Handover Prediction Reduces Ping-Pong Switching in Wireless Terminals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current handover techniques in communication systems, such as DAPS and LTM, face challenges in reducing interruption time and terminal complexity, leading to increased ping-pong handovers and signaling overhead due to frequent cell switching based on L1 measurement results, which degrades handover performance.
Innovation Solution
A method that predicts the best cell based on signal strength measurements and utilizes previous handover information to classify and reduce ping-pong handovers by applying a signal strength offset to the source cell, thereby minimizing unnecessary handovers and signaling overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If DAPS handover technique is used to reduce interruption time, then interruption time is reduced to 0 ms, but terminal complexity increases due to requiring multiple PDCP and RLC/MAC/PHY entities
Solution Approach 1:
The patent extracts the complexity of maintaining multiple protocol stack entities from the terminal by implementing prediction-based handover selection at the network side. The network predicts the target cell and prepares handover parameters in advance, so the terminal only needs to execute the predetermined handover without managing multiple active protocol stacks simultaneously, thus reducing terminal complexity while maintaining 0ms interruption time.
Solution Approach 2:
The network performs preliminary actions by predicting the best target cell using machine learning models before handover is triggered. The source base station prepares handover parameters and configures the terminal with prediction-based handover conditions in advance, allowing the terminal to execute handover smoothly without real-time complexity of evaluating multiple cells, thereby reducing interruption time and terminal complexity.
2Speed
If frequent cell switching is performed based on L1 measurement results, then handover responsiveness is improved, but ping-pong handovers increase and handover performance degrades
Solution Approach 1:
The system implements feedback mechanisms where the network receives handover performance information from terminals and uses this feedback to continuously optimize machine learning models for handover prediction. The network adjusts prediction parameters based on actual handover outcomes, reducing ping-pong handovers while maintaining responsive handover execution, thus improving both responsiveness and reliability.
Solution Approach 2:
The patent changes the parameters used for handover decision-making from simple L1 measurement results to comprehensive parameters including historical handover data, terminal mobility patterns, and network conditions. Machine learning models process these parameters to predict optimal handover timing and target cells, reducing unnecessary handovers while maintaining fast response to genuine handover needs, thereby improving handover performance without sacrificing responsiveness.
3Measurement precision
If signal strength measurement is performed for handover decision, then handover accuracy is improved, but signaling overhead increases due to frequent measurement reporting
Solution Approach 1:
The network performs preliminary prediction of handover candidates using machine learning models before triggering actual handover procedures. This prediction phase filters out cells that are unlikely to be optimal targets, so subsequent measurement reporting only needs to focus on a reduced set of candidate cells. This maintains handover accuracy by still performing measurements on relevant cells while significantly reducing signaling overhead by eliminating measurements on non-candidate cells.
Data Source
AI summary
Disclosed are a method and an apparatus for improving mobility management performance by reducing unnecessary handovers in a communication system. a method of a terminal may comprise: receiving cell prediction configuration information; generating measurement results by performing measurements on a serving cell and neighbor cell(s) based on signal strength measurement configuration information; and generating a result of predicting a best cell by performing best cell prediction based on the measurement results.