CAREER: Scalable Techniques for Validating Broadband Infrastructure Availability and Performance

Access to reliable broadband Internet shapes who can participate fully in modern life, from schoolwork to healthcare to job opportunities. Yet understanding where reliable Internet is actually available turns out to be surprisingly hard: directly measuring every community is too expensive and slow to do at national scale. This project develops techniques for learning the patterns behind where and how Internet service providers build their networks, making it possible to estimate where reliable Internet is available, identify gaps, and support efforts to ensure public money intended to connect communities reaches those that need it most. Beginning from the observation that Internet networks are planned and built according to geographic, economic, and regulatory constraints, this project develops a modeling framework for broadband infrastructure that learns spatial and temporal patterns in how Internet service providers deploy their networks. Drawing on techniques from Internet measurement and machine learning, the project is organized around three predictive tasks: forecasting how providers change claimed coverage over time, inferring likely service availability from observable spatial and demographic features, and assessing the validity of claimed service performance. By reframing broadband measurement as a prediction problem, this work expands the methodological toolkit available for reasoning about Internet infrastructure at scale under uncertainty. The project strengthens public oversight of broadband deployment by developing open tools that help researchers, communities, and regulators to evaluate whether Internet service is available where providers claim it is. The models, datasets, and software produced in this project will be integrated into public dashboards for use by local, state, Tribal, and federal policymakers; community advocates; individuals; and researchers. The project also establishes a Broadband Data Clinic, an experiential learning program in which undergraduate students work with community partners on real-world broadband data challenges. This clinic will provide hands-on training at the intersection of networking, data science, and public-interest technology, while delivering free technical assistance to communities facing broadband data challenges. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2542894 | Program: 01002930DB NSF RESEARCH & RELATED ACTIVIT,01002627DB NSF RESEARCH & RELATED ACTIVIT,01003031DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Shaddi Hasan | Institution: Virginia Polytechnic Institute and State University, BLACKSBURG, VA | Award Amount: $389,891 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2542894 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2542894.html