Network Problem Analytics for Automated AI Solution Mapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current network automation solutions require operators to manually track and trigger specific AI/ML solutions for each network problem, limiting scalability due to the complexity and density of modern mobile networks.

Innovation Solution

An apparatus and method for detecting and analyzing network problems, enabling automated classification and mapping to appropriate solutions through a problem analytics function that includes problem detection, classification, and solution search, using a processor and memory configured to perform these tasks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If operators deploy and trigger specific AIML solutions for each network problem, then network automation capabilities are provided, but operator workload and system complexity increase

Engineering Contradiction:
Improvenetwork automation capabilityVSAvoidsystem complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The patent introduces a problem analytics function as an intermediary layer between network monitoring and AIML solution deployment. This function automatically analyzes network problems, classifies them into predefined categories, and triggers appropriate AIML solutions without operator intervention. The intermediary handles the complexity of matching problems to solutions, reducing operator workload while maintaining high automation capability.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system enables self-service by allowing the network management system to automatically detect, analyze, and resolve its own problems. The problem analytics function continuously monitors network conditions, identifies issues, and autonomously triggers appropriate AIML solutions, eliminating the need for operators to manually track and trigger each solution.

Inventive Principle:
Principle #25Self-service

2Reliability

If operators manually track and trigger each AIML solution, then specific network problems can be solved, but scalability is limited

Engineering Contradiction:
Improveproblem resolution accuracyVSAvoidscalability
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent implements preliminary action by pre-defining problem categories and their corresponding AIML solutions in a structured framework. When a network problem occurs, the system automatically matches it against predefined categories and triggers the appropriate solution without requiring operator analysis or decision-making. This preliminary structuring enables scalable automation while maintaining reliable problem resolution.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The problem analytics function serves multiple purposes: it monitors network conditions, detects problems, classifies them into categories, and triggers appropriate solutions. This multi-functional approach consolidates what would otherwise require separate manual processes, enabling the system to handle diverse network problems across different domains through a single automated framework.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If the system provides detailed problem analysis and classification, then solution accuracy improves, but processing time increases

Engineering Contradiction:
Improveproblem classification accuracyVSAvoidanalysis time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the problem analysis process into distinct stages: problem detection, problem classification into predefined categories, and solution triggering. Each stage operates independently with clearly defined inputs and outputs. This segmentation allows the system to perform detailed analysis when necessary while maintaining fast processing for routine problems by following predefined classification paths.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4319086A1Communication network
Publication Date: 2024.02.07 NOKIA TECHNOLOGIES OY
  • EP4319086A1 patent drawingFigure 1
  • EP4319086A1 patent drawingFigure 2
  • EP4319086A1 patent drawingFigure 3

AI summary

An apparatus of a communication network acquires information with respect to a problem related to the communication network, based on a request for problem analysis received from a consumer. The apparatus performs the problem analysis based on the information, comprising checking whether or not the information matches at least one of a plurality of problem profiles and, in case the information does not match at least one of the plurality of problem profiles, conducting learning of a new problem profile related to the problem based on the information. The apparatus returns a report on the problem analysis performed to the consumer.