The Study On The Technique Of Ancillary Diagnosis Client Malfunction Of Campus Network

In recent years, the dimensions of campus network become more and wider by the rapidly developing of computer technique. In the every life, people increasingly depended on the network. It is more important that keeping the network unimpeded and steady.In current the network had become more and more complicate and isomeric, as well as the population of internet users are incising quickly and the business of network be emerge in endlessly causing the data flow of network greatly increased.All of those factors greatly increased the difficulty and Challenge of networks maintenance.This article describes the Simple Network Management Protocol and combined with the practical situation of campus network the fault diagnosis system. This system was fit for almost every campus network. After analyzing and comparing each merit and demerit and the application area of them of correlation technology, this system use the methods of Case-based reasoning to diagnosis faults. The main research work and are as following:1. The first of all, analyze for the characteristics of campus network and the difficulties of fault diagnosis. Study on Network Alarm management. The Gather of fault information was used the methods of" active polling and in tempt-base".2. Mainly studied the case-based reasoning for case representation and case library management and the case retain and the case retrieval, modification and retention methods improved. Main work is improvements of the case retrieval algorithm: KNN.3. In case management, for the each of case to set a parameter called occur frequency. And the case in descending order by occur frequency. This increases the case's retrieval speed.This system is a C / S model that installed on the user terminal, and able to complete interruption of the network under the diagnosis of network failures, with a high value.