On the Path to High Precise IP Geolocation: A Self-Optimizing Model

TL;DR

This paper introduces a self-optimizing IP geolocation model that enhances accuracy by selecting optimal landmarks and approximating network distances using road networks, evaluated in real-world European scenarios.

Contribution

It proposes a novel self-optimizing approach for IP geolocation, including landmark selection and network distance approximation with real-world validation.

Findings

Improved location accuracy in real-world tests

Effective landmark optimization strategy

Enhanced network distance estimation using road networks

Abstract

IP Geolocation is a key enabler for the Future Internet to provide geographical location information for application services. For example, this data is used by Content Delivery Networks to assign users to mirror servers, which are close by, hence providing enhanced traffic management. It is still a challenging task to obtain precise and stable location information, whereas proper results are only achieved by the use of active latency measurements. This paper presents an advanced approach for an accurate and self-optimizing model for location determination, including identification of optimized Landmark positions, which are used for probing. Moreover, the selection of correlated data and the estimated target location requires a sophisticated strategy to identify the correct position. We present an improved approximation of network distances of usually unknown TIER infrastructures using the road network. Our concept is evaluated under real-world conditions focusing Europe.