Evaluation Model for IP Network Routing Decision based on PCA-ANN

Abstract

As an act of moving information across an internetwork from one source to one destination, routing is vital for network activities. The IP (Internet protocol) network is the so-called �best-effort� communication network in which the best-effort delivery is provided for hosts. Accordingly, routing decision is closely related with a wide variety of network applications. This study presents one called PCA (�principal component analysis�)-ANN (�artificial neural networks�) evaluation model for routing decision. The PCA technology is used to reduce the dimension of the route measurement data to present the most effective structure of the route measurement data for the ANN to perform further evaluation. The proposed PCA-Based model is evaluated by experiments. Experimental results show its potentiality.