Internal Network Slice Mapping With Unsupervised Learning for Diverse QoS

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The complexity of network slice management in 5G networks is exacerbated by the limited number of specified slice types, varying QoS requirements, and the need for abstraction to maintain competitive advantage, leading to inefficiencies and increased operational complexity.

Innovation Solution

A network slice internal recommendation node uses unsupervised learning to dynamically map network service requests to internal slice instances based on historical data, allowing for fine-grained management and hiding internal details from consumers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a limited number of network slice types are specified, then system complexity is reduced, but the ability to meet diverse QoS requirements and support various use cases is limited

Engineering Contradiction:

Improvenetwork slice management complexityVSAvoidnetwork slice type diversity

Core Design Contradiction:

Device complexityVSAdaptability or versatility

Solution Approach 1:

An AI-based network slice selection function is introduced as an intermediary between network service requests and internal network slice instances. This intermediary dynamically maps service requests to appropriate slice instances based on learned patterns from historical data, thereby supporting diverse QoS requirements without increasing the number of explicitly specified slice types. The AI model acts as a mediator that translates varied service requirements into mappings onto a limited set of internal slice instances.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system changes the parameter of network slice selection from static, predefined mappings to dynamic, data-driven mappings. By using unsupervised learning on historical data, the system adapts the mapping parameters between service requests and slice instances based on observed patterns, enabling the same limited set of slice instances to serve multiple different QoS requirements through parameter-based differentiation rather than through increasing slice instance diversity.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If internal network slice instances are exposed externally, then service customization is improved, but proprietary implementation details are revealed and competitive advantage is reduced

Engineering Contradiction:

Improveservice customization capabilityVSAvoidproprietary network slice implementation details

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The AI-based network slice selection function serves as an intermediary layer that decouples external service customization requirements from internal proprietary slice instance implementations. This intermediary enables service customization at the interface level while keeping internal slice instance details hidden. The AI model learns to map service requirements to internal slices without exposing the internal structure, thereby maintaining competitive advantage while still providing customized services.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system segments the network slice management into two distinct layers: an external service interface layer that handles customization and a internal implementation layer that remains proprietary. The AI-based selection function operates at the boundary between these layers, allowing service customization to be achieved through configuration and policy at the interface level without revealing internal slice instance implementations.

Inventive Principle:

Principle #1Segmentation

3Device complexity

If manual network slice management is used, then system simplicity is maintained, but operational efficiency and adaptability to changing requirements are reduced

Engineering Contradiction:

Improvenetwork slice management structureVSAvoidnetwork slice allocation efficiency

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The system implements self-service through an AI-based network slice selection function that automatically maps service requests to appropriate slice instances without manual intervention. The AI model learns from historical data and autonomously makes allocation decisions, thereby improving operational efficiency and adaptability to changing requirements while maintaining a relatively simple management structure. The self-learning capability eliminates the need for complex manual management procedures.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system incorporates feedback mechanisms where the AI-based selection function continuously learns from historical network service requests and slice instance performance data. This feedback loop enables the system to improve its mapping accuracy over time and adapt to changing requirements automatically. The learned patterns from historical data are used to optimize slice instance selection, thereby improving productivity without increasing structural complexity.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12369112B2Network slice internal mapping using unsupervised learning and related apparatuses

Publication Date: 2025.07.22 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

AI summary

A method performed by a first network node for a telecommunications network is provided. The method includes receiving a request for a network service. The method further includes dynamically mapping the request for the network service to an internal network slice instance based on unsupervised learning of historical data.