I am not presently recruiting any Ph.D. student. Master's and undergraduate
students looking for directed study projects are always welcome to drop by
for a chat.
Internet Performance Estimation using Big Data
Is it possible to combine the millions of RTT and bandwidth
measurements per second, collected as part of delivery by content and
cloud storage providers, to estimate real-time performance to anywhere
in the Internet? In addition to helping predict the performance
experience of a newly established client, could the estimated
real-time performance also help improve the estimates of clients that
rely on their own past experience to predict future experience?
Applying topic modeling and other statistical analysis and
visualization tools to analyze historical data for trends and
Students may also find the pointers I've collected useful.