21st ITC Specialist Seminar on Multimedia Applications - Traffic, Performance and QoE Program

Tuesday, March 2

11:00 AM - 11:45 AM

Keynote

11:00 "Future TV Services Conveying an Enhanced Sense of Reality" (invited).
Eisuke Nakasu (NHK)

1:00 PM - 3:00 PM

QoE

13:00 Switching of Transmission Method of Sense of Force for Remote Haptic Control Systems
Yutaka Ishibashi (Nagoya Institute of Technology, Japan)
In this paper, we propose a scheme which dynamically switches the transmission method (i.e., one-way or two-way) of the sense of force according to network latency for remote haptic control systems. The proposed scheme chooses the two-way transmission of the sense of force when the network latency between the master and slave terminals is small, and it selects the one-way transmission of the sense of force when the network latency between the terminals is large. We demonstrate the effectiveness of the proposed scheme by subjective assessment. Assessment results show that there exists the optimum switching time depending on the work contents.
13:30 Switching Scheme of Group Synchronization Control in Multipoint Communications
Kazuki Hosoya (Nagoya Institute of Technology, Japan); Yutaka Ishibashi (Nagoya Institute of Technology, Japan); Shinji Sugawara (Nagoya Institute of Technology, Japan); Kostas Psannis (University of Macedonia, Greece)
In this paper, we propose a scheme which switches group (or inter-destination) synchronization control among three states dynamically according to work contents in multipoint communications with voice and video. In one state, group synchronization control is not exerted. In another state, the conventional group synchronization control is performed. In the other state, the group synchronization control considering difference of conversation roles is carried out. The proposed scheme can also switch conversation roles. By examining the timing from the generation of voice and video until the output, we demonstrate that proposed scheme can switch among the three states correctly, and the scheme can switch conversation roles adequately.
14:00 Quality-of-Experience Beyond MOS: Experiences with a Holistic User Test Methodology for Interactive Video Services
Sebastian Egger (Telecommunications Research Center Vienna (ftw.), Austria); Peter Reichl (Telecommunications Research Center Vienna (ftw.), Austria); Michal Ries (Vienna University of Technology, Austria)
Interactive video services, like video telephony, social TV or on-line gaming, are about to become a significant part of the service portfolio for telecommunication service providers. As the future commercial success of these services will depend essentially on their end-to-end quality as perceived by the end user, appropriate Quality-of-Experience (QoE) measurement methods are of paramount importance. The aim of this paper is to provide an application-oriented re-evaluation of actual QoE metrics for these services and to design a test methodology for evaluating user perceived experience which goes beyond standard Mean Opinion Score (MOS) metrics. To this end, we describe a specific test scenario designed for assessing the currently recommended question sets from two ITU recommendations. We determine the goodness of fit of these question sets to the user perceived quality dimension. Altogether, the resulting reduced set of items (questions) provides a significant step towards a more realistic methodology for assessing audio-visual quality perception.
14:30 Quality of Experience-Related Differential Equations and Provisioning-Delivery Hysteresis
Markus Fiedler (Blekinge Institute of Technology, Sweden); Tobias Hossfeld (University of Wuerzburg, Germany)
Churn of revenue-generating and dissatisfied users has become a major point of concern for service providers and network operators. As services rely on interconnecting networks, service performance and thus user satisfaction depend on network performance. Consequently, it is of outmost importance to understand the relationships between user perception, captured by quantitative Quality of Experience (QoE) parameters, and network performance, described by Quality of Service (QoS) parameters. This paper provides insights into fundamental relationships between QoE and QoS, formulated as partial differential equations describing changes in QoE with respect to specific QoS parameters. A set of illustrating examples is given. Furthermore, the different impacts of provisioning and degree of success or failure of delivery on QoE are discussed, leading to QoE provisioning-delivery hysteresis. This hysteresis provides a striking motivation for employing elastic adaptation mechanisms to available resources instead of suffering from uncontrolled data loss.

3:30 PM - 5:00 PM

Traffic modeling

15:30 Co-optimization of Transmission Parameters of PHY and MAC layers for IEEE 802.11n
Kaijie Zhou (SANYO Electric Co., Ltd., Japan); Akiomi Kunisa (SANYO Electric Co., Ltd., Japan)
This paper presents a co-optimization method of the physical (PHY) and medium access control (MAC) layers for IEEE 802.11n. With analytical models, the system goodput, or effective throughput, at the MAC-Service Access Point (SAP) is derived as functions of the controlled parameters of the PHY and MAC layers under the packet fragmentation and aggregation respectively, where the parameters consist of the packet size, modulation and channel coding scheme (MCS), and transmission try limit. Among the candidates for the transmission strategy performed according to the parameters, the optimal one with the highest goodput can be selected. The accuracy of the analytical model used for generating the goodput is verified through the simulation. The simulation result shows the proposed co-optimization method can increase the goodput by 15 Mbps at most, compared with the conventional packet-length adaptation and adaptive MCS methods.
16:00 Heavy and Non-heavy Users of Video Content Services: Dimorphic Demand Traffic
Shinsuke Shimogawa (NTT, Japan); Hiroshi Saito (NTT, Japan); Satoshi Kamei (NTT Service Integration Laboratories, Japan)
User demand for web-based video services is investigated. Heavy users and non-heavy users are classified by detecting the structural change of the frequency of the number of views a month. As the service becomes successful, the number of heavy users is small but occupies a large portion of the total traffic volume. Therefore, we can estimate the demand by monitoring small number of users. In addition, we show that the daily fluctuation (coefficient of variation) of the number of views of each user has strong dependence on the number of days in which the user views at least once. On this basis, we suggest that this number of days can be forecasted by observing the daily fluctuation of the number of views for a user. In particular, applying this forecasting method to heavy users will produce important information for the management of a successful service.
16:30 Analysis and Modeling of Mobile Telephone Traffic Triggered by Earthquakes
Hideyuki Koto (KDDI R&D Laboratories, Inc., Japan); Hajime Nakamura (KDDI R&D Laboratories Inc., Japan)
This paper proposes a traffic models for earthquake-triggered mobile telephone traffic. It is well known that people use mobile phone to make emergency and urgent calls after earthquakes. This paper reveals the statistical correlation between the scale of the earthquakes and the characteristics of the increased mobile phone traffic after earthquakes. Firstly, we propose an analysis model to characterize the features of the traffic increased after earthquakes. In addition, we make some hypothesis regarding the proposed features of the traffic in correlation with the scale of the earthquakes. Traffic data obtained from commercial mobile telephone network during various earthquakes are analyzed. Specifically, traffic data of over 100 subscriber switches from 21 cases of earthquakes monitored over the past 4 years are analyzed. The obtained results verify the hypothesis and the statistical correlations are revealed. Secondly, based on the analyzed results, we propose a traffic forecasting model. Some applications of the proposed forecasting model are described.

6:30 PM - 8:30 PM

Banquet

Wednesday, March 3

9:00 AM - 10:30 AM

P2P

9:00 Hierarchical Traffic Analysis and Modeling for P2P Streaming
Hiroyuki Kitada (Shibaura Institute of Technology, Japan); Takumi Miyoshi (Shibaura Institute of Technology, Japan); Takeshi Kurosawa (NTT, Japan); Masayuki Tsujino (NTT, Japan); Motoi Iwashita (NTT, Japan); Hideaki Yoshino (NTT, Japan)
This paper analyzes traffic characteristics dispersed by peer-to-peer (P2P) video streaming services, which have been a recent source of annoyance for Internet service provides since they generate a large amount of data. We focus on PPStream, a popular P2P video delivery service, and capture several day-long packet streams using a personal computer running PPStream. By statistically analyzing the measured data, we identified the distributions of a connection numbers, traffic volume, and flow characteristics. Through hierarchical observations, the following conclusions are obtained: (1) the geographical distribution of international connections follows a power-law relationship, (2) downstream traffic and connections are greater, and also more remarkable, than those upstream, (3) flow interarrival follows an exponential distribution, (4) flow duration follows a log-normal distribution, and (5) flow volume follows a logarithmic distribution.
9:30 Supporting Scalable Video Codecs in a P2P Video-on-Demand Streaming System
Simon Oechsner (University of Wuerzburg, Germany); Thomas Zinner (University of Wuerzburg, Germany); Jochen Prokopetz (University of Wuerzburg, Germany); Tobias Hossfeld (University of Wuerzburg, Germany)
There are currently two complementing and concurrent trends in the Internet. The first is the rise of video streaming and especially Video-on-Demand (VoD) as the most popular application for end users, expressed by a continuously increasing amount of video traffic. In this context, P2P is seen as a promising technology for providing video content efficiently to the users while reducing the cost for the content providers. The second trend is the diversification of end user devices used for watching these videos. This is reflected by the Scalable Video Codec (SVC) extension to the state-of-the-art H.264 video codec, which allows for a stream to be separated into substreams of varying quality and size. In this paper, we propose and evaluate a P2P VoD architecture based on the Tribler application which is enhanced to support such a SVC video. Thus, the proposed system is able to adapt the video quality on-the-fly to the network situation and access capabilities of the user devices.
10:00 Evaluating Global Traffic Engineering in Peer-to-peer Content Distribution and Edge Caches
Satoshi Kamei (NTT Service Integration Laboratories, Japan); Hiroshi Saito (NTT, Japan)
Traffic controls that try to keep P2P content distribution traffic inside each ISP are evaluated through computer simulation. The main example of this traffic control is P4P/ALTO. P4P/ALTO is a mechanism that requests the collaboration of ISPs and P2P service providers. It uses information provided by each ISP at the choice of a peer. The evaluation results show that, for a high hit ratio of contents in the ISP, P4P/ALTO can be effective. However, for a low hit ratio, the effectiveness is limited. This is because P4P/ALTO uses information limited to each ISP and does not use information beyond the ISP. Therefore, it is not effective if it does not find target content inside the ISP. This suggests that more extensive collaboration beyond the boundary of each ISP may be needed.

10:50 AM - 11:50 AM

New Applications and QoS

10:50 Low-Density Parity-Check Code Extensions Applied for Broadcast-Communication Integrated Content Delivery
Hosei Matsuoka (NTT DoCoMo, Japan); Akira Yamada (NTT DoCoMo, Inc, Japan); Tomoyuki Ohya (NTT DoCoMo, Japan)
This paper presents LDPC (Low-Density Parity-Check) code extensions suitable for integrated systems of communications and broadcasting where the lost data in broadcasting is repaired through communications channels with retransmission requests. The first extended scheme is the retransmission request data selection. When the transmission data is encoded with non maximum distance separable codes, the amount of retransmission traffic changes depending on which data is requested. The scheme selects the most effective data for LDPC decoding. The second extended scheme offers unequal error protection; the head of content is most strongly protected. It enables users to recover the lost data while viewing the head of the contents. We implemented these extension schemes on the LDPC staircase code on a binary erasure channel. The experimental results show that the proposed scheme can reduce the amount of retransmission traffic and also increase the loss resiliency of the head of contents.
11:20 Towards QoE Management for Scalable Video Streaming
Thomas Zinner (University of Wuerzburg, Germany); Osama Abboud (TU Darmstadt, Germany); Oliver Hohlfeld (TU Berlin, Germany); Tobias Hossfeld (University of Wuerzburg, Germany); Phuoc Tran-Gia (University of Wuerzburg, Germany)
Video streaming applications are a major driver for the evolution of the future Internet. In this paper we introduce a framework for QoE management for video streaming systems based on H.264/SVC codec, the scalable extension of H.264/AVC. A relevant feature is to control the user perceived quality of experience (QoE) by exploiting parameters offered by SVC. A proper design of a control mechanisms requires the quantification of the main influence parameters on the QoE. For this purpose, we conducted an extensive measurement study and quantified the influence of i) video resolution, ii) scaling method, iii) network conditions in terms of packet loss and iv) video content types on the QoE by means of the SSIM and PSNR full-reference metrics. Further, we discuss the trade-off between these different control knobs and their influence on the QoE.

13:00 PM - 14:30 PM

Consolidated session with the IN/NS workshop

13:00 Overview
Hiroyuki Morikawa (the University of Tokyo)
13:10 "QoE assessment and monitoring of IPTV" (invited)
Akira Takahashi (NTT)
13:50 "Trends on video distribution services and related standardization activities" (invited)
Satoshi Miyaji (KDDI)