Simulating and analyzing wireless railway control networks using NS-2

The research presented in this dissertation investigates wireless communication protocols used by the rail industry. Several wireless railway control protocols are analyzed with a specific focus on security issues. Suggestions for improvements and vulnerability mitigations that could be implemented within those legacy technologies are provided. The research then focuses on the Advanced Train Control System (ATCS) protocol and provides a detailed analysis of that protocol, along with modeling experiments and suggestions for improvement.;ATCS is a popular protocol for wireless control and data transmission in the rail industry. The latter half of the dissertation details the creation of a framework, written and created by the author, for modeling and simulating the ATCS protocol. The ATCS simulator was based on the popular Network Simulator 2 (NS-2) program, providing a familiar environment to those already familiar with NS-2, but very little of the original NS-2 system is used in the ATCS simulator (less than 20%).;The ATCS simulator models wireless dynamic network routing based on three types of transmitters found in ATCS networks. Simulations account for traffic control, packet priority, packet acknowledgement, packet collision, and signal propagation based on detailed terrain elevation maps. Multiple simulation runs can be graphed together to show the results when network configurations are changed. Results can also be loaded into industry-standard ATCS protocol visualization tools like ATCSMon.;Using the ATCS simulator, the author modeled communications over a section of track in Illinois used in the North American Joint Positive Train Control (NAJPTC) project. This model was used to predict network traffic loads, packet timings, packet loss, and signal strength. The simulator was also used to predict the outcomes of NAJPTC base station failures and relocations. Modeling relocations provided a mechanism for improving the NAJPTC project. Original base station locations were fed into a Monte Carlo process where each station was randomly moved and the signal strength was recorded by receiving stations. The results show that the original base station locations were not optimal for NAJPTC performance. Thus, the ATCS simulator has proven to be an effective modeling tool and the research presented in this dissertation shows the system's viability in designing and configuring wireless railway control networks.