AI-Based Mobility Load Balancing for Fluctuating Base Station Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mobility load balancing (MLB) policies in communication systems like LTE and 5G NR are not robust due to frequent fluctuations in network load, leading to continuous updates and reconfigurations, which limits network performance improvement.
Innovation Solution
A load balancing policy determining method using AI technology for accurate prediction of network load, enabling formulation of a robust MLB policy based on comprehensive load information to balance network load effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If MLB policy is formulated based on recently reported resource usage, then the policy can be updated quickly, but the robustness of the MLB policy deteriorates due to frequent fluctuations in network load
Solution Approach 1:
The patent applies preliminary action by formulating MLB policies based on predicted future load information rather than just recent historical data. The network device predicts load information for a future time period and uses this prediction to create policies that are robust against load fluctuations, rather than continuously updating policies based on recently reported resource usage that changes frequently.
2Adaptability or versatility
If MLB policy is updated continuously based on reported resource usage, then the policy adapts to current conditions, but frequent base station configuration changes and UE reconfigurations occur
Solution Approach 1:
The patent implements periodic action by determining MLB policies based on predicted load information for specific future time periods rather than continuously updating based on every reported resource usage change. This periodic approach based on prediction intervals reduces the frequency of configuration changes and UE reconfigurations while maintaining adaptability through regular policy updates.
3Loss of energy
If resource usage is reported with periodicity T, then reporting overhead is reduced, but MLB policy formulated at moment t cannot optimize network performance when load fluctuates between t and t+T
Solution Approach 1:
The patent resolves this contradiction by performing preliminary prediction of load information for the future period t to t+T. Instead of waiting for the next periodic report at t+T, the system proactively predicts what the load will be during this interval and formulates policies in advance, maintaining network performance optimization without increasing reporting frequency or overhead.
Data Source
AI summary
A load balancing policy determining method and apparatus are provided including: A first prediction module obtains first information of at least one first base station, where the first information is used to perform load prediction on the first base station, and the first information includes information about a first cell of the first base station and information about a first terminal device located in the first cell. The first prediction module determines load prediction information of the at least one first base station based on the first information of the at least one first base station and obtains load prediction information of at least one second base station, and further determines a load policy of the at least one first base station based on the load prediction information of the at least one first base station and the load prediction information of the at least one second base station.