Implementation and Evaluation of Flow-level Network Simulator for Large-scale ICN Networks

Abstract

In recent years, ICN (Information-Centric Networking) that focuses on the data being transferred, rather than hosts exchanging the data, has been attracting attention as one of the promising next-generation Internet architectures. It has developed that fluid model of large-scale ICN networks, which is aimed at analyzing the performance of transport layer protocols in ICN networks. In this paper, we present a flow-level ICN sim-ulator called FICNSIM (Fluid-based ICN SIMulator), which is based on the numerical solver for ICN fluid models. In particular, we introduce two types of FICNSIM implementations: a highly customizable implementation in the Python language and a high-performance implementation in the Julia language. Furthermore, through several experiments, we evaluate the effectiveness of our FICNSIM implementation. Consequently, we show that our implemented FICNSIM can perform a high-speed simulation execution compared to a conventional packet-level ICN simulator.