Retrospective IP Address Geolocation for Geography-Aware Internet Services.

Area of Science:

• Computer Science
• Internet Engineering
• Network Security

Background:

• Retrospective geolocation of IP addresses is challenging due to dynamic internet infrastructure and limited historical location data.
• Understanding past IP address locations is crucial for various analytical and security applications.

Purpose of the Study:

• To analyze the accuracy and feasibility of retrospective geolocation for IPv4 and IPv6 addresses over a five-year period.
• To develop and introduce a method for handling missing historical IP geolocation databases.
• To propose a parametric model for estimating IP address location lifetime.

Main Methods:

• Analysis of retrospective geolocation data for IPv4 and IPv6 addresses spanning five years.
• Development of an approach to address gaps in historical IP geolocation databases.
• Implementation of a parametric model to estimate the timing of past IP address location changes.

Main Results:

• Retrospective IP address geolocation is reliable for up to a few years, with observed differences between IPv4 and IPv6.
• The proposed model effectively estimates the duration of IP address locations in the past.
• The study validates the utility of retrospective geolocation in practical scenarios.

Conclusions:

• Retrospective IP address geolocation is a viable technique for historical analysis, particularly within a few years' scope.
• The developed methods and models enhance the reliability and applicability of historical IP location data.
• Findings support applications in social studies, system analysis, security investigations, geotargeted content, and e-commerce identity theft prevention.