DocumentCode :
412594
Title :
Using an evolutionary algorithm for bandwidth minimization
Author :
Lim, Andrew ; Rodrigues, Brian ; Xiao, Fei
Author_Institution :
Dept. of Ind. Eng. & Eng. Manage., Hong Kong Univ. of Sci. & Technol., Kowloon, China
Volume :
1
fYear :
2003
fDate :
8-12 Dec. 2003
Firstpage :
678
Abstract :
In this paper, we propose an integrated genetic algorithm with hill climbing to solve the matrix bandwidth minimization problem, which is to reduce bandwidth by permuting rows and columns resulting in the nonzero elements residing in a band as close as possible to the diagonal. Many algorithms for this problem have been developed, including the well-known CM and GPS algorithms. Recently, Marti et al., (2001) used tabu search and Pinana et al. (2002) used GRASP with path relinking, separately, where both approaches outperformed the GPS algorithm. In this work, our approach is to exploit the genetic algorithm technique in global search while using hill climbing for local search. Experiments show that this approach achieves the best solution quality when compared with the GPS algorithm, tabu search, and the GRASP with path relinking methods, while being faster than the latter two newly-developed heuristics.
Keywords :
bandwidth compression; computational complexity; genetic algorithms; matrix algebra; minimisation; reachability analysis; search problems; CM algorithm; GPS algorithm; GRASP; bandwidth reduction; evolutionary algorithm; genetic algorithm; global search; heuristics; hill climbing; local search; matrix bandwidth minimization problem; matrix permutation; path relinking; tabu searching; Bandwidth; Computer science; Engineering management; Evolutionary computation; Global Positioning System; Industrial engineering; Labeling; Minimization methods; Symmetric matrices; Technology management;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Evolutionary Computation, 2003. CEC '03. The 2003 Congress on
Print_ISBN :
0-7803-7804-0
Type :
conf
DOI :
10.1109/CEC.2003.1299641
Filename :
1299641
Link To Document :
بازگشت