Portal de Periódicos da CAPES

Artigo Produção Nacional Revisado por pares

BIBTEX RIS

Heuristic planning algorithm for sharing restoration interfaces in OTN over DWDM networks

2020; Elsevier BV; Volume: 61; Linguagem: Inglês

10.1016/j.yofte.2020.102426

1095-9912

A. V. S. Xavier, J.C. Silva, Joaquim F. Martins-Filho, Carmelo J. A. Bastos-Filho, Daniel A. R. Chaves, R.C.A. Almeida, Danilo Araújo, José Augusto Correa Martins,

Advanced Photonic Communication Systems

Due to the expansion of telecommunication services, the demand for technologies that improve transmission quality has been growing in recent years. One of these technologies is optical transport networks (OTNs), which works as a layer above the DWDM (Dense Wavelength Division Multiplexing) network to enable the most efficient use of resources. This work proposes an OTN network planning solution over DWDM based on heuristic algorithms. The main objective is to plan the allocation of OTN network interfaces (the most costly element) to meet the performance and resilience requirements. We implemented two versions of the heuristic algorithm. The first version, called OTN-PASRIN-Max, delivers a robust solution to ensure all planning requirements, but the heuristic does not guarantee a solution that deploys the minimum number of resources. The second version, which is our main proposal, is called OTN-PASRIN and returns a solution that simultaneously maximizes resilience and minimizes the deployment of resources through better reuse of components, which causes a reduction in the final cost of the project. We evaluate our proposal in two scenarios, a small network with 4 OTN nodes, and a larger network with 16 OTN nodes. For the small network, we compared the results of the proposed algorithm with an exhaustive search, and the results were very close in all traffic conditions. In the larger network, we performed a comparison between our proposal and the other version (OTN-PASRIN-Max). The results showed that OTN-PASRIN saves a considerable amount of resources (approximately 20%) in all studied scenarios.