Session
Poster Session 1
Abstract
On the basis of the small satellite Flying Laptop and its derived platform FLP2, this paper introduces a generic methodology of simulating a satellite on-board computer within an on-board data-handling network. The concept has been developed in order to take on some of the challenges towards a reference testing facility for small satellites. Although the development of a reference testing facility has been the initial motivation, this work does not intend to provide a full concept of such an architecture in any kind, but detail solutions to specific problems down that road. The first challenge is to bind the controller-in-the-loop as close as possible to the satellite system simulator without unnecessary interfaces in between. This means that an appropri- ate simulator needs to provide all required interfaces, as it has to simulate all data-handling devices that are connected to the on-board computer in some way. Unfortunately, the work- ing principle of a state-of-the-art satellite system simulator would causes latencies that rule out a proper simulation of an on-board data-handling, which is why a faster solution must be found here. The second challenge is the derivation of a generic modeling approach for data-handling devices, which supports portability across different simulation projects. A solution here is restricted though by the variety of hardware architectures and protocols especially on trans- port and application layer, which is why this work sticks to the application of SpaceWire and the Remote Memory Access Protocol. Verification of developed models is the third and last challenge met in the scope of this paper. Models of a satellite system simulator are implemented mostly in an object-oriented fashion. Testing object-oriented software does not work straightforward as a matter of principle. It is even more comprehensive, when the respective models are multi-threaded. Therefore, a profound approach of testing these data-handling models must be introduced.
Development and Testing of Fast and Portable Data-Handling Models in a Synchronous Small Satellite Simulator
On the basis of the small satellite Flying Laptop and its derived platform FLP2, this paper introduces a generic methodology of simulating a satellite on-board computer within an on-board data-handling network. The concept has been developed in order to take on some of the challenges towards a reference testing facility for small satellites. Although the development of a reference testing facility has been the initial motivation, this work does not intend to provide a full concept of such an architecture in any kind, but detail solutions to specific problems down that road. The first challenge is to bind the controller-in-the-loop as close as possible to the satellite system simulator without unnecessary interfaces in between. This means that an appropri- ate simulator needs to provide all required interfaces, as it has to simulate all data-handling devices that are connected to the on-board computer in some way. Unfortunately, the work- ing principle of a state-of-the-art satellite system simulator would causes latencies that rule out a proper simulation of an on-board data-handling, which is why a faster solution must be found here. The second challenge is the derivation of a generic modeling approach for data-handling devices, which supports portability across different simulation projects. A solution here is restricted though by the variety of hardware architectures and protocols especially on trans- port and application layer, which is why this work sticks to the application of SpaceWire and the Remote Memory Access Protocol. Verification of developed models is the third and last challenge met in the scope of this paper. Models of a satellite system simulator are implemented mostly in an object-oriented fashion. Testing object-oriented software does not work straightforward as a matter of principle. It is even more comprehensive, when the respective models are multi-threaded. Therefore, a profound approach of testing these data-handling models must be introduced.
