Between Small Complex and Large Simplistic Network Simulators

Abstract

Presently, discrete-even network simulation market is split among a number of popular simulators, such as NS-x, OM-NET++, Opnet, and maybe a few others. These tools are very common in research that targets small to medium size networks. Whenever one needs to simulate a large network of many tens of thousands nodes, the ability to use traditional network simulators is questionable. Since performance, and specifically, execution speed, is the number one priority in traditional network simulation, simulation runs are normally implemented as solid executables running in a memory space, to which there is a physical limit. This paper proposes a different paradigm of simulation that specifically aims at the ability to simulate large networks while supporting models in nodes with any level of complexity.