Philipp Ittershagen and Kim Grüttner and Wolfgang Nebel
ACM Transactions on Design Automation of Electronic Systems (TODAES)
The rise of mixed-critical embedded systems imposes novel challenges on the specification, development, and functional validation in a design flow. In the emerging dynamic scheduling context of mixed-criticality platforms, the system behaviour needs to be estimated in an early step in the design flow to assess the integration impact, especially for quality of service-driven, low-critical subsystems. We provide a modelling and integration flow for specifying, estimating, and evaluating software functions, ranging from an initial executable specification to an implementation candidate on an MPSoC. Based on a data-driven model to evaluate dynamic resource consumption effects of high-critical subsystems and the scheduling overhead, we propose a systematic method for constructing workload models of high-critical software components on the target. Our proxies provide an integration environment for low-critical functions by mimicking the high-critical temporal behaviour on the target. By integrating a low-critical video encoding subsystem with a benchmark suite as the high-critical subsystem we show that the performance model allows for evaluating end-to-end execution times in the low-critical function with an average error of 0.37 % and the application proxy only introduces a maximum error of 1.14 % in a performance evaluation.
5 / 2018
Embedded Multi-Core systems for Mixed criticality applications in dynamic and changeable real-time environments