Video On Demand

1
Video On Demand
2
Video on Demand

• One video server
• Many video data
• Many clients
• Client want to watch at any time

3
Assumptions

• Constant bitrate stream
• Perfect network transport

4
Unicast Solution

• One channel per client
• No start-up latency
• No client buffer
• Low client bandwidth
• Large server bandwidth
• Not scalable

5
Multicast Solution

• Batching
• aggregate client requests
• serve using multicast
• clients have to wait
• No client buffer
• Low client bandwidth
• Scheduling Policies for an On-Demand Video
  Server with Batching
• Dan, Sitaram, Shahabuddin, IBM

6
Multicast Solution

• User-centered approach
• Scheduling data based on user requests
• Data-centered approach
• Dont care about user
• Just broadcast popular video

7
Multicast Solution

• Batching
• Staggered Broadcast

8
Staggered Broadcast
Video
C0
C1
C2

9
Staggered Broadcast

• 2 hour video
• 5 minutes waiting time
• Number of channels
• 2 x 60 / 5 24
• Required bandwidth

1.5Mbps x 24 36Mbps
10
Multicast Solution

• Batching
• Staggered Broadcast
• clients have to wait
• No client buffer
• Low client bandwidth
• Huge server bandwidth

11
Multicast Solution

• Batching
• Staggered Broadcast
• Periodic Broadcast

12
Periodic Broadcast
Video
C0
C1
C2

13
Pyramid Broadcast
Video
C0
C1
C2

14
Pyramid Broadcast
Video
C0
C1
C2

15
Analysis of Pyramid Broadcast

• Notations
• B Total available bandwidth
• Bv Bandwidth of video
• Tv Total length of each video
• K Number of segments per video
• Ti Length of segment i
• ? Factor in geometric series

16
Channel Bandwidth
playback time Ti
i
i1
i1
Doh! Just miss it!
download time Ti1Bv/Bi
Download time for segment i1 needs to be smaller
than Ti for it to arrive in time.
17
Channel Bandwidth
18
? 2
19
Start-up Latency

• Worst case waiting time

20
Optimal ?
T1
?
2.5
21
Storage Requirements
22
Pyramid Broadcast

• Large client bandwidth (?KBv)
• Huge client buffer (7080 Tv)

23
Permutation-based Pyramid Broadcast
C0
C1
C2
24
Channel Bandwidth
playback time Ti
i
i1
i1
i1
i1
Doh! Just miss it!
download time Ti1Bv/Bi
X
X needs to be smaller than Ti for segment i1 to
arrive in time.
25
Channel Bandwidth
26
Client Latency
27
Storage Requirement

• One channel at a time
• Can pause and wait

28
Storage Requirement
pause
k-1
k-1
k
k
resume
Within time X, better not consume all data in
buffer.
X
29
Storage Requirement
pause
k-1
k-1
k
k
resume
Within time X, better not consume all data in
buffer.
X
30
Storage Requirement
31
Comparisons
Scheme Storage Servers BW Clients BW
Pyramid 70 ?KBv 4-5 Bv
Pyramid 20 (?p)KBv 2-3 Bv

Carter, Long and Paris Video on Demand
Broadcasting Protocols
32
Pyramid Broadcasting
33
Skyscraper Broadcasting

• Observations
• storage requirement is affected by size of the
  largest chunk
• So, lets limit the size of the largest chunk!

34
Pyramid
Skyscraper
35
Skyscraper Broadcasting

• Uses series
• 1 2 2 5 5 12 12 25 25 52 52 W W W

36
Skyscraper Example
37
Skyscraper Example
38
Comparisons
Scheme Storage Servers BW Clients BW
Pyramid 70 ?KBv 4-5 Bv
Pyramid 20 (?p)KBv 2-3 Bv
Skyscraper 10 KBv 1-2 Bv
Carter, Long and Paris Video on Demand
Broadcasting Protocols
39
Other schemes

• Pagoda Broadcasting 1 3 5 15 25 75 125
• Harmonic Broadcasting
• Equal segment size, varies bandwidth instead!

40
Multicast Solution

• Batching
• Staggered Broadcast
• Periodic Broadcast
• Sending rate playback rate
• May need multiple channels
• Need additional client buffer
• Need to wait

41
Multicast Solution

• Batching
• Staggered Broadcast
• Periodic Broadcast
• Patching

42
Patching
Time
mcast
unicast
Client Request
43
Patching
Patching Window W
Time
mcast
mcast
Client Request
44
Grace Patching

• if W lt B
• client buffer videoW .. end
• 30 minutes video
• 1 client arrival per minute
• Total data delivered

45
Scenario 1 B 15mins

• 30 minutes video
• 1 client arrival per minute
• Total data delivered

46
Scenario 2 B 5mins

• 30 minutes video
• 1 client arrival per minute
• Total data delivered

47
Scenario 3 B 2mins

• 30 minutes video
• 1 client arrival per minute
• Total data delivered

48
Summary

• Batching (User Centered)
• Staggered Broadcast (Data Centered)
• Periodic Broadcast (Data Centered)
• Patching (True VOD)