Miniature UAVs &

Future Electronic Warfare

Dr Anthony Finn
Dr Kim Brown
Dr Tony Lindsay
EW & Radar Division
DSTO, Edinburgh, SA 5111
Abstract
A capable adversary will use a combination
of SAR, ISAR, early warning, and MTI
radars, as well as EO, IR, and ESM sensors
networked together using tactical data links
to provide a complete situational awareness
picture. For a small defence force such as
the ADF, which has historically depended
upon a few high value assets to act as force
multipliers, this means that its platforms
will be difficult to conceal and
commensurately easy to target. In order to
protect against this we are undertaking a
research project aimed at prosecuting an
EW campaign through the use of small,
inexpensive, autonomous, cooperative air
vehicles. This combination of attributes
provides a capability that, even if it is
possible for an adversary to target the
individual vehicles, due to the distributed,
autonomous, adaptive, and robust nature of
the system it is difficult to counter
effectively.
The current program focuses on the
location and suppression of targets such as
radars through the use of a geographically
distributed and heterogenous mix of sensors
that are autonomous and capable of
cooperative behaviour. The individual
sensors are relatively unsophisticated,
however, the observation system that is
created through the cooperation and
adaptive networking of these sensors
provides sufficient process gain to achieve
results similar to those of significantly more
expensive centralised systems, but with
the added advantage of achieving
robustness to an adversary's counter-
measur es whi l st si mul t aneousl y
maintaining operational capability at a
reduced cost.
The program comprises a trials program
that uses the Australian-made Aerosonde
UAV, the study of team formation,
distributed sensing, data fusion, sensor
resource and energy management, and
communication link control based on the
concept of cooperating machines, the
development of intelligent agent algorithms
based on the self-organising behaviour
observed in ant colonies, and the
implementation of these algorithms within a
simulation environment.
UAV's can contribute in all aspects of
Electronic Warfare, from jamming and
Suppression of Enemy Air Defence
(SEAD) to Electronic Support Measures
(ESM), and Signals Intelligence (SIGINT).
The inherent range advantages enjoyed by
EW/ES payloads make them the natural
sensor of choice for cross-cueing payloads
with shorter ranges and/or more restricted
fields of view such as SAR or EO/IR
sensors. EW fits for UAVs can also include
SIGINT payloads, or defensive sensors that
can perform a SIGINT role. For example, a
Radar Warning Receiver (RWR) can be a
source of vital information, particularly
when related to imagery information to
form a more complete or accurate
Situational Awareness picture or when
updating the Electronic Order of Battle. The
key is the integration of the inputs from all
of the vehicle's sensors or all of the sensors
on all of the vehicles, in the case of smaller
more distributed systems that use a
heterogenous mix of sensors.
1. Introduction
Electronic Warfare (EW) is vital to all types
of military operation. The ADFs capacity
to identify, characterise, locate, exploit, and
suppress the electromagnetic emissions of
an adversary is crucial to this objective as it
allows for the establishment and mapping
of the adversarys electronic order of battle.
In particular, integrating EW information
(which often includes ID and intent) with
surveillance and imagery data (which does
not) provides a very much more complete
situational awareness (SA) picture.
Unfortunately, all too often EW is over-
looked - until the shooting starts.
Unmanned Aerial Vehicles (UAVs) have a
vital role to play in the prosecution of EW
campaigns. Similarly, EW has a vital role to
play in the protection of UAVs. To exploit
this synergistic relationship we need to
make use of the latest miniaturised EW
equipment. For example, a communications
jammer or an Electronic Surveillance (ES)
receiver can supplement or even replace a
UAVs main payload. An EW module such
as a Radar Warning Receiver (RWR) can
also be used to provide the UAV with some
advance threat warning. Alternatively, the
UAV and its payload can form some part of
a higher-level total capability.
This paper discusses:
a) Issues relevant to the physical and
electromagnetic vulnerabilities of UAVs,
b) The potential for EW payloads to
simultaneously enhance the capabilities and
reduce the vulnerabilities of these UAVs,
c) The potential applications of smaller
UAVs in an EW campaign.
The paper also describes a DSTO program
of work that is exploring the trade-offs
between the larger, more sophisticated,
platform-centric UAV/payload options and
the smaller, cheaper, distributed, network
centric options that are available.
2. UAV Vulnerabilities
In Afghanistan the UAVs appear to have
achieved excellent results for few losses.
However
Our adversaries are not blind to this,
The opposition in Afghanistan do/did not
have access to sophisticated air defences;
When the ADF goes up against a more
capable adversary the UAVs will attract a
great deal of unwelcome attention;
Shoot-downs may not result in loss-of-life
or hostages, but they do represent loss of
opportunity and expenditure of resources;
There are very few political ramifications
for an adversary destroying one of these
UAVs should the level of hostilities be less
than open conflict;
EO/IR payloads are effectively blinkered
by their narrow fields of view;
The cost of the airframe is relatively high,
slow and relatively easy to target, but not
sufficiently expensive enough to merit the
inclusion of capable EW self-protection
systems;
This leaves the UAVs vulnerable to 30-
40 year old SAM technologies (UAV losses
during the Kosovo campaign are thought to
be approximately one every 3 days in part
due to an air defence that was typical of the
Soviet pattern from the 1960s era [1]);
Put an emitter such as a SAR on board
and, from an EW perspective, the UAV
becomes a slow moving, highly luminous
sitting duck.
Another problem for the larger UAVs is
that their development process parallels that
of larger (manned) aircraft, which stresses
longer life, a high level of maintainability, a
multi-role capability, and high reliability.
The resulting systems are expensive with
life-cycle costs and logistic complexities
approaching those of manned aircraft.
3. Miniature UAVs
UAV platform technologies (eg. guidance,
airframes, control, etc.) are now sufficiently
mature in the small-to-medium sized range
to be considered as part of the ADFs
experimental force mix options. Moreover,
because UAVs of this size are now being
fielded, and a range of technologies for new
and advanced payloads is readily available,
we can expect a rapid evolution in the
available platform and sensor mix. The
mission characteristics expected of these
UAVs and their associated concepts of
operations are also likely to change
accordingly.
These Affordably Expendable alternatives
[2] offer the prospect of a capability that
allows the warfighter to conduct the high
value, high risk missions that are beyond
the capability (or justifiability) of other
systems. It relies upon the notion that the
useful life of the capability is a function of
its constituent payloads and technologies
rather than the physical life of the airframe.
The situation is similar to that of the
modern computer. Early computers were
physically large, essentially stand-alone
devices that were expensive to produce and
maintain. Todays computers, on the other
hand, are small and inexpensive platforms
that are extensively networked and, when
they outlive their usefulness due to
advances in technology, are discarded even
though they still function.
There is, of course, no free lunch: smaller,
less expensive, lighter UAVs are generally
less capable than their larger, more strategic
counterparts. Moreover, they also carry less
capable payloads. However, this may be
offset by the increased affordability of the
systems, our ability to network the UAVs
and sensors to derive process gain
1
, and our

1
A simple example of process gain is depth
perception from stereo-scopic vision.
capacity to withstand losses due to conflict
and malfunction.
Given the recent advances in sensor
miniaturisation, data links, and fusion
technologies, it may well be that a
combination of several smaller UAVs
standing in at lower altitudes with a few
manned platforms (or larger UAVs)
standing off and at higher altitudes provides
the optimal mix for the largest range of
applications. Whatever the eventual
solutions, it is clear that the full breadth of
payloads and CONOPS is only just
beginning to be explored.
In addition to the exploration/evaluation of
operational concepts either in their own
right or (appropriately scaled) for their
larger counterparts, as mini-UAVs and
their associated payloads can be acquired or
developed at a fraction of the cost of the
larger systems, they can also be used as a
framework to explore more experimental
concepts such as Network Centric Warfare
(NCW).
4. AVATAR Key Initiative
NCW exploits concepts such as information
superiority to provide a competitive edge in
warfare. NCW emcompasses the ability to
collect, process, and disseminate an
uninterrupted flow of information while
exploiting and/or denying an adversarys
ability to do the same [3]. Under the
Autonomous Vehicle Advanced Tactical
Applications Research (AVATAR) Key
Initiative, DSTO plans to use formations of
networked, autonomous vehicles as a model
to explore the capability edge obtained
when networking multiple platforms and
sensors.
AVATAR is aimed at developing a multi-
di sci pl i nar y, cr oss- envi r onment al
framework for demonstrating the advanced
applications of autonomous, uninhabited
vehicles [4]. Its focus is on experimentation
in the hands of the warfighter. The aim is to
investigate the potential for advanced
capabilities through the exploitation of new
technologies and novel concepts of
operation for networked, multi-platform,
autonomous vehicles, sensors, and
effectors.
To develop the airborne aspect of this
experimental capability and to explore the
capability edge as a function of technology
innovation and application innovation, we
required the development of a flexible and
accessible autonomous aerial vehicle
infrastructure. The Electronic Warfare
Division (EWRD) of the Defence Science
and Technology Organisation (DSTO)
recently signed a contract with Aerosonde
Ltd to provide this infrastructure through
the use of their mini-UAVs. The
capabilities of the Aerosonde UAV are
shown in Table 1 (for more details on the
UAV, readers are advised to see the website
www.aerosonde.com.au).
The six Aerosonde UAV's purchased under
the contract will be used to conduct a series
of trials over a two-year experimental
program. In addition to funding purchase of
the airframes, the contract makes provision
for the trials program, some experimental
payloads, a ground control/support element,
and a launch and recovery system.
It is important to note that AVATAR is
aimed at experimentation and fielding
proof-of-concept systems in order to inform
future capability analysis, acquisition or
upgrade. It is not aimed at defining final
capability solutions. The significant
restrictions associated with limited payload
and power can be viewed as either a
technology challenge or simply as factors
that limit the scale of the experiment (eg
data link range or data rate). Clearly, some
CONOPS and payloads will not scale down
to enable sensible experimentation.
Specification
Weight 13 - 15 kg
Wing span 2.9 m
Engine 24 cc, Avgas, 1kw
Fuel Consump 180g/hr level flight
Full Fuel Load 5 kg
Navigation GPS/DGPS
Communication
Range via UHF
200 km depending on
height and terrain
On board power
generation
40 Watt continuous
60 Watt peak
Payload
Computer
Supports Serial
Interface Input
Main Payload
Bay Area (can
be adapted)
100mm Length
120mm Width
180mm Depth
Performance
Speed, Climb 18 32 ms
-1
Climb > 2.5 ms
-1
Endurance,
Range
No Payload > 30 hrs
Clear air >3000km
Altitude Range 100 m to > 6000 m
Payload Maximum 5 kg (gives
approx 10-hour flight)
Landing & take
off distance
Less than 300 m
Take off speed Average 90 km/hr
Operation
Staff for
Launch and
Recovery
3 ~ Controller,
Engineer,
Pilot/Maintenance
Staff for Flight
Command
1 Person for several
aircraft
Ground
Equipment
Proprietary Staging
Box, 2 PCs, GPS
Flight Fully autonomous, or
under Base Command.
Launch and
Recovery
Launch from car roof
rack, land on belly
Max speed 31 m/s (110 km/hr)
Ground & air
comms
UHF or Satcomms
to/from Aerosonde
Table 1: Aerosonde UAV Specifications
Nevertheless, the opportunities offered by
the miniaturisation of the UAVs and their
payloads mean that there is considerable
potential for new approaches to old
problems and that we must explore very
different cost-capability equations.
The overall AVATAR program goal is to
analyse, develop, and field - in conjunction
with all that wish to participate - a varied
range of baseline operational concepts,
flight and sensor algorithms, and payloads
suitable for small, tactical UAVs. A crucial
part of this research will be the
demonstration of combined platform-UAV
operations at a relatively sophisticated level
(i.e. UAV experimentation in conjunction
with ADF assets).
5. Mini EW Payloads
There are a variety of potential payloads
suitable for mini-UAVs. These include:
communications & electronic intelligence
(SIGINT) payloads, communications and
radar jammers, electro-optic, infra-red, and
MAW sensors, MTI and SAR radars, BDA
sensors, comms relays, EW self-protection
suites, chemical, biological, & nuclear
detectors, target designators, and horizon
extenders. Only a selection of the EW
payloads is discussed here.
Electronic Surveillance (ES) Payloads
ES and SIGINT sensors can be a source of
vital information, particularly when related
to imagery information to form a more
complete or accurate Situational Awareness
picture or when updating the Electronic
Order of Battle. The key is the integration
of the inputs from all of the sensors on all
of the platforms. Moreover, as the sensor
only has to receive and process signals it
does not require large amounts of power to
operate it. Consequently, ES payloads scale
well to the power constraints of mini-
UAVs.
Figure 1: Geolocation error ellipses for 0.5
deg. rms DF sensors onboard 2 platforms
with stand-off range of 100km. AOI (100km
x 100km) is bounded by blue line.
Figure 2: Geolocation error ellipses for 5
deg. rms DF sensors onboard 8 platforms
with stand-in capability. AOI (100km x
100km) is bounded by blue line.
Mini-UAVs are not considered a practical
replacement for the larger strategic UAVs,
although by networking their sensors we
may derive greater capability. As an
example, Figure 1 and Figure 2 show a
comparison between the 50% uncertainty
bounds for geolocating
2
emissions of
interest from two platforms with 0.5degree
rms Direction Finding (DF) capability and
eight platforms with 5degree rms DF
capability, respectively. The more accurate
sensors are nominally placed on board

2
Geolocation is the process of determining,
either directly or indirectly (eg. through DF), the
location of an emitter of interest.
high-value assets and must therefore
standoff at a range of 100km, whereas the
less capable sensors, which are significantly
smaller and cheaper, are placed on more
expendable platforms and may therefore
stand-in (their cost means that we are also
able to afford more of them). Analysis of
the figures shows that system using the less
accurate sensors has errors around 50% less
than those of the more expensive system.
As another example, Figure 3 shows the
trade-off between geolocation error and
range for a variety of Direction Finding
(DF) sensor accuracies. The figure clearly
shows that situational awareness and even
targeting level accuracies are achievable
using ES/ELINT sensors with very modest
DF capabilities, if stand-in capability can be
achieved.
Figure 3: Geolocation Error (m) versus
Range for 0.1-15 degree rms DF sensors
enjoying optimum geometry.
Geolocation may be achieved in a number
of ways other than using DF techniques and
the reader is referred to [5] & [6] for more
details on these techniques. By and large,
however, the accuracy of the techniques is
heavily dependent upon the separation
distance between the sensors and their
geometry relative to the emitter.
Clearly, if geometry is a major component
of the geolocation problem then, for a
single platform, the time that it takes to
obtain a given level of accuracy is
dependent upon the time taken to
manoeuvre to obtain this geometry. These
manoeuvres can be very time-consuming
and the emitter may move some
considerable distance in the intervening
periods. Using multiple platforms
geolocation can be performed in near real
time. Moreover, the distribution of the
payloads means that the system geometry
can be adapted more quickly to favour a
small geolocation error. This provides a
double incentive to obtain geolocation
using multiple platforms.
Figure 4: Relative geolocation error versus
the number of DF sensors located around
an emitter [6].
Electronic Attack (EA) Payloads
A jamming platform must stand off at a
considerable range from a target to allow
for its own protection. Because it must
stand off, it requires a large amount of
power. By reducing the size of the platform
and the need to protect it, we are able to
stand in, which means that we need
significantly less power to jam a given
target. In addition to this, because the stand-
in jammer is closer to its target its
transmissions cover a smaller area and the
potential for electro-magnetic fratricide is
also significantly reduced.
Figure 5 shows the Jammer-to-Signal Ratio
(JSR) as a function of range for a 100W
jammer and 10kW radar transmitting into a
20dB directional antenna. The radar returns
1
0
0
0
k
m
1
2
h
r
1
0
h
r
7
h
r
0
h
r
Tim
e
on
stati
on
are based on the detection of a target with a
radar cross-section roughly the size of a
(non-stealth) strike fighter.
Figure 5: Jammer-to Signal Ratio versus
Range for a 100Watt noise jammer against
10kW radar with 20dB antenna gain
attempting to detect a target of 1m
2
RCS.
The red, green, and blue lines are for target
ranges of 1, 10, and 100km.
The figure shows that the achievable JSR
from a 100W jammer located 10km from
the radar and protecting the target at a range
of 10km is equivalent to a 10kW jammer
located 100km from the radar attempting to
protect the same target. When one considers
that many modern weapons systems have
ranges in excess of 100km and that
miniature UAVs are also hard to detect
(and hence target) it makes the technologies
a very attractive potential alternative. In
addition to this, even if the mini-UAVs are
detected, targeted, and engaged, the UAVs
have such small IR signatures and RCS
there is still no guarantee that the weapons
will fuse correctly and destroy the UAVs.
Horizon Extenders
In addition to placing the sensor/processing
capabilities onboard the UAV, it can be
treated as a flying antenna. In this case,
the UAV is effectively an electromagnetic
bent pipe comprising a receive antenna,
a modest amount of processing capacity,
some time stamping & signal amplification,
and a (directional) transmit antenna.
Combined with knowledge of the UAVs
location and a high gain antenna on the
ground (and considerably more processing
power than is available onboard the UAV)
we have the capacity for significant horizon
extension.
Figure 6: Radio line of sight vs. altitude.
Figure 6 shows the rapid increase in the
potential line of sight (LOS) as a function
of UAV height. It can be seen that even for
very modest altitudes of about 500m, LOS
of around 100km are achievable.
Operational Issues
Figure 7: Endurance capabilities of the
Aerosonde UAV in geographic perspective
for a 2kg payload (12 hrs on station ~ black
continuous circle to 0 hrs on station ~ red
dotted circle)
Figure 7 and Figure 8 show the typical
endurance capabilities and coverage that
could be achieved using the ES payloads
referred to above when flown on the
Aerosonde mini-UAV. Control of the UAV
in the area of operations could then be
undertaken either by a commander at the
launch site or by one based further forward
(eg. onboard a ship or tactically deployed).
1000 500 0
km
1000 500 0 1000 500 0
km
Figure 8: Typical ES sensor coverage for
two areas of interest - the range estimates
assume a launch from Darwin, a 2kg
payload, and an altitude of 3.5km. The time
on station is estimated as 15 hours.
6. DSTO Trials
In February and June of this year, DSTO
conducted a series of trials to demonstrate
potential operational concepts for miniature
UAVs involved in maritime EW operations.
The mini-UAVs were designed, built, and
operated by Aerosonde Ltd. The payloads
were designed, built, and operated by
DSTO. The payloads tested were:
EA ~ 100MHz bandwidth noise jammer,
RF repeater ~ a test target of controlled
radar cross-section used with EA payload,
ES ~ 218 GHz superheterodyne receiver
The combined objectives of the trials were:
1. Demonstrate successful launch of the
Aerosonde UAV carrying the EW payloads,
2. Demonstrate hand-over of control of the
UAV to a surface combatant and back again
to the land-based launch/recovery crew,
3. Determine detectability of the Aerosonde
UAV using the ships radars and ES,
4. Determine the effectiveness of the EA
payload installed on an Aerosonde UAV,
5. Determine the effectiveness of the ES
payload onboard an Aerosonde UAV,
6. Demonstrate the surface combatants
extended horizon capability by enabling
it to control the UAV-based ES payload and
receive/process surface emissions detected
over the horizon from the ship,
7. Demonstrate injection of the ES data
collected by the payload onboard the UAV
into DSTOs EXC3ITE network,
8. Demonstrate the TECHINT value-add
of the ELINT/ES product within EXC3ITE,
9. Demonstrate the dissemination of this
product to the Joint Command Support
Environment (JCSE) in near real time.
10. Demonstrate successful landing back at
Jervis Bay carrying the EW payloads.
Figure 9 shows a schematic of the DSTO
trials conducted in June and February.
LOW BW LOW BW
COMMS COMMS
VALUE ADD TO VALUE ADD TO
ELINT PRODUCT ELINT PRODUCT
JCSE JCSE
EXC3ITE EXC3ITE
ACCESS ACCESS
VIA WEB VIA WEB
TOOLS TOOLS
TACTICAL TACTICAL
ESM/ELINT ESM/ELINT
SA- 6
RESTRICTED -Mock-Up Data
Map shows spatial features
including SA-6 site
Clock shows temporal activity
of SA-6 site (from ELINT)
CONTROL CONTROL
LINKS LINKS
TACTICAL TACTICAL
USERS USERS
LOW BW LOW BW
COMMS COMMS
VALUE ADD TO VALUE ADD TO
ELINT PRODUCT ELINT PRODUCT
JCSE JCSE
EXC3ITE EXC3ITE
ACCESS ACCESS
VIA WEB VIA WEB
TOOLS TOOLS
TACTICAL TACTICAL
ESM/ELINT ESM/ELINT
SA- 6
RESTRICTED -Mock-Up Data
Map shows spatial features
including SA-6 site
Clock shows temporal activity
of SA-6 site (from ELINT)
CONTROL CONTROL
LINKS LINKS
TACTICAL TACTICAL
USERS USERS
Figure 9: Schematic of mini-UAV trials
The trials are reported in detail elsewhere
[7] & [8]. The February trial comprised EA
and a shake-down of the ES payloads.
The June trial may be summarised as:
At launch, a pilot at the Aerosonde Launch
and Recovery Site (ALRS) manually flew
the UAV before switching it to its
autonomous mode. Control of the UAV and
its payload was then passed to another land-
based control site or to the ship, depending
upon the trial. A DSTO controlled emitter
was located, as were a number of other
marine radars. The radars and transmitter
were used as test signals for the ES
payload. Signals information received at the
UAV was passed to the ship and the land-
based control sites. On being received at the
land-based control site, the information was
injected into DSTOs EXC3ITE network
for TECHINT value-add and dissemination
to the JCSE in near real time.
One of the navigation radars was used as a
test-radar for the EA payload. A second
UAV, with an RF repeater payload, was
used as a controlled radar cross-section and
flown in conjunction with the UAV with
the EA payload. The navigation radar was
successfully jammed (Figure 10).
Figure 10: Navigation radar display
showing jamming strobe due to EA payload
onboard mini-UAV
All of the objectives referred to above were
achieved successfully.
7. Future EW
The ADFs current approach to EW is
platform-centric in nature and requires each
platform to rely upon its own sensors. This
approach requires extremely fast responses.
It does, however, allow those responses to
be tailored to the relevant threat. Achieving
that response requires rapid detection and
classification of the emitters and is typically
obtained from an onboard ESM receiver
loaded with the appropriate threat libraries.
Unlike platform-based EW, which is
tactical, short-range, reactive, requires very
fast responses to the threats, and is often the
last defence option, Force Level EW
3
can
be operational, pre-emptive, and long range
in nature, requires more modest response
times and is an early defence option.
For the ADF, the areas in which FLEW is
likely to make a impact include [9]:
The surveillance and monitoring of a
region of interest (including the detection,
identification, geolocation, recording, and
analysis of electromagnetic signals that are
potentially linked to an adversarys sensors
or C3I systems),
EA for counter-surveillance or counter-
acquisition of radars, imaging, or C3I
systems in both a strike context and a
defensive role,
The use of anti-radiation missiles
(ARMs) to destroy hardware and suppress
enemy air defences,
The development of operational concepts,
doctrine, and tactics (for the use of FLEW),
The assessment and allocation of
priorities to requirements for equipment
acquisitions,
Through-life-support to maintain, sustain,
and enhance EW-capable datalinks and
applications that exploit shared EW data
(such as situational awareness, geolocation,
and remote/third-party targeting of EA and
ARMs),
The development and maintenance of
databases for EA techniques for counter-
C4ISR and the provision of ES and EA
reprogramming data, tactics, advice, and

3
Force Level EW is defined as the offensive or
defensive application of EW in such a manner
as to have effects beyond the immediate self-
protection of a platform. In other words, where
EW is used to attack or protect a group of assets
or individuals within a group, excluding the act
of self-protection. It is achieved by networking
sensors, data fusion centres, command and
control systems, and weapons systems.
analysis via secure communications from
support areas to forward areas,
The assessment of derived sensor and C3
arrangements of potential adversaries for
vulnerability to EA,
The investigation of the vulnerability to
EA of the ADFs C4ISR arrangements and
the implementation of appropriate
protective measures,
The research, development, and
application of both hardware and software
technologies to produce and enhance
integrated EW situational awareness and
EA capabilities, and
Liaison with civilian authorities on
matters relating to an operational EW
capability.
The implementation of FLEW is more
complex than simply widening the
communications bandwidth and linking the
platforms. It includes the interpretation and
application of information of uneven
quality and timeliness, the purification of
information to preserve the quality of the
network and prevent error propagation, and
the coordination of assets to obtain a
synchronised response. In addition to this, it
must be achieved within a framework of
finite resources [10].

Ph
ysi
cal
E
M
Cy
be
rsp
ace
Battle
space
Effectors
Agents
Decision
Makers
Data Links
Sensors
Data Links
Figure 11: Relationship between Force
Level EW entities and elctromagnetic
conflict.
The coordination and synchronisation of
assets has the potential for improving EW
operations. However, it also increases the
matrix of options available to the ADF. The
study of Complex Adaptive Systems (CAS)
shows that the most striking benefits come
not from the direct linkages and
connections between the entities, but from
the vastly larger number of indirect
connections that are established [11].
The requirements for FLEW transcend the
traditional environmental boundaries of air,
land, and sea and also the operational
boundaries of tactical and strategic. In both
cases, this is because the sub-systems
involved in delivering the EW product
occupy multiple domains. Consequently, it
will be necessary to analyse the use and
integration of these entities within a joint
operations framework that spans both the
tactical and strategic environments. For
these reasons, it is instructive to break the
battlespace into three overlapping domains:
physical space, cyberspace, and electro-
magnetic space. For each of these domains
the means by which the ADF may affect an
outcome may be broken into four blocks
(Figure 11):
a) Sensors, which provide information to
form the basis of situational awareness,
b) Agents, which perform selected tasks
delegated by the effectors/decision-makers,
c) Decision-makers, which exercise C2, and
d) Effectors, which execute the plans and
prosecute the engagements.
These are the standard entities referred to in
NCW texts, and FLEW draws heavily from
some of the concepts in this area. The
architecture that supports FLEW is heavily
dependent upon the nature and capabilities
of the datalinks used to support the sensor-
processing, value-adding, and C2 processes.
In the FLEW concept of operations the
participants are heavily dependent upon a
network of continuously adapting systems
that are capable of making strategic choices
about how to survive and achieve their
goals in a dynamic environment. The agents
are enabled to perform a range of selected
tasks delegated by other participants,
decision makers, or controllers. The power
of the networked force is thus derived from
the strong networking of a well informed,
but geographically dispersed force.
To visualise the potential benefits of FLEW
and UAV networks it is instructive to
analyse the Reactive EW
4
and Geolocation
problems. In a platform-centric engagement
the ESM receiver is co-located with its EA
assets and radar sensors and the receiver is
allocated only a limited duty cycle because
of the look-through problem
5
. Clearly, the
probability of intercepting the threat signals
is a function of the duty cycle (amongst
other things). Moreover, the response delay
of the platforms countermeasures is a
function of the sensor and processing
latencies. Achieving a response to multiple
threats within a limited duty cycle is
therefore extremely challenging, as rapid
detection and classification of the emitters
is needed. Geolocation is similarly affected.
First, if DF techniques are used the level of
accuracy that can be achieved within a
given time frame is restricted, as the
platform must spend time manoeuvring to
obtain the necessary geometry. Second, the
probability of intercepting certain types of
signals is obviously affected by the look-
through problem. If, instead of DF, TDOA
techniques are used to geolocate the signals,
the geometry is similarly unfavourable as
the sensors antennas are limited in their
distribution to the extremes of the platform.
Let us now consider the situation where the
ES sensors are distributed across a number
of geographically dispersed platforms. By
placing dedicated sensors on the platforms
and combining the information available
from each of them we are able to improve
(if not completely remove) the look-through
problem. Similarly, we are able to improve

4
Reactive EW is the deployment of electronic
countermeasures in response to an EW threat
(eg. jamming a seeker on an incoming missile).
5
Look-through is the inability to receive and
transmit at the same time due to the co-location
of the ESM and jammer/radar.
the likely detection of signals through the
cumulative probability of intercept of
multiple sensors. On the other hand, the
information must be passed over finite
bandwidth communications links and can
be expected to experience network delays.
Thus, the multi-platform case offers
reductions in the response delay to a threat,
but only if the gains in sensor and
processing latency are not overwhelmed by
network latencies. This may require new
datalinks and more intelligent ways of
passing information.
Clearly, in terms of achieving specific
levels of geolocation accuracy within a
given time frame, for any fixed level of
payload capability, the multi-platform case
provides superior results because of the
distributed geometry. That is, in the case of
DF techniques, multiple cuts or fixes
are available simultaneously and in the case
of TDOA techniques, the dispersal of the
sesnors antennas is no longer restricted to
location on a single platform (which means
that longer baselines may be used).
In addition to the above, in the Force Level
EW environment it is possible to use a
heterogenous mix of sensors to exploit both
the diversity of the available information as
well as its geographic distribution. For
instance, we can use a wideband IFM
6
receiver to cross-cue one or more
narrowband superheterodyne
7
receivers. As
before, combining the signal time-of-arrival
information with the UAV locations
provides a geolocation capability. This
time, however, the Probability of Intercept
(POI) is enhanced and the description of the

6
Instantaneous Frequency Measurement (IFM)
receivers can typically monitor a wide band of
frequencies continuously (eg all spectrum from
2-18GHz), but they provide poor information on
the signals received.
7
Superhetrodyne receivers monitor only a
narrow band of frequencies (eg 1GHz segment),
but provide more accurate signal information.
signal likely to be more accurate (and the
commensurate signal ID).
If we now combine the control of these
payloads with the autonomous control of
the UAVs through the use of Intelligent
Agents
8
and provide these agents with a
communications architecture that allows the
information to be passed from agent to
agent we have a system that is potentially
able to adapt to its dynamic environment.
Depending upon the nature of the corporate
and individual goals of the UAVs and their
payloads the structure of the network may
vary greatly, but it allows the strengths of
the individual agents to be combined into a
single cohesive team [11]. This intelligent,
autonomous, adaptive, and co-operative
behaviour is an important component of the
NCW and FLEW concepts and DSTO has a
program of work that is exploring a range
of options pertinent to their development.
Currently, these concepts are implemented
in simulation environments, such as the
FLEWSE (Force Level EW Simulation
Environment ~ pronounced floozy [12]).
8. Conclusions
EW needs UAVs and UAVs need EW.
The synergistic relationship can provide the
ADF with enhanced Situational Awareness,
a significantly superior capacity to conduct
EW operations, and much better protected
UAVs.
Miniature UAVs are of particular interest
to EW operations as these UAVs are likely
to undertake some of the most dull, dirty,
dangerous, and impossible missions (D3I).
The affordability of these smaller UAVs

8
An Intelligent Agent is an autonomous
algorithm that, having been provided with some
over-arching mission objective, senses, reasons,
and acts upon its environment independent of
any external command structures. A typical
deployment philosophy might be go over there,
fix the problem, and dont do any harm.
also provides us with the opportunity to
acquire or develop suitable payloads for
these missions. In addition, it allows for the
development, experimentation with, and
evaluation of (scaled) operational concepts
for the larger UAVs and more advanced
concepts, such as NCW.
DSTO has embarked on the AVATAR Key
Initiative. The initiative is aimed at
developing a multi-disciplinary, cross-
environmental framework for demon-
strating the advanced applications of
autonomous, uninhabited vehicles. The aim
is to investigate the potential for advanced
capabilities through the exploitation of new
technologies and novel concepts of
operation for networked, multi-platform,
autonomous vehicles, sensors, and
effectors. Its focus is on experimentation
in the hands of the warfighter.
Ackowledgements
The authors would like to thank Dr Don
Koks and Mr Richard Bowyer of EWRD
for their helpful contributions to this paper.

References
[1] Steven J. Zaloga, The Evolving SAM
Threat: Kosovo & Beyond, J. Electronic
Defence: pp45-50, May 2000
[2] Tactical EW Division, NRL, Affordably
Expendable UAVs, Paper prepared for US
Secretary of Navy, April 1999
[3] Alberts, D., J. Gartska, & F. Stein,
Net work Cent ri c Warf are, CCRP
Publication Series, Washington DC, US
Department of Defense, 1999
[4] Lindsay, A.C., AVATAR Key Initiative -
Proposal for Scope, Minute to CDS, April
2002

[5] TTCP-EWS-TP4, Next Generation
Tactical ES Geolocation Technologies,
Proceedings of Workshop held at NRL,
Washington, April 1999
[6] D. Koks, Passive Geolocation for
Multiple Receivers with No Initial State
Estimate, DSTO Report, DSTO-RR-0222,
November 2001
[7] K. Brown, Miniature UAV EW Payload
Navy Trials, Draft DSTO Report, Sept 2002
[8] C. Howard & K. Mason, Force Level
EW And Situational Awareness For Future
Littoral Warfare, Land Warfare Conference
2002, Brisbane, 22-24 October, 2002
[9] Rockliff, S., An Assessment of R&D to
be undertaken for Force Level EW, DSTO
Report DSTO-TR-1018, Oct 2000
[10] Sciortino, J. et al, The Role of Artificial
Intelligence and Complex Adaptive Systems
in Network Centric Warfare, Proceedings
of the 45
th
Joint EW Conference, Monterey,
California, May 2000
[11] R.S. Bowyer & R.E Bogner, Agent-
Enabled Sensor Networks for Target
Observation, Advanced Information
Systems in Defense, Edited by Prof. L.C.
Jain, published by Springer-Verlag, 2002
[12] Finn, A. & G. Chalmers, Operations
Analysis Methods for Force Level
Electronic Warfare, DSTO Report, DSTO-
TR-0298, June 2002