Distributed Flight Simulation Architecture using a PC/Linux Cluster

Until recently, real-time, high fidelity flight simulation architectures could only be realized through expensive proprietary hardware and software. High-end specialized computer systems have provided the only platforms available to satisfy the math modeling and 3D graphics requirements for rotorcraft handling qualities, crew systems, and training. In many cases, the software architectures of these computer systems are equally as specialized. Compounding already expensive simulation architecture are the maintenance costs incurred over the life cycle of the specialized hardware and software. The potential appeal of inexpensive open source simulation architectures, while still maintaining real-time characteristics, is undeniable versus the expensive proprietary methods described. Today it is the belief of many computer scientists that this inexpensive open source architecture is capable of both achieving and exceeding the goals of any existing and future simulation. This can be realized through a distributed computing environment using high-end PCs, running a flavor of both Windows and Linux operating systems, all communicating via a gigabit Ethernet infrastructure. This paper describes an inexpensive open source PC/Linux-Ethernet flight simulation architecture which is currently being used to support the requirements of engineering and marketing. Linux was the operating system of choice since its Unix-like quality makes transitioning current simulation applications from the high-end workstation environment to a PC platform relatively straightforward. This paper will focus on the specific porting issues of current systems, such as graphics capabilities and performance under Linux, and also more general issues including ease of installation, availability of source files, cost and timing analyses, and available commercial off-the-shelf (COTS) packages and components to support transition and maintenance activities. Obstacles encountered during the porting effort will be identified and discussed.