A Comparison of Layering and Stream Replication Video Multicast Schemes

1
A Comparison of Layering and Stream Replication
Video Multicast Schemes

• Taehyun Kim and Mostafa H. Ammar

2
Content

• Research Goal
• Replication VS Layering
• Experimental Comparison
• Results
• Conclusion

3
Research Goal

• A systematic comparison of video multicasting
  schemes designed to deal with heterogeneous
  receivers
• Replicated streams
• Cumulative layering
• Non-cumulative layering

4
Stream Replication

• Multiple video streams
• Same content with different data rates
• Receiver subscribes to only one stream
• Example
• SureStream of RealNetworks
• Intelligent streaming of Microsoft

5
Replicated Stream Multicast

• R1, R2 and R3 are from different domain
• Receivers subscribe to only one stream
• R1 joins the high quality stream (8.5Mbps)
• R2 receives the medium quality stream (1.37Mbps)
• R3 joins the low quality stream (128kbps)

6
Cumulative Layering

• 1 base layer enhancement layers
• Base layer
• Independently decoded
• Enhancement layer
• Decoded with lower layers
• Improve the video quality
• Example
• MPEG-2 scalability modes

7
Non-Cumulative Layering

• Video is encoded in two or more independent
  layers
• Receiver can join any subset of the video layer
  without joining the layer 1 multicast group
• Example
• Multiple description coding (MDC)

8
Layered Video Multicast

• R1 subscribes to all video layers (10 Mbps)
• R2 joins enhancement layers 1 and the base layer
  (1.5 Mbps)
• R3 just receives the base layer (128kbps)

9
Layering or Replication?

• Common wisdom states
• Layering is better than replication
• However, it depends on
• Layering bandwidth penalty
• Specifics of encoding
• Protocol complexity
• Topological placement of receivers

10
Layered Video Multicast

• Considering 20 overhead, the data rates
  contributing to the video quality are 8Mbps,
  1.2Mbps and 102.4Kbps
• Stream Replication video quality are 8.5Mbps,
  1.37Mbps and 128kbps

11
Bandwidth Penalty

• Information theoretic results
• Recent results showed that the performance of
  layered coding is not better than that of
  non-layered coding
• Increase the number of layers gt significant
  quality degradation
• Packetization overhead
• Enhancement layers carry
• Picture header
• GoP information
• Macroblock information

12
Experimental Comparison

• Non-layered streams has better video quality
• Difference in data rates ranges from 0.4 at
  27.7dB PSNR to 117 at 23.2dB PSNR
• For a good quality video, the overhead is around
  20

13
Providing a Fair Comparison

• Need to insure that each scheme is optimal
• Two dimensions
• Stream assignment algorithm
• Determine the reception rate of each receiver by
  aggregating the data rates of the assigned
  streams
• Rate allocation algorithm
• Determine the data rate of each stream
• Goal
• Maximize the bandwidth utilization by each scheme
  for
• a given network
• a particular set of receivers and
• given available bandwidth on the network links

14
System Model

• Model the network by a graph G (V, E)
• V is a set of routers and hosts
• E is a set of edges representing connection links

• Isolated rate
• The reception rate of the receiver if there is no
  constraint from other receivers in the same
  session

15
Stream Assignment

• Cumulative layering
• Define
• ?i is the data rate of a stream and m is the
  number of layers
• Assign as many layers as possible
• Compute the isolated rates
• Assign that does not exceed the isolated
  rate

16
Stream Assignment

• Stream replication
• Define
• ?i is the data rate of a replicated stream and m
  is the number of replicated streams
• Set of receivers assigned to stream i,
• Two objectives
• Minimum reception rate for all receivers is
  greater than zero
• Maximum
• Greedy algorithm
• Allocate ?1 to all receivers to satisfy the
  minimum reception rate constraint
• Receiver is assigned a stream that has not been
  assigned and has the maximum value of group size
  and stream rate product

17
Stream Assignment

• Non-cumulative layering
• Define
• ?i is the data rate of a non-cumulatively layered
  stream and m is the number of streams
• Set of receivers assigned to stream i,
• Two objectives
• Minimum reception rate for all receivers is
  greater than zero
• Maximum

18
Rate Allocation

• Cumulative layering
• Optimal receiver partitioning algorithm (Yang,
  Kim and Lam 2000) determines the optimal rates of
  layer i, ?i
• Receivers are partitioned into K groups (G1,
  G2,, GK)
• Objective is to maximize the sum of receiver
  utilities
• Dynamic programming algorithm is used to find an
  optimal partition
• For a given partition, an optimal group
  transmission rate can be determined
• Stream replication
• Stream rates, ?i, are allocated based on the
  optimal cumulative layering rate

19
Rate Allocation

• Non-cumulative layering
• Receiver can subscribe to any subset of layers
  without joining the base layer
• ?1,2,4 gt isolated rates of 1,2,3,4,5,6,7
• 2m-1 different link capacities with m
  non-cumulative layers
• ?i are allocated based on ?i gt

20
Performance Metrics

• Average reception rate
• Average rate received by a receiver
• Average effective reception rate
• Amount of data received less the layering
  overhead
• Total bandwidth usage
• Adding the total traffic carried by all links in
  the network for the multicast session
• Efficiency
• total effective reception rate / total bandwidth
  usage

21
Network Topology

• Georgia Tech Internetwork Topology Models
  (GT-ITM)
• 1 server
• 1640 nodes with 10 transit domains
• 4 nodes per transit domains, 4 stubs per transit
  node, 10 nodes in a stub domain
• transit-to-transit edges 2.4Gbps
• stub-to-stub edges 10Mbps and 1.5Mbps
• transit-to-stub edges 155Mbps, 45Mbps and
  1.5Mbps
• number of layers 8
• amount of penalty 20

22
Date Reception Rate

• Cumulative layering can receive more data
• Number of layers in cumulative layering is twice
  as many as that of non-cumulative layering

Cumulative
Non-cumulative
Replication
23
Bandwidth Usage

• Bandwidth consumption of cumulatively layered
  multicasting is the largest

Cumulative
Non-cumulative
Replication
24
Effective Reception Rate

• Only 80 of data contributes to improving the
  video quality

Cumulative
Non-cumulative
Replication
25
Efficiency

• Replicated stream video multicasting is more
  efficient

Cumulative
Non-cumulative
Replication
26
Effect of Overhead

• Layering overhead of more than 7 tends to favor
  the replicated stream approach

27
Effect of the number of layers

• Efficiency of stream replication is always
  greater than that of cumulative layering
• The effect is not so significant

28
Narrow Distribution
Narrow distribution
Wide distribution

• The layering approach achieves better bandwidth
  efficiency when multiple streams share the
  bottleneck link
• In narrow distribution, the reception rates in
  Figure (a) is larger than that of Figure (b) by
  1.63Mbps

29
Efficiency

• Compared to the wide distribution results, the
  performance of replicated stream video multicast
  is degraded

Cumulative
Non-cumulative
Replication
30
Protocol Complexity

• Receiver-driven Layered Multicast (RLM)
• Receivers decide whether to drop additional layer
  or not
• Join experiment incur a bandwidth overhead
• Receivers send a join message and multicast a
  message identifying the experimental layer to the
  group
• Layered video multicasting
• Receiver can join multiple groups
• Large multicast group size
• Replicated stream video multicasting
• Receiver only join one group
• Small multicast group size

31
Average Group Size

• Group size in cumulatively layered video
  multicasting is twice as large as that in stream
  replication
• More bandwidth to multicast a message reporting
  the join experiment

32
Conclusion

• Identified the factors affecting relative merits
  of layering versus replication
• Layering penalty
• Specifics of the encoding
• Protocol complexity
• Topological placement
• Developed stream assignment and rate allocation
  algorithms
• Investigated the conditions under which each
  scheme is superior