Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/86468
Title: Efficient data and application offloading in mobile cloud computing
Authors: Li, Jiwei
Degree: Ph.D.
Issue Date: 2017
Abstract: The inherent limitations in battery size and computing capabilities are great obstacles to developing resource-intensive applications for mobile devices. The emergence of mobile cloud computing (MCC), combining the power of both cloud computing and mobile computing, heralds a promising solution to the limitations on mobile devices. In MCC, the execution of applications is distributed across mobile devices and cloud servers, taking advantage of both the mobility and cloud computing powers. However, one of the enabling technologies for MCC, wireless network communications, also poses a great challenge to the efficiency of MCC applications, as data must be transferred between mobile devices and cloud services at the cost of energy consumption. In the thesis, we attempt to mitigate the negative effects brought by wireless data transmissions in MCC. We focus on two major areas in MCC, namely data offloading and application offloading. The contributions of the thesis are threefold. First, we propose an energy efficient data communication mechanism between wearable devices and cloud servers in mobile environments. The mechanism utilizes smartphones as a middle layer to help wearable devices make wise decisions to transfer data. Second, we propose a novel pre-processing based framework for computer vision application offloading on smartphones. Our empirical experiments show that pre-processing image data before uploading can significantly reduce the cost of network transmissions, at the cost of slightly compromised accuracy. Third, we apply the pre-processing technique to the problem of broadcasting from wearable cameras. As video live streaming requires low end-to-end latency, we propose a two-phase video resolution adjustment approach, namely dynamic video recording on wearable cameras and Lyapunov based video pre-processing on smartphones. Our evaluation results show that our approach achieves up to 50% reduction in power consumption on smartphones and up to 60% reduction in average delay, at the cost of slightly compromised video quality. We believe that our exploration and proposed methods can further bridge the gap between theory and practice of MCC for various types of mobile devices.
Subjects: Hong Kong Polytechnic University -- Dissertations
Mobile computing
Cloud computing
Wireless communication systems
Pages: xvi, 135 pages : color illustrations
Appears in Collections:Thesis

Show full item record

Page views

3
Citations as of May 29, 2022

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.