Execution Scheduling Methods for Mobile Applications.

Date of Submission

December 2015

Date of Award

Winter 12-12-2016

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science

Department

Advance Computing and Microelectronics Unit (ACMU-Kolkata)

Supervisor

Banerjee, Ansuman (ACMU-Kolkata; ISI)

Abstract (Summary of the Work)

Making mobile applications energy efficient immensely builds user satisfaction. Apart from the fact that there are not many efficient techniques for evaluating energy consumption for applications on mobile devices, the methods used are static in nature. Static techniques assume that during the running of an application, no other process can run concurrently, and the concerned application has the entire CPU at its disposal. This thesis is built around three main ideas.Firstly, we propose a novel idea of measuring the energy consumption of an application running on a mobile device considering the fact that not always the entire CPU is available. This is because the application may sometimes run in the foreground when the mobile is idle and therefore, use the maximum CPU available; at other times, there maybe other tasks being run (apart from the routine background tasks) by the user for which this application is forced to run in the background. The major highlight of this work is in considering the concept of variable CPU availability in energy analysis. We have also suggested to model the energy consumption problem of a mobile phone as a finite state automaton, where our aim is to find if a state can be reached where the entire battery of the mobile phone is exhausted.As our next work, we address the problem of application scheduling on user mobile devices. Given the fact that a vast number of application may run both as in the foreground and the background, the scheduling task is a challenging one. We propose to solve this problem using clustering and iterative refinement. Results on simulated benchmarks show the efficacy of our proposal.Mobile Cloud Computing (MCC) offloading has emerged as a key way of mitigating the resource constraints of mobile devices like smartphones. In MCC offloading systems, one or more tasks of the mobile application are migrated and executed on the cloud system. For our final work, we focus on utilizing MCC to optimize applications having tasks with different levels of quality of service. We propose an algorithm to optimize the quality of service of tasks while ensuring that their execution does not exceed the given energy budget. Analysis of our algorithm shows that it provides the optimal solution in polynomial time.

Comments

ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28843416

Control Number

ISI-DISS-2015-312

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

http://dspace.isical.ac.in:8080/jspui/handle/10263/6469

This document is currently not available here.

Share

COinS