Core Backbone Loopback Distribution via BGP
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The deployment of services requiring a large footprint of Provider Edges in networks using Multiprotocol Label Switching (MPLS) leads to rapid growth in routers and Label Switch Paths, making it difficult to scale the OSPF core backbone area zero to support non-zero areas, resulting in scalability issues.
Innovation Solution
A method is introduced where loopback addresses from edge networks are received by a provider router in the core backbone network, advertised to a transport route reflector, and propagated using Border Gateway Protocol (BGP), allowing separation of loopback address distribution from signaling, enabling a common transport domain for multiple service domains.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If loopback addresses are distributed across all OSPF areas to support MPLS services, then service reachability is improved, but the number of routers and Label Switch Paths grows rapidly causing scalability issues
Solution Approach 1:
The patent segments the network into a core backbone area (area zero) and multiple non-zero service areas. Loopback address distribution is separated from service signaling by using BGP in the core backbone to carry loopback addresses, while OSPF is used within service areas for local routing. This segmentation prevents the exponential growth of LSPs across the entire network while maintaining service reachability.
Solution Approach 2:
The patent introduces BGP as an intermediary protocol between the core backbone and service areas. Instead of distributing loopback addresses directly through OSPF across all areas, BGP acts as a mediator that carries loopback address information from the core backbone to service areas, reducing the routing footprint and preventing rapid growth of routers and LSPs.
2Area of stationary object
If the OSPF core backbone area zero is scaled to support all non-zero areas, then network coverage is improved, but routing scalability deteriorates
Solution Approach 1:
The patent moves loopback address distribution to a different dimension by using BGP (a path vector protocol) instead of relying solely on OSPF (a link state protocol). This dimensional change allows the core backbone to advertise loopback addresses to multiple service areas without requiring all areas to be part of the same OSPF backbone, thereby improving routing scalability while maintaining network coverage.
3Ease of operation
If loopback address distribution is coupled with service signaling, then service setup is simplified, but control over route propagation is reduced
Solution Approach 1:
The patent segments loopback address distribution from service signaling by using separate protocol stacks: BGP for carrying loopback addresses in the core backbone, and OSPF for service area routing. This separation allows independent control of route propagation characteristics while maintaining service setup functionality.
Solution Approach 2:
The patent changes the parameters of route propagation by using BGP's path vector attributes instead of OSPF's link state database. This allows different propagation parameters (such as route policies, filtering, and control) to be applied independently to loopback address distribution while service signaling continues to operate with its own parameters.
Data Source
AI summary
A method of operating a communication network comprises receiving loopback addresses from a plurality of edge networks at a provider router of a core backbone network, the edge networks and the core backbone network being logically distinct from each other, advertising the loopback addresses to a transport route reflector element, propagating the advertisement of the loopback addresses to other provider routers of the core backbone network using a protocol for communicating between autonomous systems, and using the transport route reflector element to advertise at least one of the loopback addresses to a service route reflector element in one of the plurality of edge networks.