A biased random-key genetic algorithm for OSPF and DEFT routing to minimize network congestion

Reis, Roger; Ritt, Marcus; Buriol, Luciana S.; Resende, Maurício G. C.

doi:10.1111/j.1475-3995.2010.00771.x

Cited by 27 publications

(12 citation statements)

References 18 publications

(45 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Among meta-heuristics, BRKGA, a class of GA, has been recently proposed to effectively solve optimization problems, in particular, network related problems such as routing in IP networks and RWA in optical networks [9,[16][17][18]. Compared with other meta-heuristics, BRKGA has provided better solutions in shorter running times.…”

Section: Heuristic Algorithmsmentioning

confidence: 99%

“…The maximum amount of traffic routed through a slot is computed in constraint (15), whilst constraint (16) provides the slot with enough bit-rate and constraint (17) ensures that only one port is equipped per slot. The required switching capacity of each node is computed in constraint (18), whereas the required capacity in number of ports is obtained by constraint (19).…”

Section: Sd(d)mentioning

confidence: 99%

See 1 more Smart Citation

Survivable IP/MPLS-Over-WSON Multilayer Network Optimization

Ruiz

Pedrola

Velasco

et al. 2011

J. Opt. Commun. Netw.

View full text Add to dashboard Cite

Abstract-Network operators are facing the problem of dimensioning their networks for the expected huge IP traffic volumes while keeping constant or even reducing the connectivity prices. Therefore, new architectural solutions able to cope with the expected traffic increase in a more cost-effective way are needed. In this work, we study the survivable IP/multi-protocol label switching (MPLS) over wavelength switched optical network (WSON) multilayer network problem as a capital expenditure (CAPEX) minimization problem. Two network approaches providing survivability against optical links, IP/MPLS nodes, and opto-electronic port failures are compared: the classical overlay approach where two redundant IP/MPLS networks are deployed, and the new joint multilayer approach which provides the requested survivability through an orchestrated interlayer recovery scheme which minimizes the over-dimensioning of IP/MPLS nodes. Mathematical programming models are developed for both approaches. Solving these models, however, becomes impractical for realistic networks. In view of this, evolutionary heuristics based on the biased random-key genetic algorithm framework are also proposed. Exhaustive experiments on several reference network scenarios illustrate the effectiveness of the proposed approach in minimizing network CAPEX.Index Terms-Integer linear programming; Multilayer planning; Survivable multilayer networks.

show abstract

Section: Heuristic Algorithmsmentioning

confidence: 99%

Section: Sd(d)mentioning

confidence: 99%

Survivable IP/MPLS-Over-WSON Multilayer Network Optimization

Ruiz

Pedrola

Velasco

et al. 2011

J. Opt. Commun. Netw.

View full text Add to dashboard Cite

show abstract

“…Como o tempo de computação dessa instância é significativamente maior, uma atribuição de fatias de iterações em tempo de compilação associada a grandes fatias é capaz de resultar em desempenhos melhores. Essa tendência pode ser também identificada para a instância c100, que tem o maior speedup com a política static (2,26). Em relação ao tamanho das fatias de iterações para a instância c100, a granularidade fina resultou no melhor speedup para a política static, enquanto que, para as políticas dynamic e guided a granularidade grossa foi mais eficiente (speedups de 2,23 e 2,22, respectivamente).…”

Section: Impacto Da Paralelização No Tempo De Execuçãounclassified

O impacto da paralelização com OpenMP no desempenho e na qualidade das soluções de um Algoritmo Genético

Gressler

Cristina

2014

Revista Brasileira De Computação Aplicada

View full text Add to dashboard Cite

Resumo: O Problema do Roteamento de Veículos (PRV) é um problema combinatório de difícil solução, aplicável tanto para logística de empresas de transporte quanto para melhor ocupação das vias públicas. Resolvê-lo testando todas as combinações possíveis (método de força bruta) torna-se inviável à medida que o problema escala, pois demanda um tempo de computação muito grande. Os Algoritmos Genéticos (AG) são meta-heurísticas capazes de encontrar soluções em um tempo computacional aceitável. Entretanto, mesmo os AG podem demandar um elevado tempo de processamento, dependendo das configurações utilizadas. Com a evolução das arquiteturas computacionais e a difusão das arquiteturas multicore, o uso da programação multithread torna-se uma alternativa para reduzir o tempo envolvido na solução de problemas combinatórios. Este artigo objetiva acelerar a resolução do PRV por meio da paralelização do AG com OpenMP, que é um padrão amplamente difundido para programação multithread. Nossos resultados atingiram um speedup acima de 2, utilizando 4 threads em um processador quadcore. Esse ganho está limitado à forma como o AG está implementado. Além do impacto no desempenho do AG também comprovou-se que o uso do OpenMP não afeta a qualidade das soluções. Adicionalmente, o uso do OpenMP permitiu que o AG encontrasse melhores soluções devido ao aumento do número de evoluções computadas num mesmo intervalo de tempo.Palavras-chave: Algoritmos genéticos. Desempenho. OpenMP. Qualidade das soluções. Abstract:Routing Vehicle Problem (RVP) is a combinatorial problem, hard to solve, used to improve the logistics of transport enterprises as well as to improve the traffic in the public ways. To solve it testing all combinations (brute force method) became unfisible as the problem scale, because it demands a large computing time. Genetic Algorithms (GA) are meta-heuristics able to find solutions in an acceptable computing time. However, even GA can demand a large computing time as they are set. Computional architectures evolution and the multicore difusion became the multithread programming an alternativ to reduce GA time. This article aims to speed up the RVP solution through the GA parallelization using OpenMP, which is a popular standard to multithreading programming. Our results show an speedup up to 2 for 4 threads in a quadcore processor. This gain is limited according to our GA is implemented. Beside the performance impact, we also show that the usage of OpenMP did not affect the solutions quality. Furthermore, OpenMP allow the GA to find better solutions because it make possible to increase the number of evolution in an time slice.

show abstract

“…Among meta-heuristics, BRKGA, proposed in [26], has proven to effectively solve network problems such as IP routing [30], routing in single layer [31] and multilayer [32] optical networks. Essentially, BRKGA is a class of genetic algorithm where a set of individuals, called a population, evolves over a number of generations with the aim to produce high quality solutions in short running times.…”

Section: Brkga Heuristic Algorithmsmentioning

confidence: 99%

Elastic Spectrum Allocation for Time-Varying Traffic in FlexGrid Optical Networks

Klinkowski

Ruiz

Velasco

et al. 2013

IEEE J. Select. Areas Commun.

137

View full text Add to dashboard Cite

Elastic flexgrid optical networks (FG-ON) are considered a very promising solution for next-generation optical networks. In this article we focus on lightpath adaptation under variable traffic demands in FG-ON. Specifically, we explore the elastic spectrum allocation (SA) capability of FG-ON and, in this context, we study the effectiveness of three alternative SA schemes in terms of the network performance. To this end, we formulate a Multi-Hour Routing and Spectrum Allocation (MH-RSA) optimization problem and solve it by means of both Integer Linear Programming (ILP) and efficient heuristic algorithms. Since, as numerical results show, the effectiveness of SA schemes highly depends on the traffic demand profile, we formulate some indications on the applicability of elastic SA in FG-ON.

show abstract

A biased random-key genetic algorithm for OSPF and DEFT routing to minimize network congestion

Cited by 27 publications

References 18 publications

Survivable IP/MPLS-Over-WSON Multilayer Network Optimization

Survivable IP/MPLS-Over-WSON Multilayer Network Optimization

O impacto da paralelização com OpenMP no desempenho e na qualidade das soluções de um Algoritmo Genético

Elastic Spectrum Allocation for Time-Varying Traffic in FlexGrid Optical Networks

Contact Info

Product

Resources

About