Loading...

Real-Time Scheduling in Distributed Fault-Tolerant and Low Energy Embedded Systems

Aminzadeh, Soheil | 2009

490 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 39867 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Ejlali, Alireza
  7. Abstract:
  8. Scheduling of real-time processes is often one of the greatest challenges in designing distributed embedded systems. Low energy consumption and fault tolerance are key objectives in the design of embedded systems. However, these objectives are at odds, and there is a trade-off between them. A large part of this thesis includes a comprehensive study on previous researches on energy consumption and fault tolerance in hard real-time embedded systems. Real-time systems usually use system level energy management methods, i.e., dynamic voltage scaling (DVS) and dynamic power management (DPM). Also hard real-time systems often use replication to achieve fault tolerance. In this thesis, we investigate the impact of system level energy management methods on both the reliability and energy consumption of hard real-time systems which use replication for fault tolerance. In this analysis, we have considered four various existing energy management methods: i) Classic DPM, ii) Classic DVS, iii) Postponement method: a variation of DPM which is only applicable to replicated systems, and iv) Hybrid method: a combination of postponement and DVS. Based on the comparative study, we have provided guidelines so that designers can decide which energy management method is more suitable for a given application. For example, we have shown that when reliability is the main concern, the postponement method is the most preferable. However, when the energy consumption is the primary concern, the hybrid method may be more appropriate. The last part of this thesis presents a new approach for scheduling and binding processes of hard real-time task-graphs on a distributed embedded system. Our approach is divided to two separate off-line and on-line phases. In the off-line phase, for each process we first determine a time interval that the process must be executed within, and then we assign the process to a processing node. The results of off-line scheduling are called “Scheduling Tables” that are saved in processors’ memories. On-line scheduler works based on these tables, and uses slack reclamation for DVS. In the proposed scheduling method the off-line scheduler reserves some slack in case a fault occurs at runtime. The on-line scheduler can exploit these slacks to tolerate faults by re-executing the faulty process. The results of experiments show that our approach is very effective from the viewpoints of reliability and energy consumption as compared to previous methods, particularly when deadline constraints are tight.

  9. Keywords:
  10. Energy Management ; Fault Tolerance ; Real Time System ; Embedded System

 Digital Object List

  • محتواي پايان نامه
  •   view

 Bookmark

...see more