5G Network Configuration Optimization for Autonomous Vehicle Mobility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for modeling 5G network configurations for CAM services, particularly for autonomous vehicle movements, are manually intensive and iterative, making it impractical to optimize performance across various use cases.
Innovation Solution
A method using a deep neural network to correlate 5G network configuration parameters with CAM performance metrics, allowing for predictive modeling of performance without manual guesswork, by training the network with sensor data from vehicles and storing these correlations in a data structure for querying new configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If manual observation and iterative experimentation are used to model the impact of 5G network configuration on CAM service performance, then the modeling can be performed with simple tools, but the process becomes too manually intensive to be realistic
Solution Approach 1:
The patent replaces manual observation and iterative experimentation (mechanical human processes) with machine learning models and automated simulation systems. The ML model learns from simulation data to predict CAM service performance for different 5G network configurations, eliminating the need for repeated manual testing and observation.
Solution Approach 2:
The patent performs preliminary actions by generating extensive simulation data across various 5G network configurations and using this data to train machine learning models in advance. This pre-computed knowledge base enables rapid performance prediction without requiring real-time manual experimentation for each new configuration scenario.
2Adaptability or versatility
If multiple 5G network configuration parameters are tested to optimize performance for different use cases, then the adaptability to various CAM services improves, but the complexity of the modeling process increases
Solution Approach 1:
The patent creates a universal machine learning model that handles multiple 5G network configuration parameters and various CAM service types through a single system. The model is trained on diverse simulation data covering different use cases (autonomous vehicles, drones, remote surgery), enabling it to generalize and predict performance across multiple scenarios without requiring separate modeling approaches for each use case.
Solution Approach 2:
The patent systematically varies multiple 5G network configuration parameters (bandwidth, latency, antenna settings, base station density) in simulations to generate training data. The machine learning model learns the relationships between these parameters and CAM service performance, enabling efficient optimization without manually testing each parameter combination.
3Measurement precision
If extensive simulation data is generated to train the machine learning model, then the prediction accuracy for new configurations improves, but the data processing requirements and training time increase
Solution Approach 1:
The patent employs periodic action by training the machine learning model on pre-generated simulation data in advance, rather than performing real-time analysis when configuration changes occur. The model is periodically retrained or updated with new simulation data as needed, enabling rapid prediction for new configurations without requiring extensive real-time computation.
Data Source
AI summary
5G connected and automated mobility (CAM) services modeling for optimization includes defining different 5G communications network configuration parameters which configure a 5G cellular communications network that encompasses a geographic region supporting autonomous vehicle movements. The method further includes assigning different values to each of the different parameters in different sets of network configurations and correlating each of the different sets with corresponding CAM performance metrics of the autonomous vehicle movements. The method yet further includes storing the correlated sets in a data structure. Finally, the method includes querying the data structure with a new set of values for the parameters and receiving in response to the querying, correlated ones of the corresponding CAM performance metrics.