Model-based deployment of mission-critical spacecraft applications on multicore processors

A variety of complex spacecraft applications, such as autonomous maneuvers based on image recognition, can benefit from the increased performance of multicore processors. On the other hand the redundant cores can also be used for fault-tolerance. Spacecraft missions more and more require a balanced trade-off between power, performance and reliability. Finding an optimal trade-off for each mission phase leads to new engineering challenges, especially regarding the efficient and safe deployment of software applications to hardware resources. We propose a model-based approach for the construction of software deployment schemes, and apply it to a spacecraft use case with two different mission phases to illustrate the benefits of such model-based software deployment.