A hybrid genetic algorithm with Lamarckian individual learning for tasks scheduling.

Zhong Yi-wen,Yang Jian-gang

DOI: https://doi.org/10.1109/ICSMC.2004.1400858

2004-01-01

Abstract:Efficient application scheduling is critical for achieving high performance in parallel multiprocessor systems. The tasks scheduling problem is NP-hard in general. In order to obtain optimal or suboptimal solutions, a large number of scheduling heuristics have been presented in the literature. The most studied heuristics are based on list heuristics. In recent years, genetic algorithm (GA), as a power tool to achieve global optimal, has been successfully used in this field. This paper presents a new hybrid genetic algorithm to solve the tasks scheduling problem both for homogeneous and heterogeneous computing systems. It uses genetic algorithm to evolve tasks dispatching priority queue, and uses a list scheduling to decode the queue into a schedule. In order to remedy the GA's weakness in fine-tuning, this paper uses neighborhood search method to improve the fitness of the individuals of each generation, based on Lamarckian theory in the evolution. The simulation results comparing with two genetic algorithms and two list algorithms, both from the literature, show that this new GA produces encouraging results in terms of quality of solution and time complexity.