Conditional Handover Connectivity Graphs for NTN Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing handover mechanisms in non-terrestrial networks (NTN) face challenges such as high propagation distances leading to delayed and inaccurate handover decisions, resulting in connectivity issues and increased energy consumption, while federated learning is hindered by data privacy concerns and inconsistent data distribution across telecommunications operators.
Innovation Solution
A distributed machine learning approach using federated learning and graph neural networks to generate connectivity graphs that predict optimal handover decisions by incorporating UE context and network attributes, minimizing handover failures and energy consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional handover procedures are used, then network coverage can be maintained, but handover failures occur in high-speed scenarios and the process is interruptive
Solution Approach 1:
The system performs preliminary actions by pre-establishing connectivity graphs and identifying alternative network nodes before handover is needed. Machine learning models continuously learn and update connectivity relationships in advance, so when handover becomes necessary, the UE can immediately switch to a pre-identified alternative node without interruption, resolving the contradiction between reliability and productivity.
Solution Approach 2:
The patent implements dynamic connectivity graphs that are continuously updated based on real-time network conditions and UE movement patterns. The machine learning models adapt to changing environments dynamically, adjusting connectivity predictions and alternative node selections based on current speed, location, and network state, enabling reliable handovers in high-speed scenarios without interruptive processes.
2Adaptability or versatility
If machine learning models are trained offline only, then training resource requirements are reduced, but the models cannot adapt to dynamic network conditions
Solution Approach 1:
The system implements periodic action by performing offline training at scheduled intervals to build base model knowledge, then continuously performing lightweight online updates at regular intervals to adapt to changing network conditions. This periodic combination of comprehensive offline training and incremental online learning enables model adaptability while managing system complexity through structured, rhythmic update cycles rather than continuous heavy processing.
Solution Approach 2:
The patent introduces an intermediary layer of pre-trained machine learning models that serve as mediators between offline training data and real-time network conditions. These models process and interpret complex network states, providing simplified predictions and recommendations that reduce the computational burden on the network while maintaining high adaptability to dynamic conditions.
3Ease of operation
If connectivity information is not shared, then network security is maintained, but conditional handover cannot be performed
Solution Approach 1:
The system applies local quality by selectively sharing connectivity information on a need-to-know basis. Instead of broadly exposing network topology data, the patent shares specific connectivity graph information and alternative node identifiers only with UEs that require conditional handover. This localized information sharing enables handover operations for specific users while maintaining network security by limiting exposure to only those entities that require it.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The disclosed aspects relate to the use of distributed machine learning models in conditional handover procedures. There is provided a method for a distributed machine learning assisted conditional handover procedure for a connected device, the method including receiving, by the connected device, one or more measurement configurations from a source station; transmitting, by the connected device, one or more measurement reports to the source station; receiving and storing, by the connected device, one or more CHO commands from the source station, the one or more CHO commands having at least one triggering condition for CHO to one or more candidate target cells, wherein the one or more CHO commands are determined by inputting the one or more measurement reports into a distributed ML-generated reference model; and evaluating, by the connected device, whether the at least one triggering condition in any of the one or more CHO commands is fulfilled.