Perceptual Quality Assessment of P2P Assisted Streaming Video for Chunk-level Playback Controller Design Tom Z.J. Fu, CUHK W. T. Leung, CUHK P. Y. Lam,

Slides:



Advertisements
Similar presentations
1 P2P Layered Streaming for Heterogeneous Networks in PPSP K. Wu, Z. Lei, D. Chiu James Zhibin Lei 17/03/2010.
Advertisements

Layered Video for Incentives in P2P Live Streaming
Tuning Skype Redundancy Control Algorithm for User Satisfaction Te-Yuan Huang, Kuan-Ta Chen, Polly Huang Proceedings of the IEEE Infocom Conference Rio.
Playback delay in p2p streaming systems with random packet forwarding Viktoria Fodor and Ilias Chatzidrossos Laboratory for Communication Networks School.
LOGO Video Packet Selection and Scheduling for Multipath Streaming IEEE TRANSACTIONS ON MULTIMEDIA, VOL. 9, NO. 3, APRIL 2007 Dan Jurca, Student Member,
Cooperative Overlay Networking for Streaming Media Content Feng Wang 1, Jiangchuan Liu 1, Kui Wu 2 1 School of Computing Science, Simon Fraser University.
Confused, Timid, and Unstable: Picking a Video Streaming Rate is Hard Published in 2012 ACM’s Internet Measurement Conference (IMC) Five students from.
A simple model for analyzing P2P streaming protocols. Seminar on advanced Internet applications and systems Amit Farkash. 1.
QoE Assessment in Olfactory and Haptic Media Transmission: Influence of Inter-Stream Synchronization Error Sosuke Hoshino, Yutaka Ishibashi, Norishige.
Silberschatz, Galvin and Gagne  2002 Modified for CSCI 399, Royden, Operating System Concepts Operating Systems Lecture 19 Scheduling IV.
An Approach to Evaluate Data Trustworthiness Based on Data Provenance Department of Computer Science Purdue University.
Resilient Peer-to-Peer Streaming Paper by: Venkata N. Padmanabhan Helen J. Wang Philip A. Chou Discussion Leader: Manfred Georg Presented by: Christoph.
Sang-Chun Han Hwangjun Song Jun Heo International Conference on Intelligent Hiding and Multimedia Signal Processing (IIH-MSP), Feb, /05 Feb 2009.
Network Coding for Large Scale Content Distribution Christos Gkantsidis Georgia Institute of Technology Pablo Rodriguez Microsoft Research IEEE INFOCOM.
A Layered Hybrid ARQ Scheme for Scalable Video Multicast over Wireless Networks Zhengye Liu, Joint work with Zhenyu Wu.
Scalable and Continuous Media Streaming on Peer-to-Peer Networks M. Sasabe, N. Wakamiya, M. Murata, H. Miyahara Osaka University, Japan Presented By Tsz.
Service Differentiated Peer Selection An Incentive Mechanism for Peer-to-Peer Media Streaming Ahsan Habib, Member, IEEE, and John Chuang, Member, IEEE.
Peer-to-Peer Based Multimedia Distribution Service Zhe Xiang, Qian Zhang, Wenwu Zhu, Zhensheng Zhang IEEE Transactions on Multimedia, Vol. 6, No. 2, April.
Designing QoE experiments to evaluate Peer-to-Peer streaming applications Tom Z.J. Fu, CUHK Dah Ming Chiu, CUHK Zhibin Lei, ASTRI VCIP 2010, Huang Shan,
Multiple Sender Distributed Video Streaming Thinh Nguyen, Avideh Zakhor appears on “IEEE Transactions On Multimedia, vol. 6, no. 2, April, 2004”
PBS: Periodic Behavioral Spectrum of P2P Applications Tom Z.J. Fu, Yan Hu, Xingang Shi, Dah Ming Chiu and John C.S. Lui The Chinese University of Hong.
The Effectiveness of a QoE - Based Video Output Scheme for Audio- Video IP Transmission Shuji Tasaka, Hikaru Yoshimi, Akifumi Hirashima, Toshiro Nunome.
Understanding Mesh-based Peer-to-Peer Streaming Nazanin Magharei Reza Rejaie.
A Distributed Search Service for Peer-to-Peer File Sharing in Mobile Application Presented by Tony Sung On Loy, MC Lab, CUHK IE 1 A Distributed Search.
Performance Evaluation of Peer-to-Peer Video Streaming Systems Wilson, W.F. Poon The Chinese University of Hong Kong.
Performance metrics and configuration strategies for group network communication Tom Z. J. FU Dah Ming Chiu John C. S. Lui.
How to Turn on The Coding in MANETs Chris Ng, Minkyu Kim, Muriel Medard, Wonsik Kim, Una-May O’Reilly, Varun Aggarwal, Chang Wook Ahn, Michelle Effros.
Peer-to-peer Multimedia Streaming and Caching Service by Won J. Jeon and Klara Nahrstedt University of Illinois at Urbana-Champaign, Urbana, USA.
1 How Many Packets Can We Encode? - An Analysis of Practical Wireless Network Coding Jerry Le, John C.S. Lui, Dah Ming Chiu Chinese University of Hong.
Choosing an Accurate Network Model using Domain Analysis Almudena Konrad, Mills College Ben Y. Zhao, UC Santa Barbara Anthony Joseph, UC Berkeley The First.
CUHK Analysis of Movie Replication and Benefits of Coding in P2P VoD Yipeng Zhou Aug 29, 2012.
Some recent work on P2P content distribution Based on joint work with Yan Huang (PPLive), YP Zhou, Tom Fu, John Lui (CUHK) August 2008 Dah Ming Chiu Chinese.
Research on cloud computing application in the peer-to-peer based video-on-demand systems Speaker : 吳靖緯 MA0G rd International Workshop.
Feb. 22, 2005 EuroIMSA A Hybrid Video Streaming Scheme on Hierarchical P2P Networks * Shinsuke Suetsugu Naoki Wakamiya, Masayuki Murata Osaka University,
Chun-Yuan Chang, Cheng-Fu Chou * and Ming-Hung Chen Presenter: Prof. Cheng-Fu Chou National Taiwan University
Rate-distortion modeling of scalable video coders 指導教授:許子衡 教授 學生:王志嘉.
DELAYED CHAINING: A PRACTICAL P2P SOLUTION FOR VIDEO-ON-DEMAND Speaker : 童耀民 MA1G Authors: Paris, J.-F.Paris, J.-F. ; Amer, A. Computer.
Distributing Layered Encoded Video through Caches Authors: Jussi Kangasharju Felix HartantoMartin Reisslein Keith W. Ross Proceedings of IEEE Infocom 2001,
1 Requirements for the Transmission of Streaming Video in Mobile Wireless Networks Vasos Vassiliou, Pavlos Antoniou, Iraklis Giannakou, and Andreas Pitsillides.
Segment-Based Proxy Caching of Multimedia Streams Authors: Kun-Lung Wu, Philip S. Yu, and Joel L. Wolf IBM T.J. Watson Research Center Proceedings of The.
An Empirical Evaluation of VoIP Playout Buffer Dimensioning in Skype, Google Talk, and MSN Messenger Chen-Chi Wu, Kuan-Ta Chen, Yu-Chun Chang, and Chin-Laung.
Prediction Assisted Single-copy Routing in Underwater Delay Tolerant Networks Zheng Guo, Bing Wang and Jun-Hong Cui Computer Science & Engineering Department,
Content Clustering Based Video Quality Prediction Model for MPEG4 Video Streaming over Wireless Networks Asiya Khan, Lingfen Sun & Emmanuel Ifeachor 16.
1 P2P Layer Streaming for Heterogeneous Networks in PPSP K. Wu, Z. Lei, D. Chiu Kent Kangheng Wu 9/11/2010.
Network Instruments VoIP Analysis. VoIP Basics  What is VoIP?  Packetized voice traffic sent over an IP network  Competes with other traffic on the.
MULTI-TORRENT: A PERFORMANCE STUDY Yan Yang, Alix L.H. Chow, Leana Golubchik Internet Multimedia Lab University of Southern California.
Adaptive Multi-path Prediction for Error Resilient H.264 Coding Xiaosong Zhou, C.-C. Jay Kuo University of Southern California Multimedia Signal Processing.
Effects of P2P Streaming on Video Quality Csaba Kiraly, Luca Abeni, Renato Lo Cigno DISI – University of Trento, Italy
Interaction of Overlay Networks: Properties and Implications Joe W.J. Jiang Dah-Ming Chiu John C.S. Lui The Chinese University of Hong Kong.
1 Presented by Jari Korhonen Centre for Quantifiable Quality of Service in Communication Systems (Q2S) Norwegian University of Science and Technology (NTNU)
Department of Communication and Electronic Engineering University of Plymouth, U.K. Lingfen Sun Emmanuel Ifeachor New Methods for Voice Quality Evaluation.
Cross-Layer Optimization in Wireless Networks under Different Packet Delay Metrics Chris T. K. Ng, Muriel Medard, Asuman Ozdaglar Massachusetts Institute.
University of Plymouth United Kingdom {L.Sun; ICC 2002, New York, USA1 Lingfen Sun Emmanuel Ifeachor Perceived Speech Quality.
2007/03/26OPLAB, NTUIM1 A Proactive Tree Recovery Mechanism for Resilient Overlay Network Networking, IEEE/ACM Transactions on Volume 15, Issue 1, Feb.
A Simple Model for Analyzing P2P Streaming Protocols Zhou Yipeng Chiu DahMing John, C.S. Lui The Chinese University of Hong Kong.
Efficient P2P Search by Exploiting Localities in Peer Community and Individual Peers A DISC’04 paper Lei Guo 1 Song Jiang 2 Li Xiao 3 and Xiaodong Zhang.
CS Spring 2014 CS 414 – Multimedia Systems Design Lecture 18 – Multimedia Transport (Part 1) Klara Nahrstedt Spring 2014.
SocialTube: P2P-assisted Video Sharing in Online Social Networks
A Robust Luby Transform Encoding Pattern-Aware Symbol Packetization Algorithm for Video Streaming Over Wireless Network Dongju Lee and Hwangjun Song IEEE.
1 Push-to-Peer Video-on-Demand System. 2 Abstract Content is proactively push to peers, and persistently stored before the actual peer-to-peer transfers.
Cooperative Mobile Live Streaming Considering Neighbor Reception SPEAKER: BO-YU HUANG ADVISOR: DR. HO-TING WU 2015/10/15 1.
SHADOWSTREAM: PERFORMANCE EVALUATION AS A CAPABILITY IN PRODUCTION INTERNET LIVE STREAM NETWORK ACM SIGCOMM CING-YU CHU.
Video Quality Assessment and Comparative Evaluation of Peer-to-Peer Video Streaming Systems Aditya Mavlankar Pierpaolo Baccichet Bernd Girod Stanford University.
Network and Systems Laboratory nslab.ee.ntu.edu.tw Yipeng Zhou, Dah Ming Chiu, and John C.S. Lui Information Engineering Department The Chinese University.
Content aware packet scheduling in peer-to-peer video streaming By: Reza Motamedi Advisor: Hamid Reza Rabiee.
A Comparison of RaDiO and CoDiO over IEEE WLANs May 25 th Jeonghun Noh Deepesh Jain A Comparison of RaDiO and CoDiO over IEEE WLANs.
Video Caching in Radio Access network: Impact on Delay and Capacity
Saving Bitrate vs. Users: Where is the Break-Even Point in Mobile Video Quality? ACM MM’11 Presenter: Piggy Date:
Cost-Effective Video Streaming Techniques Kien A. Hua School of EE & Computer Science University of Central Florida Orlando, FL U.S.A.
The Fundamental Role of Hop Distance in IEEE 80
Presentation transcript:

Perceptual Quality Assessment of P2P Assisted Streaming Video for Chunk-level Playback Controller Design Tom Z.J. Fu, CUHK W. T. Leung, CUHK P. Y. Lam, CUHK Dah Ming Chiu, CUHK Zhibin Lei, ASTRI PV 2010, Hong Kong

Introduction and motivation Chunk-level impairment model Experiments with various Chunk Receiving Patterns (CRP) Heuristic on satisfaction function Future work and conclusion Outline

Internet streaming service becomes popular 1.C/S mechanism, 2.P2P mechanism, mostly implemented.  CDN,  single/multiple tree-based application layer multicast,  peer-to-peer streaming (live streaming / VoD). The evaluation for the above two mechanisms are quite different. Introduction and motivation

1. For the C/S mechanism,  The simple end-to-end link model is suitable for abstraction, link condition metrics:  Packet loss rate  End-to-end packet transmission delay  etc. Packet-level impairments are considered

Introduction and motivation 2.For the P2P assisted mechanism,  The simple end-to-end link model is not suitable:  The transmission pattern is dynamic and complicated. P2P mechanism forms overlay topology

Introduction and motivation 2. For the P2P assisted mechanism,  The simple end-to-end link model is not suitable: 1.The transmission pattern is dynamic and complicated. 2.The granularity of packet-level impairment is too fine, causing mismatch with system design: a)Important building blocks of P2P mechanism are based on chunks: chunk selection, peer selection, buffer map management  affecting the changing of the overlay topology b)System-wide performance metrics are chunk level:  Average playback (dis) continuity;  Buffer filling probability, etc c)In each client side, Playback decision is on top of chunk (How to make decisions with QoE considered?) Therefore, we need chunk-level impairment model !

Chunk-level impairment model

Video encoder – Different media codec, transmission rate could be chosen at the video encoder component Network transmission – chunk level impairment module  Chunk maker – responsible for organizing video stream packets into chunks.  Chunk-level distortion generator – three different ways are designed to implement chunk-level distortion generator  Chunk buffer manager – manages and keeps the received chunks in a local chunk-level buffer (serving other peers later)  Playback controller (client software) – make playback decision for each chunk. Video decoder – After being decoded by the video decoder component, the processed videos (PVS) are then displayed in the monitors to the users. Chunk-level impairment model

Various distortion sources: a) Peers’ dynamic behaviors b) Peers’ network condition (Uplink and Downlink) c) Chunk/peer selection strategy d) Scheduling algorithm, etc…… Consider these factors  too complicated What we can do  make abstraction Chunk receiving pattern (CRP) is the equivalent distortion effects of all the distortion sources considered together. Chunk-level impairment model

General model of chunk-receiving pattern a) r i (x) – download percentage of chunk, any non-decreasing function b) L i – chunk size (all equal to l in the example below) c) X i s – starting downloading time d) X i c – complete downloading time e) X i d – desired playback time, depend on previous chunk f) X i p – real play back time, playback decision Three conditions: r 1 (x): non-delayed chunk, playback normally. r 2 (x): delayed chunk, wait and playback after completion r 3 (x): delayed chunk, wait and playback before completion Chunk-level impairment model

Three ways to generate chunk-receiving patterns 1.Live experiments Collecting and recording the CRP for each chunk a real- life experiment (not repeatable) 2.Simulation results Collecting and record CRPs from the simulation trace in a large network with a large number of users. (Simulation is repeatable under same settings, CRPs are following certain distributions) 3.Artificial generating Manually generate CRPs (by implementing r i (x) with certain increasing curves and parameters, completely repeatable)  suitable for subjective testing studies Chunk-level impairment model The parameter space for subjective testing should not be too large!

In the previous work, we start from the simplest form of CRP: the step function In the previous experiment: Step function assumes that X i s = X i c for all chunks. Two types of chunks: non-delayed (X i c X i d ); Simple playback strategy: X i P = X i d + LWT Non-delayed, Normal playback Delayed, wait and normal playback Delayed, wait and skip Experiments under various CRPs: I

In the previous work, we start from the simplest form of CRP: the step function  Fixed LWT = 3 seconds  Two types of chunk delay D (= X i c – X i d ) distribution 1.Short delay: uniformly distributed in [0, 2] seconds; or 2.Long delay: all equals to 3 seconds, ( = LWT). LWT D2D2 D3D3

Experiments under various CRPs: I In the previous work, we start from the simplest form of CRP: the step function  Fixed LWT = 3 seconds  Two types of chunk delay D (= X i c – X i d ) distribution 1.Short delay: uniformly distributed in [0, 2] seconds; or 2.Long delay: all equals to 3 seconds, ( = LWT).  Number of delay chunks is determined by applying different values of average discontinuity (d = 1 - c): 0, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60% Experiment settings I: 1.50 source video clips (News, Music videos, Movie Trailers and Sports) with average length of 30 seconds; 2.30 subjects (16 males and 14 females), age range ( ); 3.Absolute Category Rating (ACR) with hidden reference as assessment

Experiments under various CRPs: I In the previous work, we start from the simplest form of CRP: the step function Subjective assessment results for each processed video sequence MOS value (left), DMOS value (right): The meaning for Mean Opinion Score (MOS) and DMOS:

Experiments under various CRPs: I In the previous work, we start from the simplest form of CRP: the step function Comparison between short and long chunk delay distribution Insights from the comparison: 1. PVSes with long delay distribution obtain higher MOSes than those with short delay distribution when average d is same. 2. Subjects care more about the number of screen freezing events than the duration of each freezing event.

In this work, we are trying more complicated form of CRPs: the piecewise linear function In this experiment, we assume: Two types of chunks: non-delayed (X i c X i d ); For delayed chunks, three piecewise linear patterns; Playback strategy with D: X i p = X i d + D (D <= LWT) R(D): Chunk completeness when playback (depend on D and Pattern) Experiments under various CRPs: II LWT R A (D) R B (D) R C (D) XidXid X i c =X i d + LWTX i p = X i d + D

In this work, we are trying more complicated form of CRPs: the piecewise linear function In this experiment:  Fixed probability of delayed chunks = 0.1; i.e., 3 out of 30 chunks are delayed;  Fixed LWT = 4 seconds;  Fixed X i c = X i d + LWT;  5 playback decisions: D = X i p – X i d = 0, 1, 2, 3, 4 seconds.  Real implementation of R(D): Experiments under various CRPs: II XidXid X i c =X i d + LWT LWT X i p = X i d + D R A (D) R B (D) R C (D)

In this work, we are trying more complicated form of CRPs: the piecewise linear function Experiments under various CRPs: II Experiment settings II: 1.56 source video clips (News, Music videos, Movie Trailers and Sports) with average length of 30 seconds; 2.42 subjects (19 males and 23 females), age range ( ); 3.Absolute Category Rating (ACR) with hidden reference as assessment

In this work, we are trying more complicated form of CRPs: the piecewise linear function Experiments under various CRPs: II Insights from the comparison among four video categories: 1. News earns the highest scores in all patterns, followed by MV. This is due to the nature of these two categories. 2. The overall difference among three patterns are not very big, probably because we fixed the number of delayed chunks. Patten A Patten B Patten C

We try to model the trade off between Waiting time D and Chunk completeness R. Heuristic Satisfaction function: S(R, D) = alog (R) + bD c, a >= 0, b 0 So we have: This is one possible choice, other forms may investigate in future. Heuristic on satisfaction function

After applying the three patterns to the heuristic satisfaction function with proper parameters: Experiments under various CRPs: II Insights from the comparison among S-function and experiment results: 1. The S-function does not match the results so well, more proper form of S-function needs to be studies and investigated. 2. However, the optimal D values for three patterns are similar as shown in both figures. It indicates that, usually, for pattern like A, small waiting time is better while patterns like C, longer waiting time is more preferred.

Future work  Carry out more experiments with different chunk receiving patterns and parameter settings, e.g. number of delayed chunks, LWT, etc.  Investigate various forms of S-functions to fit the results.  Consider layered coding in P2P streaming (another form of D and R) Conclusion  Chunk-level impairment model is proposed for P2P mechanism.  We introduce two sets of QoE Experiments by applying chunk-level impairment model.  The results are preliminary but still get some interesting insights. Future work and conclusion

The end Thanks! Q & A