微型电脑应用

研究了一种基于云计算和机器学习的新型入侵检测系统MOVCIDS (mobile visualization connectionist IDS)。该系统能够有效检测到神经网络结构检测网络中的异常情况。通过可视化方式对识别出的网络异常情况进行展示,同时给出流量异常预警。利用神经网络和贝叶斯滤波器,建立海量流量数据集预测模型,获取数据内部结构的可视化信息,利用可视化界面展示组织映射。MOVCIDS系统可以从任何移动设备访问,为网络管理员提供更多的访问方法,同时满足计算机网络的持续监控、监督和可视化展示需求。通过实验方式验证入侵检测策略的有效性,在不同的实际环境中测试检测模型,同时验证了时间维度对网络入侵检测的重要性,对网络安全及入侵处理效率的提升具有重要意义。

This study introduces and describes a new intrusion detection MOVCIDS(mobile visualization connectionist IDS) system. The system uses the neural network structure of cloud computing and machine learning to detect abnormal conditions in the network. An early warning can be given for traffic anomaly, dynamics, and variability through visualization facilities, and the identification of anomaly can be handled by computer networks. These self-organizing maps are described through a mobile visual interface based on neural networks and Bayesian filters through the prediction of massive flow data sets to provide visual information on the internal structure of the data. The MOVCIDS can be accessed from any mobile device to give network administrators more access to support continuous visualization, monitoring, and monitoring of computer networks. The effectiveness of the proposed intrusion detection strategy is verified and tested in different environments. In addition, this paper also emphasizes the importance of time dimension to intrusion detection and the processing ability of intrusion detection system to time dimension.