The Study and Implementation of an Intelligent Data Acquisition Middleware for Supporting the Performance Management of Modern Switched Local Area Networks

87

As the dramatically progress of new network technologies and rapidly growing of Internet users, the architecture of local area networks becomes more and more complicated. The best way to a network manager to control and manage network problems is to have the better network management skills and employ efficient network tools to assist him. This thesis establishes a data acquisition middleware system based on open system interconnection (OSI) reference model and telecommunication management network (TWN) standards to collect the traffic on a modern Ether switched local area network to support a network management software for performance evaluation and analysis. In addition, the system is implemented as a Web-based three-tier model on a Common Object Request Broker Architecture (CORBA) platform. Therefore, it makes the task to manage a switched LAN become much easier. The system consists of three subsystems. The first one is the user interface subsystem, which is implemented in Web-base with user graphic interface. The second one is the data acquisition subsystem, which is implemented complying with SNMP standards and having RMON function residing in each network device. The network traffic data needed to performance evaluation and analysis is collected and stored into a database. The third one is the data connection subsystem. It is implemented by applying CORBA technology and written in JAVA programming language. The network traffic data are thus collected and transferred to the user interface subsystem through the data acquisition subsystem. The goal of this tool is to construct a fundamental switched LAN management system for future development of network management performance support. The main goal of this construction is to provide a useful middleware to efficiently collect network traffic data for a generic network management tool to proceed its performance evaluation and analysis.