Application-Specific pQoS Model for Mobile Communications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting quality of service (QoS) in mobile communication, particularly for autonomous vehicles, face inaccuracies due to varied data traffic properties and application-dependent key performance indicators (KPIs), leading to biased prediction models that are not tailored to specific application needs.
Innovation Solution
A method that determines predictive quality of service (pQoS) by receiving data property and KPI indications from applications, using these to create tailored pQoS models that account for different data types, communication protocols, encoding, and communication patterns, and adapts to feedback for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a general pQoS prediction model is used for all applications, then the system complexity is reduced and ease of operation is improved, but the prediction accuracy deteriorates due to varied data traffic properties and application-specific KPIs
Solution Approach 1:
The patent segments the general pQoS prediction model into application-specific models. Each application receives a tailored pQoS model that considers its specific data traffic properties and KPIs, rather than using a single general model for all applications. This segmentation resolves the contradiction by improving accuracy through specialization while managing complexity through modular application-specific models.
Solution Approach 2:
The patent applies local quality by customizing pQoS models according to specific application requirements. Different applications receive models with different parameters and characteristics suited to their specific needs (e.g., autonomous vehicles, streaming services). This allows each application to have optimized prediction accuracy without requiring all applications to use complex customized models.
2Measurement precision
If application-specific pQoS models are created for different data traffic properties and KPIs, then prediction accuracy is improved, but the device complexity and difficulty of detecting and measuring increases
Solution Approach 1:
The patent applies preliminary action by having applications indicate their data properties and KPIs in advance before pQoS prediction. This allows the system to pre-configure appropriate models and parameters based on the application's specific requirements, reducing the complexity of real-time detection and measurement while maintaining high prediction accuracy.
Solution Approach 2:
The patent implements feedback mechanisms where applications provide information about their data properties and KPIs, and the system uses this feedback to select or configure appropriate pQoS models. This feedback loop simplifies the measurement process by using application-provided information rather than requiring complex system-wide detection and measurement of all parameters.
3Measurement precision
If pQoS models are trained on real data transfer without knowing exact traffic properties, then the system is easier to operate, but the prediction accuracy deteriorates due to biased prediction models
Solution Approach 1:
The patent applies preliminary action by collecting and analyzing application-specific data properties and KPIs before training pQoS models. This preliminary configuration ensures that models are trained with accurate information about the specific application requirements, improving prediction accuracy while maintaining ease of operation through automated model selection and configuration based on pre-collected application information.
Data Source
AI summary
A device and a method determine a predictive quality of service (pQoS) in mobile communication, wherein an indication of at least one data property concerning data to be exchanged between an application of a terminal and an external device is received from the application. A communication link is determined between the application and the external device. A pQoS model is determined based on the indicated data property and the communication link; a pQoS is determined by applying the PQOS model; and providing the determined pQoS to the application. Disclosed embodiments also include receiving from the application of the terminal an indication of at least one key performance indicator (KPI) that is applied by the application to measure an actual quality of service (QoS), and determining the pQoS model further based on the indicated at least one KPI.