Phrenology Forum

1
Synthetic Natural Environment (SNE) Conceptual
Reference Model
Dr. Paul A. Birkel The MITRE Corporation pbirkel_at_
mitre.org
2
Background

• HLA defines strong framework for establishing
  data (or syntactic) interoperability
• Relatively weak on how federates should leverage
  framework to achieve semantic interoperability
• Supports negotiation process for mutually
  agreeing on FOM
• Provides interface to common data transport
  mechanism
• Deemed necessary and sufficient to meet
  interoperability goals of military system
  modelers ...
• Once a FOM has been negotiated and federates have
  been upgraded to communicate data in terms of
  that FOM.
• Simulated military systems, alone, have limited
  value
• SNE is context within which military system
  models operate
• Supports basic state Where are you?
• Mediates interactions Can I see you?

3
SNE What is it?
The representation of the physical world within
which all models of military systems exist and
interact. It includes both data and models
representing the elements of the environment and
their effects on military systems, and models of
the impact of military systems on environmental
variables (e.g. contrails, dust clouds from
moving vehicles, spoil from combat engineering).
4
Authoritative Representation

• MS Master Plan defines Executive Agents
• Terrain (TMPO), Ocean (OEA), Air Space (AS EA)
• Focus primarily on data requirements, production,
  distribution
• Limited (unsuccessful) forays into run-time
  modeling
• SEDRIS standard SNE data model and data
  dictionary
• Data distribution solution pre-exercise
• Provides both semantic objects and data transport
• DIS/SIW have tackled run-time issues under guise
  of the environmental architecture concluding
  either
• Environmental modeling has demands incommensurate
  to those of military system modeling (e.g.,
  building / tank), or
• A literal reuse of existing approaches to
  distributed military system modeling can be
  applied to distributed environmental modeling
  (e.g., cloud tank)

5
Missing?

• A standard system (or process) model for use as a
  reference against which to compare alternative
  environmental run-time system descriptions,
  components, decompositions, and data flows.
• A proposal follows ...
• Not a specific environmental architecture
• Whether youre a federation, or federate,
  designer
• A conceptual reference model framework within
  which alternative environmental architectures and
  designs can be
• Dissected, described, compared, and contrasted
  in order to
• Progress towards the complex goal of SNE
  interoperability
• One design does not fit all
• We propose that one conceptual framework can
  fit all

6
Overview

• 1. Present the SNE conceptual reference model,
• 2. Explore its structure using functional and
  interface examples,
• 3. Describe how several contemporary MS problems
  map to that model, and
• 4. Postulate how use of this model could help the
  SNE MS community better focus their efforts
  towards
• Achieving effective reuse of data and software,
• Application/federate interoperability, and
• Cost/effective future system development.

7
Conceptual Reference Model Schema

• The physical manifestations of military systems
• Is the vehicle structure part of the
  Environmental Ground Truth?
• YES
• Provides measurable attributes/values describing
  the physical environment
• E.g., responds to illumination (whether solar or
  target illuminator)
• E.g., interacts with other structures according
  to Newtonian laws
• It is a physical, space-occupying mass

8
Schema Elaboration

• Component Models
• The collection of algorithms used to model
  military equipment (men and material).
• These are generally the space-occupying physical
  manifestations of the military systems, minimally
  including location and extent.
• Behavior Models
• The collection of algorithms used to model how
  military equipment is employed in response to
  existing conditions (e.g., doctrine, tactics,
  planning, problem solving).
• Virtual simulations replace these models by
  direct participants / trainees.

• Environmental State
• The collection of data defining the measurable
  attributes/values used to describe the SNE at a
  point in time.
• Specific values may, however, represent
  time-varying conditions.
• This state would be initialized via a SEDRIS
  transmittal, then evolve in response to other
  inputs.
• Environmental Models
• The collection of algorithms used to derive
  additional SNE-related data (e.g. emergent
  properties such as geometric occlusion along a
  path), or evolve the state of the SNE over time
  in response to internal and external factors.

9
Conceptual Reference Model

• Environmental Ground Truth forms the gaming board
• Military System Models comprise the gaming pieces
• The rules of play are tightly coupled to the
  configuration of the gaming pieces
• The gaming board is constantly changing
  configuration as a result of gaming piece
  state/actions as well as on its own ...

10
CRM Elaboration

• Environmental Effect
• A direct influence of the environment on sensors
  (active and passive), weapons countermeasures,
  and/or units/platforms (e.g., ducting of acoustic
  energy, mud).
• May be as simple as a pass-through of
  environmental state in the immediate vicinity of
  the military system model .
• Environmental Impact
• A direct influence on the environment caused by
  active sensors, weapons countermeasures, and/or
  units/platforms (e.g., crater, shipping noise).
• Environmental Internal Dynamics
• The algorithms modeling physical processes that
  cause spatial and temporal variations in
  environmental state variables.
• These processes may be represented by either
  distinct data sets that define the environmental
  state at a sequence of times or by algorithms
  implementing mathematical models of some form
  (e.g., deterministic or stochastic differential
  equations).
• Almost all MS training systems today have
• limited effects,
• no impacts, and
• (at best) limited prescriptive internal dynamics

11
Acoustic Data Processing

• Conceptually, energy and mass in the environment
  should be treated identically
• Modeled at all locations (in all directions, at
  all frequencies, )
• Resulting measured values represented as
  Environmental Data
• Conceptually, we propose that the above sequence
  of processing stages takes place, even if only in
  the mind of the system designer

12
Critical Point!

• The end-to-end trace of data flow
• Ship Component Model to
• Ship Acoustic Signature Environmental Impacts
  model to
• Acoustic Energy Environmental Data to
• Acoustic Propagation Loss Environmental Effects
  model, and finally to
• Passive Sensor Component Model
• affords greater opportunity for comparing
  alternative process flows, models, and
  implementations than merely stating
• ship entities have acoustic signatures which
  affect sonars
• Unfortunately, the SNE community is currently in
  the latter situation and progress in developing a
  common understanding has been limited to date.

13
STOW TAOS Total Atmosphere and Ocean Services

• Internal Dynamics is often scripted
  (pre-computed) due to comp cost
• But could be as simple as an ephemeris driving
  solar illumination

14
Contemporary SNE Problems

• Specifying Simulations
• How much SNE is enough? For what purpose? How
  to separate data from models? How to decompose
  the specification process? Traceability?
• Federating Simulations
• Good News few existing simulations include
  robust SNE
• Bad News every simulation has some SNE and no
  one really understands the semantic implications
  of establishing data interoperability
• Multi-Resolution Modeling
• Trade-off between Military System Model
  resolution and number executing
• Desire to validly federate constructive and
  virtual simulations
• Does reduced Military System Model resolution
  require reduced SNE resolution? What are the
  implications of SNE dynamics cross-federation?
• Command and Control
• Another Virtual Simulator federate (e.g.,
  trainee-crewed TOC)
• Alternatively, a Live-Virtual (bullet firing)
  federation ...
• Non-DoD Applications
• Any computer application that references SNE data

15
Requirements Implications

• Usually missing
• The specification of environmental data elements
  which Military System Models must respond to,
• How they will react, and
• Which environmental data elements they will
  impact.

16
Wheres the Beef?

• The CRM clearly shows the multiple
  inter-relationships between Military Systems
  Models and the Environmental Ground Truth
• Improves our ability to define the specification
  problem(s)
• Clean separation of the Military Systems Models
  from the Environmental Ground Truth leads to a
  well-specified interface
• Implications for parallelizing system development
  and potential reuse
• Recognizing the inherent decomposition of the
  Environmental Ground Truth into State/Data and
  Models encourages good specifications
• Of State/Data to be populated from SEDRIS
  transmittals
• And subsequently maintained over the course of
  the simulation
• Separately from the Models (algorithms) to be
  reused (or implemented de novo)
• Better specifications are encouraged and
  potential reuse enhanced
• To the degree that reuse occurs, cross-federation
  interoperability improves
• To the degree that good specifications are
  prepared, the development of the SOM, and
  subsequently the FOM, are simpler and more likely
  to be correct and complete!
• A better way to think about specifying,
  developing, and reusing both data and system
  components within simulations and federations
• Consistent use should lead to reduced life-cycle
  costs and improved interoperability across the
  MS community

17
Consistency and Interoperability
Consistency Agreement (with statistical limits)
between results (final or intermediate) of
different simulations of the same scenario.
18
Multi-Resolution Interoperability
19
Multi-Resolution Military Models

• Whether the conjecture is true or
  false is not the issue here
• That the CRM facilitates Diagram and Discuss,
  is and this is otherwise very
  difficult to do today!

20
Example C4ISR Application
21
SNE Conceptual Reference Model

• Supports Diagram and Discuss
• Focus SNE community efforts in the area of
  environmental modeling system description and
  design (architecture?)
• Leverage individual/system experience in
  descriptive form
• Serve as basis for prescriptive approach to
  next-gen systems
• Provide focus to next iteration of MS Master
  Plan
• Context for future MS SNE EA architecture
  related efforts
• Increased rigor for developmental systems (e.g.,
  JSIMS)
• Design and documentation
• Basis for establishing future interoperability
• Remains a work-in-progress, like SEDRIS in FY96
• Based only on experienced perspectives
• Needs concerted (funded) effort
• Judiciously elaborate and test against new
  concepts
• Application to problems of immediate relevance