Multicast Traffic Rerouting via Acknowledgment-Based Join
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multicast routing protocols experience packet loss and duplication during network events like link restorations and additions, as they struggle to transition from old to new multicast trees without incurring significant side effects.
Innovation Solution
Implementing an acknowledgment-requesting join message mechanism that ensures the new multicast tree is established before pruning the old tree, thereby minimizing packet loss and duplication by waiting for acknowledgment before switching traffic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a node sends a prune message to remove itself from an old multicast tree and then sends a join message to add itself to a new multicast tree, then the node can transition to the new tree, but packet loss occurs during the transition period and the node may be temporarily cut off from multicast traffic
Solution Approach 1:
The node sends the join message to the new multicast tree before sending the prune message to the old multicast tree. This preliminary action ensures that the node is already connected to the new tree before disconnecting from the old tree, preventing any gap in multicast traffic reception and eliminating packet loss during the transition.
2Reliability
If a node sends a join message on the new tree before sending a prune message on the old tree to prevent packet loss, then packet delivery continuity is maintained, but duplicate packets are received via both trees for a short period
Solution Approach 1:
The node monitors for duplicate packets during the transition period. When duplicates are detected, the node selectively discards them while maintaining reception of legitimate packets. This feedback mechanism allows the node to tolerate the temporary overlap period without suffering from duplicate packet issues, effectively resolving the contradiction between continuity and duplication.
3Speed
If the multicast routing protocol quickly establishes a new multicast tree after a failure, then convergence speed is improved and packet loss due to failure is minimized, but the protocol cannot converge as quickly after restorations or additions
Solution Approach 1:
Upon detecting a restoration or addition event, the node proactively sends a join message to the new multicast tree before the protocol completes its standard convergence process. This preliminary action allows the node to be among the first to receive traffic on the new tree, effectively reducing packet loss during restoration scenarios where traditional protocols struggle to converge quickly.
Data Source
AI summary
Various techniques for exchanging control messages in order to gracefully reroute multicast traffic are disclosed. For example, one method involves sending a join message for a multicast group towards a root of a new multicast tree and forwarding multicast traffic, addressed to the multicast group, on a current multicast tree until an acknowledgment corresponding to the join message is received. The new multicast tree can be identified in response to detection of a topology change within the network. Until the acknowledgment is received, multicast traffic that is received via the new multicast tree can be dropped.