Physical Link Load Control via Virtual Metric Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for managing congestion in communication networks, particularly those using Interior Gateway Protocol (IGP), are inefficient and complex, leading to computational overload and unsuitable for large networks due to the need for real-time calculation of metrics for virtual links, which can cause temporary routing loops and cascade effects.
Innovation Solution
A method for controlling the state of charge of physical links by modifying metrics associated with virtual links, allowing autonomous decision-making by physical nodes to adjust weights and prioritize traffic, thereby redistributing load and avoiding global context calculations, implemented on a per-virtual network basis to mitigate congestion and maintain service quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dynamic allocation of topology by class coupled with quality of service management is implemented, then quality of service for virtual networks is improved, but computational load increases significantly
Solution Approach 1:
The patent segments the network management problem by treating each virtual network separately. When congestion is detected on a physical link, the system identifies and modifies metrics only for virtual links belonging to that specific physical link, rather than recalculating metrics for all virtual networks globally. This segmentation reduces computational complexity while maintaining QoS guarantees for affected virtual networks.
Solution Approach 2:
The patent applies local quality by making autonomous decisions at each physical node based on local congestion conditions. Each physical node independently monitors its own link utilization and adjusts virtual link metrics locally without requiring global network state information or complex centralized calculations. This local approach reduces overall computational load while effectively managing QoS.
2Productivity
If real-time calculation of metrics for virtual links is performed, then congestion control is improved, but temporary routing loops and cascade effects occur
Solution Approach 1:
The patent prepares metric values for virtual links in advance, storing them as predefined options. When congestion is detected, the system switches to pre-calculated alternative metric values rather than performing real-time calculations. This preliminary preparation eliminates the need for complex real-time computations that could cause routing loops, while still providing effective congestion control through pre-planned metric adjustments.
Solution Approach 2:
The patent enables physical nodes to make autonomous decisions about metric adjustments based on local congestion conditions. Each physical node independently determines when and how to modify virtual link metrics without requiring coordination with other nodes or centralized control. This self-service approach simplifies the control mechanism and reduces the risk of cascade effects that can occur with synchronized global adjustments.
3Productivity
If virtual networks are multiplexed on physical links, then physical link occupancy is optimized, but congestion on physical links affects multiple virtual links
Solution Approach 1:
The patent changes the metric parameter of virtual links dynamically based on the load state of underlying physical links. When congestion is detected on a physical link, the system modifies the metric values of affected virtual links to redirect traffic away from the congested link. This parameter change allows the system to maintain high physical link occupancy by multiplexing virtual networks while protecting virtual link stability through adaptive metric adjustments.
Data Source
Figure 1~6
Figure 2~3
Figure 4a~4b
AI summary
The invention relates to a method for monitoring a load status of a physical link between two physical network nodes (6j, ..., 67) of a physical network (2). A physical link supports a plurality of virtual links, a virtual link connects two virtual nodes (812, ..., 872) belonging to a virtual network (41, ..., 43), and first and second metrics are related to said virtual link. Said monitoring method includes the following steps, implemented by one of the two physical network nodes: a step of detecting a load status of the physical link; a step of identifying a virtual link for which a common metric is the first metric; and a step of sending a command to the virtual node, supported by the physical node and to which the identified virtual link is connected, said command requiring change of the common metric, from the first metric to the second metric, for said virtual link.