TY - GEN
T1 - A hybrid genetic algorithm/variable neighborhood search approach to maximizing residual bandwidth of links for route planning
AU - Kandavanam, Gajaruban
AU - Botvich, Dmitri
AU - Balasubramaniam, Sasitharan
AU - Jennings, Brendan
PY - 2010
Y1 - 2010
N2 - This paper proposes a novel approach to performing residual bandwidth optimization with QoS guarantees in multi-class networks. The approach combines the use of a new highly scalable hybrid GA-VNS algorithm (Genetic Algorithm with Variable Neighborhood Search) with the efficient and accurate estimation of QoS requirements using empirical effective bandwidth estimations. Given a QoS-aware demand matrix, experimental results indicate that the GA-VNS algorithm shows significantly higher success rate in terms of converging to optimum/near optimum solution in comparison to pure GA and another combination of GA and local search heuristic, and also exhibits better scalability and performance. Additional results also show that the proposed solution performs significantly better than OSPF in optimizing residual bandwidth in a medium to large sized network.
AB - This paper proposes a novel approach to performing residual bandwidth optimization with QoS guarantees in multi-class networks. The approach combines the use of a new highly scalable hybrid GA-VNS algorithm (Genetic Algorithm with Variable Neighborhood Search) with the efficient and accurate estimation of QoS requirements using empirical effective bandwidth estimations. Given a QoS-aware demand matrix, experimental results indicate that the GA-VNS algorithm shows significantly higher success rate in terms of converging to optimum/near optimum solution in comparison to pure GA and another combination of GA and local search heuristic, and also exhibits better scalability and performance. Additional results also show that the proposed solution performs significantly better than OSPF in optimizing residual bandwidth in a medium to large sized network.
UR - http://www.scopus.com/inward/record.url?scp=77954709882&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-14156-0_5
DO - 10.1007/978-3-642-14156-0_5
M3 - Conference contribution
AN - SCOPUS:77954709882
SN - 3642141552
SN - 9783642141553
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 49
EP - 60
BT - Artificial Evolution - 9th International Conference Evolution Artificielle, EA 2009, Revised Selected Papers
T2 - 9th International Conference on Artificial Evolution, EA 2009
Y2 - 26 October 2009 through 28 October 2009
ER -