時空間トラヒックの異常検知のためのストリームデータ分析手法の研究(学内特別研究および国外研修)--(学内特別研究費報告書)時空間トラヒックの異常検知のためのストリームデータ分析手法の研究(学内特別研究および国外研修)--(学内特別研究費報告書) A Study on Stream Data Analysis for Anomally Traffic Detection with Space and Time Variation(Researches and Overseas Activities)--(Research Reports)
Network monitoring for high speed networks shared by the increasing internet traffic becomes crucial issue. Especially, anomaly traffic detection technology for high speed network is one of the most important issues, because the cyber-attacks have been more serious, and future increases in IoT devices will result in a structural traffic change with space and time variation. In order to detect anomaly traffic and take quick actions, detailed stream data including packet header information such as packet capture (pcap) data is needed. Therefore the data reduction methods have essentials for reducing original huge data size caused by increasing traffic on high speed network. This article is the first-year report on our ongoing research on sketch scheme as an effective data reduction method for anomaly traffic detection.