Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute...

34
NASA Aeronautics Research Institute Separa&on Pla,orm for Integra&ng Complex Avionics (SPICA) NASA Aeronau&cs Research Mission Directorate (ARMD) FY12 LEARN Phase I Technical Seminar November 13–15, 2013

Transcript of Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute...

Page 1: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Separa&on  Pla,orm  for  Integra&ng  Complex  Avionics  (SPICA)    

NASA  Aeronau&cs  Research  Mission  Directorate  (ARMD)  FY12  LEARN  Phase  I  Technical  Seminar    

November  13–15,  2013  

Page 2: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

SPICA  

November  13–15,  2013     2  

Architecture�Integration

Aircraft�HW�Architecture

Navigation

Route�Planning

Communications

Scheduler

Mem Mgr

Thread�Mgr

Msg Router

SMT,�other�solvers

AADL�Models

Allocation

Scheduling

System�Architecture�Specification

Configured�Safe�Secure�RTOS

Navigation

Commun

ications

Route�Planning

Separation�Platform�for�Integrating�

Complex�Avionics

Aircraft�HW�Architecture

Hosted�Control�Applications Domain�Specific�Components

Sel

ect

System�Safety�RequirementsSystem�Security�Requirements Rate,�Jitter,�Duration

RTOS�modelMutual�ExclusionPerformanceIntegrity,�AvailabilitySynchronization

Integrated�Vehicle

Requirements

Multi-core, multi-processingDistributed sensing, actuationPhysical separationBandwidthHierarchical Memories

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 3: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Presenta&on  Outline  

•  The  problem  •  Technical  approach  •  Results  of  Phase  I  •  Impact  •  Next  steps  

November  13–15,  2013     3  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 4: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

787  Common  Data  Network  

November  13–15,  2013     4  

Diagram  showing  where  the  common  core  system  (CCS)  is  connected  throughout  the  787  aircraV.  Most  of  what  is  noted  in  the  fuselage  are  the  21  or  so  remote  data  concentrators  that  GE  provides  and  are  advanced  sensors  to  the  CCS.  Source:  GE  Avia&on.  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 5: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

November  13–15,  2013     5  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 6: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

A380  IMA  

November  13–15,  2013     6  

Source: http://www.artist-embedded.org/docs/Events/2007/IMA/Slides/ARTIST2_IMA_Itier.pdf

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 7: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Avionics  Hardware  Example  

CPM: Core Processing Module IOM: Input / Output Module LRM: Line Replaceable Module

SAFEbus® Design: Honeywell International, Brendan Hall et al, "Ringing Out Fault Tolerance." (DSN'05) pp 298-307.

         IOM1  

         IOM2  

         CPM

1            CPM

2            CPM

3            CPM

4          IO

M3  

       IO

M4  

SAFEbus®  

ARINC  429  

ARINC  629  

ARINC  429  

ARINC  629  

Airplane Information Management System (AIMS) Cabinet

HOST – Input / Output Module (IOM)

429   629  

HOST Core Processing Module (CPM)

BIU: Bus Interface Unit ARINC 629: communications ARINC 429: communications

November  13–15,  2013     7  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 8: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Technical  approach  

•  Modeling  the  en&re  aircraV  avionics  in  SAE’s  AADL  open-­‐source  standard  •  Formal  model  of  relevant  constraints  •  Using  a  Sa+sfiability  Modulo  Theories  (SMT)  solver,  extrac&ng  informa&on  

directly  from  the  AADL  model.  

November  13–15,  2013     8  

Innova&ons:  –  Complete,  consistent  set  of  constraints  defining  a  correct  schedule  

(not  an  algorithm  or  a  priority  scheme)  suppor&ng  a  wide  range  of  architectures  and  protocols  

–  Using  a  generic  solving  engine  (yices  SMT  solver)  to  generate  schedules  

–  Solving  mul&ple  levels  of  a  hierarchical  scheduling  problem,  all  at  the  same  &me  and  in  the  same  model  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 9: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Phase  I  Output  

As  a  result  of  this  project,  we  have  generated:  –  a  formal  specifica&on  of  the  complete  set  of  constraints,  sufficient  to  

represent  a  wide  range  of  different  avionics  architectures  and  problems,  

–  a  large  set  of  test  problems  for  input  to  the  yices  SMT  solver,  demonstra&ng  the  use  of  those  constraints,  along  with  output  results  and  performance  data,  

–  a  tunable  test  problem  generator,  automa&cally  genera&ng  problem  instances  in  yices  input  format,  and  

–  an  exemplar  aircraV  avionics  architecture,  rendered  in  both  diagrams  and  AADL.  

These  ar&facts  are  available  under  SBIR  data  rights  for  government  use.    In  addi&on,  we  intend  to  turn  the  forthcoming  Phase  I  final  report  into  one  or  more  technical  papers  for  conference  submission.  

November  13–15,  2013     9  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 10: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Modeling  the  en&re  aircraV  avionics  

•  Hierarchical  organiza&on  •  Asynchronous  boundaries  •  ARINC  429,  653,  659,  664  •  Varying  latencies,  rates,  cri&cality  •  Shared  memory,  buffers,  and  buses  •  …  

November  13–15,  2013     10  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 11: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

e500.X03c  

e500core  

e500core  

e500core  

e500core  

PCI

L3  

Hardware  Parts  

11  

AFDX  VME   Ethernet  PCI  

intel   i5   i5core  

RAM   ROM   MASS  

Arm   arm7   arm7core  

arm7.X03c  

arm7core  

arm7core  

arm7core  

arm7core  

PCI

L2  Freescale   e500mc   e500core  

L[123]  

AFDX_FO  

TI   TI320C  

November  13–15,  2013     NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 12: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Modules  

November  13–15,  2013     12  

Cabinet  Switch  

arm7.X03c  

ROM  RAM  

PCI  

AFDX Core  Processing  Module  

e500.X03c  

ROM  RAM  

AFDX

PCI  

Network  Interface  (FOX)  

AFDX arm7.X03c  

ROM  RAM  

PCI  AFDX_FO

ACS   FOX  

Remote  Data  Concentrator  

AFDX_FO

FOX  

&320.X03c  

ROM  RAM  

I2C  (?)  

429

A

D CAN (asynch boundary)

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 13: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Core  Processing  Module  Detail  

November  13–15,  2013     13  

Core  Processing  Module  

e500.X03c  

ROM  RAM  

AFDX

PCI  

Core  Processing  Module  

ROM  

RAM  

e500.X03c  PCI  

Bridge  AFDX_FO

MemBus  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 14: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

2

Core  Computer  

November  13–15,  2013     14  

Core  Computer  

AFDX_FO

Core  Processing  Module  

e500.X03c  

ROM  RAM  

AFDX

PCI  

Core  Processing  Module  

e500.X03c  

ROM  RAM  

AFDX

PCI  

8

AFDX  

Network  Interface  (FOX)  

AFDX arm7.X03c  

ROM  RAM  

PCI  AFDX_FO

2

Cabinet  Switch  

arm7.X03c  

ROM  RAM  

PCI  

AFDX 2

Core  Computer  

AFDX_FO CPMs  (8)  

Switch  

FOX  (2)  AFDX  

(symbol)

(Implementation)

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 15: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Adven&um  “X-­‐03c”  

•  Next  Gen  Commercial  transport  •  Twin  engine  •  Long  haul  •  Narrow  body  •  Mixed  passenger  /  cargo  

November  13–15,  2013     15  

0        3  X

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 16: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

X-­‐03c  Hardware  Architecture  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     16  

AFDX

Common Cabinets

Data Concentrator

Data Concentrator

Sensors  

Actuators  

Sensors  

Actuators  

DA/AD  

DA/AD  

Page 17: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

X-­‐03c  Hardware  Architecture  (AADL)  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     17  

Core  Computer  

AFDX_FO CPMs  (8)  

Switch  

FOX  (2)  AFDX  

Core  Computer  

AFDX_FO CPMs  (8)  

Switch  

FOX  (2)  AFDX  

RDC  

AFDX_FO 429

A

D CAN Sensors  

Actuators  

Sensors  

Actuators  

RDC  

AFDX_FO 429

A

D CAN

RDC  

AFDX_FO 429

A

D CAN

Forward EE bay

Aft EE bay

Page 18: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Formal  model  of  relevant  constraints  

Latency  Jirer  Preemp&on  Over/Undersampling  Asynchronous  boundaries  Task  grouping/varying  context-­‐switch  &mes  Inter-­‐  and  intra-­‐frame  &ming  constraints  Shared  memory  Resource  assignment  …  

November  13–15,  2013     18  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

For  the  full  set  and  formal  defini&ons,  see  the  Final  Report.  

Page 19: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Task  Grouping  and  Preemp&on  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     19  

Page 20: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Communica&on  Latency  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     20  

Latency:�Preemption

• Producer�and�consumers�can�be�interrupted�or�preempted• In�that�case,�Latency�is�still�measured�from�the�producer begin event�to�consumer end event

2013 ©�2009Ͳ2013�Adventium�Proprietary 13

Time

Producer�A

Data�transfer�ABConsumer�B

Latency�AB Latency:�Over�Sampling

Oversampling�can�be�used�to�reduce�latencyWithin�a�synchronous�temporal�domain,�latency�is�measured�to�the�first�consumer�instance

2013 ©�2009Ͳ2013�Adventium�Proprietary 14

Time

Producer�A�(5�Hz)

Consumer�B�(10�Hz)

Latency�AB

Page 21: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Latencies  Across  Asynch.  Boundary  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     21  

LatencyͲJitter:�Asynchronous�Boundaries

The�minimum�latencyͲjitter�guarantee�across�an�asynchronous�boundary�equals�Jitter�+�period�of�the�slowest�rate�of�the�producer/consumer

2013 ©�2009Ͳ2013�Adventium�Proprietary 19

Time

Task�A

Task�B

10�ms 160�ms

Frames�in�phase�sync(large�latency)

Frames�out�of�phase�sync(min�latency)

Page 22: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Using  a  SMT  solver  

•  Effec&ve  combina&on  of  logical  and  mathema&cal  reasoning  •  Based  on  technologies  demonstrated  to  scale  to  millions  of  variables  

and  constraints  

November  13–15,  2013     22  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

   ;;;  Jobs  on  a  given  cpu  may  not  overlap  (assert  (or  (/=  pJ1  pJ2)  (<=  fJ1  sJ2)  (<=  fJ2  sJ1)))  (assert  (or  (/=  pJ1  pJ3)  (<=  fJ1  sJ3)  (<=  fJ3  sJ1)))  (assert  (or  (/=  pJ1  pJ4)  (<=  fJ1  sJ4)  (<=  fJ4  sJ1)))  (assert  (or  (/=  pJ1  pJ5)  (<=  fJ1  sJ5)  (<=  fJ5  sJ1)))  

Page 23: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Undersampling  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     23  

0 1000 2000 3000 4000

12

/home/redman/Adventium/spica/Design/experiments/smt/test−cases/test8.mat

time

resource

t4(1)t1(1) t1(2) t1(3) t1(4)

t2(1) t2(2) t2(3) t2(4)

t3(1) t3(2)

partition123

Page 24: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Dataflow  and  Resource  Assignment  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     24  

0 1000 2000 3000 4000

12

3

/home/redman/Adventium/spica/Design/experiments/smt/test−cases/test14.mat

time

resource

t1(1) t2(1) t1(2) t2(2) t1(3) t2(3) t1(4),t6(1) t2(4)

t3(1) t4(1) t3(2) t4(2) t3(3) t4(3) t3(4) t4(4)

t5(1) t5(2) t5(3) t5(4)

partition123

Page 25: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Mul&ple  Buses  and  Asynch.  Boundary  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     25  

0 500 1000 1500 2000 2500 3000

12

34

56

78

910

1112

13

/home/redman/Adventium/spica/Design/experiments/smt/test−cases/test15.mat

time

resource

t01(1) t01(2) t01(3),t16(1)

t02(1) t02(2) t02(3)

t03(1) t03(2) t03(3)

t04(1) t04(2) t04(3) .. .

t05(1) t05(2) t05(3)

t06(1) t06(2) t06(3)

t07(1) t07(2) t07(3)

t08(1) t08(2) t08(3)

t09(1) t09(2) t09(3)

t10(1) t10(2) t10(3)

t11(1) t11(2) t11(3)

t12(1) t12(2) t12(3) .

t13(1) t13(2) t13(3)

t14(1) t14(2) t14(3)

t15(1) t15(2) t15(3)

partition12345

678910

1112131415

Page 26: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Scaling  For  Different  Problems  

November  13–15,  2013     26  

● ● ● ● ● ● ● ●

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

● ● ● ● ● ● ● ●

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

● ● ● ● ● ● ● ●

Test cases scaling

Multiplier

Tim

e to

sch

edul

e (s

ec)

1 2 3 4 5 6 7 8

010

030

050

070

090

011

00●

●

●

test1test2test4test5test6test7test8test9test10test11test14test15test16

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 27: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Scaling:  Processor  Load  vs.  Time  

November  13–15,  2013     27  

● ● ● ● ● ● ●● ● ●

●● ●

●

●

●

●

●

050

010

0015

00

Load versus Time

Load

Aver

age

time

to s

ched

ule

(sec

)● 0 flows, UNsat inc.

0 flows, SAT only5 flows, UNsat inc5 flows, SAT only

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 28: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Impact:    NASA  Programs  

•  Integrated  Systems  Research  -­‐-­‐  inves&ga&ng  the  avionics-­‐level  integra&on  of  novel  func&ons,  hardware  systems,  and  architectures.    

•  Avia&on  Safety  -­‐-­‐  This  program  includes  assurance  for  flight-­‐cri&cal  systems,  including  managing  the  complexity  of  architec&ng,  valida&ng,  and  verifying  the  correct  func&oning  of  increasingly  complex  avionics.  SPICA’s  output  is  a  concrete  schedule  which  can  easily  be  verified  to  sa&sfy  requirements  governing  execu&on  &mes,  latencies,  and  sampling  rates,  as  well  as  more  complex  issues  such  as  metastable  communica&ons  across  an  asynchronous  boundary.    

•  Orion  -­‐-­‐  SPICA  is  developing  relevant  capabili&es  for  other  complex,  networked  vehicular  systems.  For  example  NASA’s  Orion  MPCV  uses  several  of  the  protocols  and  standards  SPICA  is  designed  to  address.  

November  13–15,  2013     28  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 29: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Impact:    Outside  of  NASA  

1.  Adven&um  is  part  of  the  System  Architecture  Virtual  Integra&on  (SAVI)  consor&um  as  a  tool  vendor  partner.  SAVI  is  an  Aerospace  Vehicle  Systems  Ins&tute  (AVSI)  program,  with  membership  from  industry,  government,  and  academia.    

2.  The  Phase  I  proposal  included  lerers  of  support  from  Lockheed  Mar&n  and  the  Army  Avia&on  and  Missile  Research  Development  and  Engineering  Center  (AMRDEC).  

3.  Adven&um  has  a  current  contract  suppor&ng  the  Army  in  the  development  of  an  Architecture-­‐Centric  Virtual  Integra&on  Process  for  Future  Ver&cal  LiV  mission  systems.  

November  13–15,  2013     29  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 30: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Next  steps  

Technical  issues  •  Mul&-­‐core,  more  generally,  e.g.,    conten&on  for  on-­‐board  cache  •  Integra&ng  mul&ple  scheduling  approaches  •  Other  protocols,  as  needed  

Matura&on  •  Scaling  •  Finalize  transla&on  from  AADL  to  SMT  input  format  •  Different  avionics  architectures  

November  13–15,  2013     30  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 31: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Mul&-­‐core  

•  Characteris&cs  addressed  in  Phase  I  –  Shared  IO  –  Shared  buffers  –  Task  alloca&on  to  discrete  processing  resources  –  (A)synchronous  communica&ng  processes  

•  Issues  deferred  to  Phase  II  –  Memory  conten&on  from  different  cores,  including  various  levels  of  

cache  –  Migra&on  between  cores  –  (virtualiza&on)  

November  13–15,  2013     31  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 32: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Integra&ng  Different  Schedulers  

•  SPICA  produces  a  sta+c  schedule  that  is  mathema&cally  guaranteed  to  sa&sfy  the  input  constraints.  

•  Dynamic  schedulers  are  specified  in  terms  of  a  set  of  schedulability  constraints.  

•  Current  integra&ons  provide  a  sta&c  alloca&on  within  which  the  dynamic  scheduler(s)  have  control.  

•  SMT  is  specifically  designed  to  incorporate  specialized  types  of  constraints  

So:    Is  it  possible  to  specify  schedulability  constraints  in  a  decomposable  form,  such  that  the  resul&ng  alloca&on  may  take  several  forms,  but  is  in  any  case  guaranteed  to  be  schedulable?    For  example,  is  one  alloca&on  more  efficient  than  another  at  accommoda&ng  sporadic,  high-­‐priority,  low-­‐latency  tasks  and  s&ll  providing  the  required  guarantees  for  other  tasks?  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    November  13–15,  2013     32  

Page 33: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

Scaling  

•  Current  system  solves  problems  involving  dozens  to  hundreds  of  constraints,  in  minutes.  

•  SMT  technology  has  demonstrated  performance  on  millions  of  constraints  •  Growth  in  solving  &me  with  problem  size  is  reasonable.  

Next  steps:  •  Search  control  •  Problem  reformula&on  

November  13–15,  2013     33  

In  previous  work,  we  have  demonstrated  several  orders  of  magnitude  increase  in  problem  size  and  decrease  in  solving  &me.  

NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    

Page 34: Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA))...NASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e ion ft HW e tion e Planning tions Scheduler gr ad

NASA Aeronautics Research Institute

The  End  

November  13–15,  2013     34  NASA  Aeronau&cs  Research  Mission  Directorate  FY12  LEARN  Phase  I  Technical  Seminar    


SPICA FPI envelopes – current status

SPICA FPI envelopes – current status

SPICA FAR-Infrared Instrument SAFARIaramis.obspm.fr/PNCG08/talks/SAFARI-PNGPNC-April08.pdf · majeur en cosmologie. 2 avril 2007, Paris SPICA FAR-Infrared Instrument : SAFARI Denis.Burgarella@oamp.fr

SPICA FAR-Infrared Instrument SAFARIaramis.obspm.fr/PNCG08/talks/SAFARI-PNGPNC-April08.pdf · majeur en cosmologie. 2 avril 2007, Paris SPICA FAR-Infrared Instrument : SAFARI [email protected]

Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA)) › sites › default › files › BODDY_LEARN.pdfNASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e

Separaon)Plaorm)for)Integrang) Complex)Avionics)(SPICA)) › sites › default › files › BODDY_LEARN.pdfNASA Aeronautics Research Institute SPICA) November)13–15,)2013)) 2 e

SPICA White Paper

SPICA White Paper

ALMA-SPICA Synergies

ALMA-SPICA Synergies

SPICA : the concept pre-2013 › ~levrier › Files › SPICA-SAFARI-2015.05.04.pdf · • Summer/Fall 2013 – JAXA/L + ESA/M4 ! replan for SPICA 2.0 ... Refine mission/instrument

SPICA : the concept pre-2013 › ~levrier › Files › SPICA-SAFARI-2015.05.04.pdf · • Summer/Fall 2013 – JAXA/L + ESA/M4 ! replan for SPICA 2.0 ... Refine mission/instrument

[CII] mapping of the diffuse ISM with SPICA / SAFARI

[CII] mapping of the diffuse ISM with SPICA / SAFARI

Sludge dewatering Wastewater separaon Polymer dosing

Sludge dewatering Wastewater separaon Polymer dosing

Suggested Bench-Check Procedures for a SPICA Injection Pump

Suggested Bench-Check Procedures for a SPICA Injection Pump

Integrang)FrameNetand) MetaNet) - Linguisticslinguistics.berkeley.edu/cognetwork/documents/FNMNintegration.pdf · Integrang)FrameNetand) MetaNet) Miriam)R.)L.)Petruck) miriamp@icsi.berkeley.edu)

Integrang)FrameNetand) MetaNet) - Linguisticslinguistics.berkeley.edu/cognetwork/documents/FNMNintegration.pdf · Integrang)FrameNetand) MetaNet) Miriam)R.)L.)Petruck) [email protected])

Spica sunon cubicles & cabins

Spica sunon cubicles & cabins

SPICA Mission Requirement Document (MRD) ver. 3.5 draft

SPICA Mission Requirement Document (MRD) ver. 3.5 draft

Stellar Code™ Spica - Humanity Healing Networkhumanityhealing.net/.../2010/08/Stellar-Code-Spica... · Spica or Alpha Virginis, (α) Virgo, is the luminary of Virgo, the constellation

Stellar Code™ Spica - Humanity Healing Networkhumanityhealing.net/.../2010/08/Stellar-Code-Spica... · Spica or Alpha Virginis, (α) Virgo, is the luminary of Virgo, the constellation

SPICA Science for Transiting Planetary Systemsesppro.mtk.nao.ac.jp/~narita/seminar/090602spicaws_narita.pdf · SPICA Science for Transiting Planetary Systems ... Lab tests for characterization

SPICA Science for Transiting Planetary Systemsesppro.mtk.nao.ac.jp/~narita/seminar/090602spicaws_narita.pdf · SPICA Science for Transiting Planetary Systems ... Lab tests for characterization

Educaonal Research on Teaching: Integrang with …...Educaonal Research on Teaching: Integrang with your Research & Teaching Program Tuesday July 11th Kaatje Kra1 with contribu8ons

Educaonal Research on Teaching: Integrang with …...Educaonal Research on Teaching: Integrang with your Research & Teaching Program Tuesday July 11th Kaatje Kra1 with contribu8ons

Spica Sunon Catalogue new (2)

Spica Sunon Catalogue new (2)

The SAFARI Imaging Spectrometer for the SPICA …...Netherlands Institute for Space Research / Netherlands Organisation for Scientific Research The SPICA satellite Tracing the evolution

The SAFARI Imaging Spectrometer for the SPICA …...Netherlands Institute for Space Research / Netherlands Organisation for Scientific Research The SPICA satellite Tracing the evolution

B-BOP: The SPICA far-IR imaging polarimeter

B-BOP: The SPICA far-IR imaging polarimeter

Spica Star

Spica Star

B-BOP, the SPICA Imaging Polarimeter

B-BOP, the SPICA Imaging Polarimeter