A delay deviation tolerance IP geolocation method with error estimation.

Area of Science:

• Computer Science
• Network Engineering
• Geographic Information Systems

Background:

• IP geolocation is crucial for location-based services, but reliable error estimation remains a challenge.
• Existing methods often struggle with accuracy due to network delay variations and lack robust error estimation.
• Accurate geolocation error estimation is vital for assessing the trustworthiness of location-based network service results.

Purpose of the Study:

• To propose a novel IP geolocation method with reliable error estimation capabilities for IP targets.
• To address the limitations of existing algorithms in accurately estimating geolocation errors.
• To improve the precision and reliability of IP target location estimation in real-world networks.

Main Methods:

• A metropolitan area network (MAN) topology is extracted using a training set of landmark paths.
• Router area centers are inferred by searching governed landmarks and calculating minimum covering circles.
• Geolocation errors are quantified using a verification path set, calculating a router area radius based on mean error and covering circle radius.
• IP target location is estimated by measuring its path and comparing it with the established MAN.

Main Results:

• The proposed method demonstrates improved error estimation accuracy compared to existing algorithms.
• Experimental results show finer geolocation granularity and a lower median error.
• The approach effectively handles delay inflation issues common in actual networks.

Conclusions:

• The novel IP geolocation method provides reliable error estimation and enhanced accuracy.
• This technique offers a significant improvement over traditional IP geolocation algorithms.
• The findings are validated through experiments across 12 cities in China.