Network AI/ML Capability Activation With Partial Scope Progression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies lack a flexible and efficient method for activating artificial intelligence and machine learning capabilities in network environments, particularly in 5G systems, making it difficult to predict and quantify their benefits without implementation.
Innovation Solution
A method for obtaining and managing activation levels of AI/ML capabilities in network devices, allowing for partial and gradual activation, with performance criteria checks and adjustments based on predefined scopes and network contexts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AI/ML capabilities are fully activated in network devices, then network performance and intelligence are improved, but it is difficult to predict and quantify benefits without implementation, and performance degradation may occur
Solution Approach 1:
The patent segments the AI/ML capability activation into multiple predefined ordered activation levels (e.g., level 1, level 2, level 3) where each level represents a gradually increasing scope of activation. This allows the network device to activate AI/ML capabilities incrementally rather than all at once, enabling performance evaluation at each stage and reducing the risk of overall performance degradation while providing a structured approach to manage the complexity of activation.
2Ease of operation
If AI/ML capabilities are activated without predefined activation levels, then deployment is simplified, but fine-tuned control and controlled testing of partial scopes are not possible
Solution Approach 1:
The patent introduces dynamic activation levels that can be adjusted based on performance criteria. The network device can transition between different activation levels (e.g., from level 1 to level 2 to level 3) as performance criteria are met, providing both structured control and adaptability. This dynamic approach maintains ease of operation through automated level progression while enabling fine-tuned control over the activation scope.
Solution Approach 2:
The patent changes the parameter of activation scope by defining multiple predefined ordered activation levels with progressively increasing scopes. Each level represents a different parameter state of the AI/ML capability activation, allowing the system to control and test partial scopes systematically while maintaining a structured and manageable process.
3Reliability
If performance criteria checks are implemented before full deployment, then performance degradation is avoided, but additional evaluation steps and time are required
Solution Approach 1:
The patent implements preliminary performance criteria checks at each activation level before proceeding to the next level. The network device evaluates whether performance criteria are met at level 1 before activating level 2, and so on. This preliminary action approach ensures performance stability by catching issues early while minimizing time loss through structured, incremental evaluation rather than comprehensive testing only at full deployment.
Data Source
AI summary
A disclosed aspect concerns a network device (101) comprising at least one processor (510), at least one memory (520) including computer program code (570), the at least one memory and computer program code configured to, with the at least one processor, cause the network device at least to perform: obtaining (201) information descriptive of a capability of a network entity (103) of another network device (102), said network entity comprising a capability for providing an output based on an artificial intelligence and/or machine learning inference function related to one or more network objects or object types; sending (203, 404), to the other network device, a request for a partial activation of the capability. Other aspects concern the other network device, methods respectively associated with each network devices and computer-readable media storing appropriate computer program code.