Survivable DWDM optical mesh transport network design via genetic algorithms
Kavian, Y. S., Leeson, Mark S., 1963-, Hines, Evor L., Ren, W., Rashvand, H. F. and Naderi, M. (2009) Survivable DWDM optical mesh transport network design via genetic algorithms. In: Hayworth, G. I., (ed.) Reliability Engineering Advances. New York: Nova Publishers, pp. 331-354. ISBN 978-1-60692-329-0Full text not available from this repository.
Official URL: https://www.novapublishers.com/catalog/product_inf...
Acceptable service provision in the presence of failures and attacks is a major issue in the design of next generation dense wavelength division multiplexing (DWDM) networks. Survivability is provided by the establishment of spare lightpaths for each connection request to protect the working lightpaths, which is an NP-hard Problem. This chapter presents a genetic algorithm (GA) solver for the routing and wavelength assignment problem with working and spare lightpaths to design survivable optical networks in the presence of a single link failure. Lightpaths are encoded into chromosomes made up of a fixed number of genes equal to the number of entries in the traffic demand matrix. Each gene represents one valid path, and is thus coded as a variable length binary string. After crossover and mutation, each member of the population represents a set of valid but possibly incompatible paths and those that do not satisfy, the problem constraints are discarded. The best paths are then found by use of a fitness function and these are assigned the minimum number of wavelengths according to the problem constraints. The proposed approach has been evaluated for dedicated path protection and shared path protection. Simulation results show that the GA method is efficient and able to design survivable real-world DWDM optical mesh networks. The GA solver is further tuned for both a bandwidth optimization scheme (BOS) and a delay optimization scheme (DOS). The performance of the new GA based resiliency model is evaluated for benchmark networks, and simulation results show that the method displays excellent performance in solving this complex, multi-constrained problem for BOS and DOS.
|Item Type:||Book Item|
|Subjects:||T Technology > TK Electrical engineering. Electronics Nuclear engineering|
|Divisions:||Faculty of Science > Engineering|
|Place of Publication:||New York|
|Book Title:||Reliability Engineering Advances|
|Editor:||Hayworth, G. I.|
|Number of Pages:||24|
|Page Range:||pp. 331-354|
|Access rights to Published version:||Restricted or Subscription Access|
Actions (login required)