Wireless QoS Policy for AI/ML Data Transfer Time Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack effective solutions for ensuring higher quality-of-service (QoS) assurance in application layer data transfer, particularly for planned and event-driven scenarios, which are crucial for Application Artificial Intelligence (AI)/Machine Learning (ML) data transfer, as they do not adequately support multiple time windows and QoS performance data analytics.
Innovation Solution
A wireless communication method is proposed that supports multiple time windows and extends QoS performance data analytics by allowing pre-negotiation between Application Functions (AF) and the 5G system (5GS) to agree on desired time windows with specific QoS parameters, enhancing AAMDT mechanisms to include additional QoS parameters such as packet delay budget, packet loss rate, and traffic rate, and enabling dynamic policy adaptation to changing network conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current technologies are used for application layer data transfer, then basic data transfer functionality is provided, but quality-of-service (QoS) assurance for planned and event-driven scenarios is insufficient
Solution Approach 1:
The patent segments the data transfer process into multiple time windows, allowing different QoS parameters to be applied to different segments. This enables precise control over when and how data is transferred, improving QoS assurance for specific scenarios while maintaining overall system reliability.
Solution Approach 2:
The patent introduces dynamic QoS parameter adjustment based on event-driven triggers and network conditions. QoS parameters are no longer static but can be modified in real-time based on predefined events and current network state, enhancing both reliability and adaptability simultaneously.
2Reliability
If traditional data transfer mechanisms are used, then simple connectivity is achieved, but advanced QoS parameters such as packet delay budget, packet loss rate, and traffic rate control are not supported
Solution Approach 1:
The patent extends the AAMDT mechanism to handle multiple QoS parameters (packet delay budget, packet loss rate, traffic rate) within a unified framework. This multi-functional approach allows the same mechanism to control various aspects of QoS without requiring separate systems for each parameter, balancing complexity with comprehensive control.
Solution Approach 2:
The patent enables pre-negotiation of QoS parameters between Application Functions and the 5G system before actual data transfer. By establishing QoS agreements in advance, the system can ensure reliable parameter control without requiring complex real-time adjustments during data transfer, reducing operational complexity.
3Adaptability or versatility
If static data transfer policies are applied, then implementation simplicity is maintained, but dynamic adaptation to changing network conditions and event-driven responses are not enabled
Solution Approach 1:
The patent implements feedback mechanisms that monitor network conditions and event triggers, automatically adjusting data transfer policies based on current state. This closed-loop approach enables dynamic adaptation without manual intervention, maintaining ease of operation while achieving high adaptability through automated policy management.
Solution Approach 2:
The system enables self-service policy adaptation where the network automatically adjusts QoS parameters based on predefined events and current conditions without requiring manual policy management. This autonomous behavior simplifies operation while providing dynamic adaptation to changing scenarios.
Data Source
AI summary
A wireless communication method for data transfer includes receiving, by a function entity, a first request, wherein the first request indicates quality-of-service (QoS) parameters associated with a list of time windows for Application Artificial Intelligence (AI)/Machine Learning (ML) Data Transfer (AAMDT) and selecting, by the function entity, an AAMDT policy based on the QoS parameters.