Segment Routing MPLS Point-to-Multipoint Path Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network communication technologies face scalability issues with point-to-multipoint (P2MP) paths in Multiprotocol Label Switching (MPLS) networks, as they require stateful routing protocols that increase memory and processing resources, limiting the number of paths and degrading network performance.
Innovation Solution
Implementing a Segment Routing (SR) mechanism that uses multicast Segment Identifiers (SIDs) with a number of branches and SIDs fields to create and maintain stateless P2MP paths, allowing nodes to forward packets based on label stacks without maintaining states, thereby reducing memory and processor usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If stateful routing protocols are used to support P2MP paths, then P2MP communication functionality is achieved, but memory and processing resources increase significantly
Solution Approach 1:
The patent segments the P2MP path into multiple unidirectional point-to-point paths using segment routing technology. Each segment is identified by a segment identifier (SID), allowing the network to forward packets along segmented paths without maintaining global P2MP states at intermediate nodes. This segmentation enables P2MP functionality while reducing resource consumption by distributing path information in the packet header rather than maintaining it in node memory.
Solution Approach 2:
The patent introduces segment identifiers (SIDs) as intermediaries that carry path information through the network. Instead of nodes maintaining state information about P2MP paths, the SIDs act as mediators that encode the necessary routing information in the packet header, allowing intermediate nodes to forward packets based on label switching without maintaining P2MP path states.
2Adaptability or versatility
If forwarding tables are expanded to support more paths, then network scalability improves, but processing resources to search forwarding tables increase
Solution Approach 1:
The patent replaces the traditional mechanical forwarding table lookup process with a label switching mechanism based on segment identifiers. Instead of searching through expanding forwarding tables, nodes perform O(1) label matching operations. The segment list in the packet header provides explicit path instructions, eliminating the need for complex forwarding table searches and reducing processing power requirements as the number of paths increases.
3Reliability
If state maintenance is implemented at each node for P2MP paths, then path forwarding accuracy improves, but device complexity increases
Solution Approach 1:
The patent extracts path state information from node memory and places it in the packet header as a segment list. Each node only needs to process the top segment identifier in the list, performing simple label popping and forwarding operations. This extraction eliminates the need for complex state maintenance at each node while preserving forwarding accuracy, as the complete path information travels with the packet.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed is a mechanism implemented by an ingress node in a network. The mechanism comprises receiving a packet in Multiprotocol Label Switching (MPLS) format. A segment list is pushed onto the packet. The segment list describes a segment routing point to multipoint (SR P2MP) path as a list of multicast segment identifiers (SIDs) each comprising a multicast adjacency label. The packet is transmitted on a downstream interface along the SR P2MP path.