Loading...

Samei, Farzad | 2010

585 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 40844 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Ejlali, Alireza
  7. Abstract:
  8. With new technologies, processor power density is dramatically increased which results in high temperature. Temperature has significant impact on reliability, performance, power consumption and cooling costs. This alarming trend underscores the importance of temperature management methods in system design. Recently, various research efforts have been focused on addressing thermal issues. Due to mobility, cost and size constraints and reliability requirements, the embedded systems do not warrant use of modern cooling mechanisms such as heat sink and fan. This caused the thermal issues to be more prominent in embedded domain. This study aims at introducing thermal management techniques for distributed embedded systems. We introduce an analytical model for thermal behavior of a repetitive task sequence with real-time constraints. This model includes both transient and steady-state thermal behaviors. By exploiting this model, we propose an algorithm to reduce peak temperature in a distributed embedded system under real-time constraints. Considering thermal blocks inside the tasks we distributed the slack time more attentively and achieved more temperature reduction. We also present Schedule-Swapping, a technique for reducing peak temperature in distributed embedded systems while satisfying real-time constraints. Contrary to many other approaches, our proposed technique does not use slack time for reducing power dissipation but leaves it to be used by recovery mechanisms. In addition, we show that by changing the power distribution over time, peak temperature can diminish. According to this observation, we introduce a DVS-based scheme which suggests executing some portions of task with lower voltage and frequency and executing the other portions with the maximum voltage and frequency while satisfying deadlines.

  9. Keywords:
  10. Task Migration ; Embedded System ; Distributed Embedded System ; Thermal Management ; Voltage/Frequency Scaling ; Temperature Reduction

 Digital Object List

 Bookmark

No TOC