inverse synthetic aperture radar simulator implementation ... · the theoretical representation of...

T KOSTIS et al: INVERSE SYNTHETIC APERTURE RADAR SIMULATOR IMPLEMENTATION

IJSSST, Vol. 10, No. 3 ISSN: 1473-804x online, 1473-8031 print 44

Inverse Synthetic Aperture Radar Simulator Implementation for an Extended Naval Target for Electronic Warfare Applications

Theodoros G. Kostis Dept. of Information &

Communication Systems Engineering University of the Aegean

Karlovassi, Samos 83200,Greece. e-mail: [email protected]

Sokratis K. Katsikas Dept. of Technology Education &

Digital Systems University of Piraeus

150 Androutsou Street, Piraeus 18534, Greece. e-mail: [email protected]

Abstract—Airborne Inverse Synthetic Aperture Radar systems provide excellent surveillance abilities for imaging naval targets. Considering this as a maritime surveillance threat we derive a simulator system in this paper that can be used to synthesize false targets that would appear as credible military fleets. The resulting electromagnetic signature synthesis allows the implementation of returns that appear to backscatter from areas of high reflectivity as arranged in military superstructures instead of commercial counterparts. The conceptual model for this simulation is constructed by considering standard representation theory. The modelling approach follows standard abstraction degrees for the three-dimensional synthetic environment, target fidelity and inverse scattering. The pace engine that acts as the computing moving force for the simulation is mathematically presented. The inverse scattering signal follows standard ISAR theoretical guidelines. The data exchange standard between the various elements is presented. Issues in implementation and project success are discussed by drawing upon the verification and validation analyses.

Inverse Synthetic Aperture Radar; Simulator; Conceptual Model; Complex System; Extended Naval Target.

I. INTRODUCTION

High Range Resolution (HRR) radar systems mounted on airborne platforms provide enhanced surveillance functions of ship targets [8]. The superior resolution is performed by diverse Inverse Synthetic Aperture Radar (ISAR) modes of operation [35]. Such ISAR modes have the ability to resolve many features on a target thus producing high resolution imaging details in the slant (x) and cross (y) range dimensions of the target. Thus the imaging output is two dimensional and even three-dimensional with more processing in contrast to conventional radar which can only resolve a target as an intensity point on a scope. There is a lot of recent, active and enthusiastic research motivation for ISAR systems countermeasures because the results have big practical applications to the military remote sensing community [6] [18]. We choose to employ mathematical modelling and subsequent simulator implementations of HRR radar systems as they are important tool in this countermeasure research process because at least they allow the formation of feasibility studies for designs that use a complete three-dimensional synthetic environment.

In this manner the simulator is placed in a virtual worldspace that can reproduce a real-life scenario, such as an attack of a sea-skimming missile to a naval friendly asset. This reproduction can create a false naval target and perform deception tasks on the threat signal in real-time.

Such simulator system implementations provide an added value because they enable design engineers to envision novel just-in-time countermeasure systems designs and moreover allow the drafting of appropriate proofs of concept [14][15][16]. Therefore the problem that this paper is trying to address is the conceptual and implementation steps towards the design of a High Range Resolution Radar (HRR) simulator for electronic warfare purposes.

The preferred method to deceive an ISAR system is the use of a transponder which can analyze the threat signal, inject a false reflectivity solution and send back the result to the threat antenna [27, p. 180]. An important function in a false target generator is the creation of the target’s reflectivity solution. We propose that this solution can be furnished by a simulator and its supporting database of results. We argue that such a false target generator can be achieved by using a virtual reality approach supported by a corresponding database system.

In this paper the target is a naval vessel of considerable superstructure dimensions (battleship class) which is tracked by an airborne threat signal. Using an ISAR system the threat sensor is able to resolve many reflectivity centres on the target classifying the ship as an extended target. The common practice for the simulation of an extended target is the multiple scatterers model. In this manner the target’s radar cross section (RCS) is represented by many pointlike reflectors that together synthesize its electromagnetic identity back at the radar receiver. For example high reflectivity centers like gun placements in lower superstructure coordinates create high



valued outliers in the ISAR image that do not correspond to the real height coordinate [11]. The simulator should be able to cater for this case in order to be able to produce vessels that could be classified as military candidates. Furthermore the target should be able to advance into the virtual battlespace in a way that produces inconsiderable effects on the ISAR imagery process. Because any translational motion of the target is not a major contributor to the ISAR imaging output due to its physical nature. Only pitch and roll motions must be allowed to synthesize the high resolution image [32]. Otherwise the threat itself will drop the track as non-credible. Based on the above we craft the outcome of the simulation to encompass all the above issues in order to represent a credible electromagnetic return to the threat signal.

The paper is organized as follows: In Section 2 the conceptual model of the application is introduced. In Section 3 the modelling approach that is deemed able to cover the aims of the project is described. Section 4 explains the data exchange standard that was utilized for this project. Section 5 deals with relevant verification and validation issues. In Section 6 the discussion about thequality focus of the model and project is presented. Finally in Section 7 concluding remarks are given.

II. CONCEPTUAL MODEL

The Conceptual Model (CM) is the first step in the task of producing an abstract representation for a system and a draft for its corresponding software model using algorithms and data [2]. We start by applying the properties of an intellectual problem to our case [4, p.23]. The goal which can tell us how to judge the outcome would be the output of the simulator. All findings should correspond to existing verified and validated ISAR simulators found in the related work review. Therefore the initial state as well will be dictated by existing literature. Our main objective then would be to ensure that an improvement has been achieved towards the authoring of an ISAR simulator especially for electronic warfare purposes.

Now we have set up the primary mechanism for the successful communication among the Simulation Engineers (SEs), like software designers, system engineers and system analysts and the Subject Matter Experts (SMEs), such as engineers who specialize in the field of the simulation context. Analytically the SMEs are tasked with guiding the complete process of the simulation since they can predict out of experience the right steps and avoid any representational pitfalls that can lead to a product unfit for its intended purpose.

The Conceptual Model deals with the context of the representation and the methodology that will be later followed in a concise and structured manner. We have comprised a set of operations as a toolkit, called FB-16, in order to capture the representation and methodology

issues that can be utilised to transform the initial states and resources of the input and produce a proper output using the FB-16 toolkit as the transfer function. We have paid attention to specific constraints that govern the realm of ISAR Imaging in order for the output to be realistic. This is important because we aim to use the outcome of the project in order to recreate a realistic target as seen by high-resolution electromagnetic eyes [13]. In order to determine the final draft of the Conceptual Model there are five steps that have to be completed: Related Work Review (RWR), Application Domain Definition, Problem Space Decomposition, Entity Abstraction Degree and finally Entity Relationship Identification [22].

A. Related Work Review (RWR)

[7] have clearly presented the mathematical basis of the Inverse Synthetic Aperture process. [29] have described a technique used to simulate ISAR images of a ship model while under angular motions such as yaw, pitch and roll. [25] have presented the theoretical analysis of SAR techniques as can be applied to ISAR imaging of ship targets. Emphasis is given in the exploitation of information resulting from the point spread function. Also foundations are laid towards the study of interference effects (glint). [9] have introduced the ISARLAB software package which is a comprehensive set of functions that emulate the particular functions of an ISAR system. [5] have developed a simulation program which can generate ISAR images of ships. The method is based on the localization of dominant scatterers and has applications in evaluating the performance of automatic ship classifiers. [19] have investigated the acquisition of top or side view ISAR images with the proper cross range scaling. The technique is based on the measurement of slopes of the two main feature lines of the ship, which are the center line and the stern line. This process has the advantage of using only the acquired image to complete its tasks. [20] have investigated methods to obtain three dimensional radar cross section (RCS) images using the ISAR concept. Results are provided towards the degradations effect of specular multipath effects on the final image. [26] have described a method of ISAR image classification based on a comparison of Range-Doppler imagery to existing three dimensional ship reference models. This technique uses a sequence of ISAR images in order to estimate the dominant ship motion. In all above indicative work there is no mention of the computing force that provides the motion of the radar and target platforms. Our work makes an attempt to fill in the details of an ISAR simulation analysis in a virtual reality environment.



B. Application Domain Definition (ADD)

The task of Application Domain Definition is given to the SMEs that have authoritative information about the actual situational context. Usually at this point in time the SMEs will hold several meetings with the SEs and discuss the theoretical and practical milestones that have to be observed during the course of the project. Usually at this stage the SEs will have only superficial knowledge about the subject matter. On the other hand SEs that can perform the task of SMEs are valuable for any particular situation.

An historical angle was introduced in order to make the work more appealing to the involved parties.

Specifically the theatre of operations is the airborne assault on the Deutsche Kriegsmarine Schlachtschiff BISMARCK by the Royal Air Force Swordfish swarm in the Battle of the Atlantic, back in 1941. The introduction of an historical perspective to the project provides organization, stimulation and inspiration towards its completion and further maintenance. Therefore in this particular case the SMEs should first explain to the SEs the operational behaviour of an ISAR system mounted on a Swordfish biplane when viewing the Bismarck, as shown in Figure 1.

Figure 1. FB-16 Simulation geometry draft

The theoretical representation of Inverse Synthetic Aperture Radar systems is the dual of a Synthetic Aperture Radar system in spotlight mode, as shown in Figure 2(a) [Wehner, 1998]. And to introduce the wider concept of radar imaging we have included Figure 2(b). From Figure 2(b) we derive the final simulation theoretical basis which is shown in Figure 2(c).

The ADD for this project deals with the data collection radar images of Slant Range vs Doppler Frequency for a target under Spotlight Mode. These radar images are known as Inverse Synthetic Aperture Radar (ISAR) images. There are two major techniques employed for the creation of an ISAR image. First the concept of Range-Doppler (RD) imaging states that it is possible to achieve a two-dimensional view of a ship target at sea by using the relevant RD algorithm. This algorithm is phase preserving and is best suited for applications using continuously collected data. The second concept is the motion compensation of the scatterers. This method corrects any unnecessary phase components of the reflected waveforms

so the image can be processed by the focus imaging method. Drafting our simulator we should provide a solution to the threat signal that is motion compensated.

It is now assumed that the threat radar’s signal bandwidth is sufficient for the successful range resolution of the numerous superstructure elements. In this case the signal signature of the target for wideband illumination in the time domain is its Slant Range Profile (SRP). The associated frequency domain translation of the SRPs produces the last step of the simulation, which is the ISAR image.

The final touch to the ADD is the theoretical foundation of the ISAR image data format which is called Polar Reformatting (PR). PR is an interpolation process where a set of samples uniformly spaced in cylindrical coordinates is fitted into a rectangular grid. Therefore the output of the simulation is the interpolation of the ISAR waves to a rectangular grid. Again we draft our solution to correspond to a polar reformatting process.



(a)

TargetPlatform

RadarPlatform

Transmitter Typeie Tx Parameters

Waveform Selection

Antenna Typeie phased array

parabolic

Receiver Typeie Rx Parameters

InformationExtraction from

Signals Received

Propagation of Radar Waves Modelingie. Parabolic - Ray Tracing

Hybrid - Custom - Custom Hybrid[Refraction - Diffraction]

correlationconvolutionprocedures

Glint

Ground Clutterand other clutter types

Target Typeie. rcs

3D EngineSystem Dynamics

DirectionalInterference

ie. JammerChaff

ElectromagneticInterference

Dynamic Errorsie. Platform MotionScanning Motion

Vibration

MotionProvider

MotionProvider

All Other Noiseie. equipment noise

General Simulation Parametersie Code Specific Considerations

Sampling Rate

MatchedFilter

Results Formation

(b)

(c)

Figure 2. Theoretical Inverse Synthetic Aperture Radar process (a) Theoretical Explanation (b) Complete Model (c) ISAR Subset

C. Problem Space Decomposition (PSD)

The entities and processes that must be represented for the successful accomplishment of the simulation are defined. A basic memorandum of understanding between the SMEs and the SEs regarding the level of detail that the

simulation should encompass must be drawn at this moment.

For this particular project a crude list of entities as shown in Table 1 would involve the radar transmit and receive functions, the target properties, the environmental impact on the emitted signal and the digital signal processor simulator after the radar receive block.



TABLE I. ENTITIES

Simulation Reality 1 Target cartesian coordinates

plus inherent amplitude & phase

Target physical properties & electromagnetic signature

2 Radar slant range and cross range cartesian coordinates with respect to the center of the target

Most prominent appears to be the middle of the ship

3 Sea-level distance from radar to target

FM height finder radar (altimeter) on airborne platform

4 Radar operational parameters

ISAR system particulars

5 Aspect angle from radar to target

Change of aspect angle provides the resolutions ability

6 Pace Engine Target movement due to forces of nature

7 ISAR processor details Range-Doppler processing

8 ISAR system output Slant Range Profile and ISAR Image of target

Now we can draw the necessary associations between

the entities and come up with the corresponding processes, as shown in Table 2. Again as above the comparison

between the reality and the simulation is strongly taken into account.

TABLE II. PROCESSES

Simulation Reality A Provide information to the

Pace Engine of target Cartesian coordinates to the pace engine

Physical presence and movement of target

B Provide information to the Pace Engine of radar two-dimensional (slant ranger and cross range) coordinates

Physical presence and movement of radar

C Provide information to the Pace Engine of radar’s third (height range) dimensional coordinates

Measurement - captures reality with a sensor

D Provide information to the ISAR Processor about the radar’s operational parameters

Instrumentation – operational information

E Aspect angle variation Caused by changes in target/radar location

F From Pace Engine to rotated points database

Caused by changes in time

G From Pace Engine Database to ISAR Processor

Recording Process – processes history of target in computer memory

H From points database to ISAR processor

Computer process – Range-Doppler Processing - translates reality to computer memory



D. Entity Abstraction Degree (EAD)

The representational abstraction of the involved entities is finalized in this step. The level of accuracy, precision, resolution and fidelity of the entities and processes is determined. For this project the level of aggregation should involve a target that is made out of one-hundred and forty-five points. The ISAR system will be mainly operating at 5 GHz with a Pulse Repetition

Frequency of 1KHz. Other operational values can easily be accepted by the simulator.

E. Entity Relationship Identification (ERI)

The relationships among the entities are identified in this design phase. It is ensured that all constraints and boundary conditions are properly imposed by the simulation context. All operational and functional requirements are taken into consideration, as shown in Figure 3.

Figure 3. Entity Relationship (E-R) Identification

III. MODELLING APPROACH

First the physical situation is analyzed. In this case a phased array antenna is transmitting and receiving a collection of numerous agile beams towards the target rapidly scanning it in azimuth and elevation. Also the digital signal processing block of the radar can produce results either in sequential or parallel modes. Therefore a simulator design is needed for a parallel-to-parallel or a parallel-to-sequential physical process. In more detail the system is divided into its building block, which may be call datatom (from data and atom). All data for one point is processed and the resulting information, which may be called informatom (from information and atom), is stored into one or many databases. The datatom can be defined as the simplest configuration that can provide satisfactory verification in a

simulation process. The informatom can be defined as the simplest result or group of results that can provide satisfactory validation in a simulation process. Then when all points are processed and the complete databases are available, graphing results are produced [13].

A. Modelling the Synthetic Environment

All objects like radars and targets exist in a three dimensional matrix which is defined as the Synthetic Environment, as shown in Figure 4. Typical types of objects that exist in the Synthetic Environment include Radar Platforms (airborne or terrain based), Target Platforms (airborne-ground based), Environment Volumes (clouds) and Terrain Platforms (sea, land). Moreover those platforms can be stationary or in motion.

Figure 4. Synthetic Environment Modelling



B. Modelling the Naval Target Platform

Each reflector is characterized by its polarization profile for wideband illumination by a monostatic radar, which is the simplest situation where a single sensor is acting as the surveillance function. It is safe to assume a plausible solution for the radar cross section of each of these points and arbitrarily create the input data to the model. Analytically in Figure 5(a) the upper layer of the vessel is defined. We call this the single layer model because it includes only the topmost reflective parts of the target’s silhouette. For example the mast of the vessel is 52 meters but the mounted antenna reflective surface is at 49 meters. In this case this would be the topmost reflective point. But a ship is a three dimensional target and

scattering points that are located below the topmost parts have an important role in the overall ISAR image. Therefore we coded more points situated lower in the three dimensional target mesh. In Figure 5(b) we present the enhancement of the target model where another middle lower layer of the superstructure is set. Now there are reflectivity points emanating closer to the deck level of the target as opposed to the topmost points of the target [15].

We conduct the above approach in order to add more reality to the overall simulation effort. Because the radar sensor strongly captures returns from all scattering centres of the target vessel in all slant, cross and height coordinates.

(a) (b)

Figure 5. Extended target modelling (a) Single Layer Model (b) Multiple Layer Model

Using the above approach the vessel is divided into

many pointlike scattering points. The main point is block 73 which is in the middle of the vessel.

From this point all others are calculated either by adding or subtracting a standard value, as shown in Figure 6.

Figure 6. Point scatterers modelling



C. Modelling the Pace Engine

The Pace engine is the computing moving force that proceeds the points on the ship in time to new locations from their initial values depending on the (ordained by the SMEs & SEs) motion of the ship. Therefore the movement of the target can be represented by a dynamical system. Modelling of the Pace Engine is done by utilizing the theory of affine transformations, whereby a three-dimensional point on the ship is multiplied by appropriate 4 by 4 matrices that provide motion effects, as explained in an excellent descriptive manner by [1]. For example a point in a three dimensional plane represented by a 4 by 4 matrix is multiplied consecutively with other 4 by 1 matrices that perform roll, yaw, pitch and translational transformations in space so to produce new motion coordinates in 3D space that can be easily translated into 2D space. This transformation is necessary because the ISAR process has a two-dimensional outcome. In order to explain this in a better way we use an example of a 4 by 4

matrix approach used in order to pitch the target into time [1] as shown in Figure 7. By multiplying the transfer function matrix with roll, depression angle and aspect angle matrices we can perform any motion possible by a ship vessel. An important point must be raised here by considering that translational motion does not play a major part in ISAR imaging and should be treated in this manner. We defend the above approach and provide for the translational motion of the vessel because we claim that we can lay the foundations for a false target generator system. Such a generator should be able to move the false ship target into space and therefore we consider the furnishing of translational motion to be of great importance in our case. The complete mathematical solution is shown in (1) in full form and in (2) in block form. The contents of the movement permutation matrix are outlined in (3). Respectively the aspect angle and depression angle matrix is given by (4), the roll and pitch matrices are given by the sets of (5).

Figure 7. Pitch motion block - affine transformations with time advance

( ) cos 0 sin _ _ _ ( )( ) 0 1 0 _ _ _ ( )

[1]( ) sin 0 cos _ _ _ ( )

1 0 0 0 1 1

x new translation unit in x x oldy new translation unit in y y oldz new translation unit in z z old

θ θ

θ θ

⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥ ⎢ ⎥= ×⎢ ⎥ ⎢ ⎥ ⎢ ⎥−⎢ ⎥ ⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦ ⎣ ⎦

Output Transfer Function Input

_ P _

_ ( ) [2]0 1T

Movement ermutation Matrix tShip Position t xrotation ⎡ ⎤

= = ⎢ ⎥⎣ ⎦

where,



[Movement_Permutation_Matrix] = [Aspect_Angle-Depression_Angle_Matrix]x[Roll_Matrix]x[Pitch_Matrix] [3]

cos cos cos sin sin_ _ _ sin cos 0 [4]

sin cos sin sin cos

1 0 0 cos 0 sin_ 0 cos sin , _ 0 1 0 [5]

0 sin cos sin 0 cos

d a d a dAspect Angle Depression Angle Matrix a a

d a d a d

Roll Matrix Pitch Matrixθ θ

θ θθ θ θ θ

−⎡ ⎤⎢ ⎥− = −⎢ ⎥⎢ ⎥⎣ ⎦

⎡ ⎤ ⎡ ⎤⎢ ⎥ ⎢ ⎥= − =⎢ ⎥ ⎢ ⎥⎢ ⎥ ⎢ ⎥−⎣ ⎦ ⎣ ⎦

D. Modelling the Inverse Scattering

The radar platform is characterized by the radar operational parameters: wavelength, type of modulation, transmission power and antenna aperture. These parameters can be set to many different values depending on the simulation scenario. The main point is the simulation of the range calculation between the radar and

the target points as they move and rotate in the virtual worldspace. This digital world is described in Figure 8 and follows well established and recent geometrical models in the field of ISAR imaging [24]. And for the ISAR calculations we have used the mathematical approach used by [7] which is sound and fully describes the involved imaging mathematics.

Figure 8. Complete Simulator (xrotation function provides all rotation kind of functions)

Moreover for the target resolution cell calculation we

followed the modelling approach which can be found in [21, p. 98] called the Cookie-Cutter footprint calculation method. This is a simple way to represent the radar coverage of the target and is deemed adequate for this representation. The radar’s transmitted angle pattern is made to have an equal diameter at all the major beamwidth points of 3, 10 and 20dB. The resulting pattern is an elliptical cone in rectilinear space. Those cones offer the footprint which can be seen in Figure 9(a) from above.

We discuss Figure 9(a) further by stating that the resulting footprints are the outcome of the polar format algorithm usually implemented in conventional ISAR systems, as [3, p.105] explains. The two-dimensional sampled data space encompasses samples that are stored in a polar coordinate form in the radar’s memory. Moreover the spatial occupation of the received data is calculated using the small-angle approximation again as in [3]. This way the radar achieves its azimuth resolution which dictates the afore-mentioned data storage requirements.



003

007

004

012017027

058

022032037

034

047057

054

062

074

078

082

094

098

102

048

014

008

005

009

013018

019

023

024

028

029

033038

039

042

043

044049

052

053

079 059

063

064

067

068

069

072

073

077

083

084

087

088

089

097 092

093

099

103

104

107

108

109

112

113

114

117

118

119

122

123

124

127

128

129

132

133

134

137

138

139

143

144

145

45o

Radar Line of SightLOS

RS10

Zaxis

Xaxis

Yaxis

Resolution Cells Inclusion for ISAR Calculations

RS05

RS01

RS15

RS20

RADAR

001

002

006

010

011

020

021 016

015025045

026

030050 035040055060065070075080085090095100105110115120125130135140

036041 031046051066111 106 096 071 061 056081 076091 086136 101116121131 126

142

141

(a)

(b)

Figure 9. (a) Theoretical implementation (Polar Format Approximation)

(b) Software implementation (ISAR Function)

The final step of the simulation is the way to represent the signal processing block after the receiving block, which usually is a Phased Array Antenna system. After reviewing the theory of ISAR and phased array antennas the solution of the signal processing block representation is the division of the vessel into 33 resolution cells. Each cell contains the raw addition of the amplitudes and initial phases of the target with the phase delays that are added by the distance calculator block. The resolution cells are

depicted in Figure 9(b). A similar explanatory figure can be found in [30].

There are certain shortcomings related to the fact that the ISAR method is inherently dependent on the target’s changing its viewing angle to the radar [35, p.435]. First the cross range dimension scale is a direct function of the target’s aspect angular rotation rate. Second the ISAR image plane does not reveal the true aspect of the target. And finally target dwell time required to produce a given cross-range resolution is dependent o the target’s aspect



rotation rate relative to the radar. Therefore high valued reflectors in lower superstructure height will distort the shape of the ship, as the shape corresponds to a large degree to the battleship’s silhouette but is actually a measure of the electromagnetic backscatter emanating from the superstructure. Therefore it should never be

confused with the physical dimensions of the target. In other words the Range-Doppler method will exhibit higher reflectance on parts of the target that may be gun placements or antennas, leading to a purely microwave type of observation [11].

IV. DATA EXCHANGE STANDARD

The adoption of a standard for the communication protocol between the input data, the transfer functions and the output of the project increases the possibility of a successful realization, as can be seen in [23]. For this project the propagation of information between the system blocks has a custom made form. There are three main advantages. First it is easy to work through the creation

and compilation phases of implementation without having to learn a current protocol. Then debugging is easier to conduct because the authors had full control during the creation process. Finally in this manner it is much easier to upgrade and extend the system thus making it able to easily adapt to future demands. An example of a packet structure that is used is presented in Figure 10.

TSR

SlantRange

TCR

CrossRange

THR

HeightRange

TID

TransactionID

CR

ScattererAmplitude

IP

ScattererPhase

Example of a Frequency Domain Packet

Target Scatterer Coordinates

n

PulseNumber

RADAR PROPERTIESTARGET PROPERTIES

RSR

SlantRange

RCR

CrossRange

RHR

HeightRange

Radar Antenna Coordinates

φrot

TargetRotation

Angle

ENVIRONMENT &MOVEMENT ENGINE PROPERTIES

u

SpeedthroughMedium

λ

Wavelength

θ

AspectAngle

θrot

RadarRotation

Angle

Output Packet Transfer Function Input Packet

[slant range profile packet] = MOVEMENT ENGINE [frequency domain packet]

[range-doppler packet] = RADAR PROCESSOR [slant range profile packet]

Figure 10. Description of information tokens

Each point creates32 movement points

Get target point's x,y and z coordinates

Pace Engine : Affine Transformations

Get radar antenna point's x,y and zcoordinates

A

Distance Calculator (32 distances) Get target point's reflectivity and phaseinformation

B

Slant Range Profiles(Time Domain)

ISAR Image(Frequency Domain)

Figure 11. State machine implementation

Packets are declared as either input or output. The slots

of the packets are strictly assigned to the corresponding kind of information. For example slot 2 in an input packet

can never be anything else than a slant range parameter of a scatterer. The next step is the translation of the above graph into the corresponding information propagation



model that the computer simulator can understand, as shown in Figure 11.

V. IMPLEMENTATION AND PROJECT SUCCESS

The primary method for building the simulator is the gradual complexity increase approach. In this manner three different data sets of increasing complexity were used for the creation of initial results. The first and simplest set is a parallelogram shaped target that is comprised of point scatterers all of unity amplitude and zero phase. The resulting output from the signal processor block is presented in Figure 12(a) and 12(b). The second model involved the same amplitude-phase pairs but now with added phase delay effects caused by the physical distance between the radar and the target. The resulting output from the signal processor block is presented in Figure 12(c) and 12(d). The final initial set is diverse in the amplitude-phase pairs. Now these pairs have values that intuitively correspond to an actual ship’s reflectance properties. The resulting output from the signal processor

block looks like usual Slant Range Profiles and corresponding ISAR images for a ship vessel. Analytically the final result of a reflectivity solution that resembles a single-layered vessel for test purposes is presented in Figure 13(a) for the slant range profile and Figure 13(c) for its corresponding ISAR image. For the solution to have more credibility we have the enhanced version of the multiple layered ship which is presented in Figure 13(b) for the slant range profile and Figure 13(d) for its corresponding ISAR image.

A. Verification Analysis

Verification is a quality process used to evaluate whether or not a system complies with the initially stated algorithmic specifications, regulations and conditions. In other words it is the process that establishes the mathematical agreement of the final product to the physical phenomenon that it has been called upon to represent.

0 5 10 15 20 25 30 351

1.5

2

2.5

3

3.5

4

4.5

51 TEST MODEL - Slant Range Profiles Antenna 1 - 32 Pulses

Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35

0

20

40

60

80

100

120

140

1601 TEST MODEL - ISAR Image Antenna 1 - 32 Pulses

Resolution Cell ID

Dop

pler

0 5 10 15 20 25 30 35

0.5

1

1.5

2

2.5

3

3.5

4

4.5


Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35

-20

0

20

40

60

80

100

120

140


Resolution Cell ID

Dop

pler

(a) (b) © (d)

Figure 12. Uniform amplitude/phase model Model 1 (a) Slant Range Profile (b) ISAR imageModel 2 (c) Slant Range Profile (d) ISAR image

0 5 10 15 20 25 30 350

0.5

1

1.5

2

2.5x 10

4 3 MORE MODEL C - SRP Ant 1 (32)

Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35

0

0.5

1

1.5

2

2.5x 10

4 E1 MULTILAYER MODEL - SRP Ant 1 (32)

Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35-1

0

1

2

3

4

5

6

7x 10

5 3 MORE MODEL C - ISAR Ant 1 (32)

Resolution Cell ID

Dop

pler

0 5 10 15 20 25 30 35-1

0

1

2

3

4

5

6

7x 10

5 E1 MULTILAYER MODEL - ISAR Ant 1 (32)

Resolution Cell ID

Dop

pler

(a) (b) © (d)

Figure 13. Diverse amplitude/phase single layer model with added phase delay effects Single Layer (a) Slant Range Profile (b) ISAR image Multiple Layer (c) Slant Range Profile (d) ISAR image



The method to verify the model involved the diversity

in increasing complexity of the major mathematical expressions within each component [10]. The expressions that were refined towards better resolution were the range finder equation, the polarization properties, the signal attributes and the environmental response of transmit receive block and the digital signal processing block algorithm. Therefore verification for this project is done at component level.

B. Validation Analysis

Validation is the process of establishing documented evidence that provides a high degree of assurance that a system accomplishes its intended requirements. The

process involves accreditation actions like acceptance and suitability certificates that the system model is compatible with the physics aspects that was designed to implement. Validation is an intuitive process that involves Subject Matter Engineers and external customers. The organization model should be validated and tested against known or physically expected results in order to ascertain its authority. The validation loop that was followed is shown in Figure 14 [31]. After the loop we obtained results as shown in Figure 15 that follow the simulation complexity history. These results show the output of the three systems in a sequence and help to debug any validation issues. Therefore we conducted the validation for this project at system level.

Setup Target ReceiveSurfaces that are

uniform in amplitudeand phase (Uniform

RCS)

Disregard all otherpossible amplitude

attenuation and phasechanges due to thedistance betweenRadar and Target

Hit Target ReceiveSurfaces with theRadar Transmit

Surface

Move the data receivedfrom the Target ReflectSurfaces to the Signal

Processing Blocks.

Results should showstrong signs of validity.

Complicate theTransmit, Reflect andall Receive Surfaces

one step at a time

Complicate theEnvironment interaction

one step at a time.

Figure 14. Validation procedure

0 5 10 15 20 25 30 351

1.5

2

2.5

3

3.5

4

4.5


Resolution Cell ID

Reflectance Value

0 5 10 15 20 25 30 350

20

40

60

80

100

120

140


Resolution Cell ID

Doppler

0 5 10 15 20 25 30 350.5

1

1.5

2

2.5

3

3.5

4

4.5


Resolution Cell ID

Reflectance Value

0 5 10 15 20 25 30 35-20

0

20

40

60

80

100

120

140


Resolution Cell ID

Doppler

0 5 10 15 20 25 30 350

0.5

1

1.5

2

2.5x 10


Resolution Cell ID

Reflectance Value

0 5 10 15 20 25 30 35-1

0

1

2

3

4

5

6

7x 10


Resolution Cell ID

Doppler

Figure 15. Simulation complexity history



The project can be extended by adding a second antenna at a different depression angle in order to provide three-dimensional imaging. The results are shown in Figure 16 and reveal another fact that enhances the credibility of the simulator. High intensity points in lower superstructure coordinates appear as having greater height than in reality and that the same target viewed from two antennas separated by any distance produces a different Range-Doppler signature [11][12][13]. The latter fact is normal and forms the basis of the Interferometry technique where two R-D images are combined in order to provide a three-dimensional representation of the target [28].

Moreover another validation milestone has to be confirmed. We added translation effects in our pace engine and have found, as shown in Figure 17, that only roll and pitch motions contribute to the final outcome of ISAR imaging. Translational motion does not alter the result so the solution already contains global angle alignment for the range-bin alignment stage [34] and minimum entropy methodology for the phase adjustment stage [35].

VI. DISCUSSION

An ISAR simulator system and its corresponding organizational details like its components, pace engine and interfaces for an extended target has been described. Verification was done at component level whereas validation was performed at system level. Gradual model complexity strengthens both of the above testing procedures. We argue that data created by a simulator can be used to build upon a new system design without the need to utilize real experimental data as long as certain milestones are observed. In this case we have made certain that the input phase follows correct practices, the transfer function phase conforms to theoretical and

practical guidelines and the output phase conforms to valid results.

Analytically the verification part is conducted by explaining the datatom of the simulator. And the validation part is performed by explaining the corresponding informatoms of the simulator. The datatom can be defined as the simplest configuration that can provide satisfactory verification in a simulation process. And the informatom can be defined as the simplest result or group of results that can provide satisfactory validation in a simulation process [13]. Furthermore the tests keep track of all the prior understanding of how the software implemented the target scientific theory. Datatoms are created by process tests which ensure that each process is reflected in the code and informatoms are created by canonical results which preserve the empirical support of the theory [17].

Looking at all the results that the simulator has produced we can see that the theoretical aspects of military ship imaging is certified. Namely :

Only roll and pitch motions contribute to the final outcome of ISAR imaging. Translational motion does not alter the result.

High reflectivity in lower superstructure coordinates produces high Doppler in the ISAR image as expected. This function can be used to produce an image that more resembles a military target.

The superstructure of the target is made of different layers that each contribute to the final ISAR image.

RCS solution is dependent on aspect and depression angle configurations. For the same aspect angle the result must be different when the target is viewed by antennas mounted at different positions and therefore different depression angles, as in real situations.

The computing force that moves the false target is based on elements from the virtual reality and computer game industries.

0 5 10 15 20 25 30 350

0.5

1

1.5

2

2.5x 10


Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35

-1

0

1

2

3

4

5

6

7x 10


Resolution Cell ID

Dop

pler

0 5 10 15 20 25 30 350

0.5

1

1.5

2

2.5x 10


Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35

-1

0

1

2

3

4

5

6

7x 10


Resolution Cell ID

Dop

pler

Figure 16. ISAR Shortcomings Effect



0 5 10 15 20 25 30 350

0.5

1

1.5

2

2.5x 10

4 3 MORE MODEL C [15 Knots] - SRP Ant 1 (32)

Resolution Cell ID

Ref

lect

ance

Val

ue

0 5 10 15 20 25 30 35-1

0

1

2

3

4

5

6

7x 10

5 3 MORE MODEL C [15 Knots] - ISAR Ant 1 (32)

Resolution Cell ID

Dop

pler

Figure 17. Translation Impact (should be minimum)

II. CONCLUSIONS

The method of generating ISAR images for the purpose of generating false targets for electronic warfare purposes is described in this work which addresses both the simulation engineering and the radar engineering communities. The overall implementation is constructed like a computer game. The threat radar and the friendly assets are moved while interacting into a three dimensional environment. Simply the target returns are carefully crafted to realistic levels by considering ISAR imaging considerations. This approach allows the use of simulators as active elements in electronic warfare applications. The next step in this work would be to introduce more phase error into the generated ISAR return signal because the enquiring sensor is expecting such responses, like target angular glint effects.

ACKNOWLEDGMENTS

This project was initially supported by the Engineering & Physical Sciences Council (EPSRC), United Kingdom. For this occasion we would like to thank Prof. Griffiths of Defence Academy of the United Kingdom and Prof. Baker of University College London for valuable guidance and support. Also we express our gratitude to Mrs M Diakaki for helping with the challenging data processing requirements of the necessary many Bismarck models. Further the project was supported by a grant by the University of the Aegean. This work forms part of the first author’s PhD thesis in simulator systems design for Synthetic Aperture and Inverse Synthetic Aperture Radar systems with Prof. Katsikas as main supervisor.

REFERENCES

[1] Abrash M., 1997, Graphics Programming Black Book, Coriolis

Group, ISBN 1-576-10174-6. [2] Boccara N., 1985, Modelling complex systems, Springer Verlag,

ISBN 9780387158853. [3] Carrara W. G., Goodman R. S., Majewski R. M., 1995, Spotlight

Synthetic Aperture Radar, Artech House, ISBN 0-89006-728-7. [4] Dunleavy P., 2003, Authoring a PhD: How to plan, draft, write and

finish a doctoral thesis or dissertation, Palgrave MacMillan, ISBN 1-4039-1191-6.

[5] Emir E., Topuz E., 1997, Simulation of ISAR images of ships for localization of dominant scatterers, In Radar 97 (Conf. Publ. No. 449) 14-16 Oct 1997 Page(s):273 – 275.

[6] Fouts et. Al, 2002, A Single-Chip False Target Radar Image Generator for Countering Wideband Imaging Radars, IEEE Journal of Solid-State Circuits, Vol. 37, No. 6, June 2002.

[7] Given J.A. Schmidt W.R., 2005, Generalized ISAR - Part I: an optimal method for imaging large naval vessels, In IEEE Transactions on Image Processing, Publication Date: Nov. 2005, Volume: 14, Issue: 11, On page(s): 1783-1791, ISSN: 1057-7149, INSPEC Accession Number: 8622151, Digital Object Identifier: 10.1109/TIP.2005.857283

[8] Hajduch G., Le Gaillec J.M., Garello R., Airborne High Resolution ISAR Imaging of ship targets at sea, IEEE Transactions on Aerospace and Electronic Systems, 2004, 40, (1), pp. 378-384.

[9] Haywood B., Kyprianou R., Zyweck A., 1994, ISARLAB: a radar signal processing tool, In IEEE International Conference on Acoustics, Speech & Signal processing, Vol. 5.

[10] Kleb B., 2007, Toward Scientific Numerical Modeling, NATO RTO AVT-147 Symposium on Computational Uncertainty in Military Vehicle Design, Athens, Greece.

[11] Kostis T. G., Baker C.J., Griffiths H.D., 2006, An Interferometric ISAR System Model for Automatic Target Identification, EUSAR 2006, Dresden, Germany.

[12] Kostis T. G., Katsikas S.K., 2007, Three-Dimensional Multiple Layer Extended Target Modeling for ISAR Studies in Target Identification, Panhellenic Conference on Informatics, Patras, Greece.

[13] Kostis T. G., 2008, Simulator Implementation of an Inverse Synthetic Aperture Radar System for an Extended Naval Target in a Three Dimensional Synthetic Environment, UKSIM 2008, pp.366-371, Tenth International Conference on Computer Modeling and Simulation, 2008.

[14] Kostis T. G., 2008, Glint Effects Simulation for an Extended Naval Target using an Interferometric-ISAR System Model, In European



Synthetic Aperture Radar Conference (EUSAR 2008), Friedrichschafen, Germany.

[15] Kostis T. G., Galanis K. G., Katsikas S. K., 2008, Simulator Implementation of an IF-ISAR System for Studies in Target Glint, In Panhellenic Conference on Informatics, pp.140-144, 2008.

[16] Kostis T. G., 2008, Proof of Concept for the Extensibility Attribute of an ISAR Simulator for Studies in Target Glint, Imaging Systems & Techniques, Chania, Greece.

[17] Lane P. C. R. and Gobet F., 2008, A Methodology for Developing Computational Implementations of Scientific Theories, EUROSIM / UKSIM08, Cambridge, England, 2008.

[18] Li C., Zhu D. 2008, The Detection of Deception Jamming against SAR based on Dual-Aperture Antenna Cross-Track Interferometry, IEEXplore.

[19] Ling W., Daiyin Z., Zhaoda Z., 2008, Image-based scaling for ship top view ISAR images, Journal of Electronics (China) Publisher Science Press, co-published with Springer-Verlag GmbH ISSN0217-9822 (Print) 1993-0615 (Online) Issue, Volume 25, Number 1 / January, 2008 DOI10.1007/s11767-006-0071-z Pages 76-83

[20] Lord R. T., Willie N., Gaffar M. Y. A., 2006, Investigation of 3-D RCS Image Formation of Ships Using ISAR., In European Synthetic Aperture Radar Conference, (EUSAR 2006).

[21] Lynch D., 2004, Introduction to RF Stealth, Sci-tech Publishing, ISBN 9781891121210.

[22] Neugebauer E., Steinkamp D., 2007, Representation, pp. 2-1~2-4, NATO Modeling and Simulation Group, RTO MSG-067 Lecture Series, Athens, Greece.

[23] OTA-BP-ISS-136, 1994, U.S. Congress, Office of technology Assessment, Virtual Reality and Technologies for Combat Simulation – Background Paper.

[24] Pastina D., Spina C., 2008, Slope-based frame selection and scaling techniques for ship ISAR imaging, IET Signal Processing, 2, (3), pp.265-276.

[25] Porter, N.J. Tough, R.J.A., 1994, Processing schemes for hybrid SAR/ISAR imagery of ships, In IEE Colloquium on Radar and Microwave Imaging, Nov 1994 Page(s):5/1 - 5/5.

[26] Rice F., Cooke T., Gibbins D., 2006, Model based ISAR ship classification, Elsevier, Digital Signal Processing, Volume 16, Issue 5, September 2006, Pages 628-637, DASP 2005.

[27] Schleher D. C., 1986, Electronic Warfare in the Information Age, Artech House, ISBN 0-89006-526-8.

[28] Seybold, J. S.; Bishop, S. J., 1996, Three-dimensional ISAR imaging using a conventional high-range resolution radar, In Proceedings of the 1996 IEEE National Radar Conference, Issue ,Page(s):309 – 314, DOI 10.1109/NRC.1996.510699

[29] Shillington, K.R., Jahans, P.A. Buller, E.H. Tunaley, J.K.E., 1991, An ISAR simulator for ships, In Antennas and Propagation Society International Symposium, 1991, Page(s):1032 - 1035 vol.2.

[30] Son Sok J, Flores B. C., Gabriel T., 2002, Range Doppler Imaging and Motion Compensation, Artech House.

[31] Sornette D., Davis A., Ide K., Kamm J.R., 2007, Theory and Examples of a New Approach to Constructive Model Validation, NATO RTO AVT-147 Symposium on Computational Uncertainty in Military Vehicle Design, Athens, Greece.

[32] Stimson G., 1998, Introduction to Airborne Radar, Sci-Tech Publishing ISBN 1-891121-01-4.

[33] Wang J., Kasilingham D., 2003, Global range alignment for ISAR, IEEE Transactions on Aerospace and Electronic Systems, 39, (1), pp. 351-357.

[34] Wang J., Liu X., Zhou F., 2004, Minimum entropy phase adjustment for ISAR, IEE Proceedings on Radar Sonar & Navigation, 151, (4), pp.203-209.

[35] Wehner D. R., 1998, High Resolution Radar, Artech House, ISBN 0-89006-727-9, 1994.

Theodoros G. Kostis is a Telecommunications Engineer at the Department of Information Systems and Network Infrastructure, University of the Aegean, Network Operations Center in Athens, Greece. He is a Chartered Electrical Engineer both in the United Kingdom as appointed by the Engineering Council United Kingdom (ECUK) and in Greece as appointed by the Technical Chamber of Greece (TEE-TGC). He has earned a BEng (Hons) degree in Telecommunications and Computer Networks Engineering from London South Bank University, UK and a Master’s degree of Science in Microwaves and Optoelectronics from University College London, UK. He is currently a PhD candidate at the University of the Aegean, Department of Information and Communication Systems in the field of conceptual and applied modelling for novel high range resolution radar systems.

Sokratis K. Katsikas was born in Athens, Greece, in 1960. He received the Diploma in Electrical Engineering from the University of Patras, Patras, Greece in 1982, the Master of Science in Electrical & Computer Engineering degree from the University of Massachusetts at Amherst, Amherst, USA, in 1984 and the Ph.D. in Computer Engineering & Informatics from the University of Patras, Patras, Greece in 1987. Currently he is a Professor with the Dept. of Technology Education and Digital Systems of the University of Piraeus, Greece and a member of the Board of the Hellenic Authority for Information and Communication Security and Privacy. From 1990 to 2007 he was with the University of the Aegean, Greece, where he served as Rector, Vice-rector, Department head, Professor of the Department of Information & Communication Systems Engineering and Director of the Information & Communication Systems Security Lab. His research interests lie in the areas of information and communication systems security and of estimation theory and its applications. He has authored or co-authored more than 150 journal publications, book chapters and conference proceedings publications in these areas. He is serving on the editorial board of several scientific journals, he has authored/edited 20 books and has served on/chaired the technical programme committee of numerous international conferences.



Enlarged Versions of Detailed Figures

Figure 1. FB-16 Simulation geometry draft

Figure 4. Synthetic Environment Modelling



TargetPlatform

RadarPlatform

Transmitter Typeie Tx Parameters

Waveform Selection

Antenna Typeie phased array

parabolic

Receiver Typeie Rx Parameters

InformationExtraction from

Signals Received

Propagation of Radar Waves Modelingie. Parabolic - Ray Tracing

Hybrid - Custom - Custom Hybrid[Refraction - Diffraction]

correlationconvolutionprocedures

Glint

Ground Clutterand other clutter types

Target Typeie. rcs

3D EngineSystem Dynamics

DirectionalInterference

ie. JammerChaff

ElectromagneticInterference

Dynamic Errorsie. Platform MotionScanning Motion

Vibration

MotionProvider

MotionProvider

All Other Noiseie. equipment noise

General Simulation Parametersie Code Specific Considerations

Sampling Rate

MatchedFilter

Results Formation

Figure 2. Theoretical Inverse Synthetic Aperture Radar process (b) Complete Model



Figure 2. Theoretical Inverse Synthetic Aperture Radar process

(c) ISAR Subset



(a)

Figure 5. Extended target modelling (a) Single Layer Model (b) Multiple Layer Model



(b)

Figure 5. Extended target modelling

(a) Single Layer Model (b) Multiple Layer Model

inverse synthetic aperture radar simulator implementation ... · the theoretical representation of...

Documents