EWEF Electronic Warfare Engagement Framework

1
EWEFElectronic Warfare Engagement Framework

• Presented By
• Alan Smith (DERA)
• tel [phone]
• email ssew1rg4_at_dera.gov.uk
• Duncan Clark (IBM)
• tel [phone]
• email Duncan_Clark_at_uk.ibm.com

2
Agenda

• EWEF Overview
• EWEF Architecture
• Modelling Aspects
• Implementation Aspects
• Performance Results
• Conclusions

3
EWEF Overview - Function

• Electronic Warfare Engagement Framework
• Evaluation of EW Techniques and Tactics
• Integrated System Response requiring
• Threat / ASW Missile representation (RF and IR)
• Sensor - ESM / Radar / Signature
• Countermeasures (Techniques) ECM, Decoys, Chaff
• EW Command and Control (tactics)

4
(No Transcript)
5
Typical EWEF Applications
EW Picture / Network / Co-operative Triangulation
Counter Targetting
CNGF(2)
Chaff
Soft-kill
Co-ordination
Multiple Missiles
CVS - Self Defence
Active On-Board
CNGF(1)
Active Decoy - DLH
Multiple Waves of Missiles
Passive Decoy DLF
6
Study Objectives

• Evaluate the utility of HLA in EWEF for
• Improving Performance by distribution
• Interoperating with other Simulations (e.g. Hard
  Kill models)
• Interfacing with Stealth Viewers and other
  Visualisation Tools
• Interoperating with other Naval Technology
  Demonstrators
• Particular Issues to be addressed
• Time management Options
• Scenario Management and Distribution Control
• Data Interface standards including DIS / RPR FOM
• Performance

7
EWEF Overview - Implementation

• Object Oriented Design using OMT
• Implemented in C
• Running on PCs under Windows NT
• Standalone Workstation Design
• Time ordered Event based Simulation
• Based on DERA Simulation Framework (DROMAS)
• Enhanced Framework for EW modelling

8
Simulation Layers
Model

• ESM
• ECM

MDL
Model Framework

• Position
• Messaging
• Interactions

MF
Simulation Framework
SF

• Base Classes
• Scheduler
• Streaming

9
EWEF Component Architecture
Component(Platform)
10
EWEF Platform Architecture
Motion Model
Platform 1
ECM
Communications Highway
ESM
CMS
- Illumination Port
Communications Highway acts as a network allowing
message routing between equipment models on the
platform
EWC2
- Emission Port
Radar
- Position Port
- Message Port
11
EWEF Propagation Architecture
Propagation Model applies the propagation algorith
ms to emissions to create illuminations ready
for equipment model to process.
RF Propagation
Range
Platform 1
Platform 2
Path 1
Range
RF Propagation
- Illumination Port
- Position Port
- Emission Port
12
Modelling Aspects

• Scenario Object Hierarchy
• Position Updates
• Propagation Modelling
• Electromagnetic Emissions Representation
• Messaging
• C2 Modelling

13
Scenario Platform Definition

• DEFINITION EWEF example HLA representation
• Interfaces Platform PhysicalEntity
• Code Class Ship Platform B MilitarySeaSurfacePlat
  form
• Data Class Type42.gen Entity.EntityType
• Identity Scenario.scn Entity.Marking
• Four levels for defining an object in EWEF

14
FOM Object Representation

• EWEF Object RPR FOM Class (Derived from)
• Ship MilitarySeaSurfacePlatform
• Missile MunitionEntity
• Chaff / Decoy MilitaryAirLandPlatform
• Subsystem EmbeddedSystem
• Radar/Emitter EmitterSystem
• Message Link EWEFLink (EmbeddedSystem)
• Emission EWEFEmission (EmitterBeam)

15
Position Modelling and Updates

• Uses Dead Reckoning and RPR FOM fields
• Position, Velocity, Accn, Orientation, AngVel
• Flat earth geometry (X North, Y East, Z down)
• Filtering conversion options for
  interoperability

Position Changes If change
is significant Send UpdateAttributeValues

ReflectAttributeValues received Synchronise
Time Position Changes
Position Port
Filter Convert
Entity Attribute
Synch Convert
Position Port
16
Propagation Modelling with HLA
Emission
RF Propagation
Federate1
Range
Entity
Platform 2
Path 1
Entity
Platform 1
Path 1
Federate 2
Range
Emission
RF Propagation
- Position Port
- Emission Port
- Illumination Port
17
Electromagnetic Emissions
ERP ERP Dot
Elevation Elevation Dot Azimuth Azimuth Dot
Frequent Updates
BeamShape Polarisation
InFrequent Updates
Frequency
Time (within PRI)
Frequency PulseWidth PRI for each pulse in PRI
sequence
Delay Delay Dot
Time (Pulse to Pulse)
18
Messaging - Connection
19
Messaging - Sending messages
Message Link
Message Link
ESM
EWC2
20
C2 Modelling
21
C2 Modelling 1
22
C2 Modelling 2
23
Implementation Aspects

• Object Distribution Management
• Object Interaction Management
• Attribute Updates Position,Emissions
• Interactions Link connectivity,Messages
• Time management
• Interoperability with other systems

24
Object Distribution Management

• How to control which Federate models what
• Object Reflection Options
• Reflect all Objects of Subscribed FOM Class
• Reflect all Objects of specified EntityTypes
• Reflect identified Objects with specified marking
• EWEF Solution
• Each federate scenario identifies reflected
  platforms
• Each federate scenario identifies reflected
  EntityTypes for dynamically created platforms
  (e.g. chaff and decoys)

25
Examples of Scenario Distribution
Federate 1 Scenario Represented
Federate 2 Scenario Represented
26
Object Interaction Implementation

• EWEF Interaction Ports minimises intrusion of HLA
  into model
• Dead Reckoning used to minimise attribute update
  frequency
• Need to separate the dead reckoning reference
  time from the time management time.
• Attributes separated into groups that get updated
  together
• eg Emission - Az, El, Delay, ERP change with
  Geometry change Waveform, Beamshape
  change with Mode change
• Conversion between EWEF and FOM representation
  undertaken by interface class NOT an EWEF class.
  (I.e. mirrored objects)
• Interactions only used for message connectivity
  and sending

27
Time Management (1)

• Next Event Request
• Suited to event based simulation - NER (Time of
  next EWEF event)
• Performance inadequate due to
• Extensive RTI network traffic to calculate LBTS /
  next event
• Very fine granularity of EWEF events
• Time Advance Request
• Adopted a timestepping approach, caching all
  changes / events up to Timestep Boundaries
• Performance was variable
• limited by timestep (small timestep implies slow
  but better accuracy)
• Scenarios have intervals with little happening
  (tried to NER for events gt timestep)

28
Time Management (2)

• Lookahead
• Achieving any significant lookahead was difficult
• Timestep approach allowed a lookahead equal to
  timestep
• Performance improved significantly if a lookahead
  equivalent to 2 timestep was allowed
• Delivery Order
• Time Stamped Order Delivery used initially to
  maintain causality
• TSO delivery with Timestep approach forced
  delivery latency to no advantage
• RO delivery allowed internal assessment of
• Relative Federate Times (for possible load
  balancing)
• Significant Breakdowns in causality

29
Time Management Summary

• Separate the fine grain internal events from
  external events
• Use a timestepped approach for external events
• Use the timestep to control the granularity of
  the simulation and hence the lookahead and
  performance.
• Use a TAR system to keep federate time
  synchronised to within the timestep.
• The TAR system also allows faster than real time
  to be constrained to work with real time
  simulations
• Ensure that the synchronisation latency executes
  in parallel with the simulation execution. (e.g.
  At T9 . Federate TARs to 10 and immediately
  starts processing time up to 10 seconds while the
  TAG to 10 is being sought.)

30
Interoperability with Stealth Viewer - SIMDIS

• Four options for interoperating investigated
• HLA Federate logging positions and beams to file,
  played back through SIMDIS
• HLA Federate sending positions and beams over
  proprietry multicast interface
• EWEF conversion to DIS and hence interfaced to
  SIMDIS - Not persued due to
• The need for additional DIS software
• The lack of representation in DIS for radar beams
  and gates
• Use of HLA to DIS Gateway and hence interface to
  SIMDIS - Not persued due to
• The need for another set of RTI s/w compatible
  (SGI, Sun, PC)
• The need to change the FOM to include the EWEF
  aspects

31
Performance Results

• Configuration
• Based on RTI 1.0-3
• RTI Fedex on Sun Sparc ,
• Each Federate on a different PC
• Investigations of
• Distribution Overhead
• Scenario Distribution Effects
• Time Management Systems
• Causality and Resolution
• Load Balancing

32
Performance Assessment Scenario
Propagation
Seekers Activate at 40 seconds Missiles fly past
at 80 seconds approx Scenario 100 seconds duration
RCS ESM C2
Radar
RCS only
Active Seeker Guidance
33
Distribution Overhead
34
Scenario Distribution Effects
35
Time Management Effects
36
Causality and Resolution

• Dead Reckoning reduces need for causal events
• Applied to position and E/M signals
• Timestepping introduces granularity errors
• Can be minimised by selecting timestep
• Need to ensure that federate times do not differ
  too greatly
• May need to change timestep to cope with
  different effects (e.g. Scan)
• Interactions and Messages should be causal
• TSO delivery always introduces latency
• Found ourselves rejecting the HLA causal system
  in favour of fast as possible delivery, with
  resolution defined by timestep.

37
Load Balancing
Average Received Event Time wrt Federate local
Time -0.25 secs - 0.2 secs
Received Event Count (0-14)
38
Conclusions

• HLA provides a useful means for
• Performance Improvement for Analytical Models
• Interoperating with Other Systems
• EW Modelling can benefit from DIS / HLA
  techniques
• Dead Reckoning
• Existing Data Standards
• Scenario Management needs careful
• Configuration Control
• Performance Optimisation