Loading...

Energy-Aware Fault Tolerant Mapping of Mixed-Criticality Tasks on Heterogeneous Multicores

Safizade Amzajerdi, Amir Hassan | 2020

388 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52706 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Hessabi, Shaahin
  7. Abstract:
  8. Increasing complexity of modern embedded systems has caused tasks with different criticality levels to run on a shared hardware platform, called mixed-criticality systems. In these systems, it is important to use fault tolerant techniques to achieve reliability targets. Moreover, the limited battery life in many applications and real-time requirements make energy-aware reliability management and guaranteeing deadline constraints a major challenge. Checkpointing with rollback recovery is a good option to ensure reliability of these sys- terns because of its low cost and relatively low time overhead. However, improper use of this technique can impose high energy overhead and cause deadline misses. In the other hand, using heterogeneous multicore platforms is increasing and these platforms can be used to overcome mixed-criticality challenges. In this study, first, an approach is introduced to determine the number and interval of checkpoints according to multiple operational states, and then a schedulability test based on demand bound functions is proposed for fault-tolerant mixed-criticality tasks. Finally, a checkpoint time-aware and checkpoint energy-aware map- ping algorithm is proposed for mixed-criticality tasks. The evaluation results show that using the proposed mapping algorithm can increase schedulabilty while the energy does not change appreciably.
  9. Keywords:
  10. Mapping ; Mixed Criticality ; Energy Aware Fault Tolerance ; Heterogeneous Multicors ; Checkpointing with Rollback Recovery

 Digital Object List

 Bookmark

No TOC