AI/ML Inference Function Management via Operations Service

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current AI/ML inference management in wireless communication networks lacks explicit visibility and efficient management capabilities for AI/ML inference functions, particularly in 5G systems, leading to suboptimal performance and operational challenges.

Innovation Solution

The implementation of Information Object Classes (IOCs) managed via an operations and notifications service, similar to those defined in 3GPP TS 28.532, provides explicit views and management capabilities for AI/ML inference functions, allowing consumers to create, modify, and manage AI/ML inference functions and related ML entities through generic provisioning management services.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If AI/ML inference functions are deployed in 5G networks without explicit management capabilities, then the network can support advanced intelligence applications, but the visibility and operational control of these functions remain insufficient

Engineering Contradiction:

ImproveAI/ML inference function supportVSAvoidManagement visibility

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The patent introduces an intermediary management layer that sits between the AI/ML inference functions and the network operations. This intermediary provides standardized interfaces and abstraction layers that enable operators to manage complex AI/ML functions through simplified, explicit control mechanisms without directly interfacing with the underlying complexity

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The management system is segmented into distinct functional components including separate entities for creating, configuring, monitoring, and controlling AI/ML inference functions. This segmentation allows each management aspect to be handled independently through dedicated interfaces and procedures, improving overall visibility and operational control

Inventive Principle:

Principle #1Segmentation

2Adaptability or versatility

If generic provisioning management services are used to manage AI/ML inference functions, then standardization and interoperability improve, but specific AI/ML management capabilities may be generalized

Engineering Contradiction:

ImproveStandardizationVSAvoidManagement precision

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The patent applies local quality by tailoring specific management capabilities and parameters to the unique requirements of AI/ML inference functions within the standardized framework. While using generic provisioning management services for overall structure and interoperability, the system incorporates AI/ML-specific attributes, metrics, and control mechanisms at the local level to maintain precise management of inference function performance, configuration, and operations

Inventive Principle:

Principle #3Local quality

Data Source

PatentEP4529114A1Artificial intelligence/machine learning (ai-ML) inference function management

Publication Date: 2025.03.26 INTEL CORP

Figure 1
Figure 2~3
Figure 4

AI summary

A computer-readable medium may include instructions which, if executed by a processor, may cause the processor to: receive a request from a management service, MnS, consumer to manage an artificial intelligence and/or machine learning, AI/ML, inference function; manage the AI/ML inference function; and respond to the MnS consumer to indicate the result of the management of the AI/ML inference function.