Please use this identifier to cite or link to this item:
|Title:||Observations on using Genetic Algorithms for Routing and Wavelength-Assignment (RWA) in All-optical networks|
|Citation:||Proceedings of the National Conference on Computer Science and Technology, 11-12 Nov 2006, KIET, Ghaziabad|
|Abstract:||All-optical networks based on wavelength division multiplexing (WDM) have emerged as a promising technology for network operators to respond to an increased demand for broadband service. All-optical networks consist of optical fiber links and nodes. Each node has a dynamically configurable optical switch or router, which supports wavelength based switching or routing. Configuring these optical devices across the network enables one to establish all-optical connections, or lightpaths, between source and destination nodes. In order to establish a lightpath, the network needs to decide on the route and the wavelength(s) for the lightpath. Given a set of connections, the problem of setting up lightpaths by routing and assigning a wavelength to each connection is called the Routing and Wavelength Allocation (RWA) problem. There are two variations of the problem: static and dynamic. In the static case the set of desired connections is known beforehand; in the dynamic case the connection requests arrive based on some stochastic process. The RWA problem can be simplified by decoupling it in four subproblems: the topology subproblem, the lightpath routing subproblem, the wavelength allocation subproblem and traffic routing subproblem. All-optical WDM networks are characterized by multiple metrics (hop-count, cost, delay), but generally routing protocols only optimize one metric, using some variant of shortest path algorithms (e.g., the Dijkstra, All-pairs and Bellman-Ford algorithms). The multicriteria RWA problem has been solved combining the relevant metrics or objective functions. This is achieved by defining integrated link cost functions (which embed the different metrics) or using a single cost objective function defined as a weighted sum of objective functions. This paper outlines the application of GA to solve a RAW problem. We have proposed a modified genetic algorithm with reference to the basic model proposed by Zhong Pan. The modifications proposed are change of fitness function by adding a new parameter hopcount and a two-point crossover operation is used instead of single-point crossover. The performance of both the algorithms are studied with respect to the time taken for making routing decision, number of wavelengths required and cost of the requested lightpaths. The modified algorithm performed better than the existing algorithm with respect to the time and cost parameters on standard networks such as ARPANET, and NSFNET. It has been observed that the proposed modified Genetic algorithm performed better than the existing algorithm with respect to the time and cost parameters.|
|Description:||Copyright for this article belongs to KIET, Ghaziabad|
|Appears in Collections:||Journal Articles|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.