Real-time video streaming is a significant and growing segment of today's Internet traffic, particularly for the edge. In the meantime, the expectations of high quality of experience (QoE) from end users are also increasing dramatically. However, achieving high QoE is difficult because lossy links with unpredictable latency are prevalent in practice. This means the burden is on the transport layer to minimise the transport-induced delay, especially when delivering real-time or delay-sensitive payload.
Currently, video applications use either UDP or TCP as payload carrier, but neither of which is well suited to meet a wide range of application requirements. TCP has become the main transport protocol for video streaming traffic because of feasibility of existing global infrastructural deployment. However, its reliable protocol, e.g., loss recovery mechanism, can lead to poor performance when the transmission encounters packet losses that are not caused by congestion. As an alternative, UDP could offer better delay performance due to its best effort transmission mechanism. However, this jeopardises video quality as packets can be lost during transit or arrive out of order. This extends to many edge focused applications such as data analytics, high speed data queries. etc.
Edge computing brings new opportunities and challenges. This project aims at building a resilient transport protocol for supporting a large variety of low-latency edge applications.
Entry requirements
Students should have a first degree in computer science or related areas.
Applicants must meet the minimum English language requirements. Further details are available on the international website.
How to apply
All applications should be made online. Under school/department name, select 'Computer Science'. Please quote reference CO/PT-Un1/2021.
The deadline for applications is 30th June 2021.