[Acpc-l] 11.12.00: Informatikkolloquium
Katrin Seyr
seyr@dbai.tuwien.ac.at
Wed, 06 Dec 2000 14:31:15 +0100
--=====================_19131625==_.ALT
Content-Type: text/plain; charset="iso-8859-1"; format=flowed
Content-Transfer-Encoding: quoted-printable
Die Technische Universit=E4t Wien und die =D6sterreichische
Computergesellschaft laden gemeinsam zu folgenden 2 Vortr=E4gen im Rahmen=
des
Kolloquiums des Fachbereiches Informatik ein und bitten um Weiterleitung
an Interessierte:
**********************************************************************
1. V o r t r a g
Thema: Probabilistic Scheduling Guarantees
Prof.Dr. A. Burns
Department of Computer Science
University of York, UK
alan.burns@cs.york.ac.uk
http://www.cs.york.ac.uk/~burns/
Zeit: 11.Dezember 2000, 16:30 p=FCnktlich
Ort: Zemanek H=F6rsaal, Favoritenstra=DFe 11/Erdgeschoss
**********************************************************************
2. V o r t r a g
Thema: A R I N C - 6 5 9 / S A F E b u s=AE -
Safety critical systems in the Aerospace Industry
Kevin DRISCOLL
Senior Staff Scientist Honeywell Technology Center
Safety-Critical, Secure, and Real-Time Systems Architectures
Zeit: 11.Dezember 2000, 17:45 s.t.
Ort: Zemanek H=F6rsaal, Favoritenstra=DFe 11/Erdgeschoss
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D
Probabilistic Scheduling Guarantees
Abstract:
Hard real-time systems are usually required to provide an absolute
guarantee that all tasks will always complete by their deadlines. In this
talk we introduce the notion of a probabilistic guarantee. Two aspects are
investigated: imprecise knowledge of Computation Times and Fault
Tolerance. The techniques of extreme value estimation is explained for
predicting the execution times of tasks on modern processors. For fault
tolerance, schedulability analysis is used together with sensitivity
analysis to establish the maximum fault frequency that a system can
tolerate. The fault model is then used to derive a probability
(likelihood) that, during the lifetime of the system, faults will not
arrive faster than this maximum rate. The framework presented is a general
one that can accommodate transient `software' faults, tolerated by
recovery blocks or exception handling; or transient `hardware' faults
dealt with by state restoration and re-execution.
****************************************************************************=
**********
Prof.Dr. A. Burns
Department of Computer Science, University of York, UK
Biography:
Alan Burns is a Head of Department and Professor of Real-Time Systems in
the Department of Computer Science, University of York, U.K. He graduated
in Mathematics from the University of Sheffield in 1974, undertook his PhD
at the University of York before taking up a tenured post at the
University of Bradford. He joined the University of York again in 1990 and
was promoted to a personal chair in 1994. Together with Professor Andy
Wellings he heads the Real-Time Systems Research Group at the university;
a group that has currently 4 faculty members, 9 Postdoc researchers and 7
PhD students. He has served on many Programme Committees and has been PC
chair and General Chair for RTSS. He is currently the Chair of the IEEE
Technical Committee on Real-Time Computing. He has published widely (over
250 papers and articles, and 10 books) and works on a number of research
areas within the real-time field.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
A R I N C - 6 5 9 / S A F E b u s=AE -
Safety critical systems in the Aerospace Industry
A b s t r a c t:
ARINC-659 is the first standard backplane bus for commercial avionics
systems. It was designed to meet the highest levels of safety criticality
while allowing untrusted hardware to be connected to the bus and untrusted
software to execute on the same processors as safety critical software. The
design requirements and detailed design features of ARINC-659 will be
discussed as well as its first application as the SAFEbus backplane for the
Boeing 777 Aircraft Information Management System (AIMS). This highly
integrated system comprises most of the cockpit functions in the Boeing 777
airplane, some of these functions which have the highest safety criticality
and some are completely untrusted. The concept of "robust partitioning" in
time and space will be introduced and the mechanisms used by ARINC-659 to
support robust partitioning will be discussed.
=
***************************************************************************=
**************
Kevin DRISCOLL
(Senior Staff Scientist Honeywell Technology Center)
(Safety-Critical, Secure, and Real-Time Systems Architectures)
B i o g r a p h y:
From 1971 to 1976, Mr. Driscoll was a electronic cryptography specialist=
for
the U.S. Army's Communication Command and Army Security Agency. He taught
at the cryptography school in Fort Monmouth, New Jersey. In 1977, he joined
Honeywell's Systems and Research Center. He was a major designer of both
ARINC 659 and PI-bus (the only standard military avionics backplane bus).
He helped design the VHSIC TM bus, the predecessor of IEEE 1149. He
pioneered the use of self-checking pairs which is now a common fault
tolerance technique, developed a fault tolerant fiber optic mesh
communications system, designed the first multi-tasking 1553 bus controller
and designed the only ultra-reliable 1553. He has contributed to the
electronics architecture design of: National Aerospace Plane (NASP), Space
Defense Initiative (SDI), Light Helicopter Experimental (LHX), Boeing's 777
(AIMS and ADIRS), Advanced Launch System (ALS), Honeywell's vetronics and
unmanned vehicle programs (air, land, and underwater). His current
interests are cryptography and real-time fault-tolerant systems.
--=====================_19131625==_.ALT
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<html>
<br>
Die Technische Universit=E4t Wien und die =D6sterreichische<br>
Computergesellschaft laden gemeinsam zu folgenden 2 Vortr=E4gen im Rahmen
des<br>
Kolloquiums des Fachbereiches Informatik ein und bitten um
Weiterleitung<br>
an Interessierte:<br>
<br>
**********************************************************************<br>
&nbs=
p;
<x-tab> </x-tab>
<u>1. V o r t r a g<br>
<br>
</u> Thema: Probabilistic
Scheduling Guarantees<br>
<br>
&nbs=
p;
Prof.Dr. A. Burns<br>
&nbs=
p;
Department of Computer Science<br>
&nbs=
p;
University of York, UK<br>
&nbs=
p;
alan.burns@cs.york.ac.uk<br>
&nbs=
p;
<a href=3D"http://www.cs.york.ac.uk/~burns/"=
eudora=3D"autourl">http://www.cs.york.ac.uk/~burns/</a><br>
<br>
Zeit: 11.Dezember 2000, 16:30 p=FCnktlich<br>
Ort: Zemanek H=F6rsaal, Favoritenstra=DFe 11/Erdgeschoss<br>
<br>
**********************************************************************<br>
<br>
<x-tab> </x-tab><x-tab> =
</x-tab>
<u>2. V o r t r a g<br>
<br>
</u><font face=3D"Arial, Helvetica"> &nbs=
p;
Thema: A R I N C - 6 5 9 / S A F E b u s=AE - <br>
</font> &nb=
sp;
<font face=3D"Arial, Helvetica">Safety critical systems in the Aerospace
Industry<br>
<br>
</font> <br>
<font face=3D"Arial, Helvetica"> &n=
bsp; =
Kevin DRISCOLL <br>
Senior Staff Scientist Honeywell Technology
Center<br>
Safety-Critical, Secure, and Real-Time Systems
Architectures <br>
<br>
<br>
</font>Zeit: 11.Dezember 2000, 17:45 s.t.<br>
Ort: Zemanek H=F6rsaal, Favoritenstra=DFe 11/Erdgeschoss<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D<br>
&nbs=
p;
<br>
<x-tab> </x-tab><x-tab> =
</x-tab>Probabilistic
Scheduling Guarantees<br>
<br>
<u>Abstract:<br>
</u>Hard real-time systems are usually required to provide an
absolute<br>
guarantee that all tasks will always complete by their deadlines. In
this<br>
talk we introduce the notion of a probabilistic guarantee. Two aspects
are<br>
investigated: imprecise knowledge of Computation Times and Fault<br>
Tolerance. The techniques of extreme value estimation is explained
for<br>
predicting the execution times of tasks on modern processors. For
fault<br>
tolerance, schedulability analysis is used together with=20
sensitivity<br>
analysis to establish the maximum fault frequency that a system can<br>
tolerate. The fault model is then used to derive a probability<br>
(likelihood) that, during the lifetime of the system, faults will
not<br>
arrive faster than this maximum rate. The framework presented is a
general<br>
one that can accommodate transient `software' faults, tolerated by<br>
recovery blocks or exception handling; or transient `hardware'
faults<br>
dealt with by state restoration and re-execution.<br>
<br>
****************************************************************************=
**********<br>
<x-tab> </x-tab><x-tab> =
</x-tab><x-tab> &=
nbsp; </x-tab>Prof.Dr.
A. Burns<br>
<x-tab> </x-tab>Department
of Computer Science, University of York, UK<br>
<br>
<u>Biography:<br>
</u>Alan Burns is a Head of Department and Professor of Real-Time Systems
in<br>
the Department of Computer Science, University of York, U.K. He
graduated<br>
in Mathematics from the University of Sheffield in 1974, undertook his
PhD<br>
at the University of York before taking up a tenured post at the<br>
University of Bradford. He joined the University of York again in 1990
and<br>
was promoted to a personal chair in 1994. Together with Professor
Andy<br>
Wellings he heads the Real-Time Systems Research Group at the
university;<br>
a group that has currently 4 faculty members, 9 Postdoc researchers and
7<br>
PhD students. He has served on many Programme Committees and has been
PC<br>
chair and General Chair for RTSS. He is currently the Chair of the
IEEE<br>
Technical Committee on Real-Time Computing. He has published widely
(over<br>
250 papers and articles, and 10 books) and works on a number of
research<br>
areas within the real-time field.<br>
<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D<br>
<font face=3D"Arial, Helvetica"> <br>
<x-tab> </x-tab><x-tab> =
</x-tab>A
R I N C - 6 5 9 / S A F E b u s=AE - <br>
<x-tab> </x-tab>
Safety critical systems in the Aerospace Industry<br>
<br>
<u>A b s t r a c t:</u> <br>
ARINC-659 is the first standard backplane bus for commercial avionics
<br>
systems. It was designed to meet the highest levels of safety criticality
<br>
while allowing untrusted hardware to be connected to the bus and
untrusted <br>
software to execute on the same processors as safety critical software.
The <br>
design requirements and detailed design features of ARINC-659 will be
<br>
discussed as well as its first application as the SAFEbus backplane for
the <br>
Boeing 777 Aircraft Information Management System (AIMS). This highly
<br>
integrated system comprises most of the cockpit functions in the Boeing
777 <br>
airplane, some of these functions which have the highest safety
criticality <br>
and some are completely untrusted. The concept of "robust
partitioning" in <br>
time and space will be introduced and the mechanisms used by ARINC-659 to
<br>
support robust partitioning will be discussed. <br>
<br>
</font>***************************************************************=
**************************<br>
<font face=3D"Arial, Helvetica"> <x-tab> &=
nbsp; </x-tab><x-tab> </=
x-tab><x-tab> </x-tab>Kevin
DRISCOLL <br>
<x-tab> </x-tab>(Senior
Staff Scientist Honeywell Technology Center) <br>
<x-tab> </x-tab>(Safety-Criti=
cal,
Secure, and Real-Time Systems Architectures) <br>
<br>
<u>B i o g r a p h y:</u> <br>
From 1971 to 1976, Mr. Driscoll was a electronic cryptography specialist
for <br>
the U.S. Army's Communication Command and Army Security Agency. He taught
<br>
at the cryptography school in Fort Monmouth, New Jersey. In 1977, he
joined <br>
Honeywell's Systems and Research Center. He was a major designer of both
<br>
ARINC 659 and PI-bus (the only standard military avionics backplane bus).
<br>
He helped design the VHSIC TM bus, the predecessor of IEEE 1149. He=20
<br>
pioneered the use of self-checking pairs which is now a common fault
<br>
tolerance technique, developed a fault tolerant fiber optic mesh <br>
communications system, designed the first multi-tasking 1553 bus
controller <br>
and designed the only ultra-reliable 1553. He has contributed to the
<br>
electronics architecture design of: National Aerospace Plane (NASP),
Space <br>
Defense Initiative (SDI), Light Helicopter Experimental (LHX), Boeing's
777 <br>
(AIMS and ADIRS), Advanced Launch System (ALS), Honeywell's vetronics and
<br>
unmanned vehicle programs (air, land, and underwater). His current <br>
interests are cryptography and real-time fault-tolerant systems.<br>
<br>
<br>
</font></html>
--=====================_19131625==_.ALT--