by licence restrictions and EMC considerations as well as the need to
avoid detonation of certain types of fuse.

Propagation issues

• The attenuation losses in materials rapidly increase with frequency, which means that most systems
operate at frequencies in the range 300 MHz to 1.5GHz. The use of transmitted frequencies above

© Copyright ERA Technology 2008

 Page 10 of 11

2GHz is unlikely to provide useful performance in real world conditions and will severely limit
depth performance.
• Most path losses are such that penetration is generally limited to 50cm depth of cover.
• The propagation losses decrease as the fourth power of range to landmine for far field conditions.
• The propagation losses may decrease at lower rates depending on the landmine dimensions for near
field boundary conditions.
• The received signal may be augmented by induction and quasi-stationary contributions for
landmines within the near field.
• At 1GHz the total losses in typical soils mean that, in ideal conditions, detection ranges of 20-30cm
are feasible.
• In dry soils the dielectric contrast between the soil and mine reduces and this can make the detection
of mines with minimal air voids more difficult.
• Most GPR systems will achieve optimum performance in terms of range when the antennas are
operated in close proximity to the ground. As the antenna to ground spacing increases, the antenna
radiation pattern results in reduction of the received signal from small landmines and increased
vulnerability to clutter from free space sources.

Clutter and scattering

• Rough surfaces, ruts, potholes etc degrade the signal to clutter ratio and reduce the system
performance. Better target and clutter recognition and identification is needed to improve the signal
to clutter performance.
• The angular response of mines that are tilted relative to the ground surface may not be co-incident
with their physical position and this may cause errors in location.

Array based radars

• The attenuation losses in materials will reduce the effectiveness of multi-look antenna arrays by
effectively putting a window taper across the array.
• Stand off SAR radar systems have fundamental limits to performance at shallow grazing angles,
which constrains their forward look range to between 10 and 20m.

Acknowledgment

The author gratefully acknowledges the support of the Directors of ERA Technology for this work, which
was carried out under a company-funded programme.

References

[1] Daniels D J, An Assessment of the fundamental performance of GPR against buried landmines, SPIE Detection
and Remediation Technologies for Mines and Minelike Targets Xll, Paper 6553-16, SPIE 2007, 13 April,
2006,Orlando, Florida.
[2] Daniels D J, Curtis P, MINEHOUND trials in Bosnia, Angola and Cambodia. Proceedings of the SPIE Defense
and Security Conference 2006, 17-23 April, 2006,Orlando, Florida.
[3] Doheny R C et al, Handheld Standoff Mine Detection System (HSTAMIDS) field evaluation in Namibia,
Proceedings of the SPIE Defense and Security Conference, 16-21. April 2006, Orlando, FL USA.
[4] Wong D C, Lam H. Nguyen L H, and. Gaunaurd G C, Time-frequency analysis for radar classification of land
mine images, J. Electron. Imaging 16, 033014 (2007).
[5] Lopera, O.; Milisavljevie, N.; Daniels, D.; Macq, B, Time-frequency domain signature analysis of GPR data for
landmine identification, Advanced Ground Penetrating Radar, 2007 4th International Workshop 27-29 June 2007
Page(s): 159 – 162.
[6] Daniels D J, Curtis P, Lockwood O, Classification of landmines using GPR, IEEE RadarCon 2008, 26-30 May
2008 Rome Italy.
[7] Wilson J, Gader P, Lee W, Frigui H, Ho K, A Large-Scale Systematic Evaluation Of Algorithms Using Ground-
Penetrating Radar For Landmine Detection And Discrimination, IEEE Transactions On Geoscience And Remote
Sensing, Vol. 45, No. 8, August 2007.

© Copyright ERA Technology 2008

 Page 11 of 11

[8] Simonson K Statistical considerations in designing tests of mine detection systems 1 Measures related to the
probability of detection, Sandia Report SAN98-1769/1.
[9] Voles R, Confidence in trials of landmine detection systems, Mathematics Today April 2000.
[10] Daniels D J, Ground Penetrating Radar, IET London 2004 , ISBN 0 86341 360 9.

© Copyright ERA Technology 2008

View publication stats