RELAY - Resource utilization in time-dependent large-scale distributed systems

Increasing heterogeneity of end-systems and the increasing speed of consumer access networks make large scale distributed multimedia applications such as streaming, gaming and Internet telephony increasingly popular. Higher bandwidth, interactivity and more symmetrical traffic patterns change the challenges that lie in developing and operating an affordable infrastructure. Users' demands for quality of service remain, as does the lack of end-to-end resource reservation. We therefore investigate system-level approaches that provide quantifiable better support for a class of distributed systems in spite of the lack of control over the infrastructure, at a predictable cost.

RELAY will help to ensure the maintenance of the existing and the deployment of new services. We will design, implement and evaluate mechanisms and tools to improve resource utilization, increase throughput, reduce/hide latency and support soft QoS. One of our main goals is to integrate and combine mechanisms to get more scalable, less resource demanding, high performance systems for time-dependent large-scale distributed multimedia systems.

We consider architectural, kernel and protocol support for reduced resource consumption in servers and intermediate systems, algorithms for the allocation of data and functions to servers and intermediate systems, and investigate combinations of performance-enhancing mechanisms such that they do not counteract each other.

RELAY members are currently involved in

• the Verdione project (the technical part of The World Opera)
• the MiSMoSS project
• the ADIMUS project
• the HyStream project
• the ROMUS project
• the iAd SFI
• the AMPS activity
• the GEMINI project

 Active and graduated students:

• Master students
• PhD students

 Additional information:

• Goals and methods of the RELAY group
  
• Checklist for applicants