Discrete-event simulation has been widely used to analyze communication networks, due to the increased number of networks in existence, and the inability of mathematical modeling to handle their greater complexity. Although there is a large number of simulation packages designed specifically for modeling communication systems with a point-and-click GUI for entering network models and parameters, large networks modeling involving many objects, interconnections, and regular structures can be a very tedious and errorprone exercise. In this article, we present NAMS, an automated approach for communication networks modeling, simulation, and performance evaluation employing commercial simulation tools and programing interfaces. NAMS receives the configuration of the network to be modeled or simulated in a specific format, uses the built-in models for the key hardware devices and protocols in the network simulation tool, puts these models together to form a high level network model representing the received network configuration, compiles the network model into an executable simulation, generates the simulation environment, generates an appropriate performance evaluation methodology, executes the simulation, and reports the performance indices back to the system operator.