DocumentCode :
2665214
Title :
Solving a Ring Star Problem Generalization
Author :
Mauttone, Antonio ; Nesmachnow, Sergio ; Oivera, A. ; Amoza, Franco Robledo
Author_Institution :
Comput. Sci. Inst., Univ. de la Republica, Montevideo, Uruguay
fYear :
2008
fDate :
10-12 Dec. 2008
Firstpage :
981
Lastpage :
986
Abstract :
The Capacitated m-Ring Star Problem (CmRSP) consists of finding a set of m rings, each of them including the central depot, a subset of customers, and a set of optional nodes used to diminish the costs of the network design. The rings must be node-disjoint (except for the central depot) in order to provide node-survivability to the network. Customers that are not part of the rings must be directly connected to nodes in the rings (as pendants). An additional constraint is that no ring (the cycle itself and pendants) can have more than Q customers. The objective is to minimize the sum of routing and connection costs. The CmRSP is NP-Hard and has important applications in the metropolitan optical fiber network planning. This work presents a metaheuristic approach to solve the CmRSP. The algorithm was tested over a set of 90 benchmark instances accomplishing in all cases optimal or near-optimal solutions.
Keywords :
combinatorial mathematics; optical fibre networks; optimisation; search problems; telecommunication network planning; NP-hard problem; capacitated m-ring star problem; combinatorial optimization; connection costs; metropolitan optical fiber network planning; network design; networks survivability; ring star problem generalization; tabu search; telecommunication networks design; Benchmark testing; Communication networks; Computer science; Costs; NP-hard problem; Optical fiber networks; Routing; Telephony; Topology; Traveling salesman problems; GRASP; Tabu Search; combinatorial optimization; ring star problem;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational Intelligence for Modelling Control & Automation, 2008 International Conference on
Conference_Location :
Vienna
Print_ISBN :
978-0-7695-3514-2
Type :
conf
DOI :
10.1109/CIMCA.2008.160
Filename :
5172759
Link To Document :
بازگشت