US-20260129496-A1 - COMMUNICATION METHOD AND APPARATUS

Abstract

This application provides a communication method and apparatus. The method may include: A first communication apparatus receives configuration information of a first neural network; the first communication apparatus determines the first neural network based on the configuration information of the first neural network; the first communication apparatus obtains first channel information based on channel information obtained through measurement and the first neural network, where a data amount of the first channel information is less than a data amount of the channel information obtained through measurement; and the first communication apparatus sends the first channel information, where the first channel information is used to obtain second channel information by using a second neural network, and the second channel information is used for data transmission.

Inventors

• Jian Wang
• Chen Xu
• Tianhang YU
• Rong Li
• Jun Wang

Assignees

• HUAWEI TECHNOLOGIES CO., LTD.

Dates

Publication Date

20260507

Application Date

20251219

Priority Date

20200713

Claims (20)

1. 1 .- 20 . (canceled)
2. 21 . A method, comprising: receiving, by a first communication apparatus, a first index value, wherein the first index value is used to determine a configuration parameter of a first neural network, and the first index value corresponds to a group of configuration parameters in a configuration parameter set of the first neural network; determining, by the first communication apparatus, the first neural network based on the first index value; obtaining, by the first communication apparatus, first channel information based on channel information obtained through measurement and the first neural network, wherein a data amount of the first channel information is less than a data amount of the channel information obtained through measurement; and sending, by the first communication apparatus, the first channel information, wherein the first channel information is used to obtain second channel information by using a second neural network, and the second channel information is used for data transmission.
3. 22 . The method according to claim 21 , wherein the configuration parameter of the first neural network comprises a structure of the first neural network and a parameter in the first neural network; the structure of the first neural network comprises one or more of: a type of the first neural network, a layer quantity of the first neural network, a node quantity of the first neural network, or a node connection manner of the first neural network; and the parameter in the first neural network comprises one or more of: a weight matrix, a weight vector, a bias matrix, a bias vector, or an activation function.
4. 23 . The method according to claim 21 , wherein the method further comprises: sending, by the first communication apparatus, first capability information, wherein the first capability information indicates a processing capability of the first communication apparatus.
5. 24 . The method according to claim 21 , wherein the receiving, by the first communication apparatus, the first index value comprises: receiving, by the first communication apparatus, first indication information, wherein the first indication information indicates to feed back the first channel information, and the first indication information further comprises the first index value.
6. 25 . The method according to claim 21 , wherein before the sending, by the first communication apparatus, the first channel information, the method further comprises: sending, by the first communication apparatus, a first request message, wherein the first request message is used to request a first time-frequency resource, the first time-frequency resource is used to transmit the first channel information, and the first request message further indicates the data amount of the first channel information; and receiving, by the first communication apparatus, second indication information, wherein the second indication information indicates configuration information of the first time-frequency resource.
7. 26 . The method according to claim 21 , wherein before the receiving, by the first communication apparatus, the first index value, the method further comprises: sending, by the first communication apparatus, a first request message, wherein the first request message is used to request a first time-frequency resource, and the first time-frequency resource is used to transmit the first channel information; and the receiving, by the first communication apparatus, the first index value comprises: receiving, by the first communication apparatus, second indication information, wherein the second indication information indicates configuration information of the first time-frequency resource, and the second indication information further comprises the first index value.
8. 27 . A method, comprising: sending, by a second communication apparatus, a first index value, wherein the first index value is used to determine a configuration parameter of a first neural network, and the first index value corresponds to a group of configuration parameters in a configuration parameter set of the first neural network; receiving, by the second communication apparatus, first channel information that is fed back, wherein the first channel information is obtained based on channel information obtained through measurement and the first neural network, and a data amount of the first channel information is less than a data amount of the channel information obtained through measurement; and performing, by the second communication apparatus, data transmission based on second channel information, wherein the second channel information is obtained based on the first channel information by using a second neural network.
9. 28 . The method according to claim 27 , wherein the configuration parameter of the first neural network comprises a structure of the first neural network and a parameter in the first neural network; the structure of the first neural network comprises one or more of: a type of the first neural network, a layer quantity of the first neural network, a node quantity of the first neural network, or a node connection manner of the first neural network; and the parameter in the first neural network comprises one or more of: a weight matrix, a weight vector, a bias matrix, a bias vector, or an activation function.
10. 29 . The method according to claim 27 , wherein before the sending, by the second communication apparatus, the first index value, the method further comprises: determining, by the second communication apparatus, the first index value and the second neural network.
11. 30 . The method according to claim 29 , wherein before the determining, by the second communication apparatus, the first index value and the second neural network, the method further comprises: receiving, by the second communication apparatus, first capability information, wherein the first capability information indicates a processing capability of a first communication apparatus.
12. 31 . The method according to claim 27 , wherein the sending, by the second communication apparatus, the first index value comprises: sending, by the second communication apparatus, first indication information, wherein the first indication information indicates to feed back the first channel information, and the first indication information further comprises the first index value.
13. 32 . The method according to claim 27 , wherein before the receiving, by the second communication apparatus, the first channel information that is fed back, the method further comprises: receiving, by the second communication apparatus, a first request message, wherein the first request message is used to request a first time-frequency resource, the first time-frequency resource is used to transmit the first channel information, and the first request message further indicates the data amount of the first channel information; and sending, by the second communication apparatus, second indication information, wherein the second indication information indicates configuration information of the first time-frequency resource.
14. 33 . The method according to claim 29 , wherein before the sending, by the second communication apparatus, the first index value, the method further comprises: receiving, by the second communication apparatus, a first request message, wherein the first request message is used to request a first time-frequency resource, and the first time-frequency resource is used to transmit the first channel information; and the sending, by the second communication apparatus, the first index value comprises: sending, by the second communication apparatus, second indication information, wherein the second indication information indicates configuration information of the first time-frequency resource, and the second indication information further comprises the first index value.
15. 34 . The method according to claim 27 , wherein the sending, by the second communication apparatus, the first index value comprises: periodically sending, by the second communication apparatus, the first index value; or sending, by the second communication apparatus, the first index value when determining that decoding performance of first data is lower than a preset threshold, wherein the first data is sent by the second communication apparatus based on the second channel information.
16. 35 . A communication apparatus, comprising: at least one processor, wherein the at least one processor is configured to read an instruction in a memory to enable the communication apparatus to perform: receiving a first index value, wherein the first index value is used to determine a configuration parameter of a first neural network, and the first index value corresponds to a group of configuration parameters in a configuration parameter set of the first neural network; determining the first neural network based on the first index value; obtaining first channel information based on channel information obtained through measurement and the first neural network, wherein a data amount of the first channel information is less than a data amount of the channel information obtained through measurement; and sending the first channel information, wherein the first channel information is used to obtain second channel information by using a second neural network, and the second channel information is used for data transmission.
17. 36 . The communication apparatus according to claim 35 , wherein the configuration parameter of the first neural network comprises a structure of the first neural network and a parameter in the first neural network; the structure of the first neural network comprises one or more of: a type of the first neural network, a layer quantity of the first neural network, a node quantity of the first neural network, or a node connection manner of the first neural network; and the parameter in the first neural network comprises one or more of: a weight matrix, a weight vector, a bias matrix, a bias vector, or an activation function.
18. 37 . The communication apparatus according to claim 35 , wherein the communication apparatus is further enabled to perform: sending first capability information, wherein the first capability information indicates a processing capability of the communication apparatus.
19. 38 . The communication apparatus according to claim 35 , wherein the receiving the first index value comprises: receiving first indication information, wherein the first indication information indicates to feed back the first channel information, and the first indication information further comprises the first index value.
20. 39 . The communication apparatus according to claim 35 , wherein before the sending the first channel information, the communication apparatus is further enabled to perform: send a first request message, wherein the first request message is used to request a first time-frequency resource, the first time-frequency resource is used to transmit the first channel information, and the first request message further indicates the data amount of the first channel information; and receive second indication information, wherein the second indication information indicates configuration information of the first time-frequency resource.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application is a continuation of U.S. patent application Ser. No. 18/153,606, filed on Jan. 12, 2023, which is a continuation of International Patent Application No. PCT/CN2021/100486, filed on Jun. 17, 2021, which claims priority to Chinese Patent Application No. 202010667227.7, filed on Jul. 13, 2020. All of the afore-mentioned patent applications are hereby incorporated by reference in their entireties. TECHNICAL FIELD This application relates to the communication field, and more specifically, to a communication method and apparatus. BACKGROUND In a massive multiple-input multiple-output (MIMO) technology, a network device may reduce interference between a plurality of terminal devices and interference between a plurality of signal streams of a same terminal device by using a precoding technology. Therefore, signal quality is improved, spatial multiplexing is implemented, and spectrum utilization is improved. For example, a terminal device may determine, in a manner such as channel measurement, a precoding matrix that adapts to a downlink channel, and expects that, through feedback, the network device obtains a precoding matrix that is the same as or similar to a precoding vector determined by the terminal device. Alternatively, the network device may determine, for example, in a manner such as channel measurement, a precoding matrix that adapts to an uplink channel, and expects that, through feedback, the terminal device obtains a precoding matrix that is the same as or similar to a precoding vector determined by the network device. Because there is a large amount of data of channel information obtained through channel measurement, if the channel information obtained through channel measurement is directly fed back, large feedback overheads are caused. Therefore, to reduce feedback, channel information is usually quantized according to a fixed codebook and then fed back. However, the fixed codebook may not match a real channel. As a result, a system throughput is not optimal. SUMMARY This application provides a communication method, to implement channel information feedback and improve a system throughput. According to a first aspect, a communication method is provided. The method may include: A first communication apparatus receives configuration information of a first neural network; the first communication apparatus determines the first neural network based on the configuration information of the first neural network; the first communication apparatus obtains first channel information based on channel information obtained through measurement and the first neural network, where a data amount of the first channel information is less than a data amount of the channel information obtained through measurement; and the first communication apparatus sends the first channel information, where the first channel information is used to obtain second channel information by using a second neural network, and the second channel information is used for data transmission. Based on the foregoing technical solution, a second communication apparatus sends the configuration information of the first neural network to the first communication apparatus, so that the second neural network used by the second communication apparatus matches the first neural network used by the first communication apparatus, and the second communication apparatus can further restore, based on the second neural network, channel information compressed by the first neural network. In this process, the first communication apparatus compresses, by using the first neural network, the channel information obtained through measurement. Therefore, the first communication apparatus may feed back the compressed channel information (the first channel information) with low overheads. Correspondingly, because the first communication apparatus feeds back the channel information obtained through measurement, the channel information (the second channel information) restored by the second communication apparatus better matches a real channel. Therefore, in a process in which the second communication apparatus performs data transmission based on the second channel information, a system throughput can be improved. With reference to the first aspect, in some implementations of the first aspect, the configuration information of the first neural network is a first index value, the first index value is used to determine a configuration parameter of the first neural network, and the first index value corresponds to a group of configuration parameters in a configuration parameter set of the neural network. With reference to the first aspect, in some implementations of the first aspect, the configuration information of the first neural network includes a configuration parameter of the first neural network. With reference to the first aspect, in some implementations of the first aspect, the configuration parameter of the first n