-
8/9/2019 Adapting EW to Radar Waveforms
1/4
ADAPTING ELECTRONIC WARFARE RADAR
WAVEFORMS
L. R. Falk
Swedish Defence Research Agency, 16490 Stockholm, [email protected]
Keywords: electronic warfare, radar, network, deception,
information flooding
Abstract
Modem methods
of
electronic warfare must consider that
many sources
of
information are available to sensor systems.
Situation awareness is created in a network of sensors and
human operators by a process analysed in this paper. Each
step in this process suggests a method of electronic warfare.
The analysis is based on a division of radars into two basic
groups, surveillance and tracking radars, depending on the
amount of information received. Human knowledge is used in
surveillance systems and this makes information flooding an
effective method of attack. Distraction with noise and chaff
and deception with decoys may be preferred in other cases.
The conclusions have been tested in trials performed with the
Swedish air surveillance system during staff training.
1 Introduction
The development of electronic warfare is affected by the large
amount of information available to modem sensor systems.
Sensors are organized in networks and can combine data with
knowledge obtained from human sources.
This process has been used for many years in air surveillance
systems and is known to produce good results. A network is
flexible and resilient to jamming because information is
distributed among the nodes. The problem
of
attacking a
network is addressed by investigating how an optimized
analysis is performed and how it can be affected [1-3].
Networks appear in many forms but some conclusions can be
drawn by applying information theory [1,5,6]. The idea is to
study how sensor information is used and how human
knowledge is included to improve the result. Traditional
electronic warfare is based on locating weaknesses in the
enemy system. Modem warfare is based on the assumption
that the opponent is using methods close to optimum to
analyse available information. Such an analysis leads to
certain general conclusions concerning the future of
electronic warfare.
The possibility of having abundant sensor information
analysed in an optimum fashion led to the idea
of omplete
situ tion w reness [3]. Military concepts like Network
Centric Warfare NCW and Network Enabling Capability
NEC have been accepted and correspond to a technical
revolution in communication and data storage.
From the point
of
view
of
electronic warfare this development
means that the traditional methods
of
jamming individual
sensors and communication lines must be replaced by the
complex obj ective of disturbing situation awareness. This
description is general enough
to
cover many situations, like
the traditional problem of protecting a ship or aircraft from
missiles. In that case the pre-programmed hypothesis and
choices of a target seeker are described in terms of situation
awareness.
2 Information from sensors
The amount of information available to military commanders
increased dramatically after WW
Electromagnetic waves
propagate in straight lines over long distances and this allows
one to create sensors like radar and optical system with
enormous range and great accuracy.
Airborne platforms carrying such long range sensors can
collect enormous amounts of data by mapping the ground.
They are also useful for early warning and surveillance tasks,
but in this case the expected number
of
targets is smaller and
human operators must enter the process.
Radars and optical sensors are also used to track targets. In
this case less information is needed but the time between data
updating is shorter. This difference is important, since it
allows one to divide radars and other sensors into two basic
groups: surveillance and tracking systems use different time
periods and deliver different amounts
of
data.
A surprisingly small data flow is required to guide a gun or
missile, usually less than 1000 bits/s [7]. Surveillance radars
deliver considerably more, say 1-1000 Mbits per period of
search, if they are searching for targets. The time of
measurement is longer than for trackers and human operators
are used to introduce knowledge into the process. This
method allows one to solve complex problems like
identifying unknown targets and assessing their intention.
Anti-aircraft systems fall between these two groups. They
contain both surveillance and tracking systems and are in fact
early examples of networks, just like fighter control and air
-
8/9/2019 Adapting EW to Radar Waveforms
2/4
Targets and non-targets detection)
2 Friends and non-friends FoF)
3. Threats and non-threats RoE)
surveillance systems. These systems involve several types o delimiting a problem and defining logically complementary
radars and optical sensor with different ranges and accuracies, alternatives as required y the Bayesian method [6]) is very
which must be combined in a network. hard in practice. One can compare the following tasks that are
presented in raising order
o
difficulty, where operators have
to decide between:he poi nt is that hu man oper ators can affect the pr ocess in
surveillance systems since they are slow. Human operators
assist the system by introducing additional knowledge, which
is rarely possible for tracking systems.
The fundamental question is how one can use data in a sensor
network to create situation awareness. This question will be
analysed by considering the optimum process for evaluating
available information.
is difficult to present a logical solution o these problems to
a c omput er a nd h uman o perato rs are th us requi red , th ou gh
computers perform much o t he p re pa ra to ry work. We will
not
try
to solve this difficult problem, but rather point out how
situation awareness can be destroyed when sensors are
supported by human knowledge.
Shannon s theory o in fo rmat io n sho ws t ha t the op ti mu m
method o processing information under stationary conditions
is to ap ply B aye si an p ro bab ili ty t he ory [6]. The p ro ce ss is
complicated but in principle one can include various forms o The preceding description o how information is used to
human knowledge into the analysis. generate situation awareness makes it possible to identify
some general principles o electronic warfare.
Thi s th eoreti ca l c onc lusio n will be u se d be lo w, t hou gh in
practice one would only use Bayesian processing at certain
stages. The optimal process is very slow and should be
replaced y other methods whenever possible.
From our point o view a sensor network is a system
c ol le ct in g a nd p ro ce ssin g l arge a mou nt s o data. Human
inf orm ation is im port ant b ut the am ount is small a nd this
makes a description in terms
o
information useful.
B ay esia n an al ysis prescrib es a p ro cess wh ic h ag re es wi th
com mon sense [3,6] and is guaranteed to b e optimal i the
following steps are used. All information is described in terms
o probability.
The basic methods o destroying information are
di ssimul at ion a nd distract ion. D issimul at ion co nsists in
concealing a target, while distraction is used to divert
attention from the object. These principles are used in biology
to inter pr et how animals act to hide from carnivores. The
same method can be applied to electronic warfare, since
military targets are also isolated objects appearing in
complicated environments. The process
o
c onfusin g a n
o pp on en t is most easily d escrib ed in a spac e de te rmin ed b y
the sensors. The structure
o
an information space will not be
ver y simple in general, bu t radars and optical systems are
constructed to produce few ambiguities when a single target
is observed.
Delimit and define all possible alternatives.
2. Describe their logical relations.
3. Collect data and assess their uncertainty.
4. Calculate the probability o all alternatives.
5. Formu la te a ssessmen t in t erms o probability.
6. Make a decision based on probability and acceptable
risks.
In this process the last step o forming a decision is separated
from the process
o
creating situation awareness. would be
difficult to obtain probability distributions for all alternatives
but this is rarely required in practice.
The point is that the process is separated into two stages.
Human operators only contribute to the first steps 1-2), while
th e foll owin g step s 3 -5) sho ul d b e h and le d y a computer.
This is true o tracking systems and missile seekers which are
programmed in advance to handle possible incidents. They
onl y per for m steps 3-5, while surveillance systems m us t
handle unexpected incidents with the assistance o human
operators. Steps 1-2 are time-consuming, since the process o
Figure
single target in information space
The simplest way o concealing a target is to use noise o r
r andom signals to cover it. This well-known principle is
di sad va nt age ou s wh en in format io n spa ce is l arge, w hi ch
-
8/9/2019 Adapting EW to Radar Waveforms
3/4
happens when many different sensors are observing the same
target. It is possible to calculate how unfavorable noise
becomes by comparing the volume o all space with the
observation cells surrounding the Such calculations
are performed in simple jamming cases to compare the
spectral width
o
the jammer with the effective spectral width
o
the radar signal. Traditional electronic warfare is almost
exclusively concerned with calculations concerning the effect
o noise.
The simplified diagram suggests another form o electronic
warfare based on distraction.
an attractive decoy is formed
it can divert attention from the target by indicating higher
probability for the decoy. This method works both for
surveillance and tracking radars, but information space is
smaller for tracking radars, since they will filter away
unnecessary data. The tactical effect o a decoy is stronger for
a tracking radar,
i
it is close enough, especially since less
time is available to correct an incorrect decision.
Figure Decoy and target in information space
A Bayesian calculation shows that the best method is to use a
single but convincing decoy to divert attention from the real
target. The most famous decoy operation was performed in
preparation for D-day in 1944.
German HQ regarded Calais as the principle target and this
belief was supported by various means. The operation was
successful, but this case must be regarded as exceptional. The
decoy was accepted because the Allies could read the German
secret telegrams and correct their signals to support the notion
that Calais was the real target. The basic reason for success
was that German
HQ believed the Enigma cipher machine to
be secure. This belief made the false information obtained
from double-cross spies appear trustworthy since they always
answered the right questions.
It is often difficult to foresee how an opponent would react to
a false target, unless it is an exact copy. Several operations
similar to D-day did not work because signals were never
noticed [4]. Decoy operations require detailed knowledge
o
the system and consequently long preparation. A typical case
is the Israeli attack in the Beqaa valley in June 1982, where
Syrian anti-aircraft systems were provoked to reveal their
positions by unmanned decoys
A third basic principle obtained from the simplified diagram
is a combination
o
dissimulation and distraction, produced
by creating several moderately credible false targets that
temporarily divert attention from the real target.
This principle can be used against surveillance radars. The
idea is to overwhelm the system with a large number
o
hypotheses that must be tested by human operators that can
handle only a few doubtful targets usually about one
millionth
o
the cells observed .
a sufficient number o false
targets is introduced and pass the machine filter they will
occupy the human operators by information flooding. The
basic difference between surveillance and tracking radars is
used here by exploiting that human operators will still handle
logical problems concerning identity and intention.
Information flooding has been successfully tested in trials
performed with the Swedish air surveillance system. This
method requires less detailed knowledge about the system
than decoys. Information flooding is possible i the targets are
sufficiently similar to attract attention and pass the initial
machine filters [1,2].
interesting point, confirmed by calculation, is that sensor
inputs used to produce false targets must correspond to the
sensors regarded as most reliable by operators. Otherwise
false targets rarely produce any effect.
Figure
3
real target surrounded by false targets will
distract human operators by informationflooding
-
8/9/2019 Adapting EW to Radar Waveforms
4/4
This principle is confirmed by tests and simulation. In fact,
classified tests show good agreement with the theoretical
conclusions. The theoretical view presented here should not
be expected to produce new methods
of
electronic warfare.
This subject has been thoroughly investigated by simulations,
but is useful to understand why certain general principles
seem to apply in all cases. Moreover, the difficulty of
producing effective jamming confirms that sensor networks
are resilient to electronic attacks.
5 Conclusions
Some simple conclusions follow from the analysis.
Noise is simple but often ineffective against large
networks.
Deception requires detailed knowledge and simple
situations.
Information flooding affects human operators.
Networks are resil ient to jamming.
The best protection is offered by training the
operators against electronic attacks.
Tests confirm theoretical conclusions.
Acknowledgements
The author is deeply indebted to his colleague Per Hyberg for
numerous discussions during this work. e re both indebted
to Olle MaIm and Michal Herre at the Swedish air
surveillance system for cooperation and support during
simulator trials in the course of staff training.
References
[1] L. Falk: The Benefits of Deception , MilTech 2
Conference, Stockholm, pp. 101-108 2005).
[2] L. Falk: Jamming the network: The Benefits of
Deception , AOC Conference, London 2006).
[3] L. Falk: Situational awareness and electronic deception
with historic examples , Stockholm Contributions in Military
Technology 2007, ed. Martin Norsell, pp. 83-98 2008).
[4] M. Howard: Strategic Deception in the SecondWorld
War Norton 1990).
[5] P Hyberg: Network Centric Warfare and Information
Theory , Journal ofElectronic Defense, Vol 28, Dec 2005).
[6] E T. Jaynes: Probability theory: The logic of science
Cambridge University Press 2003).
[7]
Kjellgren: A simple study
of
the information
requirements for missile guidance , Acquisition, tracking and
pointing XVII,
M K
Masten and
L
A. Stockum, Editors,
Proceedings of SPIE, vol. 5082, pp. 77-86 2003).
