Letras M, Morales-Reyes A, Cumplido R (2016)
Publication Type: Journal article, Original article
Publication year: 2016
Pages Range: 899 - 910
URI: http://www.sciencedirect.com/science/article/pii/S0925231215015933
DOI: 10.1016/j.neucom.2015.05.128
Architectures design for Genetic Algorithms (GAs) has proved its effectiveness to tackle hard real-time constrained problems that require an optimization mechanism in one of their phases. Most of these approaches are problem dependent and cannot be easily adapted to other problems. Moreover, GAs based architectures preserve the algorithmic structure of a panmictic population in a sequential GA and therefore they are similar to a software implementation. Recently, a combination of GAs, both sequential and parallel and reconfigurable devices such as FPGAs have been merged to create GAs based parallel hardware architectures. This study proposes a novel hardware architectural framework that implements a fine-grained or cellular GAs while maintaining a toroidal connection among individuals within the population. Achieving massive parallelism is limited by available resources; therefore, the proposed architectural design implements a segmentation strategy that partitions the entire decentralized population while maintaining original algorithmic interaction among solutions. The proposed architecture aims at preserving fine-grained GAs algorithmic structure while improving resources usage. It also allows flexibility in terms of population and solutions representation size and the evaluation module containing the objective function is interchangeable.
APA:
Letras, M., Morales-Reyes, A., & Cumplido, R. (2016). A scalable and customizable processor array for implementing cellular genetic algorithms. Neurocomputing, 899 - 910. https://doi.org/10.1016/j.neucom.2015.05.128
MLA:
Letras, Martin, Alicia Morales-Reyes, and René Cumplido. "A scalable and customizable processor array for implementing cellular genetic algorithms." Neurocomputing (2016): 899 - 910.
BibTeX: Download