SDN Controller MPLS Label Assignment for Point-to-Point Connectivity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network technologies incur significant overhead and processing costs due to the need for full mesh connectivity and signaling of all MPLS labels among endpoint devices in a VLAN, especially when only point-to-point connections are required, leading to unnecessary routing table updates and maintenance of BGP-VPLS state information.
Innovation Solution
A method that provides non-signaled point-to-point connectivity by moving the responsibility of assigning MPLS labels to a centralized SDN controller, eliminating the need for signaling all MPLS labels and reducing the burden on LERs to maintain BGP-VPLS state information, allowing for vendor-independent label management and reduced routing table complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full mesh connectivity and signaling of all MPLS labels among endpoint devices is implemented, then connectivity between endpoint devices is achieved, but processing overhead and routing table complexity increase significantly
Solution Approach 1:
The patent segments the network into customer VLANs and provider core networks, with customer edge devices handling only local VLAN routing while provider edge routers handle inter-VLAN traffic. This segmentation allows endpoint devices to maintain small routing tables for their own VLAN without needing routes to all other endpoints in the network.
Solution Approach 2:
The patent introduces provider edge routers as intermediaries between customer endpoint devices. These intermediary routers perform MPLS label swapping and routing lookups, eliminating the need for endpoint devices to maintain full mesh routing tables. The provider network absorbs the routing complexity while customers enjoy simplified local routing.
2Loss of information
If all MPLS labels are signaled among endpoint devices, then complete routing information is available, but processing overhead and maintenance burden increase
Solution Approach 1:
The patent extracts the MPLS label signaling and routing information maintenance functions from customer endpoint devices and relocates them to provider edge routers. Endpoint devices only need to signal labels for their immediate connections, while provider routers handle the complex inter-VLAN label distribution and routing table maintenance.
Solution Approach 2:
The provider network performs self-service routing by automatically performing MPLS label swaps at provider edge routers based on destination VLAN and endpoint identifiers. This eliminates the need for endpoint devices to manually maintain routing tables for all network endpoints, as the provider network dynamically handles route establishment and updates.
3Adaptability or versatility
If BGP-VPLS state information is maintained by LERs for all endpoints, then full mesh connectivity is supported, but device memory and processing requirements increase
Solution Approach 1:
The patent segments BGP-VPLS state information maintenance between customer edge devices and provider edge routers. Customer devices maintain minimal state for their local VLAN, while provider routers maintain the comprehensive BGP-VPLS routing tables for inter-VLAN traffic. This segmentation distributes memory requirements according to actual needs at each network level.
Solution Approach 2:
The patent adds an MPLS label dimension to the routing architecture, allowing routing decisions to be made based on MPLS labels rather than traditional IP routing tables. This dimensional change enables provider routers to perform efficient label-based forwarding without requiring endpoint devices to maintain large routing tables, effectively moving the memory burden to the provider network infrastructure.
Data Source
AI summary
A method and apparatus for providing a point-to-point connection are disclosed. The method queries for a next available label for a first provider edge router and a next available label for a second provider edge router, performs a first configuration at the first provider edge router and a second configuration at the second provider edge router, wherein the performing the first configuration comprises configuring a first interface and configuring a label for using at least one tunnel by a second interface, wherein the performing the second configuration comprises configuring a third interface and configuring a label for using the at least one tunnel by a fourth interface, and performs a first mapping for the first provider edge router from the first interface to the second interface, and a second mapping for the second provider edge router from the third interface to the fourth interface.


