Research On Network Intelligent Fault Diagnostic Techniques

With the development of computer science and communication network, the scale of network is growing larger, together with the emergence of more and more network applications. Owing to simple function, complex operation and lower efficiency, the old network troubleshooting system already cannot meet the demands of carrier development. Now, the most important problem is how to manage network effectively, securely and make network extensible. In order to perform high efficiency and reliability, it is very important for us to set up a perfect network troubleshooting system. This paper brings layering-decentralization intelligence into the field in which we can make it possible for the failure automatic location and diagnosis. According to this paper, the main achievements are as follows:(1) Firstly, it describes a new framework which takes advantage of the mobile agent to the management of distributed network. Secondly, it describes the information flow of key modules in this framework and also implements. Finally it tests the new framework and proves the efficiency.(2) Give a new fault diagnosis model of physics layer based on event correlation, which set fault diagnosis into two phases: Failure locating and failure diagnosing. This paper uses the failure relation algorithm in the first phase, and the case based reasoning in the other, which will definitely improve the efficiency of failure diagnosis. Based on the event correlation, Fault Diagnosis is applied to extract the root failure sets by using graph theory and adjacency matrix. According to the relationship of failures, this paper gives a method to determine the source of failure and make failure filtration and location function effectively.(3) Knowledge systems technologies are introduced into linker layer diagnosis system ,and a knowledge base related to the network fault diagnosis is constructed, which enable the system to utilize the experience and knowledge of experts. Network status features are captured through feature extraction from running status information of the network elements. Since the essence of fault diagnosis is a kind of mapping, an artificial neural network (ANN) model is adopted to deal with the mapping relation, categorizing the network faults. The knowledge base provides discipline samples for the artificial neural network model. In order to eliminate redundancy of the knowledge rules in the knowledge base, means based on rough sets theories is undertaken for reductionof knowledge, which results in efficient training of the ANN model and more precision of the fault diagnosis.(4) A kind of load prediction and congestion control policy based on FNN (fuzzy neural networks) is proposed NOW (Networks of Workstation). Load balancing is one of the key problems in NOW. The normal technology of load balancing, such as master-slave scheduler and threshold scheduler, always assign the task based on the present load of workstation. So, the resource utility is low and control of load balancing is lagged. In the paper, a fuzzy neural network scheme, which processes good capability of processing inaccurate information and learning, is provided to solve these limitations. The results of simulations show that the FNN scheme is effective.(5 ) A framework of SVM based Network Fault Detection System of Application Layer is proposed. The function, mechanism and realization of the components of this framework are discussed in the paper. By means of distance metric of heterogeneous datasets, the feature data of network are preprocessed. Based on guaranteed estimators, we estimate the size of test set. Thus we not only avoid bad train result for lack of examples, but also reduce the training time and improve the efficiency of training. During the training, by means of fuzzy mathematics, considering the effect of different network data features to the classification, a weight method is brought forward. It improves the accuracy of network fault detection. The problem of low detection accuracy of some types of faults for the imbalance of training examples is res...