Source Routed Multicast Using Sub-Tree Identifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Source routed forwarding solutions for multicast traffic are inefficient due to large packet headers, which limit the size of multicast trees and increase overhead, making them unsuitable for large networks.
Innovation Solution
Implementing source routed multicast forwarding using multicast sub-tree identifiers, where a network controller computes and updates multicast trees by dividing them into sub-trees and using sub-tree identifiers in the packet headers to reduce overhead and adapt to changes in network topology.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If source routed forwarding uses fixed size header label for each link, then each intermediate node can determine the next hop, but the packet header becomes too long and generates too much overhead
Solution Approach 1:
The patent segments the multicast tree into multiple sub-trees, each identified by a unique sub-tree identifier. Instead of including a header label for every single link in the entire multicast tree, the header now contains only a sub-tree identifier that represents a group of links. This segmentation reduces the header size from being proportional to the total number of links to being proportional to the number of sub-trees, thereby resolving the contradiction between ease of operation and header length.
2Loss of information
If source routed forwarding includes a header label entry for each link, then the complete path is represented, but the overhead increases and limits the size of multicast trees
Solution Approach 1:
The patent merges multiple link representations into a single sub-tree identifier. Instead of having separate header labels for each individual link in the multicast tree, multiple links that belong to the same sub-tree are grouped together and represented by a single sub-tree identifier. This merging reduces the quantity of header overhead while preserving the essential path information needed for forwarding, as the sub-tree identifier still uniquely identifies the group of links that the packet should traverse.
3Ease of manufacture
If fixed size header label approach is used, then implementation is simple, but it becomes inefficient and limits the size of multicast trees
Solution Approach 1:
The patent introduces a dynamic approach where the header structure adapts to the size and complexity of the multicast tree. Instead of using a fixed-size header label for each link that becomes inefficient for large trees, the system dynamically uses sub-tree identifiers that scale with the network size. The header size now depends on the number of sub-trees rather than the total number of links, allowing the system to efficiently handle both small and large multicast trees without being constrained by fixed-size limitations.
Data Source
AI summary
A controller transmits a data packet to a node in a source routed forwarding network having a plurality of nodes configured to transfer data packets to one another via a plurality of links. The data packet includes a header. The header includes a source routed hop list defining a path of the data packet. The data packet is associated with a flow of data packets from a source to one or more destinations. The hop list includes a sub-tree identifier indicative of a multicast sub-tree.