Study 354 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA The - - PDF document

TIiE ANALYSIS OF SlRATEGlC COMMUNICATIONS USING PETRI-NETS zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I

The Task - Study 354 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I Question X2

Master Question X2 “Justify Military SATCOM”

• Sponsor: ACDS(C1S)
• Timescale: 2003 - 2013
• SAG approved scenarios

Study Dates: Jan 94 -Jan 95 DRA able to provide data inputs Page 1

The Context zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• DOAC'S View zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA
• f c31

Intelligence

Sitreps

affect combat units

Page 2

I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

354 and the

c31

Strategy

, J p 1

Represent Key

pjzjiii

Decisions zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

and

• Information

The Role of SATCOM zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

long distance c o n " 'Must' be

• beyond range of terrestrial HFNHF links
• Secure - not exploitable
• Survivable - not prone to jamming or disruption
• Safe to use - enemy not able to locate users
• Capable of supporting forces worldwide

Versatile Page 3

Model comms infrastructure in 2010 Identify IERs in specific scenarios Identify the 'background' requirement Determine mismatch between requirement and MOE: Satisfaction of Requirement: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• Proportion of Info zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

IERs supported

• Proportion of Context IERs supported
• Value of Intel IERs supported

capability

I

BrigadelSF ManpackslSubs

I

Tri-service 100+ 'nodes' SATCOM + 6 other comms systems ScenariolTime dependent IERs Petri-Net approach

...

.

. . ..

.

..

. .

. . . . . . . .

. , . . , . .

Page 4

The Application of Coloured Petri-Nets

simulation technique with these properties zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• networks of nodes and links explicitly modelled
• queues explicitly modelled
• competition for resources explicitly modelled
• 'tokens' move from one node to another along links
• tokens have properties (attributes)
• nodes can act on tokens depending on their

properties

Approach

Satellite Capacity

Sai: Caoacitv

'ransmitter :Message Queue zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Page zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

5

e zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I

t zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I

• ur Model

Users and decision nodes = TRANSITIONS Comms bearers = PLACES Data streams (IERs) generated by users from script = TOKENS COMCEN transition

• checks IER attributes against available bearers
• checks connection to recipient
• checks bearer capacity
• passes IER to most appropriate bearer

Bearer transition allocates capacity and power to IER

Features of the Analysis

5 different satellite options 3 different scenarios

• jamming
• EMP

Sensitivity to assumptions on

• channel allocation
• loading

Page zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

6

1 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

4 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Features of the Model

Up to 14 parallel channels per satellite

• Each channel represented by a separate Transition
• Transmitters allocated to

channels depending on scenario

• Different properties at different frequencies, reflecting IER

EHFlUHF and SHF

token attributes

Link budget calculations Dynamic allocation of power between

• Carried out in

each channel transition when an IER enters

beamslchannels

• Channel Transitions use up power which may come from a zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

common 'pot'

Further Features of the Model

Users compete for channels with spare capacity Users seek alternative comms bearers

• COMCEN

transition looks at remaining capacity on each available bearer

• COMCEN

transition checks all bearers against IER attributes and most efficient use of power

Excess users thrown off the system

• COMCEN transition may allocate IERs to a Wastebin'

Page 7

Measures of Effectiveness zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

.

• Typical Comms MOE - 'Delays on

messages' - not relevant at this level Preferred MOE - 'How does SATCOM affect the battle?' - impractical Next best - 'Ability of system@) to meet the [commander's] requirement'

• Classify unsent IERs by typdpriority
• Intel IERs used to determine value of

resuiting intelligence to commander

• Other IERs (operationa1,contextual)

assessed as a straightforward percentage of total requirement

1 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I

Form of Results zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

1

f zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

1 1

1111111

.._.__________________________________._

CI

Total Loading

j

:

i

C0dj

x x x

i

:

:

. . . . . . . . . .

~ ~ ~

• ~

n ~ e

j

: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ; ;

j

:

j j

Amount of Requlrement Met :

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

"

. . . . . . . . . . . . . . i

Page 8

• I

I : zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Progress to Date

Completed:

• Identified nodes in the HLSG scenarios
• Obtained approval for Background Scenario
• Designed Petri-net model
• Determined zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

comms infrastructure

• Determined IERs
• Determined design of satellite options
• Buiit model
• Analysis (since July)
• Reporting (from

November)

On-going: Projected

Challenges

Long timescale

• f satellite compared to other

comms systems impacts on

• Quantity and quality of IER data
• Lack of infrastructure data

Lack of agreed future policy for military Data collection much harder than predicted Long-term aim of measuring influence

• f

communications

• n combat effectiveness

SATCOM Page zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 9

Summary zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

1 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I

Task to justify MILSATCOM Joint DOACIDRA study Method zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• Explicit simulation of all strategic bearers and IERs
• Petri-Net model

On schedule to report Jan 95 Long term capability to model strategic comms being built up Page 10

. . . . .. . . . . . . . . . . . . . .

LYSlS OF STRATEGIC COMMUNICATIONS USING PETR1-NETS

. . . . . . . . . . . . . . . . . . . .

ISMOR '94

. '

• The Task -

Study 354 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

I Question zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

X 2 Master Question X2 "Justify Military SATCOM"

• Sponsor: ACDS(C1S)
• Timescale: 2003 - 2013
• SAG approved

scenarios

Study Dates: Jan 94 -Jan 95 DRA able to provide data inputs Page 1

The Context zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• DOAC's View of C31

Page 2

I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

354 and the C? Strategy The Role of SATCOM

Provide long distance zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

comms

‘Must’ be

• beyond range zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA
• f terrestrial HFNHF links
• Secure - not exploitable
• Survivable - not prone to jamming or disruption
• Safe to use - enemy not able to locate users
• Capable of supporting forces worldwide

Versatile Page 3

Method

Model comms infrastructure in 2010 Identify IERs in specific scenarios Identify the 'background' requirement Determine mismatch between requirement and MOE: Satisfaction

• f Requirement: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA
• Propoltion of Info IERs supported
• Proportion zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA
• f Context IERs supported
• Value of Intel IERs supported

capability zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

. > g

• qw,

,&

"T,:..

r"

FromMODHQdownto

BrigadelSF ManpackslSubs

h

Tri-service 100+ 'nodes' SATCOM + 6 other comms

I

systems ScenarioKime dependent IERs Petri-Net approach Page 4

The Application of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

. . . .. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

:E

. ...

• L

Coloured Petri-Nets

Coloured Petri-nets are a graphical dynamic simulation technique with these properties

• networks of nodes and links explicitly modelled
• queues explicitly modelled
• competiiion for resources explicitly modelled
• ‘tokens’ move from one node to another along links
• tokens have properties (attributes)
• nodes can act on tokens depending on their

properties

Approach

Satellite Capacity Transmitter

(Messme Queue

Manager ‘

I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Y

Wastebin zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Page 5

Users and decision nodes zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

= TRANSITIONS

Comms bearers = PLACES Data streams (IERs) generated by users from script = TOKENS COMCEN transition

• checks IER attributes against available bearers
• checks connection to recipient
• checks bearer capacity
• passes IER to most appropriate bearer

Bearer transition allocates capacity and power to IER

Features of the Analysis

5 different satellite options 3 different scenarios Represent effects of

• jamming
• EMP

Sensitivity to assumptions on

• channel allocation

Page zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 6

Features of the Model

Up to 14 parallel channels per satellite

• Each channel represented by a separate Transition
• Transmitters allocated

to channels depending on scenario

• Different properties at different frequencies, reflecting IER

EHFIUHF and SHF

token attributes

Link budget calculations

Dynamic allocation of power between

• Carried out In each channel transition when an IER enters

beamdchannels

• Channel Transitions use up power which may come from a

Further Features of the Model

Users compete for channels with spare capacity Users seek alternative comms bearers Excess users thrown off the system

• COMCEN transition looks at remaining capacity on each

available bearer

• COMCEN transition checks all bearers against IER

attributes and most efficient use zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• f

power

• COMCEN transition may allocate IERs to a ’wastebln’ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Page zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 7

Measures of Effectiveness zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

. Typical Comms MOE - 'Delays on messages' - not relevant at this level Preferred MOE - 'HOW does SATCOM affect the battle?' - impractical Next best - 'Ability of system@) to meet the [commander's] requirement'

• Classify unsent IERs by type/priority
• Intel IERs used to determine value of

resulting intelligence to commander

• Other IERs (operationa1,contextual)

assessed as a straightforward percentage of total requirement zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• yd."lb. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Form of Results

11111

Satellite Loading

a l

_ _ _ _ _

Total Loadlng

niin m

ER6 not sent

. . . . . . . . . . . . . . . . . . . . . . . .

"

i

i j

X X X zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA i j

i i zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

• 0. 0 .o

i j

i

. . . . .

" ' ~ ~ n c e

'

1 .

!

Amount of Requirement Met i

.. ..

i

i

i i . . .

. . . . . . . . . . . . . . . . .

"

. . . . . . . . . "."

"I

Page 8

. . zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Progress to Date " 1

a

Completed:

• Identified nodes in the HLSG scenarios
• Obtained approval zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

for Background Scenario

• Designed Petri-net model
• Determined comms infrastructure
• Determined IERs
• Determined design of satellite options
• Buik model
• Analysis (since July)
• Reporting

(from November)

On-going: Projected

nlW.d%id1M16

Challenges

Long timescale of satellite compared to other comms systems impacts

• n
• Quantity and quality of IER data
• Lack zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA
• f

intastructure data

Lack of agreed future policy for military Data collection much harder than predicted Long-term aim of measuring influence

• f

communications

• n combat effectiveness

SATCOM

Page 9

. .

I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA

Summary

Task to justify MILSATCOM Joint DOACIDRA study Method

• Explicit simulation of all strategic bearers and IERs
• Petri-Net model

On schedule to report Jan 95 Long term capability to model strategic comms being built up

Page 10