Routing Engine Health Prediction Using Support Vector Machine
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large computer networks face challenges in quickly converging to a stable state after network faults due to temporary oscillations in routing information, leading to issues like intermittent network connectivity and increased packet loss and latency.
Innovation Solution
A machine learning-based control system is implemented in routers to predict the health of routing engines, allowing for proactive switchover to secondary engines before failure, reducing the need for manual intervention and improving health status assessment accuracy by using a comprehensive view of routing engine health indicators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional rule-based health monitoring is used for routing engines, then the system structure is simple, but the health status assessment accuracy is insufficient
Solution Approach 1:
The patent replaces traditional rule-based health monitoring mechanisms with a machine learning-based health determination system. The routing engine collects operational parameters and feeds them to a health determination module that uses machine learning algorithms to predict health status, substituting simple threshold checks with intelligent predictive analysis that adapts to complex routing engine behaviors
Solution Approach 2:
The system transitions from monitoring a limited set of parameters with fixed thresholds to collecting and analyzing multiple operational parameters (CPU usage, memory usage, temperature, error rates) dynamically. The health determination module processes these varying parameters to generate comprehensive health assessments, allowing the system to adapt to changing operational conditions
2Reliability
If proactive switchover based on machine learning health prediction is implemented, then network stability improves, but computational resources increase
Solution Approach 1:
The health determination module continuously predicts routing engine health status before actual failures occur. When the predicted health status indicates potential failure, the system proactively initiates switchover to backup routing engines, preventing network disruptions rather than reacting after failures happen
Solution Approach 2:
The routing engine performs self-health assessment by collecting its own operational parameters and feeding them to the health determination module. The system autonomously monitors its own state, predicts potential failures, and triggers switchover operations without external intervention, reducing the need for manual monitoring and management
3Loss of time
If manual intervention is used for routing engine health monitoring, then system complexity is low, but response time to failures increases
Solution Approach 1:
The system implements continuous feedback loops where the routing engine collects operational parameters, the health determination module predicts health status, and the system automatically responds to predicted failures. This closed-loop feedback mechanism enables real-time monitoring and automatic response, eliminating delays associated with manual intervention
Solution Approach 2:
The routing engine autonomously performs health self-assessment and triggers switchover operations when predicted failures are detected. The system monitors itself, evaluates its own health status using machine learning predictions, and automatically initiates failover procedures without requiring human operators to detect or respond to failures
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This disclosure describes techniques that include determining the health of one or more routing engines included within a router. In one example, this disclosure describes a method that includes performing, by a first routing engine included within a router, routing operations, wherein the router includes a plurality of routing engines, including the first routing engine and a second routing engine; receiving, by a computing system, data including health indicators associated with the first routing engine; applying, by the computing system, a machine learning model to the data to determine, from the health indicators, a health status of the first routing engine, wherein the machine learning model has been trained to identify the health status from the health indicators; and determining, by the computing system and based on the health status of the first routing engine, whether to switch routing operations to the second routing engine from the first routing engine.