Progress In Electromagnetics Research, Vol.

122, 245268, 2012

A NEW UNMANNED AERIAL VEHICLE SYNTHETIC

APERTURE RADAR FOR ENVIRONMENTAL MONITORING
V. C. Koo1 , Y. K. Chan1, * , V. Gobi1 , M. Y. Chua1 ,
C. H. Lim1 , C. S. Lim1 , C. C. Thum1 , T. S. Lim1 , Z. Ahmad2 ,
K. A. Mahmood2 , M. H. Shahid2 , C. Y. Ang1 , W. Q. Tan1 ,
P. N. Tan1 , K. S. Yee1 , W. G. Cheaw1 , H. S. Boey1 ,
A. L. Choo1 , and B. C. Sew1
1 Faculty

of Engineering & Technology, Multimedia University, Jalan

Ayer Keroh Lama, Bukit Beruang, Melaka 75450, Malaysia
2 Malaysian

Remote Sensing Agency, No. 13, Jalan Tun Ismail, 50480

Kuala Lumpur, Malaysia
AbstractA new Unmanned Aerial Vehicle (UAV) Synthetic
Aperture Radar (SAR) has been developed at Multimedia University,
in collaboration with Agency of Remote Sensing Malaysia. The SAR
operates at C-band, single V V -polarization, with 5 m 5 m spatial
resolution. Its unique features include compact in size, light weight,
low power and capable of performing real-time imaging. A series
of field measurements and flight tests has been conducted and good
quality SAR images have been obtained. The system will be used for
monitoring and management of earth resources such as paddy fields, oil
palm plantation and soil surface. This paper reports the system design
and development, as well as some preliminary results of the UAVSAR.

1. INTRODUCTION
Radar is the abbreviation for Radio Detection and Ranging. It
operates by radiating electromagnetic energy through a transmitting
antenna and detecting the reflected or scattered signal from the
target [1]. SAR is an imaging radar which utilizes relative motion
between an antenna and the target under observation to synthesize
a very long antenna via signal processing [2]. As compared to
Received 26 September 2011, Accepted 2 November 2011, Scheduled 18 November 2011
* Corresponding author: Yee Kit Chan (ykchan@mmu.edu.my).

246

Koo et al.

the conventional real aperture radar, SAR can obtain finer spatial
resolution particularly in azimuth direction. The concept of SAR
can be traced back to early 1950s when Carl Wiley proposed a
Doppler beam sharpening system to improve the azimuth resolution
of radar [3, 4]. Today, SAR has become an important tool for
microwave remote sensing because of its capability to operate day
and night, and in nearly all weather conditions [5]. It has wide
range of applications, including sea and ice monitoring [6], mining [7],
oil pollution monitoring [8], oceanography [9], snow monitoring [10],
terrain classification [11] and so on. The potential of SAR in a diverse
range of applications has led to the development of a number of
airborne and spaceborne SAR systems.
An unmanned aerial vehicle (UAV) is an unpiloted aircraft
typically found in the area of surveillance, reconnaissance as well as
military missions. The earliest UAV was found in 1916 [12], followed
by a number of remote-controlled airplanes developed during and after
World War I for military applications. UAV is currently one of the
mostrapid growth research areas due to its increased importance in
military operations [13]. The role of a UAV system falls mainly in
reconnaissance and surveillance missions that utilize remote sensing
instruments such as electro-optical (EO) sensors, infrared (IR) sensors
and synthetic aperture radars (SAR). The increased popularity of a
UAV system compared to a conventional aircraft is due to a few factors.
There is no risk of losing human life especially for risky operations.
The over loading of a manned aircraft has to be limited to 10 g due
to human physical limitations but it can be increased up to 20 g for
UAV. There is also no concern on human fatigue and boredom in a long
UAV mission. The weight of the human pilots is roughly equivalent to
15% of the effective load of a military aircraft, while the pilot related
supporting and emergency systems cost as much as 50% of the total
cost of the aircraft [13]. Therefore the size of a UAV is only about 60%
of the size of a manned aircraft with the same performance and the
price for a UAV system is expected to be 40% less than the cost of a
manned solution [13, 14].
In terms of remote sensing applications, UAV-based imaging
radar such as SAR has great potential due to its all-day, all-weather
capabilities. As compared to conventional airborne or space-borne
SAR systems, UAV SARs have several advantages which include low
cost, low risk and timely operations.This has led to the development
of a number of UAV SAR systems in recent years.
A low cost, compact and low power micro-SAR (SAR) system
has been developed by the Brigham Young University (BYU) [15].
The SAR is based on a linear frequency modulated continuous

Progress In Electromagnetics Research, Vol. 122, 2012

247

wave (LFM-CW) configuration, which has smaller size and power

requirements as compared to a conventional pulsed SAR system. This
enables the BYU SAR to fly on a small UAV, further reducing the cost
of operation. The SAR was designed to image Arctic sea ice. The
target UAV has a 6-foot wingspan, flies at low-altitudes (< 300 m),
with limited power to supply to the payload. European Aeronautic
Defence and Space Company (EADS) has developed a miniaturised
SAR (MISAR) sensor that was specially designed for UAVs with
stringent payload size, weight and power (SWP) constraints. The
LUNA UAV of the German manufacturer EMT was chosen as the
reference platform. LUNA has a take-off weight in the order of
35 kg and its payload weight is restricted to 4 kg. The flying
altitude of LUNA is between 3002000 m with the velocity of 10
40 m/s. This UAVSAR will be operated at stripmap mode with swath
width of 5001000 m. In order to achieve SWP requirement, the
FMCW configuration is employed. The Jet Propulsion Laboratory
(JPL) has designed and developed a L-band SAR for repeat pass
deformation measurements on a UAV platform [1618]. The SAR
system is operated at a wavelength of 0.2379 m (L-band, 1.26 GHz)
and bandwidth of 80 MHz. The pulse duration is designed at 40 msec
with PRF of 350 Hz (interleaving H and V transmit polarizations). It
is a fully polarimetric and repeat pass interferometry SAR system. The
UAVSAR is able to obtain a range swath of 16 km with look agle of
30 to 60 degree. The system will nominally operate at 45,000 ft. Both
ALTAIR and the Proteus can be used for this SAR system.
Research Establishment for Applied Science (FGAN) is in
the progress of developing ARTINO (Airborne Radar for Three
dimensionally Imaging and Nadir Observation) [1921]. It comprises
a low flying UAV, which will be able to map a directly overflown
scene into a high resolution three-dimensional image by looking
perpendicularly downward. With a wingspan of only 4m and a modular
structure the UAV can be split up into small segments and easily
transported to the area of interest. Due to the light weight of only
25 kg it is possible to fly at a slow speed of approximately 15 m/s. Since
ARTINO is a low flying UAV, the 3D imaging radar can be designed
for short range. Hence, the average transmitting power is only a few
Watts.
A commercial available UAVSAR Lynx SAR has developed
by Sandia National Laboratories [22, 23]. Lynx is a state-of-theart, ultra-fine-resolution, real-time SAR and ground moving target
indicator (GMTI) radar. Sandia National Laboratories collaborated
with General Atomics (GA) to design and build the Lynx system
to incorporate General Atomics design requirements for operation

248

Koo et al.

on their Predator, IGNAT, or Prowler II unmanned aerial vehicles

(UAVs). AN/APY-8 Lynx II is a lightweight version of Lynx operating
on the US Air Force Predator RQ-1 UAV. Lynx is a Ku-band
multimode radar. Its SAR modes include real-time 0.1 m resolution
spotlight and 0.3 m resolution stripmap modes. The radar can also scan
a large or small area for moving objects, detecting targets at speeds
typical of vehicular movements (1070 kph). The Lynx SAR operates
in the Ku-Band anywhere within the range 15.2 GHz to 18.2 GHz, with
320 W of transmitter power.
Many other UAVSAR systems are designed for government
agency, research institution and private industry. An experimental
millimetre band (34 GHz frequency) Synthetic Aperture Radar (SAR)
sensor has been developed by Polytechnic University of Madrid,
Spain [24]. It is a short-range (2 Km) high-resolution (3030 cm2 ) SAR
system with bandwidth of 675.8 MHz and PRF of 1300 Hz. Research
Institute of China Electronics Technology Group Corporation is also
designed a 35 GHz, 8 mm wavelength UAV-borne SAR [25]. The
resolution achieved by this UAVSAR is proved by flight experiments
that have reached 0.85 m 0.45 m. A Ka band, fully-polarized SAR
radar is developed by the Lincoln Laboratory. This millimeter-wave
SAR imaging system has the bandwidth of 600 MHz in order to assure
0.3 m range resolution. The strip is 440 m wide, the spot imaging
area is 150 m 150 m, and the azimuth resolution is 0.3 m [25]. Italy
Salenia Corporation has developed a UAV-borne MMW SAR which
can perform ground mapping, detection, tracking and imaging of slowspeed objects. The operating frequency is 35 GHz. The SAR system
weight 60 kg and takes up 0.05 m3 . The waveguide slotted antenna
is 1 m long, 0.24 m wide and 2 kg weight. Pulse width 6 s and
bandwidth is 40 MHz. The range resolution corresponding to the pulse
compression is 5 m. Microwave Remote Sensing Laboratory, Center
for Environmental Remote Sensing, Chiba University is currently
developing its Circular Polarization SAR (CP-SAR) sensor which can
be used in UAV as well as satellite based platform [26]. This sensor
is operated with center frequency on L band (1.27 GHz) and 10 MHz
of chirp pulse bandwidth. The sensor is designed as a low cost,
simple, light, strong, low power, low profile configuration to transmit
and receive left-handed circular polarization (LHCP) and right-handed
circular polarization (RHCP), where the transmission and reception
are working in RHCP and RCHP + LCHP, respectively.
In late 2008, a new UAVSAR project was initiated at Multimedia
University (MMU), Malaysia, in collaboration with Agency Remote
Sensing of Malaysia (ARSM). This project was started after
preparatory studies in the previous year [2731]. The proposed system

Progress In Electromagnetics Research, Vol. 122, 2012

249

is a C-band, V V -polarization, miniature SAR mounted on a small

UAV. The UAVSAR system is designed to operate at low altitudes
with low transmit power in order to optimize the operating cost. The
construction of the SAR sensor was completed in early 2010. A series
of indoor and outdoor testing and measurements has been carried
out to verify the performance of the system. At the end of 2010,
flight missions have been conducted at Mersing, Johor. High quality
SAR images were obtained which demonstrated the capabilities of the
UAVSAR for remote sensing applications in Malaysia.
This paper presents the system design and development, as well as
some preliminary results of the UAVSAR. Section 2 outlines the major
design issues and considerations for UAVSAR. Section 3 describes
the development of various SAR subsystems including the microstrip
patch antenna, the radar electronics, the embedded SAR controller
and processor, and the UAV platform. Section 4 presents some of the
preliminary results obtained by the UAVSAR, which include groundbased SAR experiments and fight tests. Finally, Section 5 concludes
the paper.
2. DESIGN CONSIDERATIONS
The primary goal of this research project is to develop an UAV based
SAR system which capable to illuminate terrain and construct the
image of the scanned area. High level design consideration have been
carefully considered and presented in this section.
2.1. Operating Frequency
C-band (48 GHz) frequency is widely used for high-resolution land
imaging, agricultural monitoring and ocean observations. In this
frequency band, incident wave tends to be reflected more by vegetation
canopies rather than the surface layer. This penetration characteristic
allows the canopy biomass and soil below the canopy to be analyzed.
The center frequency of the UAVSAR is selected at 5.3 GHz,
which is within the allowable spectrum (52505460 MHz) defined by
International Telecommunication Union (ITU) for Earth Exploration
Satellite System (EESS) [32]. Beside the reason of having low
absorption in the atmospheric-window region, the size of a C-band
antenna is considerably small and most of the RF components are
easily available.

250

Koo et al.

2.2. Mode of Operation

The two most common SAR imaging modes are stripmap and spotlight.
The stripmap is a standard mode of SAR operation, widely used by
airborne SAR sensors where a strip (swath) to the side of the aircraft
is imaged. On the other hand, a spotlight SAR steers its antenna
beam to continuously illuminate a specific terrain patch during data
collection. The spotlight mode is suitable to collect fine-resolution data
from localized areas, while the stripmap mode is more efficient when
used for coarse-resolution mapping of large regions. In our design, the
stripmap mode is the preferred choice.
2.3. Polarization
Single polarization mode is proposed for simple classification and
multi-temporal change detection. V V -polarization is the preferred
configuration since it is sensitive to the vegetations vertical canopy
structure, thus providing the opportunity for crop type and growth
stage discrimination.
2.4. Dynamic Range
Based on the measurement results reported in numerous literatures,
it is found that typical values of scattering coefficients for various
categories of terrain fall in the range from 0 dB to 30 dB [33].
Therefore, a wide dynamic range (> 30 dB) is needed to accommodate
the measurement of various types of terrain.
2.5. Resolution
Typical resolution of airborne SAR ranges from 1 m to 20 m. It depends
mostly on the application requirements. Since the main objective is to
monitor earth resources such as vegetation fields, resolution of 5 m5 m
for both range and azimuth direction will be adequate.
2.6. Antenna Requirements
A planar, lightweight, small size, linearly polarized microstrip patch
array panel is proposed. The center frequency for the array is set at
5.3 GHz, with more than 80 MHz bandwidth. The directive gain of the
panel should be more than 15 dBi for good detection in the presence of
noise. Due to the limited space of the UAV platform, the antenna will
be mounted directly underneath the fuselage. The beamwidth will be
tilted 25 off nadir for side-looking operation.

Progress In Electromagnetics Research, Vol. 122, 2012

251

2.7. Signal Processor

The on-board facility consists of a high-speed analog-to-digital
converter, a front-end processor, and a high-density digital data
recorder to store the raw data. The SAR images should be produced by
both ground processing facility and onboard processor. Real-time data
capturing, recording and processing are desired for timely monitoring
during flight mission.
2.8. Motion Sensing
The irregular motion of the aircraft due to atmospheric turbulence will
seriously affect the image quality. In particular, the change of altitude
causes the image out of focus. In this system, an embedded inertial
navigation system (INS) and global positioning system (GPS) are used
to provide integrated flight information.
2.9. UAV Platform
A small UAV will be used to install the radar system. The platform
should support true ground speed of approximately 3040 m/s with
operating altitude at about 1000 m. The desired payload is at least
20 kg.
The constraints that are imposed onto the design of the remote
sensing instruments for UAVs are mainly weight, power consumption
and volume. The aircraft should have a lot of reserve internal volume
and electrical power so that it is well-suited for a variety of payloads.
The SAR system can be designed to fit existing EO/IR sensor mounting
as the mounting will become common for remote sensing platform and
it certainly helps interpretability. The UAV performance such as flight
speed and altitude also dictate the design of the SAR system. In this
project, Aludra MK1 has been chosen as the SAR platform.

Figure 1. Aludra MK1.

252

Koo et al.

Aludra MK1 is developed by UST (Unmanned Systems

Technology), Malaysia with the dimension of 4.27 m and wingspan
of 4.88 m. The image of Aludra MK1 is shown in Figure 1. The
typical speed of MK1 is 35 m/s and up to maximum speed of 61 m/s.
Endurance of this UAV is about 3 hrs. However limited payload weight
is a challenge. The working space is about 27 cm (W ) 35 cm (L)
25 cm (H) and maximum payload of 20 kg. Thus extra attentions need
to be taken for the dimension and weight of the SAR sensor and
antenna size. Therefore a new chassis of the SAR sensor need to be
design and constructed for installing all SAR subsystem.
Based on the design considerations described above, the
specifications of the UAVSAR system are determined and are
summarized in Table 1. Figure 2 shows the geometric configuration
of the UAVSAR. The standard image size produced is 900 m (range)
by 1500 m (azimuth).
Table 1. UAVSAR system specifications.
System Parameters

Specifications

Mode of Operation

Stripmap

Operating Frequency

5.3 GHz (C-band)

Modulation

LFM Pulse

Bandwidth

80 MHz

Pulse Repetition Frequency

1000 Hz

Pulse Width

10 s

Polarization

VV

Antenna Gain

15 dBi

Antenna Size

1 m 0.3 m

Spatial Resolution

5m 5m

RCS Dynamic Range 30 dB

0 dB to 30 dB

SNR

> 10 dB

Platform Height, h

1000 m

Swath Width

900 m

Nominal Platform Speed

35 m/s

Data Take Duration

12 hour (50 sec per scene)

Operating Platform

UAV, Aludra MK1

Overall Sensor Weight

< 20 kg

Overall Sensor Dimension

<27 cm (W ) 35 cm (L) 25 cm (H)

Peak Transmit Power

50 W

Progress In Electromagnetics Research, Vol. 122, 2012

40

253

1000 m

25 o

912 m

Figure 2. Geometric configuration of the UAVSAR system.

GPS
Antenna

GPS
Data

Motion sensing data

Motion Sensing Unit
PRF

Rx Window
Control Signal

Antenna
System

10 MHz Ref
80 MHz LFM
I and Q signals

Tx/Rx
signal

FPGA-based
Embedded
SAR
Controller

RF Subsystem
Data Recorder

From UAV
main power
supply

Power Supply Module

(+15 V/+28 V)

Down-converted
received I and Q
signals

Down-link
Antenna

Embedded SAR
Processor

Data Downlink

Figure 3. Functional block diagram of the SAR sensor.

3. SYSTEM IMPLEMENTATION
Figure 3 shows the functional block diagram of the UAVSAR Sensor.
In this design, a Field-Programmable Gate Array (FPGA) based
SAR controller is used to produce the required linear FM (LFM) signals
and other necessary timing signals. The LFM I and Q signals are
up-converted and amplified in the RF subsystem before transmitting

254

Koo et al.

out via an antenna system. The antenna system is implemented by

using microstrip patch arrays. The returned echoes are amplified and
down-converted into an intermediate frequency (IF), at which they
are further digitized and stored in the data recorder. A replica of
the IF signal is fed into an embedded SAR processor for real-time
image formation. The resultant image will be transmitted to the
ground station via a 2.4 GHz data downlink unit. An integrated
Inertial Measurement Unit (IMU) with Global Positioning System
(GPS) receiver is also developed to provide the necessary platform
motion information.
3.1. RF Subsystem
The block diagram of the RF subsystem is shown in Figure 4. Basically
it can be divided into two sub-sections: the transmitter and the
receiver. At the transmitter front-end, the I (In-phase) and the Q
(quadrature-phase) chirp signals are mixed with the carrier frequency
(5.3 GHz) via an up-convert mixer. The output of the mixer is
amplified by a power amplifier (PA). A RF switch is used to turn
on the transmitting window, where its on-time duration is selectable
based on mission requirements. The signal is then fed to a bandpass
filter to reject unwanted signal outside the desired frequency band,
and further amplified by a high power amplifier (HPA) to boost up the
power for long range transmission. The average output power is about
160 mW for nominal operating range of 1 km.

Figure 4. Functional block diagram of the SAR RF subsystem.

Progress In Electromagnetics Research, Vol. 122, 2012

255

At the receiver front-end, a diode limiter is used to provide

protection for any sudden power spikes from the transmission. A fast
switching pin diode switch is employed to turn on the receiving window
at certain range-gate interval, and to prevent saturation due to leakage
signals from the transmitter. Two low noise amplifiers (LNAs) are
cascaded to provide sufficient amplification for the returned echoes.
Finally, the mixer down-converts the received signal to IF I and Q
signals. The prototype of the RF subsystem is shown in Figure 5.
3.2. Antenna
The technical specifications of the SAR antenna are summarized in
Table 2. Design and fabrication of the antenna is based on composite
laminates. The antenna is fabricated as a single layer system whereby
the patches and the feeds are assembled on one side of the surface,
while the base of the plate attached to the undercarriage of the UAV
will act as the ground plane.

Figure 5. Prototype of the SAR RF subsystem.

Table 2. Antenna technical specifications.
System Parameters
Operating Frequency
Bandwidth
Polarization
Antenna Gain
Platform Height, h
Incident Angle
Azimuth Beamwidth
Elevation Beamwidth

Specifications
5.3 GHz (C-band)
80 MHz
VV
15 dBi
1000 m
25
4
40

256

Koo et al.

The UAVSAR antenna operates at 5.3 GHz with bandwidth of

80 MHz and transmits signal in single V polarization. This antenna
adopts 3 layers configuration with the top layer acts as radiating layer,
middle layer acts as ground plane with feeding network is designed on
bottom layer. Probe fed method has been utilized where the EM signal
is fed from feeding network to radiating elements through a probe. The
UAVSAR antenna is divided into 3 sub panels for the ease of fabrication
and then joined by power dividers. The back and front views of the
final antenna system are depicted in Figures 6 and 7, respectively.
The azimuth array is formed with 24 radiating elements to produce
a narrow pencil beam of 3 . The elevation radiation pattern of the
antenna shows the main lobe is located at 30 away from center
with beamwidth of 24 to provide wider swath width. Before it can
be installed for flight mission, the performance of the antenna has
been verified in the laboratory via VNA as well as radiation pattern
measurement using anechoic chamber. The measured result shows the
return loss value at 5.3 GHz achieve 32.381 dB. From the radiation
pattern, the highest power level of sidelobe appears at 45 to 55
with the maximum sidelobe power level of 15 dB which help to
overcome left-right ambiguity issue. Figure 8 shows the radiation
pattern of the antenna which is well match with our simulation result.

Figure 6. Back view of the fabricated antenna.

Figure 7. Front view of the fabricated antenna.

Progress In Electromagnetics Research, Vol. 122, 2012

257

Figure 8. Measured radiation pattern of SAR antenna.

Figure 9. Block diagram of SAR motion sensing unit.

3.3. Motion Sensing Unit
An embedded motion sensing unit has been developed to collect the
required information for motion compensation. These include data
such as time, longitude, latitude, altitude, 3-axis accelerations and
3-axis angular velocity (raw, pitch and yaw). The block diagram of
the SAR motion sensing unit is illustrated in Figure 9. The motion
sensing of SAR sensor is achieved by integrating both global positioning

258

Koo et al.

system (GPS) and strapdown inertial moment unit (IMU). The GPS
is able to provide instantaneous position data on the SAR sensor
with no accumulation errors, but suffers a relatively low data rate
(5 Hz). On the other hand, the strapdown IMU provides higher data
rate (167 Hz) for position, velocity and orientation estimations, but
are subject to sensors accumulation errors. Thus, the integration of
GPS and IMU suppresses the shortcomings of each individual system,
namely, the GPSs low data rate and IMUs accumulation errors. Also,
the integration combines the advantages of both systems, such as the
high accuracy positioning data in GPS and the short term stability of
IMU.
As shown in the block diagram of Figure 9, the data from GPS and
IMU were compared and feed into an extended Kalman filter (EKF)
for optimal errors prediction and estimation. The estimated errors
will be used to correct the original measurements. Due to data rate
inconsistency between IMU and GPS, the data fusion event will only
take place when GPS data was available. In the event of no GPS
data the IMU will correct itself using the last predicted errors. The
constructed motion sensing unit is shown in Figure 10.

4 cm

11 cm

15 cm

Figure 10. The integrated IMU-GPS motion sensing unit.

High Speed SAR

Processor

Raw Data in

Figure 11.
recorder.

DAQ
Unit

Wireless
Downlink

High Density
Harddisk

Overview of the embedded SAR processor and data

Progress In Electromagnetics Research, Vol. 122, 2012

259

3.4. Embedded SAR Processor and Data Recorder

Figure 11 depicts the overview of the embedded SAR processor and
data recorder. A high speed data acquisition unit (DAQ) is used to
digitize the down-converted IF I and Q signals from the RF receiver.
These raw data are directly stored onto a high density harddisk.
Meanwhile, a sample of the stored data (level-0) will be retrieved from
the storage and processed by the embedded SAR processor in real-time.
The output of the on-board SAR processor is level-1 images, which will
be sent to the ground receiving station via a wireless downlink.
There are a few stripmap mode imaging formation algorithms
such as the Range Doppler algorithm (RDA), Chirp Scaling algorithm
(CSA) and Wave Number Algorithm/Omega-K Algorithm (kA) [34].
RDA is a simple and mature SAR image processing algorithm with
a relatively small amount of computation load and it is suitable
for real-time processing. Basically, it divides the SAR echo signal
processing into two one-dimensional processes in cascade, i.e., range
compression followed by azimuth compression. In summary, real-time
signal processing algorithms for airborne SAR can be implemented by
RDA. RDA block diagram is shown in Figure 12. Therefore, RDA
has been selected as the real time image formation algorithm for our
UAVSAR system.

Figure 12. RDA block diagram.

260

Koo et al.

Figure 13. Block diagram of enhanced UAVSAR data recording and

processing system.
The embedded SAR processor is implemented using a High
Performance Computer (HPC). The selected motherboard has a Core2based CPU and it is capable of supporting up to 4 GB of RAM.
It also has a built-in CUDA (Compute Unified Device Architecture)
compatible graphic processing unit. A 12-bit high-speed ADC card is
employed for the implementation of the DAQ unit. This ADC card
has two-channel inputs and it is capable of simultaneous sampling
up to 250 MHz. The digitized data are transferred to the HPC via
PCI express interface at 1.25 GBytes/s data rate. Figure 13 shows
the block diagram of the enhanced UAVSAR data acquisition and
processing system. It is worth mentioning that UAVSAR raw data is
acquired using the ADC with a designed data rate of 7.5 Mbytes/s.
After the conversion, the ADC stored the data in HPCs Random
Access Memory (RAM) temporarily, where the HPC will excerpt the
information out of the RAM and saved it into local mass storage
device. Here lies the first bottleneck of the designed system, where
a typical magnetic hard disk should not be used in UAV flight test due
to the vibration of the flight might potentially damaged the hard disks
spinning disk with its read/write heads. This bottleneck can be resolve

Progress In Electromagnetics Research, Vol. 122, 2012

261

by replacing the magnetic hard disk with solid-state drive (SSD). A

SSD is a mass storage device uses microchip to store data. It contains
no moving part during data retaining as compared to magnetic hard
disks spinning disk, and its data rate is also faster than magnetic
hard disk. The second bottleneck lies on the physical data bus
congestion in real time UAVSAR recording and processing application.
However, enhancement can be made to resolve this bottleneck by
proper arrangement of two SSDs into the system. As shown in
Figure 13, the HPC is designed in such a way that the UAVSAR
raw data recording and processing always taking place at different
SSD, thus reducing the SSDs data bus congestion and enhancing the
UAVSAR data processing efficiency.
3.5. System Installation
As shown in Figure 14, all the SAR subsystems except the antennas
(SAR antenna, GPS antenna and data downlink antenna) are mounted
inside the UAV compartment. The overall dimension of the SAR sensor
is 27 cm (W )35 cm (L)25 cm (H), and the total weight of the system
(SAR sensor plus antenna) is less than 20 kg. The SAR platform used
is capable of operating at a nominal ground speed of 35 m/s at 1000 m.
Figure 15 shows the bottom view of UAV platform, where the SAR
antenna is placed directly underneath the UAVs fuselage.

Figure 14. The SAR sensor

mounted inside the UAV compartment.

Figure 15. The SAR antenna

mounted at the bottom of the
UAVs fuselage.

262

Koo et al.

4. PRELIMINARY RESULTS
A series of field experiment has been conducted in year 2010 to verify
the performance of the UAVSAR system. These include point target
measurement, and ground-based SAR experiments using a moving
platform. Flight tests were finally conducted at the end of 2010 to
evaluate the overall performance of the SAR sensor.

82.5 m

starting point
20 m

20 m

Sensor Moving Direction

SAR
Sensor

Figure 16. Experimental setup for point target measurement.

Figure 17. The SAR image of a trihedral placed at 82.5 m.

Progress In Electromagnetics Research, Vol. 122, 2012

263

4.1. Point Target Measurement

Figure 16 shows the experimental setup for point target measurement.
The test site was a public field with low background clutter. A trihedral
corner reflector was placed on a polystyrene structure of about 1.5 m
in height, located 82.5 m from the SAR transmitting antenna. The
measurement was conducted by employing the stop-and-go approach,
where the SAR sensor was manually moved to a pre-defined cross range
(azimuth) position for every single range bin data capturing. The total
cross range travelled was 40 m, which yields 400 points of azimuth data
with 0.10 m spacing. Figure 17 shows a SAR image generated from this
experiment. A strong target is clearly shown at location 82.5 m. The
measured 3 dB azimuth resolution is 0.78 m, and the peak sidelobe level
is 12.23 dB.
4.2. Ground-based SAR Experiment
In this experiment, the SAR system was mounted onto a truck.
The truck was then travelled at a constant speed to perform SAR
imaging. The target of interest was a housing area at distance (slant
range) more than 1 km. Figure 18 shows the processed image where
multiple strong targets are observed at distance more than 1 km. This
preliminary ground experiment has successfully verified the overall
imaging capability of the system.

Figure 18. The SAR image of multiple targets (at distance > 1 km).

264

Koo et al.

Figure 19. The SAR image of the calibration site at Mersing.

4.3. Fight Tests
Preliminary flight tests have been conducted in December 2010 at
Mersing, Malaysia. During the 6-day flight mission, more than 200
sets of SAR raw data were collected. Eleven trihedral corner reflectors,
arranged in three groups, were used as the point targets for external
calibration. All the trihedral corner reflectors were placed in a straight
line and perpendicular to the flight path of the UAV. Figure 19 shows
the SAR image of the calibration site. Signatures such as the point
targets, UAV runway, pit crew tent and ground control system are
clearly observed.
Figures 20 and 21 show two samples of SAR images captured
on Dec. 5th and Dec. 8th, 2010, respectively, with comparison to the
Google Earth map of the same site. Clear signatures of river, roads,
urban and forested areas are observed.

Progress In Electromagnetics Research, Vol. 122, 2012

265

Range

Range

Azimuth

Azimuth

(a)

(b)

Figure 20. (a) Google earth

map. (b) SAR image-1 at Mersing
site.

(a)

(b)

Figure 21. (a) Google earth

map. (b) SAR image-2 at Mersing
site.

5. CONCLUSIONS
A miniature SAR system operated on a small UAV platform has
been successfully designed and developed. Various design parameters
have been carefully studied and determined in the system level
design. Implementation of SAR system with especial attention to the
dimension and size of the sensor chassis and antenna due to constrains
of the UAV in particularly the limitation in weight and volume. A
series of filed measurement and flight test has been conducted to verify
the performance of the UAVSAR sensor. This new system will serve
as a test-bed for demonstrating SAR technology and acquiring data
for environmental monitoring in Malaysia.
ACKNOWLEDGMENT
This project is funded by Agency of Remote Sensing Malaysia under
Ministry of Science, Technology and Innovation, Malaysia.

266

Koo et al.

REFERENCES
1. Skolnik, M. I., Radar Handbook, McGraw-Hill, New York, 1970.
2. Curlander, J. C. and R. N. McDonough, Synthetic Aperture
Radar: System and Signal Processing, A Wiley Interscience
Publication, 1991.
3. Wiley, C. A., Synthetic aperture radar A paradigm for
technology evolution, IEEE Trans. on Aerospace and Electronic
Systems, Vol. 21, 440443, 1985.
4. Ausherman, D. A., A. Kozma, J. L. Walkeretal, H. M. Jones, and
E. C. Poggio, Developments in radar imaging, IEEE Trans. on
Aerospace and Electronic Systems, Vol. 20, No. 4, 363440, 1984.
5. Ulaby, F. T., R. K. Moore, and A. K. Fung, Microwave Remote
Sensing-active and Passive, Vol. 2, Addison Wesley, 1981.
6. Drinkwater, M. K., R. Kwok, and E. Rignot, Synthetic aperture
radar polarimetry of sea ice, Proceeding of the 1990 International
Geoscience and Remmote Sensing Symposium, Vol. 2, 15251528,
1990.
7. Lynne, G. L. and G. R. Taylor, Geological assessment of SIRB imagery of the amadeus basin, IEEE Trans. on Geosc. and
Remote Sensing, Vol. 24, No. 4, 575581, 1986.
8. Hovland, H. A., J. A. Johannessen, and G. Digranes, Slick
detection in SAR images, Proceeding of the 1994 International
Geoscience and Remmote Sensing Symposium, 20382040, 1994.
9. Walker, B., G. Sander, M. Thompson, B. Burns, R. Fellerhoff,
and D. Dubbert, A high-resolution, four-band SAR testbed with
real-time image formation, Proceeding of the 1986 International
Geoscience and Remmote Sensing Symposium, 18811885, 1996.
10. Storvold, R., E. Malnes, Y. Larsen, K. A. Hgda, S. E. Hamran,
K. M
uller, and K. A. Langley, SAR remote sensing of snow
parameters in Norwegian areas Current status and future
perspective, Journal of Electromagnetic Waves and Applications,
Vol. 20, No. 13, 17511759, 2006.
11. Kong, J. A., S. H. Yueh, H. H. Lim, R. T. Shin, and J. J. van
Zyl, Classification of earth terrain using polarimetric synthetic
aperture radar images, Progress In Electromagnetics Research,
Vol. 3, 327370, 1990.
12. Taylor, J. W. R., Janes Pocket Book of Remotely Piloted Vehicles,
Collier Books, New York, 1977.
13. Yong, J. M., Secrets of Military Unmanned Aerial Vehicle, Defense
University Publisher, China, 2004.

Progress In Electromagnetics Research, Vol. 122, 2012

267

14. Wall, R., Trial Run: Sigint Version, Not Maritime Patrol,
Attracts Initial German Interest in Global Hawk, 4243, McGrawHill, USA, 2003.
15. Zaugg, E. C., D. L. Hudson, and D. G. Long, The BYU
SAR: A small, student-built SAR for UAV operation, The 2006
International Geoscience and Remote Sensing Symposium, 411
414, 2006.
16. Hensley, S., K. Wheeler, G. Sadowy, et al., Status of a UAVSAR
designed for repeat pass interferometry for deformation measurements, IEEE MTT-S International Microwave Symposium Digest, 14531456, 2005.
17. Wheeler, K., S. Hensley, Y. Lou, T. Miller, and J. Hoffman, An
L band SAR for repeat pass deformation measurements on a UAV
platform, Proceedings of the IEEE Radar Conference, 317322,
2004.
18. Rosen, P. A., S. Hensley, K. Wheeler, G. Sadowy, T. Miller,
S. Shaffer, R. Muellerschoen, C. Jones, H. Zebker, and S. Madsen,
UAVSAR: A new NASA airborne SAR system for science and
technology research, IEEE Conference on Radar, 2229, 2006.
19. Rosen, P. A., S. Hensley, K. Wheeler, G. Sadowy, T. Miller,
S. Shaffer, R. Muellerschoen, C. Jones, and S. Madsen, and
H. Zebker, UAVSAR: Airborne SAR system for research, IEEE
A&E Systems Magazine, 2128, 2007.
20. Weib, M. and J. H. G. Ender, A 3D imaging radar for small
unmanned airplanes ARTINO, Radar Conference, 209212,
2005.
21. Weiss, M., O. Peters, and J. Ender, A three dimensional SAR
system on an UAV, IEEE International Geoscience and Remote
Sensing Symposium, 53155318, 2007.
22. Wells, L., K. Sorensen, A. Doerry, and B. Remund, Developments
in SAR and IFSAR systems and technologies at sandia national
laboratories, IEEE Aerospace Conference Proceedings, Vol. 2,
10851095, 2005.
23. Tsunoda, S. I., F. Pace, J. Stence, M. Woodring, W. H. Hensley,
A. W. Doerry, and B. C. Walker, Lynx: A high-resolution synthetic aperture radar, IEEE Aerospace Conference Proceedings,
Vol. 5, 5158, 2000.
24. Saldana, R. R. and F. P. Martinez, Design of a millimetre synthetic aperture radar (SAR) onboard UAVs, IEEE International
Conference on Electronics, Circuits and Systems, 15, 2007.
25. Peihong, R., High-resolution 8 mm radar imaging technique,

268

26.

27.

28.
29.

30.

31.
32.
33.
34.

Koo et al.

International Conference on Radar, 13, 2006.

Wissan, Y. V., I. Firmansyah, P. Rizki Akbar, J. T. Sri Sumantyo,
and H. Kuze, Development of circularly polarized array antenna
for synthetic aperture radar sensor installed on UAV, Progress
In Electromagnetics Research C, Vol. 19, 119133, 2011.
Chan, Y. K., M. K. Azlindawaty, V. Gobi, B. K. Chung, and
H. T. Chuah, The design and development of airborne synthetic
aperture radar, IEEE International Geoscience and Remote
Sensing Symposium, Vol. 2, 518520, Jul. 2000.
Koo, V. C., Y. K. Chan, V. Gobi, T. S. Lim, B.-K. Chung, and
H.-T. Chuah, The MASAR project: Design and development,
Progress In Electromagnetics Research, Vol. 50, 279298, 2005.
Chan, Y. K., B.-K. Chung, and H.-T. Chuah, Transmitter and
receiver design of an experimental airborne synthetic aperture
radar sensor, Progress In Electromagnetics Research, Vol. 49,
203218, 2004.
Gobi, V., B. K. Chung, and H. T. Chuah, Design of a
microstrip patch antenna for airborne SAR applications, Journal
of Electromagnetic Waves and Applications, Vol. 19, No. 12, 1687
1701, 2005.
Chua, M. Y. and V. C. Koo, FPGA-based chirp generator
for high resolution UAV SAR, Progress In Electromagnetics
Research, Vol. 99, 7188, 2009.
International Telecommunication Unions (ITU) World Radiocommunication Conference 1997 (WRC-97).
Ulaby, F. T. and M. C. Dobson, Handbook of Radar Scattering
Statistics for Terrain, Artech House, Norwood, 1989.
Cumming, I. G. and F. H. Wong, Digital Processing of Synthetic
Aperture Radar Data, Artech House, Norwood, 2005.