Offline Terminal AI Triggering for Timely Network Response

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In current distributed AI architectures, terminal devices switched to an offline state due to low priority or power conservation needs cannot timely respond to AI operations, affecting network performance optimization.

Innovation Solution

A network device sends a trigger message with scheduling information to terminal devices in an offline state, enabling them to perform AI operations by matching group identifiers and time-frequency resources, thus optimizing network AI capability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If terminal devices are switched to offline state to reduce power consumption, then power consumption is reduced, but response time to AI operations deteriorates

Engineering Contradiction:
Improvepower consumptionVSAvoidresponse time
Core Design Contradiction:
Use of energy by moving objectVSLoss of time

Solution Approach 1:

The terminal device performs AI operations in advance when in online state, completing model training and data processing before switching to offline state. This preliminary action ensures that AI tasks are already executed or prepared, eliminating the need for immediate response during offline state and thus resolving the contradiction between power saving and response time

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The terminal device alternates between online and offline states in periodic cycles. During online periods, it performs AI operations and communicates with the network; during offline periods, it conserves power. This periodic pattern allows the device to meet AI operation requirements during active periods while achieving power savings during inactive periods

Inventive Principle:
Principle #19Periodic action

2Use of energy by moving object

If terminal devices are switched to offline state to reduce power consumption, then power consumption is reduced, but network performance optimization deteriorates

Engineering Contradiction:
Improvepower consumptionVSAvoidnetwork performance
Core Design Contradiction:
Use of energy by moving objectVSProductivity

Solution Approach 1:

AI operations such as model training, data processing, and optimization tasks are executed in advance when the terminal device is online and has full computing resources. This preliminary execution ensures that network performance optimization is achieved before the device goes offline, preventing any negative impact on overall network productivity

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The terminal device autonomously determines when to perform AI operations based on its own state and network requirements. It self-manages the timing of model training and data processing, ensuring that optimization tasks are completed at appropriate moments without requiring continuous network connection, thus maintaining network performance while enabling power saving

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP4336935B1Ai operation processing method and related apparatus
Publication Date: 2025.12.31 HUAWEI TECH CO LTD
  • EP4336935B1 patent drawingFigure 1a~1c
  • EP4336935B1 patent drawingFigure 1d~2
  • EP4336935B1 patent drawingFigure 3

AI summary

This application discloses an AI operation processing method and a related apparatus. After a terminal device has been switched to an offline state, a network device can trigger, in a timely manner, the terminal device in the offline state to respond to and perform an AI operation, thereby participating in optimization of a network AI capability and improving network performance. The processing method may include: obtaining a first group identifier; obtaining a trigger message sent by a network device, where the trigger message includes first control information, the first control information indicates a first time-frequency resource, the first time-frequency resource carries N pieces of first scheduling information, each of the N pieces of first scheduling information includes a respective second group identifier and AI operation information, and a terminal device in a user group indicated by the second group identifier needs to perform an AI operation corresponding to the AI operation information; determining target AI operation information when the first group identifier matches a target group identifier; and starting a target AI operation based on the target AI operation information.