UE ML Group Index Transfer Across RAN Handover Context Release

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In current 5G RAN architectures, essential UE-specific information is lost during handovers, leading to inefficiencies in AI/ML model outputs and inter-node RAN actions due to the lack of methods to collect and store UE-specific information after context release.

Innovation Solution

Implement a UE ML group index system that allows for the identification and tracking of UE groups across network entities, enabling the collection and storage of UE-specific measurement data and feedback information even after handover, facilitating efficient AI/ML model training and inference.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If UE context is released during handover to free network resources, then network resource utilization is improved, but UE-specific information is lost which deteriorates AI/ML model precision and RAN procedure stability

Engineering Contradiction:

Improvenetwork resource utilizationVSAvoidUE-specific information

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent segments UE information into two parts: UE context (released after handover) and UE ML group index (retained in target gNB). This segmentation allows the network to free resources while preserving essential AI/ML identification data for future feedback operations.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The target gNB receives and stores a copy of the UE ML group index from the source gNB during handover. This copying mechanism ensures that the identification information survives context release and enables subsequent feedback reporting without requiring the original UE context to persist.

Inventive Principle:

Principle #26Copying

2Loss of information

If UE-specific information is retained after handover to maintain AI/ML model precision, then information availability is improved, but network resource consumption increases

Engineering Contradiction:

ImproveUE-specific informationVSAvoidnetwork resource consumption

Core Design Contradiction:

Loss of informationVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential UE ML group index from the complete UE context and transfers it to the target gNB. This extraction approach retains the minimum necessary information for AI/ML operations while avoiding the overhead of storing entire UE contexts, thus balancing information availability with resource efficiency.

Inventive Principle:

Principle #2Taking out (Extraction)

3Ease of operation

If resource status procedure is used to collect neighboring node information, then information collection is simplified, but UE-specific information cannot be collected leading to reduced measurement precision

Engineering Contradiction:

Improveinformation collectionVSAvoidUE-specific measurement precision

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent performs preliminary action by including the UE ML group index in the handover request message before the handover is executed. This ensures that the target gNB is pre-configured with the identification information needed to collect UE-specific feedback data afterward, combining operational simplicity with measurement precision.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20260046727A1Configuration of UE context surviving during ai/ML operation

Publication Date: 2026.02.12 NOKIA TECHNOLOGIES OY

AI summary

It is provided an apparatus comprising one or more processors and a memory storing instructions that, when executed 2024/027980 by the one or more processors, cause the apparatus to: submit, by a source network entity of a RAN, a request to a target network entity, the request including at least a user equipment machine learning, UE ML, group index, wherein the UE ML group index indicates that an UE is part of an UE ML group; receiving, by the source network entity, a response message from the target network entity based at least on the indicated UE ML group index.