Crowdsourced multimedia content: Resource allocation and data transmission

Author: 
Date created: 
2015-12-15
Identifier: 
etd9382
Keywords: 
Crowdsourced multimedia
Resource allocation
Abstract: 

Recent years have witnessed the rapid growth of crowdsourced multimedia services, such as text-based Twitter, image-based Flickr, and video streaming-based Twitch and YouTube live events. Empowered by today's rich tools for multimedia generation/distribution, as well as the growing prevalence of high-speed network and smart devices, most of the multimedia contents are crowdsourced from amateur users, rather than from commercial and professional content providers, and can be easily accessed by other users in a timely manner. Since cloud computing offers reliable, elastic and cost-effective resource allocation, it has been adopted by many multimedia service providers as the underlying infrastructure. In this thesis, we formulate the cloud resource allocation in crowdsourced multimedia services as a standard network utility maximization (NUM) problem with coupled constraints, in which real-time user interaction is a fundamental issue, and develop distributed solutions based on dual composition. We further propose practical improvements for the content generation and big data processing of crowdsourced multimedia services in a cloud environment. Crowdsourced multimedia services also rely on convenient mobile Internet access, since mobile users occupy a large portion of both content generators and content consumers. The rich multimedia content, especially images and videos, put significant pressure on the infrastructure of state-of-the-art cellular networks. Device-to-device (D2D) communicationthat smartly explores local wireless resources has been suggested as a complement of great potential to support proximity-based applications. In this thesis, we jointly consider resource allocation and power control with heterogeneous quality of service (QoS) requirements from diverse multimedia applications.

Document type: 
Thesis
Rights: 
This thesis may be printed or downloaded for non-commercial research and scholarly purposes. Copyright remains with the author.
File(s): 
Supervisor(s): 
Jiangchuan Liu
Department: 
Applied Sciences: School of Computing Science
Thesis type: 
(Thesis) Ph.D.
Statistics: