Machine Learning Routing Safety Net for Traffic Disruption Control

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Network administrators are hesitant to fully automate routing decisions using machine learning-based predictive engines due to the lack of a mechanism to ensure acceptable performance, as inappropriate behavior can lead to traffic disruptions or unnecessary rerouting, and there is a need for a balanced approach between aggressive and conservative automation.

Innovation Solution

A safety net engine is introduced to monitor and adjust the operation of a machine learning-based predictive routing engine by comparing its behavior to a behavioral policy, detecting detrimental actions, and adjusting its operation when it violates the policy, allowing for a tradeoff between risk-taking and conservative behaviors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If routing decisions are fully automated using machine learning-based predictive engine, then predictive failure detection and proactive routing become possible, but network administrators cannot ensure acceptable performance and may experience traffic disruptions due to inappropriate automated decisions

Engineering Contradiction:

Improvepredictive failure detection capabilityVSAvoidtraffic disruption from inappropriate routing decisions

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

A safety net engine is introduced as an intermediary component between the predictive routing engine and the network traffic. This safety net engine monitors the behavior of the predictive routing engine, compares its decisions against learned behavioral patterns, and can override or adjust decisions that deviate from expected behavior, thereby preventing traffic disruptions while preserving the benefits of automated predictive routing

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback mechanisms where the safety net engine continuously monitors the performance and behavior of the predictive routing engine. By comparing actual routing decisions against predicted behavioral patterns and performance metrics, the system can detect anomalies and adjust operations to maintain acceptable performance levels, creating a closed-loop control system that prevents harmful outcomes

Inventive Principle:

Principle #23Feedback

2Productivity

If aggressive automated control is implemented, then proactive routing and failure detection improve, but unnecessary rerouting may occur which negatively affects traffic

Engineering Contradiction:

Improveproactive routing efficiencyVSAvoidunnecessary traffic rerouting

Core Design Contradiction:

ProductivityVSObject-generated harmful factors

Solution Approach 1:

The safety net engine applies partial action by selectively intervening only when the predictive routing engine's behavior deviates from acceptable patterns. Rather than completely disabling aggressive automation, the system allows most automated decisions to proceed while applying corrective action only to problematic cases, thus maintaining high productivity while filtering out harmful unnecessary rerouting

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The system dynamically adjusts operational parameters of the predictive routing engine based on monitored behavior. By changing parameters such as routing thresholds, confidence levels, or intervention triggers when abnormal behavior is detected, the system can prevent unnecessary rerouting while maintaining aggressive proactive routing efficiency during normal operation

Inventive Principle:

Principle #35Parameter changes

3Stability of the object's composition

If conservative manual control is maintained, then traffic stability is preserved, but predictive failure detection and proactive routing capabilities are lost

Engineering Contradiction:

Improvenetwork traffic stabilityVSAvoidpredictive failure detection capability

Core Design Contradiction:

Stability of the object's compositionVSReliability

Solution Approach 1:

The system implements dynamic control where the degree of automation transitions from conservative manual control to aggressive automated control based on real-time conditions. The safety net engine continuously assesses the predictive routing engine's behavior and adjusts the level of automated intervention dynamically, allowing the system to operate in a hybrid mode that preserves both traffic stability and predictive capabilities

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS12375389B2Safety net engine for machine learning-based network automation

Publication Date: 2025.07.29 CISCO TECHNOLOGY INC

AI summary

In one embodiment, a device obtains data regarding routing decisions made by a machine learning-based predictive routing engine for a network. The device determines, based on the data regarding the routing decisions, a behavior of the machine learning-based predictive routing engine. The device compares the behavior of the machine learning-based predictive routing engine to a behavioral policy for the machine learning-based predictive routing engine. The device adjusts operation of the machine learning-based predictive routing engine, when the behavior of the machine learning-based predictive routing engine violates the behavioral policy.