Title: Particle swarm optimisation of memory usage in embedded systems

Authors: Jose L. Risco-Martin, Oscar Garnica, Juan Lanchares, J. Ignacio Hidalgo, David Atienza

Addresses: Department of Computer Architecture and Automation, Facultad de Informatica, Universidad Complutense de Madrid, C/Prof. Jose Garcia Santesmases, s/n, 28040 Madrid, Spain. ' Department of Computer Architecture and Automation, Facultad de Informatica, Universidad Complutense de Madrid, C/Prof. Jose Garcia Santesmases, s/n, 28040 Madrid, Spain. ' Department of Computer Architecture and Automation, Facultad de Informatica, Universidad Complutense de Madrid, C/Prof. Jose Garcia Santesmases, s/n, 28040 Madrid, Spain. ' Department of Computer Architecture and Automation, Facultad de Informatica, Universidad Complutense de Madrid, C/Prof. Jose Garcia Santesmases, s/n, 28040 Madrid, Spain. ' Embedded Systems Laboratory (ESL), Ecole Polytechnique Federale de Lausanne (EPFL), Station 11, ESL-IEL-STI-EPFL, 1015 – Lausanne, Switzerland; Department of Computer Architecture and Automation, Facultad de Informatica, Universidad Complutense de Madrid, C/Prof. Jose Garcia Santesmases, s/n, 28040 Madrid, Spain

Abstract: In this paper, we propose a dynamic, non-dominated sorting, multi-objective particle-swarm-based optimiser, named hierarchical non-dominated sorting particle swarm optimiser (H-NSPSO), for memory usage optimisation in embedded systems. It significantly reduces the computational complexity of others multi- objective particle swarm optimisation (MOPSO) algorithms. Concretely, it first uses a fast non-dominated sorting approach with O(mN2) computational complexity. Second, it maintains an external archive to store a fixed number of non-dominated particles, which is used to drive the particle population towards the best non-dominated set over many iteration steps. Finally, the proposed algorithm separates particles into multi sub-swarms, building several tree networks as the neighbourhood topology. H-NSPSO has been made adaptive in nature by allowing its vital parameters (inertia weight and learning factors) to change within iterations. The method is evaluated using two real world examples in embedded applications and compared with existing covering methods.

Keywords: embedded systems; dynamic memory optimisation; particle swarm optimisation; PSO; high performance systems; multi-objective optimisation; evolutionary computation.

DOI: 10.1504/IJHPSA.2008.024205

International Journal of High Performance Systems Architecture, 2008 Vol.1 No.4, pp.209 - 219

Published online: 29 Mar 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article