SRv6 Path Computation via Virtual Topology Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of upgrading all devices in a network to support Segment Routing over Internet Protocol Version 6 (SRv6) is cumbersome due to the large number of devices and the need for hardware upgrades, which is costly and time-consuming, making it difficult for operators to establish paths that meet service level agreements.
Innovation Solution
A method to determine a new network topology by identifying and utilizing only nodes that support SRv6, ignoring or removing nodes that do not, and measuring transmission overheads to establish paths that minimize computational burden on management devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all devices in the network are upgraded to support SRv6, then path computation capability is improved, but upgrade cost and time increase significantly
Solution Approach 1:
The network is segmented into SRv6-capable nodes and non-SRv6 nodes. The patent creates a virtual topology that includes only SRv6 nodes, separating them from the rest of the network. This allows path computation to be performed independently on the SRv6 segment without requiring upgrades to the entire network, thus improving path computation capability while avoiding the time cost of upgrading all devices.
Solution Approach 2:
The patent extracts SRv6-capable nodes from the overall network topology to form a dedicated virtual topology. By taking out only the necessary SRv6 nodes and their connections, the system enables path computation for SRv6 services without requiring the presence or upgrade of non-SRv6 nodes, thereby improving capability while minimizing upgrade time and cost.
2Reliability
If all devices in the network are upgraded to support SRv6, then path computation capability is improved, but upgrade cost increases
Solution Approach 1:
The network is divided into SRv6-capable segments and non-SRv6 segments. By creating a virtual topology that includes only SRv6 nodes, the patent enables path computation capability improvement in the SRv6 segment without requiring costly upgrades to the entire network, thus reducing overall upgrade cost while maintaining the desired capability.
Solution Approach 2:
The patent extracts only the SRv6-capable nodes from the network to form a virtual topology for path computation. This extraction approach allows the system to achieve path computation capability without investing in upgrading non-SRv6 nodes, thereby significantly reducing upgrade costs while still providing the required functionality.
3Measurement precision
If the complete network topology is used for path computation, then path accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the network topology into a virtual topology containing only SRv6 nodes and their relevant connections. This segmentation reduces the computational scope from the entire network to only the SRv6-capable portion, thereby reducing computational complexity while maintaining path accuracy for SRv6 services within the segmented domain.
Solution Approach 2:
The patent extracts the essential SRv6 nodes and their interconnections from the complete network topology to create a simplified virtual topology. This extraction removes unnecessary nodes and edges that would increase computational complexity, while preserving the accuracy needed for computing paths between SRv6 nodes.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
This application provides a path determining method and a related device. The method includes: determining N1 first-type nodes from N nodes included in a first network topology, where the N nodes include the N1 first-type nodes and N2 second-type nodes, and the first-type node supports SRv6; determining a second network topology corresponding to the first network topology, where the second network topology includes the N1 first-type nodes but does not include the N2 second-type nodes, and the target topology structure includes M first-type target paths; determining transmission overheads of the M first-type target paths, where a transmission overhead of an ith first-type target path is a smallest value of transmission overheads of a Ki paths; and performing path computation based on the transmission overheads of the M first-type target paths and the second network topology. In the foregoing technical solution, a new network topology may be established by using an existing network topology structure, to reduce difficulty in path computation, thereby reducing a burden of a management device.