Recent civil airborne platforms are produced using Integrated Modular Avionics (IMA). IMA promotes both sharing of execution and communication resources by the avionics applications. Designs following IMA decrease the weight of avionics equipment and improve the whole system scalability. However, the price to pay for these benefits is an increase of the system's complexity, triggering a challenging system integration process. Central to this integration step are the timing requirements of avionics applications: the system integrator has to find a mapping of applications and communications on the available target architecture (processing modules, networks, etc.) such as end-to-end delay constraints are met. These challenges stress the need for a tool capable of evaluating different integration choices in the early design stages of IMA. In this paper, we present and formalize the problem of spatial and temporal integration of an IMA system. Then, we focus on the temporal allocation problem which is critical to ensure a proper timely behavior of the system. Two main properties are presented to ensure perfect data transmission for hard real-time flows. To quantify the quality of a set of valid temporal allocations, CPM utilization and communication robustness performance criteria are defined. We show on an example that both criteria are antagonist and that they can be leveraged to choose an allocation that either improves the system computing performance or the robustness of the network.
[1]
Jérôme Ermont,et al.
Analyzing End-to-End Functional Delays on an IMA Platform
,
2010,
ISoLA.
[2]
Paul Prisaznuk,et al.
ARINC Specification 653, Avionics Application Software Standard Interface
,
2006,
Avionics.
[3]
Jérôme Ermont,et al.
Methods for bounding end-to-end delays on an AFDX network
,
2006,
18th Euromicro Conference on Real-Time Systems (ECRTS'06).
[4]
Christian Fraboul,et al.
End-to-end delay analysis in an Integrated Modular Avionics architecture
,
2013,
2013 IEEE 18th Conference on Emerging Technologies & Factory Automation (ETFA).
[5]
Christian Fraboul,et al.
Improving the Worst-Case Delay Analysis of an AFDX Network Using an Optimized Trajectory Approach
,
2010,
IEEE Transactions on Industrial Informatics.
[6]
Jérôme Ermont,et al.
Latency and freshness analysis on IMA systems
,
2011,
ETFA2011.
[7]
Mathieu Grenier,et al.
Configuring the communication on FlexRay - the case of the static segment ⁄
,
2008
.
[8]
Christian Fraboul,et al.
Modeling and simulation of integrated modular avionics
,
1998,
Proceedings of the Sixth Euromicro Workshop on Parallel and Distributed Processing - PDP '98 -.
[9]
Olivier Brun,et al.
Partition Scheduling on an IMA Platform with Strict Periodicity and Communication Delays
,
2010
.