Research And Implementation Of Intelligent Fault Recovery System On Terabit Router

The effectiveness and power of fault management system correlate with the availability and reliability of managed network. Fault detection is to identify a fault after it occurred, and fault correlation is to locate a fault after it is detected. Fault localization is to determine the location of the fault. The aim of fault diagnosis is to effectively diagnose the fault and give the causation and to recover it. Basing the requirements of 863 Terabit router project, this thesis presents the realization scheme of intelligent fault recovering system.The main contributions of this thesis are as follows.i) Do research further on the fault detection technology. Because the proactive polling has some disadvantages, such as big management expenses, high polling latency, and non-real-time fault information, a new dynamic polling algorithm that can meet the real-time requirement is given.ii) Do discuss on some kinds of fault correlation technology and do compare with their advantages and disadvantages. A new fault correlation algorithm, the device fault correlation algorithm which is based on graph is put forward.iii) Do study detailedly on two kinds of fault diagnosis technology, traditional expert system and artificial neural network system. These fault diagnosis technology has its own advantage and disadvantage and has some limitation in parctice. In pratical use, we should combine these fault diagnosis technology and make effectively fault diagnosis.iv) Based on the forementioned analysis and according to the requirements of 863 Terabit router project, we synthetically use traditional expert system and artificial neural network and present muti-agent intelligent fault recovery system. This scheme works well in the fault management in practice.