Radio Node Configuration Using RL Search Space Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication networks lack effective solutions for optimizing energy efficiency by synchronizing multiple radio network nodes based on dynamic traffic demands and mobility patterns, leading to computational challenges and suboptimal energy usage.
Innovation Solution
A method using a reinforcement learning model with a reduced search space, employing importance factors and similarity matrices to dynamically optimize the configuration of radio network nodes, considering environmental factors and real-time traffic patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If traditional energy optimization methods are used for individual radio network nodes, then implementation simplicity is maintained, but energy efficiency is suboptimal due to lack of collaborative optimization
Solution Approach 1:
The patent segments the large feature space into multiple smaller subspaces using clustering algorithms. Each cluster represents a group of similar network states, allowing the system to optimize energy consumption within manageable segments rather than attempting to optimize the entire state space at once, thus reducing computational complexity while maintaining collaborative optimization benefits
Solution Approach 2:
The patent transforms the continuous feature space into discrete clusters through parameter changes in the clustering algorithm. This discretization converts the complex continuous optimization problem into a manageable discrete problem, enabling collaborative energy optimization across multiple nodes without overwhelming computational requirements
2Measurement precision
If the complete feature space is considered for optimization, then optimization accuracy is maximized, but computational complexity increases significantly
Solution Approach 1:
The patent extracts only the most relevant features and state combinations that actually impact energy consumption, rather than processing the complete feature space. This extraction is achieved through clustering which identifies and focuses on significant patterns, maintaining optimization accuracy by concentrating computational resources on the most influential parameters
Solution Approach 2:
The patent applies partial action by considering only the necessary subset of features required for effective energy optimization. Rather than exhaustively analyzing all possible features, the clustering approach identifies and processes only the critical subset needed to achieve optimal energy efficiency, reducing computational complexity while preserving accuracy
3Device complexity
If individual radio network nodes are optimized independently, then implementation complexity is reduced, but overall energy efficiency deteriorates due to lack of coordination
Solution Approach 1:
The patent merges the optimization perspectives of multiple individual radio network nodes into a unified collaborative optimization framework. By clustering nodes based on similar characteristics and optimizing them together, the system achieves better overall energy efficiency while keeping implementation complexity manageable through the structured clustering approach
Data Source
AI summary
A method is herein provided, performed by a network node for handling configuration of radio network nodes in a wireless communication network. The network node such as a O&M node or similar calculates a configuration for one or more radio network nodes by using a machine learning model with a search space of parameters, wherein the search space is reduced based on an importance factor for parameters of the radio network nodes and/or the wireless communication network.