Solution for real-time scheduling in distributed systems using genetic algorithms

Damian Cristian SILIȘTEANU1, Bogdan Costel MOCANU1, Mihnea Horia VREJOIU2, Florin POP2,1
1 Faculty of Automatic Control and Computer Science, Politehnica University of Bucharest
2 National Institute for Research and Development in Informatics – ICI Bucharest,,, /

Abstract: In recent times, distributed systems have demonstrated their efficiency in processing a large number of tasks in as little time as possible. In this context, the scheduler has the most influence because it manages how the respective tasks are processed on various resources. We set out to treat a planning problem with tasks and resources as main entities. Between these two entities we established certain constraints to model a real problem, and to solve the planning problem we opted for the development of a genetic algorithm. To evaluate the performance of the algorithm, we analyzed how it manages to reduce the processing time and at the same time fulfill the set constraints. We also analyzed the impact of adding additional resources in the schedule for the same tasks.

Keywords: distributed systems, tasks, resources, scheduler, constraints, optimization, genetic algorithms.

View full text

Damian Cristian SILIȘTEANU, Bogdan Costel MOCANU, Mihnea Horia VREJOIU, Florin POP, Solution for real-time scheduling in distributed systems using genetic algorithms, Romanian Journal of Information Technology and Automatic Control, ISSN 1220-1758, vol. 32(3), pp. 33-50, 2022.