Data set transmission methods, reception methods, communication devices and storage media

Abstract

An embodiment of the present application provides a data set transmitting method, a receiving method, a communication device, and a storage medium, wherein the data set transmitting method is used in a first communication device and includes: obtaining a first data set including L0 samples; obtaining a second data set including L1 samples from the first data set based on a first sample parameter and / or a second sample parameter corresponding to at least one sample; and transmitting the second data set, where L0 and L1 are positive integers, and L1 is less than or equal to L0.

Description

[Technical Field]

[0001] This application is based on and claims priority from a Chinese patent application bearing application number 202310401595.0 and filing date April 7, 2023, the entire contents of which are incorporated herein by reference.

[0002] The present application relates to communication technology, and in particular to a method for transmitting and receiving a data set, a communication device and a storage medium. [Background technology]

[0003] Wireless communication systems are widely applied in people's daily lives and production. For example, wireless communication systems are used for video transmission, voice calls, wireless positioning, communication between devices in factories, and communication between vehicles and other devices in connected cars. With technological development and higher demands on daily life and production, increasingly higher requirements are being placed on the transmission reliability, capacity, and transmission speed of wireless communication technologies, which is driving the rapid development of the wireless communication field. Currently, various advanced technologies such as multi-antenna technology, artificial intelligence technology, and intelligent surface technology can be applied to wireless communication systems or are expected to be applied in the future.

[0004] Artificial intelligence (AI) technology is a field of computer science that includes devices, components, and functional modules with self-learning capabilities. Currently, AI technology can be widely applied to wireless communication systems, such as AI-based channel estimation, AI-based channel state information feedback, AI-based positioning, AI-based spatial and / or frequency domain beam prediction, AI-based channel information prediction, and AI-based channel coding. In various advanced data-driven or data-plus-model-driven technologies, including AI, the acquisition and transmission of high-quality or tailored data sets is crucial and directly determines the application effectiveness of the technologies. Therefore, how to acquire and transmit data sets containing multiple high-quality samples is currently an urgent problem that needs to be solved. Summary of the Invention [Problem to be solved by the invention]

[0005] The present application provides a method for transmitting and receiving a data set, a communication device and a storage medium, which can obtain a data set of high quality or meeting requirements. [Means for solving the problem]

[0006] According to a first aspect, an embodiment of the present application provides a data set transmission method for use in a first communication device, the transmission method comprising: obtaining a first data set including L0 samples; obtaining a second data set including L1 samples from the first data set based on first sample parameters and / or second sample parameters; and transmitting the second data set, where L0 and L1 are positive integers, and L1 is less than or equal to L0.

[0007] According to a second aspect, an embodiment of the present application provides a method for receiving a data set for use in a second communication device, the receiving method comprising: receiving a second data set comprising L1 samples, the second data set being derived by the first communication device from the first data set comprising L0 samples based on first sample parameters and / or second sample parameters, where L0 and L1 are positive integers and L1 is less than or equal to L0.

[0008] According to a third aspect, an embodiment of the present application provides a communication device, the communication device comprising: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the method for transmitting a dataset according to the first aspect or the method for receiving a dataset according to the second aspect is realized.

[0009] According to a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having stored thereon a processor-executable program, the processor-executable program being used, when executed by the processor, to implement the method for transmitting a dataset according to the first aspect or the method for receiving a dataset according to the second aspect.

[0010] According to a fifth aspect, an embodiment of the present application provides a computer program product having stored thereon program instructions which, when executed by a computer, cause the computer to perform the method for transmitting a dataset of the first aspect or the method for receiving a dataset of the second aspect. [Effects of the Invention]

[0011] In an embodiment of the present application, a first dataset including L0 samples can be obtained, a second dataset including L1 samples can be obtained from the first dataset based on the first sample parameters and / or the second sample parameters, and the second dataset can be sent, where L0 and L1 are positive integers, and L1 is less than or equal to L0, that is, a high-quality or suitable second dataset can be screened from the first dataset based on the first sample parameters and / or the second sample parameters, thereby improving the model training effect and further improving the model performance. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a schematic diagram of the architecture of a communication system according to an embodiment of the present application; [Figure 2] 1 is a flowchart of a method for transmitting a data set according to an embodiment of the present application; [Figure 3] FIG. 2 is a schematic diagram of obtaining a second data set from a first data set according to an embodiment of the present application; [Figure 4] FIG. 1 is a schematic diagram showing that not all samples within a measurement window are reported according to an embodiment of the present application. [Figure 5] FIG. 1 is a schematic diagram of determining samples to be reported or discarded based on frequency domain constraint relationships according to an embodiment of the present application; [Figure 6] FIG. 10 is a schematic diagram of feedback when samples are discarded or reported by bitmap according to an embodiment of the present application; [Figure 7] 1 is a flowchart of a method for receiving a data set according to an embodiment of the present application; [Figure 8] 1 is a schematic diagram of the structure of a communication device according to an embodiment of the present application; DETAILED DESCRIPTION OF THE INVENTION

[0013] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be described in more detail below with reference to the drawings and examples. The specific examples described herein are only for the purpose of interpreting the present application, and are not intended to limit the present application.

[0014] In the description of the embodiments of this application, when terms such as "first," "second," etc. are used, they are merely intended to distinguish technical features and should not be understood as indicating or implying relative importance, implicitly indicating the number of the indicated technical features, or implicitly indicating the priority relationship of the indicated technical features. "At least one" means one or more, and "multiple" means two or more. "And / or" describes a relationship between related objects and indicates that a three-way relationship can exist. For example, "A and / or B" may indicate that A exists alone, that A and B exist simultaneously, and that B exists alone. Here, A and B may be singular or plural. The " / " character typically indicates that the context object is in an "or" relationship. "At least one of the following" and similar expressions refer to any group of these terms, including any group of single or multiple terms. For example, at least one of a, b, and c may represent a, b, c, a and b, a and c, b and c, or a, b, and c, where a, b, and c may be singular or plural.

[0015] The technical features of each embodiment of the present application described below may be combined with each other if they do not conflict with each other.

[0016] In one embodiment, artificial intelligence (AI) includes devices, components, software, modules, models, functional modules, functional functions, etc. that have self-learning capabilities such as machine learning (ML), deep learning, reinforcement learning, transfer learning, deep reinforcement learning, and meta-learning.

[0017] In one embodiment, the artificial intelligence may be realized by an artificial intelligence network (also called a neural network), which includes multiple layers, each layer including at least one node. For example, the neural network may include an input layer, an output layer, and at least one hidden layer. Each layer of the neural network may include, but is not limited to, at least one of a fully connected layer, a dense layer, a convolutional layer, a transposed convolutional layer, a direct connection layer, an activation function, a normalization layer, and a pooling layer.

[0018] In one embodiment, each layer of a neural network may include one sub-neural network, such as a residual network block (Resnet block), a dense network (Densenet block), or a recurrent neural network (RNN). The artificial intelligence network can be realized by a model, where the model refers to a data stream passing through multiple linear or nonlinear components from the original input of the sample to the output destination. Here, the model includes a neural network model, a non-artificial intelligence module for processing information or a corresponding model thereof, and a functional component or function that maps input information to output information (where mapping includes linear mapping and nonlinear mapping). Furthermore, the model includes a model structure and model parameters. For example, if the model is a neural network model, the neural network model includes a neural network model structure and / or neural network model parameters. The neural network model structure is used to describe the structure of the neural network, and the neural network model parameters are used to describe the values ​​of the parameters of the neural network. One neural network model structure may correspond to multiple neural network model parameters, i.e., the neural network model structures may be the same, but the values ​​of the corresponding neural network model parameters may be different.

[0019] Furthermore, the neural network model structure may be abbreviated as "model structure," and the neural network model parameters may be abbreviated as "network parameters" or "model parameters." A model structure defines the architecture of a network, such as the number of neural network layers, the size of each layer, the activation function, the link status, the convolution kernel and size, the convolution step, and the convolution type (e.g., 1D convolution, 2D convolution, 3D convolution, hollow convolution, transposed convolution, separable convolution, packet convolution, and dilated convolution), while the network parameters are the network weights and / or offsets and their values ​​of each layer in the neural network model. A model structure may correspond to multiple different neural network model parameter values, thereby enabling adaptation to different scenes. The neural network model parameters may be obtained through online or offline training. For example, the neural network model parameters may be obtained by inputting at least one sample, a label, and a trained neural network model.

[0020] However, as artificial intelligence technology continues to develop and advance, its application to wireless communication systems has become one of the current research hotspots. Here, artificial intelligence typically uses models to perform prediction tasks. Data features are input into the model and processed through multiple linear or nonlinear components to map to an output target. The quality of a model is directly related to the performance of the artificial intelligence. However, to obtain a model that meets the requirements, training using a large number of samples is required. One sample typically consists of N features and M labels, where N is a positive integer and M is an integer greater than or equal to 0. Multiple samples constitute a dataset. To demonstrate the value of artificial intelligence in wireless communication systems, it is important to obtain high-quality or meet the requirements. Therefore, how to obtain high-quality or meet the requirements is currently a technical issue that requires research and solution.

[0021] Each model corresponds to one Model Indicator (Model ID) or model Identity (Model ID), where the model identity may also be called a model index, a first identity, a functional identity, or a model indicator, etc.

[0022] Based on this, the embodiments of the present application provide a data set transmission method, a data set reception method, a communication device and a storage medium, which can obtain a high quality or required data set.

[0023] The data set transmitting method and data set receiving method according to the embodiments of the present application may be used in various communication systems of communication networks such as 2G (2nd Generation, second generation mobile communication technology) networks, 3G (3rd Generation, third generation mobile communication technology) networks, 4G (4th Generation, fourth generation mobile communication technology) networks, 5G (5th Generation, fifth generation mobile communication technology) networks or future mobile communication networks, for example, Global System for Mobile Communications (GSM) or any other second generation cellular communication system, or Universal Mobile Telecommunications System (UMTS) based on basic Wideband Code Division Multiple Access (W-CDMA), High-Speed ​​Packet Access (HSPA) systems, Long Term Evolution (LTE) systems, advanced LTE systems, systems based on the IEEE 802.11 standard, IEEE 802.12 standard, IEEE 802.13 standard, IEEE 802.14 standard, IEEE 802.15 standard, IEEE 802.16 standard, IEEE 802.17 standard, IEEE 802.18 standard, IEEE 802.19 ... The present invention may be applied to at least one of the following systems: a system based on the 802.15 standard and / or a fifth generation (5G) mobile or cellular communication system, or a future mobile communication system. However, the embodiments are not limited to the systems given in the above examples, but a person skilled in the art can apply the solution to other communication systems having the required attributes.

[0024] 1, which is a schematic diagram of a communication system architecture according to an embodiment of the present application, the communication system 100 includes a plurality of communication devices, and can perform wireless communication between different communication devices using air interface resources. Specifically, the communication system 100 includes at least one network side device and at least one receiving side device, and the at least one network side device and / or the at least one receiving side device can connect with a third party device, for example, by a method such as a wireless signal, and can transmit a data set and / or a model to the third party device, where the model may be provided in a base station and / or a terminal.

[0025] Specifically, as shown in FIG. 1, a communication system 100 includes a network side device 110, a first receiving device 120, a second receiving device 121, and a third receiving device 122, and communication is established between the network side device 110, the first receiving device 120, the second receiving device 121, and the third receiving device 122.

[0026] Here, wireless communication between communication devices includes reference signal transmission or data transmission between communication devices, where transmission includes sending or receiving. When transmitting a reference signal between communication devices, the receiving device (i.e., the first communication device) is the device that receives the reference signal, and the network side device (i.e., the second communication device) is the device that transmits the reference signal.

[0027] In the downlink, the first communication device may be a network side device, and the second communication device may be a receiving device. In the uplink, the first communication device may be a receiving device, and the second communication device may be a network side device. When the communication method of the two communication devices is communication between devices, the first communication device and the second communication device may both be network side devices, or both may be receiving devices.

[0028] Furthermore, the first communication device may be referred to as a first communication node or a first node, and the second communication device may be referred to as a second communication node or a second node, and there are no specific limitations here.

[0029] The network side equipment includes, but is not limited to, a base station, and the base station may be a base station in Long Term Evolution (LTE), an evolved base station (Evolutional Node B, eNB or eNodeB) in Long Term Evolution advanced (LTEA), a base station in a 5G network, or a base station in a future communication system, and the base station may include various network side equipment such as various macro base stations, micro base stations, home base stations, WLANs, routers, Reconfigurable Intelligent Surfaces (RISs), Wireless Fidelity (WIFI) equipment, or primary and secondary cells, and may be a location management function (LMF) equipment, but the embodiments of the present application are not limited thereto.

[0030] The receiving device includes, but is not limited to, a terminal, where the terminal is a device having wireless transmission and reception capabilities, and may be deployed on land, including indoors or outdoors, handheld, worn, or vehicle-mounted, or on water (e.g., a steamship, etc.), or in the air (e.g., an airplane, balloon, satellite, etc.). The terminal may be a mobile phone, a tablet PC, a computer with wireless transmission and reception capabilities, a virtual reality (VR) terminal, an augmented reality (AR) terminal, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in remote medical, a wireless terminal in a smart grid, a wireless terminal in transportation safety, a wireless terminal in a smart city, or a wireless terminal in a smart home, etc. The embodiments of the present application do not limit application scenarios. A terminal may also be referred to as a user, User Equipment (UE), access terminal, UE unit, UE station, mobile station, mobile station, distant station, remote terminal, mobile equipment, UE terminal, wireless communication equipment, UE agent, or UE device, and the embodiments of this application are not limited in this respect.

[0031] The above-described communication system is intended to more clearly explain the technical solutions of the embodiments of the present application, and does not constitute a limitation on the technical solutions provided by the embodiments of the present application. With the evolution of system architecture and the emergence of new business scenarios, the technical solutions provided by the embodiments of the present application can be similarly applied to similar communication systems.

[0032] In order to better transmit data or signals, the first communication device needs to obtain measurement parameters. Here, the measurement parameters for the reference signal of the first communication device may include channel state information or other parameters for describing the channel, where the channel state information includes a Channel State Information-Reference Signal Resource Indicator (CSI-RS Resource Indicator, CRI), a Synchronization Signals Block Resource Indicator (SSBRI), a Layer 1 Reference Signal Received Power (L1-RSRP or RSRP), a Differential RSRP, a Layer 1 Reference Signal Signal-to-Interference Noise Ratio (L1-SINR or SINR), a Differential L1-SINR, a Reference Signal Received Quality (RSRQ), a Channel Quality Indicator (CQI), a Precoding Matrix Indicator (PMI), a Layer Indicator (LI), a Rank Indicator (RI), and precoding information.

[0033] The precoding information includes a first type of precoding information, which is precoding information configured based on a conventional channel feature matrix or a quantized value of the feature matrix, for example, precoding information realized based on a codebook method, and the precoding matrix indicator is one in the codebook-based precoding information. The precoding information further includes precoding information realized based on a non-codebook method, for example, a second type of precoding information.

[0034] In one embodiment, channel state information corresponding to channel information can be transmitted between a first communication device (e.g., a terminal) and a second communication device (e.g., a base station) through first-class precoding information. For example, if the first-class precoding information is precoding information realized based on a codebook method, the codebook may be a codebook for N antennas in LTE, where N=2, 4, 8, 12, 16, 24, 32, etc., or may be a type I codebook, type II codebook, type II port selection codebook, enhanced type II codebook, enhanced type II selection codebook, or further enhanced type II selection codebook in NR (New Radio), where the codebook includes L codewords, and the main idea is that the base station and the terminal store the L codewords based on a predetermined formula, table, or dictionary, where L is an integer greater than 1 and is generally greater than the number of transmit antennas, and the codewords may be a vector or a matrix. If the codeword is a matrix, the matrix includes r columns, each of which is also a vector, where the columns of the matrix may be orthogonal to each other.

[0035] In one embodiment, the vector constituting the codeword may be a 0-1 vector, i.e., the entire vector has only one value of 1 and the other values ​​are zero, or the vector constituting the codeword may be a DFT (Discrete Fourier Transform) vector, or the vector constituting the codeword may be obtained by performing a tensor product (Kronecker product) on two or more DFT vectors, or the vector constituting the codeword may be obtained by linearly combining two or more DFT vectors multiplied by different phase rotations, or the vector constituting the codeword may be obtained by performing a tensor product (Kronecker product) and multiplication by phase rotation on two or more DFT vectors.

[0036] In one embodiment, the base station or the terminal may search for L codewords to find a codeword that best matches the channel information of the terminal, and transmit data or signals using the codeword as the optimal codeword, where the codeword that matches the channel information includes at least one of the following codewords: a codeword that has the smallest distance from the channel information, a codeword that is most relevant to the channel information, a codeword that has the smallest distance from the optimal right singular vector or matrix of the channel information, a codeword that is most relevant to the optimal right singular vector or matrix of the channel information, and a codeword that has the largest signal-to-interference-and-noise ratio calculated from the channel information.

[0037] In one embodiment, the antenna may be a physical antenna, a logical antenna, a transmitting antenna, or a receiving antenna, or an antenna pair including a transmitting antenna and a receiving antenna, or a uniform linear array, or a uniform planar array, for example, including Ng rows and Mg columns of array elements / antennas, where Ng and Mg are both positive integers, or a uniform circular array, a non-uniform linear array, a non-uniform planar array, a non-uniform circular array, a directional antenna, an omnidirectional antenna, a dual-polarized antenna, or a single-polarized antenna, etc., not listed further here. Also, a port and an antenna may be interchangeable concepts.

[0038] In one embodiment, a first communication device (e.g., a terminal) and a second communication device (e.g., a base station) may transmit channel state information corresponding to the channel information through second-class precoding information, where the second-class precoding information acquires the channel state information based on AI, and the channel state information acquired by the base station and the terminal through an encoder of a self-encoder, where the self-encoder includes an encoder and a decoder, where the encoder is located on the terminal side and the decoder is located on the base station side. Specifically, the terminal compresses the channel information H acquired by the encoder to obtain compressed channel information H1, quantizes the compressed channel information H1, and then feeds it back to the base station. After receiving the quantized H1, the base station dequantizes the quantized H1 to obtain compressed channel information H1, and then inputs the compressed channel information H1 to the decoder, and the decoder decompresses the compressed channel information H1 to restore H.

[0039] Here, the above H may include K0 elements. The terminal selects K elements from H as H1, quantizes H1, and then feeds back the quantized H1 to the base station. The base station receives these K quantized elements, dequantizes them, inputs the dequantized K elements into the network model. The network model outputs K0 elements to recover H, thereby obtaining the precoding matrix of H. Here, K and K0 are integers greater than 1, and K < K0. Here, the K elements selected from H1 or H obtained by the compressor are all type-2 precoding information. For simplicity, the quantized H1 is also called type-2 precoding information.

[0040] Note that the type-2 precoding information may be a precoding matrix different from the type-1 precoding information generated by other non-AI methods, or the type-2 precoding information may be a precoding matrix other than the type-1 precoding information. Specifically, there is no limitation here.

[0041] Here, the channel information in each of the above embodiments is information for describing the channel environment between communication devices obtained based on a reference signal (e.g., CSI-RS). This channel information may be a time-domain channel matrix or a frequency-domain channel matrix. The channel information is a single complex matrix (i.e., a channel matrix), and the size of this channel matrix is related to the number of transmit antennas Nt, the number of receive antennas Nr, and the resource element (RE). For example, there is at least one Nr*Nt channel matrix in one physical resource block (Physical Resource Block).

[0042] In one embodiment, the beam parameter information may be a layer 1 reference signal received power or differential RSRP corresponding to at least one beam, or a layer 1 reference signal signal to interference and noise ratio or differential SINR corresponding to at least one beam, or a reference signal received quality corresponding to at least one beam, or a beam angle (i.e., AOA (Angle Of Arrival), ZOA (Zenith angle Of Arrival), AOD (Angle Of Departure), ZOD (Zenith angle Of Departure), etc.) corresponding to at least one beam. The AOA may also be a horizontal angle of arrival, the ZOA may also be a vertical angle of arrival, the AOD may also be a horizontal angle of departure, the ZOD may also be a vertical angle of departure, or may be a transmit beam index corresponding to the at least one beam, or may be a receive beam index corresponding to the at least one beam, or may be a transmit beam and receive beam pair index corresponding to the at least one beam, where the transmit beam and receive beam pair index is abbreviated as beam pair index or beam pair, or a beam domain receive power mapping corresponding to the at least one beam. MAP, BDRPM), or may be a channel state information-reference signal resource indicator corresponding to at least one beam, or may be a synchronization signal block resource indicator or other reference signal resource indicator, for example, SRSRI, corresponding to at least one beam, or may be a combination of at least two of beam parameter information such as RSRP, RSRQ, SINR, beam angle, transmit beam index, receive beam index, beam pair index, CRI, and SSBRI corresponding to at least one beam, or may be a linear value of any one of RSRP, RSRQ, and SINR, or may be a logarithmic value (i.e., decibel value (DB)) of any one of RSRP, RSRQ, and SINR.

[0043] In one embodiment, the location parameter information includes, but is not limited to, transmission time-related information, angle-related information, received signal quality-related information, and multipath-related information, where the transmission time-related information includes at least one of a time of arrival (TOA), a reference signal time difference (RSTD), a relative time of arrival (RTOA), a receive-transmit time difference (Rx-Tx time difference), and a transmit-receive time difference (Tx-Rx time difference). The angle-related information includes at least one of an angle of arrival, a departure angle, an arrival zenith angle, and an arrival azimuth angle, where the departure angle includes a departure zenith angle and a departure azimuth angle (AOD). The received signal quality-related information includes at least one of a reference signal received power, a SINR, a CQI, and an SNR. The multipath-related information includes at least one of the following information: the number of increased paths, the relative delay of the increased paths, the power of the increased multipaths, the time domain response of the increased multipaths, the real and imaginary parts of the time domain response of the increased multipaths, the path with the strongest power, the first path with the strongest power, the N paths with the strongest power, the times and / or RSRPs corresponding to the N paths with the strongest power, the N paths within the time window, the times and / or RSRPs corresponding to the N paths within the time window, the N paths greater than a threshold, the times and / or RSRPs corresponding to the N paths greater than a threshold, and a line-of-sight / non-line-of-sight indicator (LoS / NLoS indicator).

[0044] In one embodiment, the beam parameter information or the position parameter information may be a subset of the channel state information, that is, the beam parameter information belongs to the channel state information, the position parameter information belongs to the channel state information, and the channel state information belongs to the measurement parameters.

[0045] In another embodiment, the measurement parameters, channel state information, and beam parameter information all belong to the measurement results, or the processing results, or the generated results.

[0046] In one embodiment, to transmit measurement results, e.g., channel state information, at the physical layer, the UE and the base station define a report (e.g., a CSI report or a CSI report configuration), where the report defines at least one of information such as a time-frequency resource for feeding back CSI, a report quality (reportQuantity) included in the CSI, a time-domain type (reportConfigType) of the CSI feedback, a channel measurement resource, an interference measurement resource, and a measured bandwidth size. Here, the CSI report (i.e., the CSI report) can be transmitted on uplink transmission resources, which include the PUSCH and the PUCCH. However, since the CSI report may include time-domain characteristics, the CSI report may be a periodic CSI report (P-CSI), an aperiodic CSI report (AP-CSI), or a semi-persistent CSI report (SP-CSI).

[0047] Generally, the number of bits for P-CSI transmission is relatively small, and P-CSI is usually transmitted on the PUCCH. The number of bits for A-CSI transmission is relatively large, and P-CSI is usually transmitted on the PUSCH. SP-CSI may be transmitted on either the PUSCH or the PUCCH. Here, P-CSI based on PUCCH transmission is generally configured by higher layer signaling (Radio Resource Control, RRC), and SP-CSI based on PUCCH transmission is also configured and activated or deactivated by higher layer signaling (RRC and / or MAC CE). However, SP-CSI based on PUSCH transmission is activated or deactivated by physical layer signaling (Downlink control information, DCI). A-CSI is triggered by DCI. Meanwhile, DCI is generally transmitted on a physical downlink control channel (PDCCH).

[0048] In the embodiments of the present application, the feedback CSI may also be referred to as transmission CSI or transmitted CSI, for example, channel state information (CSI) is fed back or transmitted on uplink transmission resources, where the uplink transmission resources and corresponding CSI are all indicated by one channel state information report (i.e., CSI report).

[0049] In one embodiment, the base station configures N CSI reports that the terminal needs to feed back to the base station through higher layer signaling and / or physical layer signaling, where each CSI report has an identity (ID), which is referred to as a CSI report ID. The terminal may select M CSI reports from the N CSI reports based on its own computing or processing capability and a request from the base station, and feed back at least one CSI report from the M CSI reports based on uplink feedback resources, where N and M are positive integers and M<=N.

[0050] Furthermore, among the MC CSI reports that need to be fed back, the feedback resources of at least two CSI reports collide, where the feedback resource collision of two CSI reports means that the transmission resources (e.g., PUCCH or PUSCH) corresponding to the two CSI reports to be fed back have at least one symbol that is the same and / or at least one subcarrier that is the same.

[0051] In an embodiment of the present application, feeding back or transmitting a CSI report refers to feeding back channel state information of a CSI report configuration. For example, feeding back or transmitting a CSI report refers to transmitting content that needs to be transmitted in a CSI report configuration via transmission resources.

[0052] In one embodiment, information, such as channel state information or data sets, may be transmitted by a higher layer. The transmission resources for transmitting uplink data by the higher layer include two scheduling schemes, namely, a dynamic grant (DG) scheme and a configured grant (CG), where the configured grant includes two types, namely, a configured grant type 1 (Configured Grant Type 1) and a configured grant type 2 (Configured Grant Type 2).

[0053] Exemplarily, the terminal may transmit data using parameters determined by the configured license type 1. The base station may transmit parameters such as the time-frequency resource location, the period of the CG resource, the number of Hybrid Automatic Repeat Request (HARQ) processes using the CG resource, and the MCS of the configured license type 1 to the terminal through RRC signaling, and the terminal stores these parameters as a configured uplink grant. After the RRC signaling configures the grant type 1, the terminal can perform uplink data transmission using the transmission resource corresponding to this configured license.

[0054] For example, the terminal may transmit data using parameters determined by the configuration license type 2. The base station may transmit parameters such as the period of the CG resource of the configuration license type 2, the number of HARQ processes using the CG resource, and the MCS table to be used to the terminal through RRC signaling, but parameters such as the time-frequency resource location and the MCS index value are transmitted to the terminal by the network device through DCI, and the terminal stores these parameters as a configuration uplink grant, and the base station activates or deactivates the terminal's operation of transmitting uplink data through physical layer signaling.

[0055] In a possible embodiment, the indicators (i.e., Indicators) of various parameters in all the above examples may be referred to as indices (i.e., Indexes) or Identifiers (IDs), which are completely equivalent concepts. For example, the resources of a wireless system may include, but are not limited to, an index corresponding to one reference signal resource, a reference signal resource group, a reference signal resource configuration, a channel state information report, a CSI report set, a terminal, a base station, a panel, a neural network, a sub-neural network, a neural network layer, a precoding matrix, a beam, a transmission scheme, a transmission scheme, a reception scheme, a module, a model, a functional module, etc.

[0056] In one embodiment, the first communication device may transmit the identity of one or a set of resources to the second communication device through various higher layer signaling and / or physical layer signaling. Specifically, when the first communication device is a terminal and the second communication device is a base station, the first communication device may feed back the identity of one or a set of source identities to the second communication device through various higher layer signaling and / or physical layer signaling, and the second communication device may indicate the identity of one or a set of source identities to the first communication device through various higher layer signaling and / or physical layer signaling.

[0057] Here, the higher layer signaling may include, but is not limited to, Radio Resource Control (RRC), Media Access Control (MAC CE), and other signaling other than physical layer signaling, such as LTE Positioning Protocol (LPP) higher layer signaling, NR Positioning Protocol A (NRPPPa) higher layer signaling, and LTE Positioning Protocol A (LPPa) higher layer signaling, where LPP is also used in the NR positioning protocol. Meanwhile, the physical layer signaling may be transmitted on a Physical Downlink Control Channel (PDCCH) or a Physical Uplink Control Channel (PUCCH).

[0058] In one embodiment, a slot may be a slot or a subslot (mini slot). One slot or subslot includes at least one symbol. Here, the symbol is a time unit in one subframe, frame, or slot, and may be, for example, an Orthogonal Frequency Division Multiplexing (OFDM) symbol, a Single-Carrier Frequency Division Multiple Access (SC-FDMA) symbol, an Orthogonal Frequency Division Multiple Access (OFDMA) symbol, etc.

[0059] In one embodiment, a first communication device or a second communication device needs to transmit a reference signal (RS) to calculate channel state information or perform channel estimation, mobility management, positioning, etc. The reference signal includes, but is not limited to, a channel-state information reference signal (CSI-RS). The channel-state information reference signal may include a zero-power CSI-RS (ZP CSI-RS) and a non-zero-power CSI-RS (NZP CSI-RS), a channel-state information-interference measurement signal (CSI-IM), a sounding reference signal (SRS), a synchronization signal block (SSB), a physical broadcast channel (PBCH), and a synchronization signal block / physical broadcast channel (SSB / PBCH). Here, the NZP CSI-RS may be used to measure channel or interference, the CSI-RS may be used for tracking, and the CSI-RS for tracking is called a tracking reference signal (CSI-RS for Tracking, TRS), the CSI-IM is generally used to measure interference, and the SRS is used to measure uplink channels. Furthermore, a set of resource elements (REs) included in the time-frequency resources for transmitting reference signals is called a reference signal resource, e.g., a CSI-RS resource, an SRS resource, a CSI-IM resource, and an SSB resource. The SSB includes a synchronization signal block and / or a physical broadcast channel.

[0060] In one embodiment, in order to save signaling overhead, etc., multiple reference signal resources may be divided into multiple sets (e.g., a CSI-RS resource set, a CSI-IM resource set, and an SRS resource set), where a reference signal resource set includes at least one reference signal resource, but multiple reference signal resource sets may come from the same reference signal resource (e.g., a CSI-RS resource setting and an SRS resource setting, where the CSI-RS resource setting may be integrated with the CSI-IM resource setting and referred to as the CSI-RS resource setting), and parameter information is configured by this reference signal resource.

[0061] In one embodiment, the base station may configure measurement resource information for acquiring various measurement parameters such as channel state information, where the measurement resource information is N Channel Measurement Resource (CMR) information and / or C M C N and C M and are all positive integers. Specifically, the base station may configure the measurement resource information in a report configuration or a reporting setting.

[0062] In a possible embodiment, one piece of channel measurement resource information may include at least one channel reference signal resource, for example at least one CSI-RS resource setting or at least one SRS resource setting, and one piece of interference measurement resource information may include at least one interference reference signal resource, for example at least one CSI-IM resource setting.

[0063] In a possible embodiment, one piece of channel measurement resource information may include at least one channel reference signal resource set, for example at least one CSI-RS resource set or at least one SRS resource set, and one piece of interference measurement resource information may include at least one interference reference signal resource set, for example at least one CSI-IM resource set.

[0064] In a possible embodiment, one piece of channel measurement resource information may include at least one channel reference signal resource, for example, at least one CSI-RS resource or at least one SRS resource, and one piece of interference measurement resource information may include at least one interference reference signal resource, for example, at least one CSI-IM resource.

[0065] In one embodiment, the beams include transmit beams, receive beams, receive and transmit beam pairs, and transmit and receive beam pairs.

[0066] A beam may be understood as a resource, such as a reference signal resource, a reference signal resource set, a transmitting end spatial filter, a receiving end spatial filter, a spatial filter, a spatial receiving parameter, a transmitting side precoding, a receiving side precoding, an antenna port, an antenna weight vector or an antenna weight matrix.

[0067] Correspondingly, the beam index may be replaced with a resource index (e.g., a reference signal resource index) because a beam can perform transmission binding with several time-frequency code resources.

[0068] The beam may be a transmission (transmission / reception) scheme, which may include spatial division multiplexing, frequency domain / time domain diversity, beamforming, etc. The base station may perform quasi-colocation (QCL) configuration for two reference signals and report the configuration to the terminal to describe channel characteristics. Here, the quasi-colocation parameters include at least Doppler spread, Doppler shift, delay spread, average delay, average gain, and spatial parameters (Spatial Rx parameters or Spatial parameters). The spatial parameters may include spatial reception parameters, beam angles, spatial correlation of reception beams, average delay, and correlation of time-frequency channel responses (including phase information). The spatial domain filtering may be at least one of a DFT vector, a precoding vector, a DFT matrix, and a precoding matrix, or may be a vector formed by a linear combination of multiple DFT vectors, or may be a vector formed by a linear combination of multiple precoding vectors. Vector and vector may be interchangeable concepts.

[0069] A beam pair may also be a combination including one transmit beam and one receive beam.

[0070] Referring to Figure 2, Figure 2 is a flowchart of a data set transmission method according to an embodiment of the present application, which may be used in a first communication device, such as the receiving device shown in Figure 1. The data set transmission method includes, but is not limited to, step S110, step S120, and step S130.

[0071] Step S110: A first data set is obtained.

[0072] In this step, the first dataset includes L0 samples, also referred to as parameter sets, each parameter set including at least one of at least one first sub-parameter set and / or at least one second sub-parameter set. In one embodiment, the first sub-parameter set is also referred to as the sample's feature, and the second sub-parameter set is also referred to as the sample's label. In one embodiment, the sample includes M0 labels and M1 features, where M0 and M1 are non-negative integers.

[0073] In one embodiment, a sample may include one feature and one label, e.g., a sample in supervised learning, or a sample may include only one feature and no label, e.g., a sample in unsupervised learning, or a sample may include multiple features and one label, e.g., a sample in a multi-input single-output supervised learning network model, or a sample may include one feature and multiple labels, e.g., a sample in a single-input multi-output supervised learning network model.

[0074] Here, the features may be arrays, and the labels may also be arrays. However, the array may also be a vector, a matrix, or a tensor with more than two dimensions. Each element in the array may be a discrete value or a real value, for example, a real value between 0 and 1, or a real value between -0.5 and 0.5.

[0075] Here, the sample features and / or labels include channel matrix information, which includes channel state information and / or beam metric parameters, or the sample features and / or labels include only channel state information, or the sample features and / or labels include only beam metric parameters.

[0076] In one embodiment, a normalization process needs to be performed on the elements in the array corresponding to the labels or features so that the network model can converge more quickly. The so-called normalization is a value that normalizes the element values ​​in the array to the interval greater than or equal to a and less than or equal to b. In one example, a=-0.5, b=0.5, and in another example, a=0, b=1.

[0077] For example, normalization may be achieved by dividing an element in an array by the number with the largest absolute value in the array element, or by dividing an element in an array by the variance in the array element, or by dividing an element in an array by a fixed value (e.g., the maximum value of all elements in all samples), or by dividing an element in an array by a statistical value (e.g., the statistical variance of all elements in all samples). Normalization may be achieved by one-hot encoding for index values, such as beam index, CRI, SSBRI, etc.

[0078] In one embodiment, to obtain the first data set, the second communication device transmits a reference signal to the first communication device, and after receiving the reference signal, the first communication device measures the received reference signal to obtain the first data set.

[0079] Step S120: Obtain a second data set from the first data set based on the first sample parameters and / or the second sample parameters.

[0080] Here, the second dataset includes L1 samples, the first sample parameters and the second sample parameters are both parameters corresponding to samples in the first dataset, and each sample or set of samples in the first dataset corresponds to at least one of the first sample parameters and / or the second sample parameters. The first sample parameters are used to screen the L0 samples for samples that satisfy a predetermined condition. The second sample parameters are used to sort the samples that satisfy the predetermined condition and obtain a second dataset from the sorted samples, for example, by selecting the top L1 samples to form the second dataset. Here, L0 and L1 are positive integers, and L1 is less than or equal to L0.

[0081] In one embodiment, the first sample parameter and / or the second sample parameter are first sample parameters and / or second sample parameters corresponding to at least one sample.

[0082] In some embodiments, if a sample satisfies a preset condition, the sample is also a valid sample.

[0083] In some embodiments, each sample further corresponds to at least one first sample parameter, which may include one or more parameters in the first subparameter set or one or more parameters in the second subparameter set. The first sample parameters may further include parameters for describing sample quality, characteristics, application scenarios, channel characteristics, etc., such as reference signal received power, signal-to-noise ratio, signal-to-interference-plus-noise ratio, reliability, inference time, measurement time, and moving speed for evaluating sample quality. In some embodiments, the first sample parameters may be first sample parameters corresponding to one sample or may be first sample parameters corresponding to a set of samples. In one example, the parameters corresponding to the first sample, such as RSRP, SINR, CQI, etc., may represent the link quality of the terminal at the corresponding time when the sample is collected, or may be broadband, and one sample may only correspond to RSRP, SINR, and CQI. The probability and reliability here may be evaluation indicators for the label of this sample, and the sample corresponds to the reliability or probability.

[0084] In one example, a sample or a set of samples corresponds to at least one second sample parameter, including, but not limited to, a signal-to-interference-plus-noise ratio, a reference signal received power, a channel rank, a modulation and decoding scheme, a channel quality indicator, a mobile speed, a bit error rate, and a block error rate.

[0085] As used herein, unless otherwise specified, signal-to-interference-plus-noise ratio includes signal-to-interference-plus-noise ratio, signal-to-noise ratio, signal-to-interference ratio, etc. A high-quality sample may also be referred to as a satisfactory sample or a useful sample.

[0086] In one embodiment, the first communication device may measure a reference signal resource set corresponding to each sample to obtain first sample parameters and / or second sample parameters.

[0087] Step S130: Transmit the second data set.

[0088] As an example, if L0=100 and L1=10, when a first communication device acquires a first data set including 100 samples, it can acquire a second data set including 10 samples from the first data set based on the first sample parameters and / or the second sample parameters, and finally transmit this second data set to the second communication device.

[0089] In this embodiment, by adopting the above-mentioned method for transmitting datasets including steps S110 to S130, the embodiment of the present application can obtain a first dataset including L0 samples, then obtain a second dataset including L1 samples from the first dataset based on the first sample parameters and / or second sample parameters, and finally transmit the second dataset, where L0 and L1 are positive integers, and L1 is less than or equal to L0. That is, the first communication device can screen a high-quality or suitable second dataset from the first dataset based on the first sample parameters and / or second sample parameters, thereby improving the model training effect and further improving the model performance.

[0090] In one embodiment, obtaining the second data set from the first data set based on the first sample parameters may specifically include obtaining the second data set from the first data set based on the first sample parameters and obtaining a first threshold group, and then obtaining the second data set from the first data set based on the first sample parameters corresponding to at least one sample and the first threshold group.

[0091] Here, the at least one sample may be at least one sample in the first data set, and the first sample parameter may include at least one parameter. Correspondingly, the first threshold group may include at least one threshold group, and one threshold group may include one or more thresholds. Here, each parameter corresponds to one first threshold group. The values ​​of the threshold groups corresponding to different parameters may be the same or different, and are not specifically limited herein.

[0092] Below are some examples of using thresholds to determine whether a sample is valid.

[0093] In one example, the first sample parameter is L1-RSRP, and the first threshold is RSRP0. At this time, if L1-RSRP is equal to or less than RSRP0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if L1-RSRP is greater than RSRP0, it can be determined that the sample satisfies the preset condition or is a valid sample. In another example, the first sample parameter is L1-SINR, and the first threshold is SINR0. At this time, if L1-SINR is equal to or less than SINR0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if L1-SINR is greater than SINR0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is CQI, and the first threshold is CQI0. At this time, if CQI is equal to or less than CQI0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if the CQI is greater than CQI0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is MCS, and the first threshold is MCS0. At this time, if the MCS is less than or equal to MCS0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if the MCS is greater than MCS0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is moving speed V, and the first threshold is V0. At this time, if the moving speed V is greater than V0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if the moving speed V is less than V0, it can be determined that the sample satisfies the preset condition or is a valid sample.In yet another example, the first sample parameter is a confidence level Z, and the first threshold is Z0. In this case, if the confidence level Z is equal to or less than Z0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if the confidence level Z is greater than Z0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is a predicted probability P, and the first threshold is P0. In this case, if P is equal to or less than P0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if P is greater than P0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is a measurement time MT, and the first threshold is MT0. In this case, if MT is equal to or less than MT0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if MT is greater than MT0, it can be determined that this sample satisfies a preset condition or is a valid sample.

[0094] In yet another example, the first sample parameter is an inference time T, and the first threshold is T0. At this time, if T is smaller than T0, it can be determined that the sample satisfies the preset condition or is a valid sample. Otherwise, if T is greater than T0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. In yet another example, the first sample parameter is a throughput TH, and the first threshold is TH0. At this time, if TH is equal to or smaller than TH0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if TH is greater than TH0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is a bit error rate BER, and the first threshold is BER0. At this time, if the BER is equal to or greater than BER0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if the BER is smaller than BER0, it can be determined that the sample satisfies the preset condition or is a valid sample. In yet another example, the first sample parameter is a block error rate (BLER), and the first threshold is BLER0. In this case, if the BLER is equal to or greater than BLER0, it can be determined that the sample does not satisfy the preset condition or is an invalid sample. Otherwise, if the BLER is less than BLER0, it can be determined that the sample satisfies the preset condition or is a valid sample. In these embodiments or examples, a sample not satisfying a quality requirement may be referred to as a sample that does not satisfy the requirement or a sample that does not satisfy the requirement. This means that the sample is not applied to a training model, not applied to model monitoring, or not applied to model inference. Therefore, the terminal does not need to feed back the sample. That is, a sample that does not satisfy the requirement cannot be included in the dataset.

[0095] In one example, it is necessary to simultaneously determine whether a sample satisfies a requirement (e.g., a quality requirement) based on K first sample parameters and K first sample parameter thresholds, where a sample is not deemed to satisfy the requirement unless each first type of first sample parameter is equal to or greater than its corresponding threshold, or each second type of first sample parameter is equal to or less than its corresponding threshold. For example, in one example, before a sample is considered to satisfy the condition, at least two of the following conditions must be simultaneously satisfied: the L1-RSRP corresponding to the sample is equal to or greater than RSRP0, or the L1-SINR corresponding to the sample is equal to or greater than SINR0, or the CQI corresponding to the sample is equal to or greater than a threshold CQI0, or the MCS corresponding to the sample is equal to or greater than a threshold MCS0, or the probability P corresponding to the sample is equal to or greater than a threshold P0, or the reliability Z corresponding to the sample is equal to or greater than a threshold Z0, the throughput TH corresponding to the sample is equal to or greater than a threshold TH0, or the inference time T corresponding to the sample is equal to or less than T0, or the moving speed V corresponding to the sample is equal to or less than a threshold V0, or the measurement time MT corresponding to the sample is equal to or less than a threshold MT0, the bit error rate BER corresponding to the sample is equal to or less than a threshold BER0, and the block error rate BLER corresponding to the sample is equal to or less than BLER0.

[0096] For the first sample parameter X (mainly the first type of first sample parameter, such as the reference signal received power RSRP, the signal-to-interference-and-noise ratio SINR, the modulation and coding scheme MCS, the channel quality indicator CQI, the reliability, the prediction probability, the cosine similarity, the position parameter information, the beam parameter information, the throughput), a second threshold Y1 and a third threshold Y2 are provided, and Y1 is smaller than Y2. In one example, if X is less than or equal to Y1, it can be determined that the sample quality of the sample does not meet the preset condition or is an invalid sample; otherwise, if X is greater than or equal to Y2, it can be determined that the sample quality of the sample meets the preset condition or is a valid sample. In another example, if X (mainly referring to the first sample parameter of the second type, including but not limited to at least one of the moving speed, the measurement time, the inference time, the bit error rate, and the block error rate) is greater than or equal to Y2, it can be determined that the sample quality of the sample does not meet the preset condition or is an invalid sample; otherwise, if X is less than or equal to Y1, it can be determined that the sample quality of the sample meets the preset condition or is a valid sample.

[0097] In one example, if the first sample parameter L1-RSRP is equal to or less than a second threshold RSRP2, it is determined that the sample does not satisfy the preset condition or is an invalid sample. If the L1-RSRP is equal to or greater than a third parameter threshold RSRP3, it is determined that the sample satisfies the preset condition or is a valid sample, where RSRP2 is smaller than RSRP3. In another example, if the first sample parameter L1-SINR is equal to or less than a second threshold SINR2, it is determined that the sample does not satisfy the preset condition or is an invalid sample. If the L1-SINR is equal to or greater than a third parameter threshold SINR3, it is determined that the sample satisfies the preset condition or is a valid sample, where SINR2 is smaller than SINR3. In another example, if the first sample parameter CQI is equal to or less than a second threshold CQI2, it is determined that the sample does not satisfy the preset condition or is an invalid sample. If the CQI is equal to or greater than a third parameter threshold CQI3, it is determined that the sample satisfies the preset condition or is a valid sample. Here, CQI2 is smaller than CQI3. In yet another example, if the first sample parameter MCS is equal to or smaller than a second threshold MCS2, it is determined that the sample does not satisfy the predetermined condition or is an invalid sample. If MCS is equal to or larger than a third parameter threshold MCS3, it is determined that the sample satisfies the predetermined condition or is a valid sample. Here, MCS2 is smaller than MCS3. In yet another example, the first sample parameter is a moving speed V, and if V is equal to or smaller than a second threshold V2, it is determined that the sample satisfies the predetermined condition or is a valid sample. If V is equal to or larger than a third parameter threshold V3, it is determined that the sample does not satisfy the predetermined condition or is an invalid sample. Here, V2 is smaller than V3. In yet another example, the first sample parameter is a probability P corresponding to the sample, and if P is equal to or smaller than a second threshold P2, it is determined that the sample does not satisfy the predetermined condition or is an invalid sample.If P is equal to or greater than a third parameter threshold P3, it is determined that the sample satisfies the preset condition or is a valid sample, where P2 is smaller than P3. In yet another example, the first sample parameter is confidence Z, and if Z is equal to or less than a second threshold Z2, it is determined that the sample does not satisfy the preset condition or is an invalid sample. If Z is equal to or greater than a third parameter threshold Z3, it is determined that the sample satisfies the preset condition or is a valid sample, where Z2 is smaller than Z3.

[0098] In yet another example, the first sample parameter is an inference time T, and if T is equal to or greater than a second threshold T3, it is determined that the sample does not satisfy the predetermined condition or is an invalid sample. If T is equal to or less than a third parameter threshold T2, it is determined that the sample satisfies the predetermined condition or is a valid sample, where T2 is less than T3. In yet another example, the first sample parameter is a measurement time MT corresponding to the sample, and if MT is equal to or greater than a second threshold MT3, it is determined that the sample does not satisfy the predetermined condition or is an invalid sample. If MT is equal to or less than a third parameter threshold MT2, it is determined that the sample satisfies the predetermined condition or is a valid sample, where MT2 is less than MT3.

[0099] In some cases, the first sample parameter X is between Y1 and Y2. Whether the sample satisfies the requirements needs to be further determined based on the capabilities of the terminal, the capabilities or complexity of the model, functions, etc.

[0100] In an embodiment of the present application, a high quality or acceptable second data set can be screened from a first data set based on a first sample parameter and a first threshold group.

[0101] 3 , each parameter in the first sample parameters may be compared with a corresponding threshold parameter in a first threshold group. If a parameter does not match the threshold parameter, the sample corresponding to this parameter may be discarded, and the remaining set of samples in the first dataset may be included in the second dataset. The present application may screen out and discard poor-quality samples from the first dataset based on the first sample parameters and the first threshold group, thereby obtaining a high-quality or acceptable second dataset. The present application may compare each parameter in the first sample parameters with a corresponding threshold parameter in the first threshold group. If a parameter matches the threshold parameter, the sample corresponding to this parameter may be included in the second dataset. The present application may screen out high-quality or acceptable samples from the first dataset based on the first sample parameters and the first threshold group.

[0102] In frequency domain prediction based on an artificial intelligence algorithm, the deployed artificial intelligence model typically predicts channel state information or beam information of one or more frequency domains (i.e., prediction windows) based on channel measurement results of multiple frequency domains (i.e., observation windows), where the measurement results obtained by a first communication device (e.g., a terminal) measuring a reference signal resource set within one frequency domain range once are referred to as one sample (or frequency domain parameter subset).

[0103] For training and fine-tuning a frequency-domain prediction model, the dataset that needs to be acquired mainly consists of two parts: model input data and model label data. Here, the model input data consists of multiple frequency-domain samples within an observation window, and the model label data consists of one or more frequency-domain samples within a prediction window. Training this frequency-domain prediction model requires the complete collection of all samples within the same observation and prediction windows. If a poor-quality sample exists among these samples, causing it to be discarded, or if a frequency-domain sample is not collected for some reason (e.g., because a reference signal resource is not transmitted or measured), all other related frequency-domain samples should be discarded.

[0104] Therefore, as shown in Figure 4, if any sample is discarded or the number of measured valid samples exceeds a predetermined threshold within any one measurement window configured by a first communication device (e.g., a terminal) or a second communication device (e.g., a base station), all samples within this measurement window should not be reported, that is, all samples within this measurement window should not be transmitted to the first communication device.

[0105] According to this, in one embodiment, the first data set may include at least one sample group, where the sample group includes K samples associated with a frequency domain, where K is a positive integer. In this case, obtaining the second data set from the first data set based on the first sample parameters may specifically include obtaining a first threshold group, determining whether the sample group satisfies a predetermined condition based on the first sample parameters corresponding to at least one sample of the sample group and the first threshold group, and then including the samples in the sample group that satisfy the predetermined condition in the second data set.

[0106] A sample group containing K samples that are related in the frequency domain may be referred to as a related sample set.

[0107] In one embodiment, determining whether a sample group satisfies a predetermined condition based on a first sample parameter corresponding to at least one sample of the sample group and a first threshold group, the predetermined condition includes at least one of: the number of valid samples in the sample group is equal to or greater than K; the ratio of the valid samples in the sample group to the total number K of samples in the sample group is equal to or greater than a first predetermined ratio value; the frequency domain position interval between a sample in the sample group and a sample adjacent to the frequency domain position is smaller than N frequency domain units; there are at least K valid samples in a first predetermined frequency domain segment before the sample in the sample group; or there are at least K valid samples in a second predetermined frequency domain segment after the sample in the sample group, where the frequency domain unit includes at least one of a subcarrier interval, a number of resource blocks, a number of subbands, and a number of center frequency bands.

[0108] In the embodiment of the present application, K, K1, K2, K3, and N are positive integers, and K1, K2, and K3 are less than or equal to K. A subband is a part of a frequency band with a specific characteristic. Samples that meet the first threshold group belong to valid samples, or samples from which data (i.e., valid data) can be obtained (i.e., valid samples). The first and second preset frequency domain segments may be set according to actual circumstances and are not specifically limited herein.

[0109] Illustratively, a first threshold group is obtained, and whether the number of valid samples in each sample group is greater than or equal to K1 is determined based on the first sample parameters and the first threshold group; if the number of valid samples in the sample group is greater than or equal to K1, the samples in the sample group that meet this preset condition are included in the second data set.

[0110] For example, if a first data set includes one sample group A and one sample group B, and sample group A includes 10 frequency domain-related samples, i.e., K A = 10, sample group B contains 7 frequency domain related samples, i.e., K B Assume that K1=7. If K1=5, the number of valid samples in sample group A is determined to be 8 based on the first sample parameters and the first threshold group, and the number of valid samples in sample group B is determined to be 6 based on the first sample parameters and the first threshold group. In this case, it can be seen that sample group A and sample group B simultaneously satisfy the preset condition. At this time, the 8 valid samples in sample group A and the 6 valid samples in sample group B can be constructed as a second data set, and this second data set includes 14 samples.

[0111] Or, the first data set includes sample group A and sample group B, and sample group A includes 10 frequency domain-related samples, i.e., K A = 10, and sample group B contains seven frequency domain related samples, i.e., K BAssume it is equal to 7. When K1 = 4, based on the first sample parameter and the first threshold group, determine that the number of valid samples in sample group A is 8, and based on the first sample parameter and the first threshold group, determine that the number of valid samples in sample group B is 3. In this case, it can be seen that sample group A satisfies the preset conditions, but sample group B does not satisfy the preset conditions. At this time, discard sample group B, and construct 8 valid samples in sample group A as the second dataset. At this time, the second dataset contains 8 samples.

[0112] Exemplarily, obtain the first threshold group, and based on the first sample parameter and the first threshold group, determine whether the number of invalid samples in each sample group is K4 or more. If the number of invalid samples in the sample group is K4 or more, discard all samples in this sample group. Here, K4 < K. For example, when K = 10, the value of K4 may be 1, 2, etc. The value of K4 may also be determined by K. An invalid sample is a sample that does not conform to the first threshold group or a sample for which data (i.e., the data is invalid) has not been obtained.

[0113] Exemplarily, obtain the first threshold group, and based on the first sample parameter and the first threshold group, determine whether the number of valid samples in each sample group is K5 or more. If the number of valid samples in the sample group is K5 or more, all samples in this sample group are all included in the second dataset. Here, K5 < K. For example, when K = 10, the value of K5 may be 6, 7, 8, 9, 10, etc. The value of K5 may also be determined by K. A valid sample is a sample that can be effectively measured and conforms to the first threshold group.

[0114] Illustratively, a first threshold group is obtained, valid samples in each sample group are determined based on the first sample parameters and the first threshold group, and then it is determined whether the ratio of the number of valid samples to the total number of samples K in the sample group is greater than or equal to a first preset ratio value. If the ratio of the number of valid samples to the total number of samples in the sample group is greater than or equal to the first preset ratio value, the samples in the sample group that meet the preset condition are included in the second data set.

[0115] Here, when the first data set includes multiple sample groups, the first preset ratio value may be multiple, for example, different sample groups may correspond to different first preset ratio values, or different sample groups may correspond to the same first preset ratio value. Alternatively, when the first data set includes multiple sample groups, the first preset ratio value may be only one, that is, one first data set may correspond to one first preset ratio value. In addition, the value of the first preset ratio value may be set according to actual circumstances, and is not specifically limited herein.

[0116] For example, if a first data set includes sample group A and sample group B, sample group A has K A number of frequency domain-related samples (i.e., the total number of samples in sample group A is K A Sample group B contains K B number of frequency domain-related samples (i.e., the total number of samples in sample group B is K B and a first preset proportion value corresponding to sample group A is N1, a first preset proportion value corresponding to sample group B is N2, and a number of valid samples in sample group A is determined based on the first sample parameters and the first threshold group as L. A However, L A / K AIf N is greater than or equal to N, the samples in sample group A may be included in the second data set. Similarly, the number of valid samples in sample group B may be determined based on the first sample parameters and the first threshold group. B However, L B / K B If >= N2, the samples in sample group B may be included in the second data set, where the values ​​of N1 and N2 may be the same or different.

[0117] For example, a first threshold value is obtained, and the number of invalid samples in a sample group is determined based on the first sample parameter and the first threshold value. If the ratio of the number of invalid samples to the total number of samples in the sample group is equal to or greater than a second preset ratio value, all samples in the sample group are discarded. Here, the value of the second preset ratio value may be set according to actual circumstances and is not specifically limited herein.

[0118] Illustratively, a first threshold group is obtained, and based on the first sample parameter and the first threshold group, it is determined whether the frequency domain position interval between one sample in the sample group and a sample adjacent to the frequency domain position is smaller than N frequency domain units, and samples in the sample group that are smaller than N frequency domain units are included in the second data set.

[0119] Illustratively, a first threshold group is obtained, and whether there are at least K2 valid samples in a first preset frequency domain segment prior to one sample of the sample group is determined based on the first sample parameter and the first threshold group; if there are K2 valid samples, all samples in the sample group are included in the second data set.

[0120] Illustratively, a first threshold group is obtained, and whether there are at least K3 valid samples in a second preset frequency domain segment after one sample of the sample group is determined based on the first sample parameter and the first threshold group; if there are K3 valid samples, all samples in the sample group are included in the second data set.

[0121] In one embodiment, when a sample includes a first sub-parameter set and / or a second sub-parameter set, the first sub-parameter set corresponding to at least one sample in the sample group satisfies a constraint relationship in the frequency domain, or the second sub-parameter set corresponding to at least one sample in the sample group satisfies a constraint relationship in the frequency domain, or the first sub-parameter set and the second sub-parameter set corresponding to at least one sample in the sample group satisfy a constraint relationship in the frequency domain.

[0122] The first sub-parameter set corresponding to at least one sample in the sample group satisfies a frequency domain constraint relationship, equivalent to the feature of at least one sample in the sample group satisfying the frequency domain constraint relationship; similarly, the second sub-parameter set corresponding to at least one sample in the sample group satisfies a frequency domain constraint relationship, equivalent to the label of at least one sample in the sample group satisfying the frequency domain constraint relationship.

[0123] For training and fine-tuning the artificial intelligence model, the second communication device (e.g., a base station) may configure two related reference signal resource sets to collect model input data and model label data, respectively, and the two have a paired relationship. Since the first communication device (e.g., a terminal) may simultaneously feedback frequency domain parameter sets (i.e., measurement results, i.e., a set of multiple samples, or some parameters of multiple samples, e.g., sample features) measured based on multiple frequency domains in one report, the report needs to include frequency domain information (e.g., subband, PRB group, subcarrier) corresponding to each frequency domain parameter set. At the same time, since the model input data and the model label data have a paired relationship, if the model input data of a certain frequency domain is discarded due to poor quality, the corresponding frequency domain model label data should also be discarded.

[0124] The associated reference signal resource set is a reference signal resource set associated with a frequency, and the associated frequency may be set according to actual circumstances and is not specifically limited herein. The frequency domain information is used to determine model input data and / or model label data of a specific frequency.

[0125] Based on this, a frequency domain constraint relationship may be set. If a first subparameter set and / or a second subparameter set corresponding to at least one sample in a sample group satisfy the frequency domain constraint relationship, measurement results (i.e., samples) of two or more reference signal resource sets associated with each other may be reported together or discarded altogether without being reported. For example, referring to FIG. 5, a reference signal resource set corresponding to a first subparameter set corresponding to at least one sample in a sample group is in frequency domain resource set 1, and a reference signal resource set corresponding to a second subparameter set corresponding to at least one sample in the sample group is in frequency domain resource set 2. If the first subparameter set and the second subparameter set corresponding to a certain sample in the sample group satisfy the frequency domain constraint relationship, frequency domain resource set 1 corresponding to the first subparameter set is associated with frequency domain resource set 2 corresponding to the second subparameter set. In this case, samples corresponding to the first subparameter set and the second subparameter set that satisfy the frequency domain constraint relationship may be reported together or discarded altogether without being reported.

[0126] In one embodiment, the frequency domain constraint relationship is: In the sample group, the frequency domain position corresponding to the first sub-parameter set of one sample and the frequency domain position corresponding to the first sub-parameter set of one sample adjacent to the frequency domain position are adjacent frequency domain positions; Or, in the sample group, the frequency domain position corresponding to the second sub-parameter set of one sample and the frequency domain position corresponding to the second sub-parameter set of one sample adjacent to the frequency domain position are adjacent frequency domain positions; Or, in the sample group, the interval between the frequency domain position of the first subparameter set of one sample and the frequency domain position of the first parameter subset of one sample adjacent to the frequency domain position is smaller than N frequency domain units; Or, in the sample group, the interval between the frequency domain position of the second sub-parameter set of one sample and the frequency domain position of the second parameter subset of one sample adjacent to the frequency domain position is smaller than N frequency domain units; Alternatively, in a sample group, the frequency domain position of the reference signal corresponding to one sample and the frequency domain position of the reference signal corresponding to one sample adjacent to the frequency domain position Place , are adjacent frequency domain positions thing, Or, in the sample group, the difference between the frequency domain position of the reference signal corresponding to one sample and the frequency domain position of the reference signal corresponding to one sample adjacent to the frequency domain position is smaller than N frequency domain units; Or, in a sample group, a frequency domain position corresponding to one sample and a frequency domain position corresponding to one sample adjacent to the frequency domain position are adjacent frequency domain positions; Alternatively, in a sample group, the difference between a frequency domain position corresponding to one sample and a frequency domain position corresponding to one sample adjacent to the frequency domain position is less than N frequency domain units.

[0127] In this embodiment, the frequency domain unit may be a subcarrier interval, a PRB, a subband, or a PRB group, etc., where PRB refers to a physical resource block within a BWP (Bandwidth Part), and a PRB group refers to a set of multiple PRBs in NR. The threshold and the preset threshold values ​​may be determined according to actual conditions and are not specifically limited here.

[0128] In one embodiment, when the first dataset includes at least one first sample group and at least one second sample group related to the first sample group, obtaining the second dataset from the first dataset based on the first sample parameters may specifically include obtaining a first threshold group, determining whether the first sample group satisfies a predetermined condition based on the first sample parameters and the first threshold group, and then constructing the second dataset with samples in all the first sample groups and their corresponding second sample groups that satisfy the predetermined condition.

[0129] If the first sample group does not satisfy the predetermined condition, all the second sample groups related to the first sample group are discarded. The first sample group may be an observation window or a prediction window, and is not specifically limited here.

[0130] For example, if a first sample group corresponds to an observation window, if the first sample group does not satisfy a preset condition, all samples in the first sample group and the corresponding second sample group are discarded; or if the first sample group satisfies the preset condition, all samples in the first sample group and the corresponding second sample group are included in the second data set.

[0131] Similarly, if a first sample group corresponds to a prediction window, if the first sample group does not satisfy a predetermined condition, all samples in the first sample group and the corresponding second sample group are discarded, or if the first sample group satisfies the predetermined condition, all samples in the first sample group and the corresponding second sample group are included in the second data set.

[0132] For example, if a first sample group corresponds to an observation window and a second sample group corresponds to a prediction window, if either the first sample group does not satisfy a predetermined condition or the second sample group does not satisfy a predetermined condition, all samples in the first sample group and the corresponding second sample group are discarded; or if both the first sample group and the second sample group satisfy a predetermined condition, all samples in the first sample group and the corresponding second sample group are included in the second dataset.

[0133] In one embodiment, the first sample group includes K samples, and determining whether the first sample group satisfies a predetermined condition based on the first sample parameter and the first threshold group includes determining whether the sample group satisfies a predetermined condition based on the first sample parameter corresponding to at least one sample of the sample group and the first threshold group, the predetermined condition including at least one of: a number of valid samples in the sample group being greater than or equal to K; a ratio of valid samples in the sample group to a total number K of samples in the sample group being greater than or equal to a first predetermined ratio value; a frequency-domain position interval between a sample in the sample group and a sample adjacent to the frequency-domain position being less than N frequency-domain units; or at least K valid samples being present in a first predetermined frequency-domain segment before a sample in the sample group; or at least K valid samples being present in a second predetermined frequency-domain segment after a sample in the sample group.

[0134] Similarly, the relevant interpretation of each condition in the embodiments of the present application can be referred to the relevant content in one embodiment of determining whether the above-mentioned one sample group satisfies the preset condition, and the explanation will be omitted here.

[0135] In one embodiment, obtaining the second data set from the first data set based on the second sample parameters may specifically include sorting the samples based on at least one parameter of the second sample parameters, and obtaining the second data set from the sorted samples, where the second sample parameters include at least one parameter of a signal-to-interference-plus-noise ratio, a reference signal received power, a channel rank, a modulation and decoding scheme, a channel quality indicator, a bit error rate, a block error rate, and a mobile speed.

[0136] After sorting, the higher the sample, the higher the quality. In addition, if the second sample parameters include multiple parameters, the samples may be sorted based on the priority of different parameters. The priority of the signal-to-interference-plus-noise ratio, reference signal received power, channel rank, modulation and decoding scheme, channel quality indicator, bit error rate, block error rate, and moving speed in the second sample parameters may be set according to actual situations, and is not specifically limited herein.

[0137] In one embodiment, sorting the samples based on at least one parameter of the second sample parameters and obtaining the second data set from the sorted samples may specifically comprise sorting the samples in the first data set based on a first parameter of the second sample parameters and obtaining the second data set from the sorted samples.

[0138] For example, if the first parameter is a signal-to-interference-plus-noise ratio, the samples in the first data set can be sorted based on the signal-to-interference-plus-noise ratio corresponding to the samples, with samples having higher signal-to-interference-plus-noise ratios being ranked higher, and the top L1 samples can be selected from the sorted samples and used as a second data set.

[0139] The first parameter may be a reference signal received power, a channel rank, a modulation and decoding scheme, a channel quality indicator, or a moving speed, etc., and is not specifically limited here.

[0140] In one embodiment, sorting the samples based on at least one parameter of the second sample parameters and obtaining the second data set from the sorted samples may specifically include sorting the samples in the first data set based on a first parameter of the second sample parameters, and then sorting the samples having the same first parameter by a second parameter of the second sample parameters, and obtaining the second data set from the sorted samples.

[0141] The first parameter has a higher priority than the second parameter.

[0142] For example, if the first parameter is a signal-to-interference-plus-noise ratio and the second parameter is a reference signal received power, then the samples in the first data set are sorted based on the signal-to-interference-plus-noise ratio in the second sample parameter, and then the samples with the same signal-to-interference-plus-noise ratio are sorted by the reference signal received power in the second sample parameter, and the top L1 samples from the last sorted sample are obtained, and these L1 samples can be used as the second data set.

[0143] In one embodiment, sorting the samples based on at least one parameter of the second sample parameters and obtaining the second data set from the sorted samples may specifically include sorting the samples in the first data set based on a first parameter of the second sample parameters, then sorting the samples with the same first parameter by a second parameter of the second sample parameters, and then sorting the samples with the same second sample parameter by a third parameter of the second sample parameters, and obtaining the second data set from the sorted samples.

[0144] The first parameter, second parameter and third parameter have successively lower priorities.

[0145] For example, if the first parameter is the reference signal received power, the second parameter is the channel rank, and the third parameter is the modulation and decoding scheme, then the samples in the first data set can be sorted based on the reference signal received power in the second sample parameter, and then the samples with the same reference signal received power can be sorted by the channel rank in the second sample parameter, and then the samples with the same channel rank can be sorted by the modulation and decoding scheme in the second sample parameter, and a second data set can be obtained from the sorted samples.

[0146] In one embodiment, sorting the samples based on at least one of the second sample parameters and obtaining the second data set from the sorted samples may specifically include sorting the samples in the first data set based on a first parameter of the second sample parameters, then sorting samples with the same rank by employing at least one second parameter different from the first parameter of the second sample parameters, then sorting samples with the same rank by employing at least one third parameter different from the first and second parameters of the second sample parameters, and so on, until the top L1 samples are determined, and obtaining the second data set from the sorted samples accordingly.

[0147] In one embodiment, obtaining the second dataset from the first dataset based on the first sample parameters and the second sample parameters may specifically include obtaining a first threshold group, then obtaining a target dataset from the first dataset based on the first sample parameters corresponding to at least one sample and the first threshold group, and finally sorting the samples in the target dataset based on at least one parameter of the second sample parameters, and obtaining the second dataset from the sorted samples.

[0148] For example, if the first dataset includes L0 samples, first obtain a first threshold group, then obtain L3 target datasets from the first dataset based on the first sample parameters and the first threshold group, and then sort the samples in the L3 target datasets based on at least one parameter among the second sample parameters, and obtain L1 top-ranked samples from the last sorted sample, and compose these L1 samples into a second dataset, where L0>=L3 and L3>=L1.

[0149] In one embodiment, the first sample parameters may include at least one of a moving speed, a reference signal received power RSRP, a signal-to-interference-and-noise ratio SINR, a modulation and coding scheme MCS, a channel quality indicator CQI, a reliability, a predicted probability, a cosine similarity, a measurement time, an inference time, location parameter information, beam parameter information, throughput, a bit error rate, and a block error rate.

[0150] Furthermore, in the data collection stage, in order to simultaneously collect a large number of samples, the second communication device (e.g., a base station) may configure a reference signal resource set to be transmitted periodically and transmit the same reference signal resource set in different frequency regions, and the first communication device (e.g., a terminal) can obtain multiple frequency-domain parameter sets after measuring the reference signal resource set respectively. Since the first communication device (e.g., a terminal) may simultaneously feed back frequency-domain parameter sets (i.e., measurement results of reference signal resource sets, i.e., sets of samples) measured based on multiple frequency regions in one report, it is necessary to report the frequency-domain information of each sample implicitly or explicitly.

[0151] Based on this, in one embodiment, when transmitting the second data set, frequency domain information corresponding to at least one sample in the second data set may be transmitted, i.e., the second data set and the frequency domain information of at least one sample in the second data set may be transmitted simultaneously, where the frequency domain information includes a transmission frequency domain location and / or a bitmap (i.e., bitmap) of a reference signal resource corresponding to the at least one sample, and the bitmap is used to indicate whether the sample on the frequency domain location is valid or to be transmitted.

[0152] 6, if a sample is discarded due to poor quality (i.e., the sample is invalid), its frequency domain information feedback is 0. If a sample is reported (i.e., the sample is transmitted), its frequency domain information feedback is 1.

[0153] Referring to Figure 7, Figure 7 is a flowchart of a method for receiving a data set according to an embodiment of the present application, which may be used in a second communication device, such as the network side device shown in Figure 1. The method for receiving a data set includes, but is not limited to, step S210.

[0154] Step S210: Receive a second data set.

[0155] Here, the second data set includes L1 samples, and the second data set is obtained by the first communication device from the first data set based on the first sample parameters and / or the second sample parameters, and the first data set includes L0 samples.

[0156] In this embodiment, by adopting the dataset transmission method including the above step S210, the embodiment of the present application can receive a second dataset, where the second dataset includes L1 samples, and the second dataset is obtained by the first communication device from the first dataset based on the first sample parameters and / or the second sample parameters, and the first dataset includes L0 samples, that is, the second communication device can receive a high-quality or meet-requirements second dataset screened from the first dataset by the first sample parameters and / or the second sample parameters from the first communication device, thereby improving the model training effect and further improving the model performance.

[0157] In one embodiment, the second communication device may transmit the first threshold group to the first communication device, whereby the first communication device obtains a second data set from the first data set based on a first sample parameter corresponding to at least one sample and the first threshold group, or the first communication device obtains a target data set from the first data set based on a first sample parameter corresponding to at least one sample and the first threshold group, sorts samples in the target data set based on at least one of the second sample parameters, and obtains a second data set from the sorted samples.

[0158] In one embodiment, the second communication device may transmit a target number of feedback samples L1 to the first communication device, whereby the first communication device sorts the samples based on at least one of the second sample parameters and obtains a second data set from the sorted samples.

[0159] In one embodiment, sorting the samples based on at least one parameter of the second sample parameters and obtaining the second data set from the sorted samples comprises: sorting the samples in the first data set based on a first one of the second sample parameters and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on a first parameter of the second sample parameters, and then sorting the samples with the same first parameter by a second parameter of the second sample parameters, and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on a first parameter among the second sample parameters, then sorting the samples with the same first parameter by a second sample parameter among the second sample parameters, and then sorting the samples with the same second sample parameter by a third sample parameter among the second sample parameters, and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on a first one of the second sample parameters, and then sorting samples with the same ranking by employing at least one target parameter different from the first one of the second sample parameters, until the top L1 samples are determined, and obtaining a second data set from the sorted samples.

[0160] In one embodiment, the first sample parameters may include at least one of a moving speed, a reference signal received power RSRP, a signal-to-interference-and-noise ratio SINR, a modulation and coding scheme MCS, a channel quality indicator CQI, a reliability, a predicted probability, a cosine similarity, a measurement time, an inference time, location parameter information, beam parameter information, throughput, a bit error rate, and a block error rate.

[0161] In one embodiment, the second sample parameters include at least one of the following parameters: signal-to-interference-plus-noise ratio, reference signal received power, channel rank, modulation and decoding scheme, channel quality indicator, mobile speed, bit error rate, and block error rate.

[0162] In one embodiment, the second data set includes at least one sample group, where the sample group includes K frequency-related samples. Because the first data set was screened in sample groups, the samples within the sample groups in the second data set are also frequency-related and will not be described here.

[0163] In one embodiment, when receiving the second data set, frequency domain information corresponding to at least one sample in the second data set may also be received.

[0164] In one embodiment, the frequency domain information includes a transmission frequency domain location of a reference signal resource corresponding to at least one sample and / or a bitmap (i.e., bitmap) used to indicate whether a sample on a frequency domain location is valid or to be transmitted.

[0165] The method for receiving the data set on the second communication device side is similar to the method for transmitting the data set on the first communication device side, and therefore the description thereof will be omitted here.

[0166] The base station may combine the second data collected one or more times into a larger dataset and store the dataset locally to be used for training, fine-tuning or testing by the base station itself, or may send the collected dataset to another third party server or base station or terminal to be used for training or fine-tuning the model.

[0167] In one embodiment, one communication node (i.e., a communication device) transmits functionality or a functionality index to another communication node. For example, a base station transmits a functionality or a functionality index to a terminal (i.e., a second communication device) to notify the terminal (i.e., a first communication device) that the functionality can process information. Here, the functionality may also be referred to as a functional module, functional function, functional mapping, etc., and is used to describe the characteristics or type of an information processing method. The types of information processing methods include multiple types, such as positioning, beam management, CSI prediction, beam prediction, channel estimation, etc. The characteristics of the information processing method include, but are not limited to, a scene description of function adaptation, an input parameter description, an output parameter description, and whether the output result is a measurement parameter. Here, one function corresponds to one or more information processing methods, and each information processing method may be realized by one or more models, or one function may be realized by one or more models.

[0168] In one embodiment, a base station (i.e., a second communication device) can process information using a function. The base station transmits a reference signal to a terminal (i.e., a first communication device). After receiving the reference signal, the terminal can obtain channel matrix information H for multiple physical resource blocks based on the received reference signal. In one example, the terminal can obtain a feature vector V or a quantized value of the feature vector V for each subband based on the channel matrix information, where V is an Nt*Ns complex matrix, where Nt and Ns represent the number of transmit antennas and the rank of the channel matrix, respectively. To efficiently feed back V or H for multiple subbands, an encoder needs to compress V or H and feed back the channel state information compressed by the encoder. The base station can recover V or H to some extent by receiving the channel state information and decompressing it using a decoder at the base station side. However, the encoder and decoder need to train or fine-tune samples to achieve good performance. In this process, the higher the quality of the sample labels or the similarity to the ideal V or H, the better. However, high-precision, high-quality, or required labels require a relatively large number of quantization bits and a very large feedback overhead. However, the channel quality (e.g., SINR, CQI, RSRP, etc.) of different users varies. For example, some users are located at the center of a cell and have relatively high channel quality, while others are located at the edge of a cell and experience relatively large interference, resulting in poor channel quality and poor quality labels.

[0169] Of course, in one embodiment, the label may be at least one of the position coordinates or other position parameter information in the positioning, such as TOA, RSTD, LoS / NLoS indicator and angle information.

[0170] In one embodiment, the label may be one of the beam metric parameters in beam management, such as L1-SINR, L1-RSRP, CRI, SSBRI, etc.

[0171] In one embodiment, it is also very important that the features corresponding to the samples are of high quality or meet the requirements, as if a feature is affected by interference or noise, or by the quantization accuracy of the device that collected it, the accuracy of the device measurement, etc., this will affect the accuracy of training the model.

[0172] Also, referring to FIG. 8, an embodiment of the present application further provides a communication device 200, which includes at least one processor 201 and at least one memory 202, and the memory 202 is used to store at least one program.

[0173] The processor 201 and the memory 202 may be connected by a bus or in other ways.

[0174] The memory 202 may store non-transitory software programs and non-transitory computer-executable programs as a non-transitory computer-readable storage medium. The memory 202 may include high-speed random access memory and may further include non-transitory memory, such as at least one magnetic disk memory device, flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory 202 may optionally include memory remotely located relative to the processor 201, and the remote memory may be connected to the processor 201 via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0175] Non-transitory software programs and instructions necessary for the data set transmission method realizing the above embodiment are stored in memory 202 and, when executed by processor 201, perform the data set transmission method of the above embodiment, for example, perform steps S110 to S130 of the method in Figure 2 described above, or perform the data set reception method of the above embodiment, for example, perform step S210 of the method in Figure 7 described above.

[0176] The above-described device embodiments are merely illustrative, and the units described herein as separate components may or may not be physically separated, i.e., may be located in one place or distributed across multiple network units, and some or all of the modules may be selected to achieve the objectives of the solutions of the present embodiments according to actual needs.

[0177] In addition, an embodiment of the present application further provides a computer-readable storage medium, on which computer-executable instructions are stored. The computer-executable instructions are executed by a processor or controller, for example, by a processor in the above-mentioned device embodiment, causing the processor to perform the method for transmitting a dataset in the above-mentioned embodiment, and to perform steps S110 to S130 of the method in FIG. 2 described above, or to perform the method for receiving a dataset in the above-mentioned embodiment, and for example, to perform step S210 of the method in FIG. 7 described above.

[0178]

[0023] Note that an embodiment of the present application further provides a computer program product including a computer program or computer instructions, wherein the computer program or computer instructions is stored in a computer-readable storage medium, and a processor of the computer device reads the computer program or computer instructions from the computer-readable storage medium, and the processor executes the computer program or computer instructions, thereby causing the computer device to perform the method for transmitting a dataset in the above embodiment, for example, to perform steps S110 to S130 of the method in Figure 2 described above, or to perform the method for receiving a dataset in the above embodiment, for example, to perform step S210 of the method in Figure 7 described above.

[0179] All or some of the steps in the methods and systems disclosed above may be implemented as software, firmware, hardware, or any suitable combination thereof. Any or all physical components may be implemented as software executed by a processor, such as a central processor, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on computer-readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As is well known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (e.g., computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital multifunction disks (DVDs) or other disk storage, magnetic cartridges, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium used to store desired information and accessible by a computer. As is well known to those skilled in the art, communication media typically includes computer-readable instructions, data structures, program modules or other data in a modulated data signal, such as a carrier or other transport mechanism, and may include any information delivery media.

Claims

1. 1. A method for transmitting a data set for use in a first communication device, comprising: obtaining a first data set comprising L0 samples; obtaining a second data set including L1 samples from the first data set based on the first sample parameters and / or the second sample parameters; transmitting the second data set; wherein L0 and L1 are positive integers, and L1 is less than or equal to L0.

2. Obtaining a second data set from the first data set based on first sample parameters includes: Obtaining a first threshold group; deriving a second data set from the first data set based on the first sample parameters corresponding to at least one sample and the first threshold group.

3. Obtaining a second data set from the first data set based on second sample parameters includes:

10. The method of claim 1, further comprising sorting the samples based on at least one of second sample parameters and obtaining a second data set from the sorted samples.

4. Sorting the samples based on at least one parameter of second sample parameters and obtaining a second data set from the sorted samples may include: sorting the samples in the first data set based on a first one of the second sample parameters and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on a first parameter of the second sample parameters, and then sorting the samples with the same first parameter by a second parameter of the second sample parameters, thereby obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on the first parameter of the second sample parameters, then sorting the samples with the same first parameter by a second parameter of the second sample parameters, and then sorting the samples with the same second parameter by a third parameter of the second sample parameters, and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on the first parameter of the second sample parameters, and then sorting the samples of the same rank by employing at least one second parameter different from the first parameter of the second sample parameters until a top L1 number of samples are determined, and obtaining a second data set from the sorted samples.

5. Obtaining a second data set from the first data set based on a first sample parameter and a second sample parameter includes: Obtaining a first threshold group; obtaining a target data set from the first data set based on the first sample parameter corresponding to at least one of the samples and the first threshold group; and sorting the samples in the target dataset based on at least one parameter of second sample parameters, and obtaining a second dataset from the sorted samples.

6. The first sample parameter is:

2. The method of claim 1, comprising at least one of a moving speed, a reference signal received power RSRP, a signal-to-interference-and-noise ratio SINR, a modulation and coding scheme MCS, a channel quality indicator CQI, a reliability, a predicted probability, a cosine similarity, a measurement time, an inference time, location parameter information, beam parameter information, throughput, a bit error rate, and a block error rate.

7. The second sample parameter is:

2. The method of claim 1, comprising at least one of a signal-to-interference-and-noise ratio, a reference signal received power, a channel rank, a modulation and decoding scheme, a channel quality indicator, a mobile speed, a bit error rate, and a block error rate.

8. The method of claim 1 , wherein the first data set includes at least one sample group, wherein the sample group includes K samples associated with a frequency domain.

9. Obtaining a second data set from the first data set based on first sample parameters includes: Obtaining a first threshold group; determining whether a sample group satisfies a predetermined condition based on a first sample parameter corresponding to at least one of the samples of the sample group and the first threshold group; The method of claim 8 , further comprising: including the samples of the group of samples that satisfy a predetermined condition in the second data set.

10. In determining whether a sample group satisfies a predetermined condition based on a first sample parameter corresponding to at least one of the samples of the sample group and the first threshold value group, the predetermined condition is: the number of valid samples in the sample group is equal to or greater than K1; Or, the ratio of valid samples in the sample group to the total number K of samples in the sample group is equal to or greater than a first preset ratio value; Alternatively, the interval between a frequency domain position corresponding to one of the samples in the sample group and a frequency domain position corresponding to a sample adjacent to the frequency domain position is smaller than N frequency domain units; or there are at least K2 valid samples in a first predetermined frequency domain segment prior to said sample of one of said sample groups; or there being at least K valid samples in a second predetermined frequency domain segment after the sample of one of the sample groups; 10. The method of claim 9, wherein K, K1, K2, and K3 are positive integers, and K1, K2, and K3 are less than or equal to K.

11. the sample includes a first sub-parameter set and / or a second sub-parameter set; the first sub-parameter set corresponding to at least one sample in the sample group satisfies a frequency domain constraint relationship; Alternatively, the second sub-parameter set corresponding to at least one sample in the sample group satisfies a frequency domain constraint relationship; Alternatively, the first sub-parameter set and the second sub-parameter set corresponding to at least one sample in the sample group satisfy a frequency domain constraint relationship.

12. The frequency domain constraint relationship is In the sample group, a frequency domain position corresponding to a first sub-parameter set of one sample and a frequency domain position corresponding to a first sub-parameter set of one sample adjacent to the frequency domain position are adjacent frequency domain positions; Alternatively, in the sample group, the frequency domain position corresponding to the second sub-parameter set of one of the samples and the frequency domain position corresponding to the second sub-parameter set of one of the samples adjacent to the frequency domain position are adjacent frequency domain positions; Or, in the sample group, the interval between the frequency domain position of the first sub-parameter set of one of the samples and the frequency domain position of the first parameter subset of one of the samples adjacent to the frequency domain position is smaller than N frequency domain units; Or, in the sample group, the interval between the frequency domain position of the second sub-parameter set of one of the samples and the frequency domain position of the second parameter subset of one sample adjacent to the frequency domain position is smaller than N frequency domain units; Alternatively, in the sample group, a frequency domain position of a reference signal corresponding to one of the samples and a frequency domain position of a reference signal corresponding to one of the samples adjacent to the frequency domain position are adjacent frequency domain units; Alternatively, in the sample group, a difference between a frequency domain position of a reference signal corresponding to one of the samples and a frequency domain position of a reference signal corresponding to one of the samples adjacent to the frequency domain position is smaller than N frequency domain units; Alternatively, in the sample group, a frequency domain position corresponding to one of the samples and a frequency domain position corresponding to another of the samples adjacent to the frequency domain position are adjacent frequency domain positions; or wherein, in the sample group, a difference between a frequency domain position corresponding to one of the samples and a frequency domain position corresponding to one of the samples adjacent to that frequency domain position is less than N frequency domain units.

13. The first data set includes at least one first sample group and at least one second sample group related to the first sample group, and obtaining the second data set from the first data set based on first sample parameters includes: Obtaining a first threshold group; determining whether the first sample group satisfies a predetermined condition based on the first sample parameters and the first threshold group; and constructing the second data set with all samples in the first sample group and corresponding second sample groups that satisfy a predetermined condition.

14. The first sample group includes K samples, and determining whether the first sample group satisfies a predetermined condition based on the first sample parameters and the first threshold group includes determining whether the first sample group satisfies a predetermined condition based on the first sample parameters and the first threshold group, the predetermined condition being: the number of valid samples in the first sample group is greater than or equal to K1; or the ratio of valid samples in the first sample group to the total number K of samples in the first sample group is equal to or greater than a first preset ratio value; or a frequency domain position interval between one sample in the first sample group and a sample adjacent to the sample in the first sample group is smaller than N frequency domain units; or there are at least K valid samples in a first predetermined frequency domain segment prior to one of the samples of the first sample group; or there are at least K valid samples in a second predetermined frequency domain segment after one of the samples of the first sample group; 14. The method of claim 13, wherein K, K1, K2, and K3 are positive integers, and K1, K2, and K3 are less than or equal to K.

15. The frequency domain unit is The method of claim 10 or 14, comprising at least one of the following: subcarrier spacing, number of resource blocks, number of subbands, and number of center frequency bands.

16. transmitting the second data set comprises: The method of claim 1 , comprising transmitting frequency domain information corresponding to the second data set and / or at least one sample in the second data set.

17. The frequency domain information is a transmit frequency domain location of a reference signal resource corresponding to the at least one sample; 17. The method of claim 16, further comprising at least one of: a bitmap for indicating whether a sample on the frequency domain location is valid or not, or whether it is to be transmitted.

18. 1. A method for receiving a data set for use in a second communication device, comprising: receiving a second data set comprising L1 samples; the second data set is obtained by a first communication device based on first sample parameters and / or second sample parameters from a first data set comprising L0 samples; wherein L0 and L1 are positive integers, and L1 is less than or equal to L0.

19. The method of claim 18 , further comprising transmitting the first threshold group.

20. the second data set is obtained by a first communications device from the first data set based on first sample parameters; obtaining the second data set from the first data set based on the first sample parameter corresponding to at least one of the samples and a first threshold group; Alternatively, the method of claim 18 further comprises obtaining a target dataset from the first dataset based on the first sample parameter corresponding to at least one of the samples and the first threshold group, sorting the samples in the target dataset based on at least one parameter of second sample parameters, and obtaining a second dataset from the sorted samples.

21. Sorting the samples in the target dataset based on at least one parameter of second sample parameters and obtaining a second dataset from the sorted samples may include: sorting the samples in the first data set based on a first one of the second sample parameters and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on a first parameter of the second sample parameters, and then sorting the samples having the same first parameter by a second parameter of the second sample parameters, thereby obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on the first parameter of the second sample parameters, then sorting the samples having the same first parameter by a second sample parameter of the second sample parameters, and then sorting the samples having the same second sample parameter by a third sample parameter of the second sample parameters, and obtaining a second data set from the sorted samples; or sorting the samples in the first data set based on the first one of the second sample parameters, and then sorting the samples of the same rank by employing at least one target parameter different from the first one of the second sample parameters until a top L1 number of samples are determined, and obtaining a second data set from the sorted samples.

22. The first sample parameter is:

20. The method of claim 18, comprising at least one of a moving speed, a reference signal received power RSRP, a signal-to-interference-and-noise ratio SINR, a modulation and coding scheme MCS, a channel quality indicator CQI, a reliability, a predicted probability, a cosine similarity, a measurement time, an inference time, location parameter information, beam parameter information, throughput, a bit error rate, and a block error rate.

23. The second sample parameter is:

20. The method of claim 18, comprising at least one of a signal-to-interference-and-noise ratio, a reference signal received power, a channel rank, a modulation and decoding scheme, a channel quality indicator, a mobile speed, a bit error rate, and a block error rate.

24. 20. The method of claim 18, wherein the second data set includes at least one sample group, wherein the sample group includes the samples associated with K frequencies.

25. Receiving the second data set includes:

19. The method of claim 18, comprising receiving frequency domain information of the second data set and / or at least one of the samples in the second data set.

26. The frequency domain information is a transmit frequency domain location of a reference signal resource corresponding to the sample; and and a bitmap for indicating whether a sample on the frequency domain location is valid or not, or whether it is to be transmitted.

27. A communication device, at least one processor; at least one memory for storing at least one program; 27. A communications device, wherein at least one of the programs, when executed by at least one of the processors, implements the method for transmitting a data set according to any one of claims 1 to 17 or the method for receiving a data set according to any one of claims 18 to 26.

28. 27. A computer-readable storage medium having a processor-executable program stored thereon, the computer-readable storage medium being used to implement the method for transmitting a dataset according to any one of claims 1 to 17 or the method for receiving a dataset according to any one of claims 18 to 26 when the processor-executable program is executed by the processor.

29. 27. A computer program product comprising a computer program or computer instructions, the computer program or the computer instructions being stored on a computer-readable storage medium, the computer program or the computer instructions being read by a processor of a computing device from the computer-readable storage medium, the processor executing the computer program or the computer instructions to cause the computing device to perform the method for transmitting a dataset according to any one of claims 1 to 20 or the method for receiving a dataset according to any one of claims 18 to 26.

Images