A simulation execution manager for ns-3 encouraging reproducibility and simplifying statistical analysis of ns-3 simulations

Abstract

The typical workflow for ns-3 users consists of coming up with an experiment, translating that idea to simulation code, running multiple simulations, analyzing the outcomes, and finally plotting results. So far, the ns-3 project has not been providing tools to cover the steps from running simulations to obtaining plots: research teams typically develop their own custom solutions, and often need to learn new tools in order to reproduce results found in the literature. In this work we propose a framework that allows ns-3 users to go from their simulation script to plots in as few lines of code as possible, hiding tedious details about simulation running and result management, and leveraging Python’s widely established statistical analysis tools to quickly perform simulations, analyze their outcomes, and plot results. The code and its documentation, which have been in part developed under the Google Summer of Code 2018 program, are publicly available at [6, 7].