袁藝芳,李雁,陳緒,高永龍,席新



摘 ?要: 本文針對移動警務網絡復雜多變、數據量大的特點,提出一種基于孤立森林算法的網絡流量監測方法。該方法以網絡IP數據流為基礎,通過對IP數據流提取特征參數,并將特征參數作為輸入向量,利用孤立森林算法進行訓練以實現監測。這種方法能夠快速、有效地檢測出移動警務網絡中的異常流量,精確率高,在一定程度上對移動警務網絡的智能運維和安全防護起到重要作用。
關鍵詞: 孤立森林,算法,移動警務,網絡,流量監測
中圖分類號: TP391.0 ? ?文獻標識碼: A ? ?DOI:10.3969/j.issn.1003-6970.2019.12.051
本文著錄格式:袁藝芳,李雁,陳緒,等. 基于孤立森林算法的移動警務網絡流量監測方法研究[J]. 軟件,2019,40(12):229232
Research on Mobile Police Network Traffic Monitoring Method
Based on Isolated Forest Algorithm
YUAN Yi-fang1, LI Yan2, CHEN Xu2, GAO Yong-long2, XI Xin2
(1. Science and Technology Information Bureau of the Ministry of public security 100005, China;
2. Tianjin Public Security Bureau Science and Technology Information Office 300393, China)
【Abstract】: Mobile police network is complicated and changeable, and it has a very large amount of data to be handled. According to these characteristics, a network traffic monitoring method based on isolated forest algorithm is proposed in this paper. This method is based on the IP network data. Feature parameters are extracted for each IP data flow, and the feature parameters are taken as the input vectors for isolated forest algorithm to train isolated trees and achieve monitoring. This method can detect abnormal traffic in mobile police network quickly and effectively, which can play an important role in intelligent operation and security protection of mobile police network.
【Key words】: Isolated forest; Algorithm; Mobile police; Network; Traffic monitoring
0 ?引言
2002年以來,公安部對公安信息移動接入及應用系統安全建設進行不斷完善,全國公安系統信息化得到了很大的提升。2017年初,天津市公安局新一代移動警務系統開始建設,目前已建設完成投入運行。新一代移動警務平臺體系相對復雜、運維成本高且難度大,平臺中一旦出現問題,往往需要大量的專業人員參與,協同診斷問題,耗時長,代價高。為緩解以上問題,我們對移動警務平臺中網絡流量數據進行分析,利用一種基于孤立森林算法的網絡流量監測方法對網絡流量異常檢測。目的是可以及時發現網絡攻擊行為和網絡結構問題引起的異常流量,從而增強網絡態勢感知能力和安全防護能力,對移動警務平臺的智能化運維有重要推進作用。……