Network Preplan Assignment for Fast SRLG Failure Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication networks struggle with rapid and resilient recovery from Shared Risk Link Group (SRLG) failures due to inadequate monitoring techniques, which fail to identify specific SRLGs causing link failures, leading to potential data loss and unintended network operations.
Innovation Solution
A preplan assignment is generated for the communication network, specifying headends, reporting edges, and notification paths to rapidly identify and recover from SRLG failures by associating each demand with a headend and providing instructions for configuring the network to reroute traffic around failed links.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional monitoring techniques are used, then network operation continues normally, but SRLG failures cannot be rapidly or accurately identified
Solution Approach 1:
The system performs preliminary actions by pre-configuring reporting edges and notification paths before failures occur. Each headend is pre-assigned specific reporting edges that monitor for SRLG failures, and notification paths are pre-established to rapidly communicate failure information. This eliminates the need for complex real-time analysis when failures occur, enabling immediate identification and recovery.
Solution Approach 2:
The patent introduces intermediary elements (reporting edges and notification paths) that mediate between the physical network links and the control plane. Reporting edges act as intermediaries that detect link failures and translate them into notification messages. Notification paths serve as intermediaries that efficiently deliver failure information from endpoints to headends, isolating the complexity of failure detection from the recovery decision-making process.
2Reliability
If comprehensive monitoring of all communication links is implemented, then SRLG failures can be detected, but the number of notification messages and computational resources required increases significantly
Solution Approach 1:
The patent segments the monitoring function by assigning specific reporting edges to each headend based on their responsibility for particular demands. Instead of every node monitoring every link, each headend only processes notifications from its assigned reporting edges. This segmentation reduces the notification message volume each headend must handle and simplifies the computational burden of failure analysis and recovery decision-making.
Solution Approach 2:
The system implements local quality by making each headend's monitoring scope specific and localized to the demands it manages. Each headend has a tailored set of reporting edges assigned to it, creating a customized monitoring configuration that matches its functional responsibilities. This localized approach ensures each headend processes only the notifications relevant to its demands, reducing overall system complexity while maintaining comprehensive coverage.
Data Source
AI summary
Configuration of a communication network can include obtaining a graph representing the communication network, generation of a the preplan assignment for the communication network, and configuration of the communication network according to the preplan assignment. The graph can be associated with shared risk link groups (SRLGs) and demands. Generation of the preplan assignment can include obtaining a headend assignment between each one of the demands and a headend vertex on the graph. Generation of the preplan assignment can further include obtaining a SRLG-distinguishing assignment between each headend and a corresponding reporting edge set. The reporting edge set for a headend can include sufficient edges to distinguish among SRLGs affecting the headend. Generation of the preplan assignment can further include obtaining a notification path assignment between each headend and each reporting edge in the corresponding reporting edge set for the headend.