Next:
List of Figures
Up:
MPLS_trees
Previous:
Acknowledgements
Contents
Approvals
Abstract
Acknowledgements
List of Figures
List of Tables
Introduction
Switching technology
Circuit switching
Datagram packet switching
Virtual circuit packet switching
Virtual circuit packet switching with IP
IP-over-ATM
MPLS
Multicast
Multicast routing tree structure
Multicast with IP
Multicast with ATM
Multicast with MPLS
Contributions of this thesis
Resilience and protection in networks
Overview of rerouting
Protection at the MAC and physical layers: self-healing rings
Network layer protection
MPLS Unicast Fast Reroute
Multicast fault recovery
A multicast routing tree repair algorithm
Problem modeling
Maximization of the resilience of a tree with a single backup path
Main algorithm
Incremental version
Complexity analysis
Computation of the metrics
General case
Average case
Worst case
MPLS Multicast Fast Reroute
Overview
Link failure and recovery detection
Failure and recovery notification
Switchover and switchback
Implementation
Multicast MPLS-Linux
Unicast MPLS-Linux implementation
Multicast MPLS-Linux implementation
FIB management API
The MulTreeLDP protocol
Multicast Explicit Routing
Link failure and recovery detection
Link failure and recovery notification
Switchover and switchback
MulTreeLDP messages and TLV formats
Experiments
Hardware used for the experiments
Experiment 1: Measuring MPLS multicast throughput
Experiment 2: Measuring link failure and recovery detection times
Experiment 3: Measuring link failure and recovery notification times
Experiment 4: Measuring the tree repair time
Experiment 4.1: Measuring the service interruption time due to a link failure
Experiment 4.2: Observing duplicate packets on the tree when a failed link is repaired
Conclusion
Contributions
Directions for future work
Bibliography
About this document ...
Yvan Pointurier 2002-08-11