Granted Patent B2

US 12,732,433 · App. 18/160,791 · Granted Sep 8, 2026

Inventors: Yourui Huangfu (Hangzhou, CN); Jian Wang (Hangzhou, CN); Rong Li (Hangzhou, CN); Jun Wang (Hangzhou, CN)

Assignee: Huawei Technologies Co., Ltd.

H04L41/16H04W28/08H04W36/22

Abstract

Claims (73)

1 . A method, comprising:

obtaining, by a first communication apparatus, a first artificial intelligence (AI) model corresponding to a first task;

obtaining, by the first communication apparatus, first data corresponding to a first feature, wherein the first feature is a feature of data processed by using the first AI model, and the first data is used for drawing inferences for decision-making of the first task; and

determining, by the first communication apparatus, an inference result for the decision-making of the first task based on the first data and the first AI model,

wherein the first data includes at least one of: a channel measurement report, a user behavior profile, network performance data, or resource utilization data, and

wherein the inference result comprises a probability of each of a plurality of possible decision results for the first task, wherein the first communication apparatus sends a third request to a second communication apparatus requesting the first data, the third request comprising indication information of the first feature, and wherein the first communication apparatus is a network data analytics function (NWDAF) and the second communication apparatus is one of an access network device or a terminal device.

2 . The method according to claim 1 , wherein the first task is a task for a second communication apparatus, and the method further comprises:

sending, by the first communication apparatus, the inference result to the second communication apparatus.

3 . The method according to claim 2 , wherein the method further comprises:

receiving, by the first communication apparatus, a first request sent by the second communication apparatus, wherein the first request is used to request the inference result for the decision-making of the first task, and the first request comprises an identifier of the first task.

4 . The method according to claim 1 , wherein the obtaining, by the first communication apparatus, the first AI model corresponding to the first task comprises:

determining, by the first communication apparatus based on a first mapping relationship, the first AI model corresponding to an identifier of the first task, wherein the first mapping relationship is used to indicate a correspondence between a plurality of identifiers and a plurality of AI models.

5 . The method according to claim 1 , wherein the obtaining, by the first communication apparatus, the first AI model corresponding to the first task comprises:

sending, by the first communication apparatus, a second request to a third communication apparatus, wherein the second request comprises an identifier of the first task, and the second request is used to request to obtain the first AI model corresponding to the identifier of the first task; and

receiving, by the first communication apparatus, the first AI model sent by the third communication apparatus.

6 . The method according to claim 5 , wherein the first task is a task for a second communication apparatus, and the obtaining, by the first communication apparatus, first data corresponding to a first feature comprises:

sending, by the first communication apparatus, a third request to the second communication apparatus, wherein the third request comprises indication information of the first feature, and the second request is used to request to obtain data corresponding to the first feature; and

receiving, by the first communication apparatus, the first data sent by the second communication apparatus.

7 . The method according to claim 1 , wherein the method further comprises:

sending, by the first communication apparatus, a correspondence between an identifier of the first task, the first data, and the inference result to a fourth communication apparatus, wherein the fourth communication apparatus is configured to store training data of the first AI model.

8 . A method, comprising:

sending, by a second communication apparatus, a first request to a first communication apparatus, wherein the first request is used to request an inference result for decision-making of a first task, and the first request comprises an identifier of the first task;

receiving, by the second communication apparatus, a third request sent by the first communication apparatus, wherein the third request comprises indication information of a first feature, and the third request is used to request to obtain data corresponding to the first feature;

sending, by the second communication apparatus, first data corresponding to the first feature to the first communication apparatus; and

receiving, by the second communication apparatus, the inference result sent by the first communication apparatus, wherein the inference result is determined based on the first data and a first artificial intelligence (AI) model corresponding to the first task, and the first feature is a feature of data processed by using the first AI model,

wherein the first data includes at least one of: a channel measurement report, a user behavior profile, network performance data, or resource utilization data, and

wherein the inference result comprises a probability of each of a plurality of possible decision results for the first task, wherein the first communication apparatus sends a third request to a second communication apparatus requesting the first data, the third request comprising indication information of the first feature, and wherein the first communication apparatus is a network data analytics function (NWDAF) and the second communication apparatus is one of an access network device or a terminal device.

9 . The method according to claim 8 , wherein the method further comprises:

sending, by the first communication apparatus, a correspondence between the identifier of the first task, the first data, and the inference result to a fourth communication apparatus, wherein the fourth communication apparatus is configured to store training data of the first AI model.

10 . The method according to claim 8 , wherein the decision-making of the first task comprises a plurality of decision results, and the inference result comprises a probability of each of the plurality of decision results.

11 . The method according to claim 8 , wherein the first task comprises mobility enhancement optimization of a terminal device, and

the first data comprises at least one of the following data:

a first measurement report of the terminal device, a first user behavior profile of the terminal device, and resource utilization of the terminal device; and

the inference result for the decision-making of the first task comprises a size of a reserved resource of the terminal device.

12 . The method according to claim 8 , wherein the first task comprises access mode optimization of a terminal device, and

the first data comprises at least one of the following data:

a second measurement report of the terminal device and a random access report of the terminal device; and

the inference result for the decision-making of the first task comprises a quantity of times of two-step random access attempts of the terminal device.

13 . The method according to claim 8 , wherein the first task comprises mobility robustness optimization of a terminal device, and

the first data comprises at least one of the following information:

a third measurement report of the terminal device and a handover report of the terminal device; and

the inference result for the decision-making of the first task comprises a handover policy of the terminal device, or

the inference result for decision-making of the first task comprises a handover effect corresponding to a handover policy of the terminal device.

14 . The method according to claim 8 , wherein the first task comprises handover optimization of a terminal device, and

the first data comprises at least one of load of a plurality of cells, power consumption of the plurality of cells, second user behavior profiles of terminal devices in the plurality of cells, an average of the power consumption of the plurality of cells, or an average load of the plurality of cells; and

the inference result for the decision-making of the first task comprises a handover policy of one or more terminals in the plurality of cells.

15 . The method according to claim 8 , wherein the first task comprises radio access network notification area optimization, and

the first data comprises at least one of the following data:

a second user behavior profile of a terminal device accessing a radio access network, a location of a first cell of the radio access network, and signaling overheads of a notification area of the radio access network; and

the inference result for the decision-making of the first task comprises a probability of providing the first cell as a notification area for the terminal device and not providing the first cell as a notification area for the terminal device.

16 . The method according to claim 8 , wherein the first task comprises radio resource management policy optimization of a terminal device, and

the first data comprises at least one of the following data:

a third user behavior profile of the terminal device, a quantity of active users of a cell in which the terminal device is located, and resource utilization of the terminal device; and

the inference result for the decision-making of the first task comprises a radio resource management policy.

17 . The method according to claim 8 , wherein the first task comprises optimization of matching between an application layer of a terminal device and a radio access network, and

the first data comprises at least one of the following data:

a fourth user behavior profile of the terminal device, a rate requirement of a first application of the terminal device, and quality of service of the terminal device; and

the inference result for the decision-making of the first task comprises an adjustable rate of the first application.

18 . The method according to claim 8 , wherein the first task comprises mobility load balancing optimization of a cell, and

the first data comprises load of the cell and a fifth user behavior profile of a terminal device in the cell; and

the inference result for the decision-making of the first task comprises measurement of mobility processing of one or more terminal devices in the cell, and the mobility processing comprises at least one of cell reselection, cell handover, or cell selection.

19 . The method according to claim 8 , wherein the first task comprises coverage optimization of a cell, and

the first data comprises a minimization of drive tests MDT report of the cell and coverage of the cell; and

the inference result for the decision-making of the first task comprises a coverage adjustment policy of the cell.

20 . A communication apparatus, comprising:

at least one processor;

a transceiver; and

one or more memories coupled to the at least one processor and storing programming instructions for execution by the at least one processor to cause the communication apparatus to:

determine a first artificial intelligence (AI) model corresponding to a first task;

receive, through the transceiver, first data corresponding to a first feature, wherein the first feature is a feature of data processed by using the first AI model, and the first data is used for drawing inferences for decision-making of the first task; and

determine an inference result for the decision-making of the first task based on the first data and the first AI model,

wherein the first data includes at least one of: a channel measurement report, a user behavior profile, network performance data, or resource utilization data, and

wherein the inference result comprises a probability of each of a plurality of possible decision results for the first task, wherein the communication apparatus sends a third request to another communication apparatus requesting the first data, the third request comprising indication information of the first feature, and wherein the communication apparatus is a network data analytics function (NWDAF) and the another communication apparatus is one of an access network device or a terminal device.

Assignments (1)

Priority Claims (1)

Continuity (2)

References Cited (16)