Presentation 2019-07-11
Few-shot Learning based on Prototypical Network to Understand Area Service Level in LTE Networks
Shogo Aoki, Kohei Shiomoto, Chin Lam Eng, Sebastian Backstad,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In case a base station in mobile network malfunction, it is crucial to classify a service degradation event and identify a countermeasure to improve the service quality. However, it is difficult for operators to determine the network performance because the network performance in base station is very different depending on location and time. Classification by machine learning is considered a promising solution to this problem. However classification by machine learning requires large amounts of labeled data, which requires costly human-labor tasks to annotate large amounts of data. In order to address this problem we propose a method based on few-show learning that uses Prototypical Networks to analyze performance KPI data of base stations. Using data set obtained from a production LTE network with thousands of eNodeB data, we demonstrate that the proposed yields high detection performance while reducing the number of labeled data.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Cellular Network / eNodeB / KPI / Few-shot Classification
Paper # RCC2019-42,NS2019-78,RCS2019-135,SR2019-54,SeMI2019-51
Date of Issue 2019-07-03 (RCC, NS, RCS, SR, SeMI)

Conference Information
Committee SeMI / RCS / NS / SR / RCC
Conference Date 2019/7/10(3days)
Place (in Japanese) (See Japanese page)
Place (in English) I-Site Nanba(Osaka)
Topics (in Japanese) (See Japanese page)
Topics (in English) Communication and Networked Control for the Future Radio of the AI Age, etc
Chair Susumu Ishihara(Shizuoka Univ.) / Tomoaki Otsuki(Keio Univ.) / Yoshikatsu Okazaki(NTT) / Masayuki Ariyoshi(NEC) / Kazunori Hayashi(Osaka City Univ.)
Vice Chair Kazuya Monden(Hitachi) / Koji Yamamoto(Kyoto Univ.) / Satoshi Suyama(NTT DoCoMo) / Fumiaki Maehara(Waseda Univ.) / Toshihiko Nishimura(Hokkaido Univ.) / Akihiro Nakao(Univ. of Tokyo) / Suguru Kameda(Tohoku Univ.) / Osamu Takyu(Shinshu Univ.) / Kentaro Ishidu(NICT) / Shunichi Azuma(Nagoya Univ.) / HUAN-BANG LI(NICT)
Secretary Kazuya Monden(Kyoto Univ.) / Koji Yamamoto(NTT DOCOMO) / Satoshi Suyama(Hitachi) / Fumiaki Maehara(NTT) / Toshihiko Nishimura(Kyushu Univ.) / Akihiro Nakao(Osaka Pref Univ.) / Suguru Kameda(NTT) / Osamu Takyu(ATR) / Kentaro Ishidu(Univ. of Electro-Comm.) / Shunichi Azuma(Mie Univ.) / HUAN-BANG LI(Kagawa Univ.)
Assistant Akira Uchiyama(Osaka Univ.) / Kenji Kanai(Waseda Univ.) / Masafumi Hashimoto(Osaka Univ.) / Kazushi Muraoka(NTT DOCOMO) / Shinsuke Ibi(Doshisha Univ.) / Koichi Adachi(Univ. of Electro-Comm.) / Osamu Nakamura(Sharp) / Shinya Kumagai(Fujitsu) / Shinya Kawano(NTT) / Mai Ohta(Fukuoka Univ.) / Teppei Oyama(Fujitsu) / Kentaro Kobayashi(Nagoya Univ.) / Toshinori Kagawa(NICT) / Masateru Ogura(NAIST)

Paper Information
Registration To Technical Committee on Sensor Network and Mobile Intelligence / Technical Committee on Radio Communication Systems / Technical Committee on Network Systems / Technical Committee on Smart Radio / Technical Committee on Reliable Communication and Control
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Few-shot Learning based on Prototypical Network to Understand Area Service Level in LTE Networks
Sub Title (in English)
Keyword(1) Cellular Network
Keyword(2) eNodeB
Keyword(3) KPI
Keyword(4) Few-shot Classification
1st Author's Name Shogo Aoki
1st Author's Affiliation Waseda University(Waseda Univ.)
2nd Author's Name Kohei Shiomoto
2nd Author's Affiliation Tokyo City University(TCU)
3rd Author's Name Chin Lam Eng
3rd Author's Affiliation Ericsson Japan(Ericsson Japan)
4th Author's Name Sebastian Backstad
4th Author's Affiliation Ericsson Japan(Ericsson Japan)
Date 2019-07-11
Paper # RCC2019-42,NS2019-78,RCS2019-135,SR2019-54,SeMI2019-51
Volume (vol) vol.119
Number (no) RCC-106,NS-107,RCS-108,SR-109,SeMI-110
Page pp.pp.151-156(RCC), pp.177-182(NS), pp.173-178(RCS), pp.183-188(SR), pp.165-170(SeMI),
#Pages 6
Date of Issue 2019-07-03 (RCC, NS, RCS, SR, SeMI)