ECCM

MODELLING RADAR-ECCM
A SYSTEM APPROACH
AK Subramanian
MODELLING RADAR-ECCM
A SYSTEM APPROACH
AK Subramanian
Defence Research & Development Organisation

Ministry of Defence
New Delhi 110 011
2003
DRDO MONOGRAPH SERIES
MODELLING RADAR-ECCM: A SYSTEM APPROACH
AK Subramanian
Series Editors
Editor-in-Chief Editors
Dr Mohinder Singh Dr JP Singh, A Saravanan
Coordinator
Ashok Kumar
Asst. Editors Editorial Assts. Cover Design

Ramesh Chander AK Sen, RK Saxena Anjan Das
Sanjay Kumar Kumar Amar Nath
Production
Printing Marketing
JV Ramakrishna RK Dua
SK Tyagi Rajpal Singh
Cataloguing in Publication
Subramanian, A.K.
Modelling radar-ECCM: a system approach.
DRDO monograph series.
Includes index and bibliography.
ISBN 81-86514-12-0
1. Radar 2. ECCM I. Title (Series)
623.623
2003, Defence Scientific Information & Documentation Centre (DESIDOC), Defence

R&D Organisation, Delhi-110 054.
All rights reserved. Except as permitted under the Indian Copyright Act 1957, no
part of this publication may be reproduced, distributed or transmitted, stored in a
database or a retrieval system, in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, without the prior written permission of the
publisher.
The views expressed in the book are those of the author only. The editors or publisher
do not assume responsibility for the statements/opinions expressed by the author.
Printed and published by Director, DESIDOC, Metcalfe House, Delhi-110 054.

Dedicated
to
Gita Subramanian
for
the inspiration, help and support
CONTENTS
Preface xi
Acknowledgement xiii
CHAPTER 1
INTRODUCTION 1
1.1 Importance of ECM 1
1.2 Background 2
1.3 Basic Considerations 2
1.3.1 Jamming Interference 3
1.4 ECM-ECCM Matrix Dynamics 4
1.4.1 ECCM EfficacyThe Difficulty 5
1.4.2 Survival 5
1.4.3 Judging Efficacy of an ECCM Element 6
1.5 Assessment of the State-of-the-Art 6
CHAPTER 2
MODERN RADAR DESIGN & ITS DEPLOYMENT
ENVIRONMENT 9
2.1 Introduction 9
2.2 Radar Performance 10
2.2.1 Two Representative Radar Models 10
2.2.1.1 Low Flying Detection Model 10
2.2.1.2 Land Clutter Model for Radar Design 11
2.3 Command, Control, Communication & Intelligence 13
2.4 ECCM Modelling 14
2.4.1 Radar Systems 14
2.4.2 Communication Systems 15
2.5 Mathematical Analysis 15
2.6 Simulation 15
2.7 The EW Flywheel 16
Reference
(viii)
CHAPTER 3
ECCM EVALUATIONSOME MODELS 19
3.1 Introduction 19
3.2 ECCM Improvement Factor 19
3.3 Functional Models 20
3.4 Response to Specific ECM 21
3.5 Deployment Models 22
3.6 Weapon System Efficacy 22
3.7 Game Theory Model 22
3.7.1 Case Study 1 23
3.7.2 Case Study 2 23
References
CHAPTER 4
MODELLING FOR ELECTRONIC CONFLICT:
THE BURNTHROUGH EQUATION MODEL 27
4.1 Background 27
4.2.1 Burnthrough Range 29
4.2.2 Escort Jammer 29
4.2.3 On-board Jammer 30
4.2.4 Repeater Jammer 30
4.3 Comments on the Burnthrough Range Equation 31
4.4 A Relook at the Burnthrough Range Relationship 32
4.4.1 Dynamic Radar Range Equation 33
4.5 Search Radar Modelling for ECCM 34
4.5.1 Functions of Radar Variables 34
4.5.2 Antenna Function 36
4.5.3 Jammer Function 37
4.6 Low Flying Target Detection Model 38
References
CHAPTER 5
LOW PROBABILITY OF INTERCEPT SEARCH
RADAR MODEL 39
5.1 Radar Detection & LPI 39
5.1.1 Detection 40
(ix)
5.1.2 Search Antenna 40

5.2 Intercept Receiver 41
5.3 Interface Between Radar & Interceptor 42
5.3.1 Effect of Receiver Aperture 42
5.3.2 Conflicts 42
5.3.3 LPI Modulation Scheme 43
5.3.4 Interceptor Receiver Bandwidth 43
5.4 LPIECCM 43
5.5 LPIECCM Radar Models 44
5.5.1 Radar vs Interceptor 44
References
CHAPTER 6
ECCM IN MULTIFUNCTION RADAR 49
6.1 Introduction 49
6.2 MFR ECCM: The RF Chain 50
6.3 MFR Antenna ECCM 50
6.4 MFR Transmitter ECCM 53
6.5 MFR Receiver ECCM 54
6.6 Resistance to Anti Radiation Missile 54
6.7 ECCM Evaluation 55
6.8 Comment 56
References
CHAPTER 7
RADAR ECCM MODEL THROUGH PETRI NETS 59
7.1 Conflicts 59
7.2 Petri Net 59
7.3 Extended & Timed Petri Nets 60
7.4 ECCM Modelling Through Petri Nets 60
7.5 Concluding Remarks 67
References
CHAPTER 8
GAME THEORY MODEL APPROACH TO RADAR ECCM
EVALUATION 69
8.1 Applications to ECCM Evaluation 70
8.2 Point-Defence Surface Radar 76
(x)
8.3 Conclusion 79
References
CHAPTER 9
KNOWLEDGE-BASED INFERENCE FOR RADAR ECCM 83
9.1 ECCM Decisions 83
9.2 Acquisition of ECCM Knowledge Base 86
9.3 Knowledge-based Classification 86
9.3.1 Handcrafting 88
9.4 Knowledge-based Simulation & Evaluation 90
References
CHAPTER 10
CONCLUSION 93
Index 95
PREFACE
Conflict situations are part of human life. What this

monograph wishes to deal with is about the area of conflict that
World Wars have brought about; that of the electronic encounters,
often fraught with far-reaching consequences. History has many
tales to tell, from the advent of World Wars, Egypt-Israeli conflicts,
Lebanon-Syria-Falkland encounters to the repeated encounters in
Iraq of the present times.
That the electronic battlefields armour is Technology is
well understood. However, its application, results and evaluation
are mired in controversy, deception and secrecyto become
ultimately justifiable and be synonymous with countrys honour at
the theatre of war.
It is the intention of this monograph to attempt an analysis
on one aspect of the electronic warfarethe counter
countermeasure of a radar sensor during an electronic conflict.
Finally, to the broad question of peace time simulation initiatives
so that we are all prepared to prevent the inevitable reaction under
duress.
Bangalore AK Subramanian
Date: Dec 6, 2003
ACKNOWLEDGEMENT
Grateful thanks and appreciation to Stephen L. Johnston,

USA, and with respectful remembrance of the Late Dr John Clarke,
UK, for the association, debate and research material and for
fostering an abiding interest in me on this subject of Radar.
AK Subramanian
CHAPTER 1
INTRODUCTION
1.1 IMPORTANCE OF ECM

The importance of Electronic Countermeasure (ECM) and
Electronic Counter Countermeasure (ECCM) in strategic and
tactical electronic warfare is not overstated. The ultimate survival
of any electronic system today is measured in terms of its
operational efficiency and survivability under adverse
electromagnetic environment. Defence radar and communication
networks are made vulnerable under such heavy and concentrated
attack. Apart from the conventional ECM attack delivered through
aerial and ground resources, a nuclear strike has its own
formidable component of Electromagnetic Pulse (EMP) which can
paralyse the entire chain of electronic sensors and associated
communication channels. A new dimension in the present era is
the added threat of the cyber warfare.
Electronic
Strategic Plan Countermeasures
Signal Intelligence
(SIGINT) Stand-Off Jamming (SOJ)
Self-Protection Jamming (SPJ)
Escort Jamming
Repeaters, Spoof & Decoy
COMMINT ELINT
-9
Tactical Electronic
Intelligence Counter
Countermeasures
Electronic Detection Sensors

Support &
Measures Communication
Figure1.1. Electr
Figure1.1. onic war
Electronic far
warfar e
fare
Modelling Radar-ECCM: A System Approach
1.2 BACKGROUND
Considering its effective usage since World War II, it is not
surprising that the art of Electronic Warfare has gone through a
sea change over the last century. The real assessment of ones
might as well as weakness in this area, is unfortunately proved
only during a major conflict. Modern examples have all been well
documented: those of the conflicts in Lebanon, Golan Heights,
Falklands, and the latest in the Gulf region. Whichever be the
theatre of war in future, it is amply clear that the initial strike
would be on the vital electronic defences of a country, be it the
missile complex, communication network, major air-defence
installation or the central and field headquarters.
The concept of a countrys boundary, guarded by its natural
geographical features, stays only as an historical fact these days.
Thanks to the modern technology, weapon development and
delivery systems, no nation needs to invade another territory to
start a war. Physical occupation may come later, but the initial
threat and thrust can be delivered from a safe distance. It is
presumed that the United States of America covertly supported
bombing of Iraqi Nuclear Facility by Israeli fighter planes. American
Warships positioned in the Mediterranean, guided the Israeli
fighters as well as provided facilities for the mid-air refuelling of its
aircraft. The Falkland War and the Gulf War had strong support of
satellite surveillance.
Further, even when delineated, the boundaries of a country
have a way of expanding due to economic and political reasons.
The USSR, when it existed, had its own ring of satellite countries
and the Warsaw Pact to protect its air space and the seas.
Americas reach was even wider and farther from its own country.
In India, due to the exploration of natural resources from the sea,
the naval complement of our defence effort extends our concern on
air-defence, even farther from the existing land boundaries. All in
all, the subject of air-defence rules the world as a modern dynamic
entity, precariously balanced and vulnerable to changes.
1.3 BASIC CONSIDERATIONS

Radar and communication systems have gone through a
sea change in the last three decades. Some of the old definitions
applied to the ECM-ECCM effectiveness have proved to be
inadequate.
2
Introduction
1.3.1 Jamming Interference

Jamming interference is mostly added in a nonlinear
fashion to the signal input of a radar sensor. Broadly, it can be
classified as one of the continuous noise, random pulse noise, or
synchronized pulse noise. The effects of jamming on radar are seen
in several ways. There is reduction in the Probability of Detection
Pd , increase in the Probability of False Alarm Pfa , poor accuracy in
tracking a target, poor resolution among targets, and a general
receiver saturation leading to non-linear processing. All of these
contribute to the degradation of the radar performance and
threaten its survivability.
Jamming effects are studied in terms of their spatial,
spectral, and temporal characteristics as they victimise different
blocks of the radar system. During jamming, a typical search radar
system suffers a loss in Pd for a given Pfa or it exhibits increased
false alarm rate while trying to maintain the same Pd. In tracking
radars, the jammers intentions would be to increase the
probability of track interruption and or increase the track errors.
In guidance radars, the jammer would try to deny the guidance of
the weapon, both in terms of communication and the guidance
radar parameters.
Radars have now become multifunctional and adaptive
(Fig.1.2). In the operating environment of a multifunctional radar, the
study of the ECM effects becomes more complex. Such a radar
SEARCH TRACK GUIDANCE
MULTIFUNCTION
RADAR WITH
ADAPTIVE
CAPABILITIES
Figur
Figuree 1.2. Multifunctional radar capability
3
carries out a variety of tasks, from the initial detection of a target

to its final engagement by a guided weapon. A number of radar
parameters hence become vulnerable to the jamming threat than if
the radar had operated on a single function continuously. It is also
to be noticed that the multifunctional radar replaces the space
diversity in an air-defence complex, which would have been
available, if separate radars had been fielded to accomplish the
varied tasks of detection, tracking, and guidance.
Communication networks incorporate more and more
digital sophistication in modulation and have frequency, packet
switching facilities, error detection codes and secrecy.
Communication links relaying important data among sensors are
equally vulnerable to jamming effects. Among netted radars and in
radars relaying guidance commands, the survival of the
communication network is crucial to the effectiveness of the
air-defence complex. Some basic differences in the electronic
warfare (EW) between radar and communication network is
illustrated below (Fig. 1.3).
EW
RADAR COMMUNICATION
Transmitter and receiver Transmitter and receiver

usually co-located at different locations
Two-way range for One-way range for
transmission and reception transmission and reception
Transmitter coding Encryption for security
Easier to spoof More difficult
Figur
Figuree 1.3. Some basic dif fer
differ ences between radar and communication
ferences
network under electr onic war
electronic far
warfar e.
fare.
1.4 ECM-ECCM MATRIX DYNAMICS

The format of interaction between the countermeasure and
the counter-countermeasure is a variable by the very nature of a
conflict. To be effective, each one needs an element of surprise and
doses of deception. In practice, the turnaround time to field a new
ECM system is much shorter than that required for an appropriate
4
Introduction
ECCM response. What is more, new ECCM features cannot be

easily handled as add-on feature in a system already designed and
fielded. Therefore, the need to consider all the present and future
threats of ECM is a must before a solution is presented on the
required ECCM front. Building adequate adaptivity to the
electronic design is presently the best answer to the evolving
dynamics of ECM vs ECCM.
1.4.1 ECCM EfficacyThe Difficulty

The question How good is an ECCM against a given ECM
threat? or its converse has been engaging the attention of the
defence electronic community for many decades. In the World
War II era, the answer was mostly qualitative but it met their
requirements, as systems were essentially simple. Given the
complexities and diverse functions of present day defence
preparedness, a qualitative measure is no longer adequate.
All systems are designed on a cost-performance criterion. As such,
over-designing cannot be resorted to but must be constantly
challenged on cost-effective basis. Each ECCM function added to a
major system has to be weighed against performance vs cost, time
for introduction, compatibility with the existing functions of the
system and finally, its longevity in the ECM-ECCM conflict chain.
Further, each ECCM element has its own share of compromise on
the normal performance of the radar/communication equipment.
It is meant for survival purposes, and thus, should be judiciously
introduced. Some ECCM measures in tandem are not mutually
compatible. At this stage, the subject becomes difficult to track,
especially when one examines a case like the multifunctional radar
(MFR), which combines target search, identification, designation,
tracking, and finally guides the missiles towards it!
1.4.2 Survival
The survival of a military electronic complex is dependent
on the diverse elements in the system. The electronic sensor or
communication is one of the limbs, albeit an important one. In a
defence scenario, every element is finally tied down to a weapon
and ultimately, the total effectiveness is measured in the manner
in which this system either performs/survives or perishes. Such a
terminal evaluation, as mentioned earlier, is unfortunately
possible, only upon major encounters.
Even such evaluations are not easily tractable, given the
inevitable shroud of secrecy in the EW matters. This is more so, in
5
the ECM arena, which has assumed more glamorous and

mysterious air. Our country is no exception to this syndrome.
1.4.3 Judging Efficacy of an ECCM Element

Under the circumstances, how does one judge the efficacy
of an ECCM element? The august body of IEEE, USA, went into
the matter in the seventies. The IEEE-AES panel headed by
Stephen Johnston came up with a definition of ECCM improvement
Factor (EIF). Definition-wise, it is similar to the MTI improvement
factor in a radar receiver but is not that easily measurable, except
through results. For more complex systems connected with a
number of ECCM elements, the combined effectiveness is
measured as
(1/EIF) of a system = (1/f1)+(1/f2)+(1/f3)+ ...
where, f ns are the EIF of individual elements. For defence related
areas, in the final reckoning, EIF calculation should take into
account all the diverse elements of a system. The terminal element
in a defence system is the weapon; its platform, deployment,
guidance and delivery. Some of these will be totally non-electronic
but still needs to be factored into the final evaluation of the
systems survivability. With all its defects, EIF is the only defined
measure, as of this moment, to determine the efficacy of an ECCM
system.
1.5 ASSESSMENT OF THE STATE-OF-THE-ART

The development of ECCM techniques and their deployment
strategies have, by force, been driven by the Electronic
Countermeasure (ECM). The classical model of development, based
on thrust and counter-thrust, is no longer valid in the present
context, as the time needed for a new ECCM system development is
just not available during a developing conflict. This immediately
brings in the necessity of a continuous, full-time analysis of the
emerging technologies in the electronic warfare field. All over the
world, analytical work carried out in the electronic warfare is not
easily shared for the obvious reasons of compromise with the
national security. The solutions are not applicable globally either.
From a deeper examination of the subject of modern ECCM, devoted
to radar and communication area, the following pointers emerge:
(a) Sophisticated signal processing methods, especially on
multidimensional spectral estimation techniques
(b) Study and analysis of non-gaussian signal interference and the
applicable mitigation techniques
6
Introduction
(c) Artificial intelligence/knowledge-based inference mechanism for

assistance in real-time ECCM decisions
(d) Neural network-based simulation for creating a learning process
in the ECCM decision tree
(e) Prediction and ECCM efficacy evaluation techniques for validating
the simulation process
(f) Use of VLSI techniques in integrating diverse high-speed decision
logic, as well as providing specific ASIC solutions to equipment
needing secure personality modules. This also includes the
development of micro-electromechanical sensors (MEMs)
(g) Development of photonics and other multi-sensor systems for
building survivability in the networking of information
(h) Safeguards needed in the cyber world, when one builds computer-
based decision systems in the defence field.
7
CHAPTER 2
MODERN RADAR DESIGN & ITS

DEPLOYMENT ENVIRONMENT
2.1 INTRODUCTION
Modern radar system design looks into two major phases,
the benign as well as the hostile environment. To the former, falls
the usual classification of clutter, terrain, screening, radar
horizon, mode of propagation and interference, EMC, etc. The
resultant modification to the radar performance has been studied
and well documented over the last century of radar development
in the world.
However, a radar system under Electronic Countermeasure
(ECM) and other means of hostile electronic interference operate
altogether in a different dimension. Firstly, the situation is dynamic.
The sources and the nature of interference are many. The solutions
for survival are also varied, depending on the effectiveness one seeks
to achieve. Secondly, there is no clear understanding of what was
achieved under what circumstances, and how long such a
performance can be maintained by the radar system designer. Much
of the problem in this field is due to the inevitable secrecy that
shrouds the EW world.
It then becomes essential to examine the present and

ever evolving EW technology and techniques and compare the
performance obtained with the current radar design concepts. The
limitations of some of the system concepts and in their physical
implementation are to be understood when seeking better solutions
against EW threats. Apart from all other functional and operational
requirements, the radar design needs augmentation in robustness
of operation under hostile environment.
2.2 RADAR PERFORMANCE

As stated, the interference to radar operation under benign
environment is understood and well documented. Analysis and
solutions are invariably available for operational conditions under
Ground and weather clutter
Propagation effects, angles, and anomalous conditions
Multi-path, lobing, screening, and radar horizon effects
Most of these conditions are prominently featured when a

ground radar operates against low-flying targets. Similar situation
holds well when an airborne radar is functioning on a look-down
mode. Shipborne systems operate under various sea-states and
have solutions when they experience severe forward scatter from
the sea surface and for encountering the sea clutter. Anomalous
propagation effects are very much prevalent over the Indian
waters.
2.2.1 Two Representative Radar Models
2.2.1.1 Low flying detection model

The strength of the radar signal received by the radar
system in consideration of the earths influence is written in the
form
2/
5 min = . ( ) (2.1)
( 4)
where, F ( ) is the pattern propagation factor. This, for small

angles of elevation
= 4 (Height of the Transmitter) (Height of the
Receiver)/Wavelength = 4(Ht Hr)/.
The modified radar range equation under this condition is
Pt G t 4 H t H r 4H t H r
R max = 8 = 8 Ro (2.2)
S min Const .
where, Ro is the maximum range of the radar in free space

condition. The eighth root relationship in the modified radar range
equation is to be noted.
10
Modern Radar Design & its Deployment Environment
2.2.1.2 Land clutter model for radar design
Clutter reflectivity = o (Fc )" and o = R Sin ()

= K for R Rh
= K (Rh/R)4 for R > Rh
where, R = Radar range, Rh is the radar horizon and Fc is the two-

way propagation factor for radar clutter path geometry.
Under the hostile environment, there is a severe restriction
imposed on the performance of any given radar, unless a suite of
ECCM features have been built into its design. For example, the
graph below (Fig. 2.1), shows how the range performance can be
almost negligible against a self-screening jammer.
BURNTHROUGH RANGE
8
DATA 1
DATA 2
DATA 3
7 S Vs RANGE
5
S/J
S/J Vs RANGE
4
3
J Vs RANGE
0
0.6 0.8 1
RANGE
Figure 2.1. Behaviour of S/J ratio as a function of range
While viewing this graph, it is to be remembered that the

jammer signal varies as R2 and the radar signal as R4.
A host of relationships have been developed to describe the
radar performance under ECM1. Modelling of the environment has
11
been suggested whenever analytical expressions could be derived

to represent the hostile interference. Processing decisions through
application of Game Theory have been suggested to take into
account the nature of the electronic conflict. A modified radar
range equation, in a functional form to take care of the ECCM
decisions, has also been proposed.
In general, the nature of countermoves in the radar against
ECM would be to have
(a) Reduction in the time period of signal emission
(b) Complex radar signature
(c) Highly directional beam with non-uniform scan/random
search patterns of the antenna
(d) Agility in the antenna beam as well as in the frequency of
radiation of the transmitter
(e) Sophisticated signal processing using digitally adaptive
solutions
(f) Parameter management of the radar in peace and war
situations
(g) Space and frequency diversity in the radar deployment.
These solutions are of course bought under a price, both in
terms of the cost and in the normal performance of the radar. The
system complexity increases with added pressure on the
engineering reliability. Some of the countermoves built in the
radar are short lived in terms of time element as it is relatively
difficult to configure to a new threat as fast as it is feasible with
an ECM equipment. It is indeed difficult to visualise a radar with
a new ECCM feature in the time it takes to fit a new or modified
ECM pod in an airborne mission.
Some of the complex and sophisticated processing built as
ECCM are quickly reduced in effectiveness under ECM. For
example, the linearity of processing in a coherent radar receiver is
severely impaired due to limiting or the saturation effects imposed
at the receiver front-end by a jammer. The linear chirp [of a pulse
compression design in a low probability intercept (LPI) radar] is
often degraded by the use of STC, or due to the blanking of the
portion of the expanded pulse during jamming.
Another area in which a clear understanding is yet to
emerge is in the evaluation of a given ECCM against its
counterpart ECM. Considerable interest has been shown in the
recent past1, but there is inevitably an element of secrecy and
deliberate ambiguity on this aspect. Nor is there an accepted
12
method of assessing radars ECCM performance against a host of

simultaneously present ECMs. Some of these issues have been
examined by several authors 1.
2.3 COMMAND, CONTROL, COMMUNICATION &

INTELLIGENCE
When examining the efficacy and survivability under EW,
apart from the single radar detection device, the ECCM scenario
extends to the air-defence complex as a whole. In a typical modern
air-defence complex, radars of different types and functionality are
usually netted and supplemented with other types of sensors.
Such a scheme provides space and spectrum diversity and thus
renders itself less vulnerable. Consequent to this arrangement is
the inevitable necessity to collect, collate, and analyse critical data
for the final decision mechanism. The timely and proper
information flow to a command and control centre acts as a
lifeline during a conflict situation. Thus, communication networks
too become critical in judging the robustness of operation under
the EW threat.
Accordingly, a concentrated effort is called for in the
integration of the following knowledge elements: (a) expertise
developed under multisensor, real-time information flow coupled
with (b) real-time decision process.
This leads to the concept of electronic battlefield
management. The scenario is graphically illustrated below (Fig 2.2).
EW NETTED DATA
THREAT SENSORS FUSION
NETTED
THREAT
SENSORS
PROCESSING
NETTED
SENSORS
BATTLE
COUNTER MANAGEMENT
ACTION
Figure 2.2. Electronic battlefield management
13
The concept combines a number of functions as follows:

(a) Intelligent fusion of various sensors deployed in a complex
(b) Decoding of opponents intention through appropriate,
knowledge-based inference machines
(c) Communication to the various echelons regarding the micro
and macro decisions
As such, there is no dispute about the speed of execution

of the above procedure (a computer-based decision will always be
faster than that of a human being who needs more time to
assimilate new and emerging facts). However, the credibility of the
automatic decision is dependent on the database and knowledge
accumulated and updated over a period. This process includes
several layers of preparation and updating cycles:
_ Documenting the ever proliferating domain of threats, their
electronic signatures, decoys, and deception methods
_ Nature and contour of ones geographic borders and its past
history of conflict play an important role in the decision
mechanism. To cite an example, the vast ocean fronts
covering India from west to east and the economical resources
built on these, render themselves vulnerable during major
conflicts
Sophisticated Electronic Support Measure (ESM) systems
with fast and accurate acquisition and monitoring
capabilities, programmable functions, vast memory capacity
and replay facility for simulation/training roles. The role of
ESMs are pretty well established in the ECM arena; but
similar application should be extended in the deployment of
the ECCM response to counter the dynamic nature of the
electronic threat.
2.4 ECCM MODELLING

It is therefore, necessary to examine the issues of
survivability, with models appropriate to the present time. Few
distinct steps in the modelling process have been suggested.
2.4.1 Radar Systems

(a) Models according to functions (such as track, guidance, and
multifunction types)
(b) According to deployment (whether singly operated or in a
network environment, monostatic or bistatic mode).
14
(c) According to platforms (land-based, shipborne or airborne)

(d) According to special techniques used in the hostile/victim
systems [adaptive/electronic scanning, coded transmitters,
customised signal processing, stealth technology, anti-
radiation missiles (ARMS), decoys, etc.].
2.4.2 Communication Systems

(a) Models according to modulation types employed
(b) According to code and degree of encryption implemented
(c) Error detection and error correction schemes employed, and
(d) Deployment, networking methods.
2.5 MATHEMATICAL ANALYSIS

After describing the various models that can be conceived
for the ECCM process, it is necessary to examine the method of
analysis. The following mathematical procedures amenable to
computer oriented numerical techniques have been employed:
_ Parametric method of analysis, after identifying the
important elements of the system for ECCM operation
_ MIN-MAX theorem and application to the ECM-ECCM
matrix
_ Application of Game Theory to ECCM evaluation
_ Application of probability functions to ECCM parameters and
in their evaluations
Invoking the figure of merit in performance of some of the
well-known ECCM subsystems.
2.6 SIMULATION
Digital computer simulation combined with hardwarein-
the-loop system, is the most effective way of combining the
analytical and the practical elements of an EW scenario. In such a
hybrid combination, the solution of a given problem can be found
on a real-time basis. However, there are a number of inexact
parameters in the process.
The immunity of the system under different jamming
threats are estimated on a statistical basis. These give an
expectation value on the performance of the system under
evaluation. The success of such an analysis depends, to a great
extent, on the capability of modelling the ECM and the victim
systems through the following steps.
15
(a) The number of independent parameters of the radar and the

probability that these parameters have been subjected to
enemys ESM analysis
(b) The probability that the radar parameters will be subjected to
jamming by enemy, after the necessary reconnaissance
(c) The probability with which the victim radar parameters are
jammed.
The above three steps lead to the overall probability
of the radar being jammed. The analytical determination of the
probability with which the victim system can function effectively
is fraught with considerable difficulty. However, if the radar is
assisted by its own dedicated ESM system (in the spatial, spectral,
and temporal coordinates), it becomes feasible to programme the
radar ECCM response in a more effective and analytical manner.
2.7 THE EW FLYWHEEL

The logical step leads one to have a knowledge-based, ESM
backed, ECCM response system which can be used in the
hybrid simulation consisting of digital computer coupled to
hardware-in-the-loop test bed.
The concept here is to involve an ESM system for ECCM
decision making, as much as such a system is exploited in the
deployment of the ECM function by the enemy forces. A dynamic
cause vs effect is tracked between the systems fielded by both the
forces.
The EW is a continuous game of winning over the
opponents use of electronics. Conventionally, the ECM users
deploy Elint surveillance to detect and map the opponents use of
their electronic gears. Typically, the ESM has all the sophisticated
system of decoding the electronic signatures of the enemys
transmission. The direction finding equipment supplements this
data with the angle of arrival of the intercepted signal. Such a
decoded information, rendered on a real time basis, is then used
to select the appropriate ECM for the jammer (Fig. 2.3).
It can then be convincingly argued that the Radar
Designers should use a similar decoding mechanism in order that
their ECCM response is appropriate. As further denoted in the
figure, a friendly ESM operated at the radar end would help in
decoding the ECM signal and thus help in the proper ECCM
16
DECISION
DECODING
ENEMY
ECM
ENEMY
ESM FRIENDLY
ESM
DECISION
RADAR
DECODING AND
ECCM
Figure 2.3. Selecting appropriate ECM for the jammer
selection. Such a scheme also has the beneficial feature of having

a dynamic response which is a must in all the EW encounters.
This integrated approach to radar ECCM decision making is
required both at the time of design and during deployment, where
geographical effects can also contribute to the vulnerability analysis.
REFERENCE
1. Johnston, S.L. Radar electronic counter countermeasures.
Artech House, 1979.
17
CHAPTER 3
ECCM EVALUATIONSOME MODELS
3.1 INTRODUCTION
The concept of ECCM evaluation in a radar system was
initially proposed by Johnston1, who introduced the term ECCM
Improvement Factor (EIF). This measures the efficacy of a particular
use of ECCM (Electronic Counter Countermeasures) in a radar
system against an encountered ECM (Electronic Countermeasures).
It is a factor to be used in the evaluation, in a manner analogous to
the concept of improvement factor in an MTI processor. Although
the EIF has been introduced about three decades back, it has not
found an all-round application so far in the literatures on radar
ECCM. This is the only accepted evaluation method presently
available which is professionally recognised as a standard by the
IEEE body since 1977. Some generalised models for radar ECCM
evaluation are proposed here.
3.2 ECCM IMPROVEMENT FACTOR

By definition, EIF is the ratio of the
ECM signal required to produce a given output at the radar (with ECCM)
ECM signal required to produce the same radar output (without ECCM).
Thus, it helps in quantifying the ECCM efficacy in a system-
oriented evaluation. The Radars ECCM effect can be conveniently
viewed under the following generalised grouping:
Functional (sensor-wise)
Response to specific ECM
Deployment type (field/environment)
Total weapon system efficacy
In expanding the above models with appropriate examples,
the ECCM implementation and its evaluation can be examined
As pertaining to a search, track or a weapon guidance radar, or
According to the ECM-ECCM matrix well defined in all
literatures, or
In relation to the vulnerability experienced in a deployment

pattern, or
By determining the effectiveness and survivability as a total
weapon system against an ECM attack (such as a missile-
site radar complex with its sensors, weapons, and inter/
intra communication equipment).
The last category is what the EIF specifies as a total figure of
merit. However, the diverse subsystems and their operations (from
sensor to weapons) make it a difficult task to evaluate the result.
The nature and spread of technologies involved in such a system
makes this evaluation a complex task.
3.3 FUNCTIONAL MODELS

When broken into functional models, it becomes easier to
relate the ECM-ECCM matrix more closely to the system under
consideration.
A search radar has the primary role of searching,
identifying, and designating targets of interest. Noise (in the benign
environment) and jamming (in times of hostilities) are the two
factors it has to contend with. Figure of merit of such a radar is
easily tractable and is presently evaluated according to the
established norms of Probability of Detection Pd and Probability of
False Alarm Pf a. Signal-to-Noise Power Ratio S/N and Jammer-to-
Signal Power Ratio J/S are directly related to the above factors. A
number of simulation methods are available2,3 to test the radar
receiver system under different threats. Thus, a fair assessment of
the receiver system performance and its deterioration against a
given interference can be carried out.
However, when one goes from the receiver subsystem of the
radar to the other building blocks like the antenna and transmitter,
the evaluation against the jammer/s becomes somewhat
intractable. The performance of antennas on sidelobe jamming is a
function depending on a number of factors; not all of these can be
easily simulated in a laboratory. Performance of a transmitter
against jamming is again effectively assessed only through the
attendant antenna and the receiver systems.
So, in assessing the ECCM features of a radar system
devoted to a particular function, the model, to a large extent, is
simplified. However, there is still a large gap in the methods and
assessment procedures to be carried out for the system as a whole.
Hence, intermediate and somewhat localised parameters of
measurements have been resorted to in quantifying the ECCM
20
ECCM EvaluationSome Models
evaluation; such as processing gain, cancellation ratio, J/S

measurement, burnthrough range, self-screening range, etc. These
have drawn heavy criticism in the literature2.
Almost similar comments can be given while discussing a
tracking radar model with ECCM. This type though attracts a
different ECM-ECCM matrix. The radar function is now devoted to
tracking the target/s with a stipulated accuracy, validate the tracks
to finally guide the operation to weapon attack on the assigned
target/s. Once again, a large number of studies have been
published in the literature2 to enumerate the ECM effects and the
ECCM response of this class of radar. The quantification done is
again related to the receiver, MTI parameters and comparison
between different tracking methods against a given jamming
technique. These are however, not easily extendable to the weapon
system to which all the sensors are finally integrated. The latter
represents an altogether different class of devices, though these
may have their own components of electronics.
The Multifunction Radar (MFR) is another functional
example, though relatively complex, which is amenable to
ECCM modelling. This case effectively represents as to how the
system evaluation of EIF becomes a difficult proposition.
Sophisticated methods are required to simulate and test all
aspects of the MFR and this is an expensive and difficult
proposition.
3.4 RESPONSE TO SPECIFIC ECM

The related ECCM response of radar against an encountered
ECM is widely covered in literature4. The success of a particular
ECCM technique against an assumed ECM threat, taken as one
pair at a time, is tangible for quick analysis. Efficacy evaluation
against multiple and simultaneous threats (such as the one from a
stand-off jammer in collusion with an on-board jammer) is a matter
of different complexity. In this context, it is relevant to point out
that the EIF is supposedly analogous to the concept and
measurement of MTI improvement factor1 in radar systems.
Nevertheless, this concept has not progressed further. To cite
examples, the measurement of improvement factor in a radar
system transcends from transmitter, STALO to the receiver.
Parameters like spectrum spread through antenna scan, clutter
modelling, clutter illumination coming through sidelobes, etc.,
support the evaluation of the MTI in a radar system. Further, a
simple relationship exists to correlate the improvement factor of
21
individual constituents to the total system (1/I = 1/I1 + 1/I2 + 1/I3

..+1/In ). Evaluation like the sub-clutter visibility (SCV) relates the
improvement factor to the total system performance against clutter.
Thus, a system evaluation is made possible.
Taking this line of thinking to the area of EIF can meet with
some obstacles as a system ECCM combines a number of diverse
elements. Especially in the weapon systems, the electronics and the
armour/missile/mechanical elements co-exist and it is difficult to
determine a factor like system EIF in such cases. The most practical
thing in these circumstances is to examine the vulnerability of each
element from the EW viewpoint and take the weakest link to
represent the ultimate factor that determines the survivability of
the system.
3.5 DEPLOYMENT MODELS

The effectiveness or otherwise of ECCM in a weapon system
model depends also on its deployment pattern. One such case was
discussed in relation to the MFR model5. Loomis has detailed an
ECCM attack pattern dependent on the system deployment6.
Further, the efficacy of the ECCM model undergoes radical changes
depending on the platform on which the ECCM is deployed such as
ground, ship, or airborne radar 2.
3.6 WEAPON SYSTEM EFFICACY

There is no argument that the ultimate ECCM evaluation
will have to be judged through the total weapon system efficacy
against ECM. The magnitude and the diverse nature of this exercise
needs no elaboration as it includes the sensors, weapons,
communication and many decision channels. Modelling and
simulation play a big role in this regard, as it is cost-effective and
practical than the actual weapon system testing under different
conditions. The veracity of the model however, needs a large amount
of corroborative work. Experience built over a long period and final
validation with limited weapon trials are effective solutions to this
aspect. It is learnt from literature that the Patriot missile defence
system could be validated by Raytheon with very limited trials. This
was made possible by the Company as it was backed up by
enormous system simulation studies and mock-runs on a
simulator, which included ECCM evaluation.
3.7 GAME THEORY MODEL

Major conflicts in the conventional and nuclear war are
studied through Game Theory models. Weapon systems and their
22
effectiveness have been traditionally studied through such

well-established methods.
3.7.1 Case Study 1

A game theory model, as a strategy applied to radar
detection under ECM attack, was analysed by Nilson7 in 1959. He
described various functions of a radar model and the related
strategies against a jammer. The model chosen employed a matched
filter in the radar receiver, which he proves to be an optimum game
theoretic filter.
Concentrating on the anti-jamming strategy as applied to
the transmitter, Nilson emphasises on the optimal use of the power
spectral distribution for the radar waveform design. The various
pay-off functions chosen related to the role of radar in detection,
tracking, and velocity estimation. It is pertinent to point out that
such results obtained through game theory in 1959 find
corroboration in the work by Gager 8. Though Gager had based his
arguments altogether on different and practical consideration, yet
he still arrived at the same conclusion.
3.7.2 Case Study 2

One more example can be cited from applied game theory to
illustrate its application to ECCM modelling against deception.
Repeater jammer, decoy, and chaff play their share of the game in
electronic warfare. Surprise and deception are their end goals. Thus
their roles appear to be not very different from those observed in
conventional warfare.
Axelrod9 has attempted to model policy-oriented theory of
deception and counter-deception in war games. This modelling is to
serve four important end uses:
(a) To make a realistic assessment between two extreme
viewpoints or results
(b) To evaluate the input information which is often incomplete
and at the control of the opposition
(c) To act as a logical and conceptual guide to future action, and
(d) To evaluate the total performance in a given setting.
Borrowing on the model discussed by Axelrod, one could
extend the argument of employing minmax game theory approach
to electronic deception. The model stipulates how to maximise the
pay-off when pure or mixed strategies are encountered. The
23
pay-off involved in this case is one of oppositions penetration of

other's defence and/or other's survivability against it. Based on the
arguments given, the results are shown below (Fig. 3.1 and 3.2).
Figure 3.1 shows the response between an attacker and a
defender, both initially following pure strategies (of attacking/faking
vs reacting/not reacting). The pay-off is indicated in terms of a value
to the attacker. Assuming this as a zero-sum game, the value for the
defender will be negative of this quantity. It can be seen that the
attacker gets the best value at B and worst pay-off at D.
B
No Reaction
C
Value
to Reaction
Attacker
Fake Real
Attack
Figure 3.1. Defence

Using minmax criterion of mixed strategies by both the
parties (involving a judicious probability mix between the real and
the fake actions), the minimum pay-off that can be obtained is
shown by a dotted curve. M indicates the maximum of the
B
No Reaction
C
Value
to Reaction
Attacker
A
Decoy/ Real
Attack
Deception Target
Figure 3.2. Weapon system
24
minimum pay-off that will be the outcome of this strategy for the
attacker.
Figure 3.2 similarly indicates the situation when the
previous example is translated to show the effect of a decoy or a
deception jammer against a weapon system. The weapon is
assumed to be guided by its electronic sensors. The reasoning is
similar to that of Fig. 3.1 provided by Axelrod9.
The simple model of Fig. 3.1 can be further refined when a
warning system is added to the environment. Such a system
enhances the capability of survival. Of course, the information
available to the warning device may be incomplete or partially
correct in a situation of conflict, but its very presence is considered
significant to the decision making. Depending on the discriminatory
value attached to the warning device, the pay-off to the defender is
improved as shown below (Fig 3.3).
The basis used by Axelrod for his model coincides with the
usual radar concepts of detection and false alarm probability, with
normal distribution. In this context, the case of the ESM receiver
is realised as a warning system to aid the ECCM response.
Borrowing on the above model, the response of a weapon system
with ESM warning can be viewed with similar results in Fig. 3.3.
Concept like Constant False Alarm Receiver (CFAR) can also be
utilised to describe the strategy for ensuring an assured pay-off
under deception. It is necessary to point out that the concepts of
game theory applied both by Nilson7 and Axelrod9 have been
accepted by radar engineers and as such these concepts form
attractive modelling possibilities.
B
No Reaction
C
Pay-off
to Reaction
Attacker
D
Reaction with Warning Device
Decoy Real
Figure 3.3. Pay-off in the presence of a warning device
25
REFERENCES
Artech House, 1979, Chapter 9.2, pp. 499-501.
Artech House, USA, 1979.
3. Military electronics/countermeasures, 1980, 3 & 12.
IEEE Trans. Aerospace Elect. Syst., 1978, 14(1), 109-17.
5. Subramanian, A.K. ECCM improvement factor considerations
in a multi-function radar. In Proceedings of the International
Radar Symposium, IRSI-83, India, 1983. pp. 587-92.
6. Loomis, R. Threats and techniques. Electronic Progress, 1975,
17(3), 17-25.
7. Nilson, N.J. An application of the theory of games to radar
reception problems. In IRE Convention Record, Pt.4, 1959.
pp. 130-40.
8. Gager, C.H. The impact of waveform bandwidth upon
tactical radar design. In Proceedings of Radar-82, UK.
pp. 278-82.
9. Axelrod, R. Coping with deception: Applied game theory. In
Proceedings of the Conference at the Institute for Advanced
Studies, Vienna, 1978. pp. 390-405.
26
CHAPTER 4
MODELLING FOR ELECTRONIC CONFLICT

THE BURNTHROUGH EQUATION MODEL
4.1 BACKGROUND
The burnthrough equation is a significant parameter in the
discussion of the Radar-ECM scenario. The radar detection
depends on the R 4 relationship whereas the jammer coverage
range is determined by the R 2 dependence. The difference is due
to the physical fact that the radar has a two-way transmission
path for its detection capability; it is easily perceived at the same
time that the jammer needs only to radiate its signal and reach its
victim on a one-way transmission path. If one considers a target
carrying its own on-board Self-Protection Jammer (SPJ), the
following equations would indicate the ranges where the jammer is
dominant over the radar and the juncture when it becomes
vulnerable to its skin detection by radar in spite of its EW suite.
The signal power S received by a radar receiver is given by
1 P t G t (1 , 1 ) t ( 1 , 1 ) Ar (1 , 1 )
S= 2 (4.1)
(4) (Rt )4 L t L r
and the interfering power from an active jammer J by
2
Ar (2 , 2)Br P j G j (2 , 2) 1
J = 1 R (4.2)
4 Lr Bj Lj j
where,
Radar Parameters
Pt = Peak power of radar transmitter
Gt (1,1) = Transmitting antenna gain in the direction of the
target (function of its azimuthal and elevation

angles)
Ar (1, 1) = Effective aperture of the receiving antenna
presented in the direction of the target
Ar (2, 2) = Effective receiver antenna aperture presented in
the direction of the jammer [If the target and the
jammer are same as in self-protection
jammers (SPJs), then 1 = 2 and 1 = 2]
Br = Radar receiver bandwidth
Lt = Transmitter losses, and
Lr = Receiver losses
Jammer Parameters
Pj = Jammer power
Gj (2, 2) = Jammer antenna gain in the direction of the
radar
Rj = Jammer range to the radar
Bj = Jammer bandwidth, and
Lj = Jammer losses
Target Parameters
t (1, 1) = Target cross-section presented in the direction
of the radar
Rt = Target range to the radar.
Hence, the ratio of the jammer to the target signal power is
4
J = 4 P j G j ( 2 , 2 ) B r Ar (2 , 2 ) (R t ) Lt
S t (1, 1) 2 (4.3)
P t G t (1, 1) B j Ar (1, 1) (R j ) L j
and the significant point to note here is that the ratio is

proportional to [Rt4/Rj2].
In the case of a target carrying a jammer aboard, it is clear
that Rt = Rj , the angular directions 1 = 2 = and 1 = 2 = , thus
making the antenna receive aperture Ar(1, 1) to be equal to
Ar(2, 2). Resetting the Eqn. 4.3 for this case, the J/S ratio is
simplified as Eqn. 4.4.
28
Modelling for Electronic Conflict:The Burnthrough Equation Model
4 P G ( , )B ( R )2 L
J j j r t
= P G ( , )B L (4.4)
S t ( , ) t t j j
4.2.1 The Burnthrough Range

The Burnthrough range is denoted by Rss. Signal from the
skin echo of the target varies as (range)-4 and the interfering
jammer signal as (range)-2. Hence, the radar has to compete
against the jammer interference, especially at the long range, in
finding the target. However, as the target flies towards the radar,
the target strength increases to 12 dB when the range is halved.
The jammer power, in contrast, gains only 6 dB for a similar
situation. It is clear then, that there will be a crossover range
R=Rss , where the J/S ratio becomes unity. Below this range, the
skin reflection from the target becomes a dominant factor for the
detection criterion of the radar. Above Rss, the jammer dominates.
Thus, the value of Rss of the radar (denoted as its burnthrough
range) is a parameter which both the parties of the electronic
conflict wish to capitalise upon. The ratio Rss/Rf which is
Radars maximum range under jamming environment
Radars maximum range for the same target under
benign environment
can also be used to characterise the performance of a radar under
the hostile interference. The jammer would try to reduce this
factor while the radar designer would like to maintain this ratio as
near to unity as possible, employing suitable ECCM. Rss can then
provide a measure of performance index for the chosen ECCM
operation.
The general relationship from Eqn. 4.3 can be rearranged
to quantify the interdependence of the radar range with the
various parameters of the radar-jammer scenario:
2
4
P Gt (1, 1) B j Ar (1, 1) (R j ) L j
Rt = t t (4.5)
4 P j G j (2, 2) B r Ar (2 , 2) L t (S/J )

4.2.2 Escort Jammer

For a stand-off or escort jammer, Rt Rj and if the jammer
is not falling on the same line connecting the radar and the target,
29
the receiver antenna apertures for the signals are not equal, i.e.,
Ar (1, 1) Ar (2, 2). The above equation remains
unaltered in this case.
For the on-axis stand-off or escort jammer, the following
holds good:
Rt Rj but Ar (1, 1) = Ar (2, 2). Equation 4.5 simplifies to
G ( , ) B j
Rt4 = t P t t 1 1
( ) 2
Rj Lj

4 P j G j (2 , 2 ) Br L (S/J ) (4.6)
t
For both the conditions, the Eqns. 4.5 and 4.6 can be used
to estimate the maximum range of detection (by skin reflection) of
the target by the radar. This provides the performance index
known as the mutual screening range.
4.2.3 On-board Jammer

For a on-board jammer carried by the target, the scenario
reduces the general Eqn. 4.4 to the case of the burnthrough
range. This is also known as the self screening range for the
jammer.
4.2.4 Repeater Jammer

With regard to repeater jamming, the conflict becomes
different. The jammer here re-radiates the received energy (with its
modification and gain) along the same or different direction. The
J/S ratio is then given by suitable substitution as
(i) Condition: Rt = Rj = R, Bj = Br and Ar (1,1) = Ar (2, 2) = A.
2 (1, 1) P G (1, 1) Lj
R = t 4 t t ( , ) (S/J ) (4.7)
P j G j 2 2 L t
(ii) Condition: R t = R j = R and B j = B r but A r ( 1 , 1 ) A r ( 2, 2 ).
(Jammers repeater action is seen through the sidelobes of
the radar beam)
G (1, 1) A (1, 1) L j
2 (1, 1) P
R = t 4 t t
(
r
, 2 ) Ar (2 , 2 ) Lt (S/J ) (4.8)
P j G
j 2

(iii) Condition: The jammer collects the radar power through its
antenna and re-radiates with its own gain and sends it back
30
along the same direction. Hence, Rt = Rj = R, Ar (1, 1) =

Ar (2, 2) = A and Bj = Br.
If Gj1 = Repeaters receive antenna gain, Gx = gain
(unsaturated) of the jammer, and Gj 2 = gain of the repeaters
transmit antenna, then the total system gain G at the jammer is
given by G = Aj 1 GxGj 2 = [2/(4)][Gj 1GxGj 2]. Hence, the power
returned by the repeater jammer to the radar is
P G ( , ) (G ) ( 2 ) (G ) (G )
J = t t 1 1 j1 x j2 1 A
2 L

4 R 2 L 4 L ( 4 R )
r
t j
(4.9)
and the signal due to the targets echo is
1 P G ( , ) ( , )A
S = t t 1 1 t 1 1
( 4 )2 4
(R ) L t L r

and thus, the ratio J/S is given by
2
J = (G j1)( )(Gx )(G j 2) 1 = System gain of jammer (G )
S (4) L j t L j t (4.10)

It is seen from the above equation that with suitable

overall gain G, the jammer can easily dominate the radars
performance. If the jammer is employed on-board a stealth
system, there is a further advantage due to the reduction of the
targets echo t, and consequently, a similar reduction in the
burnthrough range Rss.
4.3 COMMENTS ON THE BURNTHROUGH RANGE

EQUATION
Johnston1 has given the most severe criticism on the
concept and the utility of the burnthrough range. He lists several
factors against the dependence on this parameter and it is
necessary to summarise the major problems as he views them:
(a) Most of the treatments on the Rss factor do not include all the
essential variables of the ECCM-ECM matrix, or the same
variables from case to case; further, these parameters are not
defined to its entire capability in their respective usage
31
(b) It does not include the effect of ECCM

(c) Effect of various antenna coverages is not reflected in such a
simple representation of Rss
(d) Fluctuation in the target cross-section during its flight is not
accounted for
(e) Modern methods of signal formulation, signal detection
criteria, flexible antenna scans, etc., are not properly reflected
in this simple relationship
(f) There may be more than one unique range, definable for the
burnthrough concept
(g) Finally, the performance index associated with burnthrough
range applies only to cases of denial ECMs. It does not cover
aspects of deception, confusion jammers and threats like anti
radiation missile (ARM).
Further analysis on these aspects is carried out in the
ensuing paragraphs.
4.4 A RELOOK AT THE BURNTHROUGH RANGE

RELATIONSHIP
A more useful form of the Eqn. 4.5 is examined in this
section to try and provide answers to the comments recorded
above.
First of all, the equation is rewritten in terms of energy
terms as
A ( , ) R j
2
4 t (1 ,1 ) E t G t (1 ,1 ) Lj
r 1 1
(Rt ) =
4 Lt E j G j ( 2 , 2 ) Ar ( 2 , 2 ) S /J

(4.11)
It is preferable to substitute energy term instead of power
in the equation. This will enable the definition of different types of
radar waveforms, with diverse spectral density distributions.
Hence, it is most suitable to address the situations under radar
ECCM.
Secondly, the energy term Et = Pav to (where, Pav is the
average power of the radar and to the duration of antenna dwell
on the target), is a more meaningful term than Pt. For the
rectangular shaped radar pulse, Et = (Pt ) [Pulse length ()]. The
jammer output energy is readily represented as Ej , jammer power
32
per unit bandwidth (W/Hz). The S/N power ratio associated with
the radar detection can be similarly modified to Es/No , where, Es
is the signal energy and No is the noise energy (i.e., noise power per
unit bandwidth).
These changes have already been introduced in the
literature2 but have not been uniformly applied as standard
parameters in the radar range equations.
One important point to analyse in the burnthrough
equation is to find ways and means to keep it flexible and
comprehensive enough to cater for the requirements of the radar
ECCM community. Thus, efforts must be made to treat the
parameters of the range equation as functions of a number of
variables, which are valuable from the ECCM front. This will
ideally suit the dynamic behaviour of the ECM vs ECCM conflict.
The resultant effect on the system performance can then be
visualised through representative modelling and computation, all
of which can now be carried out by a modern desktop computer.
4.4.1 Dynamic Radar Range Equation

A dynamic radar range equation of the following form,
operating in a denial type of ECM environment is proposed as
[F1(E )] [F2(G )] F3 (t )] [F4 (Jammer)]

4 (Rt )4 = (4.12)
[F5 (L )] [F6 (D )]
F1(E) a function of radar transmitter characteristics [energy,

waveform, modulation, wavelength, PRF
characteristics, types of transmission (continuous
wave, interrupted continuous wave, pulse, burst
mode), tunability/instantaneous bandwidth of the
source, etc. ]
F2(G) a function of radar antenna behaviour (R, , ),
(aperture/2), beamwidth, coverage patterns, variable
beam coverage and scan rate, sidelobe levels, time of
dwell, flexibility available on the antenna operational
characteristics, etc.]
F3 (t) a function of (, , ) radar cross-sectional parameters of
the target as a function of wavelength and aspect
angles
F4 (Jammer) describes all the functional variables and capabilities
associated with the jammer [frequency range,
bandwidth and energy capabilities, sweep rates,
modulation schemes, and the known/assumed
33
capabilities of its ESM suite]

F5 (L) is the function associated with the losses in the radar
as well as the jammer, with particular reference to
their variations under different modes of operation in
a ECM vs. ECCM encounter [Loss in the transmission-
reception chain, antenna losses, scan losses,
processing losses in the signal/data processing
chains, etc.], and
F6 (D) includes all the functions for various detection criteria
employed in the radar from time to time, suiting the
ECCM response.
The utility of such a form of equation for the burnthrough
relationship is examined next, with the help of suitable radar
models. The applicability of computer analysis to handle a variety
of such functions and parameters becomes self-evident.
4.5 SEARCH RADAR MODELLING FOR ECCM

Utilising the new form of Eqn. 4.12 for typical search
radar, the following functions for the radar variables can be
defined for the ECCM role.
4.5.1 Functions of Radar Variables

F1(E) represents the energy function for the transmitter.
This is dependent on a number of variables which are
representative of the ECCM capabilities in this part of the radar
system. In general, it can have a number of sub-sets, single or in
combination of the following parameters:
F1(E) function of the radiated energy from the radar [This
could be of the form: continuous/switchable frequency agility
(used as an ECCM for the de-correlation of clutter and reduction
of target glint), control of average power Pav, pulse length , time
on target to, multiple PRF, and coded pulses for LPI, etc.,].
(a) In the most simplest form, Et = Pavto. Attempting a variational
parametric analysis3,
( Pav ) (t o )
[
E t (change in energy) = Pav t o ]
Pav
+
t o

If the transmitter tube has variable average power rating or

if the antenna dwell time varies, there is a change in the radiated
energy on the target. If both could be varied as independent
variables simultaneously, the energy management has two
34
degrees of freedom (DOF's) on the target to cater for a given

detection criterion. These options give an important ECCM feature
for improving the burnthrough range.
(b) If the search radar employs a mechanically scanned
antenna, (to) = 0 as the time on target cannot be varied.
Then the transmitter parameters are the only ones available
for the energy management:
(Pav ) = [(Pt - peak power) ( - pulse length) (f r - PRF)].
( Pt ) ( ) ( f r )
= Pav + +
Pt f r
Usually the variation available in Pt is restricted. is also

restricted in simpler radar designs, to cater for a given range
resolution; but in LPI radars using pulse compression, there is
scope of increasing the pulse length (and hence Pav ) and
simultaneously achieving the range resolution on the receive
mode. Variable PRF is yet another ECCM parameter for increasing
the average energy, even a mandatory function in airborne
systems to achieve proper detection of target against clutter.
(c) In radars employing phased arrays,

Et = (Pav) + (to), as both variations are independently
possible. The additional DOF, to, can be further expanded to
include some more variables which are useful for the ECCM
function:

( t o ) = b t s
s
where, b is the solid angle subtended by the antenna beam, s is

the solid angle of the volume search of the radar, and ts scan
time/frame time of the search radar. Thus the variations due to
beamwidth (possibly as a function of elevation angle), the angle of
scan, and frame time for search can all be simulated to assess the
end result against a given scenario. Phased arrays are expensive
to implement, and hence, a cost function can also be added to
these variables to obtain a cost-effective combination of the
parameters and their extent of variation. Such a step is definitely
indicated in the case of a multifunction phased array system4.
35
4.5.2 Antenna Function

The antenna function F2(G ) can be equated to
G (, ) g(, ), where, g is the ratio of the main-lobe to the
sidelobe, as a function of and . In this form, the gain of the
radar antenna can be described as a variable function wrt the
aspect angle of the antenna-target geometry, and as such can be
used to denote a variety of functions:

[G( ,)] = constant and [G( ,) ] = constant

are variables describing the gain change of the radar antenna in

the two principal planes. Normally, in the case of a mechanically

scanned antenna,

[G ( ,) ] = cons tant = 0, as the azimuthal
beamwidth is held constant at each elevation bracket. But

[G ( , )] = constant could be used to describe aerial function
with various elevation coverages. For example, the function
G (, ), where
G (, ) = G0, 0 1 and
G (, ) = G0 Cosec2(), 1 2 would represent a
cosecant-squared shaped beam coverage in the elevation plane for
all azimuthal angles.
Or in discrete form, the expression,
G (, ) = G1 (, ) + G2 (, ) + G3 (, ) + G4 (, ) (4.13)
[where each gain function Gx (, ) exists only for a specified

elevation bracket], would be handy for describing a series of
stacked pencil beams in elevation.
In phased array apertures, control is available in both the
planes and the antenna elements are each addressable and
controllable.
Another related antenna function is the sidelobe
control g (, ). This is necessarily to be viewed along with the
main antenna gain function G (, ), as the change in one
parameter affects the other 5,6. Hence, the variation of the product
[G (, ) g (, )], when optimised for antenna gain is
36
d g g
dG
{Gg } =G
dG
+ g = g G
dG
= 0. This form of the
expression takes into account the opposing nature of the main

beam and the sidelobe gain characteristics.
4.5.3 Jammer Function

Although these variables are not under the control of the
radar, the very presence of the jammer signal through the radar
receiver, is an indication of counter action. A knowledge of the
nature of emission from the jammer [Sj ( f ) the spectral density
function] can generate some optimum ECCM response at the
radar site7. Before the jammer operation, an important function of
the opposing force is to obtain information on the radar signature
through its electronic intelligence (ELINT) system. In this respect,
a performance criterion has been mentioned8 for the radar (in the
form of delay of such information to the enemy ELINT system), by
the term robustness. This is defined as the ratio of the range at
which the ELINT receiver can detect the radar signal Re, to the
maximum range of the radar Rmax for the target carrying an
on-board ELINT system:
Re
Rr
= ( 4 ) Rr GeGre Br (S / N )r
GrGr Be (S / N )e (4.14)
where, the subscripts r and e stand for the radar and Elint
systems, respectively. It can also be observed that all the radar
parameters are from the equation governing Rss (Eqn. 4.4). In fact,
Johnston has suggested that the range Rr should actually
correspond to the burnthrough range. It is then noticed that the
robustness factor is reduced by the jammer activity:
Re GeGre Br (S / N )r (S / J )r (ERP ) j Arj

= ( 4 ) Rr (4.15)
Rr GrGr Be (S / N )e (S / N )r K Ts t R j
2
F3 ( ) and F5 (L) are factors which are difficult to prescribe

accurately. However, their estimates (mean/rms/average values),
whenever available, can be used to improve the accuracy of the
range calculations.
F6(D )the S/N energy ratio criterion for detection varies
with different functions of the radar. Associated with this is the
type of signal processing performed and the related loss/gain
37
margins. It is also possible to add the effects of sensitivity time

control (STC), by a range-dependent weighing to this function.
4.6 LOW FLYING TARGET DETECTION MODEL

The search radar function is considerably affected when
viewed against low-flying targets with their own ECMs. This is due
to the multi-path effect of the antenna beam at low elevations as
well as that due to surrounding clutter. Taking the multi-path
effect as most dominant, it has been shown9 that the ratio
1
Range at low altitude 4 H t H a
=[ ] [Range in free space ]2
Range in free space
where, Ht and Ha represent the height of the target and antenna,

respectively. Although the relationship is an approximation, it
serves to show the decrease in the range performance, which can
be duly accounted for, while formulating the burnthrough range
under the low-flying condition of the jammer platform.
REFERENCES
Artech House, 1979. pp. 493-97.
2. Skolnik, M.I. Radar handbook. pp. 1-6.
3. Gutsche, S.L. et al. Radar ECCM. Artech Book House, 1979.
pp. 531-46.
4. Subramanian, A.K. ECCM improvement factor consider-
ations in a multifunction radar. Proceedings of international
radar symposium. IRSI-83, India, 1983. pp. 587-92.
5. Rudge, A.W. et al. Radar ECCM. Artech Book House, 1979.
pp. 195-204.
6. Hsiao, J.K. Phased array sidelobe level gain, beamwidth, and
error tolerance. Proceedings of radar conference, Paris, 1984.
pp. 304-08.
7. Nilsson, N.J. An application of the theory of games to radar
reception problems. IRE convention record, Part 4, 1959.
pp. 130-40.
8. Shenoy, R.P. Evolution of radarAn Indian point of view.
Proceedings of radar conference. Paris, 1984. pp. 5-9.
9. Maksimov, M.V. Radar anti-jamming techniques. Artech
Book House, 1979. pp. 17-18.
38
CHAPTER 5
LOW PROBABILITY OF INTERCEPT

SEARCH RADAR MODEL
A search radar system has the primary function of providing

the best possible detection at the best range, over an expected
volume of coverage. A large amount of electromagnetic (EM) energy
is required to accomplish this task. This factor becomes its
vulnerability as hostile receivers can sense the radiation, study its
characteristics and use the information for jamming and other
types of electronic countermeasures. From the point of view of the
ELINT (Electronic Intelligence) system, it is to be noted that the
fundamental physical law governing one-way reception is a factor
quite favourable to it.
The design features that optimise the radar's role, both for
its intended detection function as well as for providing protection
from the interception by hostile systems, are classified under the
Low Probability of Intercept (LPI) design.
An LPI radar, by its inherent design requirement, either tries
to reduce the time available for interception or makes the
recognition of its signature a low-probability event. The LPI design
may include the following features:
Agile and flexible antenna scan programmes
Low antenna sidelobes
Variable transmitter parameters
Frequency agility
Intra-pulse coding and coherent signal processing
Bistatic operation.
5.1 RADAR DETECTION & LPI
Requirement for radar detection and LPI consideration is
not always synonymous. Hence, the designer is required to evaluate
the influence of the radar parameters in meeting the dual
requirements and arrive at an optimum, if not the best, solution.
The guiding principle is to ensure that the radar's intended role is

not seriously compromised during hostile electromagnetic
interference.
5.1.1 Detection
The performance of a surveillance radar is specified in terms
of positively detecting a target or a group of targets in a given time-
frame by searching a given volume of space. For discussion on LPI
aspects and volumetric coverage, the radar range equation is
modified1 as
Power-aperture product of the radar =
4
( 4 ) R n (SNR )
PA = (5.1)
T N
where
P = Peak power of the radar transmitter
A = Effective antenna aperture of the transmit beam
R = Range of radar
= Solid angle of the radar's volumetric coverage
n = Noise power of radar receiver/Hz
SNR = Receiver Signal-Noise Power ratio required for a given
detection criteria
= Target cross-section
T = Time taken for a single scan of the antenna
= Duty ratio
N = Number of receivers assuming that the radar
transmits in a fan beam coverage; while on receive
mode, it has multiple, independent receive beams in
the elevation to define the elevation plane better
N = 1, if transmit and receive beams are the same.
5.1.2 Search Antenna

The effective transmit antenna gain is
G = (4 )/2 = (4)/( /T ) (5.2)

where
= Wavelength of transmission
= Beam dwell time
= Solid angle of search volume of the radar
T = Scan time
Low sidelobes are part of the search radar antenna design to
cater for the detection of low-flying targets against clutter; further,
40
Low Probability of Intercept Search Radar Model
this also helps in reducing the probability of jamming through the

sidelobes. The low sidelobe design thus contributes effectively to
both the aspects of performance and survival. These designs,
however, are accompanied by decreased antenna radiating
efficiency. To compensate for this, one has to go for longer scan
times to improve the signal strength, while keeping the same
physical aperture of the antenna. There is not much of a choice in
the wavelength of operation (typically it is around L and S bands for
long range search) as large microwave power is to be generated from
the transmitter.
5.2 INTERCEPT RECEIVER

It will be equally necessary to examine the electronic warfare
concept from the interceptor platform. It wishes to capture the
radar signal for finding its coordinates as well as derive intelligence
on the nature of its signal. The sensitivity of its receiver can be
expressed 2 as

Smin = n i SNR i (G i )(B 1i ) (2B 2i )1 (5.3)
where
ni = Interceptor receiver noise power/Hz,
SNRi= Threshold signal-to-noise ratio at the interceptor
receiver
Gi = Receive antenna gain
B1i = Pre-detection receiver bandwidth
B2i = Post-detection receiver bandwidth
= A factor varying between 0 (when two bandwidths are
comparable) and 0.5 (when B1i > B2i ).
To acquire the intended radar signal from a long range, the
interceptor has only a few choices in improving its receiver
sensitivity, increasing Gi , and reducing B1i and B2i. All of these do
have practical constraints. The gain of the antenna has to be
reckoned by the demand for wide frequency range coverage. The
wide-open receiver front-end bandwidth B1i is necessarily kept
higher to improve the probability of intercept (PoI). Hence, a finer
beam of the receiver antenna, coupled with reduced pre-detection
bandwidth will mean a fall in the PoI for the interceptor. The
reduction of B2i will lead to lower measurement bandwidth (target
characteristics and resolution). There is scope for compromise
here, without seriously impairing the main goal of achieving the
desired PoI at the best possible range.
41
5.3 INTERFACE BETWEEN RADAR & INTERCEPTOR

With the relation equations concerning the radar and the
interceptor, it is now possible to analyse the interaction between
these. The influence by the jammer on the radar could be viewed
through a connecting equation:
J = Sj N Ar i Br (5.4)
where
J = Jammer power coupled to the radar
Sj = Jammer radiated power/Hz (spectral radiance)
N = Number of independent receivers on radar
Ar i = Effective receive aperture of the radar antenna
presented in the jammer direction
Br = Bandwidth of radar receiver.
5.3.1 Effect of Receiver Aperture

Even a cursory look at the above equation reveals that the
best reduction of the coupled jammer power at the radar receiver
could be achieved by presenting the least antenna aperture towards
its direction. Since, the geometry between the jammer (usually
airborne) and the radar is variable in space as well as in time
domain, the solution is to cater for a sharper main beam, with low
sidelobes for the radar antenna design. This also aids in the radar
detection sensitivity but comes with penalty on the size of the
antenna; low sidelobe synthesis essentially means a highly tapered
antenna illumination, resulting in low aperture efficiency. This can
pose a major problem for a long range search radar when it
functions as a mobile system. With multiple receivers, this is likely
to be more demanding than implementing low sidelobe design on
the transmit aperture.
The limit set on the search antenna aperture vis--vis its
usual wavelength of operation (around L & S bands) has already
been commented.
5.3.2 Conflicts
In terms of conflict between detection and LPI functions for
the radar, it is seen that though the multichannel receivers (N) are
helpful for detection, these do couple jammer power to the same
degree as the number of channels provided. It is also observed that
the radar transmit antenna should have a better gain than that of
the receive mode, to maintain LPI characteristics, though in terms
of detection criterion, the reverse holds well.
42
5.3.3 LPI Modulation Scheme

Distinct modulation schemes, like digitally coded pulse
modulation, are employed in modern radar designs. This is helpful
in detection and implementation of LPI characteristics. The long-
coded pulses with pulse compression schemes, help in reducing the
transmitter peak power requirements. Depending on the
complexities involved in the pulse coding, the transmitter signature
is sought to be guarded against classification by an eavesdropping
receiver during its listening period.
5.3.4 Interceptor Receiver Bandwidth

This receiver bandwidth, as seen from Eqn. 5.3, can vary
from
Bi = Order of 2B1i or 2B2i (when the two bandwidths are
comparable)
= (2B1i B2i )0.5 when B1i > B2i
To increase its acquisition range, the interceptor has to
improve the receiver sensitivity (Eqn. 5.3). If the post-detection
bandwidth is reduced to achieve this, the interceptor loses its ability
to identify and classify the emitter but still retains its ability for
longer-range acquisition. This could be one of the practical
compromises, which the ECCM community has to be aware of and
consider in the radar design.
5.4 LPIECCM
The LPI radar, by its inherent design requirement, tries to
reduce the time available for interception and/or make the
recognition of its signature a low-probability event. In this sphere,
the elements of design are traceable to the same basic philosophy
that governs the ECCM capability of the candidate radar system.
The essential difference though, is in the stage of implementation
during operation. The process of interception is a silent function
that precedes the eventual ECM thrust on the radar. The LPI
concept is to attract less attention (interception) from a hostile
electronic snooper, and thus, its function can be considered as a
defensive measure. The role of ECCM is to offer a counter (offensive)
response in the presence of a threat. Both measures aid in the
survival of the radar system under the EW environment.
In this context, Johnston3 has defined a new set of radar
design measures under the name counter ESM (CESM). Referring
to Fig. 5.1, it can be seen that the solution caters to the dual needs
of a LPI radar system.
43
CESM
LOW PROBABILITY LOW PROBABILITY

OF OF
INTERCEPT IDENTIFICATION
(LPIn) (LPId)
Figure 5.1. Counter ESM
5.5 LPIECCM RADAR MODELS
5.5.1 Radar LI Interceptor

From Wiley 4, the ratio of the ELINT range Re to radar range
Rr is given by
1
Re
= Rr

4 1
( )( )
G te G e Le 2

Rr G t Gr L (5.5)
r

The subscript e and r stand for ELINT and radar,

respectively and other symbologies are according to the standard
radar range equation. The factor is the ratio of ELINT receiver
input power required for detection to the power needed at the radar
input for detection. The aim of the LPI radar is to reduce the value
of the ratio given in the Eqn. 5.5, which makes Re Rr .
If the interceptor was to be on the sidelobe of the victim
radar, then
1
2
Rr = G 4 metres (5.6)

for Gte = Ge = 1, = 1 m2, Le = Lr , Gt = Gr = G, and for the limiting case
of Re = Rr .
If the interceptor comes on the main beam of the victim
radar, then under the same conditions,
1
G 2 metres
Rr = 4 (5.7)

Given the practical constraints of achieving high gain for
radar antennas (along with the sidelobe reduction), greater
44
attention is to be paid in terms of increasing , so that parity in

detection is obtained by radar against detection range of the ELINT.
This is done through the control of transmitter waveform (coding,
agility, etc.) and attendant signal processing, as shown in the next
section.
The processing gain ratio is again provided by Wiley4 as
1 1
PGr Radar B 2 2
= B [BT0] (5.8)
PGe Elint rf
Letting B = Brf , the above ratio becomes equal to [BT0]0.5, the
square root of the time bandwidth product of the radar system,
with T0 indicating the coherent processing time. Hence The ratio
now simplifies to
(S/N )e PGr BN e 1
= [BT ]2 (5.9)
(S/N )r PGe BN r
0
Thus Rr from Eqn. 5.6 is proportional to [BT0]1/4 for providing

the quiet range in a LPI radar system. This can give a clear picture
of the importance of large time bandwidth systems fielded in such
designs.
Similar inference is derivable, if one were to think of
frequency agility/waveform coding as an ECCM feature in the same
candidate radar system. Gager5 shows that the relative radar
performance against denial ECM can be summarised to:
4 Bj
Radar R r Pt G t

Jammer : 2 = 4 S /N L
P G (5.10)
R r j j
j
From the above, it becomes clear that the radar's ECCM feature
should force the Jammer bandwidth to be spread out for proper
ECM thrust. This is not a welcome solution for an on-board jammer
with restricted power elements and space. He further shows the
relative merits of different waveforms (with high and low PRF) when
used in the radar transmitter for providing ECCM capability.
Thus, there is clearly a value-added performance index
measurable here. This is both in terms of the quiet range (obtained
through LPI) and the range against jammer (obtained through
application of ECCM). These form a common design solution for the
radar designers.
45
With the advent of phased array solutions, modern radars

are no longer constrained with fixed and non-adaptive antenna
performance. Adaptive antennas provide a welcome feature in the
LPI-ECCM design.
Fleskes6 considers the beam agility in a phased array and
compares the performance of uniform random search with that of
optimised search allocation. The basic search problem usually
encountered is that of finding an allocation of effort in space that
maximises the probability of detection of the target, subject to a
given constraint on the effort. In the example chosen by Fleskes,
the target models are real or the intended ones. However, in the
context of EW, the targets are more likely to be false, elusive and
deceptive. Hence, the search models will have to be cast with certain
stochastic probability figures.
Stone7 and Dobbie8 describe a few such models and relate
how an adaptive or a semi-adaptive solution is possible for arriving
at optimum search frame time. In both cases, the principal
detection function is of the type
b (z) = 1-e [-f (z)] (5.11)

where f (z) is the search effort function, usually based on the
cumulative probability of detection. This gives diminishing return
for increasing z (effort). The optimal search condition looks for the
appropriate search cells, which increase the posterior detection
probability. The time for search is thus optimally conserved, which
is a LPI function for reducing the time available in space to a
perceived ELINT snooping system.
With an agile, discrete, and adaptive search process
available through electronic scanning, and the possibility of
resolving false targets through ECCM, a feedback loop in the scan
and detection process becomes conceivable. This could be both self-
learning as well as leading to adaptive/optimal search control.
Thus, it becomes feasible to implement an adaptive search scheme
to reduce the mean time to resolve false targets (by association and
with feedback). At the heart of such a system, would be an agile and
adaptive scheme of scan, ECCM, and other resource management,
backed by a processor for online decision9. The improvement
possible in reducing the depth of penetration of a hostile target is
clearly brought out as a result of adaptive surveillance control. This
emerges again as an intended function of a LPI radar design.
The last model which appears appropriate for consideration
under LPI-ECCM scheme, is the intra-pulse polarisation agile radar
46
(IPAR) proposed by Cohen, et al10. The IPAR model, as the authors

describe, functions with many unique characteristics. In terms of
LPI, its compatibility with various RF carriers including noise, pulse
compression and polarisation agility can all be cited to its
advantage. The diverse carriers with Doppler-invariant pulse
compression ability, coupled with multi-mode function are good
attributes to radar design for LPI and ECCM functions.
The above models relate the combinational effect of LPI and
ECCM features in a radar design. Such models need to be evaluated
for both functions, preferably at the design stage itself. The
standard EIF definition would need modification from its presently
stated position. As the denial/delay of an ELINT intercept means a
better probability for the victim radar to survive, it is essential that
LPI-ECCM are correlated, and accounted for and evaluated as two
facets of the same entity.
REFERENCES
1. Skolnik, M.I. Introduction to radar systems. McGraw Hill,
1981.
2. Schleher, D.C. Introduction to electronic warfare. Artech
House, 1986.
3. Johnston, S.L. CESMA new category of radar ECCM. IEEE
Trans. Aerospace Elect. Syst., 1995, 31(2).
4. Wiley, R.G. Electronic intelligence. In The interception of
radar signals. Artech House, 1985.
5. Gager, C.H. The impact of waveform bandwidth upon tactical
radar design. Proceedings of radar, UK, 1982. pp. 278-82.
6. Fleskes, M. Proceedings of radar, UK, 1982. pp. 12-14.
7. Stone, L.D. Theory of optimal search. Academic Press, 1975.
8. Dobbie, J.M. SIAM Jour. Appl. Maths, 1975, 28(1), 72-86.
9. Billetter, D.R. Microwave Journal, 1986, 29(1), 147-57.
10. Cohen, M.N. et al. Proceeding of MSAT-1983. USA.
pp. 483-94.
47
CHAPTER 6
ECCM IN MULTIFUNCTION RADAR
6.1 INTRODUCTION
Multifunction radar (MFR) design has become necessary
and popular in the present-day environment of air defence for
providing multiple target engagement capability. Such a radar
often combines several functions like search, track, illumination,
guidance, etc., which were handled a few decades back by
separate radars, each having a unique role. Reliability and cost-
effectiveness are added constraints to its operation. The design of
a MFR system is complex as it has to cater for diverse needs;
providing an effective ECCM suite adds another difficult
dimension.
The MFR system is characterised by building many
adaptive designs into it. This starts from the antenna end and
continues through the RF/IF chain, signal processor, data
processor, computer, and finally the display.
In terms of the ECCM modelling for this type of radar,
the following parameters are analysed. These represent typical
modern MFR operating in several countries, deployed usually in
missile air defence complex:
Phased array antenna with programmable scan/coverage
and sidelobe cancelling system
Large RF dynamic range receiver with high image rejection
IF system
Digital signal processing with programmable filters
Self-contained illumination and guidance system for missiles
Digital computer system (central as well as distributed) to
cater for various programming controls and processing needs
of the radar
Modelling Radar-ECCM A System Approach
Encrypted data communication among various components

of the air defence system
- mobile system to centralised command
- radar to missiles and vice-versa
- diagnostics and house-keeping
- multifunction displays with dynamic operator interface.
6.2 MFR ECCM: THE RF CHAIN

In terms of ECCM incorporation, the MFR-RF chain has
many challenges compared to the subsequent digital signal/data
processors:
(a) The entire microwave chain is the first vulnerable hardware
to experience ECM and its effects. It is relatively difficult to
implement/supplement further ECCM features of this chain,
once the design is frozen and its hardware been implemented.
This is largely due to the analog process that is still involved.
Due to the presence of high frequency signals, it is difficult to
get the digital processing scaled up to these ranges. (Present
GaAs technology is being tried to do direct signal processing
at the RF stages but the commercial success has been limited
to regions of L and S bands so far).
(b) The subsequent signal/data processor, computer and the
video chain, in any case, incorporate many adaptive designs,
thanks to the prevalent, cost-effective culture of digital
technology. Further, what is more important from the ECCM
angle is that these designs are easily and less expensively
adaptable and can render themselves to real-time design
cycles for further improvement and modification to combat
newer ECM threats.
6.3 MFR ANTENNA ECCM

The MFR is characterised by an antenna system
with adaptive scan/beam coverage. Present-day hardware
implementation usually leans towards some form of electronic
scanning: frequency, phase or any number of hybrid types. The
model chosen for further discussion is the phase-phase scan
model. Though limited to sector scan coverage, the phased array
systems ensure the best possible response in high velocity, multi-
target environment, with requisite data rate.
Electronic beamwidth broadening, adaptive/multiple beam
generation, scan coverage tailoring, polarisation control, etc., are
the qualities readily attributed to the phased arrays. High scan
50
ECCM in Multifunction Radar
rates with various dwell rates and some form of adaptive

cancellation to annul interfering signals are the bright ECCM
features which can be derived from the arrays.
Let the present status on these arrays be examined, in
respect of their ECCM response. In basic terms, the phased array
antenna is not a simple or an isolated subgroup of radar.
Electronically speaking, it is no longer possible to separate the
transmitter from the antenna and vice versa. (This is all the more
true with the arrays implemented with the solid state T-R
modules).
The operation of a successful phased array depends on a
number of design variables, along with its evident software
support. The adaptability inherent in the system is a major
consideration for ECCM, but some price has to be paid. The most
serious ECCM limitation in the complex electronic array system
for the MFR is the extent of frequency range of operation.
Multifunction phased array is a poor choice when long-range
search and track functions are integrated into a single radar1.
Even, combining the properties of a medium-range search radar
with a short-range track function would result in a compromise on
the choice of frequency. When a single phased array is used in this
role, its operating frequency and bandwidth are relatively
restricted in that the search and track function should be
accomplished within a single frequency band (S or C band is
normally chosen as the frequency of operation). In the earlier
schemes where separate radars were employed, the frequency
band could extend from L band (for search) to X band or Ka band
(for track). Thus, their radar spectrum occupancy was rather
extended for ECM application.
Another limitation vulnerable to the ECM would be the act
of combining many antennas together in this type of array radar.
Antennas separately functioning for search/acquisition, track,
illumination, guidance, etc., incorporate an inherent set of space-
frequency diversity system among them. This gets compromised
when a compact multifunction system is envisaged and
substituted for. It is hence essential to weigh-in the built-in ECCM
features of a MFRs segmented aerial coverage against the
erstwhile-distributed deployment background.
The third ECCM consideration for the phased array system
would be on the instantaneous bandwidth and the frequency
range of operations within the chosen band. In the previous
51
mechanically scanned structures, these did not pose a serious

limitation. Adequate provision in the feed-reflector design
normally took care of the problem. In fact, the limitation in those
radar systems mostly came from the high power transmitter end.
In the present context of phased array, it is observable that limits
to the bandwidth of operations in a frequency band is on a
comparable basis from the transmitter as well as from the phased
array itself. Especially in phase-phase array configuration, the
phase errors and setting accuracy become limiting factors for
wideband operation.
It is then imperative that the question of Broad banding is
emphasised on the array antennas. What then is the position in
this field?
A brief survey 2-5 would indicate that the phased arrays
do have theoretical and practical problems peculiar to their type
of scanning. The antenna parameters like beamwidth, gain,
sidelobe, etc., are very much dependent on the scan angle and the
frequency extent of operation. Adaptive nulling and low sidelobe
performances are again variables, which ought to be considered
in relation to scanning programme and the requirement of
bandwidth. In tactical designs, the environment of operation also
plays an important role as the phased array contains a lot of
electronic components.
It is possible to use both amplitude and phase control in a
phased array antenna for synthesis in terms of beamwidth, scan
angle, low sidelobe design, adaptive nulling, etc. At present,
available technology has removed many hardware and system
problems by processing the signals at the microwave frequency for
control/adaptivity, without having to go through a frequency
conversion. Microwave-programmable attenuators, however, are
relatively difficult to realise with a high dynamic range and
sufficient bandwidth6. Hence, it is the digital, phase-only control/
weighting which is normally employed to obtain the adaptivity in
the phased array, once a given illumination function for the
antenna has been chosen. Thus, the microwave, digital phase
shifters become an important entity of the phased array system.
This is amply evidenced by the amount of work reported in open
literature on this subject.
With phase-only weights, excellent performance has been
reported on phased arrays for low sidelobe control, adaptive
nulling, and null steering; these are important from the ECCM
52
angle. How sensitive are these to the demand of frequency extent

of operations and instantaneous bandwidth? Certain comments in
relation to the phase control performance (the most important
element in this scheme) can be cited:
Quantisation error of digital phase shifter, its insertion
phase and phase setting accuracy over the frequency
range, etc., are all functions need to be carefully controlled.
Temperature range of operation is certainly an additional
factor for high power applications and for ferrite phasors.
Problems of low sidelobe design (for ensuring ECCM) in a
tactical phase-phase steerable array have been addressed7.
Amplitude and phase setting errors give rise to correlated
and un-correlated errors, and degrade the low sidelobe
performance8.
Signal bandwidth decreases with increasing scan angle. For
better performance, subarray steering has to be attempted.
Tight control on phase and amplitude tolerance has to be
exercised on element-on-element and subarray-to-subarray
basis. Such a requirement has been detailed9 in a typical
design, for achieving 50 dB rms sidelobes, using a 5-bit
phase shifter element. It is to be noted that better settable
accuracy and addressability comes with higher bits for the
phase shifter design.
In coherent sidelobe cancellors, with I&Q channels operating,
broadband performance needs complex multipliers with delay
lines10. Performance degradation in sidelobes, to the tune of
20 dB, is seen if either 6o of phase or 1 dB of amplitude
mismatch is present between the main and the auxiliary
channels. This factor currently limits the bandwidth of these
adaptive systems. Hence, instantaneous signal bandwidths
are at the moment restricted to around 30 MHz., if high
degree of sidelobe control over frequency is to be
maintained. It is also relevant to point out that there is a
need to implement one channel per jammer in this scheme,
unlike in the adaptive phased array antenna where
simultaneous frequency domain nulling is possible.
6.4 MFR TRANSMITTER ECCM

The transmitter stage of a MFR inherently has a large
number of electronic signatures. It incorporates waveform coding
(in one form or the other) variable pulse length, variable duty ratio
53
(due to selectable PRFs, selectable radar ranges, MTI function

etc.,) and variable power levels for energy management. Thus,
these cater for the diverse needs of the MFR system. Sinsky11,
Cook12 and Weil13 have addressed the importance of waveform
coding for radar design in general. Gager 14 has discussed the
impact of transmitter waveform bandwidth upon tactical phased
array radar design. He has compared the performance of tactical
radar utilising three waveforms with wide instantaneous
bandwidth, frequency agility, and conventional bandwidth. Their
resistance to jamming environment has been examined. It can be
seen that frequency agility, (in the search role for improving the
target detection and clutter de-correlation) and wide bandwidth
(for improving tracking and resolution) are both positive attributes
of ECCM to be considered in a MFR design. A wideband waveform
gives the greatest assurance of maximum range performance
against jamming. But hardware implementation in a phased array
for accommodating wide instantaneous bandwidth (5 to 10 per cent)
require time delay steering compared to phase control. This is
more costly to implement. Hence, a compromise is usually sought
by restricting the wideband role to limited functions15.
6.5 MFR RECEIVER ECCM

The receiver front end of a modern radar is a matured
product these days. Large dynamic range (often improved with
MSTC), better linearity, resistance to de-sensitisation, rapid
tunability, adequate signal bandwidth and controlled spectrum
signatures of STALO/COHO are attributes which make the
microwave coherent receiver quite accountable in jamming
environment. Direct RF signal processing is now being attempted
with GaAs technology so that there is no IF frequency conversion.
This improves the fidelity of reception and the reduction in the
noise of the receiver due to local oscillator and intermod products.
The rest of the receiver chain is digitally implemented with its
inherent adaptivity for providing the relevant ECCM functions.
6.6 RESISTANCE TO ANTI RADIATION MISSILE

All modern radars have to contend with Anti Radiation
Missile (ARM) and decoys16. An MFR with a phased array has
advantage through a very close control over its radiation pattern.
It employs adaptable beams with rapid scan rates and diverse
electronic signatures in its transmitter output. Further, there is
an increasing demand to put restrictions on the transmitter
54
power, waveform, and mode of operation to achieve low power

density in space. The confusing radiation scheme is sought to
deny an adequate signal for lock-on and track by the ARM seeker.
Such a complexity, adaptability, and flexibility in hardware,
coupled with a viable system software support, would naturally
make the job of the ECM or ARM that much difficult. But is it
necessarily so?
The proponents of ARM aver that they have major
capability against some of the recent radar waveforms and
techniques such as frequency agility and blinking transmissions
by incorporating stored threat files17. In the Pre-brief mode (PB),
the ARM is launched to a known location, and is said to be capable
of a search, classification and identification of all emitters that it
finds. It then locks on to the greatest threat or to a pre-set emitter
identity. The MFR, due to its diverse electronic signals as well as
its limited deployment in numbers (as a result of combining the
roles of many radar into one and the cost structure involved), can
become a prime and priced target to ARM (or to the ECM, in
general). The present picture is not very clear as to which has the
upper hand18. The recent Desert war19 had indicated a score on
the side of the RPVs/decoys/ARM's, though it is known that they
were deployed against a family of old vintage radars of Soviet and
other European origins. It is also to be noted that such integrated
evaluation of a weapon system against ECM is only possible
during a major international conflict, the results of which are
often not published fully or clouded with the inevitable secrecy.
However, in the phased array-sector scan air-defence
scenario, there is a possibility of quickly re-configuring the
deployment scheme. The transmitting sequence among many
radars deployed in the complex could be switched in a random
programme so that the direction of radiation is quickly changed in
space. With one of the radars taking the role of the transmitter at
any given point of time, the other systems could be kept in the
listening mode. This gives rise to a bi-static mode of operation-a
space-diversity provision and a valuable ECCM function to
counter ARMs threat.
6.7 ECCM EVALUATION

The role and operation of a phased array implemented MFR
has been examined so far in relation to ECCM. Due to the
multifunction nature, the design of the radar is complex but the
phased array solution has a number of inherent ECCM features.
55
There are also limitations and trade-offs in these schemes, both in

the electronics as well as during deployment. How then are we to
evaluate the ECCM efficacy or the EIF20 of the system?
The answer seems to be rather complex and not
necessarily amenable to easy standardisation or interpretation.
The following points, made out by Johnston21, are relevant:
He conceived EIF as a means of representing performance of
a specific ECCM against its related ECM threat.
EIF does not apply to all ECMs, it is a function of the ECM
used against an ECCM, and may be a function of J/S ratio
too
One must have a total weapon system engagement model to
be able to determine effects of ECCM system effectiveness.
6.8 COMMENT
With the diverse roles assigned to the MFR, a large ECM-
ECCM matrix 22 is present when dealing with the ECCM
assessment. Performance figures like EIF will undergo variations
under different modes of operation of the radar with the attendant
ECCM features. This will be typical in a MFR. Thus, the evaluation
of EIF with the entire weapon system performance against ECM
seems to be a formidable task in such a radar system. At the same
time, with the kind of prohibitive cost structure associated with
the phased array system, it is necessary to have practical and
engineering guidelines when one is faced with the ECCM
implementation in a MFR. Following are some of the
considerations to seek an appropriate answer:
In the case of a highly priced radar system, like the MFR
model discussed here, performance and reliability have normally
to be balanced against acquisition cost and ease of maintenance.
A method known as structured redundancy is usually resorted
to in keeping the above considerations in an engineering balance23-24.
This principle utilises optimum redundancy at modest cost to
achieve greater prospects of mission accomplishment.
One could draw a parallel from this for the present ECCM
consideration. A structured ECCM possibly? If the word can be
coined and methods implemented, it would imply that
(a) Advantage could be taken by concentrating on those
hardware of the MFR system which are used in more than
56
one radar mode. These are prime candidates for proofing

against the ECM, and what is better, the Lifecycle Cost (LC)
will be the optimum in this regard.
(b) Those items of hardware where growth potential is built-in,
provide the best possible ground to take care of surprise
ECM.
(c) A viable ESM receiver would assist the MFR to choose,
through its computer, the proper ECCM response from its
suite. In the case of reliability and maintainability, it will be
remembered that this function is effectively carried out by
the BITE function.
The above considerations, once again, can succeed in
showing the importance of the microwave chain in a MFR system,
when ECM hardening is examined. It should lead to cost-effective
and dynamic ECCM response. Structured ECCM along with
structured redundancy should go a long way in providing cost-
effective solutions for reliability, maintainability, and what is
more, better survivability.
REFERENCES
1. Skolnik, M.I. The phased array in military radar applications.
Military Technology, 83(3), 62-72.
2. Stark, L. Microwave theory of phased array antennasA
review. Proc. of IEEE, 1974, 12, 1661-1701.
3. Knittel, O. Phased array antennaAn overview. In Radar
Technology. Chapter 21, Artech House, 1978. pp. 289-301.
4. Forrest, J.R. Phased arrays. Current technology and future
prospects. Proceedings of the 11th European Microwave
Conference, 1981. pp. 81-90.
5. Mailloux, R.J. Phased array theory and technology. Proc. of
IEEE, 1982, 3, 246-91.
6. Fulvio, Ananasso. Null steering uses digital weighting. MSN,
1981, 7, 78-94.
7. Patton, W.T. Low-sidelobe antennas for tactical radars. IEEE
International Radar Conference, 1980. pp. 243-47.
8. Hsiao, J.K. On performance degradation of a low sidelobe
phased array due to correlated and un-correlated errors.
Proceedings of Radar-82. pp. 355-59.
57
9. Brookner, Eli. A review of array radar. Microwave Journal,

1981, 10, 25-60.
10. Grant, P.M. & Cowan, C.F.N. Adaptive antennas find military
and civilian applications. MSN, 1981, 9, 97-98, 100, 102-07.
11. Sinsky. Waveform selection and processing. In Radar
Technology, Chapter 7, Artech House, 1978. pp. 123-42.
12. Cook. Large time-bandwidth radar signals. In Radar
Technology, Chapter 8, Artech House, 1978. pp. 143-48.
13. Weil. Efficient spectrum control for pulsed radar
transmitters. In Radar Technology, Chapter 27, Artech House,
1978. pp. 379-83.
14. Gager, C.H. The impact of waveform bandwidth upon tactical
radar design. Proceedings of Radar82, U.K. pp. 278-82.
15. Winnberg, J.O. Search and target acquisition radar for
short-range air defence systemA new threat environment. A
new solution. Proceedings of Radar-82, pp. 120-24.
16. Boyle, Dan. Anti-radar missiles. Interavia, 82(11), 1194-95.
17. Lake, J.S. Insight. Defence Electronics, 1983, 5, 157.
18. MST/Israeli Interview. Military Electronics Countermeasures,
1983, 2, 94-102.
19. Desert stormthe electronic battle. Carlo kopp Australian
aviation, 1993, June/July/August.
Chapter 9.2, Artech House, 1979. pp. 499-501.
Artech House, 1979.
Chapter1.1, Artech House, 1979. pp. 11-19.
23. Airforce radar balances its costs. Electronics, 1976,
5, 40-41.
24. Electronically agile radar being developed for strategic
bombers. NATOs 15 Nations, April-May 1976, p. 100.
58
CHAPTER 7
RADAR ECCM MODEL THROUGH

PETRI NETS
The interactive response between the ECM-ECCM in a radar

environment can be compared, at the first instance, to a set of
concurrent and asynchronous activities. This could be modelled as
a parallel process with a certain degree of logic and procedure. In
real terms, the entire relationship is based on information, which is
often incomplete and in most cases deceptive. ECCM response is
thus based not wholly on deterministic concepts but modelling is
still attempted with stochastic considerations.
7.1 CONFLICTS
Major criterion in the ECCM response is the ability to resolve
conflict situations. A few examples can be readily cited: Radar with
wideband frequency agility is a compromise solution when its MTI
performance is considered. Wideband antenna can only be
optimised in relation to its narrow band counterparts superior
antenna gain, efficiency and sidelobe performance figures. Faster
scanning rates and beam agility have to compete with the goals set
for high anti-clutter rejection. Thus, any modelling for ECCM
should have the impact of conflicts shown in the analysis process.
Petri nets possess the inherent advantage of illustrating such
conflicts in their directed graphs.
7.2 PETRI NET

The concept of Petri net has been developed over the
years1-3, to graphically model event-driven simulation and its
related information flow. Conflicts requiring resolution, to any level
of abstraction, can be attempted through these Petri nets. Wherever
actions have to be inhibited, due to specified constraints, these
could be graphically added to the network. The content could be
software-intensive as well as modelled with hybrid simulation.
7.3 EXTENDED & TIMED PETRI NETS

Apart from the general formulation of Petri nets, research
has also been concentrated on the extensions required in Petri nets
to simulate real-time problems. Time Basic nets4, known as TB
Nets, deal with the situations where timing of the decision is
essential to the flow of the network function. These concepts help in
modifying the generalised nets to the kind of specification language
found desirable by the user of a particular application. This, in fact,
produces the welcome opportunity for the flexibility in ones
approach to modelling. Such a flexible network formation could be
profitably explored to represent the dynamics of ECCM modelling in
radar and its allied systems.
7.4 ECCM MODELLING THROUGH PETRI NETS

To keep the level of discussion within bound, let the
proposed modelling refer to a single radar system, with its attendant
ECCM, operating in a hostile ECM scenario. The following are some
of the ECM-ECCM Petri net primitives:
Item Input Output
Places can represent An event A result

A resource request A resource enabler
A cause An effect
A condition or constraint Meet or unable to meet
the constraint (conflict)
Figure 7.1 represents some of the user-specified Petri net

primitives employed in the ECCM model. From the ECM side, one
could define diverse platforms and the nature of their threats.
Typical airborne ECM platforms are discussed in this example:
Platform Platform
Self-screening Stand-off
Jammer Jammer
(SSJ) (SOJ)
Active Passive Active Passive
Figure 7.1. ECM Petri net primitives
60
Radar ECCM Model Through Petri Nets
On the victim radar side, the primitives can be used to

describe the various responses under its ECCM role:
Radar System Radar TX
(a) (b) Radar
RX Only TX/RX Normal Frequency

(Bistatic) (Monostatic) Adaptive Agility
Resources
Radar RX
(c)
Normal Frequency
Adaptive Agility
Resources
Figure 7.2. Radar ECCM Petri net primitives

Expanding from Figs. 7.1 and 7.2, the connectivity between
the radar and the ECM platform is shown in Fig. 7.3. The
representative radar can do search and track, and the options it
has under ECCM modes (P1, P2, P3, P5) are pictured:
Platform:
SSJ Platform
with a Wideband Jammer t3
t4
t2 Friendly
Enemy t5 ESM
ESM P4 P3
P2
t1
Inhibit
Frequency
P5 Agility
TOJ
P1
Adaptive Beam
Scan
& LPI
Figure 7.3. ECM and radar ECCM shown in Petri net connectivity
61
In the P1 mode, it can employ adaptive scan and LPI

schemes, which are the normal ECCM functions when starting the
search operation of the radar. P2 has the frequency agility mode
when the jammer applies narrow band or swept-frequency
jamming. To assist the radar, there is a friendly ESM system (P3) for
dynamic analysis of the jammer signature. When the situation
demands, the radar can go on to receive-only mode by switching off
its transmitter and track the target on track-on jammer (TOJ) mode.
This is depicted by P5.
The ECM platform has the logistic support of an ESM
receiver for its dynamic decision process. This is also depicted in
Fig. 7.3 by the mode of operation under P4.
In the above example, it is assumed that the ECM and
ECCM facilities are programmed to respond instantaneously, once
the decision is taken for a particular deployment. Thus, when the
transitions t1-t5 are fired, the response-counter response chain is
immediately enabled. An inhibitor arc is conveniently shown in the
network, which silences the radar transmission, whenever it is
listening to the active jammer (track-on jammer mode). It must be
remembered that these changes in the radar operational pattern
are also constantly monitored by the ESM assisting the ECM suite.
This gives rise to the dynamic ECM-ECCM matrix response.
Advancing further, threats from two aerial platforms can be
studied simultaneously. This can occur due to the presence of two
distinct sources as depicted in Fig. 7.4. With both threats
integrated, this also provides the example and the need for
modelling concurrent but asynchronous operations. Petri nets
provide a handy and graphic visual clue to such situations.
Place P1 represents a radar, functioning under the dual
threat of P2, a Self-Screening Jammer (SSJ) and P3, a Stand-off
Jammer (SOJ). While there is a concentrated jamming on the main
lobe by the SSJ, the radar function is also seriously hampered by
the sidelobe reception of the SOJs emission. The response has to
be necessarily dynamic, but the options available to the radar at
any given instant can be suitably shown through Petri net
representation.
The typical ECM response available with the radar is
indicated under places P4, P5 and P6. Some of the conflicts in the
decision mechanism could be observed. To effectively control the
consequences of sidelobe jamming, the radar goes into sidelobe
blanking (SLB) or canceller mode (SLC). It would then prefer a
62
RX ONLY
P2 P1 P3 P2 P2 P3
ADAPTIVE
PROCESSING
P4 P5 SLB/SIC P6 TOJ P7
FREQUENCY
AGILITY CONFLICT
INHIBIT
Figure 7.4. Radar ECCM Petri net representation of concurrent and

asynchronous action and conflicts.
transmitter operation without frequency agility, to optimise its

antenna performance. This would not be ideal for the radar, which
receives a wideband jammer on its main beam. If the operator
desires that the radar leaves both these options and switch on to a
listening mode by tracking the main beam jammer on its own
emission, this still leaves its sidelobe reception vulnerable. An
adaptive solution (place P7), in terms of antenna and signal
processing, may be a better method but at a cost and associated
complexity. This Petri net representation can be taken into further
levels of abstraction, when conflict situations need deeper analysis
into the resource requests and their timing. Resource tie-up is one
more area where Petri nets become handy again for illustration.
Hitherto, the concept of timing has not entered the
discussion. The ECM-ECCM interaction is time-driven, and the
response has to be visualised dynamically. Necessary time stamps
can be assigned either to the places or to the transitions in a
Petri net5 to represent real-time problems and their analysis. One
could assign time stamps variously to indicate processing
time intervals, delays in interconnections, concurrent and
asynchronous actions, resources tie-ups, timing and duration of
inhibitor arcs, etc. A sample of such a network for ECCM, with time
factor associated with the firing of the transitions, is illustrated in
Fig. 7.5.
63
AUXILIARY MISSILE
HOMING
T4 A4 A1
TARGET RX
HOMING
T ECM
T HEAD
R
A R
N A
S N
P S
O P
O A2 A3
N T3
D T1 N T2
E D
R E
COMMAND R HOMING TOJ
MODE MODE MODE
n1 < T1 < n 2 T3 < n5 n3 = T1 max
n3 < T2 < n4 n5 < T4 < n4 T3 min = T2 n1 < n2 < n3 < n5 < n4
Figure 7.5. Radar ECCM Petri net with time stamps
The figure describes the scenario of a guided, surface-to-air

missile during its flight towards a targeted aircraft. Let it be
assumed that the missile is command-guided up to roughly two-
thirds of its full range while its terminal phase is activated through
a homing radar. Both these functions are implemented at different
frequencies, a logical ECCM measure.
The victim aircraft has its normal ESM suite along with
capability for passive and active deception jamming. A further
attribute will be the release of decoys when situation warrants.
During the command guidance, it can be relatively assumed
that the aircrafts ECM does not overwhelm the missiles function,
as the target is still the farthest in the radar-missile-aircraft
geometry. Further, due to the transponder mode in command
guidance, the ECMs effect is not very strong at this stage. As the
missile nears the target, its homing head starts locking on to the
desired course for final flight manoeuvre. The ESM system on board
the aircraft will now suitably programme the jammer for retaliation.
The active jamming mode can take several forms. From a wideband
barrage jamming to deny detection at the search phase, it could go
on to denying track information in the range, angular, and velocity
coordinates. This would make the missile homing transmitter to
64
changeover to other modulation schemes with the attendant signal

processing in its receiver or switch-off to go on to track-on jammer
(TOJ) mode. The aircrafts radar warning receiver (RWR), alerted to
this dynamics, can react similarly and temporarily suspend the
jammer operation to try and break the lock of the missile in its
trajectory. This cycle can go on, but the missile needs to patch up to
the ground station at the appropriate moment as a last resort and
failsafe operation. To complete the mission under severe ECM, the
only option left now is to take recourse to the availability of large
resources present in the guidance radar on ground station. Now it
is to be examined as to how the above cycle of events could be
visualised through Petri nets, including the timing element that is
indeed a very crucial factor here.
Turning to Fig. 7.5, transitions A1 to A3 are seen marked
with time stamps T 1 to T 3, which rather indicate time duration of
different stages of the missile flight than any particular instant of
time. Reckoning the zero reference time as that instant when the
missile leaves its pad, all other timings and duration can be logically
determined. The duration T1 refers to the command guidance phase,
which commences at the minimum range R1 and lasts until R2.
These are the designed range limits for this guidance phase in the
normal circumstances; translated in time, the duration T1 can be
estimated to lie between n1 and n2; n1<T 1<n2. The time limits, n1 and
n2 also take into account the processing delays involved in the entire
system operation during this stage. Similarly, T 2 refers to the
duration of the homing phase, which lies between the limits n3 and
n4. The minimum value n3 obviously overlaps with the terminal
phase of the command guidance, as the homing head of the missile
needs to get a lock on to the target before commencing its guidance.
For this model, it is simply assumed that n3 equals T1 (max), i.e., the
instant at which the normal command phase ends.
It is to be noted that in the absence of ECM, the command
guidance phase duration is normally T1 and the homing phase
duration is controlled by T2.
However, as described in the earlier paragraph, the active
homing phase will not enjoy unchallenged reign. This phase is often
interrupted and renewed by the missile, whenever its lock on the
target gets broken due to active ECM jammer. When encountering
ECM, the missile switches off the homing transmitter (to avoid
detection). It then tries to home on to the target by listening to the
jammer and tracks through TOJ mode. Time stamp T3 refers to this
new stage. The lower limit of T3 overlaps with the homing phase as
65
the missile guidance weaves between homing and TOJ mode. The
first instance the guidance switches on to TOJ, T3 (min) = T2, which
is less than its maximum available duration n4. It is easily seen that
the upper time limit for both homing and TOJ modes will be n4 for
mission accomplishment. However, under severe ECM conditions,
a time cushion is necessary so that the missile still has enough
moments to patch up with the ground station to resume its flight on
fallback command guidance mode. Hence, under the ECM
influence, time stamps T2 and T3 have to work with an upper time
ceiling, say n5, which is less than n4. Transition A4 represents the
auxiliary phase, where the command guidance from ground is
invoked. The time stamps T4 for this junctions works between the
limits n 5 and n4, as can easily be verified.
A further comment that can be made is on the nature of
firing of the transitions. While transitions, A1 and A2 are necessarily
enabled and fired during T1 and T2, respectively, it is not always
mandatory for A3 or A4 to act. In the normal ECM-free environment,
transitions A3 or A4 does not have a part to play, but if these do act,
these need to be fired during time stamps T 3 and T4 allotted to them.
Such transitions, which are not necessarily forced to fire but are
anyway given the choice to act only during specified time intervals,
are known under weak-time semantic4.
The Petri net representation is advanced to another level of
abstraction (Fig. 7.6), to indicate major system-level response under
ECM-ECCM interaction. Transitions with suitable weights for firing
conditions are introduced along with inhibitor arcs and conflict
situations.
The radars ECCM responses under five categories are
depicted. In this example, an attempt has been made to show the
dynamic nature of the response, usual in these conflicts. If the ECM
platform is on active mode, the radar acts in the listening mode.
The signature analysis through its ESM serves to help the ECM
platform to switch to passive mode. This, in turn, forces the radar
to actively transmit again, but now being aware of the presence of
ECM, it tries to effectively use its response from the ECCM suite. A
further level of abstraction can consider the time factor, if the radar
is employed in a multifunction role, wherein multiple threats and
the related radar responses are to be analysed. It becomes
mandatory when common resources are to be managed. It is to be
pointed out again that the nature and the duration of resource tie-
ups are conveniently modelled in timed Petri nets with time
stamped inhibitor arcs.
66
FREQUENCY AGILITY ECM

P1 INHIBIT CONFLICT
F1 F2 P2
ESM
RADAR NORMAL RADAR
TX RX
RESPONSE
RESOURCES
RADAR (v)
1 PASSIVE F1
1 2
P1
1 F2
2 2
1 P2 1
ACTIVE 1 2 4 (iv)
(i)
RADAR RESPONSE
2 UNDER WIDEBAND
ADAPTIVE
RADAR RESOURCES JAMMING
RESPONSE
ADAPTIVE RADAR
RESPONSE
2 2
2 (iii)
TOJ MODE
NORMAL RX (ii)
FUNCTION
Figure 7.6. ECCM Petri net of a radar system with weight shown for
transitions.
7.5 CONCLUDING REMARKS
The utility of Petri nets in the modelling of radar ECCM has
been discussed. Further abstraction can take into account the
statistical nature of the decision process in the selection of ECCM
against a given ECM threat. Side by side, it is necessary to develop
functional relationships to establish ECCM improvement factor
(EIF) to denote system performance and evaluation. Much work
needs to be done, especially on the EIF, whose proper
representation will require the total efficacy of the system, including
the associated weapons5. Petri nets show the potential for
contributing to these developments in the radar ECCM sphere.
REFERENCES
1. Peterson, J.L. Computing Surveys, 1977, 9(3), 223-52.
2. Agerwala, T. Computer, 1979, 12(12), 85-94.
3. Murata, T. Proc. of IEEE, 1989, 77(4), 541-80.
4. Ghezzi, et al., Proceedings of the Fifth International Workshop
on Software Specification and Design, 1989. pp. 60-67.
Artech House, USA. 1979.
67
CHAPTER 8
GAME THEORY MODEL APPROACH TO

RADAR ECCM EVALUATION
The essence of game theory is the strategic behaviour of the

individual player or decision maker in a situation of conflict. Each
player is regarded as having a series of plans or modes of behaviour
from which he can choose, i.e., the sets of strategies. The task of the
theorist is to clarify the strategic situation and determine the
optimum choices that ought to be made. Strategy must be regarded
as a theoretical concept invented by game theorists and refers to a
mode of behaviour. Game theory involves not merely the set of
strategies available to the single decision maker, but considers him
acting in a situation in which other decision makers too are acting.
What is more, each of these individual's behaviour is seen to directly
influence the behaviour of all the others. So, each player does not
find himself in a deterministic environment but rather the strategic
interdependence of the sort, "If I do this, he will do that, so I ought
to have done that, but he could have reacted as.... The parallels
with the ECCM-ECM interaction are perhaps becoming evident.
A formalisation of the interdependence of decision making is

achieved by noting the following four elements common to all
situations.
(a) Decision makers who are faced with
(b) Choices which jointly lead to
(c) Outcomes among which each decision maker gets a
(d) Payoff
There will be preferences of each decision maker for options
of various payoffs. In general, a greater payoff is to be preferred.
where a pure strategy [a pure strategy as opposed to a mixed
strategy (discussed later)] is being employed by two decision
makers, the choices and payoffs may be represented by a matrix as
shown (Fig. 8.1).
DECISION MAKER II
S1 S2 S3 S4
S1 C
DECISION
MAKER I
S2
Figure 8.1. Simple matrix
For example, if decision maker-1 (DM-1) selects strategy S1,

his opponent S3, then the outcome is represented by C. There will
often be at least two parts to C, the outcome perceived by DM-1 and
that perceived by DM-2. In many games, the sum of two parts of C
is zero and is called as zero-sum game. The outcome C may be a
series of outcomes determined by chance. An important point to
note (Fig. 8.1) is that the number of choices open to each decision
maker may not be the same. A convenient example here is that
DM-1 may have the choices to jam (ECM) or not to jam; a radar,
represented by DM-2 may have the choices of agility, change of
scan rate, CFAR or pulse compression.
8.1 APPLICATIONS TO ECCM EVALUATION

An example of a payoff matrix is shown in terms of an
electronic conflict in Fig. 8.2. The usual convention of decision
maker-1s options on the left, and his payoffs at the lower left of
each matrix element have been applied. This example is for an air-
to-air interception which the designer of the new AI radar faces. The
radar designer has the choice of including no specific ECCM
features and making a simple reliable sensor for beyond-the-
horizon detection with good long-range detection performance; or
including a jamming-strobe processor to provide some surveillance
capability against a self-screening target; or adopting a pulse-
doppler approach with somewhat lesser detection range but with
good discrimination against simple false targets. The intruder can
elect, prior to the mission, to attack without RF emission, to use
noise jamming, or use a false target generator. In this simple
matrix, the situation is not considered to be dynamic. The strategy
once adopted is retained, i.e., the radar decision maker selects his
70
Game Theory Model Approach to Radar ECCM Evaluation
ENERGY DECISION MAKER
QUIET NOISE FALSE TARGET

APPROACH JAMMING GENERATION
0 1 1
NO SPECIAL
FEATURES
RADAR
DECISION 1 0 0
MAKER
1 0 1
J-STROBE
PROCESSOR
0 1 0
1 1 0
PULSE
DOPPLER
0 0 1
Figure 8.2. Airborne interception matrix
options at the design stage, the enemy, likewise, does not change
during his mission.
In Fig. 8.2, the ordinal measure of payoff, denoted by either

1 or 0 has been used; this allows for ranking of preferences but does
not express degree of preference. If the attack is quiet, then the
preference for the designer is not to have invested in ECCM
features; this is shown in the first column payoff. Similarly, if there
is a noise jamming, the radar preference is to have J-strobe
processor; and if there is a false target generator, the pulse-doppler
processor is the preferred option.
From the matrix, the following conclusions can be drawn:

If ECM is expected, then it pays to have ECCM features.
Conversely, if the defender radar has ECCM features, it may
still be worthwhile considering a quiet approach.
If the capability of the false target generator is considered as
sophisticated rather than simple, such that it could usefully
overload even a pulse-doppler receiver, then the payoff in the
final row/final column would have been inverted.
The ordinal evaluation of payoff is quite limited in usefulness,
perhaps to guiding early design considerations when there are
enormous possibilities open.The valuable cardinal payoff is
introduced in the next matrix.
71
A final point to note from Fig. 8.2 is that it is cast in a zero-sum

form, or to use the more general term, constant-sum form. The
essence of this type of game is what one player wins is what the
other player loses. The sum of the payoffs is
constant, zero being the special case in point. It is worth
digressing briefly to debate whether warfare is a zero-sum game.
At first sight, it may be thought that an aircraft loss is a net
loss, and that with losses on both sides of the battle, the zero-
sum concept cannot apply. However, by taking the view that the
aircraft loss to one side is a score of 1 to the other side, then the
zero-sum applies. It is very common to score kills in this way.
Thus, aircraft loss to one side are scores to the other side, and
the application of techniques (such as ECCM) on one side, which
increases the effectiveness, is a score to the side employing the
technique and an equivalent loss to the other side. Here, elec-
tronic warfare as a zero-sum game will be considered. This is
also convenient because of the literature available on two-player
zero-sum games.
The next example is the design of the ECCM features of air-

defence radar that will have to detect and track and perform raids
in the presence of their supportstandoff jamming. The radar is to
replace a fixed-frequency magnetron radar with very few ECCM
circuits. The enemy is also developing an updated noise jammer
and has the options of retaining its old narrow bandwidth, but
providing substantially more power; or retaining the old power level,
but being able to generate over a wide RF band. The payoff matrix
for the radar under this situation is shown in Fig. 8.3. Now cardinal
payoffs have been employed, these are an expression
of degree of preference or utility.
In this example, the payoffs contain certain elements of both

cost and effectiveness. Thus, wideband agility is considered to be
worth twice as much as a sidelobe canceller against narrow-band
jamming, in the light of the development and ownership costs. The
chaff detector gets a negative rating because it is useless against
either ECM strategy and its expense is wasted; however, it should
not cost much to provide, so it is only a small negative payoff. On
the other hand, pulse-doppler is equally useless, but would cost a
lot more to provide. To complete this example, the wide-band agility
does not bring benefits when there is wide-band jamming and
would cost more to provide than a chaff detector. The sidelobe
canceller will probably not work so efficiently over a wide band.
Whenever possible, these payoffs should be evaluated at the weapon
72
system-level to give the highest meaningfulness to the analysis.

Again, it is to be noted, that this is a non-dynamic example, the
equipment is assumed to be deployed and used as it is
manufactured. A control provision for narrow-band/wide-band
operation is reckoned to be absent.
HIGH WIDE ROW

JAMMER POWER BAND MINIMA
NARROW
RADAR BAND
WIDE BAND AGILITY 10 -4 -4
SIDE BAND CANCELLER 5 4 4
PULSE DOPPLER -10 -10 -10
CHAFF DETECTOR -2 -2 -2
Figure 8.3. Air defence radar
It was quite easy for a radar designer to guess the payoffs,

but how will ECM payoffs be calculated? Having just declared that a
zero-sum philosophy will be adopted, this is very easy. Each payoff
total is made zero, such that a payoff of 10 for the radar designer
corresponds to -10 for the jammer designer. Accordingly, for the
rest of the paper, the radar payoff is only indicated. This follows the
usual practice of constant-sum games.
Some conclusions can be drawn from the matrix of Fig. 8.3.

It is quite obvious that there is no use for the ECCM designer (ED)
adopting a chaff detector or a pulse-doppler processor. If he thinks
it is very likely that the ECM equipment is narrow band as currently
fielded, then wide band agility is the preference. However, if it is
virtually certain that the new enemy jammer would be wide band,
then sidelobe cancellation (SLC) is to be preferred. If ED has no
perception about the enemy's ECM strategy, then he may use
max-min strategy, i.e., to put himself best-off in the worst case.
(The maximum of the row minima). Here again, SLC would be the
73
choice. If ED feels that ECM's strategy was equally likely on both

fronts, he may choose to take the best average, which indicates the
selection of wide band agility. The ECM decision maker will seek to
minimise the potential of the radar designer and follow a min-max
strategy (The minimum of the column maxima) and adopt wide
band noise jamming. For the supposition that both the decision
makers are rational, we will see the enemy fielding the noise jammer
and SLC as the radar's ECCM response. (From subsequent
material, it will be inferred that it may not pay to be too rational!).
In the example, one visualises game theory providing an

analytical approach to the uncertainty on the ECM threat. What is
to be appreciated is that in many of these conflict situations, the
strategy options open to each decision maker are known, the
general form of payoff matrix is known, but the particular choices
that need to be exercised are not known.
The next stage of complexity, and of realism, to introduce is

that of mixed strategies. Some preliminary aspects were introduced
in Chapter 3. Let the penetration of radar-guided missile by an
enemy using a decoy ECM be considered. If the enemy can induce a
guided weapon to fire at the decoy, then it can penetrate within the
reload/arming time. For any interaction, the radar may or may not
react at the threat/decoy. The simple ordinal matrix is shown in
Fig. 8.4, indicating the payoff on a zero-sum game. This is hardly
helpful for the mixed strategy case. Since this is a practical case,
there will be significant uncertainty. The radar may mistake the
decoy for a real target, or vice-versa, a real target may be classified
as a decoy.
In the graphical representation of Fig. 8.5, the horizontal axis
represents the real target situation on the left and decoy on the
ATTACKER REAL DECOY

TARGET
DEFENDER
REACTING 1 0
NOT
REACTING 0 1
Figure 8.4. Guided missile matrix
74
C REACTION
PAYOFF NO REACTION
TO
ATTACKER
D
M
REAL TARGET DECOY
PROBABILITY
Figure 8.5. Guided missile optimisation
right (The pure strategies of Fig. 8.4). The payoff to the attacker is
shown on the vertical axis; the reverse of this scale will be the payoff
to the radar by the zero-sum postulate. Point B in the figure shows
a high payoff to the attacker if the radar reacts to the decoy.
Similarly, point C in the figure is a high payoff if the radar does not
react to the real target. Points A and D represent the low payoff of
being shot at, or no reaction to the decoy. Now, pure strategies are
not viable to the attacker, for if he only uses decoys, he will never
penetrate! and if he never fakes, the radar will always visualise that
it is shooting at a real target. So, the practical case is a probabilistic
mixture of the two, represented by the horizontal axis of Fig. 8.5. By
making the judicious probability mix, the attacker will achieve a
better payoff than he would have achieved under the pure strategies
at points A or D. The point M corresponds to the maximum of the
minimum payoff curve, shown dotted in the figure.
In passing, it may be noted that Fig. 8.5 also illustrates the
most important theorem of game theory: In the two-person zero-
sum M *N matrix game under mixed strategies, the min-max is equal
to the max-min: Player 1 has the gain-floor precisely equal to player
2 loss-ceiling. This particular payoff is known as the value of the
game. The pair of optimum strategies applied in this case are known
as solution to the gain.
Now, if an ECCM is introduced to combat decoys, perhaps

staggered PRF, then a further payoff line EF is introduced as shown
75
C
REACTION
PAYOFF E NO REACTION
TO
ATTACKER
A F
REACTION WITH ECCM

D
REAL TARGET DECOY
PROBABILITY
Figure 8.6. GW radar with staggered PRF
(Fig. 8.6). Here, point E is adjudged to higher than A since this

ECCM feature may malfunction and allow a real target to penetrate
when it previously would have been engaged.
Similarly, point F is much lower than point B, since the
ECCM largely prevents the reaction to the decoy. A maximum payoff
line is again shown (dotted line) and its maximum value is
substantially lower than in Fig. 8.5. This is the value of ECCM in
reducing the effectiveness of the attack.
8.2 POINT-DEFENCE SURFACE RADAR

To demonstrate further application of game theory, the case
of a point-defence surface radar is considered. Here, a complex
multifunction/multimode radar (Fig. 8.7) might be proposed and
the selection of ECCM design features could be bewildering. Such a
radar combines search and track capabilities as well as missile
guidance functions to respond adequately to a multi-target attack.
The complexity of this case underlines the utility of the game theory
approach.
First, a higher-level matrix is drawn up, to localise design

areas for more detailed consideration. In these broad terms, the EW
threat is composed of airborne stand-off jammers (SOJs), self-
76
RADAR MULTIFUNCTION
SEARCH/TRACK/DESIGNATE/GUIDE
ECCM ADAPTIVE ADAPTIVE VARIABLE

ANTENNA SIGNAL TRANSMITTER
FEATURES PROCESSING PARAMETERS
DOMAIN SPECTRAL/TEMPORAL
SPATIAL
Figure 8.7. Features of multimode radar
screening jammers (SSJs), stealthy targets, decoys, repeater

jammers, radar-seeking missiles (ARMs) and chaff. The areas of
ECCM initially considered by the radar designer are:
A burn-through mode, which might use scan-rate control, long

integration or higher ERP.
Low sidelobes, by precision engineering, advanced design or
adaptive control of array elements.
Inclusion of passive facilities, in particular, track-on-jam (TOJ)
mode
Complex waveforms and processing, like frequency agility, pulse
compression, etc.
Considering all the above points, the matrix of Fig. 8.8 is
prepared. The values quoted in the respective columns are a bit
arbitrary, but these serve to illustrate the relative payoff merits as a
whole. While casting these values, it is also necessary to consider
the cost of implementing these ECCM features in a typical radar
system vis-a-vis their relative effectiveness. For example, the first
two rows of the matrix (Fig. 8.8) are more expensive to provide for,
perhaps three-times than those of row third and fourth.
In the above matrix, the following consideration have been

utilised:
The burnthrough mode would not be effective for the repeater
jammer and ARM threat, when compared for its effectiveness
with SOJ and stealth.
Its performance for chaff and decoy is rated to be around 50 per
cent. In the low sidelobe designs and adaptive nulling, the dis-
77
crimination against SOJ is more effective, though it is not a

solution for stealth detection.
The performance against other types of ECM is considered to be
in the range 10 per cent to 50 per cent.
The TOJ mode is not effective against stealth, decoy, and chaff,
although it can be utilised for SOJ and SSJ effectively.
Its function against repeater and ARM can be debated, but it
does provide a situation where the enemy has to locate the ra-
dar site through separate means.
The effect of frequency agility and similar complex
waveforms on stealth targets, decoys, and chaff, arises from
improved detectability of complex targets that may show deep nulls
on fixed frequency operation. The effect on active jammers due to
the necessary spreading of ECM energy is generally well accepted.
It also makes the design of an effective repeater much more difficult.
Search for the row minimum, followed by the maxmin

criterion, yields the following:
The use of complex waveforms and associated signal

processing emerges as the most fruitful approach for the given
scenario. Whatever be the approach adopted by the enemy, a score
of minimum five is achieved out of a possible maximum of ten. A
word of caution here: The spread of results does depend upon the
care with which the payoffs are ascribed, and also, the range of
ECCM
SOJ SSJ STEALTH DECOY REPEATER ARM CHAFF (ROW
RADAR ECCM MINIMUM)
HIGH ERP/
BURNTHROUGH
10 10 15 2 -10 -20 5 (-20)
EFFECTIVE
SIDELOBES
20 4 -20 2 10 10 2 (-20)
TRACK ON
JAMMING (TOJ) 5 5 -5 -5 5 5 -5 (-5)
COMPLEX
WAVEFORM 10 10 5 5 5 5 5 (5)
PROCESSING
Figure 8.8. High level matrix for point defence
78
threats assumed: For example, the poor marking of the low sidelobe
technique is due solely to the inclusion of a stealth threat. A balance
must be attained between choosing a wide enough threat spectrum
to lead to a fairly robust solution, and yet not misinterpreting the
matrix maxmin. For the ECM designer, the minmax rating of chaff
and decoy points to their promise for attacking this point-defence
radar.
One can now proceed to a more detailed analysis of this

promising ECCM approach of complex waveforms and processing.
The following features and implementations are considered for the
multifunction radar:
Frequency agility
Pulse compression
Variable transmitter parameters (e.g., PRF)
Pulse-doppler or MTD
Monopulse, lobe on receive etc to aid tracking
PRF jitter, range rate check, etc.
CFAR
Rapid AGC control
Impulsive limiters
These are entered into the ECCM-ECM matrix of Fig. 8.9
along with various types of likely ECM. Again, the payoffs illustrate
the relative merits/demerits, rather than absolute.
The preference for ECCM feature is, as now learnt, the

maximum of the row minima. There is a clear recommendation for
frequency agility followed by CFAR. Fortunately, this aligns with
the pre-conceptions of the radar community. Pulse-doppler does
not emerge as an efficient ECCM feature, so, perhaps in this
hypothetical example, it would be best treated as an anti-clutter
technique.
The analysis could be continued to determine an optimum

mix of ECCM features along the lines of what was discussed
previously wrt to Fig. 8.6. The techniques of linear programming
should be employed.
8.3 CONCLUSION
The evaluation of the efficacy of candidate ECCM features in
a military radar should be an important part of the design process.
79
Wide band Swept Long Short Spot Sidelobe Mainlobe Range Velocity Angle Row
noise frequency pulse pulse frequency repeater repeater gate gate gate minima
jammer jammer jammer jammer jammer stealer stealer stealer
Frequency 3 10 10 5 10 5 5 3 3 3 (3)
agility
Waveform 3 3 5 5 5 5 5 5 3 3 (3)
coding/pulse
compression
Variable 3 2 5 5 2 3 3 3 3 3 (3)
transmitter
pulse parameters
Pulse 5 2 2 2 10 5 5 5 5 5 (5)
Doppler
MTD 4 2 2 2 8 4 4 4 4 4 (4)
80
Monopulse 4 4 4 4 4 4 5 4 4 10 (4)
lobe-on-
receive
Jittered 3 3 2 2 3 3 3 3 3 -3 (3)
PRF/Doppler
range rate check
FTC/IAGC 2 2 4 2 2 2 2 2 2 2 (2)
Channelised 4 5 4 2 10 4 4 4 4 4 (4)
limiter
Dicke-fix 6 5 2 10 3 3 3 3 3 3 (3)
Figure 8.9. ECCM-ECM matrix with various types of possible ECM

In spite of its importance, a well-founded quantified methodology

has not hitherto been developed. The essence of the situation, to be
addressed, is one of conflict, and the established game theory is
known to be applicable. Of particular relevance is the capability of
the game theory to accommodate uncertainity. Accordingly, the use
of game theory is suggested here for ECCM analysis and design.
Some of the elements of the theory have been introduced and its
application to ECCM evaluation explained.
The utility of this approach has been demonstrated by

examples. It can be seen that the detailed choice available in ECCM
and ECM along with their effectiveness is very important, and
indeed this can be linked to both intelligence and secrecy. The game
theory approach is equally applicable to ECM design. The theory
can also accommodate degrees of cooperation as well as competition
and confrontation. For example, in a limited conflict, both sides
may recognise a need, not to interfere with certain air traffic or
navigation services. This has not been addressed here. The
employment of game theory will fulfil a long-felt need in the radar
ECCM design field.
REFERENCES
1. Nilsson, N.J. An application of the theory of games to radar
reception problems. In IRE Convention Record, Part 4. 1959,
pp. 130-40.
2. Rodinov, V.V. A game theory approach to detection of radar
signals in the presence of unknown interference. Radio Engg.
Elec. Phys., 1982, 27(9), 74-80.
3. Friedman, B. Noise and clutter rejection in radar as a problem
in non-cooperative game theory. In Proceedings of International
Symposium on Noise and Clutter Rejection
in Radar and Imaging Sensors, Tokyo, October 1984.
pp. 327-32.
4. Subramanian, A.K. ECCM evaluationSome models. In
Proceedings of International Symposium on Radar, Paris,
May 1984. pp. 523-27.
5. Von Neumann, J. & Marganstern, O. Theory of games and
economic behaviour. Princeton University Press, 1944 (Later
Wiley, 1964).
6. Owen, G. Game theory, Ed. 2. Academic Press, 1982.
81
CHAPTER 9
KNOWLEDGE-BASED INFERENCE FOR

RADAR ECCM
Modern radar implements a variety of ECCM schemes to

survive in the EW environment. Sophistication in the field of radar
ECM is perforce required to be matched with counter-response from
the radar community. A subsequent addition to the inventory in the
arena of ECM is the trend towards expert system for EW
applications1.
Under the modern EW, the time constraint imposed on the
radar ECCM response becomes crucial. The signal environment is
dense with deceptive, noisy, and often incomplete data. Hence, the
acquisition and the subsequent processing of information at the
radar site need to be viewed as an integral function.
9.1 ECCM DECISIONS

There is no standard or clear-cut procedure readily
available, so to say, in handling ECCM decision making under a
dynamic conflict. The major concern is to optimise on the resources
to issue an effective response on a real-time basis, utilising the
available knowledge at a given point of time. The input available for
such a decision, as mentioned earlier, can often be blurred by
deliberate deception, noisy and incomplete data from the prevailing
hostile environment. In such cases, a self-contained EW
surveillance process is inevitable to be linked to the radar sensor
mechanism for improving the dynamic ECCM response.
Referring to Fig. 9.1, a friendly ESM system implemented
near the radar site can serve the ECCM selection process more
dynamically. The functions of such an ESM system are significantly
different compared to the support measure employed by the ECM
community. These can be listed as
n Gather information on hostile ECM

n Nature of signals for intercept
Different from usual radar signature
More likely to be wide band, noise-modulated CW/ICW/
pulsed transmissions
Repeaters (deception)
Could also incorporate swept frequency (with variable
instantaneous bandwidth)
Standoff, escort and SPJ platforms (space diversity)
Intercept of a number of such emitters in a multi-
target environment
Range of intercept is 20 to 50 per cent more
than the maximum radar detection range.
The decoded ESM intelligence is then passed on to a
decision circuit where it is compared with a set of reference signals
generated by ECM-ECCM matrix library functions. The library
functions are generated by the knowledge engineer by interacting
with the knowledge base and the domain experts in the field.
ECM EMITTER INPUT

ECCM
RESPONSE
ECCM
DECODE COMMAND
ESM DECISION RADAR
ECM/ECCM
LIBRARY
OPERATOR
ESM DISPLAY
INTERFACE
Figure 9.1. A friendly ESM-Knowledge based ECCM System
84
Knowledge-based Inference for Radar ECCM
The radar operator (through a suitable ESM display) also verifies

the decision command on ECCM as human intelligence and
experiences play a significant role in the whole exercise. However,
the manual override provided is significant in cases where sufficient
reaction time is available.
It is clear from the above description that the main
constituents of a knowledge-based ECCM decision circuit will need:
Procedures adopted for the ESM decoding
Setting up library functions, and
Decision matrix
ESM SORTER
HARDWARE
FRIENDLY
ESM INPUT U
S
E
R KNOWLEDGE-BASED
ECCM DECISION SUPERVISORY
DOMAIN
EXPERTISE B CONTROL
U
S
SIMULATION TRAINING/
DIAGNOSTIC HARDWARE
DYNAMIC
INPUT
RADAR
FUNCTION
Figure 9.2. ESM/ECCM/knowledge based system
The database management, data flow and control can be

arranged in such a network for the best possible interaction.
As per the figure (Fig. 9.2), equal importance is also shown for the
simulation and training aspects of such a knowledge based system.
A subset of the above network is the block devoted to the
decision mechanism. To enable a reliable and credible decision
process, various data sets have to be acquired, collated, and used
in a dynamic time frame. Multisensor deployment, its data
acquisition and fusion form the major input for formulating the C 3I
block as shown in the following diagram (Fig.9.3). Another major
block is the target and ECM reference signature library. Platform-
85
MULTISENSOR FUSION C3 I
ESM INPUT
REFERENCE EW
DECISION
SIGNATURE LIBRARY DECISIONS
SUPPORT
INFERENCE
MODEL
WEAPON
FEEDBACK
SYSTEMS
Figure 9.3. Decision support
related information and deployment are also relevant in the decision

mechanism, especially for counter-weapon strategy.
9.2 ACQUISITION OF ECCM KNOWLEDGE BASE

From the beginning of the eighties, there has been sufficient
knowledge base in the radar ECCM field. Johnston2 started the
trend by the publication of an entire volume dedicated to the subject
and this has been followed by several papers presented at the
International Radar Forums and in prestigious technical journals3,4.
There have also been correspondence expressing diverse
viewpoints5,6 to keep the relevant arguments on ECCM evaluation
alive. The ECCM community is also kept well informed in the field
of ECM and other related EW topics by publication of a large
number of research papers, which update the existing knowledge
and as well as evaluate the trends for the future. While it is possible
to borrow ideas of such an emerging endeavour from the ECM
community, the basis for automatic decision circuit for ECCM
selection can often be controversial. This is due to the present
climate in this arena where a lack of definition, different
measurement standards, and diverse applications exist.
9.3 KNOWLEDGE-BASED CLASSIFICATION

Knowledge acquisition and sorting for the radar ECCM
system could be attempted in several ways7. The following are the
four essential ECCM groupings:
Functional (arranged sensor-wise)
Response to specific ECM
86
Deployment (field/environment)
Total weapon system efficacy
A further broad classification, as detailed by Schleher8,
would be to review the ECCM deployment under spatial, spectral,
and temporal planes:
Spatial Spectral Temporal
Low sidelobe design Low probability Pulse compression

of intercept
Sidelobe canceller Frequency agility PRF, pulse length
agility
Adaptive beam shaping Doppler filtering Constant false
alarm receiver
Adaptive nulling
Monopulse
Burnthrough
For the communication ECCM techniques, this could be

classified as:
Spatial Spectral Temporal

Null steering antennas Spread spectrum Time hopping
Direct sequence Chirp
Frequency hopping
LMS algorithm Error control
coding
Block codes
Hamming
BCH
Adaptive beam shaping Reed-Solomon
Convolution
codes
Viterbi decoding
Maximum SNR algorithm
It is necessary to mention that the knowledge base

representation is highly user-related, and thus, will have to be
flexible in its approach and classification.
Building of the knowledge base rests on an iterative process
of acquisition, rules production, reviews, user comments, and
updating with feedback of the respective domain experts. Flow
charts for problem solving are attempted to break complex
situations to simpler ones. Handcrafting and computer-based
induction are the two methods generally recommended in this
process. Both the methods seem to apply equally at the present
juncture.
87
9.3.1 Handcrafting
A review of the radar ECM-ECCM matrix, such as the one
published2, can help a Knowledge Engineer to go through the
handcrafting exercise for the representation of a knowledge base to
be used by an expert during problem solving. The limitation of this
method is soon perceived when representing the proper and
effective ECCM response against an environment of multiple ECM
threats; also when there are multiple choices of ECCM available for
deployment against a particular, validated threat. In both the cases,
further knowledge representation would require logical induction
and optimisation methods.
Example: Keeping the general concept as outlined above,
one such ECCM decision mechanism is examined here:
The following tree structure illustrates the decision process
for selecting the appropriate ECCM response, given a set of threats
and their decoding capabilities at a radar site. To get the example
working within bounds, a system-level ECCM decision is kept as
the goal, with limited ECCM inventory at the disposal of the decision
maker.
Referring to the Fig. 9.4, the following sequencing of events
and inference could be easily mapped:
(a) The sensors friendly ESM system identifies the threat and
determines its direction
(b) With the input from the radar and other associated sensors,
the range and the type of platform, from which the threat is
perceived, are determined
(c) The decision regarding the nature of the electronic threat
comes from the ESM signal decoder
(d) The ECCM response is then determined from the available
ECCM suite
The above inference procedure (a case of handcrafting), is
not always the best but can provide an initial estimate to the
inferencing procedure. Often, the rule-based system is inadequate
to handle dynamically varying parameters. Further, the hostile
intention inherent in the ECM-ECCM process belies such
streamlined approach. Hence, the decision mechanism becomes
one out of many possible choices.
The decision criterion under such a situation is to optimise
the basic time frame required to search and resolve. A rule-based
88
Range Data
Angle Of ECCM Radar/Other
Arrival Data Signal Decoding Sensor Input
AND
AND
Decode ECM
Decode:Wide Band/
Threat:
Narrow Band/Swept
SSJ/SOJ/Escort
Frequency
Chaff
AND
AND AND
89
SOJ
Platform
Wide Band ECM Wide Band ECM
SSJ/Escort SOJ Platform
Frequency Agility, AND SLB/SLC/

Matched Filter Adaptive AND
Response (LPI) Beam/Scan STC/MTI/
ECCM
MTD
Figure 9.4. A tree structured decision process of selecting ECCM response

inferencing then becomes a viable option for real-time ECCM

response.
9.4 KNOWLEDGE-BASED SIMULATION & EVALUATION

One of the basic building blocks is the utilisation of
simulation and evaluation of every ECCM parameter during
peacetime development cycle. As in other applications where rule-
based systems have been initially developed for diagnostic and
training routines, credibility of usage in the field of ECCM could
well be established in a similar fashion (Fig. 9.5).
A combination of sensor-oriented input can be connected
through a user intranet to link a number of processing boxes. A
supervisory/simulation control acts as the programmers tool to
relate the operators requirement to the different blocks. Depending
on the mode of operation (simulation/training or evaluation), the
operators keyboard has functional keys to relate the dynamic needs
FRIENDLY ESM DOMAIN EXPERTISE DYNAMIC INPUTS
SIGNAL ACQUISITION/
ANGLE OF ARRIVAL/ LIBRARY FUNCTION/ RADAR, ECCM STATUS
FREQUENCY KNOWLEDGE-BASED AND
MEASUREMENT/SCAN ESM/ECM ANALYSIS OPERATOR INPUT
TYPE DETERMINATION AND INFERENCE
ETC.,
USER INTRANET
SUPERVISORY CONTROL
SIMULATION,
ESM SORTER KNOWLEDGE- TRAINING RADAR
HARDWARE BASED ECCM AND FUNCTION
DECISION DIAGNOSTIC MONITOR
HARDWARE
USER INTRANET
Figure 9.5. Knowledge-based ECCM inference/simulation/training/

evaluation.
90
to the display on the video terminal. Modern facilities can also

include multimedia and graphic presentation, digital recording, and
storage and replay facilities. These help in examining the entire
simulation/training, ECCM evaluation and its efficacy, on an offline
mode for further study and improvement. Credibility on any ECCM
model, as in every other case of simulation, has to be built over
many recursive simulation and updating.
REFERENCES
1. Gaudry, E. ITT Electri. Commu., 1986, 60(2), 168-73.
Artech House, USA, 1979.
3. Proceedings of the International Conference on Radar, India
(1983), Paris (1984), Washington (1985), China (1986).
4. Li, Nengjing. Proc. of IEEE, 1984, 131(4), 417-23.
5. Johnston, S.L. Record of the CIE International Conference on
Radar, 1986. pp. 149-54.
6. Johnston, S.L. Correspondence, Pt. F, Proc. of IEEE, 1985,
132(3), 198201.
7. Subramanian, A.K. ECCM evaluationsome models In
Proceedings of Radar-84, Paris, 1984. pp. 523-27.
8. Schleher, Curtis D. Introduction to electronic warfare. Artech
House, 1986. USA.
91
CHAPTER 10
CONCLUSION
Radar ECCM and its Evaluation
In the preceding chapters, the subject of modern radar

ECCM and the aspects of its evaluation have been discussed. The
emphasis is on the need for simulation and analysis of ECM-ECCM
conflict situations with selected modelling methods.
It is hoped that this monograph would generate further

interest and study by all those scientists/engineers involved in
the modern radar design and its ECCM techniques.
INDEX
A CESM. see Counter ESM

Coherent signal processing 39
Abstraction 59 Communication network 2
Acquisition cost 56 Controlled spectrum signatures
Activities 54
asynchronous Cost-performance 5
concurrent 59 Counter ESM 43
Adaptive Countermoves 12
antennas 46 Cyber warfare 1
designs 49
nulling 52 D
Antenna
Degrees of freedom 35
function 36
Deployment models 22
gain 40
Detection 40
search 40
Anti-radiation missiles 15, probability of 3, 20
32, 54 Deterministic concepts 59
ARM. see Missiles: anti Digital signal processing 49
radiation DOF. see Degrees of freedom
Artificial intelligence 7 Domain expertise 90
Attenuators 52 E
B ECCM 1, 5
Bands acquisition 86
L 50 classification 86
S 50 decision tree 7
Bandwidth 41 decisions 83
interceptor receiver 43 designer 73
Beam dwell time 40 efficacy 5, 56
Beamwidth 33 evaluation 19, 55, 69
Benign environment 9 improvement factor 6, 19
Bistatic operation 39 MFR 50
Broad banding 52 antenna 50
Burnthrough receiver 54
equation model 27 transmitter 53
range 29 modelling 14, 60
C structured 56
ECM 1, 9
Canceller mode 62 equipment 12
Cardinal payoff 71 hardening 57
pod 12 H
specific response 21
threat 5 Handcrafting 88
ECM-ECCM Human intelligence 85
matrix 15, 20 Hybrid simulation 59
library functions 84 I
ED. see ECCM: designer
EIF 6, 19, 56, 67. see also Instantaneous bandwidth 51
ECCM improvement factor Intercept receiver 41
Electronic Interceptor 42
battlefield receiver bandwidth 43
management 13 Interface
beamwidth broadening 50 radar
intelligence 37, 39 interceptor 42
support measure 14 Intra-pulse coding 39
warfare 2 IPAR 47. see Radar: intra-
Electronic countermeasure 9 pulse polarisation agile
ELINT 39, 44. see also J
Electronic intelligence
Encrypted data communication Jammer
50 escort 29
Equations function 37
burnthrough 27 on-board 30
range 31 power 42
dynamic radar range 33 repeater 30
Escort jammer 29 self-protection 27
self-screening 62
ESM 14. see Electronic:
stand-off 62, 76
support measure
Jamming 3
counter 43
interference 3
decoding 85
sidelobe 20
receiver 25 Judging
Event-driven simulation 59 efficacy 6
EW 9 see also Electronic:
warfare L
flywheel 16 LC. see Lifecycle cost
threat 13 Lifecycle cost 57
F Linear chirp 12
Low flying detection model 10
False alarm 3 Low flying target 38
probability 3, 20 Low probability intercept 39
Frequency Low sidelobe performance 52
agility 39 LPI. see Low probability
range of operations 51 intercept
Functional models 20 modulation scheme 43
LPI-ECCM 43, 47
G
M
Game theory 12, 69
model 22 Main-lobe 36
Gulf War 2 Matrix 69
96
Index
dynamics 4 P
ECM-ECCM 20, 56
Payoff 24, 69, 74
Max-min 75
PB. see Pre-brief mode
strategy 73
Petri net 59
MEMs see Micro- Petri nets
electromechanical sensors extended 60
MFR. see Radar: multi function timed 60
antenna ECCM 50 Phased array 35
ECCM 50 antenna 49
Micro-electromechanical Polarisation control 50
sensors 7 Pre-brief mode 55
Min-max 75 Probability
strategy 74 mix 75
theorem 15 of detection 3, 20
Missiles of false alarm 3, 20
of intercept 41
anti radiation 32
Propagation 10
resistance 54
Pure strategy 69
patriot 22
Mixed strategies 24, 74 Q
Mode
Quiet range 45
look-down 10
track-on jammer 62 R
Model
Radar
detection 10
design
land clutter 11 land clutter model 11
Models detection 39
burnthrough equation 27 dynamic range equation 33
deployment 22 ECCM 83
functional 20 evaluation 93
game theory 22 intra-pulse polarisation agile
low flying target detection 38 47
MFR 22 jammer scenario 29
radar LPI 35
LPI-ECCM 44 modelling 34
Modulation 33 models 10
MTI 19, 21 multifunction 21, 49
performance 5, 10
improvement factor 21
point-defence surface 76
Multi-target attack 76
surveillance 40
Multifunctional radar 3
variables 34
Multiple beam generation 50 functions of 34
Multisensor deployment 85 Ranges
Mutual screening range 30 burnthrough 29
O Real-time problems 60
Receiver
Optimum aperture 42
redundancy 56 intercept 41
strategies 75 Receivers
Ordinal measure 71 ESM 25
97
Repeater jammer 30 Stochastic considerations 59

Resolution 59 Strategies 69
RF chain 50 pure or mixed 23
Robustness 9 Structured redundancy 56
factor 37 Sub-clutter visibility 22
RPV 55 Surveillance radar 40
Rules production 87
T
S
Techniques
Scan coverage tailoring 50 mitigation 6
SCV. see Sub-clutter spectral estimation 6
visibility Technology
Segmented aerial coverage GaAs 50, 54
51 stealth 15
Self Time
screening bandwidth product 45
jammer 62 basic nets 60
range 30 stamps 63
target 70 TOJ. see Mode: track-on jammer
Shipborne systems 10 Transitions
Sidelobe suitable weights 66
blanking 62 Transmitter 12
cancellation 73
canceller 62 V
control 53 Value-added performance index 45
Simulation 15, 90 Variational parametric analysis 34
SLB. see Sidelobe blanking VLSI 7
SLC 73. see also Sidelobe
canceller W
SOJ 76. see also Jammer:
War
stand-off
desert 55
Sorting 86
SPJ. see Jammer: self- Falkland 2
protection Gulf 2
SSJ. see Jammer: self- Weak-time semantic 66
screening Weapon system efficacy 22
STALO 21 Weighting 52
Stand-off Jammer 62 Z
Stealth
system 31 Zero-sum form 72
technology 15 Zero-sum game 72
98

ECCM

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

ECCM

Caricato da

Copyright:

Formati disponibili

MODELLING RADAR-ECCM

Defence Research & Development Organisation

MODELLING RADAR-ECCM: A SYSTEM APPROACH

Asst. Editors Editorial Assts. Cover Design

2003, Defence Scientific Information & Documentation Centre (DESIDOC), Defence

Printed and published by Director, DESIDOC, Metcalfe House, Delhi-110 054.

5.1.2 Search Antenna 40

Conflict situations are part of human life. What this

Grateful thanks and appreciation to Stephen L. Johnston,

1.1 IMPORTANCE OF ECM

Electronic Detection Sensors

1.3 BASIC CONSIDERATIONS

1.3.1 Jamming Interference

SEARCH TRACK GUIDANCE

carries out a variety of tasks, from the initial detection of a target

Transmitter and receiver Transmitter and receiver

1.4 ECM-ECCM MATRIX DYNAMICS

ECCM response. What is more, new ECCM features cannot be

1.4.1 ECCM EfficacyThe Difficulty

the ECM arena, which has assumed more glamorous and

1.4.3 Judging Efficacy of an ECCM Element

1.5 ASSESSMENT OF THE STATE-OF-THE-ART

(c) Artificial intelligence/knowledge-based inference mechanism for

MODERN RADAR DESIGN & ITS

It then becomes essential to examine the present and

2.2 RADAR PERFORMANCE

Most of these conditions are prominently featured when a

2.2.1 Two Representative Radar Models

2.2.1.1 Low flying detection model

where, F ( ) is the pattern propagation factor. This, for small

where, Ro is the maximum range of the radar in free space

2.2.1.2 Land clutter model for radar design

Clutter reflectivity = o (Fc )" and o = R Sin ()

= K (Rh/R)4 for R > Rh

where, R = Radar range, Rh is the radar horizon and Fc is the two-

While viewing this graph, it is to be remembered that the

been suggested whenever analytical expressions could be derived

method of assessing radars ECCM performance against a host of

2.3 COMMAND, CONTROL, COMMUNICATION &

Figure 2.2. Electronic battlefield management

The concept combines a number of functions as follows:

As such, there is no dispute about the speed of execution

2.4 ECCM MODELLING

2.4.1 Radar Systems

(c) According to platforms (land-based, shipborne or airborne)

2.4.2 Communication Systems

2.5 MATHEMATICAL ANALYSIS

(a) The number of independent parameters of the radar and the

2.7 THE EW FLYWHEEL

Figure 2.3. Selecting appropriate ECM for the jammer

selection. Such a scheme also has the beneficial feature of having

ECCM EVALUATIONSOME MODELS

3.2 ECCM IMPROVEMENT FACTOR

In relation to the vulnerability experienced in a deployment

3.3 FUNCTIONAL MODELS

evaluation; such as processing gain, cancellation ratio, J/S

3.4 RESPONSE TO SPECIFIC ECM

individual constituents to the total system (1/I = 1/I1 + 1/I2 + 1/I3

3.5 DEPLOYMENT MODELS

3.6 WEAPON SYSTEM EFFICACY

3.7 GAME THEORY MODEL

effectiveness have been traditionally studied through such

3.7.1 Case Study 1

3.7.2 Case Study 2