QoS Link Interdependence Learning for 5G Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional wireless communication systems fail to consider the inter-dependence between applications when managing QoS for professional applications in 5G networks, leading to suboptimal QoS management and resource allocation.
Innovation Solution
A method to learn QoS performance inter-dependence between communication links by selecting links, generating training requests, measuring QoS performance, and combining the results to obtain inter-dependence, using a joint figure of merit for optimizing application performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the conventional independent QoS management approach is used for each communication link, then the system complexity is reduced and management is simplified, but the application performance optimization is suboptimal because inter-dependence is not considered
Solution Approach 1:
The patent introduces a machine learning model as an intermediary that learns and predicts the inter-dependence relationships between communication links. This model acts as a mediator that provides insights into link inter-dependence without requiring complex direct management of all link interactions, thus maintaining relative system simplicity while enabling performance optimization through predicted relationships.
2Productivity
If the QoS performance inter-dependence between links is considered, then the resource allocation is optimized and application performance is improved, but the measurement and analysis complexity increases
Solution Approach 1:
The patent performs preliminary actions by training the machine learning model in advance using historical QoS data to learn inter-dependence patterns. This pre-learning process captures complex relationships beforehand, so that during actual QoS management, the system can quickly query predictions without performing complex real-time measurements and analyses of all link interactions.
Solution Approach 2:
The patent creates a virtual copy or model of the complex inter-dependence relationships through the machine learning model. Instead of directly measuring and managing all actual link interactions, the system uses the learned model as a copy that reproduces these relationships, simplifying the measurement and analysis process while maintaining accuracy in resource allocation decisions.
3Measurement precision
If training requests are generated and QoS performance is measured on selected links to learn inter-dependence, then the QoS management accuracy is improved, but the time and computational resources required increase
Solution Approach 1:
The patent applies partial action by selecting only specific links for training requests and performance measurement rather than all links in the system. The machine learning model is trained on a representative subset of links to learn general inter-dependence patterns, which reduces the time and computational resources required compared to measuring all possible link combinations, while still achieving sufficient accuracy for QoS management.
Data Source
AI summary
A method for learning QoS performance inter-dependence of a plurality of communication links in a network managed by a central node, wherein the plurality of wireless links interface with a common application applying a joint figure of merit, comprising:selecting at least two links among the plurality of communication links;generating training requests respectively for each selected link;measuring QoS performance on each selected link for the training requests; andcombining the measured QoS performance of the selected links, so as to obtain the QoS performance inter-dependence between the selected links.