39
Information Technology and C4ISR 14 March 2011 James H. Lawton, PhD C4ISR/IT European Office of Aerospace R&D Air Force Office of Scientific Research AFOSR Distribution A: Approved for public release; distribution is unlimited. 88ABW-2011-0782

6. Lawton - Informatin Technology

  • Upload
    afosr

  • View
    142

  • Download
    3

Embed Size (px)

Citation preview

Page 1: 6. Lawton - Informatin Technology

Information Technology

and C4ISR 14 March 2011

James H. Lawton, PhDC4ISR/IT

European Office of Aerospace R&D

Air Force Office of Scientific Research

AFOSR

Distribution A: Approved for public release; distribution is unlimited. 88ABW-2011-0782

Page 2: 6. Lawton - Informatin Technology

2

2010 AFOSR SPRING REVIEWC4ISR/IT PORTFOLIO OVERVIEW

NAME: James H. Lawton

BRIEF DESCRIPTION OF PORTFOLIO:

International Research in C4ISR and Information Technologies

LIST SUB-AREAS IN PORTFOLIO:

• Command & Control (C2)

Planning and Scheduling

• Software and Systems

• Information Assurance

• Information Management

• Quantum Information Processing

• Complex networks

• Signal Processing (RF)Other

• Primarily AFRL (RY, RI and AFOSR)

• ONR(G), ARMY - VTT, NBU, CTU

• DARPA (Kohout) - C2 planning

Page 3: 6. Lawton - Informatin Technology

3

EOARD C4ISR/IT Topics being Emphasized, Decreased, Pursued or Transitioned

• C2 Planning and Scheduling

• Software and Systems

• Information Assurance

• Information Management

• Complex Networks

• Quantum Information Processing

• Signal Processing

Page 4: 6. Lawton - Informatin Technology

4

FY 09/10 Research Portfolio

United KingdomC2 Planning

Sensor Prognostics

MAS Comm & Control

Software Dependability

Information management

Spain HPC

GermanyStegonography

Steganalysis

GreeceAdHoc Network Modeling

Czech RepublicAdversarial Planning

Network Monitoring

PolandWireless Info Xfer

HungaryData mining

ItalyWaveform Diversity

Information management

Israel Uncertainty Management

Software regression verification

Ukraine Multimodular recurrent NNs

Bulgaria Analogical Planning

The NetherlandsInformation management

QIP (w/Physics PM)

FinlandCog Networking

Page 5: 6. Lawton - Informatin Technology

5

Scientific ChallengesC2 Planning & Scheduling

• Motivation: Need to improve the speed, quality and creativity of process

– Want to use information technology for improvements

– Computationally intractable (NP-complete)

• Specific C2 P&S questions being addressed:

Analogical reasoning (Petkov)

• People can plan/schedule in highly complex domains using experience –

How?

• Extend existing fundamental analogical reasoning model to address

complex planning domains

• Transformational: make it tractable, change the way we plan

Plan-representation (Wickler)

• Find fundamental components of a “plan”

• Define representation for mixed-initiative interaction

• Transformational: lead to true mixed-initiative planning & execution

Page 6: 6. Lawton - Informatin Technology

6

“Adaptive Problem Solving by Analogy”Georgi Petkov, New Bulgarian University

• Analogy-making is a fundamental cognitive ability that lets us perceive

the world and behave effectively by taking into account our past

experience.

• Problem solving and learning are not separate processes but run in

parallel and influence each other.

• Objective: Analogies for Planning

– Explore how certain cognitive mechanisms can be modelled within

the DUAL cognitive architecture and how they would enhance its

ability to do problem-solving by analogy.

• goal-driven and context-sensitive reasoning,

• knowledge abstraction and abstract concepts formation,

• generalization of known solutions, and

• adaptation of old memory traces.

Page 7: 6. Lawton - Informatin Technology

7

Anticipation by Analogy

?

Representation of the target; retrieval of a base; mapping;

transfer of an anticipation; action

Page 8: 6. Lawton - Informatin Technology

8

DUAL (Kokinov, 1994) Multi-agent Hybrid Architecture

• Each agents has:

– Symbolic part: represents small pieces of knowledge

(features, relations, propositions)

– Connectionist part: represents its relevance

• Integration of the Representation in Hybrid Micro-Agents

– Agent actions include: marker passing, hypothesis

generation, generalization

– Possibilities for context-dependent restructuring of the

knowledge base – the connectionist activation is considered

as power supply for the symbolic processor.

Page 9: 6. Lawton - Informatin Technology

9

Long-Term Memory and WM in DUAL Architecture and AMBR Model

focus

Goal node

Input nodeLTM

WM

Page 10: 6. Lawton - Informatin Technology

10

AMBR Reasoning System

• Reasoning component utilizing the DUAL architecture, that unifies deductive, inductive and analogical reasoning

• All processes emerge from the local interactions between the nodes; no global control is exerted;

• All processes run in parallel and influence each other:

– the retrieval process runs continuously and changes the relevance and therefore changes the mapping;

– the mapping process creates new nodes (hypotheses) and links between them and in this way it changes the spreading activation process.

• Basic mechanisms

– Spreading of activation

– Marker passing and formation of hypotheses

– Formation of anticipations (transfer)

– Constraint-satisfaction network

Page 11: 6. Lawton - Informatin Technology

11

Cued Recall in AMBR

glass

is-broken

kitchen

hplate

cause

on

binding node

yesterday

glass

kitchen

binding node

yesterday

Cue=Target Base=Old Episode

cause

on

is-broken

hplate

Page 12: 6. Lawton - Informatin Technology

12

Example 1 – “Terrorist” Simulation

terrorist Asuicide

Kamikaze Bsuicide

Page 13: 6. Lawton - Informatin Technology

13

Kamikaze Bsuicide

motivation

Prosperity of

the country

Prosperity of

the emperor

Honor of the

family

Japan

Movie

“Shogun”

England

Ireland

Page 14: 6. Lawton - Informatin Technology

14

terrorist Asuicide

Motivation

(CONCEPT)

motivationnostalgia

Un-satisfaction

Page 15: 6. Lawton - Informatin Technology

15

Base “Immigrant”

motivationnostalgia

Un-satisfaction

Ireland

Immigrant CBeating his

wife

Bulgarian

Page 16: 6. Lawton - Informatin Technology

16

Beating his

wifeImmigrant C

Open

Bulgarian

restaurantCause

Stop beating

his wife

Popularizing

traditional

culture

Stop bad

actionsCause

Is-aIs-a

Is-a

Page 17: 6. Lawton - Informatin Technology

18

“Terrorist” example demonstrates

• Analogies between relatively remote

domains;

• Using multiple analogies to re-represent the

target and to retrieve the more appropriate

remote base;

• Transfer and adapting of solutions.

Page 18: 6. Lawton - Informatin Technology

19

Scientific ChallengesSoftware and Systems

• Motivation: “Construct and deliver with confidence software-intensive systems” (Luginbuhl)

– Very difficult to formally ensure correctness and non-functional properties

– Need to extend formalisms to systems engineering

• Specific S&S problems being addressed

Software Regression Verification (Strichman)

• Tractable testing of complex software, approaching solution to Hoare‟s „Grand

Challenge‟

• Needed for legacy and other non-formally defined systems

• Transformational: Software Engineers will need understand formal representation &

specification, this is a middle step

Formal modelling of multi-core programming (Labarta)

• Define task-based programming model that auto-magically exploits parallelism and

maintains correctness

• Transformational: Many cores ARE coming (here? University of Glasgow 1000-core),

need to take advantage of them

• Transition success

Page 19: 6. Lawton - Informatin Technology

20

Software Errors

• USS Yorktown towed to port after

Windows NT divide-by-zero error

(1997)

• Mars Climate Orbiter lost due to

Metric-English unit mismatch

(1999)

• STUXNET worm: targeted

Embedded controllers at Bushehr

Nuclear Plant (2010)

Page 20: 6. Lawton - Informatin Technology

Software Regression Verification: Proving the equivalence of similar programs

Benny Godlin Ofer Strichman

Technion, Haifa, Israel

21

Page 21: 6. Lawton - Informatin Technology

22

Functional Verification

• The main pillar of the grand challenge [Hoare‟03]

– “... the construction and application of a verifying compiler that

guarantees correctness of a program before running it.”

• A more modest challenge: Regression Verification

– Develop a method for formally verifying the equivalence of two similar

programs.

– Pros:

• Default specification = earlier version.

• Computationally easier than functional verification.

– Ideally, the complexity should depend on the semantic difference

between the programs, and not on their size.

– Cons:

• Defines a weaker notion of correctness.

“For every problem that you cannot solve, there is an easier problem that

you cannot solve” (George Polya, in How To Solve It)

Page 22: 6. Lawton - Informatin Technology

23

Hoare‟s

1969

paper

has it all.

. . .

…and 6 pages later:

Page 23: 6. Lawton - Informatin Technology

24

Project Goals

• Develop notions of equivalence

• Develop corresponding proof rules

• Present a prototype for verifying equivalence of C

programs, that

– incorporates the proof rules

– sensitive to the magnitude of change rather than

the magnitude of the programs

Page 24: 6. Lawton - Informatin Technology

25

Notions of equivalence

1. Partial equivalence

Executions of P1 and P2 on equal inputs that terminate, and result in equal outputs

2. Mutual termination

Given equal inputs: P1 terminates P2 terminates

3. Reactive equivalence

Let P1 and P2 be reactive programs. Executions of P1 and P2 which read the same input sequence emit the same output sequence

4. k-equivalence

Executions of P1 and P2 on equal inputs where loops and recursions are restricted to kiterations, result in equal output

5. Full equivalence* = Partial equivalence + mutual termination

6. Total equivalence** = partial equivalence + both terminate

Note: all but k-equivalence are undecidable

* Appeared in Luckham, Park, and M. Paterson [LPP-70] / Pratt [P-71]

** Appeared in Bouge and D. Cachera [BC-97]

Page 25: 6. Lawton - Informatin Technology

26

Hoare‟s Rule for Recursion

Let A be a recursive function.

“… The solution... is simple and dramatic: to permit the use of the

desired conclusion as a hypothesis in the proof of the body itself.”

[H’71]

If

Then

Syntactically Entails

(can be derived from)

Page 26: 6. Lawton - Informatin Technology

27

//in[A]

A( . . . )

{

. . .

//in[call A]

call A(. . .);

//out[call A]

. . .

}

//out[A]

Proving partial equivalence

A B//in[B]

B( . . . )

{

. . .

// in[call B]

call B(. . .);

//out[call B]

. . .

}

//out[B]

Page 27: 6. Lawton - Informatin Technology

28

Proving partial equivalence

The premise is

decidable

• Let AUF , BUF be A,B, after replacing the recursive call with a call to (the

same) uninterpreted function (terminates and produce the same value on

same input).

• We can now rewrite the rule:

Page 28: 6. Lawton - Informatin Technology

2929

Page 29: 6. Lawton - Informatin Technology

31

Regression Verification Summary

• Regression verification is an important problem

– A solution to this problem has a better chance to succeed in the

industry than functional verification

– A grand challenge by its own right…

• This project will extend the exiting regression verification

formalism:

– increase the number of cases of proving partial-equivalence can

handle, by making the inference rules more complete

– develop automatic methods for devising invariants that will enable

us to prove partial equivalence in more cases

– develop methods for proving mutual termination between programs

Page 30: 6. Lawton - Informatin Technology

32

Recent Transitions

• AgentFly

– ARMY funding flight tests (seeking partners)

• Adversarial Planning

– Cyberspace project fully funded by AFRL

• Barcelona Supercomputing Center

– Follow-on project funding from AFRL

• Waveform Diversity (Greco)

– Transition project in low-cost sensor array (Mike

Wickes)

Page 31: 6. Lawton - Informatin Technology

33

Contact Info

James LawtonC4ISR/ITEOARDUnit 4515 Box 14APO AE 09421

Phone: +44 (0) 1895 616187 (Do not use initial 0 if calling from outside the UK) DSN 314 235-6187email: [email protected]

Page 32: 6. Lawton - Informatin Technology

Backup Slides

AgentFly Transitions

Page 33: 6. Lawton - Informatin Technology

35

Deployment on PROCERUS

video

Page 34: 6. Lawton - Informatin Technology

36

Agent-Based Computing in Distributed Adversarial Planning

Michal Pechoucek, Czech Tech Univ

• Adversarial planning: a decision-making process

through which an agent constructs a sequence of

actions (possibly consisting of a single action only)

leading to the desirable goal state of the world in an

adversarial situation, i.e., when other adversarial

agents are concurrently operating in the world.

• Interaction Stances for actions:

• Self-Interested: action of group G is self-interested if it

increases the utility of G

• Cooperative: an action of G is cooperative with H if it

changes the world in a way that increases the utility of

both groups G & H

• Competitive: an action of G is competitive with H if it

does not decrease the utility of H more than it increases

the utility of G

• Adversarial: an action of G is adversarial with H if it

lowers the utility of H more then it increases the utility

of G, or even decreases the utilities of both the groups.

• Altruistic: as an action of G that helps (increases the

utility of) H while lowering the utility of G

• Usage: Adversarial Potential occurs if most of self-

interested actions are competitive or adversarial.

Page 35: 6. Lawton - Informatin Technology

Backup Slides

Plan Representations for Distributed Planning and Execution

Page 36: 6. Lawton - Informatin Technology

38

“Plan Representations for Distributed Planning and Execution”

Gerhard Wickler, Edinburgh University

• Objective: Develop a formal model describing

the key plan representation aspects needed to

support automated planning and scheduling by

distributed software agents

• Approach: Extend Core Plan Representation

(CPR) by examining and adding concepts from

the Beliefs-Desires-Intensions (BDI) model of

software agency that explicitly account for the

agents‟ contributions to the distributed planning

problem

• Core Benefits:

– CPR/BDI structured plans:

• better indexing of experience

databases

• richer plan descriptions support

execution as intended

– structured execution protocols

• principled way of monitoring progress

and state

• principled way of dealing deviations

Plan

Plan Plan Plan

Plan Plan Plan Plan

STRATEGIC

OPERATIONAL

TACTICAL

Page 37: 6. Lawton - Informatin Technology

INCA/CPR+BDI: Shared PlansPlan

Objective

Actor

TimePoint

Resource

Evaluation

Criterion

Intention

Belief

Desire

has-a

is-a

Activity

Issue

Annotation

Constraint Signature

Parameter

Effect

Precondition

Name

Capability

Activity

Method

State-Belief

Page 38: 6. Lawton - Informatin Technology

Backup Slides

Multcore Programming

Page 39: 6. Lawton - Informatin Technology

41

“Programming models for heterogeneous multicore systems”

Jesus Labarta, Barcelona Supercomputing Center

• Background: To efficiently use the multicore machines we need:

– Understanding of application behavior

– Programming models that can express apps in a general/abstract/portable way.

– Intelligent compilers and runtime environments capable of executing very efficiently those

applications on any target machine.

• Project Objectives:

– Refine the StarSs programming model developed at BSC for multicore processors.

– Apply performance analysis tools developed at BSC to better understand parallelism in AFRL-

provided example applications

• Project Approach:

• Programming model:

– Extend dependency mechanism

– Add support for new core/device type (GPUs)

– Handling of partially aliased and strided references

– Hybrid MPI/SPMSs

• Performance analysis:

– Analyze and improve performance of AFRL-provided

apps

– Determined that tools can be extended with pattern

matching techniques