Towards network simulations credibility: Lessons from applying five key principles

Simulation is often used in evaluating network protocols in cases where implementation would be too costly or too risky. There have recently been questions raised as to whether wireless network simulations correctly model the real world. As part of a research effort to provide improved communications capabilities in maritime tactical networks, a resource reservation service (RRS) has been designed specifically for the maritime networks. In order to evaluate the protocol, simulation models of the protocol and the maritime environment were developed. In this paper, we address our efforts to ensure that our simulation results are as credible as possible. Based on the latest research, five principles for credible simulation are proposed and then applied, as an example, to our work on analysing the RRS.