Predictive Cell Handover Configuration Using AI/ML Submodels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wireless communication systems face issues of poor accuracy and high probability of handover failures during cell handover procedures due to deteriorating radio links and improving links with other cells.
Innovation Solution
A handover configuration method utilizing AI/ML models to predict cell handover events, probabilities, and measurement quantities, along with threshold and offset values, to optimize handover decisions, reducing unnecessary measurements and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional handover methods are used based on current measurement quantities, then the handover procedure is simple to implement, but the accuracy of handover to the target cell is poor and the probability of handover failures is high
Solution Approach 1:
The patent applies preliminary action by configuring handover parameters in advance based on predicted future measurement quantities and AI/ML model outputs. The network device determines configuration information including probability thresholds and offset values before the handover event occurs, using predicted measurement data from future time periods to prepare optimal handover parameters ahead of time. This allows the system to act on predictive insights rather than reactive measurements, improving handover accuracy while maintaining implementation feasibility through pre-computed configurations.
2Reliability
If AI/ML models are used to predict handover events and target cells, then the accuracy and success rate of cell handovers is improved, but the device complexity and computational requirements increase
Solution Approach 1:
The patent uses an intermediary approach by introducing AI/ML models as a mediating layer between measurement data collection and handover decision-making. The model processes input features (measurement quantities, cell information, terminal information) and outputs predicted probabilities and target cell identifiers, acting as an intelligent intermediary that translates raw data into actionable handover parameters. This intermediary layer improves reliability by providing data-driven predictions while managing complexity through modular model integration and configurable parameter thresholds.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting handover configuration parameters based on AI/ML model outputs. The system modifies probability thresholds, offset values, and selection criteria according to predicted measurement quantities and model confidence levels. This allows the handover mechanism to adapt its parameters in real-time based on predictive insights, improving success rates while managing complexity through parameter optimization rather than structural overhaul.
3Loss of time
If prediction of future measurement quantities is performed, then forward-looking handover decisions can be made, but the computational load and power consumption increase
Solution Approach 1:
The patent applies partial action by performing prediction only for the specific measurement quantities and time periods necessary for handover decisions. The system predicts future measurement quantities for configured time periods and selects only the relevant predicted values needed for determining handover parameters, rather than performing comprehensive predictions for all possible scenarios. This selective prediction approach reduces computational load and power consumption while still providing forward-looking insights to minimize handover delays.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
This application provides a handover configuration method, a terminal device, and a network device, relating to the field of wireless communication technologies. This application helps improve accuracy and a success rate of cell handover. The method includes: A first terminal device receives, in a first cell, downlink signaling from a first network device, and determines configuration information based on the downlink signaling, where the configuration information is used to configure an event related to a probability, where the probability includes a probability of cell handover or a probability that a second cell serves as a target cell, and the second cell includes at least one of the following: the first cell, at least one intra-radio access technology neighboring cell of the first cell, or at least one inter-radio access technology neighboring cell of the first cell; the configuration information indicates to determine first information based on a first submodel, where the first submodel is a part of an artificial intelligence AI/a machine learning ML model, and the first information is used for cell handover; and/or the configuration information is used to configure an event related to a future measurement quantity, where the future measurement quantity is obtained by the first terminal device through prediction.