Presentation is loading. Please wait.

Presentation is loading. Please wait.

Institute of Computer Science Department of Distributed Systems Prof. Dr.-Ing. P. Tran-Gia QoEWeb: Quality of Experience and User Behaviour Modelling for.

Published byDaisy Grant Modified over 9 years ago

Similar presentations

Presentation on theme: "Institute of Computer Science Department of Distributed Systems Prof. Dr.-Ing. P. Tran-Gia QoEWeb: Quality of Experience and User Behaviour Modelling for."— Presentation transcript:

1 Institute of Computer Science Department of Distributed Systems Prof. Dr.-Ing. P. Tran-Gia QoEWeb: Quality of Experience and User Behaviour Modelling for Web Traffic Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de

2 2 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Partner  University of Würzburg Tobias Hoßfeld, Daniel Schlosser, Thomas Zinner, Valentin Burger  Blekinge Institute of Technology Markus Fiedler, Patrik Arlos, Junaid Shaikh  France Telecom SA Sergio Beker, Denis Collange, Frédéric Guyard, Frédérique Millo  Warsaw University of Technology Zbigniew Kotulski, Wojciech Mazurczyk, Tomasz Ciszkowski  http://www3.informatik.uni-wuerzburg.de/research/projects/qoeweb/

3 3 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Motivation  User behavior strongly influences systems  e.g. selfishness, churn, or pollution in P2P systems  time-based or volume-based models in shared systems  But, current web traffic models do not consider QoE / user behavior / impatience !  Derive QoE and user behavior model for web traffic based on  active measurements in a laboratory test  passive measurements within an operator’s network  Apply model and evaluate its impact on selected examples  wireless networks with shared capacity  reputation management to react before the user reacts

4 4 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Agenda  Impact of User Behavior  Example: rate control in UMTS  Active and passive measurements  QoE and User Behaviour Modelling for Web Traffic  non-linear interdependency between QoE and QoS  timely behavior  Reputation Management  Work plan

5 5 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Example: Rate Control in UMTS Systems  Best-effort user and QoS user with guaranteed bandwidth  Time- and volume-based user: e.g. voice calls and FTP user  Impact of user behavior on performance of system? Transmission Power for QoS user?  Rate for BE user! BS x Time Transmission Power for QoS1 QoS1 QoS2 QoS4 QoS3 BE3 BE1 BE2 Transmission Power for BE3 Constant Transmission Power for DPCCH

6 6 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de A Priori Source Traffic Model of a Web User Viewing Time V Web Page WWW Session InterSessionTime I Inline Object GetReq. TCP 1 TCP 2 TCP 3 TCP 4 GetReq. Inline Object Inline Object GetReq.Inline Object Inline Object GetReq. Inline Object Main Object GetReq. Web Page Number X of Web Pages GetReq. Volume R of GetRequest Volume M of Main Object Volume O of Inline Object Number N of Inline Objects N

7 7 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Simulation: Web Users in Rate-Controlled UMTS  Different conclusions according to user behaviour model  volume-based users: rate control degenerates?!  time-based users: rate control works as expected?!  Important to get realistic models volume-based userstime-based users web users rate decreases duration of session increases number of users increases

8 8 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Basic Queueing Theory  Birth-Death-Model  Time-based users  Volume-based users 0 µ0µ0 1 λλ µ1µ1 M(1)-1M(1)M(1)+1 λλλλ µ M(1)-1 µ M(1) µ M(1)+1 µ M(1)+2 n-1n λλ µ n-1 µnµn Basic queueing theory leads to same qualitative results  understanding of system behavior  will be applied in QoEWeb

9 9 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Agenda  Impact of User Behavior  Example: rate control in UMTS  Active and passive measurements  QoE and User Behaviour Modelling for Web Traffic  non-linear interdependency between QoE and QoS  timely behavior  Reputation Management  Work plan

10 10 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Objectives of Measurements  Active measurements  quantification of user impatience due to bad network conditions  quantification of the decrease of satisfaction as a function of time or actions  disturb QoS in laboratory environment  user survey  can also be applied to interpret passive measurements  Passive measurements  investigate the statistical behavior of web traffic  analyze the correlations between the behavior of users and some network performance metrics

11 11 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Passive Measurements: Traffic Modeling  Daily behavior  Typical hours  Model of web transfers / sessions  Traffic metrics: up/down volume, type of end  Network performance criteria: throughput, loss rate, RTT  Application level performance: response time, cancelled downloads  Type of web transfers with similar characteristics  Aggregation in sessions (threshold ?)  Type of web servers  Influence of the hourly variations  Model the behavior of web users, typology

12 12 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Analysis of Correlations  Correlation between traffic metrics and performance criteria  For web transfers / sessions / users  significant performance criteria, dependence function according to  the type of transfer / session  the type of users

13 13 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Agenda  Impact of User Behavior  Example: rate control in UMTS  Active and passive measurements  QoE and User Behaviour Modelling for Web Traffic  non-linear interdependency between QoE and QoS  timely behavior  Reputation Management  Work plan

14 14 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Interdependency between QoE and QoS  Comparing iLBC and G.711 voice codecs  Similar results for both codecs regarding packet loss  IQX (exponential interdepency) cannot be rejected iLBCG.711

15 15 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Impact of Autocorrelated Delays  For different correlation factors, still exponential relationship valid  Clear impact of correlation, i.e. timely dependencies, on QoE

16 16 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Combining Active and Passive Measurements Viewing Time V Web Page WWW Session InterSessionTime I Inline Object GetReq. TCP 1 TCP 2 TCP 3 TCP 4 GetReq. Inline Object Inline Object GetReq.Inline Object Inline Object GetReq. Inline Object Main Object GetReq. Web Page Number X of Web Pages GetReq. Volume R of GetRequest Volume M of Main Object Volume O of Inline Object Number N of Inline Objects N User will abort if QoE is too bad or enjoy browsing and prolongs sessions for good QoE Content and usage of web is changing download of documents: pdf, ppt, … video streams services like chat or RDP Web page may not only be provided by a single server, but from a CDN from different service providers (Akamai, ads server, …)

17 17 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Agenda  Impact of User Behavior  Example: rate control in UMTS  Active and passive measurements  QoE and User Behaviour Modelling for Web Traffic  non-linear interdependency between QoE and QoS  timely behavior  Reputation Management  Work plan

18 18 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Reputation concept  Reputation is a proven mechanism for reflecting aggregated level of trust to network services, users, shared resources (e.g. auctioning systems, P2P networks, distributed wireless networks such as MANET; eBay, eDonkey, SecMon)  Reputation management is a feedback decision process being in charge of examining the given reputation (e.g. QoE, service performance) and triggering/enforcing remedy procedures on the on-line or threshold basis  Key features of reputation  present and historical measurements are weighted and reflect an its evolution and dynamics  in distributed P2P environments reputation is shared among network nodes reinforcing decision process  based on historical measurements estimates future expectations

19 19 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Reputation application in QoEWeb  Reputation building  For a particular Web service/Web traffic a perceived level of user satisfaction ST is expressed by QoE metrics and quantified according to the created model of user behaviour  Own experience OE of reputation is fed by ST, applying historical data shaping with WMA function   For shared reputation V service reputation SR is created with respect to credibility of recommenders IR  Reputation usage in QoEWeb  Evaluation of QoE metrics dynamic with respect to a particular Web services (web surfing, high throughput data, live streaming, interactive real time communication, etc)  Detects deterioration of networks performance before the user perceived QoE goes down below a critical level

20 20 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Agenda  Impact of User Behavior  Example: rate control in UMTS  Active and passive measurements  QoE and User Behaviour Modelling for Web Traffic  non-linear interdependency between QoE and QoS  timely behavior  Reputation Management  Work plan

21 21 Tobias Hoßfeld hossfeld@informatik.uni-wuerzburg.de Work Plan

Download ppt "Institute of Computer Science Department of Distributed Systems Prof. Dr.-Ing. P. Tran-Gia QoEWeb: Quality of Experience and User Behaviour Modelling for."

Similar presentations

802.20 Traffic Models: Status/Discussion July 22, 2003 N. K. Shankaranarayanan (Shankar) AT&T Labs-Research IEEE C802.20-03/73.

Traffic Models: Status/Discussion July 22, 2003 N. K. Shankaranarayanan (Shankar) AT&T Labs-Research IEEE C /73.
Michele Pagano – A Survey on TCP Performance Evaluation and Modeling 1 Department of Information Engineering University of Pisa Network Telecomunication.

Michele Pagano – A Survey on TCP Performance Evaluation and Modeling 1 Department of Information Engineering University of Pisa Network Telecomunication.
Lookback Scheduling for Long-Term Quality-of-Service Over Multiple Cells Hatem Abou-zeid*, Stefan Valentin, Hossam S. Hassanein*, and Mohamed F. Feteiha.

Lookback Scheduling for Long-Term Quality-of-Service Over Multiple Cells Hatem Abou-zeid*, Stefan Valentin, Hossam S. Hassanein*, and Mohamed F. Feteiha.
1 Quality of Service Issues Network design and security Lecture 12.

1 Quality of Service Issues Network design and security Lecture 12.
Ningning HuCarnegie Mellon University1 Optimizing Network Performance In Replicated Hosting Peter Steenkiste (CMU) with Ningning Hu (CMU), Oliver Spatscheck.

Ningning HuCarnegie Mellon University1 Optimizing Network Performance In Replicated Hosting Peter Steenkiste (CMU) with Ningning Hu (CMU), Oliver Spatscheck.
Bayesian Piggyback Control for Improving Real-Time Communication Quality Wei-Cheng Xiao 1 and Kuan-Ta Chen Institute of Information Science, Academia Sinica.

Bayesian Piggyback Control for Improving Real-Time Communication Quality Wei-Cheng Xiao 1 and Kuan-Ta Chen Institute of Information Science, Academia Sinica.
Playback-buffer Equalization For Streaming Media Using Stateless Transport Prioritization By Wai-tian Tan, Weidong Cui and John G. Apostolopoulos Presented.

Playback-buffer Equalization For Streaming Media Using Stateless Transport Prioritization By Wai-tian Tan, Weidong Cui and John G. Apostolopoulos Presented.
Doc.: IEEE 802.11-14/0604r1 Submission May 2014 Slide 1 Modeling and Evaluating Variable Bit rate Video Steaming for 802.11ax Date: 2014-05-12 Authors:

Doc.: IEEE /0604r1 Submission May 2014 Slide 1 Modeling and Evaluating Variable Bit rate Video Steaming for ax Date: Authors:
Chapter 10 Congestion Control in Data Networks1 Congestion Control in Data Networks and Internets COMP5416 Chapter 10.

Chapter 10 Congestion Control in Data Networks1 Congestion Control in Data Networks and Internets COMP5416 Chapter 10.
The War Between Mice and Elephants LIANG GUO, IBRAHIM MATTA Computer Science Department Boston University ICNP (International Conference on Network Protocols)

The War Between Mice and Elephants LIANG GUO, IBRAHIM MATTA Computer Science Department Boston University ICNP (International Conference on Network Protocols)
Priority Queuing Achieving Flow ‘Fairness’ in Wireless Networks Thomas Shen Prof. K.C. Wang SURE 2005.

Priority Queuing Achieving Flow ‘Fairness’ in Wireless Networks Thomas Shen Prof. K.C. Wang SURE 2005.
1 “Multiplexing Live Video Streams & Voice with Data over a High Capacity Packet Switched Wireless Network” Spyros Psychis, Polychronis Koutsakis and Michael.

1 “Multiplexing Live Video Streams & Voice with Data over a High Capacity Packet Switched Wireless Network” Spyros Psychis, Polychronis Koutsakis and Michael.
1 Network Measurements of a Wireless Classroom Network Carey Williamson Nuha Kamaluddeen Department of Computer Science University of Calgary.

1 Network Measurements of a Wireless Classroom Network Carey Williamson Nuha Kamaluddeen Department of Computer Science University of Calgary.
12/10/2006ConfidentialSlide 1 Video Streaming over UMTS: practical issues Stefan Rugel, Klaus Schäfer February 2008.

12/10/2006ConfidentialSlide 1 Video Streaming over UMTS: practical issues Stefan Rugel, Klaus Schäfer February 2008.
Towards Efficient Simulation of Large Scale P2P Networks

Towards Efficient Simulation of Large Scale P2P Networks
On Modeling Feedback Congestion Control Mechanism of TCP using Fluid Flow Approximation and Queuing Theory　 Hisamatu Hiroyuki Department of Infomatics.

On Modeling Feedback Congestion Control Mechanism of TCP using Fluid Flow Approximation and Queuing Theory　 Hisamatu Hiroyuki Department of Infomatics.
1 Traffic Sensitive Quality of Service Controller Masters Thesis Submitted by :Abhishek Kumar Advisors: Prof Mark Claypool Prof Robert Kinicki Reader:

1 Traffic Sensitive Quality of Service Controller Masters Thesis Submitted by :Abhishek Kumar Advisors: Prof Mark Claypool Prof Robert Kinicki Reader:
Traffic Sensitive Active Queue Management - Mark Claypool, Robert Kinicki, Abhishek Kumar Dept. of Computer Science Worcester Polytechnic Institute Presenter.

Traffic Sensitive Active Queue Management - Mark Claypool, Robert Kinicki, Abhishek Kumar Dept. of Computer Science Worcester Polytechnic Institute Presenter.
Traffic Characterization Dr. Abdulaziz Almulhem. Almulhem©20012 Agenda Traffic characterization Switching techniques Internetworking, again.

Traffic Characterization Dr. Abdulaziz Almulhem. Almulhem©20012 Agenda Traffic characterization Switching techniques Internetworking, again.
A Framework for the Analysis of Adaptive Voice over IP Casetti, C.; De Martin, J.C.; Meo, M. Communications, 2000. ICC 2000. 2000 IEEE International Conference.

A Framework for the Analysis of Adaptive Voice over IP Casetti, C.; De Martin, J.C.; Meo, M. Communications, ICC IEEE International Conference.

Similar presentations

Ads by Google