Data transmission method, device, equipment, storage medium and system

Abstract

The invention discloses a data transmission method, device and system, equipment and a storage medium, and belongs to the technical field of communication. The method comprises: a network device sending a downlink pilot signal to a terminal device (310); the terminal device encodes and compresses downlink information obtained based on downlink pilot signal estimation through an AI / ML model to obtain codeword information (320); the terminal device determines CSI of the downlink channel, the CSI comprising RI and / or CQI (330); the terminal device reports CSI feedback information to the network device, the CSI feedback information comprising at least one of codeword information and CSI (340); and the network device transmits downlink data on the downlink channel according to the CSI feedback information (350). The reported RI / CQI is better matched with the precoding for data transmission, so that the data transmission of the user is accurately and reasonably scheduled.

Description

Data transmission method, device, equipment, storage medium and system Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a data transmission method, apparatus, device, storage medium and system. Background Art

[0002] CSI (Channel State Information) reporting refers to the terminal device obtaining CSI based on the measurement of the downlink pilot signal (such as CSI-RS (Channel State Information Reference Signal)) sent by the network device, and reporting the above CSI to the network device according to the reporting method and uplink resources configured by the network device.

[0003] In related technologies, AI (Artificial Intelligence) / ML (Machine Learning) models are introduced into CSI reporting scenarios. For example, a terminal device measures CSI based on downlink pilot signals, compresses the CSI using an AI / ML model for encoding to obtain the compressed CSI, and then quantizes the compressed CSI into a binary bit stream and sends it to a network device. The network device then dequantizes the received binary bit stream and inputs the dequantized information into an AI / ML model for decoding to obtain the recovered CSI.

[0004] At present, the CSI reporting scenarios that introduce AI / ML models need further research.

[0005] Summary of the Invention

[0006] The embodiments of the present application provide a data transmission method, apparatus, device, storage medium, and system. The technical solution is as follows:

[0007] According to one aspect of an embodiment of the present application, a data transmission method is provided, the method being performed by a terminal device, the method comprising:

[0008] receiving a downlink pilot signal sent by a network device;

[0009] encoding and compressing downlink information estimated based on the downlink pilot signal using an AI / ML model to obtain codeword information;

[0010] Determine a CSI of a downlink channel, where the CSI includes an RI (Rank Indication) and / or a CQI (Channel Quality Indication);

[0011] Report CSI feedback information to a network device, where the CSI feedback information includes at least one of the codeword information and the CSI.

[0012] According to one aspect of an embodiment of the present application, a data transmission method is provided, the method being performed by a network device, the method comprising:

[0013] Sending a downlink pilot signal to a terminal device;

[0014] receiving CSI feedback information reported by the terminal device, where the CSI feedback information includes at least one of codeword information and CSI of a downlink channel, where the codeword information is obtained by the terminal device by encoding and compressing downlink information estimated based on the downlink pilot signal through an AI / ML model, and the CSI includes RI and / or CQI;

[0015] Transmit downlink data on the downlink channel according to the CSI feedback information.

[0016] According to one aspect of an embodiment of the present application, a data transmission device is provided, the device comprising:

[0017] A receiving module, configured to receive a downlink pilot signal sent by a network device;

[0018] a processing module, configured to encode and compress downlink information estimated based on the downlink pilot signal through an AI / ML model to obtain codeword information; and determine channel state information (CSI) of a downlink channel, where the CSI includes RI and / or CQI;

[0019] The sending module is configured to report CSI feedback information to a network device, where the CSI feedback information includes at least one of the codeword information and the CSI.

[0020] According to one aspect of an embodiment of the present application, a data transmission device is provided, the device comprising:

[0021] A sending module, configured to send a downlink pilot signal to a terminal device;

[0022] a receiving module, configured to receive CSI feedback information reported by the terminal device, the CSI feedback information including at least one of codeword information and CSI of a downlink channel, the codeword information being obtained by the terminal device by encoding and compressing downlink information estimated based on the downlink pilot signal using an AI / ML model, the CSI including RI and / or CQI;

[0023] A processing module is configured to transmit downlink data on the downlink channel according to the CSI feedback information.

[0024] According to one aspect of an embodiment of the present application, a terminal device is provided, comprising a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-mentioned data transmission method on the terminal device side.

[0025] According to one aspect of an embodiment of the present application, a network device is provided, comprising a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-mentioned data transmission method on the network device side.

[0026] According to one aspect of an embodiment of the present application, a computer-readable storage medium is provided, in which a computer program is stored. The computer program is used to be executed by a processor to implement the above-mentioned data transmission method on the terminal device side, or to implement the above-mentioned data transmission method on the network device side.

[0027] According to one aspect of an embodiment of the present application, a communication system is provided, which includes a terminal device and a network device, wherein the terminal device is used to implement the above-mentioned data transmission method on the terminal device side, and the network device is used to implement the above-mentioned data transmission method on the network device side.

[0028] According to one aspect of an embodiment of the present application, a chip is provided, which includes a programmable logic circuit and / or program instructions. When the chip is running, it is used to implement the above-mentioned data transmission method on the terminal device side, or to implement the above-mentioned data transmission method on the network device side.

[0029] According to one aspect of an embodiment of the present application, a computer program product is provided, which includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor reads and executes the computer instructions from the computer-readable storage medium to implement the above-mentioned data transmission method on the terminal device side, or to implement the above-mentioned data transmission method on the network device side.

[0030] The technical solutions provided by the embodiments of the present application may have the following beneficial effects:

[0031] The CSI compression and recovery architecture based on a bilateral AI / ML model of encoder and decoder achieves better matching of the RI / CQI reported by the terminal device to the network device with the precoding used for data transmission, enabling the network device to accurately and reasonably schedule user data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0033] FIG1 is a schematic diagram of a network architecture provided by an exemplary embodiment of the present application;

[0034] FIG2 is a schematic diagram of a CSI compression and recovery architecture based on a bilateral AI / ML model of an encoder and a decoder according to an exemplary embodiment of the present application;

[0035] FIG3 is a flow chart of a data transmission method provided by an exemplary embodiment of the present application;

[0036] FIG4 is a flowchart of a data transmission method provided by another exemplary embodiment of the present application;

[0037] FIG5 is a flowchart of a data transmission method provided by another exemplary embodiment of the present application;

[0038] FIG6 is a block diagram of a data transmission device provided by an exemplary embodiment of the present application;

[0039] FIG7 is a block diagram of a data transmission device provided by another exemplary embodiment of the present application;

[0040] FIG8 is a block diagram of a communication device provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0041] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0042] FIG1 is a schematic diagram of a network architecture 100 provided by an exemplary embodiment of the present application. The network architecture 100 may include: a terminal device 10 , an access network device 20 , and a core network device 30 .

[0043] The terminal device 10 may refer to a UE (User Equipment), an access terminal, a subscriber unit, a subscriber station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a wireless communication device, a user agent, or a user apparatus. Optionally, the terminal device 10 may also be a cellular phone, a cordless phone, a SIP (Session Initiation Protocol) phone, a WLL (Wireless Local Loop) station, a PDA (Personal Digital Assistant), a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a wearable device, a terminal device in a 5GS (5th Generation System) or a terminal device in a future-evolved PLMN (Public Land Mobile Network), etc., but the embodiments of the present application are not limited thereto. For ease of description, the above-mentioned devices are collectively referred to as terminal devices. The number of terminal devices 10 is generally multiple, and one or more terminal devices 10 may be distributed in each cell managed by an access network device 20. In the embodiments of the present application, “terminal device” and “UE” are often used interchangeably, but those skilled in the art will understand that the two generally express the same meaning.

[0044] Access network equipment 20 is a device deployed in an access network to provide wireless communication capabilities for terminal device 10. Access network equipment 20 may include various forms of macro base stations, micro base stations, relay stations, access points, and the like. In systems employing different wireless access technologies, the names of devices that provide access network equipment functions may vary. For example, in 5G NR systems, they are referred to as gNodeBs or gNBs. As communication technologies evolve, the term "access network equipment" may change. For ease of description, in the embodiments of this application, the aforementioned devices that provide wireless communication capabilities for terminal device 10 are collectively referred to as access network equipment. Optionally, a communication relationship can be established between terminal device 10 and core network equipment 30 through access network equipment 20. For example, in an LTE (Long Term Evolution) system, access network equipment 20 may be an Evolved Universal Terrestrial Radio Access Network (EUTRAN) or one or more eNodeBs within EUTRAN. In a 5G NR system, access network equipment 20 may be a Radio Access Network (RAN) or one or more gNBs within the RAN. In the embodiment of the present application, unless otherwise specified, the "network device" refers to the access network device 20, such as a base station.

[0045] The core network device 30 is a device deployed in the core network. The functions of the core network device 30 are mainly to provide user connections, user management, and service bearer, and to provide an interface to the external network as a bearer network. For example, the core network devices in the 5G NR system may include devices such as the AMF (Access and Mobility Management Function) entity, the UPF (User Plane Function) entity, and the SMF (Session Management Function) entity.

[0046] In some embodiments, the access network device 20 and the core network device 30 communicate with each other via an air interface technology, such as the NG interface in the 5G NR system. The access network device 20 and the terminal device 10 communicate with each other via an air interface technology, such as the Uu interface.

[0047] The "5G NR system" in the embodiments of the present application may also be referred to as a 5G system or an NR system, but those skilled in the art will understand its meaning. The technical solutions described in the embodiments of the present application may be applicable to LTE systems, 5G NR systems, subsequent evolution systems of 5G NR systems, and other communication systems such as NB-IoT (Narrow Band Internet of Things) systems, and this application does not limit this.

[0048] In an embodiment of the present application, the network device can provide services for a cell, and the terminal device communicates with the network device through the transmission resources (for example, frequency domain resources, or spectrum resources) on the carrier used by the cell. The cell can be a cell corresponding to the network device (for example, a base station). The cell can belong to a macro base station or a base station corresponding to a small cell. The small cells here may include: metro cell, micro cell, pico cell, femto cell, etc. These small cells have the characteristics of small coverage and low transmission power, and are suitable for providing high-speed data transmission services.

[0049] Related technologies include Type I and Type II codebooks for implementing quantized CSI feedback. CSI feedback based on the Type I codebook has low overhead, but the precoding accuracy calculated based on this codebook is low, resulting in poor channel matching and poor data transmission performance. CSI feedback based on the Type II codebook has high overhead, but the precoding accuracy calculated based on this codebook is high, which better matches the channel and can improve data transmission performance. How to further reduce feedback overhead while maintaining precoding accuracy, or how to achieve higher precoding accuracy while maintaining feedback overhead to achieve a better performance-cost trade-off, remains an unresolved issue.

[0050] Due to its powerful computational reasoning capabilities, AI technology has the ability to simulate arbitrary nonlinear functions and is currently widely used in various industries. AI technology and machine learning have also been widely studied and applied in communication systems. To address the above issues and reduce the feedback overhead of terminal devices or improve the accuracy of CSI feedback, the 3GPP (3rd Generation Partnership Project) standardization research has carried out CSI compression and recovery based on a bilateral AI / ML model of encoder and decoder. That is, the encoder model is deployed on the terminal device side to compress the downlink channel information or the feature vector corresponding to the downlink channel information, and then the decoder model is deployed on the network device side to restore the original channel information or feature vector information, as shown in Figure 2 below.

[0051] In related technologies, CSI reporting based on Type I / II codebooks reports PMI together with RI or CQI. For Type II codebooks, RI, CQI, and PMI are reported in two parts: Part I and Part II. RI and CQI are reported in Part I, while PMI is reported in Part II. The network device calculates the precoding for downlink data transmission based on the CSI (including RI, CQI, and PMI) reported by the terminal device, and determines the number of data transmission streams and the coding and modulation scheme.

[0052] For CSI compression feedback based on the bilateral AI / ML model of the encoder and decoder, the precoding of downlink data transmission is determined by the network device based on the information recovered by the decoder model. If only the encoder model is deployed on the terminal device but not the decoder model, the terminal device cannot obtain the precoding information of the downlink data transmission, and accordingly cannot directly calculate the corresponding RI and / or CQI based on the precoding obtained by the network device; if the terminal device calculates the RI / CQI based on the traditional codebook, or calculates the RI / CQI based on the ideal precoding, this will result in inconsistency with the RI / CQI calculated by the network device based on the precoding determined by the decoder model, making it impossible for the network device to accurately schedule downlink data transmission based on the RI / CQI reported by the terminal device.

[0053] FIG3 is a flowchart of a data transmission method provided by an exemplary embodiment of the present application. This embodiment uses the method as an example to illustrate the network architecture shown in FIG1. ​​The method may include at least one of the following steps 310 to 350:

[0054] Step 310: The network device sends a downlink pilot signal to the terminal device.

[0055] Accordingly, the terminal device receives the downlink pilot signal sent by the network device. The downlink pilot signal is carried in the downlink channel. The downlink pilot signal is used by the terminal device to report the CSI, and can also be called a downlink reference signal. Exemplarily, the downlink pilot signal is a CSI-RS (Channel Status Information-Reference Signal).

[0056] In step 320 , the terminal device encodes and compresses the downlink information estimated based on the downlink pilot signal through the AI / ML model to obtain codeword information.

[0057] Under the CSI compression and recovery architecture of the bilateral AI / ML model based on the encoder and decoder, the AI / ML model for encoding is deployed on the terminal device side. The AI / ML model for encoding can also be called an encoder or encoding model.

[0058] The downlink information includes downlink channel information or a feature vector corresponding to the downlink channel information. The feature vector corresponding to the downlink channel information can be extracted from the downlink channel information. The terminal device performs channel measurement and estimation based on the received downlink pilot signal to obtain the downlink channel information. Optionally, a feature vector corresponding to the downlink channel information is further obtained. The terminal device encodes and compresses the above downlink information through the AI / ML model used for encoding to obtain codeword information. The codeword information is a binary sequence consisting of 0s and 1s, or a binary bit stream. The above downlink channel information can also be called full channel information.

[0059] In some embodiments, the codeword length of the codeword information is configured by a network device. The network device sends first configuration information to the terminal device, and the first configuration information is used to configure the codeword length of the codeword information. Exemplarily, the first configuration information includes at least one codeword length corresponding to each rank value. After obtaining the above-mentioned downlink information, the terminal device generates c codewords according to the codeword length configured by the first configuration information, and quantizes the c codewords to obtain the above-mentioned codeword information, where c is a positive integer. In some embodiments, the codeword length of the codeword information can also be determined by the terminal device and reported to the network device.

[0060] In some embodiments, when the input information of the AI / ML model used for encoding (i.e., the downlink information input to the model) is the full channel information (i.e., the "downlink channel information" described above), c = 1. When the input information of the AI / ML model used for encoding (i.e., the downlink information input to the model) is the feature vector corresponding to the full channel information (i.e., the "feature vector corresponding to the downlink channel information" described above), c = 1 or v, where v represents the number of feature vectors corresponding to the full channel information, and v is a positive integer. For example, if the number of feature vectors corresponding to the full channel information is v = 2, the AI / ML model used for encoding can process the two feature vectors separately to generate two codewords, and then quantize the two codewords to obtain codeword information; alternatively, the two feature vectors can be regarded as a whole, and the AI / ML model used for encoding can process the whole to generate one codeword, and then quantize the one codeword to obtain codeword information.

[0061] In addition, the value of v can be predefined by negotiation between the terminal device and the network device. For example, if the relevant standard or protocol specifies the value of v (for example, the default value of v is 2), then both the terminal device and the network device know the value of v. For example, the value of v can be predefined as the maximum number of eigenvectors v max , v max is a positive integer. Alternatively, the value of v may be determined by the terminal device and reported to the network device via the RI. Alternatively, the value of v may be determined by the network device and sent to the terminal device in advance. This embodiment of the present application is not limited to this.

[0062] In step 330, the terminal device determines the CSI of the downlink channel, where the CSI includes RI and / or CQI.

[0063] The value of v mentioned above is also the rank value.

[0064] In the case where the rank value is predefined by negotiation between the terminal device and the network device, it can be understood that the RI is predefined by negotiation between the terminal device and the network device.

[0065] When the rank value is determined by the terminal device, the terminal device can use the following method 1 or method 2 to determine the RI.

[0066] Method 1: The terminal device determines a codebook parameter corresponding to the codeword length of the codeword information in a codebook according to the codeword length of the codeword information, and determines the CSI according to the codebook parameter and downlink information.

[0067] The codeword length of the codeword information is the codeword length of the codeword information obtained in step 320 above, denoted as c. Based on the correspondence between the codeword length of the codeword information and codebook parameters in the codebook, the terminal device determines the codebook parameters corresponding to the codeword length of the codeword information in the codebook, and then determines the RI based on the codebook parameters and downlink channel information. In some embodiments, the terminal device determines the PMI based on the codebook parameters, and determines the RI based on the PMI and downlink channel information.

[0068] Exemplarily, the codebook may be a Type I / II codebook, for example, the codebook is an eType II codebook (eType II codebook, eType II CB) or a further enhanced Type II port selection codebook (Further enhanced Type II Port selection codebook, FeType II PS CB).

[0069] Exemplarily, the correspondence between the codeword length of the codeword information and the codebook parameters in the codebook is configured by the network device, or is predefined by negotiation between the terminal device and the network device. The network device configures or predefines the correspondence between at least one set of codeword lengths and codebook parameters (or referred to as a mapping relationship) through negotiation. In the case where the above-mentioned correspondence is configured by the network device, the network device may send second configuration information to the terminal device, and the second configuration information is used to configure the correspondence between the codeword length of the codeword information and the codebook parameters in the codebook. Based on the second configuration information, the terminal device can obtain the correspondence between at least one set of codeword lengths and codebook parameters.

[0070] In some embodiments, the correspondence between the codeword length and the codebook parameters is designed in the following manner: for a set of corresponding codeword lengths and codebook parameters, the first SGCS (Square of General Cosine similarity) calculated based on the codeword information corresponding to the codeword length and the second SGCS calculated based on the codebook parameters meet the conditions. The first SGCS is an SGCS calculated based on the precoding matrix and the eigenvector of the downlink channel determined by the recovered downlink information, and the recovered downlink information is obtained by decoding the codeword information corresponding to the codeword length through the AI / ML model for decoding. The second SGCS is an SGCS calculated based on the precoding matrix and the eigenvector of the downlink channel determined by the codebook parameters. Exemplarily, the above conditions include one of the following: the first SGCS and the second SGCS are the same; or, the difference between the first SGCS and the second SGCS is less than or equal to a threshold value, which can be predefined by negotiation between the terminal device and the network device, or configured to the terminal device by the network device. Exemplarily, the calculation formula of SGCS is:

[0071]

[0072] Where K is the number of downlink channels (or sub-bands), w k represents the eigenvector of the kth downlink channel, w′ k represents the precoding matrix of the kth downlink channel determined according to the codebook parameters, H represents conjugate transpose, E represents averaging / expectation, and |||| represents modulo.

[0073] When designing the correspondence between the codeword length and the codebook parameters, the precoding matrix and the eigenvector of the downlink channel determined according to the recovered downlink information are used as w′ in the above formula respectively. k and w k , the first SGCS can be calculated; the precoding matrix determined according to the codebook parameters and the eigenvector of the downlink channel are respectively used as w′ in the above formula k and w k , the second SGCS can be calculated; if the two SGCS meet the conditions, a corresponding relationship between a set of codeword lengths and codebook parameters can be established.

[0074] In some embodiments, the codeword length of the codeword information is: the codeword length of the encoded and compressed CSI output by the AI / ML model used for encoding. In some embodiments, the codeword length of the codeword information is: the codeword length of the binary bit stream obtained by quantizing the encoded and compressed CSI output by the AI / ML model used for encoding.

[0075] In some embodiments, RI may also be determined based on a relationship other than the correspondence between the codeword length and the codebook parameter. As an alternative to method 1, the following method is used to determine the RI of the downlink channel: the terminal device determines at least one eigenvector based on the downlink information, calculates the SINR (Signal Interference Noise Ratio) at different rank values ​​based on the at least one eigenvector, and determines the RI. Exemplarily, the terminal device performs SVD (Singular Value Decomposition) on the downlink channel based on the downlink channel information to obtain at least one eigenvector, calculates the SINR at different rank values ​​based on the at least one eigenvector, and determines the RI. For example, the terminal device performs SVD decomposition on the downlink channel based on the downlink channel information to obtain r≥v eigenvectors, and then determines the appropriate v value by calculating the SINR at different v values. For example, the v value with the largest SINR is selected as the RI, or v=1, 2, ..., v is calculated in sequence. maxWhen the SINR corresponding to a certain v value meets the conditions (such as being greater than or equal to a certain threshold, or the increase compared to the SINR corresponding to the previous v value is less than or equal to a certain threshold), the v value is used as the RI.

[0076] Method 2: After receiving the codeword information reported by the terminal device, the network device sends v beam-formed downlink pilot signals to the terminal device. Correspondingly, the terminal device receives v beam-formed downlink pilot signals sent by the network device. The v beams are the v vectors contained in the precoding matrix determined based on the recovered downlink information. The recovered downlink information is obtained by decoding the codeword information using the AI / ML model for decoding, and v is a positive integer. The terminal device determines the RI based on the effective channel information estimated based on the v beam-formed downlink pilot signals.

[0077] Exemplarily, the downlink pilot signal may be a CSI-RS. The effective channel information may be determined based on the downlink channel information and the precoding matrix.

[0078] If method 2 is used to determine the RI, the network device configures at least two downlink pilot signal resources for the terminal device. One downlink pilot signal resource is used to send the downlink pilot signal described in step 310 above, and the other downlink pilot signal resource is used to send the beamformed downlink pilot signal described herein.

[0079] Exemplarily, the number of ports of the resources of the two downlink pilot signals may be the same or different. The number of ports of the resources corresponding to the beamformed downlink pilot signal is the value of v. Exemplarily, the resources of the two downlink pilot signals may belong to the same resource set or to different resource sets.

[0080] In addition, the above methods 1 and 2 can not only determine the rank value (i.e. determine RI, and the rank value is reported to the network device through RI), but also determine the channel quality value of the downlink channel (i.e. determine CQI, and the channel quality value is reported to the network device through CQI).

[0081] In step 340 , the terminal device reports CSI feedback information to the network device. The CSI feedback information includes at least one of codeword information and CSI.

[0082] CSI feedback information is used by a network device to transmit downlink data on a downlink channel. The CSI feedback information includes at least one of codeword information and CSI. In some embodiments, the CSI feedback information includes codeword information and CSI. In some embodiments, the CSI feedback information includes codeword information but does not include CSI.

[0083] In some embodiments, if the terminal device determines RI and CQI using method 1 above, the codeword information, RI, and CQI can be sent to the network device as CSI feedback information through a single reporting process. In other words, the terminal device reports the codeword information, RI, and CQI simultaneously.

[0084] In some embodiments, if the terminal device determines the RI and CQI using method 2 above, the codeword information is sent to the network device as CSI feedback information through one reporting process, and the RI and CQI are also sent to the network device as CSI feedback information through another reporting process. That is, the terminal device first determines the codeword information based on the first downlink pilot signal, and reports the codeword information to the network device. The network device recovers v eigenvectors based on the codeword information, and then sends v beamformed downlink pilot signals to the terminal device. The terminal device determines the RI and CQI based on the effective channel information estimated based on the v beamformed downlink pilot signals, and then reports the RI and CQI to the network device.

[0085] In some embodiments, codeword information and RI are sent to the network device as CSI feedback information through one reporting process, and CQI is also sent to the network device as CSI feedback information through another reporting process. In this manner, RI can be determined using method 1, and CQI can be determined using method 1 or method 2.

[0086] In some embodiments, if the RI is predefined by negotiation between the terminal device and the network device, then the RI does not need to be reported. If the terminal device determines the CQI using method 1 above, the codeword information and the CQI are sent to the network device as CSI feedback information through a single reporting process. If the terminal device determines the CQI using method 2 above, the codeword information is sent to the network device as CSI feedback information through a single reporting process, and the CQI is also sent to the network device as CSI feedback information through another reporting process.

[0087] In some embodiments, if the terminal device does not report CSI, the codeword information is reported to the network device as CSI feedback information. In this case, optionally, the CSI feedback information also includes interference information, and the interference information includes at least one of the following: interference information between cells, interference information between terminals. The codeword information and the interference information can be sent to the network device as CSI feedback information through a single reporting process; or, the codeword information is sent to the network device as CSI feedback information through a single reporting process, and the interference information is also sent to the network device as CSI feedback information through another reporting process, and this application does not limit the order in which the codeword information and the interference information are sent. In addition, if the terminal device does not report CSI, the terminal device may not need to perform the above-mentioned steps for determining CSI.

[0088] In some embodiments, the codeword information is reported to the network device in a broadband manner.

[0089] In some embodiments, the RI is reported to the network device in a broadband manner.

[0090] In some embodiments, the CQI is reported to the network device in a wideband or sub-band manner.

[0091] Step 350: The network device transmits downlink data on a downlink channel according to the CSI feedback information.

[0092] In some embodiments, the network device decodes the codeword information using an AI / ML model for decoding to obtain recovered downlink information. Based on the recovered downlink information, the network device determines the precoding matrix used to transmit downlink data on the downlink channel. Based on the precoding matrix, RI, and CQI, the network device transmits the downlink data on the downlink channel. For example, the network device determines the number of streams and modulation level for downlink data transmission based on the RI and CQI, and then uses the full channel information or v eigenvectors recovered by the decoder to determine the precoding for downlink data transmission.

[0093] The recovered downlink information can be the recovered downlink channel information (or recovered full channel information) or the feature vector corresponding to the recovered downlink channel information. It should be noted that the recovered downlink information and the downlink information used as input to the encoder model may not be identical because the quantization and dequantization processes, as well as the encoding and decoding processes, may prevent the data from being completely recovered.

[0094] In some embodiments, if the CSI feedback information does not include CSI, the network device can decode the codeword information through the AI / ML model used for decoding to obtain the recovered downlink information, determine the precoding matrix, RI and CQI used to transmit the downlink data on the downlink channel based on the recovered downlink information, and transmit the downlink data on the downlink channel based on the precoding matrix, RI and CQI. Optionally, the network device determines the precoding matrix, RI and CQI used to transmit the downlink data on the downlink channel based on the recovered downlink information and the interference information reported by the terminal device. For this method, a method is provided for a network device to determine the precoding matrix, RI and CQI for downlink transmission based on the codeword information and interference information reported by the terminal device. This method also takes into account information such as channel interference, and enables the network device to accurately and reasonably schedule user data transmission.

[0095] The technical solution provided in the embodiments of the present application is aimed at the CSI compression and recovery architecture of the bilateral AI / ML model based on the encoder and decoder, thereby achieving better matching of the RI / CQI reported by the terminal device to the network device with the precoding used for data transmission, so that the network device can accurately and reasonably schedule the user's data transmission.

[0096] In addition, the embodiments of the present application provide two methods for the terminal device to determine RI / CQI. When method 1 is used to determine RI / CQI, the network device only needs to send a downlink pilot signal once, and the RI / CQI can be reported to the network device together with the codeword information, and the entire signaling interaction process is relatively simple. When method 2 is used to determine RI / CQI, the RI / CQI and the codeword information are reported to the network device twice. Method 2 does not require advance configuration or pre-definition of the correspondence between the codeword length and the codebook parameters, and is more flexible. Moreover, regardless of the above method 1 or method 2, the RI / CQI reported by the terminal device to the network device can be better matched with the precoding matrix determined by the downlink information recovered by the decoder model, so that the network device can accurately and reasonably schedule the user's data transmission.

[0097] FIG4 is a flowchart of a data transmission method provided by another exemplary embodiment of the present application. This embodiment uses the method applied to the network architecture shown in FIG1 as an example. The method may include at least one of the following steps 410 to 470:

[0098] Step 410: The network device sends a downlink pilot signal to the terminal device.

[0099] In step 420 , the terminal device encodes and compresses the downlink information estimated based on the downlink pilot signal using the AI / ML model for encoding to obtain codeword information.

[0100] Step 430: The terminal device determines, based on the codeword length of the codeword information, a codebook parameter corresponding to the codeword length of the codeword information in the codebook.

[0101] Step 440: The terminal device determines RI and CQI based on the codebook parameters and downlink channel information.

[0102] Step 450: The terminal device reports codeword information, RI, and CQI to the network device.

[0103] In step 460 , the network device decodes the codeword information using the AI / ML model for decoding to obtain recovered downlink information, and determines a precoding matrix used to transmit downlink data on the downlink channel based on the recovered downlink information.

[0104] In step 470, the network device transmits downlink data on a downlink channel according to the precoding matrix, RI, and CQI.

[0105] For example, using the eType II CB codebook as an example, there are eight configurable codebook parameters for the eType II CB, as shown in Table 1. Each configuration represents the feedback overhead at a different rank and corresponds to a corresponding performance or an SGCS value.

[0106] Table 1 8 codebook parameter configurations for eType II CB

[0107]

[0108] In Table 1, paramCombination-r16 represents the eight codebook parameter configurations of eType II CB; L represents the number of spatial level vectors, i.e., the number of beams; p v represents the number of frequency-domain level vectors; v represents the rank of the codebook; β is used to determine the number of non-zero coefficients in the combination coefficient matrix.

[0109] When implementing CSI compression feedback using an encoder-based model, network equipment can also configure eight different codeword lengths (referred to as "code lengths") for different ranks. Each codeword length is quantized and used as input to the decoder model, ensuring that the decoder output is equal to the SCGS of the downlink channel's eigenvector and the SGCS calculated using an eType II CB codebook parameter configuration, or the difference between the two is less than or equal to a threshold. This establishes a one-to-one correspondence between different codeword lengths and different codebook parameters.

[0110] Assume that the network device configures the terminal device with a code length quantized to 40 bits (i.e., the length of the above-mentioned codeword information) under rank=1, and the codebook parameter configuration corresponding to this code length is the second one, i.e., the parameter configuration when paramCombination-r16 is 2 in Table 1.

[0111] The network device sends a downlink pilot signal (such as CSI-RS) to the terminal device. The terminal device estimates the downlink channel information based on the received CSI-RS and calculates the PMI based on the parameter configuration when the Rel-16 Type II codebook paramCombination-r16 is 2. Then, based on the PMI and the estimated downlink channel information, the RI and CQI values ​​are calculated. After that, the terminal device reports the RI, CQI and the codeword information compressed by the encoder to the network device. Among them, the RI and codeword information are reported in a broadband manner, while the CQI can be reported in a broadband or subband manner. The network device determines the precoding matrix for downlink data transmission based on the output information inferred by the decoder, that is, the v eigenvectors, and then combines the received RI and CQI to implement the scheduling transmission of downlink data.

[0112] In this embodiment, the terminal device uses the above-mentioned method 1 to determine RI and CQI. This method only requires the network device to send a downlink pilot signal once, and RI / CQI can be reported to the network device together with the codeword information. The entire signaling interaction process is relatively simple.

[0113] FIG5 is a flowchart of a data transmission method provided by another exemplary embodiment of the present application. This embodiment uses the method applied to the network architecture shown in FIG1 as an example. The method may include at least one of the following steps 510 to 580:

[0114] Step 510: The network device sends a downlink pilot signal to the terminal device.

[0115] In step 520 , the terminal device encodes and compresses the downlink information estimated based on the downlink pilot signal using the AI / ML model for encoding to obtain codeword information.

[0116] Step 530: The terminal device reports codeword information to the network device.

[0117] In step 540 , the network device decodes the codeword information using the AI / ML model for decoding to obtain recovered downlink information, and determines v eigenvectors included in the precoding matrix determined based on the recovered downlink information.

[0118] In step 550, the network device sends v beam-formed downlink pilot signals to the terminal device, where the v beams are the v eigenvectors mentioned above.

[0119] In step 560, the terminal device determines the RI and CQI using the effective channel information estimated based on the v beamformed downlink pilot signals.

[0120] Step 570: The terminal device reports RI and CQI to the network device.

[0121] In step 580, the network device transmits downlink data on a downlink channel according to the precoding matrix determined based on the downlink information recovered by decoding, and the RI and CQI reported by the terminal device.

[0122] Assume that the maximum number of downlink data transmission streams is 4, that is, v max =4. As shown in FIG5, the network device sends a downlink pilot signal (such as CSI-RS) to the terminal device for downlink channel information estimation. The terminal device calculates the v corresponding to the downlink channel information based on the estimated downlink channel information. max = 4 eigenvectors. The four eigenvectors are used as input to the encoder model, generating the corresponding c = 4 codewords. The terminal device quantizes these codewords and reports them to the network device. The network device dequantizes the received quantized codewords and uses the decoder model to infer four eigenvectors, denoted as B1, B2, B3, and B4.

[0123] The network device then sends a beamformed downlink pilot signal (such as CSI-RS) to the terminal device. The number of ports of the CSI-RS resource is 4, and each port corresponds to a beam. The port beams of the 4 CSI-RS resources are B1, B2, B3 and B4. The terminal device estimates the downlink effective channel information based on the received beamformed CSI-RS, calculates the rank=2 corresponding to the current channel and the broadband and / or subband CQI corresponding to rank=2, and the value of rank=2 is indicated by RI. The terminal device reports the RI and CQI to the network device, and the network device uses the eigenvector of rank=2 inferred by the decoder as the precoding matrix for downlink data transmission, and uses RI, CQI and the precoding matrix to implement the scheduling transmission of downlink data.

[0124] In this embodiment, the terminal device adopts the above-mentioned method 2 to determine RI and CQI, and the RI / CQI and codeword information are reported to the network device twice. Method 2 does not require advance configuration or pre-definition of the correspondence between codeword length and codebook parameters, and is more flexible.

[0125] It should be noted that the above steps performed by the terminal device can be independently implemented as a data transmission method on the terminal device side; the above steps performed by the network device can be independently implemented as a data transmission method on the network device side.

[0126] The following are device embodiments of the present application, which can be used to implement the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

[0127] Figure 6 is a block diagram of a data transmission device provided by an exemplary embodiment of the present application. This device has the functionality to implement the aforementioned exemplary method on the terminal device side. This functionality can be implemented in hardware or by hardware executing corresponding software implementations. This device can be the terminal device described above, or it can be provided within a terminal device. As shown in Figure 6, the device 600 may include a receiving module 610, a processing module 620, and a sending module 630.

[0128] The receiving module 610 is configured to receive a downlink pilot signal sent by a network device.

[0129] The processing module 620 is used to encode and compress the downlink information estimated based on the downlink pilot signal through the AI / ML model to obtain codeword information; and determine the channel state information CSI of the downlink channel, where the CSI includes RI and / or CQI.

[0130] The sending module 630 is configured to report CSI feedback information to a network device, where the CSI feedback information includes at least one of the codeword information and the CSI.

[0131] In some embodiments, the processing module 620 is configured to: determine, according to the codeword length of the codeword information, a codebook parameter corresponding to the codeword length of the codeword information in a codebook; and determine the CSI according to the codebook parameter and the downlink information.

[0132] In some embodiments, the processing module 620 is configured to: determine the PMI according to the codebook parameters; and determine the CSI according to the PMI and the downlink information.

[0133] In some embodiments, the codeword length of the codeword information is configured by the network device, or determined by the terminal device and reported to the network device.

[0134] In some embodiments, the correspondence between the codeword length of the codeword information and the codebook parameters in the codebook is configured by the network device, or is predefined by negotiation between the terminal device and the network device.

[0135] In some embodiments, for a set of corresponding codeword lengths and codebook parameters, a first SGCS calculated based on the codeword information corresponding to the codeword length and a second SGCS calculated based on the codebook parameters meet a condition; wherein, the first SGCS is an SGCS calculated based on the precoding matrix determined according to the recovered downlink information and the eigenvector of the downlink channel, and the recovered downlink information is obtained by decoding the codeword information corresponding to the codeword length through the AI / ML model for decoding, and the second SGCS is an SGCS calculated based on the precoding matrix determined by the codebook parameters and the eigenvector of the downlink channel; the condition includes one of the following: the first SGCS and the second SGCS are the same; or, the difference between the first SGCS and the second SGCS is less than or equal to a threshold value.

[0136] In some embodiments, the codeword length of the codeword information is: the codeword length of the encoded and compressed CSI output by the AI / ML model used for encoding; or, the codeword length of the binary bit stream obtained by quantizing the encoded and compressed CSI output by the AI / ML model used for encoding.

[0137] In some embodiments, the processing module 620 is configured to: determine at least one eigenvector according to the downlink information; and calculate SINRs at different rank values ​​according to the at least one eigenvector to determine the RI.

[0138] In some embodiments, the processing module 620 is used to: receive v beam-formed downlink pilot signals sent by the network device, wherein the v beams are v vectors contained in a precoding matrix determined based on the recovered downlink information, and the recovered downlink information is obtained by decoding the codeword information using an AI / ML model for decoding, and v is a positive integer; determine the CSI by estimating the effective channel information based on the v beam-formed downlink pilot signals.

[0139] In some embodiments, the network device configures at least two downlink pilot signal resources for the terminal device.

[0140] In some embodiments, the number of ports of the resources of the two downlink pilot signals is the same or different; and / or the resources of the two downlink pilot signals belong to the same or different resource sets.

[0141] In some embodiments, the codeword information, the RI, and the CQI are sent to the network device as the CSI feedback information through a single reporting process; or, the codeword information is sent to the network device as the CSI feedback information through a single reporting process, and the RI and the CQI are also sent to the network device as the CSI feedback information through another reporting process; or, the codeword information and the RI are sent to the network device as the CSI feedback information through a single reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, the codeword information and the CQI are sent to the network device as the CSI feedback information through a single reporting process; or, the codeword information is sent to the network device as the CSI feedback information through a single reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, the codeword information is reported to the network device as the CSI feedback information, and the CSI feedback information further includes interference information, and the interference information includes at least one of the following: interference information between cells and interference information between terminals.

[0142] In some embodiments, the downlink pilot signal is a CSI-RS.

[0143] In some embodiments, the codeword information is reported to the network device in a broadband manner.

[0144] In some embodiments, the RI is reported to the network device in a wideband manner, and the CQI is reported to the network device in a wideband or sub-band manner.

[0145] In some embodiments, the RI is predefined by negotiation between the terminal device and the network device.

[0146] Figure 7 is a block diagram of a data transmission device provided by another exemplary embodiment of the present application. This device has the functionality to implement the aforementioned network device-side method example. This functionality can be implemented in hardware or by hardware executing corresponding software implementations. This device can be the network device described above or can be provided within a network device. As shown in Figure 7, this device 700 may include a sending module 710, a receiving module 720, and a processing module 730.

[0147] The sending module 710 is used to send a downlink pilot signal to the terminal device.

[0148] The receiving module 720 is used to receive CSI feedback information reported by the terminal device, where the CSI feedback information includes at least one of codeword information and CSI of the downlink channel. The codeword information is obtained by the terminal device by encoding and compressing downlink information estimated based on the downlink pilot signal through an AI / ML model, and the CSI includes RI and / or CQI.

[0149] The processing module 730 is configured to transmit downlink data on the downlink channel according to the CSI feedback information.

[0150] In some embodiments, the processing module 730 is used to: decode the codeword information through the AI / ML model for decoding to obtain recovered downlink information; determine the precoding matrix used to transmit downlink data on the downlink channel based on the recovered downlink information; and transmit downlink data on the downlink channel based on the precoding matrix and the CSI.

[0151] In some embodiments, the sending module 710 is further used to: send first configuration information and / or second configuration information to the terminal device, where the first configuration information is used to configure the codeword length of the codeword information, and the second configuration information is used to configure the correspondence between the codeword length of the codeword information and the codebook parameters in the codebook; wherein the second configuration information is used by the terminal device to determine the codebook parameters in the codebook corresponding to the codeword length of the codeword information, and determine the CSI based on the codebook parameters.

[0152] In some embodiments, for a set of corresponding codeword lengths and codebook parameters, a first SGCS calculated based on the codeword information corresponding to the codeword length and a second SGCS calculated based on the codebook parameters meet a condition; wherein, the first SGCS is an SGCS calculated based on the precoding matrix determined according to the recovered downlink information and the eigenvector of the downlink channel, and the recovered downlink information is obtained by decoding the codeword information corresponding to the codeword length through the AI / ML model for decoding, and the second SGCS is an SGCS calculated based on the precoding matrix determined by the codebook parameters and the eigenvector of the downlink channel; the condition includes one of the following: the first SGCS and the second SGCS are the same; or, the difference between the first SGCS and the second SGCS is less than or equal to a threshold value.

[0153] In some embodiments, the codeword length of the codeword information is: the codeword length of the encoded and compressed CSI output by the AI / ML model used for encoding; or, the codeword length of the binary bit stream obtained by quantizing the encoded and compressed CSI output by the AI / ML model used for encoding.

[0154] In some embodiments, the sending module 710 is further used to: after receiving the codeword information reported by the terminal device, send v beam-formed downlink pilot signals to the terminal device, wherein the v beams are v vectors contained in the precoding matrix determined according to the recovered downlink information, and the recovered downlink information is obtained by decoding the codeword information through the AI / ML model for decoding, and v is a positive integer; wherein the CSI is determined by the effective channel information estimated by the terminal device based on the v beam-formed downlink pilot signals.

[0155] In some embodiments, the network device configures at least two downlink pilot signal resources for the terminal device.

[0156] In some embodiments, the number of ports of the resources of the two downlink pilot signals is the same or different; and / or the resources of the two downlink pilot signals belong to the same or different resource sets.

[0157] In some embodiments, the codeword information, the RI, and the CQI are sent to the network device as the CSI feedback information through a single reporting process; or, the codeword information is sent to the network device as the CSI feedback information through a single reporting process, and the RI and the CQI are also sent to the network device as the CSI feedback information through another reporting process; or, the codeword information and the RI are sent to the network device as the CSI feedback information through a single reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, the codeword information and the CQI are sent to the network device as the CSI feedback information through a single reporting process; or, the codeword information is sent to the network device as the CSI feedback information through a single reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, the codeword information is reported to the network device as the CSI feedback information, and the CSI feedback information further includes interference information, and the interference information includes at least one of the following: interference information between cells and interference information between terminals.

[0158] In some embodiments, the downlink pilot signal is a CSI-RS.

[0159] In some embodiments, the codeword information is reported to the network device in a broadband manner.

[0160] In some embodiments, the RI is reported to the network device in a wideband manner, and the CQI is reported to the network device in a wideband or sub-band manner.

[0161] In some embodiments, the RI is predefined by negotiation between the terminal device and the network device.

[0162] It should be noted that, when the device provided in the above embodiment realizes its function, it only uses the division of the above-mentioned functional modules as an example. In actual application, the above-mentioned functions can be assigned to different functional modules according to actual needs, that is, the content structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0163] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here. For details not described in detail in the embodiment of the device, reference can be made to the above method embodiment.

[0164] Figure 8 is a block diagram of a communication device 800 according to an exemplary embodiment. The communication device 800 may be the aforementioned terminal device or network device. The communication device 800 may include a processor 801, a receiver 802, a transmitter 803, a memory 804, and a bus 805. The processor 801 may be used to implement the function of the processing model in the above-mentioned apparatus embodiment, the receiver 802 may be used to implement the function of the receiving module in the above-mentioned apparatus embodiment, and the transmitter 803 may be used to implement the function of the sending module in the above-mentioned apparatus embodiment.

[0165] The processor 801 includes one or more processing cores. The processor 801 executes various functional applications and information processing by running software programs and modules.

[0166] The receiver 802 and the transmitter 803 may be implemented as a communication component, which may be a communication chip.

[0167] The memory 804 is connected to the processor 801 via a bus 805 .

[0168] The memory 804 may be used to store a computer program, and the processor 801 may be used to execute the computer program to implement the above-mentioned data transmission method on the terminal device side or the data transmission method on the network device side.

[0169] In addition, the memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof. Volatile or non-volatile storage devices include but are not limited to: magnetic disks or optical disks, electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), static random-access memory (SRAM), read-only memory (ROM), magnetic memory, flash memory, and programmable read-only memory (PROM).

[0170] An embodiment of the present application also provides a computer-readable storage medium, in which a computer program is stored, and the computer program is used to be executed by a processor to implement the above-mentioned data transmission method on the terminal device side, or the data transmission method on the network device side. Optionally, the computer-readable storage medium may include: ROM (Read-Only Memory), RAM (Random-Access Memory), SSD (Solid State Drives) or optical disks, etc. Among them, random access memory may include ReRAM (Resistance Random Access Memory) and DRAM (Dynamic Random Access Memory).

[0171] An embodiment of the present application also provides a communication system, which includes a terminal device and a network device. The terminal device is used to implement the above-mentioned data transmission method on the terminal device side, and the network device is used to implement the above-mentioned data transmission method on the network device side.

[0172] An embodiment of the present application also provides a chip, which includes a programmable logic circuit and / or program instructions. When the chip is running, it is used to implement the above-mentioned data transmission method on the terminal device side, or the data transmission method on the network device side.

[0173] An embodiment of the present application also provides a computer program product, which includes computer instructions, which are stored in a computer-readable storage medium. A processor reads and executes the computer instructions from the computer-readable storage medium to implement the above-mentioned data transmission method on the terminal device side, or the data transmission method on the network device side.

[0174] It should be understood that the term "indication" in the embodiments of this application can refer to direct indications, indirect indications, or an association relationship. For example, "A indicates B" can mean that A directly indicates B, for example, B can be obtained through A; it can also mean that A indirectly indicates B, for example, A indicates C, and B can be obtained through C; it can also mean that A and B are associated with each other.

[0175] In the description of the embodiments of the present application, the term "corresponding" may indicate a direct or indirect correspondence between the two, or an association relationship between the two, or a relationship between indication and being indicated, configuration and being configured, etc.

[0176] In some embodiments of the present application, "predefined" may be implemented by pre-storing corresponding codes, tables, or other methods that can be used to indicate relevant information in a device (e.g., including a terminal device and a network device), and the present application does not limit the specific implementation method. For example, predefined may refer to information defined in a protocol.

[0177] In some embodiments of the present application, the "protocol" may refer to a standard protocol in the field of communications, for example, it may include an LTE protocol, a NR protocol, and related protocols used in future communication systems, and this application does not limit this.

[0178] In this document, "plurality" refers to two or more. "And / or" describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. The character " / " generally indicates an "or" relationship between the associated objects.

[0179] The term “greater than or equal to” mentioned herein may mean greater than or equal to, or greater than, and the term “less than or equal to” may mean less than or equal to, or less than.

[0180] In addition, the step numbers described in this document only illustrate a possible execution order between the steps. In some other embodiments, the above steps may not be executed in the order of the numbers, such as two steps with different numbers are executed at the same time, or two steps with different numbers are executed in the opposite order of the diagram. The embodiments of the present application are not limited to this.

[0181] In addition, the various embodiments provided in this application can be arbitrarily combined to form new embodiments, which are all within the scope of protection of this application.

[0182] Those skilled in the art will appreciate that in one or more of the above examples, the functions described in the embodiments of the present application can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any media that facilitates the transmission of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0183] The above description is merely an exemplary embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A data transmission method, characterized in that: The method is executed by a terminal device, and includes: receiving a downlink pilot signal sent by a network device; Encoding and compressing downlink information estimated based on the downlink pilot signal using an artificial intelligence / machine learning AI / ML model to obtain codeword information; Determine channel state information (CSI) of a downlink channel, where the CSI includes a rank indicator (RI) and / or a channel quality indicator (CQI); Report CSI feedback information to a network device, where the CSI feedback information includes at least one of the codeword information and the CSI.

2. The method according to claim 1, characterized in that The determining of the CSI of the downlink channel includes: Determining, according to the codeword length of the codeword information, a codebook parameter corresponding to the codeword length of the codeword information in a codebook; The CSI is determined according to the codebook parameter and the downlink information.

3. The method according to claim 2, characterized in that The determining the CSI according to the codebook parameter and the downlink information includes: Determine a precoding matrix indicator PMI according to the codebook parameters; The CSI is determined according to the PMI and the downlink information.

4. The method according to claim 2 or 3, characterized in that For a set of corresponding codeword lengths and codebook parameters, a first regular cosine similarity squared SGCS calculated based on the codeword information corresponding to the codeword length and a second SGCS calculated based on the codebook parameters meet a condition; The first SGCS is an SGCS calculated based on a precoding matrix determined by the recovered downlink information and an eigenvector of the downlink channel, the recovered downlink information is obtained by decoding codeword information corresponding to the codeword length using an AI / ML model for decoding, and the second SGCS is an SGCS calculated based on a precoding matrix determined by the codebook parameters and the eigenvector of the downlink channel; The conditions include one of the following: The first SGCS and the second SGCS are the same; or, A difference between the first SGCS and the second SGCS is less than or equal to a threshold.

5. The method according to any one of claims 2 to 4, characterized in that The codeword length of the codeword information is: The codeword length of the encoded and compressed CSI output by the AI / ML model used for encoding; or The codeword length of the binary bit stream obtained by quantizing the coded and compressed CSI output by the AI / ML model used for encoding.

6. The method according to claim 1, characterized in that The determining of the CSI of the downlink channel includes: determining at least one characteristic vector according to the downlink information; The signal to interference plus noise ratio (SINR) at different rank values ​​is calculated according to the at least one eigenvector to determine the RI.

7. The method according to claim 1, characterized in that The determining of the CSI of the downlink channel includes: receiving v beamformed downlink pilot signals sent by the network device, where the v beams are v vectors contained in a precoding matrix determined based on recovered downlink information, the recovered downlink information is obtained by decoding the codeword information using an AI / ML model for decoding, and v is a positive integer; The CSI is determined by using effective channel information estimated based on the v beamformed downlink pilot signals.

8. The method according to claim 7, characterized in that The network device configures at least two downlink pilot signal resources for the terminal device.

9. The method according to claim 1, characterized in that The codeword information, the RI, and the CQI are sent to the network device through a single reporting process as the CSI feedback information; or, The codeword information is sent to the network device as the CSI feedback information through one reporting process, and the RI and the CQI are also sent to the network device as the CSI feedback information through another reporting process; or, The codeword information and the RI are sent to the network device as the CSI feedback information through one reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, The codeword information and the CQI are sent to the network device as the CSI feedback information through a single reporting process; or, The codeword information is sent to the network device as the CSI feedback information through one reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, The codeword information is reported to the network device as the CSI feedback information, and the CSI feedback information further includes at least one of interference information between cells and interference information between terminals.

10. The method according to any one of claims 1 to 9, characterized in that The downlink pilot signal is a channel state information reference signal CSI-RS.

11. The method according to any one of claims 1 to 10, characterized in that The codeword information is reported to the network device in a broadband manner.

12. The method according to any one of claims 1 to 11, characterized in that The RI is reported to the network device in a broadband manner, and the CQI is reported to the network device in a broadband or sub-band manner.

13. A data transmission method, characterized in that: The method is performed by a network device, and includes: Sending a downlink pilot signal to a terminal device; receiving CSI feedback information reported by the terminal device, where the CSI feedback information includes at least one of codeword information and CSI of a downlink channel, the codeword information being obtained by the terminal device by encoding and compressing downlink information estimated based on the downlink pilot signal through an artificial intelligence / machine learning AI / ML model, and the CSI including a rank indication RI and / or a channel quality indication CQI; Transmit downlink data on the downlink channel according to the CSI feedback information.

14. The method according to claim 13, characterized in that The transmitting downlink data on the downlink channel according to the CSI feedback information includes: Decoding the codeword information using an AI / ML model for decoding to obtain recovered downlink information; determining, based on the recovered downlink information, a precoding matrix used for transmitting downlink data on the downlink channel; Downlink data is transmitted on the downlink channel according to the precoding matrix and the CSI.

15. The method according to claim 13 or 14, characterized in that The method further comprises: Sending first configuration information and / or second configuration information to the terminal device, where the first configuration information is used to configure a codeword length of the codeword information, and the second configuration information is used to configure a correspondence between the codeword length of the codeword information and a codebook parameter in a codebook; The second configuration information is used by the terminal device to determine a codebook parameter corresponding to the codeword length of the codeword information in the codebook, and to determine the CSI according to the codebook parameter.

16. The method according to claim 15, characterized in that For a set of corresponding codeword lengths and codebook parameters, a first regular cosine similarity squared SGCS calculated based on the codeword information corresponding to the codeword length and a second SGCS calculated based on the codebook parameters meet a condition; The first SGCS is an SGCS calculated based on a precoding matrix determined by the recovered downlink information and an eigenvector of the downlink channel, the recovered downlink information is obtained by decoding codeword information corresponding to the codeword length using an AI / ML model for decoding, and the second SGCS is an SGCS calculated based on a precoding matrix determined by the codebook parameters and the eigenvector of the downlink channel; The conditions include one of the following: The first SGCS and the second SGCS are the same; or, A difference between the first SGCS and the second SGCS is less than or equal to a threshold.

17. The method according to claim 15 or 16, characterized in that The codeword length of the codeword information is: The codeword length of the encoded and compressed CSI output by the AI / ML model used for encoding; or The codeword length of the binary bit stream obtained by quantizing the coded and compressed CSI output by the AI / ML model used for encoding.

18. The method according to claim 13 or 14, characterized in that The method further comprises: After receiving the codeword information reported by the terminal device, sending v beamformed downlink pilot signals to the terminal device, where the v beams are v vectors contained in a precoding matrix determined according to the recovered downlink information, the recovered downlink information is obtained by decoding the codeword information using an AI / ML model for decoding, and v is a positive integer; The CSI is determined by the effective channel information estimated by the terminal device based on the v beam-forming downlink pilot signals.

19. The method according to claim 18, characterized in that The network device configures at least two downlink pilot signal resources for the terminal device.

20. The method according to claim 13, wherein The codeword information, the RI, and the CQI are sent to the network device through a single reporting process as the CSI feedback information; or, The codeword information is sent to the network device as the CSI feedback information through one reporting process, and the RI and the CQI are also sent to the network device as the CSI feedback information through another reporting process; or, The codeword information and the RI are sent to the network device as the CSI feedback information through one reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, The codeword information and the CQI are sent to the network device as the CSI feedback information through a single reporting process; or, The codeword information is sent to the network device as the CSI feedback information through one reporting process, and the CQI is also sent to the network device as the CSI feedback information through another reporting process; or, The codeword information is reported to the network device as the CSI feedback information, and the CSI feedback information further includes at least one of interference information between cells and interference information between terminals.

21. The method according to any one of claims 13 to 20, characterized in that The downlink pilot signal is a channel state information reference signal CSI-RS.

22. The method according to any one of claims 13 to 21, characterized in that The codeword information is reported to the network device in a broadband manner.

23. The method according to any one of claims 13 to 22, characterized in that The RI is reported to the network device in a broadband manner, and the CQI is reported to the network device in a broadband or sub-band manner.

24. A data transmission device, characterized in that: The device comprises: A receiving module, configured to receive a downlink pilot signal sent by a network device; a processing module, configured to encode and compress downlink information estimated based on the downlink pilot signal through an artificial intelligence / machine learning AI / ML model to obtain codeword information; determine channel state information CSI of the downlink channel, the CSI including a rank indicator RI and / or a channel quality indicator CQI; The sending module is configured to report CSI feedback information to a network device, where the CSI feedback information includes at least one of the codeword information and the CSI.

25. A data transmission device, characterized in that: The device comprises: A sending module, configured to send a downlink pilot signal to a terminal device; a receiving module, configured to receive CSI feedback information reported by the terminal device, the CSI feedback information including at least one of codeword information and CSI of a downlink channel, the codeword information being obtained by the terminal device by encoding and compressing downlink information estimated based on the downlink pilot signal through an artificial intelligence / machine learning AI / ML model, the CSI including a rank indication RI and / or a channel quality indication CQI; A processing module is configured to transmit downlink data on the downlink channel according to the CSI feedback information.

26. A terminal device, characterized in that: The terminal device includes a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the computer program to implement the method according to any one of claims 1 to 12.

27. A network device, characterized in that: The network device includes a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the computer program to implement the method according to any one of claims 13 to 23.

28. A computer-readable storage medium, characterized in that The storage medium stores a computer program, which is configured to be executed by a processor to implement the method according to any one of claims 1 to 12, or the method according to any one of claims 13 to 23.

29. A communication system, characterized in that: The communication system includes a terminal device and a network device, the terminal device is used to implement the method according to any one of claims 1 to 12, and the network device is used to implement the method according to any one of claims 13 to 23.

Images