Research Of Network Fault Detection Model Based On Danger Mode

With the wide application of computer network in people’s work and life, the expansion of network scale and the complication of network structure, the probability of network fault is more and more large. The loss caused by network fault is immeasurable, so the importance of the management network fault is significant. Network fault detection, positioning and recovery are the important content of network fault management. The complexity of network technology brought a lot of difficulties to fault detection work, and artificial immune system provides a new idea for fault detection technology, because it has many advantages.This paper intensively studies on the biological immune mechanism, and makes detailed summary on the development and application of the existing artificial immune technology. In order to reduce the fault detection rate and improve the self-adaptive capability in network fault detection, this paper put forward a kind of network fault detection method based on self-noself mode and danger mode. Main job is as follows:(1) Put forward an immune detection model based on the hazard mode, through the contradistinction of the advantages and disadvantages of the two models.(2) Intensively study on self-noself recognition model and dangerous recognition model, analyze their advantages and disadvantages. Analyze dendritic cells algorithm in detail.(3) Design the network fault detection model based on danger mode, based on the research of the network fault detection technology and artificial immune theory. This model is divided into nine modules, antigen and danger signal collection module, antigen and danger signals presenting module, DCA detection module, SNS detection module, abnormal confirm module, danger zone establish module, detectors update module, self-set update module and immune response module. Design the detailed process of each module, quantify the input and output data, and analyze the relationship between modules.(4) Design the experiment scheme. Take the small office network for example to simulate, prove the feasibility and effectiveness of the model proposed.