Abstract: In order to reduce the fault detection rate and improve the anti-fraud capacity in network fault detection, a multi-layer model for network fault detection is proposed based on the hierarchical structure of biological immune system. According to the differentiation mechanism of dendritic cells, an immune model composed of the inherent detection layer, fuzzy judgment layer and adaptive detection layer is established. It reduces the network fault detection rate through the double detection of inherent detection layer, fuzzy judgment layer and uses the adaptive detection layer to conduct self-learning on unknown faults. Experimental result shows that this model has high detection rate and low false reject rate.

Key words: network fault detection, inherent detection layer, fuzzy judgment layer, adaptability detection layer, Dendritic Cells(DCs)

摘要： 传统网络故障检测模型的误检率较高、自适应性较差。为此，提出一种基于生物免疫机制的层次检测模型。根据树突细胞分化机制，建立包含固有检测层、模糊判断层和自适应性检测层的免疫模型，使用固有检测层和模糊判断层的双重检测技术降低网络错误检测率，利用自适应性检测层对未知故障进行自我学习。实验结果表明，该模型具有较高的检测率和较低的误检率。

关键词: 网络故障检测, 固有检测层, 模糊判断层, 适应性检测层, 树突细胞

CLC Number: 

• N945.12