Abstract
Abstract
Open Shortest Path First (OSPF) is one of the most widely used intra-domain routing protocol. It is well known that OSPF protocol does not provide flexibility in terms of packet forwarding to achieve any network optimization objective. The common objective of network traffic engineering is to minimize the maximal link utilization in a network in order to accommodate more traffic and reduce the chance of congestion In this paper an algorithm, to solve congestion problem in an OSPF based best effort network, is proposed. This algorithm is called Less-congested Open Shortest Path First (LOSPF). A simple feed-forward neural network is used to predict the congestion problem in the computer network links which are over utilized. The neural network is trained to predict congested links based on a given traffic pattern. These predictions are used to modify Dijkstr's algorithm to select the shortest paths, that excluding those links.Better resource utilization is obtained by applying the resulting paths. This is achieved throughout avoiding congested links and using other links that are underutilized. Comparison between performance of the suggested algorithm and the performance of OSPF that uses Dijkstra's algorithm to find the shortest path. Simulation results show that LOSPFperformance is better in term of distributing links utilizations, and the queuing delay in the congested links.In spite of the superiority of the proposed LOSPFalgorithm, its execution time is almost same as that of the OSPF
