Channel state information transmission method, electronic equipment and storage medium

Abstract

The invention discloses a channel state information transmission method. The method comprises the steps that terminal equipment determines channel state information of a channel based on a first neural network model; and the terminal equipment sends the channel state information. The invention further discloses another channel state information transmission method, electronic equipment and a storage medium.

Description

[0001] Description of the case

[0002] This application is a divisional application of a Chinese patent with an application date of May 25, 2020, application number 202080100930.8, and invention name “A channel state information transmission method, electronic device and storage medium”. Technical Field

[0003] The present application relates to the field of wireless communication technology, and in particular to a channel state information transmission method, electronic device and storage medium. Background Art

[0004] In the related art, there is no effective solution for how to perform accurate full channel information feedback in the Multiple-Input Multiple-Output (MIMO) scenario. Summary of the invention

[0005] The embodiments of the present application provide a channel state information transmission method, an electronic device, and a storage medium. For MIMO scenarios, a terminal device can feed back accurate full channel information to a network device.

[0006] In a first aspect, an embodiment of the present application provides a channel state information transmission method, including: a terminal device determines channel state information of a channel based on a first neural network model; and the terminal device sends the channel state information.

[0007] In a second aspect, an embodiment of the present application provides a channel state information transmission method, including: a network device receives channel state information; the network device determines channel information corresponding to the channel state information based on a second neural network model.

[0008] In a third aspect, an embodiment of the present application provides a terminal device, the terminal device comprising: a first processing unit, configured to determine channel state information of a channel based on a first neural network model;

[0009] A sending unit is configured to send the channel state information.

[0010] In a fourth aspect, an embodiment of the present application provides a network device, the network device comprising: a first processing unit, configured to determine channel state information of a channel based on a first neural network model;

[0011] A sending unit is configured to send the channel state information.

[0012] In a fifth aspect, an embodiment of the present application provides a terminal device, comprising a processor and a memory for storing a computer program that can be run on the processor, wherein the processor is used to execute the steps of the channel state information transmission method executed by the above-mentioned terminal device when running the computer program.

[0013] In a sixth aspect, an embodiment of the present application provides a network device, comprising a processor and a memory for storing a computer program that can be run on the processor, wherein the processor is used to execute the steps of the channel state information transmission method executed by the above-mentioned network device when running the computer program.

[0014] In a seventh aspect, an embodiment of the present application provides a chip, comprising: a processor, configured to call and run a computer program from a memory, so that a device equipped with the chip executes the channel state information transmission method executed by the above-mentioned terminal device.

[0015] In an eighth aspect, an embodiment of the present application provides a chip, comprising: a processor, configured to call and run a computer program from a memory, so that a device equipped with the chip executes the channel state information transmission method executed by the above-mentioned network device.

[0016] In a ninth aspect, an embodiment of the present application provides a storage medium storing an executable program, which, when executed by a processor, implements the channel state information transmission method executed by the above-mentioned terminal device.

[0017] In a tenth aspect, an embodiment of the present application provides a storage medium storing an executable program, which, when executed by a processor, implements the channel state information transmission method executed by the above-mentioned network device.

[0018] In an eleventh aspect, an embodiment of the present application provides a computer program product, comprising computer program instructions, which enable a computer to execute the channel state information transmission method executed by the above-mentioned terminal device.

[0019] In a twelfth aspect, an embodiment of the present application provides a computer program product, including computer program instructions, which enable a computer to execute the channel state information transmission method executed by the above-mentioned network device.

[0020] In a thirteenth aspect, an embodiment of the present application provides a computer program, which enables a computer to execute the channel state information transmission method executed by the above-mentioned terminal device.

[0021] In a fourteenth aspect, an embodiment of the present application provides a computer program, which enables a computer to execute the channel state information transmission method executed by the above-mentioned network device.

[0022] The channel state information transmission method, electronic device and storage medium provided in the embodiment of the present application include: a terminal device determines the channel state information of a channel based on a first neural network model; the terminal device sends the channel state information. In this way, by determining the channel state information through the first neural network model, the terminal device can feedback accurate full channel information to the network device. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a schematic diagram of the feedback process of the CSI indication for this application;

[0024] Figure 2 This is a basic structural diagram of a simple neural network model for this application;

[0025] Figure 3 This is a schematic diagram of the basic structure of the deep neural network model of this application;

[0026] Figure 4 This is a schematic diagram of the structure of the convolutional neural network model of this application;

[0027] Figure 5 A schematic diagram of using a neural network model for CSI feedback in this application;

[0028] Figure 6 This is a schematic diagram of the structure of the communication system according to the embodiment of the present application;

[0029] Figure 7 A schematic diagram of an optional processing flow of the channel state information transmission method according to an embodiment of the present application;

[0030] Figure 8 This is an optional schematic diagram of channel input in a scenario where the channel characteristics of an embodiment of the present application include frequency domain information dimension, transmission information dimension, and reception information dimension;

[0031] Fig. 9 Another optional schematic diagram of channel input in a scenario where the channel characteristics of an embodiment of the present application include a time domain information dimension, a transmission information dimension, and a reception information dimension;

[0032] Fig.10 This is another optional schematic diagram of channel input in a scenario where the channel characteristics of an embodiment of the present application include a time domain information dimension, a transmission angle information dimension, and a reception angle information dimension;

[0033] Fig.11 This is another optional schematic diagram of channel input in a scenario where the channel characteristics of an embodiment of the present application include a time domain information dimension, a transmission angle information dimension, and a reception angle information dimension;

[0034] Fig.12 A schematic diagram of local channel information of a channel according to an embodiment of the present application;

[0035] Fig.13 This is a schematic diagram of the basic structure of the feature extraction unit in the embodiment of the present application;

[0036] Fig.14 This is an optional structural diagram of a feature extraction unit in an embodiment of the present application;

[0037] Fig.15 This is a schematic diagram of the structure of a redundant elimination unit according to an embodiment of the present application;

[0038] Fig.16 This is a basic structural diagram of the first neural network model of the embodiment of the present application;

[0039] Fig.17 This is a schematic diagram of an optional structure of a first neural network model for a MIMO scenario in an embodiment of the present application;

[0040] Fig.18 A schematic diagram of another optional processing flow of the channel state information transmission method according to an embodiment of the present application;

[0041] Fig.19 A schematic diagram of the structure of the information enhancement unit in the embodiment of the present application;

[0042] Fig. 20 An optional schematic diagram of a channel input to a feature recovery unit according to an embodiment of the present application;

[0043] Fig.21 A basic structural diagram of a feature recovery unit in an embodiment of the present application;

[0044] Fig. 22 A detailed structural diagram of a feature recovery unit in an embodiment of the present application;

[0045] Fig.23 A schematic diagram of a second neural network model including a residual network unit according to an embodiment of the present application;

[0046] Fig.24 A schematic diagram of the basic structure of a second neural network model for a MIMO scenario in an embodiment of the present application;

[0047] Fig.25 A detailed structural diagram of a second neural network model for a MIMO scenario according to an embodiment of the present application;

[0048] Fig.26 This is a schematic diagram of an optional processing flow for transmitting channel state information based on a neural network model according to an embodiment of the present application;

[0049] Fig. 27 This is a schematic diagram of an optional component structure of a terminal device according to an embodiment of the present application;

[0050] Fig.28 This is a schematic diagram of an optional composition structure of a network device according to an embodiment of the present application;

[0051] Fig.29 Schematic diagram of the hardware structure of the electronic device according to the embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to enable a more detailed understanding of the features and technical contents of the embodiments of the present application, the implementation of the embodiments of the present application is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present application.

[0053] Before describing the embodiments of the present application in detail, channel state information (CSI) and artificial intelligence are briefly described.

[0054] In the long term evolution (LTE) system and the new radio (NR) system, the terminal device needs to report the CSI indication to the network device, and the CSI indication is very important in both the LTE system and the NR system. The CSI indication determines the performance of data transmission in the MIMO scenario. Generally, the CSI indication in the existing system can include the indication of the channel quality indicator (CQI), precoding matrix indicator (PMI), rank indicator (RI) and other information. The feedback process of the CSI indication, such as Figure 1 As shown, the network device pre-configures the indication parameter information for CSI indication, such as which information the network device configures the terminal device needs to indicate in the CSI indication, such as CQI, PMI, and RI. At the same time, the network device configures some reference signals for CSI measurement, such as a synchronization signal block (SSB) or a channel state information reference signal (Channel-State Information Reference Signal, CSI-RS). The terminal device determines the current channel state information by measuring the above reference signals, and indicates the current channel state information to the network device according to the indication parameter information configured by the network device. The current channel state information is used by the network device to configure a reasonable and efficient data transmission method based on the current channel situation.

[0055] Artificial intelligence has become a new way for people to solve and deal with problems. Among them, artificial intelligence based on neural networks has a wide range of applications. The basic structure diagram of a simple neural network model, such as Figure 2 As shown, it includes: an input layer, a hidden layer and an output layer; wherein the input layer is used to receive data, the hidden layer is used to process the data, and the output layer is used to generate the calculation results of the neural network model.

[0056] With the continuous development of research on neural network models, neural network deep learning algorithms have been proposed. The basic structural diagram of the deep neural network model is as follows: Figure 3 As shown, the deep neural network model includes multiple hidden layers. The deep neural network model including multiple hidden layers can greatly improve the data processing capability and is widely used in pattern recognition, signal processing, optimization combination, and anomaly detection.

[0057] With the development of neural network deep learning algorithms, convolutional neural network models have also been further studied. The structural diagram of the convolutional neural network model is shown in the figure below. Figure 4 As shown, it may include: an input layer, at least one convolutional layer, at least one pooling layer, a fully connected layer, and an output layer. The introduction of convolutional layers and pooling layers can effectively control the sharp increase in the number of network parameters, limit the number of network parameters, and tap the characteristics of local structures to improve the robustness of the convolutional neural network algorithm.

[0058] The basic principle of the current wireless communication system is based on theoretical modeling and parameter selection of the actual communication environment. With the further enhancement of the requirements for the flexibility, adaptability, rate and capacity of wireless communication systems, the gains that can be brought by the working mode of wireless communication systems based on classical model theory are gradually weakening. In order to further improve the performance of wireless communication systems, it is necessary to find a new way to adopt new wireless communication methods. Taking the acquisition and indication method of CSI as an example, the prior art determines and indicates CSI through a basic model and a pre-configured set of indication parameters. This method based on quantization and codebook not only cannot accurately reflect the current actual state information of the channel, but also the measurement error of CSI is not conducive to the accurate scheduling of channel resources in future communication systems; especially in the case of the introduction of large-scale MIMO, this will lead to the limitation of the upper limit of system throughput.

[0059] One feasible way is to use a neural network model for CSI feedback. One advantage of using a neural network model for CSI feedback is that the information fed back is complete channel information, rather than processed channel information, which helps the communication system to better utilize channel information and current channel status for data transmission.

[0060] A schematic diagram of using a neural network model for CSI feedback, such as Figure 5As shown, the terminal device generates first channel information through the encoding end neural network model from the original channel information, and the decoding end neural network model generates feedback channel information (i.e., CSI) through the decoding sub-model from the first channel information. With the widespread application of MIMO scenarios, there is no effective solution for how to perform CSI feedback based on the neural network model in MIMO scenarios.

[0061] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: global system of mobile communication (GSM) system, code division multiple access (CDMA) system, wideband code division multiple access (WCDMA) system, general packet radio service (GPRS), LTE system, LTE frequency division duplex (FDD) system, LTE time division duplex (TDD) system, advanced long term evolution (LTE-A) system, new radio (NR) system, NR system evolution system, LTE on unlicensed spectrum (LTE-U) system, NR on unlicensed spectrum (NR-U) system, universal mobile telecommunication system (UMTS), worldwide interoperability for microwave access (WiMAX) communication system, wireless local area network (WLAN) system. networks, WLAN), wireless fidelity (WiFi), next generation communication systems or other communication systems, etc.

[0062] The system architecture and business scenarios described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. A person of ordinary skill in the art can appreciate that with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0063] The network equipment involved in the embodiments of the present application may be an ordinary base station (such as NodeB or eNB or gNB), a new radio controller (new radio controller, NR controller), a centralized unit, a new wireless base station, a radio frequency remote module, a micro base station, a relay, a distributed unit, a transmission reception point (TRP), a transmission point (TP) or any other device. The embodiments of the present application do not limit the specific technology and specific device form adopted by the network equipment. For the convenience of description, in all embodiments of the present application, the above-mentioned devices that provide wireless communication functions for terminal devices are collectively referred to as network devices.

[0064] In the embodiment of the present application, the terminal device may be any terminal, for example, the terminal device may be a user device for machine type communication. That is to say, the terminal device may also be referred to as a user equipment UE, a mobile station (MS), a mobile terminal, a terminal, etc. The terminal device may communicate with one or more core networks via a radio access network (RAN). For example, the terminal device may be a mobile phone (or a "cellular" phone), a computer with a mobile terminal, etc. For example, the terminal device may also be a portable, pocket-sized, handheld, computer-built-in or vehicle-mounted mobile device, which exchanges language and / or data with a radio access network. This is not specifically limited in the embodiment of the present application.

[0065] Optionally, the network device and the terminal device can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; can also be deployed on the water surface; can also be deployed on aircraft, balloons and artificial satellites in the air. The embodiments of the present application do not limit the application scenarios of the network device and the terminal device.

[0066] Optionally, the network device and the terminal device, and the terminal device and the terminal device may communicate through a licensed spectrum, or may communicate through an unlicensed spectrum, or may communicate through both a licensed spectrum and an unlicensed spectrum. The network device and the terminal device, and the terminal device and the terminal device may communicate through a spectrum below 7 gigahertz (GHz), or may communicate through a spectrum above 7 GHz, or may communicate using a spectrum below 7 GHz and a spectrum above 7 GHz at the same time. The embodiments of the present application do not limit the spectrum resources used between the network device and the terminal device.

[0067] Generally speaking, traditional communication systems support a limited number of connections and are easy to implement. However, with the development of communication technology, mobile communication systems will not only support traditional communications, but will also support, for example, device to device (D2D) communication, machine to machine (M2M) communication, machine type communication (MTC), and vehicle to vehicle (V2V) communication, etc. The embodiments of the present application can also be applied to these communication systems.

[0068] For example, the communication system 100 used in the embodiment of the present application is as follows: Figure 6 As shown. The communication system 100 may include a network device 110, which may be a device that communicates with a terminal device 120 (or referred to as a communication terminal or terminal). The network device 110 may provide communication coverage for a specific geographical area, and may communicate with terminal devices located in the coverage area. Optionally, the network device 110 may be a base station (Base Transceiver Station, BTS) in a GSM system or a CDMA system, or a base station (NodeB, NB) in a WCDMA system, or an evolved base station (Evolutional Node B, eNB or eNodeB) in an LTE system, or a wireless controller in a cloud radio access network (Cloud Radio Access Network, CRAN), or the network device may be a mobile switching center, a relay station, an access point, a vehicle-mounted device, a wearable device, a hub, a switch, a bridge, a router, a network-side device in a 5G network, or a network device in a future evolved public land mobile network (Public Land Mobile Network, PLMN), etc.

[0069] The communication system 100 also includes at least one terminal device 120 located within the coverage of the network device 110. As used herein, "terminal device" includes but is not limited to connection via a wired line, such as via a Public Switched Telephone Networks (PSTN), a Digital Subscriber Line (DSL), a digital cable, a direct cable connection; and / or another data connection / network; and / or via a wireless interface, such as, for a cellular network, a Wireless Local Area Network (WLAN), a digital television network such as a DVB-H network, a satellite network, an AM-FM broadcast transmitter; and / or a device of another terminal device configured to receive / send communication signals; and / or an Internet of Things (IoT) device. A terminal device configured to communicate via a wireless interface may be referred to as a "wireless communication terminal", "wireless terminal" or "mobile terminal". Examples of mobile terminals include, but are not limited to, satellite or cellular telephones; Personal Communications System (PCS) terminals that may combine cellular radiotelephones with data processing, fax, and data communications capabilities; PDAs that may include radiotelephones, pagers, Internet / Intranet access, Web browsers, notepads, calendars, and / or Global Positioning System (GPS) receivers; and conventional laptop and / or palmtop receivers or other electronic devices that include radiotelephone transceivers. A terminal device may refer to an access terminal, user equipment (UE), a subscriber unit, a subscriber station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a user terminal, a terminal, a wireless communication device, a user agent, or a user device. The access terminal can be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA), a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, a vehicle-mounted device, a wearable device, a terminal device in a 5G network, or a terminal device in a future evolved PLMN, etc.

[0070] Optionally, terminal devices 120 may perform device-to-device (D2D) communication.

[0071] Optionally, the 5G system or 5G network may also be referred to as an NR system or NR network.

[0072] Figure 6 One network device and two terminal devices are shown exemplarily. Optionally, the communication system 100 may include multiple network devices and each network device may include another number of terminal devices within its coverage area, which is not limited in the embodiments of the present application.

[0073] Optionally, the communication system 100 may also include other network entities such as a network controller and a mobility management entity, which is not limited in the embodiments of the present application.

[0074] It should be understood that the device with communication function in the network / system in the embodiment of the present application can be referred to as a communication device. Figure 6 Taking the communication system 100 shown as an example, the communication equipment may include a network device 110 and a terminal device 120 with communication functions. The network device 110 and the terminal device 120 may be the specific devices described above and will not be repeated here. The communication equipment may also include other devices in the communication system 100, such as a network controller, a mobile management entity and other network entities, which is not limited in the embodiments of the present application.

[0075] An optional processing flow of the channel state information transmission method provided in the embodiment of the present application is as follows: Figure 7 As shown, the following steps are included:

[0076] Step S201: The terminal device determines channel state information of a channel based on a first neural network model.

[0077] In some embodiments, the terminal device determines the channel characteristics of at least one dimension of the channel. The dimension of the channel characteristics may include at least one of the following: frequency domain information dimension, time domain information dimension, transmission information dimension, transmission angle information dimension, reception information dimension, and reception angle information dimension. It can be understood that each channel element in the channel corresponds to the channel quality under specific conditions defined by the above dimensions. The above different dimensions are described below.

[0078] For the frequency domain information dimension, the granularity of the frequency domain information dimension may include: a first number of resource blocks or a second number of subcarriers; the first number is greater than or equal to 1, or the second number is greater than or equal to 1. For example, the granularity of the frequency domain information dimension may be a RB (a is greater than or equal to 1, such as 2RB, 4RB, 8RB), or the granularity of the frequency domain information dimension may be b subcarriers (b is greater than or equal to 1, such as 4 subcarriers, 6 subcarriers, 18 subcarriers).

[0079] For the transmission information dimension, the granularity of the transmission information dimension may include: a third number of transmission antennas or a fourth number of transmission antenna ports; the third number is greater than or equal to 1, or the fourth number is greater than or equal to 1. For example, the granularity of the transmission information dimension may be one transmission antenna or one transmission antenna port; for example, in a pair of transceiver systems, 32 antennas or 32 typical transmission ports are used in the transmission direction.

[0080] For the receiving information dimension, the granularity of the receiving information dimension may include: a fifth number of receiving antennas or a sixth number of receiving antenna ports; the fifth number is greater than or equal to 1, or the sixth number is greater than or equal to 1. For example, the granularity of the receiving information dimension may be one receiving antenna or one receiving antenna port; for example, in a pair of transceiver systems, the receiving uses 4 antennas for receiving, or uses 4 receiving ports.

[0081] The above-mentioned frequency domain information dimension, transmission information dimension and reception information dimension can process three different dimensions of channel characteristics, and each channel element in the channel corresponds to the channel quality under specific conditions corresponding to the above three dimensions. For example, a channel element in a channel corresponds to the third subcarrier, the 6th transmission antenna transmits, and the 2nd reception antenna receives, and the corresponding channel quality is -18dbm. For another example, a channel element in an original channel corresponds to the third subcarrier, the 6th transmission antenna port transmits, and the 4th reception antenna port receives, and the corresponding channel quality is -14dbm.

[0082] In the scenario where the channel characteristics include frequency domain information dimension, transmission information dimension and reception information dimension, an optional schematic diagram of channel input is shown as follows: Figure 8 As shown, the channel includes K frequency domain information dimensions, M transmission information granularities and N reception information granularities; wherein the granularity of the frequency domain information dimension is k=2RB, the granularity of the transmission information dimension is m=1 transmission antenna, and the granularity of the reception information dimension is n=1 reception antenna.

[0083] In some embodiments, any one of the above-mentioned frequency domain information dimension, sending information dimension and receiving information dimension can be processed, and the processed result can be used as a dimension of the channel; the frequency domain information dimension, sending information dimension and receiving information dimension are explained below respectively.

[0084] For the frequency domain information dimension, the frequency domain information dimension can be converted into the time domain information dimension through specific processing, and the converted time domain information dimension is used as a dimension of the channel. For example, the frequency domain information dimension is subjected to Fourier-related transformation, such as IFFT transformation. At this time, the time domain information dimension corresponds to the delay information of the channel, and the granularity of the delay information includes: the number of sampling points of the seventh number of symbol lengths, or the eighth number of symbols, or the ninth number of microseconds; wherein the seventh number is greater than or equal to 1, the eighth number is greater than or equal to 1, and the ninth number is greater than or equal to 1; such as the granularity of the delay information is p1 microseconds, or p2 symbol lengths, or the number of sampling points of p3 symbols; wherein the symbol can be an orthogonal frequency division multiplexing (OFDM) symbol. For example, when a channel element in a channel corresponds to a delay of 3 microseconds, the channel quality corresponding to the 6th transmitting antenna transmitting and the 2nd receiving antenna receiving is -23dbm.

[0085] In the scenario where the channel characteristics include the time domain information dimension, the sending information dimension, and the receiving information dimension, another optional schematic diagram of the channel input is as follows: Fig. 9 As shown, the channel includes T frequency domain information dimensions, M sending information granularities and N receiving information granularities; among which, the granularity of the time domain information dimension is t = 100 microseconds, the granularity of the sending information dimension is m = 1 transmitting antenna, and the granularity of the receiving information dimension is n = 1 receiving antenna.

[0086] With respect to the transmission information dimension, the transmission information dimension can be converted into the transmission angle information dimension through specific processing, and the converted transmission angle information dimension is used as a dimension of the channel. For example, a Fourier-related transform, such as IFFT transform, is performed on the transmission information dimension. At this time, the transmission angle information dimension corresponds to the transmission angle information of the channel.

[0087] With respect to the receiving information dimension, the receiving information dimension can be converted into the receiving angle information dimension through specific processing, and the converted receiving angle information dimension is used as a dimension of the channel. For example, a Fourier-related transform, such as IFFT transform, is performed on the receiving information dimension. At this time, the receiving angle information dimension corresponds to the receiving angle information of the channel.

[0088] For example, a channel element in a channel corresponds to a delay of 100 microseconds, a transmission angle of 30 degrees, and a receiving angle of 60 degrees, and the corresponding channel quality is -23dbm. For another example, a channel element in a channel corresponds to the 9th subcarrier, a transmission angle of 30 degrees, and a receiving angle of 60 degrees, and the corresponding channel quality is -12dbm.

[0089] In the scenario where the channel characteristics include the time domain information dimension, the transmission angle information dimension, and the reception angle information dimension, another optional schematic diagram of the channel input is as follows: Fig.10 As shown, the channel includes T time domain information dimensions, M transmission angle information granularities and N reception angle information granularities; wherein the granularity of the time domain information dimension is t = 100 microseconds, the granularity of the transmission angle information dimension is m = π / 12 transmitting antennas, and the granularity of the receiving information dimension is n = π / 6 receiving antennas.

[0090] In the scenario where the channel characteristics include the time domain information dimension, the transmission angle information dimension, and the reception angle information dimension, another optional schematic diagram of the channel input is as follows: Fig.11 As shown, the channel includes K frequency domain information dimensions, M transmission angle information granularities and N reception angle information granularities; wherein the granularity of the frequency domain information dimension is k=1 subcarrier, the granularity of the transmission angle information dimension is m=π / 12 transmitting antennas, and the granularity of the receiving information dimension is n=π / 6 receiving antennas.

[0091] The channel characteristics are described above. The following describes the processing flow of the terminal device determining the channel characteristics of at least one dimension of the channel.

[0092] In some embodiments, the terminal device determining the channel state information of the channel based on the first neural network model may include:

[0093] In step a, the terminal device determines the channel characteristics of at least one dimension of the channel based on the convolutional layer of the first neural network model; and determines the channel state information based on the channel characteristics of at least one dimension.

[0094] In a specific implementation, the terminal device can determine the channel characteristics of at least one dimension of the channel based on the convolution result of the convolution kernel corresponding to the convolution layer of the first neural network and the first channel information of the channel; wherein the number of the convolution layers can be one or more, and the first channel information is the local channel information of the channel or the partial channel information of the channel. The dimension of the convolution kernel matches the dimension of the channel characteristics, which can be understood as the dimension of the convolution kernel being equal to the dimension of the channel characteristics; for example, if the dimension of the channel characteristics is 3, then the dimension of the convolution kernel is also 3. In the case of multiple convolution layers, each convolution layer corresponds to a different number of convolution kernels.

[0095] A schematic diagram of the local channel information of a channel, such as Fig.12As shown, by performing convolution operation on the convolution kernel and the local channel information, the channel feature can be obtained; wherein the dimension of the convolution kernel is the same as the dimension of the channel feature. The process of the terminal device determining the channel feature based on the convolution layer of the first neural network model can also be called channel feature extraction, which can be implemented by the feature extraction unit in the terminal device.

[0096] When the channel state information is transmitted based on the first neural network model, the basic structural diagram of the feature extraction unit is as follows: Fig.13 As shown, it can be implemented by multiple convolutional layers in the first neural network model; the n convolutional layers in the first neural network model serve as feature extraction units to obtain channel features.

[0097] An optional structural diagram of the feature extraction unit, such as Fig.14 As shown in the figure, the three dimensions of the input channel features are 32 delay units, 32 transmitting antennas and 4 receiving antennas; the first convolution layer and the second convolution layer are used as feature extraction units, wherein the first convolution layer includes 4 convolution kernels, and the size of each convolution kernel is 3*3*3; the second convolution layer includes 2 convolution kernels, and the size of each convolution kernel is 3*3*3.

[0098] In some embodiments, in the process of determining the channel state information of the channel based on the first neural network model, the terminal device may further include, after executing step a:

[0099] In step b, the terminal device filters the convolution result of the convolution layer based on the fully connected layer of the first neural network model to determine information of at least one dimension of the channel.

[0100] It can be understood that after the feature extraction unit determines the feature information of at least one dimension, the terminal device filters the feature information of at least one dimension. The process of determining the information of at least one dimension of the channel can also be called redundancy elimination, which can be implemented by the redundancy elimination unit in the terminal device. When transmitting channel state information based on the first neural network model, the redundancy elimination unit in the terminal device can be implemented by at least one fully connected layer in the first neural network model, such as Fig.15 As shown, multiple fully connected layers in the first neural network model serve as redundant elimination units to filter and screen the channel features obtained by the convolutional layer.

[0101] In some embodiments, the information output by the fully connected layer of the first neural model is the channel state information.

[0102] In some other embodiments, the terminal device may also quantize the information output by the fully connected layer of the first neural model. Therefore, in the process of determining the channel state information of the channel based on the first neural network model, after executing step b, the terminal device may also include:

[0103] In step c, the terminal device quantifies the information characterizing the channel characteristics based on the first neural network model.

[0104] In specific implementation, the terminal device can convert the data result in step b into Q-bit information based on one or more fully connected layers in the first neural network model. The terminal device can also use a specific quantization function to quantize the information characterizing the channel characteristics; such as using uniform quantization or non-uniform quantization to quantize the output information of the encoding neural network into Qbits, and then output the channel state information.

[0105] Based on the process of determining the channel state information by the terminal device described above, the basic structure of the first neural network model for determining the channel state information by the terminal device in the embodiment of the present application is as follows: Fig.16 As shown, it includes multiple convolutional layers as feature extraction units, multiple fully connected layers as redundant elimination units, and quantization units; wherein each convolutional layer includes at least one convolution kernel.

[0106] In the case where the channel belongs to a channel in a MIMO scenario, an optional structure of the first neural network model for the MIMO scenario is as follows: Fig.17 As shown in the figure, the three dimensions of the input channel features are 32 delay units, 32 transmitting antennas and 4 receiving antennas; the first convolution layer and the second convolution layer are used as feature extraction units, wherein the first convolution layer includes 4 convolution kernels, each of which is 3*3*3 in size; the second convolution layer includes 2 convolution kernels, each of which is 3*3*3 in size. After the first and second convolution layers perform feature extraction, a vector with a length of 8192 is output. The vector with a length of 8192 is filtered by a redundant elimination unit in a fully connected layer process in the first neural network, and a vector with a length of 256 is output. The vector with a length of 256 is uniformly quantized by the first neural network model to obtain channel state information.

[0107] After determining the channel state information, the terminal device performs step S202:

[0108] Step S202: The terminal device sends the channel state information.

[0109] In some embodiments, the terminal device sends the channel state information to the network device.

[0110] Another optional processing flow of the channel state information transmission method provided in the embodiment of the present application is as follows: Fig.18 As shown, the following steps are included:

[0111] Step S301: A network device receives channel state information.

[0112] In some embodiments, the network device receives channel state information sent by the terminal device.

[0113] Step S302: The network device determines channel information corresponding to the channel state information based on a second neural network model.

[0114] In some embodiments, the network device determines the channel information corresponding to the channel state information based on the second neural network model, including:

[0115] In step 3a, the network device enhances the channel state information based on the first fully connected layer of the second neural network model.

[0116] In some embodiments, the network device may enhance the channel state information by an information enhancement unit in the network device, and the information enhancement unit corresponds to at least one fully connected layer in the second neural network model. A structural diagram of the information enhancement unit, such as Fig.19 As shown, it includes multiple fully connected layers.

[0117] After step 3a, the network device determining the channel information corresponding to the channel state information based on the second neural network model may also include:

[0118] In step 3b, the network device uses the first convolutional layer of the second neural network model to determine the channel information corresponding to the enhanced channel state information.

[0119] In some embodiments, the network device determines the channel information based on the convolution result of the convolution kernel corresponding to the first convolution layer and the enhanced channel state information. In a specific implementation, the process of determining the channel information corresponding to the enhanced channel state information can be implemented by a feature recovery unit in the network device. An optional schematic diagram of the channel input to the feature recovery unit, such as Fig. 20 As shown, it includes T' channel features of the first dimension, M' channel features of the second dimension, and N' channel features of the third dimension; the convolution operation is performed on the input information using the convolution kernel of the same dimension as the channel features input by the feature recovery unit, thereby recovering the channel features. A basic structural diagram of the feature recovery unit, as shown in Fig.21As shown, the feature recovery unit corresponds to at least one first convolutional layer in the second neural network model. The dimension of the convolution kernel in each first convolutional layer matches the dimension of the channel feature; it can be understood that the dimension of the convolution kernel in the first convolutional layer is the same as the dimension of the channel feature.

[0120] A detailed structural diagram of the feature recovery unit, such as Fig. 22 As shown, the three dimensions of the information input to the feature recovery unit are 32, 32 and 4 respectively; the feature recovery unit composed of a convolutional layer performs a convolution operation on the input information to restore the channel characteristics; wherein the convolutional layer can use 2 convolution kernels, each of which has a size of 3*3*3, and the three dimensions of the channel characteristics output by the feature recovery unit are 32 delay units as the first dimension, 32 transmitting antennas as the second dimension, and 4 receiving antennas as the third dimension.

[0121] After step 3b, the network device determining the channel information corresponding to the channel state information based on the second neural network model may also include:

[0122] Step 3c: The network device uses the second convolutional layer or the second fully connected layer of the second neural network model to process the convolution result and determine the channel information.

[0123] In some embodiments, the network device merges the convolution result with an output result after the convolution result is input into the second convolution layer or the second fully connected layer.

[0124] Therefore, the feature recovery unit may also include one or more residual network units for optimizing the result of feature recovery. The residual network unit includes at least one convolutional layer or at least one fully connected layer, and the result of merging the input information of the first layer of the residual network unit and the output information of the last layer of the residual network unit is used as the output result of the residual network unit, and the residual unit is used to optimize the channel characteristics.

[0125] In the embodiment of the present application, a schematic diagram of a second neural network model including a residual network unit is shown as follows: Fig.23 As shown in the figure, the three dimensions of the information input to the feature recovery unit are 32, 32 and 4 respectively; the input information is recovered by a feature recovery unit composed of a convolution layer with two convolution kernels, and the recovered information is used as the input of the network residual unit composed of two convolution layers. The two convolution layers use 8 and 2 convolution kernels respectively, and the size of each convolution kernel is 3*3*3. The information input to the network residual unit is combined with the information output by the network residual, and the three dimensions of the channel features are 32 delay units as the first dimension, 32 transmitting antennas as the second dimension, and 4 receiving antennas as the third dimension. Finally, the channel features are output.

[0126] In some embodiments, before executing step S302, the network device may also dequantize the received channel state information based on the second neural network model.

[0127] In specific implementation, the second neural network model can be used to dequantize the received Qbit information quantized by the coding neural network model to restore the information before quantization; or a specific quantization function can be used to dequantize the received channel state information, for example, using a dequantization method corresponding to the non-uniform quantization method of the coding neural network model. Then, the channel information after quantization recovery is output.

[0128] It should be noted that if the encoding neural network model does not adopt a quantization model, the decoding neural network model does not need to adopt a dequantization module; if the encoding neural network model adopts uniform quantization, the decoding neural network model may not adopt a dequantization module.

[0129] In an embodiment of the present application, the channel information includes channel characteristics, and the dimension of the channel characteristics includes at least one of the following: frequency domain information dimension, time domain information dimension, transmission information dimension, transmission angle information dimension, reception information dimension, and reception angle information dimension. Wherein, the granularity of the frequency domain information dimension includes: an eleventh number of resource blocks or a twelfth number of subcarriers; the eleventh number is greater than or equal to 1, or the twelfth number is greater than or equal to 1.

[0130] The granularity of the transmission information dimension includes: the thirteenth number of transmission antennas or the fourteenth number of transmission antenna ports; the thirteenth number is greater than or equal to 1, or the fourteenth number is greater than or equal to 1. The granularity of the reception information dimension includes: the fifteenth number of reception antennas or the sixteenth number of reception antenna ports; the fifteenth number is greater than or equal to 1, or the sixteenth number is greater than or equal to 1. The time domain information dimension characterizes the delay information of the channel. The granularity of the delay information includes: the seventeenth number of symbol lengths, or the number of sampling points of the eighteenth number of symbols, or the nineteenth number of microseconds; the seventeenth number is greater than or equal to 1, or the eighteenth number is greater than or equal to 1, or the nineteenth number is greater than or equal to 1. The symbol includes an OFDM symbol. The channel corresponding to the channel state information includes a channel in a MIMO system.

[0131] In summary, the basic structural diagram of the second neural network model for the MIMO scenario in the embodiment of the present application is as follows: Fig.24As shown, it includes a dequantization unit for dequantizing the received channel state information; an information enhancement unit composed of multiple fully connected layers for enhancing the dequantized information; a feature recovery unit composed of multiple convolutional layers for recovering the features of the information after information enhancement, and multiple residual units for further optimizing the recovered channel features and outputting them.

[0132] The detailed structural diagram of the second neural network model for the MIMO scenario in the embodiment of the present application is as follows: Fig.25 As shown, it includes an information enhancement unit composed of a fully connected layer, which is used to enhance the received channel state information; the enhanced information is input into the residual network unit composed of 2 convolutional layers after passing through the feature recovery unit; wherein, the 2 convolutional layers in the residual network unit use 8 and 2 convolutional kernels respectively, and the size of each convolutional kernel is 3*3*3, and the result output by the residual network unit is output after passing through the convolutional layer of the 2 convolutional kernels, thereby restoring the channel characteristics to 32 delay units as the first dimension, 32 transmitting antennas as the second dimension, and 4 receiving antennas as the third dimension.

[0133] It should be noted that in the above embodiments of the present application, the fully connected layer and the convolutional layer are basic concepts in the neural network model, which will not be described in detail here. Each layer involved in the above embodiments of the present application can add a common layer to process the output results of each layer; wherein the common layer includes normalizing the output results of each layer, such as through the BatchNormalization layer; including the output results of each layer through the excitation process is a basic process within the scope of the existing neural network, and the details will not be repeated here.

[0134] The embodiment of the present application provides an optional processing flow for transmitting channel state information based on a neural network model, such as Fig.26 As shown, the network device pre-configures the indication parameter information for CSI indication, such as which information the network device configures the terminal device needs to indicate in the CSI indication, such as CQI, PMI, and RI. At the same time, the network device configures some reference signals for CSI measurement, such as SSB or CSI-RS. The terminal device measures the above-mentioned reference signal and encodes the measurement result using the first neural network model to obtain the current channel state information, and determines the indication parameter information configured by the network device to indicate the current channel state information to the network device. The network device uses the second neural network model to decode the channel state information indicated by the terminal device to obtain the channel information corresponding to the channel state information. The network device uses the decoded channel information to configure a reasonable and efficient data transmission method.

[0135] The channel state information transmission method provided in the embodiment of the present application encodes the channel state information using a first neural network model at the transmitting end of the channel state information; and decodes the channel state information using a second neural network model at the receiving end of the channel state information; and the dimension of the convolution kernel used in the convolution layer in the neural network model for encoding and decoding is the same as the dimension of the channel feature; in this way, the terminal device can provide accurate feedback of the entire channel information to the network device, rather than providing feedback on the local features of the channel.

[0136] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0137] To implement the above-mentioned channel state information transmission method, an embodiment of the present application provides a terminal device, and an optional structural diagram of the terminal device 400 is as follows: Fig. 27 As shown, including:

[0138] A first processing unit 401 is configured to determine channel state information of a channel based on a first neural network model;

[0139] The sending unit 402 is configured to send the channel state information.

[0140] In some embodiments, the first processing unit 401 is configured to determine the channel characteristics of at least one dimension of the channel based on the convolutional layer of the first neural network model; and determine the channel state information based on the channel characteristics of the at least one dimension.

[0141] In some embodiments, the first processing unit 401 is configured to determine the channel characteristics of at least one dimension of the channel based on the convolution result of the convolution kernel corresponding to the convolution layer and the first channel information of the channel; the first channel information includes partial channel information of the channel.

[0142] In some embodiments, the dimension of the convolution kernel matches the dimension of the channel features.

[0143] In some embodiments, the first processing unit 401 is further configured to filter the convolution result of the convolution layer based on the fully connected layer of the first neural network model to determine information of at least one dimension of the channel.

[0144] In some embodiments, the first processing unit 401 is further configured to quantify the information characterizing the channel characteristics based on the first neural network model.

[0145] In some embodiments, the dimension of the channel characteristics includes at least one of the following: a frequency domain information dimension, a time domain information dimension, a transmission information dimension, a transmission angle information dimension, a reception information dimension, and a reception angle information dimension.

[0146] In some embodiments, the granularity of the frequency domain information dimension includes: a first number of resource blocks or a second number of subcarriers; the first number is greater than or equal to 1, or the second number is greater than or equal to 1.

[0147] In some embodiments, the granularity of the transmission information dimension includes: a third number of transmission antennas or a fourth number of transmission antenna ports; the third number is greater than or equal to 1, or the fourth number is greater than or equal to 1.

[0148] In some embodiments, the granularity of the received information dimension includes: a fifth number of receiving antennas or a sixth number of receiving antenna ports; the fifth number is greater than or equal to 1, or the sixth number is greater than or equal to 1.

[0149] In some embodiments, the time domain information dimension represents delay information of the channel.

[0150] In some embodiments, the granularity of the delay information includes: a seventh number of symbol lengths, or the number of sampling points of an eighth number of symbols, or a ninth number of microseconds; the seventh number is greater than or equal to 1, or the eighth number is greater than or equal to 1, or the ninth number is greater than or equal to 1.

[0151] In some embodiments, the symbols include: OFDM symbols.

[0152] In some embodiments, the channel comprises a channel in a MIMO system.

[0153] To implement the above-mentioned channel state information transmission method, the embodiment of the present application provides a network device, and the optional structural diagram of the network device 500 is as follows: Fig.28 As shown, including:

[0154] The receiving unit 501 is configured to receive channel state information;

[0155] The second processing unit 502 is configured to determine channel information corresponding to the channel state information based on a second neural network model.

[0156] In some embodiments, the second processing unit 502 is configured to enhance the channel state information based on the first fully connected layer of the second neural network model.

[0157] In some embodiments, the second processing unit 502 is configured to use the first convolutional layer of the second neural network model to determine the channel information corresponding to the enhanced channel state information.

[0158] In some embodiments, the second processing unit 502 is configured to determine the channel information based on a convolution result of a convolution kernel corresponding to the first convolution layer and the enhanced channel state information.

[0159] In some embodiments, the second processing unit 502 is configured to process the convolution result using the second convolutional layer or the second fully connected layer of the second neural network model to determine the channel information.

[0160] In some embodiments, the second processing unit 502 is configured to merge the convolution result with the output result after the convolution result is input into the second convolutional layer or the second fully connected layer.

[0161] In some embodiments, the second processing unit 502 is further configured to dequantize the channel state information based on the second neural network model.

[0162] In some embodiments, the channel information includes channel characteristics, and the dimensions of the channel characteristics include at least one of the following: frequency domain information dimension, time domain information dimension, transmission information dimension, transmission angle information dimension, reception information dimension and reception angle information dimension.

[0163] In some embodiments, the granularity of the frequency domain information dimension includes: an eleventh number of resource blocks or a twelfth number of subcarriers; the eleventh number is greater than or equal to 1, or the twelfth number is greater than or equal to 1.

[0164] In some embodiments, the granularity of the transmission information dimension includes: a thirteenth number of transmitting antennas or a fourteenth number of transmitting antenna ports; the thirteenth number is greater than or equal to 1, or the fourteenth number is greater than or equal to 1.

[0165] In some embodiments, the granularity of the received information dimension includes: a fifteenth number of receiving antennas or a sixteenth number of receiving antenna ports; the fifteenth number is greater than or equal to 1, or the sixteenth number is greater than or equal to 1.

[0166] In some embodiments, the time domain information dimension represents delay information of the channel.

[0167] In some embodiments, the granularity of the delay information includes: a seventeenth number of symbol lengths, or the number of sampling points of an eighteenth number of symbols, or a nineteenth number of microseconds; the seventeenth number is greater than or equal to 1, or the eighteenth number is greater than or equal to 1, or the nineteenth number is greater than or equal to 1.

[0168] In some embodiments, the symbols include: OFDM symbols.

[0169] In some embodiments, the channel corresponding to the channel state information includes a channel in a MIMO system.

[0170] An embodiment of the present application also provides a terminal device, including a processor and a memory for storing a computer program that can be run on the processor, wherein the processor is used to execute the steps of the channel state information transmission method executed by the above-mentioned terminal device when running the computer program.

[0171] An embodiment of the present application also provides a network device, including a processor and a memory for storing a computer program that can be run on the processor, wherein the processor is used to execute the steps of the channel state information transmission method executed by the above-mentioned network device when running the computer program.

[0172] An embodiment of the present application also provides a chip, including: a processor, used to call and run a computer program from a memory, so that a device equipped with the chip executes the channel state information transmission method executed by the above-mentioned terminal device.

[0173] An embodiment of the present application also provides a chip, including: a processor, used to call and run a computer program from a memory, so that a device equipped with the chip executes the channel state information transmission method executed by the above-mentioned network device.

[0174] An embodiment of the present application also provides a storage medium storing an executable program, which, when executed by a processor, implements the channel state information transmission method executed by the above-mentioned terminal device.

[0175] An embodiment of the present application also provides a storage medium storing an executable program, and when the executable program is executed by a processor, the channel state information transmission method executed by the above-mentioned network device is implemented.

[0176] An embodiment of the present application also provides a computer program product, including computer program instructions, which enable a computer to execute the channel state information transmission method executed by the above-mentioned terminal device.

[0177] An embodiment of the present application also provides a computer program product, including computer program instructions, which enable a computer to execute the channel state information transmission method executed by the above-mentioned network device.

[0178] An embodiment of the present application also provides a computer program, which enables a computer to execute the channel state information transmission method executed by the above-mentioned terminal device.

[0179] An embodiment of the present application also provides a computer program, which enables a computer to execute the channel state information transmission method executed by the above-mentioned network device.

[0180] Fig.29 700 is a schematic diagram of the hardware structure of an electronic device (terminal device or network device) of an embodiment of the present application. The electronic device 700 includes: at least one processor 701, a memory 702 and at least one network interface 704. The various components in the electronic device 700 are coupled together through a bus system 705. It can be understood that the bus system 705 is used to realize the connection and communication between these components. In addition to the data bus, the bus system 705 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, Fig.29 Various buses are labeled as bus system 705 .

[0181] It can be understood that the memory 702 can be a volatile memory or a non-volatile memory, and can also include both volatile and non-volatile memories. Among them, the non-volatile memory can be ROM, programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic random access memory (FRAM), ferromagnetic random access memory, flash memory, magnetic surface memory, optical disk, or compact disc read-only memory (CD-ROM); the magnetic surface memory can be a disk memory or a tape memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), direct memory bus random access memory (DRRAM). The memory 702 described in the embodiments of the present application is intended to include but is not limited to these and any other suitable types of memory.

[0182] The memory 702 in the embodiment of the present application is used to store various types of data to support the operation of the electronic device 700. Examples of such data include: any computer program for operating on the electronic device 700, such as the application 7022. The program for implementing the method of the embodiment of the present application may be included in the application 7022.

[0183] The method disclosed in the above embodiment of the present application can be applied to the processor 701, or implemented by the processor 701. The processor 701 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 701. The above processor 701 may be a general-purpose processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor 701 can implement or execute the methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or any conventional processor, etc. In combination with the steps of the method disclosed in the embodiment of the present application, it can be directly embodied as a hardware decoding processor to execute, or it can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium, which is located in the memory 702, and the processor 701 reads the information in the memory 702 and completes the steps of the above method in combination with its hardware.

[0184] In an exemplary embodiment, the electronic device 700 may be implemented by one or more application specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), FPGAs, general purpose processors, controllers, MCUs, MPUs, or other electronic components to execute the aforementioned method.

[0185] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0186] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0187] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0188] It should be understood that the terms "system" and "network" are often used interchangeably in this application. The term "and / or" in this application is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this application generally indicates that the associated objects before and after are in an "or" relationship.

[0189] The above is only a preferred embodiment of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for transmitting channel state information, the method comprising: The terminal device determines channel state information of the channel based on the first neural network model; The terminal device sends the channel state information.

2. The method according to claim 1, wherein: The terminal device determines the channel state information of the channel based on the first neural network model, including: The terminal device determines a channel feature of at least one dimension of the channel based on a convolutional layer of the first neural network model; The channel state information is determined based on the channel characteristics of the at least one dimension.

3. The method according to claim 2, wherein: The terminal device determines, based on the convolution layer of the first neural network model, a channel feature of at least one dimension of the channel, including: The terminal device determines the channel feature of the at least one dimension of the channel based on a convolution result of a convolution kernel corresponding to the convolution layer and the first channel information of the channel; The first channel information includes partial channel information of the channel.

4. The method according to claim 3, wherein: The dimension of the convolution kernel matches the dimension of the channel feature.

5. The method according to any one of claims 2 to 4, wherein: The method further comprises: The terminal device filters the convolution result of the convolution layer based on the fully connected layer of the first neural network model to determine information of at least one dimension of the channel.

6. The method according to any one of claims 2 to 4, wherein: The method further comprises: The terminal device quantifies the information characterizing the channel characteristics based on the first neural network model.

7. The method according to any one of claims 2 to 4, wherein: The dimension of the channel characteristic includes at least one of the following: Frequency domain information dimension, time domain information dimension, sending information dimension, sending angle information dimension, receiving information dimension and receiving angle information dimension.

8. A method for transmitting channel state information, the method comprising: The network device receives the channel state information; The network device determines channel information corresponding to the channel state information based on a second neural network model.

9. A terminal device, comprising: A first processing unit configured to determine channel state information of a channel based on a first neural network model; A sending unit is configured to send the channel state information.

10. The terminal device according to claim 9, wherein: The first processing unit is configured to determine a channel feature of at least one dimension of the channel based on a convolutional layer of the first neural network model; The channel state information is determined based on the channel characteristics of the at least one dimension.

11. The terminal device according to claim 10, wherein: The first processing unit is configured to determine the channel feature of the at least one dimension of the channel based on a convolution result of a convolution kernel corresponding to the convolution layer and the first channel information of the channel; The first channel information includes partial channel information of the channel.

12. The terminal device according to claim 11, wherein: The dimension of the convolution kernel matches the dimension of the channel feature.

13. The terminal device according to any one of claims 10 to 12, wherein: The first processing unit is also configured to filter the convolution result of the convolution layer based on the fully connected layer of the first neural network model to determine information of at least one dimension of the channel.

14. The terminal device according to any one of claims 10 to 12, wherein: The first processing unit is further configured to quantify the information characterizing the channel characteristics based on the first neural network model.

15. A network device, comprising: A receiving unit configured to receive channel state information; The second processing unit is configured to determine channel information corresponding to the channel state information based on a second neural network model.

Images