An efficient method of modeling performance data for wireless mesh network

With the development of wireless mesh network (WMN) and its expansion of the scale, it is important to provide an efficient mechanism to convenient the network administrators. Network fault diagnosis is the key part of network management, which can support effective and reliable network operation. However, one of the problems is how to select useful information from a large scale network performance data for analyzing and diagnosing. This paper proposes a method of modeling wireless mesh network performance data based on knowledge ontology. Using the knowledge model of network performance data, combined with the decision tree algorithm, the fault diagnosis system can analyze the state of network effectively and improve the reliability of network. The paper employed the ontology to establish the network performance data model, and verified its validation on the campus network.