Improved Resource Efficient Allocation of IMA Applications to Multi-Cores

In [10] we proposed a static timing analysis for IMA applications on multi-core platforms. It produces safe bounds on tasks response times including memory interference delays and for which the number of state space to be explored does not explode. We also proposed a methodology for the automation of the software/hardware allocation which embeds the proposed timing analysis as a verification of the validity of the finally selected solution. However, some assumptions in [10] are not realistic in terms of software model, and the hardware model can be improved to reduce the pessimism of the timing analysis. We propose to improve the work in [10] so that it can better fit real avionic systems, and show that the new approach outperforms the one in [10]. The entire process eases the portability of IMA legacy software to multicore COTS without requiring any software or hardware modification.