Communication method, terminal equipment and network equipment

Abstract

The invention provides a communication method, terminal equipment and network equipment. The method comprises: a terminal device sends a monitoring report of a model to a network device, the model being used for predicting CSI, the monitoring report comprising a performance index determined based on CSI prediction data and a tag, the position of the prediction instance corresponding to the CSI prediction data and the position of the first reference signal resource corresponding to the tag have a predetermined time relationship. In the application, a terminal device sends a monitoring report of a model to a network device, the monitoring report comprises a performance index determined based on CSI prediction data and a label and / or a performance index determined based on the CSI prediction data and CSI measurement data in a prediction window, and due to the position of a prediction instance corresponding to the CSI prediction data, the prediction accuracy of the model is improved. According to the embodiment of the invention, a predetermined time relationship exists between the position of the first reference signal resource corresponding to the label and / or the CSI measurement data in the prediction window, and the time relationship enables the CSI prediction data to be matched with the appropriate label, thereby obtaining an accurate performance index, and achieving the monitoring of the performance of the model.

Description

Technical Field

[0001] This application relates to the field of communication technology, and more specifically, to a communication method, terminal equipment, and network equipment. Background Technology

[0002] Introducing models into communication systems can effectively improve communication performance. For example, channel state information (CSI) feedback based on artificial intelligence (AI) / machine language (ML) models has been proposed. When performing CSI prediction based on models, terminal devices need to utilize CSI reference signal (CSI-RS) measurement data for model training, CSI prediction, and model monitoring. This involves model input data, CSI prediction data, and ground truth labels. How to apply this data for model monitoring becomes a problem that needs to be solved. Summary of the Invention

[0003] This application provides a communication method, a terminal device, and a network device. The various aspects covered by this application are described below.

[0004] In a first aspect, a communication method is provided, comprising: a terminal device sending a monitoring report of a model to a network device, wherein the model is used to predict CSI, the monitoring report includes performance indicators determined based on CSI prediction data and labels, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window, wherein the location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of a first reference signal resource corresponding to the labels and / or the CSI measurement data within the prediction window.

[0005] In a second aspect, a communication method is provided, comprising: a network device sending a monitoring report of a model to a terminal device, wherein the model is used to predict CSI, the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window, wherein the location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of a first reference signal resource corresponding to the tags and / or the CSI measurement data within the prediction window.

[0006] Thirdly, a terminal device is provided, comprising: a transceiver module for sending a monitoring report of a model to a network device, wherein the model is used to predict CSI, and the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window, wherein the location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of the first reference signal resource corresponding to the tags and / or the CSI measurement data within the prediction window.

[0007] Fourthly, a network device is provided, comprising: a transceiver module for sending a monitoring report of a model to the terminal device, wherein the model is used to predict CSI, and the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window, wherein the location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of a first reference signal resource corresponding to the tags and / or the CSI measurement data within the prediction window.

[0008] Fifthly, a terminal device is provided, including a transceiver, a memory, and a processor, wherein the memory is used to store a program, and the processor is used to invoke the program in the memory and control the transceiver to receive or send signals so that the terminal device performs the method as described in the first aspect.

[0009] In a sixth aspect, a network device is provided, including a transceiver, a memory, and a processor, wherein the memory is used to store a program, and the processor is used to invoke the program in the memory and control the transceiver to receive or transmit signals so that the network device performs the method as described in the second aspect.

[0010] A seventh aspect provides an apparatus including a processor for calling a program from a memory to cause the apparatus to perform the method as described in the first or second aspect.

[0011] Eighthly, a chip is provided, including a processor for calling a program from memory to cause a device having the chip mounted to perform the method as described in the first or second aspect.

[0012] Ninth aspect, a computer-readable storage medium is provided having a program stored thereon that causes a computer to perform the method as described in the first or second aspect.

[0013] A tenth aspect provides a computer program product, including a program that causes a computer to perform the method as described in the first or second aspect.

[0014] Eleventhly, a computer program is provided that causes a computer to perform the method as described in the first or second aspect.

[0015] In this embodiment, the terminal device sends a model monitoring report to the network device. The monitoring report includes performance indicators determined based on CSI prediction data and labels, and / or performance indicators determined based on CSI prediction data and CSI measurement data within the prediction window. Since there is a predetermined time relationship between the location of the prediction instance corresponding to the CSI prediction data and the location of the first reference signal resource corresponding to the label and / or the CSI measurement data within the prediction window, the time relationship enables the CSI prediction data to be matched with appropriate labels, thereby obtaining accurate performance indicators and realizing the monitoring of model performance. Attached Figure Description

[0016] Figure 1 This is a system architecture example diagram of a communication system applicable to embodiments of this application.

[0017] Figure 2 This is a flowchart illustrating the communication method according to an embodiment of this application.

[0018] Figure 3 This is an example of the time relationship between the first reference signal and the first reference signal resource.

[0019] Figure 4 This is an example of the time relationship between the first reference signal and the first reference signal resource.

[0020] Figure 5 This is an example of the time relationship between the first reference signal and the first reference signal resource.

[0021] Figure 6 This is an example of the time relationship between the first reference signal and the first reference signal resource.

[0022] Figure 7 This is a diagram illustrating the alignment of predicted instances with the start of the time window.

[0023] Figure 8 This is a diagram illustrating the alignment of predicted instances with the end position of the time window.

[0024] Figure 9 This is a schematic diagram of the structure of a terminal device according to an embodiment of this application.

[0025] Figure 10 This is a schematic diagram of the structure of a network device according to an embodiment of this application.

[0026] Figure 11 This is a schematic diagram of a communication apparatus according to an embodiment of this application. Detailed Implementation

[0027] The technical solutions in this application will now be described with reference to the accompanying drawings.

[0028] Communication system

[0029] Figure 1 This is an example diagram of the system architecture of a communication system 100 applicable to embodiments of this application. The communication system 100 may include a network device 110 and a terminal device 120. The network device 110 may be a device that communicates with the terminal device 120. The network device 110 can provide network coverage for a specific geographical area and can communicate with the terminal device 120 located within that coverage area. The terminal device 120 can access a network, such as a wireless network, through the network device 110. Optionally, the communication system 100 may also include other network entities such as a network controller and a mobility management entity; this application embodiment does not limit this.

[0030] It should be understood that the technical solutions of the embodiments of this application can be applied to various communication systems, such as: fifth-generation (5G) systems, new radio (NR), long-term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, LTE time division duplex (TDD) systems, etc. The technical solutions provided in this application can also be applied to future communication systems, such as sixth-generation mobile communication systems, satellite communication systems, and so on.

[0031] In this application embodiment, the terminal device may also be referred to as user equipment (UE), access terminal, user unit, user station, mobile station, mobile station (MS), mobile terminal (MT), remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent, or user apparatus. The terminal device in this application embodiment can be a device that provides voice and / or data connectivity to a user, and can be used to connect people, objects, and machines, such as a handheld device with wireless connectivity, vehicle-mounted device, etc. Terminal devices can also be mobile phones, tablets, laptops, PDAs, mobile internet devices (MIDs), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, self-driving, remote medical surgery, smart grids, transportation safety, smart cities, and smart homes. Optionally, terminal devices can act as base stations. For example, a terminal device can act as a dispatching entity, providing sidelink signals between terminal devices in vehicle-to-everything (V2X) or device-to-device (D2D) systems. For instance, cellular phones and cars communicate with each other using sidelink signals. Cellular phones and smart home devices communicate without relaying communication signals through base stations.

[0032] In this embodiment, the network device can be a device used to communicate with a terminal device. The network device can be an access network device or a wireless access network device. For example, the network device can be a base station. The term "base station" can broadly encompass various names as follows, or can be replaced by names such as: NodeB, evolved NodeB (eNB), next-generation NodeB (gNB), relay station, transmitting and receiving point (TRP), transmitting point (TP), master station (MeNB), secondary station (SeNB), multi-standard wireless (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), positioning node, etc. A base station can be a macro base station, micro base station, relay node, donor node, or similar entity, or a combination thereof. A base station can also refer to a communication module, modem, or chip installed within the aforementioned equipment or apparatus. A base station can also be a mobile switching center, or an entity that performs base station functions in device-to-device (D2D), vehicle-to-everything (V2X), and machine-to-machine (M2M) communications, a network-side device in a 6G network, or an entity that performs base station functions in future communication systems. A base station can support networks using the same or different access technologies. The embodiments of this application do not limit the specific technologies or device forms used in the network equipment. In some deployments, the network equipment may include a CU or a DU; or, the network equipment may include both a CU and a DU. Optionally, the base station may include an AAU.

[0033] Furthermore, base stations can be fixed or mobile. For example, a helicopter or drone can be configured to act as a mobile base station, and one or more cells can move depending on the location of the mobile base station. In other examples, a helicopter or drone can be configured as a device to communicate with another base station.

[0034] Network devices and terminal devices can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; they can also be deployed on water; and they can also be deployed in the air on airplanes, balloons, and satellites. This application does not limit the scenario in which the network devices and terminal devices are located.

[0035] It should be understood that all or part of the functions of the communication device in this application can also be implemented by software functions running on hardware, or by virtualization functions instantiated on a platform such as a cloud platform.

[0036] In 5G and future wireless communication systems (e.g., 6G), the acquisition and prediction of CSI (Content Targeting) is crucial for network performance. In related technologies, terminal devices need to frequently measure CSI, for example, by measuring CSI-RS or SSB, and then feed the measurement data back to the network device. However, in massive MIMO (Massive MIMO) antenna systems or high-frequency millimeter-wave scenarios, the CSI matrix has a large dimension, resulting in extremely high feedback overhead and consuming significant uplink resources. Furthermore, because wireless channels are time-varying, especially in high-speed mobile scenarios, traditional CSI feedback suffers from processing delays. This can lead to outdated CSIs being acquired by the network device, resulting in beam alignment errors and degraded link quality.

[0037] In scenarios such as ultra-dense networks, high-frequency bands, or rapidly changing channels (e.g., high-speed rail, drones), traditional CSI estimation methods struggle to cope with rapidly changing environments. Utilizing AI / ML models, hidden patterns of channel changes can be learned from historical data, enabling more efficient predictions. Terminal devices or network devices predict future CSI using these models, reducing the amount of data required for uplink feedback. For example, only key features or model parameters can be fed back, rather than the complete CSI matrix. The model can learn the long-term statistical characteristics of the channel (e.g., spatial correlation, multipath distribution), reducing the need for frequent measurements. Using the model to capture the temporal correlation of the channel and predict CSI for several future time slots is suitable for high-speed mobile and rapidly time-varying channels. The model can be deployed on terminal devices or network devices, generating prediction results in real time, reducing the processing latency of traditional feedback methods.

[0038] CSI feedback in related technologies relies on linear models, such as minimum mean square error (MMSE). However, by employing models, complex nonlinear channel characteristics can be constructed, and sensor data (e.g., position, velocity, historical channel data) can be combined to dynamically adjust prediction strategies (e.g., switching model parameters). Furthermore, models can compress high-dimensional CSI matrices, reducing computational complexity. High-frequency channels are sensitive to obstruction; models can predict beam jamming and proactively switch to backup beams. Predicting channel abrupt changes allows for advance adjustments to scheduling strategies to meet reliability requirements. High-frequency bands rely on narrow beams, and traditional beam scanning incurs significant overhead. Predicting the movement trajectory of terminal devices through models allows for advance optimization of beam pointing, reducing beam training time and improving handover success rates.

[0039] The core value of model-based CSI prediction lies in reducing signaling overhead, decreasing feedback data volume, improving timeliness, adapting to rapidly changing channels, enhancing adaptability to complex scenarios, supporting CSI feedback in nonlinear, high-dimensional, and dynamic environments, and supporting the needs of future networks (e.g., massive MIMO, millimeter wave, URLLC). Unlike CSI feedback in related technologies, models possess self-iterative learning capabilities, enabling them to continuously optimize their performance through data-driven approaches. The model's growth rate follows a power-law relationship between computational scale and performance; as the number of model parameters increases exponentially and the scale of training data expands, its capabilities exhibit a nonlinear improvement. Furthermore, distributed training frameworks and algorithm optimization further accelerate the model's evolution cycle. With the maturity and standardization of AI technology, model-based CSI prediction will become one of the key enabling technologies for 6G intelligent wireless communication.

[0040] In model-based CSI prediction scenarios, terminal devices need to utilize measurement data from reference signals such as CSI-RS for model training, CSI prediction, and model performance monitoring. The data types involved include: model input data, CSI prediction data obtained based on the model, and ground-truth labels. The model input data comes from the measurement data of reference signals such as CSI-RS within the observation window. This measurement data can be used to predict CSI at future times. The ground-truth labels come from the actual measurement data obtained from the CSI-RS and other reference signals within the prediction window, serving as a benchmark for comparing CSI prediction results.

[0041] To monitor model performance, it's necessary to correlate the CSI prediction data within the prediction window with the actual measurement data used as tags to better compare the deviations between them. If the relationship between the CSI prediction data and the actual measurement data is unclear, causal relationships cannot be monitored. Since the observation window and prediction window transmit CSI-RS at different times, the terminal device needs to know the relationship between the CSI prediction data and the corresponding actual measurement data used as tags. For example, using CSI-RS measurement data from window t1 to t10 to predict the CSI at time t11, the location of the tag corresponding to the CSI prediction data at time t11 needs to be known. Because the wireless channel is time-varying, accurate monitoring of model performance is impossible without determining the relationship between the CSI prediction data and the actual measurement data used as tags. The lack of a correlation between the CSI prediction data within the prediction window and the actual measurement data used as tags may lead to performance degradation due to a mismatch between the model's prediction data and the tags.

[0042] Therefore, in this embodiment of the application, the terminal device sends a model monitoring report to the network device. The monitoring report includes performance indicators determined based on CSI prediction data and labels, and / or performance indicators determined based on CSI prediction data and CSI measurement data within the prediction window. Since there is a predetermined time relationship between the location of the prediction instance corresponding to the CSI prediction data and the location of the first reference signal resource corresponding to the label and / or the CSI measurement data within the prediction window, the time relationship enables the CSI prediction data to be matched with appropriate labels, thereby obtaining accurate performance indicators and realizing the monitoring of model performance.

[0043] In this application embodiment, the "model" refers to, for example, a module, algorithm, or filter used to implement a specific function. For instance, the model may include one or more of the following: a CSI compression model, a CSI generator, a CSI module, a CSI algorithm, or a CSI filter. The term "model" can also be replaced with terms such as "function." This application embodiment does not limit the function of the model; for example, the model can be used for CSI prediction, beam management, etc. Below, using a CSI prediction model as an example, the technical solution of this application embodiment is described. This CSI prediction model can also be called a CSI feedback model; for example, the model can be used to perform CSI generation, compression, quantization, and other operations.

[0044] The following is combined with Figure 2 The embodiments of this application will be described in detail below.

[0045] Figure 2 This is a flowchart illustrating the communication method provided in an embodiment of this application. Figure 2The method 200 shown can be performed by a terminal device and a network device. The terminal device may be, for example, a terminal device... Figure 1 The terminal device 120 shown may be, for example, a network device. Figure 1 The network device 110 shown is shown. Figure 2 The method 200 shown may include some or all of the following steps.

[0046] In step 210, the terminal device sends a monitoring report of the model to the network device.

[0047] Correspondingly, the network device receives monitoring reports sent by the terminal device.

[0048] The model is used to predict CSI (Conditional Skin Indicators). The monitoring report includes performance metrics determined based on CSI prediction data and labels, and / or performance metrics determined based on CSI prediction data and CSI measurement data within the prediction window. These performance metrics may include, for example, squared generalized cosine similarity (SGCS), normalized mean squared error (NMSE), or mean squared error (MSE).

[0049] This model can predict the CSI corresponding to the predicted instance within the prediction window based on the measurement data of the reference signal within the observation window, thus obtaining the corresponding CSI prediction data. To monitor the model's performance, it is necessary to compare the CSI prediction data corresponding to the predicted instance with the label, or compare the CSI prediction data with the CSI measurement data within the prediction window, to obtain the corresponding performance indicators. This requires knowing the correlation between the CSI prediction data and the label, and / or between the CSI prediction data and the CSI measurement data within the prediction window, so as to obtain the corresponding performance indicators based on the matching CSI prediction data and label, and / or the matching CSI prediction data and the CSI measurement data within the prediction window. The performance indicators mentioned below can refer to performance indicators determined based on CSI prediction data and the label, or performance indicators determined based on CSI prediction data and the CSI measurement data within the prediction window, or a combination of both.

[0050] In this embodiment, the location of the prediction instance corresponding to the CSI prediction data has a predetermined temporal relationship with the location of the first reference signal resource corresponding to the CSI measurement data within the label and / or prediction window. Therefore, based on this temporal relationship, it can be determined which prediction instance matches the CSI measurement data on which reference signal resource within the prediction window, in order to obtain a performance index (or performance metric, performance value, etc.) that accurately reflects the model performance.

[0051] Optionally, the terminal device may send information to the network device regarding the temporal relationship between the CSI prediction data obtained based on the model and the CSI measurement data within the prediction window, and / or the temporal relationship between the CSI prediction data and the tags. Optionally, the network device may configure the location of the prediction instance. Based on the location of the prediction instance and the temporal relationship, the network device may determine the tag or CSI measurement data corresponding to the prediction instance.

[0052] Monitoring the performance of the model used for CSI prediction involves two processes: inference and monitoring. The inference process, for example, refers to the terminal device measuring reference signals on reference signal resources (i.e., the second reference signal resource) configured within the observation window and generating inference results. These inference results might be CSI prediction data corresponding to the prediction instance within the prediction window, such as the predicted channel matrix and / or eigenvectors. The monitoring process, for example, refers to the terminal device measuring reference signals on reference signal resources (i.e., the first reference signal resource) corresponding to the prediction instance within the prediction window (or, may also include related data processing operations) to generate tags corresponding to the prediction instance and / or CSI measurement data within the prediction window. The reference signals may include, for example, CSI-RS or SSB reference signals, also referred to as pilot signals.

[0053] Terminal devices can report inference reports obtained during the inference process and monitoring reports obtained during the monitoring process to network devices. Monitoring and inference reports can be reported separately or jointly. The inference report includes CSI prediction data. The monitoring report includes at least one of the following types of content: monitoring results, actual CSI measurement data, and performance indicators.

[0054] For all types of performance monitoring, network devices need to instruct terminal devices on decisions regarding monitoring actions. Depending on the type of performance monitoring, terminal devices need to report the corresponding monitoring results, actual CSI measurement data, or performance indicators to the network.

[0055] For Type 1, when the monitoring report includes monitoring results, since the terminal device only reports the monitoring results (e.g., whether the model's performance is good or bad), the network device only needs to receive the high-order information processed by the terminal device, without needing the original data. The network device can implement event-triggered monitoring reports by configuring appropriate thresholds, thus saving air interface resources. This is particularly suitable for dynamic channel environments (e.g., high-speed mobile scenarios), where only key indicators can be fed back. However, since the network device cannot obtain actual CSI measurement data, it cannot verify the model's reliability. If the terminal device's model has biases, the network device cannot directly correct the model.

[0056] For type 2, when the monitoring report includes actual CSI measurement data, the true channel state can be obtained by detecting the reference signal. Network devices can compare the CSI prediction data with the actual CSI measurement data to directly evaluate the model performance and provide data support for the model training of network devices. This type of monitoring report needs to frequently report high-dimensional CSI measurement data, which consumes a lot of uplink resources and has a large delay. In rapidly changing channels, the actual CSI measurement data in the monitoring report may have become outdated.

[0057] For Type 3, when the monitoring report includes performance metrics, this report includes performance metrics corresponding to the terminal device's computation and inference results. These metrics directly reflect the model's accuracy, facilitating network decisions such as model configuration or switching. Using unified performance metrics promotes compatibility between devices from different vendors. Compared to Type 2, Type 3 terminal devices require additional statistics on performance-related data, which may increase power consumption and makes it impossible to reconstruct specific error patterns, such as local channel mutations.

[0058] The monitoring report described in this application embodiment can use any of the above types. For example, based on type 1, only the results of performance monitoring are reported, or based on type 3, the corresponding performance indicators are reported.

[0059] As an example, such as Figure 3As shown, the terminal device measures reference signals (e.g., RS1, RS2, RS3) within the observation window and performs CSI prediction based on the obtained CSI measurement data, obtaining CSI prediction data within the prediction window (e.g., prediction instance 1, prediction instance 2, prediction instance 3, prediction instance 4). This allows the terminal device and network device to anticipate upcoming changes in channel state. Simultaneously, the terminal device measures reference signals (e.g., RS4, RS5, RS6) within the prediction window and determines the performance indicators of the prediction instances (e.g., SGCS#1, SGCS#2, SGCS#3) based on the obtained CSI measurement data. SGCS#1 corresponds to prediction instance 1, SGCS#2 to prediction instance 2, and SGCS#3 to prediction instance 3. The terminal device generates inference reports and monitoring reports, which can be submitted separately at different times.

[0060] In this embodiment, the content used to determine the performance metric can be referred to as a label or as CSI measurement data within the prediction window. Hereinafter, the CSI measurement data corresponding to the prediction instance may refer, for example, to the CSI measurement data within the prediction window used to determine the performance metric corresponding to that prediction instance. In this case, the CSI measurement data can also be considered as the label corresponding to the prediction instance.

[0061] As an example, CSI measurement data whose location is the same as the location of the predicted instance can be called a tag, while CSI measurement data whose location is different from the location of the predicted instance can be called CSI measurement data within the prediction window. In other words, CSI measurement data on the first reference signal resource whose location is the same as the predicted instance is called the tag corresponding to that predicted instance, and CSI measurement data on the first reference signal resource whose location is different from the predicted instance is called CSI measurement data within the prediction window corresponding to that predicted instance, hereinafter also referred to as CSI measurement data within the prediction window. It is important to note that CSI measurement data within the prediction window is used to determine the performance index corresponding to the predicted instance, while CSI prediction data within the observation window is used to predict the CSI corresponding to that predicted instance.

[0062] For example, CSI measurement data whose corresponding first reference signal resource has the same location as the predicted instance, and CSI measurement data whose corresponding first reference signal resource has a different location from the predicted instance, can both be called tags corresponding to the predicted instance.

[0063] For example, CSI measurement data whose corresponding first reference signal resource has the same location as the predicted instance, and CSI measurement data whose corresponding first reference signal resource has a different location from the predicted instance, can both be referred to as CSI measurement data corresponding to the predicted instance.

[0064] In some implementations, the performance metrics in the monitoring report include a first performance metric and / or a second performance metric. The first performance metric is determined based on the raw values ​​of CSI measurement data within the label and / or prediction window, while the second performance metric is determined based on an estimated value obtained after processing the raw values ​​of the CSI measurement data within the label and / or prediction window (e.g., through interpolation or other algorithms). As an example, such as... Figure 4 The time relationship between the prediction instance and the reference signal is shown. Prediction instance 1 corresponds to RS4, prediction instance 3 corresponds to RS6, and prediction instance 2 has no corresponding reference signal. Therefore, interpolation can be performed based on SGCS1 and SGCS2 corresponding to prediction instance 1 and prediction instance 3. Taking linear interpolation as an example, the SGCS corresponding to prediction instance 2 is equal to (SGCS1+SGCS2) / 2.

[0065] The location of the first reference signal resource within the prediction window may be the same as or different from the location of the prediction instance. Alternatively, when calculating performance metrics based on an estimate obtained after processing (e.g., by interpolation or other algorithms) the raw values ​​of the tags and / or CSI measurement data within the prediction window, the estimated time location corresponds to the same time location as the prediction instance. This estimate is the CSI measurement data at that time location calculated (e.g., based on interpolation or other algorithms).

[0066] In some implementations, when the location of the first reference signal resource is the same as the location of the prediction instance, the monitoring report includes one of a first performance indicator and a second performance indicator. For example, it may include the first performance indicator, or it may include the second performance indicator. For instance, if the first and second performance indicators are the same, it's possible that the location of the tag corresponding to the prediction instance and the location of the CSI measurement data within the prediction window corresponding to that prediction instance are the same. In this case, the values ​​of the tag and the CSI measurement data are also the same, so only one of the first and second performance indicators needs to be reported.

[0067] In some implementations, when the monitoring report includes a first performance indicator and a second performance indicator, and the first and second performance indicators are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource. The network device can determine that the location of the predicted instance is the same as the location of the first reference signal resource based on this monitoring report. That is, if the monitoring report includes two performance indicators with the same value, it is possible that the location of the tag corresponding to the predicted instance and the location of the CSI measurement data within the prediction window corresponding to the predicted instance are the same. In this case, the values ​​of the tag and the CSI measurement data are also the same, therefore the first and second performance indicators obtained are also the same.

[0068] Optionally, the monitoring report reported by the terminal device may include one or more of the following: tags, estimated values ​​based on the tags, and performance metrics (e.g., a first performance metric and / or a second performance metric).

[0069] In some implementations, the period of the first reference signal resource and the period of the second reference signal resource can be the same or different. The measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance. For example, using... Figure 3 For example, the second reference signal resource includes resources corresponding to RS1, RS2 and RS3. The CSI measurement data obtained by measuring the reference signal on the second reference signal resource is used to predict the CSI corresponding to prediction instance 1, prediction instance 2 and prediction instance 3 at future times. The first reference signal resource includes resources corresponding to RS4, RS5 and RS6. The CSI measurement data obtained by measuring the reference signal on the first reference signal resource is used as the tags corresponding to prediction instance 1, prediction instance 2 and prediction instance 3, and / or as the CSI measurement data corresponding to the prediction instance within the prediction window, thereby obtaining the performance indicators SGCS#1, SGCS#2 and SGCS#3 of prediction instance 1, prediction instance 2 and prediction instance 3 respectively. Figure 3Taking the example that the period of the first reference signal resource is the same as the period of the second reference signal resource, that is, the interval between adjacent RSs in RS1, RS2, and RS3 is equal to the interval between adjacent RSs in RS4, RS5, and RS6. In practical applications, network devices can also configure different periods (or densities) of reference signal resources for model inference and model monitoring. For example, prediction instances are estimates of data for future times, requiring more accurate estimates within the prediction window, necessitating a higher density of reference signal resources within the observation window; similarly, in slowly changing and / or low-speed moving environments, intensive model performance monitoring is not required, allowing for sparser reference signal resources during model monitoring; furthermore, when CSI prediction data is based on historical data to capture medium- to long-term channel change trends, highly accurate prediction results are not needed, only the trend of signal state changes, thus allowing for a lower density of reference signal resources; and finally, if model monitoring supports real-time or near-real-time fine-grained monitoring, high-density reference signal resources are required to ensure the network can quickly respond to channel fluctuations and optimize link adaptation.

[0070] Because the processing and reporting cycles, reference signal resource density, and reporting times during CSI prediction and model monitoring may differ, separate monitoring and inference reports need to be configured. There may not be a one-to-one correspondence between CSI prediction instances and reference signals in the prediction window. In this case, the correspondence between reference signals and prediction instances needs to be clearly defined. When calculating performance metrics (e.g., SGCS), optionally, CSI prediction data can be matched with the measurement data of the closest reference signal to obtain the corresponding performance metric (e.g., SGCS). For example, as... Figure 4 The time-domain relationship between the prediction instance and the reference signal is shown. Prediction instance 1 corresponds to RS4, prediction instance 3 corresponds to RS6, and prediction instance 2 has no corresponding reference signal. Therefore, the SGCS corresponding to prediction instance 1 and prediction instance 3 can be reported, but the SGCS corresponding to prediction instance 2 does not need to be reported.

[0071] The following describes the relationship between the prediction instance corresponding to the prediction instance and the first reference signal resource corresponding to the label and / or CSI measurement data within the prediction window.

[0072] In some implementations, the first reference signal resource is the reference signal resource closest to the predicted instance. For example, the first reference signal resource is the reference signal resource closest to the predicted instance and located before the predicted instance; or, for another example, the first reference signal resource is the reference signal resource closest to the predicted instance and located after the predicted instance; or, for yet another example, the predicted instance is one of the first set of reference signal resources, the second reference signal is one of the second set of reference signal resources, and the second set of reference signal resources is the set of reference signal resources closest to the first set of reference signal resources.

[0073] As an example, such as Figure 5 As shown, when the predicted instance and the first reference signal resource are not time-aligned, for example, the predicted instance is located between two adjacent reference signals, the terminal device or network device can match the CSI predicted data with the measurement data of the nearest reference signal after the predicted instance when calculating performance metrics (e.g., SGCS) to obtain the performance metrics. Figure 5 As shown; alternatively, the CSI prediction data can be matched with the measurement data of the nearest reference signal prior to the prediction instance to obtain performance metrics, such as... Figure 6 As shown; alternatively, the measurement results of the set of reference signals closest to a set of prediction instances can be matched with that set of prediction instances, for example, Figure 3 The RS4, RS5, and RS6 reference signals in the middle are compared to Figure 4 The reference signals RS4, RS5, and RS6 are closer in time to the prediction instances 1, 2, and 3, therefore they can be used. Figure 3 The correspondence shown.

[0074] In some implementations, the performance metric used to evaluate the monitoring results (e.g., the second performance metric mentioned above) can be a better performance metric among the third and fourth performance metric; or, the performance metric includes both the third and fourth performance metric. The third and fourth performance metric are determined based on two labels corresponding to the two first reference signal resources closest to the predicted instance, or based on two CSI prediction data corresponding to the two first reference signal resources closest to the predicted instance. That is, the CSI prediction data can be matched with the measurement data of the two closest reference signals to obtain two corresponding performance metrics, namely the third and fourth performance metrics, and the better performance metric can be selected for reporting, or both performance metrics can be reported simultaneously and the network device can decide which performance metric to use to evaluate the model performance. Similarly, optionally, a set of CSI prediction data can be matched with the measurement data of the two closest sets of reference signals to obtain two corresponding sets of performance metrics, and the better set of performance metrics can be selected for reporting, or both sets of performance metrics can be reported simultaneously and the network device can decide which set of performance metrics to use to evaluate the model performance.

[0075] For example, the predicted instance and the first reference signal resource are reference signal resources within the same time window. That is, a time window is set, and predicted instances within the same time window are associated with the first reference signal resource; specifically, the tag and / or CSI measurement data within the prediction window corresponding to each predicted instance are located within that time window. The time window is described in detail below. Unless otherwise specified, the term "position" of the resource or window should be interpreted broadly; for example, the position of a predicted instance refers to its start, end, or intermediate position.

[0076] In some implementations, the length of the time window is equal to the length of the period of the prediction instance, or in other words, the length of the time window is equal to the interval between two adjacent prediction instances. The period of the prediction instance or the interval between two adjacent prediction instances is also called the prediction time unit.

[0077] For example, such as Figure 7 As shown, this time window can be located between the starting positions of two adjacent prediction instances, where d is the length of the time window; for example, as... Figure 8 As shown, this time window can be located between the end positions of two adjacent prediction instances. That is, this time window is the same as the prediction time unit.

[0078] Optionally, the starting position of the time window is located before the prediction instance, and the time interval between the time window and the position of the prediction instance (e.g., the starting position of the prediction instance) is equal to half the period of the prediction instance. Alternatively, the starting position of the time window is the position obtained by subtracting half the period of the prediction instance from the starting position of the prediction instance. For example, the starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance (e.g., the starting position of the prediction instance) is equal to the number of symbols corresponding to half the period; or the starting time slot of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance (e.g., the starting position of the prediction instance) is equal to the number of time slots corresponding to half the period; or the starting time of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance (e.g., the starting position of the prediction instance) is equal to the time length of half the period, where the unit of time length is, for example, milliseconds.

[0079] Alternatively, the end position of the time window may be located after the prediction instance, and the time interval between the end position of the time window and the position of the prediction instance (e.g., the end position of the prediction instance) is equal to half the period of the prediction instance. In other words, the end position of the time window is the position obtained by adding half the period of the prediction instance to the end position of the prediction instance. For example, the end symbol of the time window may be located after the prediction instance, and the number of time-domain symbols between the end symbol of the time window and the position of the prediction instance (e.g., the end position of the prediction instance) is equal to the number of symbols corresponding to half the period. Another example is that the end slot of the time window may be located after the prediction instance, and the number of time-domain symbols between the end slot of the time window and the position of the prediction instance (e.g., the end position of the prediction instance) is equal to the number of time-domain slots corresponding to half the period. Yet another example is that the end time of the time window may be located after the prediction instance, and the number of time-domain symbols between the end time of the time window and the position of the prediction instance (e.g., the end position of the prediction instance) is equal to the duration of half the period, where the duration is in milliseconds, for example.

[0080] Optionally, the predicted instance is located in the middle of the time window. For example, if the starting position of the time window is the position obtained by subtracting half the period of the predicted instance from the starting position of the predicted instance, or the ending position of the time window is the position obtained by adding half the period of the predicted instance to the ending position of the predicted instance, then if the length of the time window is equal to the length of the period of the predicted instance, the predicted instance can be located in the middle of the time window. Another example is that the length of the time window is equal to the length of the period of the predicted instance, and the middle position of the time window is the same as the middle position of the predicted instance. When the predicted instance is located in the middle of the time window, the reference signal used for model monitoring is closer to the predicted instance, and the distance between the reference signal used for model monitoring and the predicted instance is less than half the length of the time window, which helps improve the accuracy of the monitoring results.

[0081] In some implementations, the starting position of this time window is the same as the starting position of the predicted instance. For example, as... Figure 7 As shown, defining the starting position of the time window using the starting position of the prediction instance allows the time window to be aligned with the prediction time unit. That is, the time window is the time period between the starting positions of two adjacent prediction instances.

[0082] In some implementations, when the time window includes a prediction instance and multiple first reference signal resources, the first reference signal resource used to determine the performance index is one of the multiple first reference signal resources. That is, when the time window includes multiple first reference signal resources, the terminal device can select the measurement data of the reference signal on one of the first reference signal resources as the tag corresponding to the prediction instance within the time window and / or the CSI measurement data within the prediction window.

[0083] Alternatively, if the time window includes a prediction instance and multiple first reference signal resources, the monitoring report may include the measurement times (e.g., timestamps) of the reference signals on the multiple first reference signal resources within the time window. In this case, the network device can decide which first reference signal resource's measurement data to use to determine the performance metrics corresponding to the prediction instance within the time window.

[0084] In some implementations, the monitoring report also includes time indication information for the first reference signal resource. For example, this time indication information may indicate that the first reference signal resource matching the prediction instance is earlier than the prediction instance; or, for example, it may indicate that the first reference signal resource matching the prediction instance is later than the prediction instance; or, for example, it may indicate whether the time of the first reference signal resource matching the prediction instance is the same as the time of the prediction instance. Optionally, if the monitoring report does not carry time indication information for the first reference signal resource matching the prediction instance, it indicates that the time of the first reference signal resource is the same as the time of the prediction instance.

[0085] In some implementations, the monitoring report also includes measurement time information (e.g., timestamps) of the reference signal on the first reference signal resource. By reporting the measurement time of the reference signal used to determine performance metrics, network devices can understand the changing trends of the channel state, thereby enabling more accurate scheduling.

[0086] In the absence of a clearly defined alignment between the first reference signal resource and the predicted instance, if the reference signal resource in the prediction window is variable, such as a non-periodic reference signal resource, the terminal device can determine, based on the above method, which reference signal resource's CSI measurement data should be used to determine the performance index corresponding to the current sample. For example, the terminal device can determine the matching predicted instance and the first reference signal resource based on its implementation, such as matching the temporally closest predicted instance with the first reference signal resource. However, considering that implementations from different vendors may be inconsistent, leading to uncontrollable monitoring of model performance and the inability of network devices to standardize CSI prediction data, this may affect the model's generalization ability. Therefore, in this embodiment, the terminal device and the network device should have a consistent understanding of the alignment between the first reference signal resource and the predicted instance.

[0087] During CSI prediction, the monitoring report may include performance metrics (e.g., SGCS) or monitoring results corresponding to each predicted instance in the inference report. Therefore, the terminal device does not need to report a monitoring report separately for each predicted instance, but instead reports the performance metrics (e.g., SGCS) or monitoring results of multiple instances simultaneously, thereby reducing signaling overhead. For example, the monitoring report includes CSI prediction data obtained from M inference processes, where the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers. The terminal device can report the performance metrics or monitoring results of N4 instances simultaneously each time, or report the performance metrics or monitoring results of M*N4 instances simultaneously each time.

[0088] In some implementations, the terminal device can receive a CSI resource configuration (CSI-ResourceConfig) sent by the network device. The CSI resource configuration includes parameters of a second reference signal resource, and measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance. The parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

[0089] For example, the model's monitoring process can be reused as the CSI-ResourceConfig configured for the inference process; or, for another example, the model's monitoring process can be reused as the CSI-ResourceConfig configured for the inference process, but the values ​​of certain parameters used in the monitoring process need to be adjusted based on certain parameters provided in the CSI-ResourceConfig. For example, if the period of the first reference signal resource is different from the period of the second reference signal resource, the period length can be adjusted based on the period of the second reference signal resource (e.g., increasing or decreasing by a predetermined multiple) to obtain the period of the first reference signal resource.

[0090] The reporting of monitoring and inference reports can be based on different CSI reporting configurations (CSI-ReportConfig) sent by network devices. For example, a first CSI reporting configuration is used by the terminal device to send monitoring reports, and a second CSI reporting configuration is used by the terminal device to send inference reports. In some implementations, when the terminal device sends a monitoring report to the network device based on the first CSI reporting configuration, the monitoring report also includes an identifier of the second CSI reporting configuration. That is, by carrying the identifier of the reporting configuration corresponding to the inference report in the monitoring report, the monitoring report and the inference report are associated. Alternatively, the identifier of the reporting configuration corresponding to the monitoring report can also be carried in the inference report.

[0091] The above text combined Figures 1 to 8 The method embodiments of this application are described in detail below, in conjunction with... Figures 9 to 11 The present application provides a detailed description of the apparatus embodiments. It should be understood that the descriptions of the method embodiments correspond to the descriptions of the apparatus embodiments; therefore, any parts not described in detail can be found in the foregoing method embodiments.

[0092] Figure 9 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. Figure 9The terminal device 900 shown may include a transceiver unit 910. The transceiver unit 910 is used to send a monitoring report of a model to a network device, wherein the model is used to predict Channel State Information (CSI), and the monitoring report includes performance metrics determined based on CSI prediction data and tags, and / or performance metrics determined based on CSI prediction data and CSI measurement data within a prediction window. The location of the predicted instance corresponding to the CSI prediction data has a predetermined temporal relationship with the location of the first reference signal resource corresponding to the tag and / or the CSI measurement data within the prediction window.

[0093] In some implementations, the performance metrics in the monitoring report include a first performance metric and / or a second performance metric. The first performance metric is determined based on the raw values ​​of the CSI measurement data within the label and / or the prediction window, and the second performance metric is determined based on an estimated value obtained by processing the raw values ​​of the CSI measurement data within the label and / or the prediction window.

[0094] In some implementations, when the location of the first reference signal resource is the same as the location of the prediction instance, the monitoring report includes one of the first performance metric and the second performance metric.

[0095] In some implementations, when the monitoring report includes the first performance metric and the second performance metric, and the first performance metric and the second performance metric are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource.

[0096] In some implementations, the period of the first reference signal resource is the same as or different from the period of the second reference signal resource, and the measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance.

[0097] In some implementations, the first reference signal resource is the reference signal resource closest to the prediction instance; or, the first reference signal resource is the reference signal resource closest to the prediction instance and located before the prediction instance; or, the first reference signal resource is the reference signal resource closest to the prediction instance and located after the prediction instance; or, the prediction instance is one of a first group of reference signal resources, the second reference signal is one of a second group of reference signal resources, and the second group of reference signal resources is the group of reference signal resources closest to the first group of reference signal resources.

[0098] In some implementations, the performance metric is a better performance metric among the third and fourth performance metrics; or, the performance metric includes the third and fourth performance metrics; wherein the third and fourth performance metrics are determined based on two tags and / or two CSI measurement data corresponding to the two first reference signal resources closest to the prediction instance.

[0099] In some implementations, the prediction instance and the first reference signal resource are reference signal resources within the same time window.

[0100] In some implementations, the time window is located between the start or end positions of two adjacent prediction instances; or, the length of the time window is equal to the length of the period of the prediction instance.

[0101] In some implementations, the starting position of the time window is located before the predicted instance, and the time interval between the starting position and the position of the predicted instance is equal to half the period of the predicted instance; or, the ending position of the time window is located after the predicted instance, and the time interval between the ending position and the position of the predicted instance is equal to half the period of the predicted instance.

[0102] In some implementations, the starting position of the time window is located before the prediction instance, and the time interval between the time window and the position of the prediction instance is equal to half the period of the prediction instance. This includes: the starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the number of symbols corresponding to half the period; or, the starting time slot of the time window is located before the prediction instance, and the number of time slots between the time window and the position of the prediction instance is equal to the number of time slots corresponding to half the period; or, the starting time of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the time length of half the period.

[0103] In some implementations, the end position of the time window is located after the prediction instance, and the time interval between the end position of the time window and the position of the prediction instance is equal to half the period of the prediction instance. This includes: the end symbol of the time window is located after the prediction instance, and the number of time-domain symbols between the end symbol of the time window and the position of the prediction instance is equal to the number of symbols corresponding to half the period; or, the end time slot of the time window is located after the prediction instance, and the number of time-domain symbols between the end time slot and the position of the prediction instance is equal to the number of time-domain symbols corresponding to half the period; or, the end time of the time window is located after the prediction instance, and the number of time-domain symbols between the end time slot and the position of the prediction instance is equal to the time length of half the period.

[0104] In some implementations, the position of the prediction instance includes the start position, end position, or intermediate position of the prediction instance.

[0105] In some implementations, the predicted instance is located in the middle of the time window.

[0106] In some implementations, the starting position of the time window is the same as the starting position of the prediction instance.

[0107] In some implementations, when the time window includes the prediction instance and a plurality of first reference signal resources, the first reference signal resource used to determine the performance metric is one of the plurality of first reference signal resources.

[0108] In some implementations, the monitoring report further includes time indication information of the first reference signal resource, which indicates whether the first reference signal resource is earlier than the prediction instance, later than the prediction instance, or has the same time as the prediction instance.

[0109] In some implementations, the monitoring report may also include measurement time information of the reference signal on the first reference signal resource.

[0110] In some implementations, the monitoring report includes the CSI prediction data obtained in M ​​inference processes, wherein the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers.

[0111] In some implementations, the transceiver unit 910 is further configured to: receive CSI resource configuration sent by the network device, the CSI resource configuration including parameters of a second reference signal resource, and measurement data of a reference signal on the second reference signal resource being used to predict the CSI corresponding to the prediction instance; wherein the parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

[0112] In some implementations, the parameters to be adjusted include the period.

[0113] In some implementations, the transceiver unit 910 is specifically used to: send the monitoring report to the network device based on the first CSI reporting configuration, wherein the monitoring report also includes an identifier of the second CSI reporting configuration, the second CSI reporting configuration being used by the terminal device to send an inference report, the inference report including the CSI prediction data.

[0114] In some implementations, the performance metric may be of the type SGCS or NMSE.

[0115] It is understood that the transceiver unit 910 may be, for example, a transceiver 1130. Additionally, optionally, the terminal device 900 may also include a processor 1110 and a memory 1120, see details below. Figure 11 .

[0116] Figure 10 This is a schematic diagram of the network device provided in an embodiment of this application. Figure 10 The network device 1000 shown may include a transceiver unit 1010. The transceiver unit 1010 is used to receive a monitoring report of a model sent by a terminal device, wherein the model is used to predict Channel State Information (CSI), and the monitoring report includes performance metrics determined based on CSI prediction data and tags, and / or performance metrics determined based on CSI prediction data and CSI measurement data within a prediction window. The location of the prediction instance corresponding to the CSI prediction data has a predetermined temporal relationship with the location of the first reference signal resource corresponding to the tag and / or the CSI measurement data within the prediction window.

[0117] In some implementations, the performance metrics in the monitoring report include a first performance metric and / or a second performance metric. The first performance metric is determined based on the raw values ​​of the CSI measurement data within the label and / or the prediction window, and the second performance metric is determined based on an estimated value obtained by processing the raw values ​​of the CSI measurement data within the label and / or the prediction window.

[0118] In some implementations, when the location of the first reference signal resource is the same as the location of the prediction instance, the monitoring report includes one of the first performance metric and the second performance metric.

[0119] In some implementations, when the monitoring report includes the first performance metric and the second performance metric, and the first performance metric and the second performance metric are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource.

[0120] In some implementations, the period of the first reference signal resource is the same as or different from the period of the second reference signal resource, and the measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance.

[0121] In some implementations, the first reference signal resource is the reference signal resource closest to the prediction instance; or, the first reference signal resource is the reference signal resource closest to the prediction instance and located before the prediction instance; or, the first reference signal resource is the reference signal resource closest to the prediction instance and located after the prediction instance; or, the prediction instance is one of a first group of reference signal resources, the second reference signal is one of a second group of reference signal resources, and the second group of reference signal resources is the group of reference signal resources closest to the first group of reference signal resources.

[0122] In some implementations, the performance metric is a better performance metric among the third and fourth performance metrics; or, the performance metric includes the third and fourth performance metrics; wherein the third and fourth performance metrics are determined based on two tags and / or two CSI measurement data corresponding to the two first reference signal resources closest to the prediction instance.

[0123] In some implementations, the prediction instance and the first reference signal resource are reference signal resources within the same time window.

[0124] In some implementations, the time window is located between the start or end positions of two adjacent prediction instances; or, the length of the time window is equal to the length of the period of the prediction instance.

[0125] In some implementations, the starting position of the time window is located before the predicted instance, and the time interval between the starting position and the position of the predicted instance is equal to half the period of the predicted instance; or, the ending position of the time window is located after the predicted instance, and the time interval between the ending position and the position of the predicted instance is equal to half the period of the predicted instance.

[0126] In some implementations, the starting position of the time window is located before the prediction instance, and the time interval between the time window and the position of the prediction instance is equal to half the period of the prediction instance. This includes: the starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the number of symbols corresponding to half the period; or, the starting time slot of the time window is located before the prediction instance, and the number of time slots between the time window and the position of the prediction instance is equal to the number of time slots corresponding to half the period; or, the starting time of the time window is located before the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the time length of half the period.

[0127] In some implementations, the end position of the time window is located after the prediction instance, and the time interval between the end position of the time window and the position of the prediction instance is equal to half the period of the prediction instance. This includes: the end symbol of the time window is located after the prediction instance, and the number of time-domain symbols between the end symbol of the time window and the position of the prediction instance is equal to the number of symbols corresponding to half the period; or, the end time slot of the time window is located after the prediction instance, and the number of time-domain symbols between the end time slot and the position of the prediction instance is equal to the number of time-domain symbols corresponding to half the period; or, the end time of the time window is located after the prediction instance, and the number of time-domain symbols between the end time slot and the position of the prediction instance is equal to the time length of half the period.

[0128] In some implementations, the position of the prediction instance includes the start position, end position, or intermediate position of the prediction instance.

[0129] In some implementations, the predicted instance is located in the middle of the time window.

[0130] In some implementations, the starting position of the time window is the same as the starting position of the prediction instance.

[0131] In some implementations, when the time window includes the prediction instance and a plurality of first reference signal resources, the first reference signal resource used to determine the performance metric is one of the plurality of first reference signal resources.

[0132] In some implementations, the monitoring report also includes time information of the first reference signal resource, which indicates whether the first reference signal resource is earlier than the prediction instance, later than the prediction instance, or has the same time as the prediction instance.

[0133] In some implementations, the monitoring report may also include measurement time information of the reference signal on the first reference signal resource.

[0134] In some implementations, the monitoring report includes the CSI prediction data obtained in M ​​inference processes, wherein the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers.

[0135] In some implementations, the transceiver unit 1010 further includes: sending CSI resource configuration to the terminal device, wherein the CSI resource configuration includes parameters of a second reference signal resource, and measurement data of a reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance; wherein the parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

[0136] In some implementations, the parameters to be adjusted include the period.

[0137] In some implementations, the transceiver unit 101 is further configured to: send a first CSI reporting configuration to the terminal device, the first CSI reporting configuration being used by the terminal device to send the monitoring report, wherein the monitoring report further includes an identifier of a second CSI reporting configuration, the second CSI reporting configuration being used by the terminal device to send an inference report, the inference report including the CSI prediction data.

[0138] In some implementations, the performance metric may be of the type SGCS or NMSE.

[0139] It is understood that the transceiver unit 1010 may be, for example, a transceiver 1130. Additionally, optionally, the network device 1000 may also include a processor 1110 and a memory 1120, see details below. Figure 11 .

[0140] Figure 11 This is a schematic structural diagram of a communication apparatus according to an embodiment of this application. Figure 11 The dashed lines shown indicate that the unit or module is optional. Device 1100 can be used to implement the methods described in the above method embodiments. Device 1100 may be, for example, a chip, a terminal device, or a network device.

[0141] Apparatus 1100 may include one or more processors 1110. Processor 1110 may support apparatus 1100 in implementing the methods described in the foregoing method embodiments. Processor 1110 may be a general-purpose processor or a special-purpose processor. For example, processor 1110 may be a central processing unit (CPU). Alternatively, processor 1110 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors may be microprocessors or any conventional processor, etc.

[0142] The apparatus 1100 may further include one or more memories 1120. The memories 1120 store programs that can be executed by the processor 1110, causing the processor 1110 to perform the methods described in the above method embodiments. The memories 1120 may be independent of the processor 1110, or they may be integrated into the processor 1110.

[0143] The device 1100 may also include a transceiver 1130. The processor 1110 can communicate with other devices or chips via the transceiver 1130. For example, the processor 1110 can send and receive data with other devices or chips via the transceiver 1130.

[0144] This application also provides a communication system. The communication system includes the terminal device and network device described above. In some implementations, the system further includes other devices that interact with the terminal device and network device.

[0145] This application also provides a computer-readable storage medium for storing a program. This computer-readable storage medium can be applied to a terminal device or network device provided in this application, and the program causes a computer to execute the methods performed by the terminal device or network device in various embodiments of this application.

[0146] This application also provides a computer program product. The computer program product includes a program. This computer program product can be applied to a terminal device or network device provided in this application embodiment, and the program causes a computer to execute the methods performed by the terminal device or network device in the various embodiments of this application.

[0147] This application also provides a computer program. This computer program can be applied to the terminal device or network device provided in this application, and the computer program causes the computer to execute the methods performed by the terminal device or network device in the various embodiments of this application.

[0148] It should be understood that the terms "system" and "network" in the embodiments of this application can be used interchangeably. Furthermore, the terminology used in this application is only for explaining specific embodiments of this application and is not intended to limit this application. The terms "first," "second," "third," and "fourth," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. In addition, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0149] In the embodiments of this application, the term "instruction" can be a direct instruction, an indirect instruction, or an indication of a relationship. For example, A instructing B can mean that A directly instructs B, such as B being able to obtain information through A; it can also mean that A indirectly instructs B, such as A instructing C, so B can obtain information through C; or it can mean that there is a relationship between A and B.

[0150] In the embodiments of this application, "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean that B is determined solely based on A; B can also be determined based on A and / or other information.

[0151] In the embodiments of this application, the term "correspondence" can indicate a direct or indirect correspondence between two things, or an association between two things, or a relationship such as instruction and being instructed, configuration and being configured.

[0152] In this application embodiment, "predefined" or "preconfigured" can be implemented by pre-storing corresponding codes, tables, or other means that can be used to indicate relevant information in the device (e.g., including terminal devices and network devices). This application does not limit the specific implementation method. For example, predefined can refer to what is defined in the protocol.

[0153] In this application embodiment, the "protocol" may refer to a standard protocol in the field of communication, such as the LTE protocol, the NR protocol, and related protocols applied to future communication systems. This application does not limit this.

[0154] In the embodiments of this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0155] In the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0156] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0157] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0158] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0159] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can read or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital video discs (DVDs)), or semiconductor media (e.g., solid-state disks (SSDs)).

[0160] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A communication method, characterized in that, include: The terminal device sends a monitoring report of the model to the network device, wherein the model is used to predict Channel State Information (CSI), and the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within the prediction window. The location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of the first reference signal resource corresponding to the tag and / or the CSI measurement data within the prediction window.

2. The method according to claim 1, characterized in that, The performance indicators in the monitoring report include a first performance indicator and / or a second performance indicator. The first performance indicator is determined based on the raw values ​​of the CSI measurement data within the label and / or the prediction window, and the second performance indicator is determined based on the estimated values ​​obtained after processing the raw values ​​of the CSI measurement data within the label and / or the prediction window.

3. The method according to claim 2, characterized in that, When the location of the first reference signal resource is the same as the location of the predicted instance, the monitoring report includes one of the first performance metric and the second performance metric.

4. The method according to claim 2 or 3, characterized in that, When the monitoring report includes the first performance indicator and the second performance indicator, and the first performance indicator and the second performance indicator are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource.

5. The method according to any one of claims 1 to 4, characterized in that, The period of the first reference signal resource may be the same as or different from the period of the second reference signal resource, and the measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance.

6. The method according to any one of claims 1 to 5, characterized in that, The first reference signal resource is the reference signal resource closest to the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located before the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located after the predicted instance; or, The prediction instance is one of the first set of reference signal resources, the second reference signal is one of the second set of reference signal resources, and the second set of reference signal resources is the set of reference signal resources that is closest to the first set of reference signal resources.

7. The method according to any one of claims 1 to 6, characterized in that, The performance metric is the better one of the third and fourth performance metric; or... The performance indicators include the third performance indicator and the fourth performance indicator; The third performance index and the fourth performance index are determined based on two tags and / or two CSI measurement data corresponding to the two first reference signal resources that are closest to the prediction instance.

8. The method according to any one of claims 1 to 7, characterized in that, The prediction instance and the first reference signal resource are reference signal resources within the same time window.

9. The method according to claim 8, characterized in that, The time window is located between the start or end positions of two adjacent prediction instances; or, The length of the time window is equal to the length of the period of the prediction instance.

10. The method according to claim 8 or 9, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the time window and the position of the predicted instance is equal to half the period of the predicted instance; or, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance.

11. The method according to claim 10, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the starting position and the position of the predicted instance is equal to half the period of the predicted instance, including: The starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the starting symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The starting time slot of the time window is located before the predicted instance, and the number of time slots between the starting time slot and the location of the predicted instance is equal to the number of time slots corresponding to half a period; or, The start time of the time window is located before the predicted instance, and the number of time-domain symbols between the time window and the position of the predicted instance is equal to the time length of half a cycle.

12. The method according to claim 10 or 11, characterized in that, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance, including: The end symbol of the time window is located after the prediction instance, and the number of time-domain symbols between the end symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The end slot of the time window is located after the predicted instance, and the number of time slots between the end slot and the position of the predicted instance is equal to the number of time slots corresponding to half a period; or, The end time of the time window is located after the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the time length of half a cycle.

13. The method according to claim 11 or 12, characterized in that, The position of the predicted instance includes the starting position, ending position, or intermediate position of the predicted instance.

14. The method according to any one of claims 8 to 13, characterized in that, The predicted instance is located in the middle of the time window.

15. The method according to any one of claims 8 to 14, characterized in that, The starting position of the time window is the same as the starting position of the prediction instance.

16. The method according to any one of claims 8 to 15, characterized in that, When the time window includes the prediction instance and a plurality of first reference signal resources, the first reference signal resource used to determine the performance metric is one of the plurality of first reference signal resources.

17. The method according to any one of claims 1 to 16, characterized in that, The monitoring report also includes time indication information of the first reference signal resource, which indicates whether the first reference signal resource is earlier than the prediction instance, later than the prediction instance, or has the same time as the prediction instance.

18. The method according to any one of claims 1 to 17, characterized in that, The monitoring report also includes measurement time information of the reference signal on the first reference signal resource.

19. The method according to any one of claims 1 to 18, characterized in that, The monitoring report includes the CSI prediction data obtained in M ​​inference processes, wherein the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers.

20. The method according to any one of claims 1 to 19, characterized in that, The method further includes: The terminal device receives the CSI resource configuration sent by the network device. The CSI resource configuration includes parameters of the second reference signal resource. The measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance. The parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

21. The method according to claim 20, characterized in that, The parameters to be adjusted include the cycle.

22. The method according to any one of claims 1 to 21, characterized in that, The terminal device sends a monitoring report of the model to the network device, including: The terminal device sends the monitoring report to the network device based on the first CSI reporting configuration. The monitoring report also includes an identifier of the second CSI reporting configuration, which is used by the terminal device to send an inference report. The inference report includes the CSI prediction data.

23. The method according to any one of claims 1 to 22, characterized in that, The performance metrics include squared cosine similarity (SGCS) or normalized mean square error (NMSE).

24. A communication method, characterized in that, include: The network device receives a monitoring report of a model sent by a terminal device, wherein the model is used to predict Channel State Information (CSI), and the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window. The location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of the first reference signal resource corresponding to the tag and / or the CSI measurement data within the prediction window.

25. The method according to claim 24, characterized in that, The performance indicators in the monitoring report include a first performance indicator and / or a second performance indicator. The first performance indicator is determined based on the raw values ​​of the CSI measurement data within the label and / or the prediction window, and the second performance indicator is determined based on the estimated values ​​obtained after processing the raw values ​​of the CSI measurement data within the label and / or the prediction window.

26. The method according to claim 25, characterized in that, When the location of the first reference signal resource is the same as the location of the predicted instance, the monitoring report includes one of the first performance metric and the second performance metric.

27. The method according to claim 25 or 26, characterized in that, When the monitoring report includes the first performance indicator and the second performance indicator, and the first performance indicator and the second performance indicator are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource.

28. The method according to any one of claims 24 to 27, characterized in that, The period of the first reference signal resource may be the same as or different from the period of the second reference signal resource, and the measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance.

29. The method according to any one of claims 24 to 28, characterized in that, The first reference signal resource is the reference signal resource closest to the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located before the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located after the predicted instance; or, The prediction instance is one of the first set of reference signal resources, the second reference signal is one of the second set of reference signal resources, and the second set of reference signal resources is the set of reference signal resources that is closest to the first set of reference signal resources.

30. The method according to any one of claims 24 to 29, characterized in that, The performance metric is the better one of the third and fourth performance metric; or... The performance indicators include the third performance indicator and the fourth performance indicator; The third performance index and the fourth performance index are determined based on two tags and / or two CSI measurement data corresponding to the two first reference signal resources that are closest to the prediction instance.

31. The method according to any one of claims 24 to 30, characterized in that, The prediction instance and the first reference signal resource are reference signal resources within the same time window.

32. The method according to claim 31, characterized in that, The time window is located between the start or end positions of two adjacent prediction instances; or, The length of the time window is equal to the length of the period of the prediction instance.

33. The method according to claim 31 or 32, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the time window and the position of the predicted instance is equal to half the period of the predicted instance; or, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance.

34. The method according to claim 33, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the starting position and the position of the predicted instance is equal to half the period of the predicted instance, including: The starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the starting symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The starting time slot of the time window is located before the predicted instance, and the number of time slots between the starting time slot and the location of the predicted instance is equal to the number of time slots corresponding to half a period; or, The start time of the time window is located before the predicted instance, and the number of time-domain symbols between the time window and the position of the predicted instance is equal to the time length of half a cycle.

35. The method according to claim 33 or 34, characterized in that, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance, including: The end symbol of the time window is located after the prediction instance, and the number of time-domain symbols between the end symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The end slot of the time window is located after the predicted instance, and the number of time slots between the end slot and the position of the predicted instance is equal to the number of time slots corresponding to half a period; or, The end time of the time window is located after the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the time length of half a cycle.

36. The method according to claim 34 or 35, characterized in that, The position of the predicted instance includes the starting position, ending position, or intermediate position of the predicted instance.

37. The method according to any one of claims 31 to 36, characterized in that, The predicted instance is located in the middle of the time window.

38. The method according to any one of claims 31 to 37, characterized in that, The starting position of the time window is the same as the starting position of the prediction instance.

39. The method according to any one of claims 31 to 38, characterized in that, When the time window includes the prediction instance and a plurality of first reference signal resources, the first reference signal resource used to determine the performance metric is one of the plurality of first reference signal resources.

40. The method according to any one of claims 24 to 39, characterized in that, The monitoring report also includes time information of the first reference signal resource, which indicates whether the first reference signal resource is earlier than the prediction instance, later than the prediction instance, or has the same time as the prediction instance.

41. The method according to any one of claims 24 to 40, characterized in that, The monitoring report also includes measurement time information of the reference signal on the first reference signal resource.

42. The method according to any one of claims 24 to 41, characterized in that, The monitoring report includes the CSI prediction data obtained in M ​​inference processes, wherein the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers.

43. The method according to any one of claims 24 to 42, characterized in that, The method further includes: The network device sends a CSI resource configuration to the terminal device. The CSI resource configuration includes parameters of a second reference signal resource. Measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance. The parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

44. The method according to claim 43, characterized in that, The parameters to be adjusted include the cycle.

45. The method according to any one of claims 24 to 44, characterized in that, The method further includes: The network device sends a first CSI reporting configuration to the terminal device. The first CSI reporting configuration is used by the terminal device to send the monitoring report. The monitoring report also includes an identifier of a second CSI reporting configuration. The second CSI reporting configuration is used by the terminal device to send an inference report. The inference report includes the CSI prediction data.

46. ​​The method according to any one of claims 24 to 45, characterized in that, The performance metrics include squared cosine similarity (SGCS) or normalized mean square error (NMSE).

47. A terminal device, characterized in that, include: A transceiver unit is used to send a monitoring report of a model to a network device, wherein the model is used to predict channel state information (CSI), and the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window. The location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of the first reference signal resource corresponding to the tag and / or the CSI measurement data within the prediction window.

48. The terminal device according to claim 47, characterized in that, The performance indicators in the monitoring report include a first performance indicator and / or a second performance indicator. The first performance indicator is determined based on the raw values ​​of the CSI measurement data within the label and / or the prediction window, and the second performance indicator is determined based on the estimated values ​​obtained after processing the raw values ​​of the CSI measurement data within the label and / or the prediction window.

49. The terminal device according to claim 48, characterized in that, When the location of the first reference signal resource is the same as the location of the predicted instance, the monitoring report includes one of the first performance metric and the second performance metric.

50. The terminal device according to claim 48 or 49, characterized in that, When the monitoring report includes the first performance indicator and the second performance indicator, and the first performance indicator and the second performance indicator are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource.

51. The terminal device according to any one of claims 47 to 50, characterized in that, The period of the first reference signal resource may be the same as or different from the period of the second reference signal resource, and the measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance.

52. The terminal device according to any one of claims 47 to 51, characterized in that, The first reference signal resource is the reference signal resource closest to the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located before the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located after the predicted instance; or, The prediction instance is one of the first set of reference signal resources, the second reference signal is one of the second set of reference signal resources, and the second set of reference signal resources is the set of reference signal resources that is closest to the first set of reference signal resources.

53. The terminal device according to any one of claims 47 to 52, characterized in that, The performance metric is the better one of the third and fourth performance metric; or... The performance indicators include the third performance indicator and the fourth performance indicator; The third performance index and the fourth performance index are determined based on two tags and / or two CSI measurement data corresponding to the two first reference signal resources that are closest to the prediction instance.

54. The terminal device according to any one of claims 47 to 53, characterized in that, The prediction instance and the first reference signal resource are reference signal resources within the same time window.

55. The terminal device according to claim 54, characterized in that, The time window is located between the start or end positions of two adjacent prediction instances; or, The length of the time window is equal to the length of the period of the prediction instance.

56. The terminal device according to claim 54 or 55, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the time window and the position of the predicted instance is equal to half the period of the predicted instance; or, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance.

57. The terminal device according to claim 56, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the starting position and the position of the predicted instance is equal to half the period of the predicted instance, including: The starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the starting symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The starting time slot of the time window is located before the predicted instance, and the number of time slots between the starting time slot and the location of the predicted instance is equal to the number of time slots corresponding to half a period; or, The start time of the time window is located before the predicted instance, and the number of time-domain symbols between the time window and the position of the predicted instance is equal to the time length of half a cycle.

58. The terminal device according to claim 56 or 57, characterized in that, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance, including: The end symbol of the time window is located after the prediction instance, and the number of time-domain symbols between the end symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The end slot of the time window is located after the predicted instance, and the number of time slots between the end slot and the position of the predicted instance is equal to the number of time slots corresponding to half a period; or, The end time of the time window is located after the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the time length of half a cycle.

59. The terminal device according to claim 57 or 58, characterized in that, The position of the predicted instance includes the starting position, ending position, or intermediate position of the predicted instance.

60. The terminal device according to any one of claims 54 to 59, characterized in that, The predicted instance is located in the middle of the time window.

61. The terminal device according to any one of claims 54 to 60, characterized in that, The starting position of the time window is the same as the starting position of the prediction instance.

62. The terminal device according to any one of claims 54 to 61, characterized in that, When the time window includes the prediction instance and a plurality of first reference signal resources, the first reference signal resource used to determine the performance metric is one of the plurality of first reference signal resources.

63. The terminal device according to any one of claims 47 to 62, characterized in that, The monitoring report also includes time indication information of the first reference signal resource, which indicates whether the first reference signal resource is earlier than the prediction instance, later than the prediction instance, or has the same time as the prediction instance.

64. The terminal device according to any one of claims 47 to 63, characterized in that, The monitoring report also includes measurement time information of the reference signal on the first reference signal resource.

65. The terminal device according to any one of claims 47 to 64, characterized in that, The monitoring report includes the CSI prediction data obtained in M ​​inference processes, wherein the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers.

66. The terminal device according to any one of claims 47 to 65, characterized in that, The transceiver unit is also used for: The network device receives a CSI resource configuration, which includes parameters of a second reference signal resource. Measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance. The parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

67. The terminal device according to claim 66, characterized in that, The parameters to be adjusted include the cycle.

68. The terminal device according to any one of claims 47 to 67, characterized in that, The transceiver unit is specifically used for: Based on the first CSI reporting configuration, the monitoring report is sent to the network device. The monitoring report also includes an identifier of the second CSI reporting configuration, which is used by the terminal device to send an inference report. The inference report includes the CSI prediction data.

69. The terminal device according to any one of claims 47 to 68, characterized in that, The performance metrics include squared cosine similarity (SGCS) or normalized mean square error (NMSE).

70. A network device, characterized in that, include: The transceiver unit is used to receive a monitoring report of a model sent by a terminal device, wherein the model is used to predict channel state information (CSI), and the monitoring report includes performance indicators determined based on CSI prediction data and tags, and / or performance indicators determined based on CSI prediction data and CSI measurement data within a prediction window. The location of the prediction instance corresponding to the CSI prediction data has a predetermined time relationship with the location of the first reference signal resource corresponding to the tag and / or the CSI measurement data within the prediction window.

71. The network device according to claim 70, characterized in that, The performance indicators in the monitoring report include a first performance indicator and / or a second performance indicator. The first performance indicator is determined based on the raw values ​​of the CSI measurement data within the label and / or the prediction window, and the second performance indicator is determined based on the estimated values ​​obtained after processing the raw values ​​of the CSI measurement data within the label and / or the prediction window.

72. The network device according to claim 71, characterized in that, When the location of the first reference signal resource is the same as the location of the predicted instance, the monitoring report includes one of the first performance metric and the second performance metric.

73. The network device according to claim 71 or 72, characterized in that, When the monitoring report includes the first performance indicator and the second performance indicator, and the first performance indicator and the second performance indicator are the same, the monitoring report is used to indicate that the location of the predicted instance is the same as the location of the first reference signal resource.

74. The network device according to any one of claims 70 to 73, characterized in that, The period of the first reference signal resource may be the same as or different from the period of the second reference signal resource, and the measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance.

75. The network device according to any one of claims 70 to 74, characterized in that, The first reference signal resource is the reference signal resource closest to the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located before the predicted instance; or, The first reference signal resource is the reference signal resource that is closest to the predicted instance and located after the predicted instance; or, The prediction instance is one of the first set of reference signal resources, the second reference signal is one of the second set of reference signal resources, and the second set of reference signal resources is the set of reference signal resources that is closest to the first set of reference signal resources.

76. The network device according to any one of claims 70 to 75, characterized in that, The performance metric is the better one of the third and fourth performance metric; or... The performance indicators include the third performance indicator and the fourth performance indicator; The third performance index and the fourth performance index are determined based on two tags and / or two CSI measurement data corresponding to the two first reference signal resources that are closest to the prediction instance.

77. The network device according to any one of claims 70 to 76, characterized in that, The prediction instance and the first reference signal resource are reference signal resources within the same time window.

78. The network device according to claim 77, characterized in that, The time window is located between the start or end positions of two adjacent prediction instances; or, The length of the time window is equal to the length of the period of the prediction instance.

79. The network device according to claim 77 or 78, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the time window and the position of the predicted instance is equal to half the period of the predicted instance; or, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance.

80. The network device according to claim 79, characterized in that, The starting position of the time window is located before the predicted instance, and the time interval between the starting position and the position of the predicted instance is equal to half the period of the predicted instance, including: The starting symbol of the time window is located before the prediction instance, and the number of time-domain symbols between the starting symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The starting time slot of the time window is located before the predicted instance, and the number of time slots between the starting time slot and the location of the predicted instance is equal to the number of time slots corresponding to half a period; or, The start time of the time window is located before the predicted instance, and the number of time-domain symbols between the time window and the position of the predicted instance is equal to the time length of half a cycle.

81. The network device according to claim 79 or 80, characterized in that, The end position of the time window is located after the prediction instance, and the time interval between the end position and the end position of the prediction instance is equal to half the period of the prediction instance, including: The end symbol of the time window is located after the prediction instance, and the number of time-domain symbols between the end symbol and the position of the prediction instance is equal to the number of symbols corresponding to half a period; or, The end slot of the time window is located after the predicted instance, and the number of time slots between the end slot and the position of the predicted instance is equal to the number of time slots corresponding to half a period; or, The end time of the time window is located after the prediction instance, and the number of time-domain symbols between the time window and the position of the prediction instance is equal to the time length of half a cycle.

82. The network device according to claim 80 or 81, characterized in that, The position of the predicted instance includes the starting position, ending position, or intermediate position of the predicted instance.

83. The network device according to any one of claims 77 to 82, characterized in that, The predicted instance is located in the middle of the time window.

84. The network device according to any one of claims 77 to 83, characterized in that, The starting position of the time window is the same as the starting position of the prediction instance.

85. The network device according to any one of claims 77 to 84, characterized in that, When the time window includes the prediction instance and a plurality of first reference signal resources, the first reference signal resource used to determine the performance metric is one of the plurality of first reference signal resources.

86. The network device according to any one of claims 70 to 85, characterized in that, The monitoring report also includes time information of the first reference signal resource, which indicates whether the first reference signal resource is earlier than the prediction instance, later than the prediction instance, or has the same time as the prediction instance.

87. The network device according to any one of claims 70 to 86, characterized in that, The monitoring report also includes measurement time information of the reference signal on the first reference signal resource.

88. The network device according to any one of claims 70 to 87, characterized in that, The monitoring report includes the CSI prediction data obtained in M ​​inference processes, wherein the number of CSI prediction data obtained in each inference process is N, and M and N are positive integers.

89. The network device according to any one of claims 70 to 88, characterized in that, The transceiver unit further includes: Send CSI resource configuration to the terminal device. The CSI resource configuration includes parameters of a second reference signal resource. Measurement data of the reference signal on the second reference signal resource is used to predict the CSI corresponding to the prediction instance. The parameters of the first reference signal resource are obtained by adjusting the parameters of the second reference signal resource.

90. The network device according to claim 89, characterized in that, The parameters to be adjusted include the cycle.

91. The network device according to any one of claims 70 to 90, characterized in that, The transceiver unit is also used for: A first CSI reporting configuration is sent to the terminal device. The first CSI reporting configuration is used by the terminal device to send the monitoring report. The monitoring report also includes an identifier of a second CSI reporting configuration. The second CSI reporting configuration is used by the terminal device to send an inference report. The inference report includes the CSI prediction data.

92. The network device according to any one of claims 70 to 91, characterized in that, The performance metrics include squared cosine similarity (SGCS) or normalized mean square error (NMSE).

93. A terminal device, characterized in that, The device includes a transceiver, a memory, and a processor. The memory stores a program, and the processor invokes the program in the memory and controls the transceiver to receive or send signals so that the terminal device performs the method according to any one of claims 1 to 23.

94. A network device, characterized in that, The device includes a transceiver, a memory, and a processor. The memory stores a program, and the processor invokes the program in the memory and controls the transceiver to receive or transmit signals so that the network device performs the method according to any one of claims 24 to 46.

95. An apparatus, characterized in that, Includes a processor for calling a program from memory to cause the apparatus to perform the method according to any one of claims 1 to 46.

96. A chip, characterized in that, Includes a processor for calling a program from memory, causing a device on which the chip is mounted to perform the method according to any one of claims 1 to 46.

97. A computer-readable storage medium, characterized in that, It contains a program that causes a computer to perform the method according to any one of claims 1 to 46.

98. A computer program product, characterized in that, Includes a program that causes a computer to perform the method according to any one of claims 1 to 46.

99. A computer program, characterized in that, The computer program causes the computer to perform the method according to any one of claims 1 to 46.

Images