EP-4195600-B1 - METHOD AND APPARATUS FOR REPORTING AI NETWORK MODEL SUPPORT CAPABILITY, METHOD AND APPARATUS FOR RECEIVING AI NETWORK MODEL SUPPORT CAPABILITY, AND STORAGE MEDIUM, USER EQUIPMENT AND BASE STATION

Inventors

• LEI, Zhenzhu

Dates

Publication Date

20260506

Application Date

20210804

Claims (5)

1. A method for reporting Artificial Intelligence, AI, network model support capability, comprising: determining (S101) capability of supporting an AI network model, wherein the capability of supporting the AI network model comprises whether to support using the AI network model for channel estimation; and reporting (S102) the capability of supporting the AI network model using an uplink resource in a random access procedure; characterized in that said reporting (S102) the capability of supporting the AI network model using an uplink resource in a random access procedure comprises: reporting the capability of supporting the AI network model using a time-frequency resource for transmitting a preamble; wherein said reporting the capability of supporting the AI network model using the time-frequency resource for transmitting the preamble comprises: determining subsets of Physical Random Access Channel Occasions, ROs, and types of the subsets, wherein the types of the subsets comprise being used for initiating random access by a User Equipment, UE, that supports the AI network model and being used for initiating random access by a UE that does not support the AI network model; and determining a to-be-used subset of the ROs based on the capability of supporting the AI network model and the types of the subsets, and initiating random access using any RO in the to-be-used subset of the ROs; or wherein said reporting the capability of supporting the AI network model using the time-frequency resource for transmitting the preamble comprises: determining subsets of preambles and types of the subsets, wherein the types of the subsets comprise being used for initiating random access by a UE that supports the AI network model and being used for initiating random access by a UE that does not support the AI network model; and determining a to-be-used subset of the preambles based on the capability of supporting the AI network model and the types of the subsets, and reporting the capability of supporting the AI network model using a preamble in the to-be-used subset of the preambles.
2. The method according to claim 1, characterized in that prior to determining the subsets of ROs and the types of the subsets, the method further comprises: receiving the subsets of ROs and the types of the subsets which are configured by a base station.
3. The method according to claim 1, characterized in that following reporting (S102) the capability of supporting the AI network model using the uplink resource in the random access procedure, the method further comprises: based on that the capability of supporting the AI network model indicates supporting using the AI network model for channel estimation, receiving an AI model size reporting trigger instruction from a base station, wherein the support capability reporting trigger instruction indicates to report an input size of all the AI network model; and reporting the input size of all the AI network model using a Physical Downlink Shared Channel, PDSCH, scheduled by a Physical Downlink Control Channel, PDCCH, in response to the AI model size reporting trigger instruction.
4. A storage medium having computer instructions stored therein, characterized in that when the computer instructions are executed, the method of any one of claims 1 to 3 is performed.
5. A User Equipment, UE, comprising a memory and a processor, characterized in that the memory has computer instructions stored therein, and when the processor executes the computer instructions, the method of any one of claims 1 to 3 is performed.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to Chinese Patent Application No. 202010780069.6 filed on August 5, 2020, and entitled "METHOD AND APPARATUS FOR REPORTING AI NETWORK MODEL SUPPORT CAPABILITY, METHOD AND APPARATUS FOR RECEIVING AI NETWORK MODEL SUPPORT CAPABILITY, STORAGE MEDIUM, USER EQUIPMENT AND BASE STATION". TECHNICAL FIELD The present disclosure generally relates to communication technology field, and more particularly, to a method and apparatus for reporting Artificial Intelligence (AI) network model support capability, a method and apparatus for receiving AI network model support capability, a storage medium, a User Equipment (UE) and a base station. BACKGROUND An AI algorithm may be applied in channel estimation, where a process of estimating all channel values from a pilot is equated to a traditional image restoration/denoising process, and a deep learning algorithm for image restoration/denoising is adopted to complete the channel estimation. Currently, channel estimation based on AI network models is done at a UE. The UE can learn performance of each AI network model configured and a size of input and output. WO 2019/134563 A1 discloses determining a location measurement capability supported by a terminal device. S301: A positioning service function entity sends a capability request of UE to the UE. S302: The UE feeds back the capability supported by the UE to the positioning service function entity, where the capability indicates that the UE supports channel estimation and/or channel impulse response-based measurement. The positioning service function entity establishes a model of a correspondence between the measurement data and the location of the UE, and the model may be trained once in advance or periodically. WO 2020/032773 A1 discloses a method for performing channel estimation in wireless communication system and apparatus therefore. Ericsson, "Higher-layer aspects for Redcap", 3rd Generation Partnership Project (3GPP) Draft R1-2003292, Online Meeting, June 2020 discloses higher-layer aspects for Redcap. WO 2021/126907 A1 discloses a neural network configuration for wireless communication system assistance. However, how a base station learns relevant parameters of the AI network models at the UE is an urgent technical problem to be solved. SUMMARY The invention is set out in the appended set of claims. Embodiments of the present disclosure enable a base station to learn relevant parameters of an AI network model at the UE. A method for reporting AI network model support capability is provided, including: determining capability of supporting an AI network model, wherein the capability of supporting the AI network model includes whether to support using the AI network model for channel estimation; and reporting the capability of supporting the AI network model using an uplink resource in a random access procedure. Said reporting the capability of supporting the AI network model using an uplink resource in a random access procedure includes: reporting the capability of supporting the AI network model using a time-frequency resource for transmitting a preamble. Said reporting the capability of supporting the AI network model using a time-frequency resource for transmitting a preamble includes: determining subsets of Physical Random Access Channel Occasions (ROs) and types of the subsets, wherein the types of the subsets include being used for initiating random access by a UE that supports the AI network model and being used for initiating random access by a UE that does not support the AI network model; and determining a to-be-used subset of the ROs based on the capability of supporting the AI network model and the types of the subsets, and initiating random access using any RO in the to-be-used subset of the ROs. Optionally, prior to determining the subsets of ROs and the types of the subsets, the method further includes: receiving the subsets of ROs and the types of the subsets which are configured by a base station. Said reporting the capability of supporting the AI network model using a time-frequency resource for transmitting a preamble includes: determining subsets of preambles and types of the subsets, wherein the types of the subsets include being used for initiating random access by a UE that supports the AI network model and being used for initiating random access by a UE that does not support the AI network model; and determining a to-be-used subset of the preambles based on the capability of supporting the AI network model and the types of the subsets, and reporting the capability of supporting the AI network model using a preamble in the to-be-used subset of the preambles. Optionally, following reporting the capability of supporting the AI network model using the uplink resource in the random access procedure, the method further includes: based on that the capability of supporting the AI network model indicates supporting using the AI netw