Presentation 2017-11-17
Packet Loss Rate Prediction based on Dictionary during Live Video Streaming
Natsuki Kai, Hiroshi Yoshida, Koichi Nihei, Dai Kanetomo, Kozo Satoda,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Live streaming services via a mobile network have become very popular. However, playback distortion or playback interruption often occurs during a live streaming. This is because a communication throughput dynamically fluctuates due to radio strength variation and cross-traffic, and unsent data overflows a buffer on a bottleneck link, which leads a packet loss. Forward Error Correction (FEC) is one of the major method to solve a packet loss problem, and FEC can maintain a real-time communication because FEC is the way to recovery a lost packet with some redundant data. However, FEC may fail a packet recovery when a redundant data is not enough, or there may be a wasted data when a redundant data is too much. To address these problems, this paper proposes a novel method for predicting packet loss rate (PLR) in real-time during live video streaming. We predict PLR by using the dictionary of past communication throughput, transferring rate, observed actual PLR, and our unique confidence parameter. We evaluate our proposed method with a live video streaming for 400 seconds as an experiment in conditions in which the communication throughput and delay are dynamically changed by a network emulator. As a result, the accuracy of our proposed prediction method was around 74~%.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) packet loss / prediction / FEC / Wi-Fi network
Paper # NS2017-122
Date of Issue 2017-11-09 (NS)

Conference Information
Committee ICM / CQ / NS
Conference Date 2017/11/16(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Toshio Tonouchi(NEC) / Takanori Hayashi(Hiroshima Inst. of Tech.) / Hideki Tode(Osaka Pref. Univ.)
Vice Chair Yuji Nomura(Fujitsu Labs.) / Yoichi Yamashita(NTT-N) / Hideyuki Shimonishi(NEC) / Jun Okamoto(NTT) / Yoshikatsu Okazaki(NTT)
Secretary Yuji Nomura(Fujitsu) / Yoichi Yamashita(KDDI R&D Labs.) / Hideyuki Shimonishi(NTT) / Jun Okamoto(Keio Univ.) / Yoshikatsu Okazaki(Kyushu Inst. of Tech.)
Assistant Haruo Ooishi(NTT) / Kenko Ota(Nippon Inst. of Tech.) / Norihiro Fukumoto(KDDI R&D Labs.) / Ryo Yamamoto(UEC) / Kenichi Kashibuchi(NTT)

Paper Information
Registration To Technical Committee on Information and Communication Management / Technical Committee on Communication Quality / Technical Committee on Network Systems
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Packet Loss Rate Prediction based on Dictionary during Live Video Streaming
Sub Title (in English)
Keyword(1) packet loss
Keyword(2) prediction
Keyword(3) FEC
Keyword(4) Wi-Fi network
1st Author's Name Natsuki Kai
1st Author's Affiliation NEC Corporation(NEC Corp.)
2nd Author's Name Hiroshi Yoshida
2nd Author's Affiliation NEC Corporation(NEC Corp.)
3rd Author's Name Koichi Nihei
3rd Author's Affiliation NEC Corporation(NEC Corp.)
4th Author's Name Dai Kanetomo
4th Author's Affiliation NEC Corporation(NEC Corp.)
5th Author's Name Kozo Satoda
5th Author's Affiliation NEC Corporation(NEC Corp.)
Date 2017-11-17
Paper # NS2017-122
Volume (vol) vol.117
Number (no) NS-303
Page pp.pp.61-66(NS),
#Pages 6
Date of Issue 2017-11-09 (NS)