NetLogo Meets Discrete Event Simulation

Abstract

NetLogo and agent-based models more generally are traditionally implemented as time-stepped simulations. This means that everything is done at every increment of time. While this can be necessary at times, a potentially far more efficient simulation method is known as event-driven simulation. This type of simulation uses a scheduler for the times particular events will take place. Rather than doing everything every time step the simulation moves forward in time based upon the need for an event to occur. In this paper we show how to use NetLogo Tables and Arrays to create a time ordered queue and develop a simple NetLogo model implemented both as a traditional time-stepped simulation and as an event-driven simulation. We also present performance results.