Event-Driven Radio Resource Management Using Machine Learning Interference Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing network systems face inefficiencies in managing radio interference, as they often react immediately to adverse conditions, leading to resource wastage and disruption, especially when interference is temporary or caused by devices in transit, due to the lack of precise methods to differentiate between transient and permanent interference sources.

Innovation Solution

Implementing an event-driven radio resource management (ED-RRM) system using machine learning to detect and predict interference duration, allowing for intelligent decision-making on when to initiate or suppress radio resource management processes, such as channel changes, based on device type and interference patterns.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If automatic channel change processes are triggered immediately when interference is detected, then network communication reliability is improved, but computing resources and power are wasted

Engineering Contradiction:

Improvenetwork communication reliabilityVSAvoidcomputing resources and power

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The system performs preliminary classification of interference sources using machine learning models before triggering channel changes. By analyzing signal characteristics and predicting interference duration in advance, the system determines whether immediate channel switching is necessary, avoiding premature resource consumption while maintaining communication reliability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces the traditional mechanical threshold-based triggering mechanism with an intelligent machine learning-based decision system. This substitution enables nuanced judgment of interference situations, allowing the system to distinguish between transient and persistent interference, thereby optimizing the balance between reliability and resource consumption.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If automatic channel change processes are triggered immediately when interference is detected, then network communication reliability is improved, but unnecessary communication disruption occurs

Engineering Contradiction:

Improvenetwork communication reliabilityVSAvoidcommunication disruption

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary analysis of interference characteristics using machine learning models before executing channel changes. By predicting interference duration and assessing its significance, the system delays or prevents unnecessary channel switching, thereby reducing communication disruption while maintaining reliability when truly needed.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system continuously monitors interference patterns and uses machine learning models to learn from historical data. This feedback mechanism allows the system to improve its judgment over time, accurately distinguishing between transient interference that doesn't require channel changes and persistent interference that does, thereby minimizing unnecessary disruptions.

Inventive Principle:

Principle #23Feedback

3Measurement precision

If machine learning processes are used to predict interference duration and classify device types, then resource management precision is improved, but device complexity increases

Engineering Contradiction:

Improveinterference prediction precisionVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces machine learning models as intermediary components between interference detection and channel change decisions. These models act as intelligent mediators that analyze signal characteristics, classify device types, and predict interference duration, providing precise measurements without requiring complex logic in the core control system.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system uses machine learning models to create virtual representations of interference patterns and device behaviors. By training models on historical data, the system creates simplified copies of complex interference scenarios, enabling accurate prediction and classification without directly implementing complex physical analysis mechanisms.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS11006430B2Event-driven radio resource management triggering

Publication Date: 2021.05.11 CISCO TECHNOLOGY INC

AI summary

Techniques for a machine learning event-driven radio resource management (ED-RRM) module which provides analysis of interference events in a network system are described. The network system reacts to the detection of interference in a network system and uses multiple input parameters to determine a type of device creating the interference, a predicted duration of the interference, and an ED-RRM decision on whether to activate the radio resource management (RRM) processes to alter the network transmission or block the RRM processes to conserve system resources for the network system.