Machine-Learning Change Impact Management for Open RAN Networks

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Managing the impact of change requests in a telecommunication network, particularly in Open RAN architectures, is challenging due to the complexity of interactions among various network elements from different vendors, leading to time-consuming and error-prone manual reviews, incomplete information, and the inability to identify hidden software bugs and localized problems.

Innovation Solution

An Impact Management system using machine learning models to automatically generate a risk analysis report for change requests, considering previous changes and network element interactions, to identify and mitigate potential risks, thereby reducing errors and improving efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual review processes are used to manage change requests in Open RAN networks, then comprehensive human expertise can be applied, but the process becomes time-consuming and error-prone

Engineering Contradiction:

Improveaccuracy of risk assessmentVSAvoidtime for change request processing

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical review processes with an automated machine learning-based impact management system. The system uses trained ML models to automatically analyze change requests, assess risks, and predict impacts on network elements, substituting human manual analysis with automated computational processes that are both faster and equally or more accurate.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The impact management system performs self-service by automatically gathering information about network elements, change requests, and potential impacts without requiring manual intervention. The system autonomously queries network databases, analyzes the change request details, and generates risk assessments independently.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If comprehensive information gathering is performed for change request analysis, then more accurate risk assessment is achieved, but the complexity of the process increases

Engineering Contradiction:

Improvecompleteness of risk analysisVSAvoidcomplexity of analysis process

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-training machine learning models on historical network data, change requests, and their outcomes before actual change request analysis. This pre-training phase prepares the system to automatically handle comprehensive information gathering and analysis during operational use, reducing the apparent complexity during actual change management.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The machine learning models serve as intermediaries between the complex network infrastructure and the change request management process. The ML models automatically process and synthesize comprehensive information from multiple network elements, change request details, and historical data, presenting simplified risk assessments and recommendations to operators.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If traditional change management processes are used in multi-vendor Open RAN environments, then vendor-specific expertise can be utilized, but the ability to identify hidden software bugs and localized problems is reduced

Engineering Contradiction:

Improvevendor compatibilityVSAvoiddetection of hidden bugs

Core Design Contradiction:

Adaptability or versatilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent segments the change request analysis process into distinct analytical components handled by specialized machine learning models. Different ML models analyze specific aspects such as software compatibility, hardware interactions, and network protocol compliance, allowing the system to detect hidden bugs and localized problems across multi-vendor environments by examining each segment independently and comprehensively.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS20250337667A1Impact-driven management of change requests in mobile network

Publication Date: 2025.10.30 RAKUTEN MOBILE INC

AI summary

Provided are apparatus, method, and device for automatically predict the impact when applying changes in a network. According to example embodiments, the apparatus may be configured to: receive a change request to be applied to a target network element in a network; obtain one or more previous change requests that have been applied to the target network element and information related to the target network element; generate a risk analysis report based on the method of procedure and the obtained one or more previous change requests and information using one or more machine learning models; and determine whether to apply the change request to the target network element based on the generated risk analysis report using the one or more machine learning models; wherein the risk analysis report may include data related to risks of applying the change request to the target network element on the network.