Estimating hidden network topology using hop matrix

Abstract

Estimating network topology is a non-trivial task. Different from the state-of-art methods such as traceroute-like active probing, network tomography and network protocols analysis based passive measurement, this work proposes a novel method to estimate the network hidden topology using time to live (TTL) values extracted from IP packets. TTL contains the information about hops of the forward paths which can be fused to reconstruct the topology. The proposed estimating algorithm is described and its performance is illustrated using topology simulation data. Experimental results demonstrate the effectiveness of the newly developed technique.