Charging Policy Management for Split AI Task Offloading

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing communication networks face challenges in managing charging functions for user equipment engaging in work task offloading, particularly in split artificial intelligence/machine learning (AIML) model processing scenarios, where computation resources are shared across devices.

Innovation Solution

The proposed solution involves techniques for managing charging functionalities in a communication network by defining charging policies that incentivize user equipment to share computation resources. These policies include credits for computation work offloaded, based on factors like the number of layers computed, usage frequency, performance KPIs, and priority of requests.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If user equipment shares computation resources through work task offloading, then network performance and resource utilization are improved, but charging function management becomes more complex

Engineering Contradiction:

Improvenetwork performanceVSAvoidcharging function management

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The charging management system is segmented into multiple specialized network functions including CHF (Charging Function), CTF (Charging Trigger Function), and ABMF (Account and Balance Management Function). Each function handles specific aspects of charging management for offloaded tasks, dividing the complex management challenge into manageable components that can operate independently yet cooperatively.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediary network functions that mediate between user equipment engaging in work task offloading and the charging system. These intermediaries translate complex offloading scenarios into standardized charging events, simplifying the management burden while maintaining comprehensive charging capabilities for shared computation resources.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If charging policies are designed to incentivize computation resource sharing, then user participation in work task offloading increases, but policy management complexity increases

Engineering Contradiction:

Improveuser participationVSAvoidpolicy management

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

Charging policies are designed to be dynamic and adaptable rather than static. The system can adjust charging parameters in real-time based on network conditions, user behavior patterns, and resource utilization metrics. This dynamic approach incentivizes user participation while keeping policy management tractable through automated adjustment mechanisms.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The charging management system incorporates feedback loops that monitor user participation in work task offloading and automatically adjust charging policies accordingly. This feedback mechanism enables the system to respond to user behavior changes while maintaining manageable policy complexity through rule-based automatic adjustments rather than manual policy rewriting.

Inventive Principle:

Principle #23Feedback

3Reliability

If comprehensive charging data is collected for fair compensation, then user trust and consent increase, but data collection and processing overhead increases

Engineering Contradiction:

Improveuser trustVSAvoiddata processing overhead

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent extracts and isolates specific charging-relevant data from the broader data generated during work task offloading. Rather than collecting and processing all possible data, the system selectively extracts only the information necessary for fair compensation calculation, reducing processing overhead while maintaining user trust through transparent and targeted data usage.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Different levels of data collection and processing are applied based on local requirements. The system collects comprehensive data where needed for fair compensation while using simplified data processing in scenarios where basic charging information suffices. This localized approach to data quality ensures user trust is maintained where necessary while minimizing overall processing overhead.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20250150294A1Managing network functions associated with user equipment task offloading

Publication Date: 2025.05.08 NOKIA TECHNOLOGIES OY

AI summary

Techniques are disclosed for managing one or more network functions associated with user equipment data exchange functionalities. While not necessarily limited thereto, disclosed techniques are well suited for implementation for managing charging functions associated with work task offloading for user equipment engaging in split artificial intelligence/machine learning (AIML) model processing in a communication network environment.