Can Vivaldi help in IP geolocation?

Abstract

The paper deals with IP geolocation based on communication latency measurement. The aim of IP geolocation is to estimate the geographical location of an IP-enabled node. Latency-based IP geolocation methods measure latency from a set of landmarks with the known geographical position to a target with an unknown position. When the latency values are known, the target position is estimated using multilateration. A disadvantage is that for each target’s position estimation, a new latency measurement is required. In order to avoid this, it has been proposed to employ a latency prediction method, such as Vivaldi, to predict the latency between a target and a landmark and, thus, reduce the number of latency measurements. In this paper, we investigate this proposal in terms of location accuracy and efficiency. The conclusion of the paper gives an indicative answer about the credibility of Vivaldi for its use in IP geolocation.