An Airspace Simulator for Separation Management Research

Air Traffic Management (ATM) systems are undergoing a period of major transformation and modernisation, requiring and enabling new separation management (SM) methods. Many novel SM functions, roles and concepts are being explored using ATM simulators. Commercial simulators are capable, high-fidelity tools, but tend to be complex and inaccessible. The Airspace Simulator is a fast-time, discrete event simulator originally designed for exploratory ATM research. This thesis describes the redevelopment of the Airspace Simulator into a simulation platform better suited for researching and evaluating SM in future airspace. The Airspace Simulator-II has the advantage of new functionality and greater fidelity, while remaining high-speed, accessible and readily adaptable. The simulator models FMS-like spherical earth navigation and autopilot flight control with an average cross track error of 0.05 nmi for waypoint-defined routes in variable wind-fields. Trajectories are computed using the BADA v3.8 tabulated database to model the performance of 318 aircraft types. The simulator was demonstrated with up to 4000 total aircraft, and trajectories for 300 simultaneous aircraft were computed over 900 times faster than real-time. Datalink and radio-telephony communications are modelled between the air traffic and ATM systems. Surveillance is provided through ADS-B-like broadcasts, and an algorithm was developed to automatically merge instructions from conflict resolution systems with existing flight plans. Alternate communication, navigation, and separation modes were designed to permit the study of mixed-mode operations. Errors due to wind, navigational wander, communication latencies, and localised information states are modelled to facilitate research into the robustness of SM systems. The simulator incorporates a traffic visualisation tool and was networked to conflict detection and resolution software through a TCP/IP connection. A scenario generator was designed to automatically prepare flight plans for a large variety of two-aircraft encounters to support stochastic SM experiments. The simulator, scenario generator, and resolver were used for the preliminary analysis of a novel concept for automated SM over radio-telephony using progressive track angle vectoring.