基于自相似模型的智能网络质量管理系统

INTELLIGENT NETWORK QUALITY MANAGEMENT SYSTEM BASED ON SELF-SIMILAR MODEL

传统网络质量管理方法存在不能自适应网络环境变化,无法判断未知网络异常的不足。针对这种情况,以自相似模型为理论基础开发设计智能网络质量管理系统。该系统对用户网络流量行为建立自相似数学模型,然后引入图形图像学中图表相似度的比较方法,通过对比当前用户行为与模型,进而判断网络是否存在异常,并在自动或人工模式下启动相应干预操作。实际应用结果表明,相比较传统网络质量管理方法与传统的自相似模型基础下判断网络异常的方法而言,该系统有效地提高了网络质量管理的智能化水平和可靠性。

Traditional network quality management methods have the deficiencies of self-adapt to the changes in network environment and cannot determine unknown network anomalies. In view of this, we designed this intelligent network quality management system based on self- similar model. The system builds the self-similar mathematical model for users＇ network traffic behaviours, and then introduces the comparative method of image and graphics similarities in graphic iconography, by comparing the current user behaviour with the model it further determines whether the network is abnormal, and starts the correlated intervention operation in either automatic mode or manual mode. Actual application results show that compared with traditional network quality management methods and traditional method which determines the network anomaly based on self-similar model, this system effectively improves the intelligence level and reliability of network quality management.