Optimising Networks For Ultra-High Definition Video

The increase in real-time ultra-high definition video services is a challenging issue for current network infrastructures. The high bitrate traffic generated by ultra-high definition content reduces the effectiveness of current live video distribution systems. Transcoders and application layer multicasting (ALM) can reduce traffic in a video delivery system, but they are limited due to the static nature of their implementations. To overcome the restrictions of current static video delivery systems, an OpenFlow based migration system is proposed. This system enables an almost seamless migration of a transcoder or ALM node, while delivering real-time ultra-high definition content. Further to this, a novel heuristic algorithm is presented to optimise control of the migration events and destination. The combination of the migration system and heuristic algorithm provides an improved video delivery system, capable of migrating resources during operation with minimal disruption to clients. With the rise in popularity of consumer based live streaming, it is necessary to develop and improve architectures that can support these new types of applications. Current architectures introduce a large delay to video streams, which presents issues for certain applications. In order to overcome this, an improved infrastructure for delivering real-time streams is also presented. The proposed system uses OpenFlow within a content delivery network (CDN) architecture, in order to improve several aspects of current CDNs. Aside from the reduction in stream delay, other improvements include switch level multicasting to reduce duplicate traffic and smart load balancing for server resources. Furthermore, a novel max-flow algorithm is also presented. This algorithm aims to optimise traffic within a system such as the proposed OpenFlow CDN, with the focus on distributing traffic across the network, in order to reduce the probability of blocking.

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

University of Essex

Year dc:date.issued

2016

Author dc:creator

• Farrow, Paul

• Q Science (General)
• QA75 Electronic computers. Computer science
• T Technology (General)

Language dc:language

en