TITLE:
Research on the Anomaly Detection Method in Intelligent Patrol Based on Big Data Analysis
AUTHORS:
Xiaoqing Deng
KEYWORDS:
Big Data, Intelligent Patrol, Anomaly Detection
JOURNAL NAME:
Journal of Computer and Communications,
Vol.7 No.8,
August
21,
2019
ABSTRACT: The network anomaly detection in intelligent patrol is based on the trigger of a single threshold of network element performance parameters in patrol task, which has a high false alarm rate and low efficiency. In order to effectively and accurately integrate network performance, this paper proposes to mine network element performance data and network element log information in the integrated automatic patrol to detect network anomalies. Because log files have a large amount of data and a variety of types, and log data has a complex structure and contains large implied information. The relationship between network anomalies and time can actively discover through the analysis of the log files. Therefore, big data mining and classification can greatly improve the efficiency of data processing. However, the accuracy of finding network anomalies is insufficient only for log analysis. Therefore, this paper puts forward the performance indexes collected in the log analysis and patrol inspection system and adopts the sequence analysis algorithm to detect network anomalies, so as to improve the accuracy and efficiency of detection.