JDEP as a Case Study for Developing Distributed HLA Simulations

1
JDEP as a Case Study for Developing Distributed
HLA Simulations

• 25 April 2005
• Jackson Ludwig

2
Overview

• Introduction to Distributed Simulation
• Overview of HLA
• Problem Study JDEP SIAP
• Goal of Simulation
• Simulation Architecture
• FOM Federation Agreements
• Execution
• Conclusions

3
Distributed Simulation

• Reasons to do distributed simulation
• Need to link simulations belonging to others
• Need to utilize multiple computers for parallel
  processing
• Want to model a function, but lack expertise
• Want to standardize a model or promote reuse
• Distributed simulations can be spread across
  several computers at one site or computers at
  geographically dispersed sites
• Technologies to support distributed simulation
• Distributed Interactive Simulation (DIS)
• Aggregate Level Simulation Protocol (ALSP)
• Test Training Enabling Architecture (TENA)
• High Level Architecture (HLA)
• Xtensible Modeling Simulation Framework (XMSF)
• Others

4
Distributed Simulation

• Advantages
• Allows simulationist to model functions outside
  area of expertise
• Gain speed through parallel processing
• Parallel development efforts
• Ability to link models that cant be relocated
  elsewhere
• Risk mitigation (at least some parts will
  probably work)
• Disadvantages
• Networking issues
• Data representations formatting
• Models may require extensive rework to
  interoperate
• Misinterpretations during distributed development
• Can be slower than single models

5
High Level Architecture

• DoD and IEEE standard for distributed simulation
• Models are referred to as federates, the
  simulation with all the models is called a
  federation
• Run Time Infrastructure (RTI)
• Cross platform middleware application
• Regulates data exchanges object ownership
• Manages time advancement
• Time stepped
• Event stepped
• Regulated/unregulated time
• Monitors state of the federation
• Federation Object Model (FOM)
• Describes contents of federation
• Object attribute names, definitions,
  interaction parameter definitions, data type
  definitions

6
HLA BasicsObjects, Interactions, Time

• HLA takes an object-oriented view of simulation
• Objects
• Used to represent physical entities
• Attributes describe current state of the object
• Attribute inheritance from above
• Only owner of object can update attributes
• Interactions
• Used to represent transient things (data)
• Parameters convey the information in the
  interaction
• Parameter inheritance from above
• Federates subscribe to the interactions of
  interest
• Time
• In unregulated time each federate advances time
  at its own pace
• In regulated time federates may choose their own
  time steps and request their desired next time,
  but the RTI then advances time to the minimum
  next time requested by a federate
• Federates process updates FIFO for unregulated
  time and time stamp order for regulated time
• In regulated time, federation speed is determined
  by the slowest federate

7
Case Study ProblemSingle Integrated Air Picture
(SIAP)

• Develop a common air picturein the face of
• Multiple platforms each with a mission
  computer, operating in different domains (E-2C,
  AWACS, Aegis, TPS-59, Patriot, etc)
• Radar, ESM on each platform providing unique
  data to its mission computer
• IFF on each platform providing unique
  identification information to its mission
  computer
• Navigation system on each platform deriving its
  own view of location
• Communications disparate (Link-16, Link-11)
  without total interconnectivity

AWACS Airborne Warning and Control System ESM
Electronic support measures IFF Identification
Friend or Foe
8
Simulation ConceptJoint Distributed Engineering
Plant (JDEP)

• Government sponsor developing cross-platform
  software module to improve the air picture
• Goal Create a simulated environment that
  exercises and evaluates the System Under Test
  (SUT)
• Realistic data feeds
• Realistic data timing
• Predictable behavior
• Repeatable behavior
• Allows interoperability testingto be conducted
  at same time asdevelopment
• Intent to create a set ofmodels that can be
  reusedby others

9
Federation Development

• Federation development concurrent with real
  systemdevelopment
• Federation to be located at one site
• Worked with sponsor to select the models needed
  and general requirements
• Using conservative time stepped federation
• Most models to be used are new one legacy model
  will be included after heavy modification
• Need high fidelity data formatting actual data
  can be of lower fidelity
• Once the federation composition decided, then
  developed FOM and Federation Agreement document

10
JDEP FOM
11
Sample JDEP FOM Entries

• Object
• Platform EulerAngles, FederateID, InstanceID,
  LinearAcceleration, LinearVelocity, Orientation,
  PlatformType, SequenceNumber, Side, TimeStamp,
  WorldLocation
• Aircraft TailNumber, AircraftType
• Interaction
• SimulationService TimeStamp, FederateID,
  PlatformType, InteractionCounter
• RealWorldMessage
• ESMMeasurement AzimuthAccuracy, EmitterType,
  ESMNumber, MeasurementAzimuth, MeasurementTime
• Some attributes/parameters are complex or
  enumerated datatypes

12
Federation Agreements

• FOM defines how models exchange data but does not
  describe the specifics and formatting of the data
• Explains process for creating the FOM and making
  future changes
• Describes the federation structure, goals and
  objectives, roles of various models
• Defines naming conventions, data representations
  (big/little endian, etc), coordinate systems,
  terrain models
• States version of RTI to be used, time management
  scheme, data distribution needs
• Various other responsibilities for models in
  federation

13
Simulation Test Harness Federation JDEP
Infrastructure Build (IBuild)

• Represents onboard radar system (E2C, AWACS, etc._

Sensor

• Represents Interrogation Friend or Foe (IFF)

IFF
Mission Computer Program (MCP)

• Represents SIAPs Integrated Architecture
  Behavior Model

• Represents onboard navigation (accelerometer,
  ring-laser gyro, etc.) .

NAV
HLA RTI

• Communications Simulations, primarily Link 16

Communications
Communications

• Generates remote tracks on other host platforms

Remote Host (MCP) Emulator

• Scenario Driver Represents flight paths of all
  aircraft in scenario

Truth Driver
Test Utilities
3D Viewer

• Test Utilities (Control and monitoring of
  federation, data analysis)

HLA Results
HLA High Level Architecture RTI Run-Time
Infrastructure IABM Integrated Architecture
Behavior Model MCP Mission Computer
Program CRSD Common Reference Scenario Driver
14
JDEP Federation Execution
Mission Computer Program (MCP)
Truth Driver
NAV
IFF
Radar
Communications

Send ObjectGround Truth
Receive Comms Msgs
Send Nav PerceivedPosition
Send Radar Detections
Send IFF Detections
Send Comms Msgs
RTI
15
Data Distribution Management

• Every ship, aircraft, ground station has a
  mission computer, navigation, radar or other
  sensor, IFF, and communications
• With many platforms in the federation, a lot of
  data gets transferred between models
• Data distribution management increases efficiency
  by filtering messages so models only get messages
  meant for them
• Works by defining regions and models only receive
  updates and interactions for its region
• In the JDEP federation, all modelsthat are on
  the same platformhave the same region

Each Platform Is a Region
16
Conclusion

• Distributed simulations allows multiple models to
  be combined in a single simulation
• HLA provides a mechanism for linking models to
  form distributed simulations
• Having a good Federation Agreements document is
  as important as a well defined FOM
• HLA will not solve all integration issues
• Traditional simulation issues such as fidelity
  and speed concerns will still exist