SEARCH
ISTC-CC NEWSLETTER
RESEARCH HIGHLIGHTS
Ling Liu's SC13 paper "Large Graph Processing Without the Overhead" featured by HPCwire.
ISTC-CC provides a listing of useful benchmarks for cloud computing.
Another list highlighting Open Source Software Releases.
Second GraphLab workshop should be even bigger than the first! GraphLab is a new programming framework for graph-style data analytics.
ISTC-CC Abstract
IdleChat: Enabling High Bandwidth Real-time Applications in Residential Broadband Networks
In ACM SIGMOBILE Mobile Computing and Communications Review (MC2R) (invited paper), July 2011.
Ramya Raghavendra1, Michael Kaminsky2, Konstantina Papagiannaki3, Srinivasan Seshan4, Elizabeth Belding5
1IBM Research, Hawthorne
2Intel Labs Pittsburgh
3Telefonica Research
4Carnegie Mellon University
5University of California Santa Barbara
Residential broadband links are characterized by low and variable upload capacity and large latencies, leading to poor video performance. Aggregating multiple backhaul links using 802.11 Access Points has been considered as a solution to increasing the backhaul capacity limit. In this paper, we revisit the problem of backhaul aggregation in the context of interactive applications. We present the IdleChat architecture that uses wireless connectivity to aggregate idle broadband connectivity in a neighborhood to overcome the limits of individual asymmetric broadband uplinks. In contrast to prior work, IdleChat aggregates in a traffic-aware manner, focussing specifically on low-latency, real-time link aggregation in support of interactive applications such as video chat. IdleChat schedules packets in a deadline-aware manner to improve interactivity. We evaluate IdleChat both in an experimental testbed in our lab and in a residential deployment. Our results show that use of IdleChat in existing deployments can improve video SNR by up to a factor of 10 db in the presence of a single neighboring AP. This results in video quality improvement from a MOS score of bad to excellent.
FULL PAPER: pdf