Tuning SoC platforms for multimedia processing: identifying limits and tradeoffs
Proceedings of the 2nd IEEE/ACM/IFIP international conference on Hardware/software codesign and system synthesis
Abstracting functionality for modular performance analysis of hard real-time systems
Proceedings of the 2005 Asia and South Pacific Design Automation Conference
Probabilistic performance risk analysis at system-level
CODES+ISSS '07 Proceedings of the 5th IEEE/ACM international conference on Hardware/software codesign and system synthesis
Cyclic dependencies in modular performance analysis
EMSOFT '08 Proceedings of the 8th ACM international conference on Embedded software
State-based scheduling with tree schedules: analysis and evaluation
Real-Time Systems
Hi-index | 0.00 |
Many real-time embedded systems process event streamswhich are composed of a finite number of different eventtypes. Each different event type on the stream would typicallyimpose a different workload to the system, and thusthe knowledge of possible correlations and dependenciesbetween the different event types could be exploited to gettighter analytic performance estimations of the completesystem. We propose an abstract stream model to characterizesuch an event stream. The model captures the needed informationof all possible traces of a class of event streamsand can hence be used to obtain hard bounded worst-caseand best-case estimations of a system. We show how theproposed abstract stream model can be obtained from aconcrete stream specification, and how it can be used forperformance analysis. The applicability of our approachand its advantages over traditional worst-case performanceanalysis are shown in a case study of a multimedia application.