Logo des Repositoriums
Zur Startseite
  • English
  • Deutsch
Anmelden
  1. Startseite
  2. SuUB
  3. Dissertationen
  4. Dependable and energy-efficient cyber-physical systems by graceful degradation
 
Zitierlink DOI
10.26092/elib/1552

Dependable and energy-efficient cyber-physical systems by graceful degradation

Veröffentlichungsdatum
2022-05-18
Autoren
Schmidt, Robert  
Betreuer
Garcia-Ortiz, Alberto  
Gutachter
Fey, Görschwin  
Zusammenfassung
Factory automation, civil infrastructure monitoring, medical wearables, process-
and environmental control are essential applications of computer systems in
a modern society. By integration of computer systems with the environment,
society transfers major trust to such cyber-physical systems, which are required
to dependable deliver their expected service. Designing dependable computer
systems is a challenge since the inception of the computer, but today’s applications
limit one major resource required for dependability: Energy.
Where early dependable systems could spend large energy budgets, today’s
cyber-physical systems are often battery powered and required to be energy-
efficient. To enable their widespread adoption in society, we developed a
design approach which integrates low-power with dependable system design. We
leverage the energy efficiency of modern commodity components by safeguarding
them with temporal redundancy.
The resulting cyber-physical systems are energy-efficient and yet dependable,
and their real-time guarantees empower the integration of mixed-criticality
functions with verifiable quality of service guarantees. Our static, design-time
and certification-friendly approach is a breeze for designers and certification
authorities, as no assumptions about hardware features and error probabilities
are necessary, which allows application- and cost optimal component selection.
Schlagwörter
dependability

; 

low power

; 

real-time
Institution
Universität Bremen  
Fachbereich
Fachbereich 01: Physik/Elektrotechnik (FB 01)  
Dokumenttyp
Dissertation
Zweitveröffentlichung
Nein
Lizenz
https://creativecommons.org/licenses/by/4.0/
Sprache
Englisch
Dateien
Lade...
Vorschaubild
Name

2022schmidt.pdf

Size

3.34 MB

Format

Adobe PDF

Checksum

(MD5):264baeca6406eb65afdbd41b7ee3950d

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science

  • Datenschutzbestimmungen
  • Endnutzervereinbarung
  • Feedback schicken