PEGASE - A Robust and Efficient Tool for Worst-Case Network Traversal Time Evaluation on AFDX

2011-01-2711

10/18/2011

Event
Aerospace Technology Conference and Exposition
Authors Abstract
Content
Avionics systems distributed on AFDX networks are subject to stringent real-time constraints that require the system designer to have techniques and tools to guarantee the worst case traversal time of the network (WCTT) and thus ensure a correct global real-time behavior of the distributed applications/functions. The network calculus is an active research area based on the (min,+) algebra, that has been developed to compute such guaranteed bounds. There already exists several academics implementations but no up to date industrial implementation. To address this need, the PEGASE project gathers academics and industrial partners to provide a high quality, efficient and safe tool for the design of avionic networks using worst case performance guarantees.
The PEGASE software is an up-to-date software in the sense that it integrates the latest results of the theories, in tight cooperation with academics researchers. The PEGASE software is also a safe tool: it relies on a strong mathematical background, its algorithms are described in a formal document, and it shares unitary tests with some academics tools. The PEGASE software is a usable tool: in particular, it has been designed from the requirements expressed by embedded network engineers. Finally, the PEGASE software is an efficient tool: it provides bounds close to the actual real worst-case, avoiding over provisioning of resources. This paper presents the mathematical background of the tool, its architecture and some first results on realistic case studies.
Meta TagsDetails
DOI
https://doi.org/10.4271/2011-01-2711
Pages
11
Citation
Boyer, M., Migge, J., and Fumey, M., "PEGASE - A Robust and Efficient Tool for Worst-Case Network Traversal Time Evaluation on AFDX," SAE Technical Paper 2011-01-2711, 2011, https://doi.org/10.4271/2011-01-2711.
Additional Details
Publisher
Published
Oct 18, 2011
Product Code
2011-01-2711
Content Type
Technical Paper
Language
English