Handover Optimization Using Trainable Module for Cell Border Failures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication networks face challenges with abrupt handover failures at cell borders, particularly in LTE networks, where mobile terminals experience radio coverage holes and rapid handovers, leading to inefficient handover processes and reduced system performance.
Innovation Solution
Implementing a self-detection and self-optimization mechanism that uses a trainable software module or database to evaluate and optimize handovers by exchanging border location-based information between base stations, including signal strength measurements and UE history information, to identify problematic handover locations and reject unsuitable handovers, thereby selecting the best base station for maintaining a stable connection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If handover is performed to target cell with strongest signal, then connection quality is improved, but rapid handover failures occur at cell borders
Solution Approach 1:
The source base station performs preliminary evaluation of neighbor cells using a trained software module or database before executing handover. This pre-evaluation identifies cells that are likely to cause rapid handover failures, allowing the system to avoid selecting problematic target cells even if they have the strongest signal at the moment.
Solution Approach 2:
The system uses feedback from past handover outcomes to continuously improve future handover decisions. By analyzing whether previous handovers to specific cells resulted in rapid failures, the trained software module learns to avoid problematic cells, creating a closed-loop control system that improves reliability over time.
2Adaptability or versatility
If multiple neighbor cells are prepared as handover candidates, then handover flexibility is improved, but all preparations are cancelled when target confirms successful handover
Solution Approach 1:
Multiple neighbor cells are pre-evaluated and ranked before handover execution. The trained software module prepares a prioritized list of candidate cells in advance, so when rapid handover failure occurs, the system can immediately switch to the next best candidate without重新启动 the entire evaluation process.
Solution Approach 2:
The handover candidate list is dynamically maintained and updated based on real-time conditions and historical performance. When a handover failure occurs, the system dynamically switches to alternative candidates from the pre-prepared list, making the handover process adaptive rather than static.
3Measurement precision
If UE history information is transmitted in handover request, then successive handover detection is improved, but abrupt handover failures at specific locations cannot be detected
Solution Approach 1:
The system adds a spatial dimension to handover analysis by incorporating location information (such as GPS coordinates or base station triangulation data) alongside UE history information. This allows the trained software module to identify problematic geographic locations and avoid handovers to cells that cover those areas, detecting abrupt failures based on location patterns rather than just sequential handover counts.
Solution Approach 2:
A trained software module or database acts as an intermediary between raw handover data and decision-making. This intermediary processes both UE history information and location data, synthesizing them to identify patterns of abrupt failures at specific locations that would be invisible to simple successive handover detection.
4Productivity
If self-detection and self-optimization mechanism is implemented, then system performance is improved, but data usage increases for database or training
Solution Approach 1:
The system implements partial self-optimization by using a trained software module that can operate with limited training data initially, providing immediate improvement. The training process is incremental, using only the data necessary to achieve acceptable performance levels rather than requiring exhaustive training datasets, thus balancing productivity gain with data consumption.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to methods of using a trainable software module used in the handover procedure employing the location based information for the self optimization of wireless communication networks, in particular for the self optimization of cellular mobile networks. Embodiments of the invention address the problems of radio link failures and rapid handovers immediately after transferring a wireless appliance from one base station apparatus to the other at certain cell border locations by evaluating which base station apparatus a wireless appliance should have its wireless connection transferred to using either a database or a trainable software module.