BroadMap—A novel network measurement framework

Professor: Arpit Gupta

Description: BroadMap, a novel network measurement framework designed to simplify the collection of high-fidelity broadband quality datasets at scale. Additionally, BroadMap will augment broadband quality assessments with critical contextual information such as the type of speed tiers, last-hop medium, bottleneck link location, etc.—enabling more precise attribution of degraded network quality. These insights are vital not only for users but also for policymakers and nonprofits working to address broadband inequalities.

Preferred Qualifications: Networking basics (knowledge of basics concepts, like TCP, speedtest, etc.), Python programming skills, Basic understanding on ML workflow (data collection, pre-processing, model training, evaluation/validation), Knowledge of LLMs (i.e., how they work, and how to train LLMs from scratch) is a big plus