One Decoding Step

1
Path Protection Routing withSRLG Constraints to
Support IPTV in WDM Mesh Networks

Meeyoung Cha mycha_at_an.kaist.ac.kr With W. Art
Chaovalitwongse Zihui Ge, Jennifer Yates Sue
Moon
IEEE Global Internet Symposium 2006
2
Terminology

• IPTV (Internet Protocol TV) Distribution of
  broadcast TV traffic using IP technology
• WDM Optical Mesh Network WDM (wavelength
  division multiplexing) links OXC (optical
  cross-connects/switches)
• SRLG (Shared Risk Link Group) A set of links
  that can potentially fail due to a single cause

3
SRLG (Shared Risk Link Group)

• Layered architecture
• Link failure in one layer ? multiple failures in
  the upper layer
• Two disjoint links may belong to a common SRLG

4
Motivation

• Growing need for efficient and resilient IPTV
  design
• 4 million IPTV subscribers in 2005
• Optical networks evolved to mesh structure
• economical and reliable backbone
• new OXC technology supports multicasting!
• Idea use optical infrastructure to support IPTV

5
Service Architecture of IPTV
SHE
Super Head-End (SHE)
Backbone Distribution Network
VHO
How can we provide IPTV servicesover the
backbone network?
Regional Network
Broadcast TV
VHO
Regional Network
Video Hub Office (VHO)
Regional Network
customers
2 SHEs and 40 VHOs across the US
6
Service Requirements of IPTV

• Reliability against failures
• High availability
• Resiliency against single node / link failures
• Two physically disjoint paths from SHEs to VHOs
• Capital expenditure of design
• Each link associated with port / transport cost
• Find minimum cost multicast trees

7
Path Protection Routing Problem

• Fast failover not yet supported in optical
  multicasting.

SHE
SHE
Backbone
VHO
VHO
VHO
VHO
VHO

• How to create two trees such that the total cost
  is minimized and each VHO has physically
  disjoint paths connecting SHEs?

8
Integer Programming Formulation
Minimize total cost
Flow conservation
SRLG diversity
9
Applying our IP Formulation

• Dataset2 SHE / 40 VHO locations in the US
• IP formulation amenable to realistic topologies!
• Comparison against
• Active Path First heuristic for SRLG-diverse
  design
• Link diverse design less reliability
• Source diverse design

10
Cost Comparison Across Designs
Most reliable
Most Reliable
Reduced reliability
Reduced reliability
cost

• More economical than heuristic.
• Cost for increased reliability affordable.

11
Conclusions

• This is the first work to consider supporting
  IPTV on optical mesh network with SRLG
  constraints.
• We propose compact integer programming
  formulation.
• NP-completeness shown
• Compact representation of SRLG diversity
• We evaluate our method using realistic
  topologies.
• Real topologies amenable to our method
• Minimum design cost
• SRLG diversity shown affordable

END
12

Thank you very much!
13
Design Space Assumptions
optical
Technology
Hub-and-spoke
Service layer topology
Link-capacity
Dedicated
Cost
Fast-failover
multicast
physically-diverse paths
14
Link-diverse vs. SRLG-diverse

• Example 2-source 3-destination
• Link diversity is not resilient to SRLG failures
• SRLG diversity is more expensive than link
  diversity

X
X
15
Tools for supporting mathematical formulation

• Integer programming modeling
• GAMS http//www.gams.com/
• Tutorials available online
• IP solver
• ILOG, CPLEX http//www.ilog.com/.
• Not free.

16
Future Directions

• Incorporate facility location
• Given VHO locations where to place two SHEs?
• Guaranteed path performance
• Can we guarantee latency bound on paths?
• Plasticity and scalability
• Routing strategy for new VHOs
• Placing additional SHEs
• Optimal solutions may not be available
  (inevitable / large search space)