Loading...

Reducing the Energy Overhead of Replication Mechanisms in Distributed Embedded Systems

Salehi Khanghahbar, Mohammad | 2010

727 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 40979 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Ejlali, Alireza
  7. Abstract:
  8. Distributed embedded systems are widely used in safety-critical applications, such as avionics platform and flight control system. Faults in such applications may cause catastrophic effect. Therefore, these systems must be fault-tolerant and must be designed not only to detect faults but also to recover from faults. Fault tolerance is normally realized on multi-processor system via temporal duplication (time-redundancy) or spatial duplication (hardware-redundancy) depending on the availability of the slack time. Slack time is defined as the difference between the deadline and execution time. Duplication is a common technique to achieve fault tolerance in safety-critical applications, but may lead to considerable energy overhead. In this project we propose a low-energy duplication technique for fault-tolerant hard real-time systems. In this technique, we have two identical processors, and the system has two operation phases: DWC (duplication with comparison) phase, and recovery phase. In the DWC mode, identical single-processor schedules are executed on the processors and the results are compared for fault detection. During the DWC mode, we use dynamic voltage scaling (DVS) for both the processors to reduce the energy consumption. Once a fault has been detected, the system switches to the recovery mode where a dual-processor schedule is used for recovery executions at the maximum supply voltage, and then the system switches back to the DWC mode. We provide an online energy management method which uses a slack reclamation scheme, specifically developed for the proposed technique. We have physically implemented the technique and compared it with typical time-redundancy techniques. Experimental results show that, as compared to time-redundancy techniques, the proposed technique either provides the same level of fault coverage with up to 19% energy saving, or provides considerably better fault coverage but with up to 9% energy saving
  9. Keywords:
  10. Energy Management ; Embedded System ; Distributed Embedded System ; Fault Detection ; Tasks Replication

 Digital Object List

 Bookmark

No TOC