A dynamic k-routing algorithm in wavelength-routed optical networks
S. Siddiqui, Jing Wu, Hussein T. Mouftah, Michel Savoie
Abstract
S. Siddiqui, Jing Wu, Hussein T. Mouftah, Michel Savoie
Abstract
Routing in a wavelength-routed network can be static or dynamic. In static routing, routes for each source-destination pair in a network are predetermined, whereas in dynamic routing, a route is computed dynamically for a connection request as it arrives in the network. Dynamic routing incorporates current network state in path computation to ensure an optimal path selection. In this paper, we propose an adaptive routing algorithm, a dynamic k-routing algorithm that computes a least cost path for a connection request. The algorithm computes at most k paths before a connection request is blocked. This paper assumes no wavelength conversion in the network and that the link state information is available to each node in the network. We compare the blocking performance of the proposed adaptive routing algorithm with fixed shortest path routing, fixed alternate shortest path routing and dynamic routing (using the shortest path with available capacity) and show that this algorithm performs better in terms of blocking probability and network utilization.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Routing in a wavelength-routed network can be static or dynamic. In static routing, routes for each source-destination pair in a network are predetermined, whereas in dynamic routing, a route is computed dynamically for a connection request as it arrives in the network. Dynamic routing incorporates current network state in path computation to ensure an optimal path selection. In this paper, we propose an adaptive routing algorithm, a dynamic k-routing algorithm that computes a least cost path for a connection request. The algorithm computes at most k paths before a connection request is blocked. This paper assumes no wavelength conversion in the network and that the link state information is available to each node in the network. We compare the blocking performance of the proposed adaptive routing algorithm with fixed shortest path routing, fixed alternate shortest path routing and dynamic routing (using the shortest path with available capacity) and show that this algorithm performs better in terms of blocking probability and network utilization.
Key concepts: Static routing, Equal-cost multi-path routing, Link-state routing protocol, Computer science, Dynamic Source Routing, Path vector protocol, Multipath routing, Computer network