for VoD in the Cloud Chen Wang 1,2 , Hyong Kim 1 , Ricardo Morla 2 1 - - PowerPoint PPT Presentation

▶
for vod in the cloud
SMART_READER_LITE
LIVE PREVIEW

for VoD in the Cloud Chen Wang 1,2 , Hyong Kim 1 , Ricardo Morla 2 1 - - PowerPoint PPT Presentation

Aug 29, 2022 453 likes •686 views

QoE Driven Server Selection for VoD in the Cloud Chen Wang 1,2 , Hyong Kim 1 , Ricardo Morla 2 1 Department of ECE, Carnegie Mellon University 2 Faculdade de Engenharia de Universidade do Porto IEEE CLOUD 2015, New York, USA 1 Challenges

slide-1
SLIDE 1

QoE Driven Server Selection for VoD in the Cloud

Chen Wang1,2, Hyong Kim1, Ricardo Morla2

1Department of ECE, Carnegie Mellon University 2Faculdade de Engenharia de Universidade do Porto

IEEE CLOUD 2015, New York, USA

1

slide-2
SLIDE 2
  • Cloud for large-scale VoD: Elasticity, Scalability, Flexibility
  • Performance impact due to VM interference

– The performance of video server in a VM varies – The user experience on the video server varies

Challenges

2

Hardware（CPU, disk, memory, network）

Virtualization Layer

64-bit OS 64-bit OS 64-bit OS

slide-3
SLIDE 3

Problem Statement

3

Which video server in the Cloud can provide the best Quality-of-experience (QoE) for a user request?

slide-4
SLIDE 4

Our Objectives

  • Select a server providing the Best QoE

– What is the criteria to select server

  • Existing System: the lowest network latency/server load
  • The Best Server Performance Metric ≠ The best user QoE

– When to select server

  • Existing system: before the start of streaming
  • The QoE at the start of streaming ≠ The QoE in 10 min

– Who selects server

  • Existing system: local DNS server
  • Client himself knows better.
  • Neighboring clients might know better.
  • Scalability

– millions of users, thousands of servers.

4

slide-5
SLIDE 5

Our Proposed System

  • Best QoE

– What: QoE gives the best perception of server performance

  • QoE based Server Monitoring & Server Selection

– When: before the downloading of each video chunk

  • Adaptive Server Selection per chunk

– Who: clients and their neighbors.

  • Cooperation among nearby clients on QoE based

server monitoring

  • Scalability --- Agent based System

– Agents perform distributed control. – Serve user requests locally.

5

slide-6
SLIDE 6

System Design

6

Production Cloud Environment

Cache Agent

  • Discover K candidate servers
  • K servers to client

Client Agent

  • Monitor client’s QoE on Candidate Servers
  • Adaptive server selection
  • Cooperative clients share QoE of Servers

S1 S2 S3 S4 S5 C1 C2 C3 C4 C5 C6

Cooperation

slide-7
SLIDE 7

System Operation

7

S1 S2 S3 S4 S5

1. Location aware

  • verlay
  • f

cache agents 2.Multi-Candidate Content Discovery, CST

CST(S5) Srv1 Srv2

S5 S3

S3 S4

★

S5 S2

★ Videos

3.Connect to the local cache agent. 4.K candidate servers for a video request. 5.QoE driven Adaptive server Selection 6.Cooperation among client agents.

 ★   ★

slide-8
SLIDE 8

S4

★ Videos  

S1 S5

★

Multi-Candidate Content Discovery (MCCD)

8

★ 

S3 S2

CST(S3)

Cand1 Cand2

S3

S3

★

CST(S2)

Cand1 Cand2

S2

 ★

S2 S2 S2 S3 S3 CST(S5)

Cand1 Cand2

S5

 ★

S5

S5

★

S2 S5 S5 S3 S3

slide-9
SLIDE 9

QoE Model

  • Streaming Scheme: DASH
  • Factors impacting QoE per video Chunk

– Bitrate of chunk: – Freezing time:

  • Existing QoE Model

– Logarithm Law: – Logistic Model:

9

r

t

Qvideo_quality(r) = a1ln a2r r

max

a1,a2,c1 ฀ c3

are positive fitted coefficients.

3

freezing 1 2

( ) 5 1 5

c

Q t c t c t t                  

slide-10
SLIDE 10

Our Chunk based QoE Model

  • Chunk based QoE Model

10

Decreasing Bitrate Freezing

slide-11
SLIDE 11

QoE driven Server Selection

  • What: Criterion of Server Selection

11

QoE

Network Latency Server Load Others

Candidate Server 1 Candidate Server 2

Can1 Can2 Latest QoE

High Interference Low Latency

slide-12
SLIDE 12

Adaptive Server Selection

  • When: Adaptive Server Selection per Chunk

– Dynamic Interference

12

Candidate Server 1 Candidate Server 2

Can1 Can2 Chunk 1 Chunk 2 Chunk 3

Dynamic Interference Low Latency

slide-13
SLIDE 13

Cooperative Server Selection

  • Who: Neighbors know better

13

Candidate Server 1 Candidate Server 2

Can1 Can2 Latest QoE

Low Latency

Can1 Can2 Latest QoE

slide-14
SLIDE 14

Comparison Methods

  • Client streaming from 2 candidates

– DASH: Streaming from the closest server

  • DNS based Server Selection + DASH streaming

– QAS-DASH: QoE + Adaptive + DASH – CQAS-DASH: QoE + Adaptive + Cooperative + DASH

14

slide-15
SLIDE 15

Google Cloud Experiment

15

Cache Agent Client Agent Client Agent attached to Cache Agent Location Aware Cache Agent Overlay

slide-16
SLIDE 16

Google Cloud Experiment

16

Session QoE: The average chunk QoE in a video session.

< 3.4 80% QoE DASH 10% 3.5 QAS- DASH <1% 3.62 CQAS- DASH 0% 3.68

slide-17
SLIDE 17

Simulation

  • Simulation in Simgrid

17

Video Servers Clients

slide-18
SLIDE 18

Simulation Results

18

< 3 90% QoE DASH > 40% 2.9216 QAS- DASH 0% 3.1822 CQAS- DASH 0% 3.5004 CQAS Improves 90% QoE >20%

slide-19
SLIDE 19

Conclusion

  • QoE + Adaptability

– QoE: a good indicator of server performance.

– Adaptability: improve user experience in Cloud environment

– Closest DASH  QAS-DASH:

  • Google Cloud: ~3.5  >3.6 (~80th percentile session QoE)
  • Simulation: 2.9216  3.1822 (8.92% in 90th percentile

session QoE)

  • Cooperation

– Cooperation effectively help server selection in clients – CQAS-DASH (QoE + Adaptability + Cooperation)

  • Google Cloud: Doubled bitrate for 80% video sessions
  • Simulation: >20% in 90th percentile session QoE

19

slide-20
SLIDE 20

Netflix Titles

  • http://netflixcanadavsusa.blogspot.mx/

– Netflix Canada: 4499 movies/shows – Netflix USA: 8791 movies/shows

20

slide-21
SLIDE 21

Capacity Limit

  • Google Cloud

– We throttle the maximum bandwidth of each server to 4 Mbps to emulate the server

  • verloading that would happen in real systems.
  • Simulation

– Link to server: 50Mbps – Backbone link: 250Mbps

21