This research explores Geofeeds' validation, adoption, and development patterns as a future potential for the improvement of IP geolocation. Geofeeds are independently published lists of IP address ranges linked with geographic data, enabling the possibility for more accurate location data than using traditional methods. Geofeeds have the ability to improve the accuracy of IP geolocation, which is important for such purposes as online security, content delivery, and targeted services. But as great as this potential is, they're still new and not widely adopted or fully understood.
We started by verifying the Geofeed data through an automated process that begins with checking for duplicate entries to eliminate redundancy early on. Using the geofeed-validator tool, we then identified standard violations and formatting errors. After examining hundreds of thousands of entries, we found common issues such as incorrect IP prefixes and subdivision codes. We also quantified the frequency of these problems and identified the network organizations responsible. This gave us a clear picture of the most frequent errors and highlighted key areas for improving data quality.
The project's second section examined adoption. To determine how popular Geofeeds are, we mapped Geofeed entries to their corresponding Autonomous System Numbers (ASNs) using daily data snapshots from CAIDA. We kept track of how many ASes took part, how many IP ranges were covered, and how Geofeed coverage and errors changed over time. Between August 2024 and April 2025, we saw significant growth, especially in the deployment of IPv6, but we also noticed that most ASes still only published a small amount of Geofeed data.
This project is unique because it emphasizes automation to support continuous analysis. Prior research did not delve into error types but instead examined snapshots taken every few weeks. In addition to closing that gap, our work offers tools that researchers and network operators can use to more consistently track Geofeed adoption and data quality. Our method is more automated, more thorough, and more appropriate for long-term monitoring than previous studies. This project establishes the foundation for future IP geolocation that is more transparent and dependable by enhancing the way we validate and analyze Geofeeds.