Title :
A neural heuristic for access network planning
Author :
Amoza, Franco Robledo
Author_Institution :
Dept. de Investig. Operativa, Univ. de la Republica, Montevideo
fDate :
March 30 2009-April 2 2009
Abstract :
The greedy randomized adaptive search procedure (GRASP) is a well-known metaheuristic for combinatorial optimization. In this paper, we introduce a GRASP for designing the access network topology of a wide area network (WAN). This problem is NP-hard, and can be modeled as a variant of the steiner problem in graphs. The proposed GRASP employs a random neural network (RNN) model in the local search phase, in order to improve the solutions delivered by the construction phase. Experimental results were obtained on 210 problem instances of different topological characteristics, generated using the problem classes in the SteinLib repository, and with known lower bounds for their optima. The algorithm obtained good results, with low average gaps with respect to the lower bounds in most of the problem classes, and attaining the optimum in 48 cases (more than 22% of the test-set).
Keywords :
computational complexity; greedy algorithms; neural nets; search problems; wide area networks; NP-hard problem; access network planning; combinatorial optimization; greedy randomized adaptive search procedure; local search phase; neural heuristic; random neural network; wide area network; Algorithm design and analysis; Character generation; Costs; Network topology; Neural networks; Recurrent neural networks; Spine; Switches; Testing; Wide area networks;
Conference_Titel :
Hybrid Intelligent Models and Applications, 2009. HIMA '09. IEEE Workshop on
Conference_Location :
Nashville, TN
Print_ISBN :
978-1-4244-2758-1
DOI :
10.1109/HIMA.2009.4937822