基于大数据分析的网络异常流量检测

The network anomaly traffic detection based on large data analysis

为了取得较好的网络异常流量检测效果,利用云计算技术对大数据环境下的网络异常流量检测方法进行研究。将网络异常流量检测任务分配给多个子任务计算平台进行,提高了网络异常流量检测的准确率和效率。对网络异常流量检测效果进行测试,结果表明:该检测方法的平均准确率相比常规单机检测方法提高了17. 08%,误报率降低了65. 7%,能够满足目前大数据环境下对网络异常流量检测的要求;其检测耗时仅为常规单机网络异常流量检测方法的8. 81%,具有更好的实时性。

In order to detect the network traffic anomaly better,it uses the cloud computing technology to study the network anomaly traffic detection method under the large data environment,and tests the network traffic anomaly detection effect. The abnormal network traffic detection task in large data is allocated to many sub task computing platforms. The results show that the average accuracy of the proposed detection method is improved by17. 08%,the false alarm rate is reduced by 65. 7%,the accuracy of the proposed method is higher,the false alarm rate is lower,and the detection performance is better. This method can meet the requirements of the network abnormal traffic detection under the current large data environment. Using the proposed network anomaly traffic detection method,the detection time is only 8. 81% of the conventional single network anomaly traffic detection method,which has better real-time performance compared with the conventional method.