Allocating Test for Real-Time task in Heterogeneous Distributed Computing Systems

E. FERRO; R. CAYSSIALS; O. ALIMENTI; JAVIER OROZCO

Benidorm, España

Congreso; 9th IASTED International Conference Artificial Intelligence and Soft Computing; 2005

IASTED

This paper addresses the problem of allocating and scheduling periodic task to processing nodes in distributed Real-Time systems subject to task precedence and timing constraints. We proposed three techniques to find an optimal allocation in real time environments that minimizes the real-time traffic over the network and balances the processor utilization factor. First of them is the Tabu Search technique, secondly the Evolutionary Computation and finally the Evolutionary Computation tuned by Fuzzy Logic. Tasks interchange data among them and they are entirely executed in heterogeneous processors connected through a communications channel. There are placement, schedulability, communications and precedence constraints that have to be verified in order to validate a tentative assignment. Several examples are present to show the effectiveness and practicality of all proposed algorithms.