Deep Neural Network QoS Prediction Map Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radio communication networks face challenges in providing high-resolution Quality of Service (QoS) predictions, especially for mobile traffic applications, as existing methods rely on coarse measurements and lack detailed environmental data, leading to inaccurate cellular coverage maps.
Innovation Solution
An apparatus and method utilizing a trained artificial deep neural network to propagate low-resolution QoS maps and environment information, such as high-definition maps and base station locations, to generate high-resolution QoS prediction maps, even with sparse measurements, by exploiting correlations between coverage maps and environmental properties, and incorporating randomized noise to fill in missing information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coarse measurements are used for QoS prediction, then measurement cost is reduced, but prediction accuracy deteriorates
Solution Approach 1:
The system performs preliminary actions by collecting environment information (building data, terrain, base station locations) and training a neural network model in advance. This pre-processing enables the model to infer high-resolution QoS predictions from sparse measurements during actual operation, resolving the contradiction between measurement quantity and prediction accuracy.
Solution Approach 2:
A trained neural network model serves as an intermediary between sparse measurements and high-resolution QoS predictions. The model translates low-resolution input data into detailed coverage maps by leveraging environmental correlations, eliminating the need for extensive direct measurements while maintaining high prediction accuracy.
2Measurement precision
If extensive data rate measurements are collected, then QoS prediction accuracy is improved, but measurement time and complexity increase
Solution Approach 1:
The system creates a virtual copy of the physical environment by collecting environment information (building footprints, terrain data, base station locations) and using it to train a neural network. This digital model allows rapid QoS prediction without requiring extensive physical measurements, significantly reducing measurement time while maintaining accuracy.
Solution Approach 2:
Environment information and neural network training are performed in advance, creating a ready-to-use predictive model. During actual operation, the system only needs to input sparse measurements into the pre-trained model, eliminating time-consuming data collection and processing while delivering accurate predictions.
3Measurement precision
If high-resolution QoS maps are generated through traditional methods, then prediction accuracy is improved, but device complexity and processing requirements increase
Solution Approach 1:
A trained neural network model acts as an intermediary processing layer that automatically performs the complex task of translating sparse measurements into high-resolution QoS maps. The model encapsulates complex environmental correlations and processing logic, simplifying the overall system architecture while delivering high-resolution outputs without requiring complex processing infrastructure.
4Reliability
If sparse measurements are used, then measurement cost is reduced, but coverage map reliability deteriorates
Solution Approach 1:
The trained neural network serves as a reliable intermediary that compensates for sparse measurements by leveraging environmental correlations learned during training. The model infers missing information by analyzing relationships between environment features and QoS patterns, maintaining coverage map reliability even with limited input data.
Solution Approach 2:
The system changes the parameter representation by transforming sparse measurement data into comprehensive QoS predictions through the neural network. The model processes low-resolution input parameters and outputs high-resolution coverage maps, effectively changing the data density parameter while maintaining reliability through learned environmental correlations.
Data Source
AI summary
An apparatus (100) for determining a high-resolution QoS prediction map for a first radio communications network of a first environment is provided. The apparatus (100) comprises: a first input unit being configured to determine or provide environment information characterizing the first environment; a second input being configured to determine or provide a low-resolution QoS map associated with the first radio communications network of the first environment or a second radio communications network of a second environment; and a determination unit being configured to propagate the low-resolution QoS map and the environment information through a trained artificial deep neural network, wherein low-resolution QoS map and the environment information are provided as input parameters in an input section of the trained artificial deep neural network, and wherein the high-resolution QoS prediction map for the first radio communications network is provided in an output section of the trained artificial deep neural network.