Loading...

A Real-Time and Energy-Efficient Decision Making Framework for Computation Offloading in Iot

Heydarian, Mohammad Reza | 2018

642 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 51501 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Fazli, Mohammad Amin
  7. Abstract:
  8. Based on fog computing paradigm, new applications have become feasible through the use of hardware capabilities of smart phones. Many of these applications require a vast amount of computing and real-time execution should be guaranteed. Based on fog computing, in order to solve these problems in is necessary to offload heavy computing to servers with adequate hardware capabilities. On the other side, the offloading process causes time overhead and endangers the real-timeliness of the application. Also, because of the limited battery capacity of the handheld devices, energy consumption is very important and should be minimized.The usual proposed solution for this problem is to refactor the classes inside the code to make them offloadable in two server and client versions that are run on server and handheld device respectively. By analyzing the energy and performance profiles of these classes, the decision for offloading or not offloading is done for each one. In this thesis, an algorithm is introduced that using all the necessary tools, automatizes the process of finding class computation profile. This algorithm makes the basis of a complete and independent framework for offloading decision making. All the steps required for enabling offloadability and implementing this framework in a generic Android application are explained. The contribution of this work is proposing an experimental execution phase that in it, the framework calculates the necessary performance and energy profiling by executing simultaneously on client and server and provides inputs for decision algorithm. This way more autonomy is made possible. If the hardware be able to execute parts of the process on within their response time in real-time, this algorithm finds the optimum execution mode and uses a heuristic for decreasing energy consumption.It is proved that in this special case by using this offloading framework, convergence time decreased from 21 seconds in local execution to 12 seconds and frame rate increases from 4 fps in offloaded execution and 14 fps in local execution to 19 fps in selective offloading, that is a remarkable improvement. Also, the energy consumption of the app is decreased and battery draining time from 90% to 80% increased from 11 minutes and 30 seconds local execution to 12 minutes and 45 seconds in selective offloading
  9. Keywords:
  10. Green Computing ; Internet of Things ; Offloading ; Energy Efficiency ; Edge Computing ; Fog Computing

 Digital Object List

 Bookmark

No TOC