CN-122029854-A - Capability sending method, capability obtaining method and communication device

Abstract

A capability sending method, an acquisition method and a communication device, wherein the sending method comprises the steps of sending learning capability information of a second communication device to a first communication device, wherein the learning capability information comprises first information which indicates the learning capability information of the second communication device aiming at a target task, and the learning capability information of the second communication device aiming at the target task characterizes the capability which is distributed by the second communication device aiming at the target task and related to a machine learning task. The second communication device sends the learning capability of the task granularity to the first communication device, so that the first communication device obtains the learning capability of the second communication device aiming at the specific task, and the first communication device and the second communication device can be configured with a proper working mode related to learning.

Inventors

• WANG JIAN
• LI RONG

Assignees

• 华为技术有限公司

Dates

Publication Date

20260512

Application Date

20231011

Claims (20)

1. A capability transmission method, comprising: The method comprises the steps of sending learning capability information of a second communication device to a first communication device, wherein the learning capability information comprises first information which indicates the learning capability information of the second communication device aiming at a target task, and the learning capability information of the second communication device aiming at the target task characterizes the capability of the second communication device aiming at the target task and distributed by the learning capability information of the second communication device and related to a machine learning task.
2. The method of claim 1, wherein the learning capability information of the second communication device for the target task includes one or more of processor type information, computing capability information, storage capability information, acceleration capability information, compilation capability information, quantization capability information, energy consumption limitation information, support format information, and support task information.
3. The method according to claim 1 or 2, wherein the learning ability information of the second communication device for the target task further comprises: And indicating information of one or more AI model information corresponding to the target task.
4. The method of claim 3, wherein the learning capability information for the target task for the second communication device further includes indication information of context information applicable to each of the one or more AI models.
5. The method of any of claims 1-4, wherein transmitting learning capability information of a second communication device to a first communication device further comprises second information indicating common learning capability information of the second communication device.
6. The method of claim 5, wherein the common learning capability information of the second communication device includes one or more of processor type information, computing capability information, storage capability information, compiling capability information, quantization capability information, energy consumption limitation information, support format information, and support task information of the second communication device.
7. A method according to claim 5 or 6, wherein the first information comprises first sub-information for indicating whether each learning ability of the second communication device for a target task is determined based on a corresponding common learning ability in the common learning ability information.
8. The method according to claim 5 or 6, wherein the first information comprises one or more second sub-information, each of the one or more second sub-information being for indicating each learning ability of the second communication device for a target task; The second communication equipment comprises first learning ability aiming at each learning ability of a target task, second sub-information corresponding to the first learning ability is blank or second sub-information corresponding to the first learning ability does not exist, and the first learning ability is determined based on the corresponding public learning ability in the public learning ability information.
9. The method according to any one of claims 1-8, further comprising: And receiving a learning ability query request message sent by the first communication device, wherein the learning ability query request message comprises first query information, and the first query information is used for indicating and querying the learning ability information of the second communication device aiming at a target task.
10. The method of claim 9, wherein the first query information comprises a target task query list in which one or more target tasks to be queried are recorded.
11. The method according to claim 9 or 10, wherein the learning ability query request message further comprises second query information indicating a query for common learning ability information of the second communication device.
12. The method according to any of claims 1-11, wherein the learning ability information of the second communication device is transmitted by radio resource control signaling.
13. A method for capability acquisition, comprising: Receiving learning capability information of a second communication device from the second communication device, wherein the learning capability information comprises first information which indicates learning capability information of the second communication device for a target task, and the learning capability information of the second communication device for the target task characterizes the capability of the second communication device for the target task, which is distributed by the second communication device and relates to a machine learning task.
14. The method of claim 13, wherein the learning capability information of the second communication device for the target task includes one or more of processor type information, computing capability information, storage capability information, acceleration capability information, compilation capability information, quantization capability information, energy consumption limitation information, support format information, and support task information.
15. The method according to claim 13 or 14, wherein the learning ability information of the second communication device for the target task further comprises: And indicating information of one or more AI model information corresponding to the target task.
16. The method of claim 15, wherein the learning capability information for the target task for the second communication device further includes indication information of context information applicable to each of the one or more AI models.
17. The method according to any of claims 13-16, wherein the learning capability information of the second communication device further comprises second information indicating common learning capability information of the second communication device.
18. The method of claim 17, wherein the common learning capability information of the second communication device includes one or more of processor type information, computing capability information, storage capability information, compiling capability information, quantization capability information, energy consumption limitation information, support format information, and support task information of the user device.
19. A method according to claim 17 or 18, wherein the first information comprises first sub-information for indicating whether each learning ability of the second communication device for a target task is determined based on a corresponding common learning ability in the common learning ability information.
20. The method according to claim 17 or 18, wherein the first information comprises one or more second sub-information, each of the one or more second sub-information being for indicating each learning ability of the second communication device for a target task; The second communication equipment comprises first learning ability aiming at each learning ability of a target task, second sub-information corresponding to the first learning ability is blank or second sub-information corresponding to the first learning ability does not exist, and the first learning ability is determined based on the corresponding public learning ability in the public learning ability information.

Description

Capability sending method, capability obtaining method and communication device Technical Field The present application relates to the field of communications technologies, and in particular, to a capability sending method, a capability obtaining method, and a communications device. Background In order to cope with future general prospects of intelligence, the intelligentization will further evolve at the wireless network architecture level, the artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) will be fused with the wireless network further deeply, the intelligence of network endogenous is realized, and meanwhile, the terminal intelligentization is also included to cope with the following new requirements and new scenes. Terminal type is diversified, terminal connection is more flexible and intelligent, the terminal type is diversified, super internet of things (Supper internet of things, provider IOT) (internet of things, internet of vehicles, industry, medical treatment are the same), mass connection is achieved, terminal connection is more flexible, and the terminal has certain AI capability. Network endogenous intelligence-the network will provide computing and AI services in addition to traditional communication connection services to better support the generic Hui Xing, real-time and high-security AI services. These new requirements and new scenarios bring about changes in wireless network architecture and communication modes. The existing framework for querying and reporting User Equipment (UE) capabilities in the fifth generation mobile communication technology (the 5 th generation, 5G) New Radio (NR) communication technology only relates to communication-related capabilities, and does not consider learning-related capabilities. Disclosure of Invention The embodiment of the application provides a capability sending method, a capability obtaining method and a communication device, which realize learning capability obtaining with tasks as granularity, so that a network and a terminal can be configured with a proper working mode related to learning. In a first aspect, the present application provides a capability transmission method, which may be performed by an apparatus. The apparatus may be a device (such as a terminal device, and also a network device), or may be a component of a device (such as a chip or a system-on-a-chip or a circuit), which is not limited by the present application. The method includes sending learning capability information of a second communication device to a first communication device, the learning capability information including first information indicating learning capability information of the second communication device for a target task, the learning capability information of the second communication device for the target task characterizing a machine learning task-related capability assigned by the second communication device for the target task. The learning capability information of the second communication device for the target task is sent to the first communication device, so that the first communication device obtains the learning capability of the second communication device with the task as granularity, and the learning capability information of the second communication device supporting the specific task is obtained, so that the first communication device and the second communication device can be configured with a proper working mode related to learning. The capability sending method can be applied to a wireless communication system, the wireless communication system comprises a first communication device and a second communication device, the first communication device and the second communication device are in wireless communication connection, wherein the first communication device can be network devices (such as a base station or core network devices), the second communication device can be terminal devices (such as UE), namely, the terminal devices send learning capability information of the terminal devices to the network devices, the capability sending method can be also called a capability reporting method, the first communication device can be the terminal devices, the second communication device can be the network devices, namely, the network devices send learning capability information of the network devices to the terminal devices, so that the terminal devices acquire the learning capability information of the network devices for specific tasks, the terminal devices can select the network devices with higher learning capability through the specific tasks to access the network, and better service for acquiring the specific tasks is achieved. The method for transmitting the capability provided by the application is described below by taking the first communication device as a network device and the second communication device as a terminal device as an example, and the method can also be called as a capability reporting method. For example, a