IP geolocation estimation using neural networks with stable landmarks

The ability to accurately determine the geographic location of an arbitrary IP address has potential in many applications. Previous methods based on observing the relationship between network delay and physical distance are inaccurate. Methods based on delay similarity are more accurate, but inefficient because they need information on a large number of landmark nodes near the destination to be collected and maintained. We propose a method that can overcome both problems. Our method maintains a stable collection of network observers and landmark nodes that covers the target area. Observers are network nodes from which we can issue measurement command such as ping and traceroute. Landmarks are IP addresses which are reachable from observers and for which the physical locations are well-known. With measurement results collected from these landmarks, we trained two-tier neural networks that estimated the geolocation of arbitrary IP addresses. Our experiments demonstrate that a high accuracy similar to earlier methods retained with a limited number of landmarks. More specifically, the median error of our estimation is 4.1 km on the datasets with 1547 landmarks across US territory. The median error decreased to 3.7 km on half of the test regions which contain more than 100 landmarks.