| 講演名 | 2023-09-21 Extracting context from external events using auxiliary information from network traffic data Eilaf M.A Babai(Kyushu Univ.), Koji Okamura(Kyushu Univ.), |
|---|---|
| PDFダウンロードページ |  PDFダウンロードページへ |
| 抄録(和) | Network traffic flow data is the main source of information about the network status, and it is constantly monitored to ensure networkoperation. However, traffic data features alone are insufficient for network traffic prediction in network management, especiallywhen traffic has a random nature due to influence of external factors. This is because network data provides limited view aboutuser context compared to the network context and status. Hence, augmenting contextual information about users’ situation cansignificantly improve the accuracy of network traffic prediction. Such information can be extracted by correlating spatio-temporalfeatures obtained from traffic data with occurring external events such as online fairs and lectures that have matching attributes. In this work, we focus on a large campus network and aggregate inbound traffic into country wise time series based on auxiliarygeo-temporal features obtained from IP addresses. We then cluster these time series to investigate common patterns and identifywhich external events are influencing significant fluctuations in network traffic for these countries. Finally, we model these events toexamine the predictive capabilities of this contextual data. |
| 抄録(英) | Network traffic flow data is the main source of information about the network status, and it is constantly monitored to ensure networkoperation. However, traffic data features alone are insufficient for network traffic prediction in network management, especiallywhen traffic has a random nature due to influence of external factors. This is because network data provides limited view aboutuser context compared to the network context and status. Hence, augmenting contextual information about users’ situation cansignificantly improve the accuracy of network traffic prediction. Such information can be extracted by correlating spatio-temporalfeatures obtained from traffic data with occurring external events such as online fairs and lectures that have matching attributes. In this work, we focus on a large campus network and aggregate inbound traffic into country wise time series based on auxiliarygeo-temporal features obtained from IP addresses. We then cluster these time series to investigate common patterns and identifywhich external events are influencing significant fluctuations in network traffic for these countries. Finally, we model these events toexamine the predictive capabilities of this contextual data. |
| キーワード(和) | Network traffic / Context / Events / Prediction / Goelocation / time series clustering |
| キーワード(英) | Network traffic / Context / Events / Prediction / Goelocation / time series clustering |
| 資料番号 | IA2023-13 |
| 発行日 | 2023-09-14 (IA) |
| 研究会情報 | |
| 研究会 | IA |
|---|---|
| 開催期間 | 2023/9/21(から2日開催) |
| 開催地（和） | 北海道大学 |
| 開催地（英） | Hokkaido Univeristy |
| テーマ（和） | インターネット運用・管理, ネットワークアーキテクチャ, 通信プロトコル, IoT, 一般 |
| テーマ（英） | Internet Operation and Management, Network Architecture, Communication Protocols, IoT, etc. |
| 委員長氏名（和） | 秋山 豊和(京都産大) |
| 委員長氏名（英） | Toyokazu Akiyama(Kyoto Sangyo Univ.) |
| 副委員長氏名（和） | 作元 雄輔(関西学院大) / 渡辺 俊貴(NEC) / 屏 雄一郎(KDDI) |
| 副委員長氏名（英） | Yusuke Sakumoto(Kwansei Gakuin Univ.) / Toshiki Watanabe(NEC) / Yuichiro Hei(KDDI) |
| 幹事氏名（和） | 大平 健司(阪大) / 坂野 遼平(工学院大) / 野林 大起(九工大) |
| 幹事氏名（英） | Kenji Ohira(Osaka Univ.) / Ryohei Banno(Kogakuin Univ.) / Daiki Nobayashi(Kyushu Inst. of Tech.) |
| 幹事補佐氏名（和） | 小谷 大祐(京大) / 中村 遼(福岡大) / 中村 遼(東大) |
| 幹事補佐氏名（英） | Daisuke Kotani(Kyoto Univ.) / Ryo Nakamura(Fukuoka Univ.) / Ryo Nakamura(Univ. of Tokyo) |
| 講演論文情報詳細 | |
| 申込み研究会 | Technical Committee on Internet Architecture |
|---|---|
| 本文の言語 | ENG |
| タイトル（和） | |
| サブタイトル（和） | |
| タイトル（英） | Extracting context from external events using auxiliary information from network traffic data |
| サブタイトル（和） | |
| キーワード(1)（和/英） | Network traffic / Network traffic |
| キーワード(2)（和/英） | Context / Context |
| キーワード(3)（和/英） | Events / Events |
| キーワード(4)（和/英） | Prediction / Prediction |
| キーワード(5)（和/英） | Goelocation / Goelocation |
| キーワード(6)（和/英） | time series clustering / time series clustering |
| 第 1 著者 氏名（和/英） | Eilaf M.A Babai / Eilaf M.A Babai |
| 第 1 著者 所属（和/英） | Kyushu University(略称：Kyushu Univ.) Kyushu University(略称：Kyushu Univ.) |
| 第 2 著者 氏名（和/英） | Koji Okamura / Koji Okamura |
| 第 2 著者 所属（和/英） | Kyushu University(略称：Kyushu Univ.) Kyushu University(略称：Kyushu Univ.) |
| 発表年月日 | 2023-09-21 |
| 資料番号 | IA2023-13 |
| 巻番号（vol） | vol.123 |
| 号番号（no） | IA-193 |
| ページ範囲 | pp.11-18(IA), |
| ページ数 | 8 |
| 発行日 | 2023-09-14 (IA) |