Open Collections

Mobile edge cloud : computation scheduling and caching

University of British Columbia

Mobile edge cloud has been proposed to accommodate computational intensive application on mobile devices without augmenting their computational capacities and to combat with the growing traffic volume in the network. The main concept of mobile edge cloud is to bring computation and storage near to the end users, serving users’ requests locally, and reducing wide area network latency. Although mobile edge cloud improves network performance and users’ quality of experience, it brings several challenges due to possessing limited resources. The unified focus of the thesis is to design mechanisms for mobile edge cloud to maximally exploiting the computation and storage resources at the edge of the network to minimize network traffic and latency. In the first part of the thesis, we design a distributed computational resource sharing mechanism where femtocell access points (FAPs) share their resources with neighbouring FAPs and form local clouds. The aim of forming such local femto-clouds is to serve computational requests locally, reducing the data transfer delay and improving users’ quality of experience. The resource sharing problem is formulated as an optimization problem and a myopic procedure is presented that enables FAPs to collaboratively find its solution in a distributed fashion. In the second and third part of the thesis, we focus on designing caching mechanisms for mobile edge network. It has been illustrated that almost 60% of the data traffic results from the asynchronous requests for some popular content. By caching those few popular content at the edge of the network, demand for the same content can be served locally, resulting in a reduction in the data traffic volume and downloading delay in the network. In the second part of the thesis, we construct a caching scheme that accounts for content popularity prediction and properties of the network (e.g. cache size, bandwidth, and network topology). The caching scheme is further extended in the final part of the thesis that includes content popularity prediction errors and routing mechanism in the caching decision. For the caching schemes mixed-integer linear programming is used to compute where to cache the content in the network to minimize content downloading delay.

Text

2018-05-28

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/66140

Electrical and Computer Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/