Cognitive Management and Control of Optical Metropolitan Area Networks
We seek to develop a new way of managing optical Metropolitan Area Networks (MAN) motivated by increasingly bursty and unstructured high-volume traffic. By using cognitive techniques we can 1) detect predict and estimate these traffic loads and 2) optimize resource scheduling and load balancing across the network. This project specifically focuses on designing algorithms to dynamically achieve high network performance developing theoretical models for these algorithms and running simulations to confirm the results.

I’m incredibly interested in how we can derive simplicity and elegance from chaotic systems and this SuperUROP project explores exactly that! We’re investigating optical networks that are home to surprisingly bursty and unstructured traffic. It will be exciting to infer (from this ‘unknown’ traffic) how best manage the network. My background is an intersection of distributed systems neuroscience and CS theory.