Sequential and Parallel Cellular Automata-Based Scheduling Algorithms
IEEE Transactions on Parallel and Distributed Systems
IEEE Transactions on Parallel and Distributed Systems
SBRN '06 Proceedings of the Ninth Brazilian Symposium on Neural Networks
| Hi-index | 0.00 |
Static Task Scheduling Problem (STSP) in multiprocessors is a NP-Complete problem. HLFET is one of the simplest well-known heuristics designed to deal with it. Meta-heuristics like genetic algorithms and simulated annealing had also been applied to this problem. Cellular Automata (CA) have been recently used to solve STSP. The main feature of CA-based scheduling is the extraction of knowledge while scheduling an application and its subsequent reuse in other instances. An evolutionary algorithm is applied to search for efficient CA rules in learning phase. Previous works showed this approach is promising. However some desirable features have not been successfully exploited yet, such as: (i) the massive parallelism inherent to CA, (ii) the usage of an arbitrary number of processors and (iii) the reuse of evolved rules with competitive results. This paper presents a new model called SCAS-H (Synchronous Cellular Automata Scheduler initialized by Heuristics). Its major innovation is the usage of a heuristic based on HLFET to start up the CA rule evolution. Program graphs found in literature and others randomly generated were used in the experiments. Results show that SCAS-H overcame related models both in scheduling results as computational performance and it presented competitive results with meta-heuristics.