In distributed Hard-Real-Time systems the correctness of a computation strictly depends on the results and on the computing time to produce them. We introduce a mathematical model to check the schedulability of a set of parallel programs represented through Data Flow precedence graphs. Schedulability checking involves a mapping subproblem, which is NP-complete. We transform the schedulability problem into an optimization one and present a solution based on Stochastic Neural Networks.
Schedulability checking in real-time systems using neural networks
Davoli Renzo;Tamburini Fabio;
1995
Abstract
In distributed Hard-Real-Time systems the correctness of a computation strictly depends on the results and on the computing time to produce them. We introduce a mathematical model to check the schedulability of a set of parallel programs represented through Data Flow precedence graphs. Schedulability checking involves a mapping subproblem, which is NP-complete. We transform the schedulability problem into an optimization one and present a solution based on Stochastic Neural Networks.File in questo prodotto:
Eventuali allegati, non sono esposti
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
