Synthetic Data Generation for RAN Configuration Recommendation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of configuring radio access network (RAN) parameters in 5G mobile networks makes manual optimization impractical, and existing automated solutions face challenges due to sparse data input, leading to poor performance of machine learning-based recommender systems.
Innovation Solution
A generative adversarial network (GAN) is used to generate synthetic data, trained alongside a discriminative model, to supplement sparse data inputs for machine learning processes recommending RAN configurations, enhancing the data availability and accuracy of recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual tuning of CM parameters is performed, then optimization precision can be achieved, but productivity is severely limited and loss of time increases
Solution Approach 1:
The patent creates a virtual copy of the RAN environment through synthetic data generation. Instead of manually tuning parameters in the live network, the system generates synthetic configurations that replicate real network conditions, allowing automated ML models to learn optimal parameter settings without time-consuming manual intervention while maintaining optimization precision
Solution Approach 2:
The system performs preliminary actions by pre-generating synthetic training data and pre-training ML models offline before deployment. This preliminary preparation enables the automated system to quickly recommend optimal CM parameters when deployed, avoiding the need for time-consuming manual tuning during actual network operation
2Ease of operation
If SON functions with expert rules are used, then ease of operation is improved, but device complexity increases and loss of time occurs due to slow closed-loop optimization
Solution Approach 1:
The patent replaces the mechanical expert rule-based SON system with a data-driven machine learning approach. Instead of relying on manually crafted expert rules that require complex maintenance, the system uses ML models trained on synthetic data to automatically recommend CM parameters, simplifying the operational complexity while maintaining ease of use
Solution Approach 2:
The system changes the fundamental parameter of optimization speed by using pre-trained ML models that can rapidly process network conditions and generate recommendations, eliminating the slow closed-loop iteration inherent in traditional SON functions while reducing the complexity of maintaining expert rules
3Ease of operation
If ML-based solutions are implemented with sparse data input, then ease of operation is improved, but measurement precision deteriorates due to insufficient training data
Solution Approach 1:
The patent creates synthetic copies of real network data to augment the sparse available data. By generating artificial training samples that replicate the statistical properties and relationships of real CM parameters, the system provides sufficient training data for ML models to learn accurate recommendations while maintaining the ease of automated operation
Solution Approach 2:
The system performs preliminary data preparation by pre-generating synthetic training data before model training. This preliminary action ensures that the ML models receive adequate training data upfront, improving recommendation accuracy without requiring additional data collection during operation, thus maintaining ease of deployment
Data Source
AI summary
A technique for generating synthetic data as input for a machine learning process that recommends radio access network, RAN, configurations is presented. An apparatus implementation is configured to generate synthetic data from a noise input, using a trained generative machine learning model, wherein the generative machine learning model has been trained together with a discriminative machine learning model as adversaries based on non-synthetic data. The non-synthetic data comprises non-synthetic configuration management, CM, parameter values, non-synthetic RAN characteristic parameter values and non-synthetic performance indicator values. The synthetic data is in the same form as the non-synthetic data and comprises synthetic configuration management, CM, parameter values, synthetic RAN characteristic parameter values and synthetic performance indicator values. The apparatus is also configured to output the synthetic data for the machine learning process.