Performance prediction methods for address lookup algorithms of IP routers

Abstract

Many address lookup methods for an IP router have been recently proposed to improve a packet forwarding capability, but their performance prediction is very limited because of lack of considering actual traffic characteristics. It is necessary to consider actual traffic to predict more realistic performances on routers, specially in case of layers 3 and 4 switches whose performances are more influenced by flow characteristics. In this paper, we propose new methods for predicting the router’s performance based on the statistical analysis of the Internet traffic. We also present an example of its application to the existing table lookup algorithms, and show that simulation results based on our method can provide accurate performance prediction.