Network node and method for determining channel state information

Abstract

The embodiments of the present disclosure provide a method performed by a network node and a network node, which involves the field of artificial intelligence. The method comprises: determining a prediction region corresponding to a user equipment (UE), according to position information of the UE; obtaining historical channel state information (CSI) from a plurality of reference position points corresponding to the prediction region; and determining CSI of the UE, using an artificial intelligence (AI) network, based on the obtained historical CSI. Alternatively, the above method executed by the network node may be executed by using artificial intelligence models.

Description

NETWORK NODE AND METHOD FOR DETERMINING CHANNEL STATE INFORMATION

[0001] The present application relates to a field of mobile communication technology, and more specifically, the present application relates to a method for determining Channel State Information (CSI) of a User Equipment (UE) according to historical CSI and a network node.

[0002] CSI is one of the most important concepts in the field of wireless communication, the CSI may reflect characteristics of wireless channel, the subjected comprehensive effects of various factors such as multipath loss, scattering, fading, and shadowing during a propagation process of a historical signal from a transmitter to a receiver. A base station may accurately determine specific link channel characteristics by collecting CSI information reported by a UE.

[0003] Accurately grasping channel condition by collecting CSI information of the UE is crucial for improving data transmission rate, but as an antenna size of Multiple-Input Multiple-Output (MIMO) antennas increases, channel overhead required for reporting the CSI information becomes very large, and how to reduce the overhead of CSI reporting is a great challenge for a 5G and a 5.5G network.

[0004] The above information is presented as background information only to assist with an understanding of the disclosure. No determination has been made, and no assertion is made, as to whether any of the above might be applicable as prior art with regard to the disclosure.

[0005] In order to at least solve the above-mentioned problems existing in the related art, the present application provides a method performed by a network node and a network node.

[0006] According to a first aspect of an embodiment of the present application, a method performed by a network node is provided, which includes: determining a prediction region corresponding to a first UE, according to position information of the first UE; predicting Channel State Information (CSI) of the first UE, according to historical CSI of a plurality of reference position points corresponding to the prediction region.

[0007] Alternatively, the determining the prediction region corresponding to the first UE, according to the position information of the first UE, includes: determining, from among at least one prediction region within a coverage region of the network node, a prediction region to which a position of the first UE belongs, according to the position information of the first UE.

[0008] Alternatively, the predicting the CSI of the first UE, according to the historical CSI of the plurality of reference position points corresponding to the prediction region, includes: selecting the plurality of reference position points from among reference position points of the prediction region, according to a CSI prediction accuracy requirement and / or processing capability of the network node.

[0009] Alternatively, the selecting the plurality of reference position points from among the reference position points of the prediction region, according to the CSI prediction accuracy requirement and / or the processing capability of the network node, includes: in response to the number of the reference position points of the prediction region being less than or equal to a first threshold value, selecting all reference position points of the prediction region; in response to the number of the reference position points of the prediction region being greater than the first threshold value, selecting the plurality of reference position points from among the reference positions of the prediction region, according to the position information of the first UE, the CSI prediction accuracy requirement, and / or the processing capacity of the network node.

[0010] Alternatively, the selecting the plurality of reference position points from among the reference positions of the prediction region, according to the position information of the first UE, the CSI prediction accuracy requirement, and / or the processing capacity of the network node, includes: in response to a first number of reference position points determined according to the CSI prediction accuracy requirement and / or the processing capacity of the network node being less than the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose positions are closest to the first UE and of which the number being the first number of reference position points; in response to the first number of first reference position point determined according to the CSI prediction accuracy requirement and / or the processing capacity of the network node being greater than or equal to the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose position are closest to the first UE and of which the number being the first threshold value.

[0011] Alternatively, there is a UE in a connected state at each of the plurality of reference position points previously or currently.

[0012] Alternatively, the number of the plurality of reference position points is greater than or equal to a second threshold value, wherein the second threshold value is determined according to a CSI prediction accuracy requirement, the first threshold value is determined according to a processing capacity of the network node, and the first threshold value is greater than the second threshold value.

[0013] Alternatively, the predicting the CSI of the first UE, according to the historical CSI of the plurality of reference position points corresponding to the prediction region, includes: predicting the CSI of the first UE through a first Artificial Intelligence (AI) network, according to the position information of the first UE, the historical CSI of the plurality of reference position points, and reporting position information of the historical CSI.

[0014] Alternatively, the predicting the CSI of the first UE through the first AI network, according to the position information of the first UE, the historical CSI of the plurality of reference position points, and the reporting position information of the historical CSI, includes: preprocessing at least one historical CSI reported within a preset time interval prior to the current time of each reference position point, reporting position information of the at least one historical CSI, the position information of the first UE, through a preprocessing layer of the first AI network, to obtain channel information and position difference information corresponding to each reference position point; extracting feature information from input information through a embedding layer of the first AI network, wherein the input information includes the channel information and the position difference information corresponding to each reference position point; predicting the CSI of the first UE according to the feature information through a recovery layer of the first AI network.

[0015] Alternatively, the preprocessing layer further preprocesses reporting time information of the at least one historical CSI to obtain time difference information corresponding to each reference position point, wherein the input information further includes the time difference information.

[0016] Alternatively, the extracting the feature information from the input information through the embedding layer of the first AI network includes: performing feature extraction on the input information according to the number of antennas, to obtain feature information for an antenna dimension.

[0017] Alternatively, the position difference information is a position difference determined according to the reporting position information of the at least one historical CSI of each reference position point and the position information of the first UE, and / or the time difference information is a time difference determined according to the reporting time information of the at least one historical CSI of each reference position point and the time information of the predicted CSI of the first UE.

[0018] Alternatively, the method further includes: configuring a UE, that is not located at a reference position point, in the prediction region within a coverage region of the network node to not report CSI information to the network node.

[0019] Alternatively, each prediction region is obtained by: determining channel spatial consistency between respective position points within a coverage region of the network node, according to historical CSI received within the coverage region; dividing the coverage region to obtain at least one prediction region according to the channel spatial consistency.

[0020] Alternatively, the determining the channel spatial consistency between the respective position points within the coverage region of the network node, according to the historical CSI received within the coverage region, includes: identifying at least one hotspot region within the coverage region for each time period, according to the historical CSI received within the coverage region; performing channel spatial consistency analysis between the respective position points in each hotspot region for each time period, according to the historical CSI received within each hotspot region for each time period.

[0021] Alternatively, the performing the channel spatial consistency analysis between the respective position points in each hotspot region for each time period includes: determining channel spatial consistency coefficients between each position point and other position points within each hotspot region for each time period, according to the historical CSI received within each hotspot region for each time period; determining a channel spatial consistency distance of the each position point in each direction, according to the channel spatial consistency coefficients between the each position point and the other position points, wherein the channel spatial consistency distance represents a maximum distance at which the channel spatial consistency coefficients is greater than a predetermined threshold value, wherein a portion of the position points in each hotspot region for each time period are selected as candidate position points for dividing of the prediction region.

[0022] Alternatively, the dividing the coverage region to obtain the at least one prediction region according to the channel spatial consistency includes: performing the following operations for each hotspot region for a current time period in the coverage region: determining a plurality of candidate reference position points according to respective candidate position points in a current hotspot region; determining a channel consistency space of each candidate reference position point, according to channel spatial consistency distances of each candidate reference position point and a third threshold value; determining a prediction region in the current hotspot region, according to the channel consistency space of each candidate reference position point.

[0023] Alternatively, the determining the prediction region in the current hotspot region, according to the channel consistency space of each candidate reference position point includes: for any n candidate reference position points among the plurality of candidate reference position points in the current hotspot region: if there is an intersection among intersection regions of respective channel consistency spaces of the any n candidate reference position points, an union region of the intersection regions is divided as one prediction region, wherein candidate reference position points that decide a coverage area of the union region are set as reference position points; if there is no intersection among the intersection regions of the respective channel consistency spaces of the any n candidate reference position points, each of the intersection regions is divided as one prediction region, respectively, wherein candidate reference position points that decide a coverage area of each of the intersection regions are set as reference position points, wherein n is a positive integer greater than or equal to the second threshold value.

[0024] Alternatively, the dividing the coverage region to obtain the at least one prediction region according to the channel spatial consistency further includes: in response to the network node having pre-acquired environmental information, being capable of Line-of-Sight (LOS) monitoring and / or UE speed estimation, pre-classifying each of the plurality of candidate reference position points, according to the environmental information, monitored LOS information and / or estimated UE speed magnitude and direction information, wherein a result of the pre-classification is used to divide the coverage region to obtain the at least one prediction region, wherein the any n position points are from the same classification in the result of the pre-classification.

[0025] Alternatively, the dividing the coverage region to obtain the at least one prediction region according to the channel spatial consistency includes: dividing the coverage region to obtain the at least one prediction region through a second AI network, based on the channel spatial consistency and position information of the respective position points in the coverage region.

[0026] Alternatively, the dividing the coverage region to obtain the at least one prediction region according to the channel spatial consistency further includes: in response to the network node having pre-acquired environmental information, being capable of Line-of-Sight (LOS) monitoring and / or UE speed estimation, dividing the coverage region to obtain the at least one prediction region through a third AI network, based on the channel spatial consistency, the position information of the respective position points in the coverage region, the environmental information, monitored LOS information and / or estimated UE speed magnitude and direction information.

[0027] Alternatively, the third threshold value is determined based on the CSI prediction accuracy requirement.

[0028] Alternatively, the method further includes: updating prediction regions within the coverage region of the network node.

[0029] Alternatively, the updating the prediction regions within the coverage region of the network node includes: in response to a UE located outside the prediction region within the coverage region of the network node moving into a first prediction region, classifying the UE as a reference position point, or a predictable position point of which CSI information is capable of being predicted.

[0030] Alternatively, the classifying the UE as the reference position point or the predictable position point of which the CSI information is capable of being predicted includes: generating a second prediction region according to all reference position points of the first prediction region and this UE; in response to a region range of the first prediction region being less than a region range of the second prediction region, classifying this UE as the reference position point of the second prediction region; in response to the region range of the first prediction region being equal to the region range of the second prediction region, classifying this UE as the predictable position point of which the CSI information is capable of being predicted, in the first prediction region, and configuring this UE not to report the CSI information to the network node.

[0031] Alternatively, the updating the prediction regions within the coverage region of the network node includes: in response to expiration of a prediction region generation period, determining whether a degree of aggregation of a plurality of UEs located outside the prediction regions within the coverage region of the network node exceeds a fourth threshold value; in response to the degree of aggregation of the plurality of UEs exceeding the fourth threshold value, dividing the coverage region to obtain a new prediction region based on channel spatial consistency of the plurality of UEs, wherein the plurality of UEs are set as reference position points or predictable position points of which the CSI information is capable of being predicted.

[0032] Alternatively, the updating the prediction regions within the coverage region of the network node includes: for each prediction region within the coverage region, determining whether a CSI prediction accuracy of the each prediction region is lower than a fifth threshold value, according to CSI received from predictable position points within the each prediction region and CSI predicted for the predictable position points; in response to a CSI prediction accuracy of any prediction region is lower than the fifth threshold value, deleting this prediction region, and for reference position points and predictable position points of this prediction region, performing re-division for a prediction region.

[0033] Alternatively, the updating the prediction regions within the coverage region of the network node includes: in response to a shift of a time period occurring, updating all prediction regions and all reference position points within the coverage region.

[0034] Alternatively, the updating the prediction regions within the coverage region of the network node includes: periodically performing the following operations: determining CSI prediction accuracy of each predictable position point within the coverage region; in response to a proportion of the predictable position points within the coverage region of which the CSI prediction accuracy does not satisfy a predetermined requirement, being greater than a sixth threshold value, updating all prediction regions and all reference position points within the coverage region; in response to the proportion of the predictable position points within the coverage region of which the CSI prediction accuracy does not satisfy the predetermined requirement, being less than or equal to the sixth threshold value, updating prediction regions corresponding to the predictable position points of which the CSI prediction accuracy does not satisfy the predetermined requirement, and reference position points of these prediction regions.

[0035] Alternatively, the updating all prediction regions and all reference position points within the coverage region includes: deleting all prediction regions and all reference position points in the coverage region; and re-determining prediction regions and reference position points in each hotspot region for a current time period according to the each hotspot region and candidate position points in the each hotspot region.

[0036] Alternatively, the updating the prediction regions within the coverage region of the network node includes: in response to at least one new UE appears at an edge of the first prediction region within the coverage region of the network node, expanding the first prediction region and increasing reference position points of the expanded first prediction region.

[0037] According to a second aspect of an embodiment of the present application, an method performed by a network node is provided, which includes: determining a plurality of reference position points from among a set of reference position points, according to position information of first UE; predicting CSI of the first UE, according to historical CSI of the plurality of reference position points, wherein the set of reference position points is determined from among respective position points within a coverage region of the network node according to channel spatial consistency coefficients between the respective position points, CSI report frequencies and CSI report periods of the respective position points.

[0038] Alternatively, determining the set of reference position points from among the respective position points includes: determining the channel spatial consistency coefficients between the respective position points, according to historical CSI received at the respective position points; determining weights corresponding to CSI validity of the respective position points, according to the channel spatial consistency coefficients between the respective position points, the CSI reporting frequencies and the CSI reporting periods of the respective position points; selecting a plurality of position points from among the respective position points as the set of reference position points, according to the weights corresponding to CSI validity of the respective position points.

[0039] Alternatively, the selecting the plurality of position points from among the respective position points as the set of reference position points, according to the weights corresponding to CSI validity of the respective position points includes: selecting, from among the respective position points, a plurality of position points having the largest sum of weights and the smallest total number of position points as the set of reference position points, wherein a channel spatial consistency coefficient between each of the respective position points other than the set of reference position points and reference position points of which the number is at least a second threshold value in the set of reference position points is greater than a third threshold value.

[0040] Alternatively, the determining the plurality of reference position points from among the set of reference position points, according to position information of first UE, includes: selecting the plurality of reference position points from among the set of reference position points, according to the CSI prediction accuracy requirement and / or the processing capability of the network node.

[0041] Alternatively, the selecting the plurality of reference position points from among the set of reference position points, according to the CSI prediction accuracy requirement and / or the processing capability of the network node, includes: in response to a first number of reference position points determined according to the CSI prediction accuracy requirement and / or the processing capability of the network node being less than a first threshold value, selecting, from among the set of reference position points, reference position points whose position are closest to the first UE and of which the number being the first number of reference position points; in response to the first number of reference position points determined according to the CSI prediction accuracy requirement and / or the processing capability of the network node being greater than or equal to a first threshold value, selecting, from among the set of reference position points, reference position points whose position are closest to the first UE and of which the number being the first threshold value.

[0042] Alternatively, there is a UE in a connected state at each of the plurality of reference position points previously or currently.

[0043] Alternatively, the predicting the CSI of the first UE, according to the historical CSI of the plurality of reference position points, includes: predicting the CSI of the first UE through a first AI network, according to the position information of the first UE, the historical CSI of the plurality of reference position points, and reporting position information of the historical CSI.

[0044] According to a third aspect of an embodiment of the present application, a network node is provided, which includes: a transceiver for transmitting and receiving a signal; and a processor coupled to the transceiver and configured to perform the method performed by the network node as described above.

[0045] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium storing instructions is provided, wherein the instructions, when being executed by at least one processor, cause the at least one processor to perform the method performed by the network node as described above.

[0046] The beneficial effects brought by the technical solutions provided by the embodiments of the present application will be described in the later section in combination with specific optional embodiments, or may be learned from descriptions of the embodiments, or may be learned from implementation of the embodiments.

[0047] In order to more clearly and easily explain and understand technical solutions in embodiments of the present application, a brief introduction will be given to the accompanying drawings required in the description of the embodiments of the present application below.

[0048] FIG. 1 is a flowchart illustrating a method performed by a network node according to an exemplary embodiment of the present application.

[0049] FIG. 2 is a schematic diagram illustrating classification of points within a coverage region of a network node according to an exemplary embodiment of the present application.

[0050] FIG. 3A is a flowchart illustrating a process of dividing a coverage region of a network node to obtain at least one prediction region according to an exemplary embodiment of the present application.

[0051] FIG. 3B is a diagram illustrating an example of collecting historical CSI by a network node according to an exemplary embodiment of the present application.

[0052] FIG. 3C is a diagram illustrating an example of dividing time into time periods according to an exemplary embodiment of the present application.

[0053] FIG. 3D is a diagram illustrating an example of hotspot region division according to an exemplary embodiment of the present application.

[0054] FIG. 3E is a flowchart illustrating a process of performing channel spatial consistency analysis for respective position points in each hotspot region for each time period, according to an exemplary embodiment of the present application.

[0055] FIG. 3F is a diagram illustrating an example of calculating a channel spatial consistency coefficient between position points according to an exemplary embodiment of the present application.

[0056] FIG. 3G is a diagram illustrating a process for determining a channel spatial consistency coefficient by an interpolation method according to an exemplary embodiment of the present application.

[0057] FIG. 3H is a diagram illustrating an example of channel spatial consistency distances under a predetermined threshold according to an exemplary embodiment of the present application.

[0058] FIG. 3I is a diagram illustrating an example of determining typical position points from among respective position points in each hotspot region for one time period, according to exemplary embodiments of the present application.

[0059] FIG. 4A is a diagram of a result illustrating one example of results of spatial channel consistency analysis according to an exemplary embodiment of the present application.

[0060] FIG. 4B is a flowchart illustrating a process for determining a prediction region according to an exemplary embodiment of the present application.

[0061] FIG. 4C is a schematic diagram illustrating a process for determining typical position points according to an exemplary embodiment of the present application.

[0062] FIG. 4D is a diagram illustrating an example of determining a geographic region at which a channel spatial consistency coefficient is greater than a third threshold value ｐ or any candidate reference position point, according to exemplary embodiments of the present application.

[0063] FIG. 5 is a schematic diagram illustrating prediction region division according to an exemplary embodiment of the present application.

[0064] FIG. 6A is a diagram illustrating a process for determining reference position points according to an exemplary embodiment of the present application.

[0065] FIG. 6B is a diagram illustrating a process for determining reference position points according to another exemplary embodiment of the present application.

[0066] FIG. 6C is a diagram illustrating a process for determining reference position points according to another exemplary embodiment of the present application.

[0067] FIG. 6D is a diagram illustrating a process for determining reference position points according to another exemplary embodiment of the present application.

[0068] FIG. 6E is a diagram illustrating an example of a CSI reporting setting to a UE by a network node according to an exemplary embodiment of the present application.

[0069] FIG. 6F is a diagram illustrating an example of determining a prediction region to which a UE to be predicted belongs, according to an exemplary embodiment of the present application.

[0070] FIG. 7A is a flowchart illustrating a process for predicting CSI of the UE to be predicted according to historical CSIs of a plurality of reference position points in the prediction region, according to an exemplary embodiment of the present disclosure.

[0071] FIG. 7B is a schematic diagram illustrating a CSI prediction process according to a first AI network according to an exemplary embodiment of the present application.

[0072] FIG. 7C is a diagram illustrating a structure of an embedding layer according to an exemplary embodiment of the present application.

[0073] FIG. 7D is a diagram illustrating a process of data processing by a first AI network according to an exemplary embodiment of the present application.

[0074] FIG. 8A is a flowchart illustrating a process of predicting a CSI of the UE to be predicted according to historical CSI of a plurality of reference position points corresponding to the prediction region, according to another exemplary embodiment of the present disclosure.

[0075] FIG. 8B is a schematic diagram illustrating a CSI prediction process according to a first AI network according to another exemplary embodiment of the present application.

[0076] FIG. 9A illustrates a schematic diagram of a process for updating a prediction region according to an exemplary embodiment of the present application.

[0077] FIG. 9B is a diagram illustrating a process for updating prediction regions and / or reference position points according to an exemplary embodiment of the present application.

[0078] FIG. 9C is a diagram illustrating a process of a shift of a time period occurring according to an exemplary embodiment of the present application.

[0079] FIG. 9D is a detailed flowchart illustrating a method performed by a network node according to an exemplary embodiment of the present application.

[0080] FIG. 10 is a schematic diagram illustrating an overall process of a method performed by a network node according to an exemplary embodiment of the present application.

[0081] FIG. 11 is a schematic diagram illustrating a process of a method performed by a network node according to another exemplary embodiment of the present application.

[0082] FIG. 12 is a flowchart illustrating a method performed by a network node according to another exemplary embodiment of the present application.

[0083] FIG. 13 is a flowchart illustrating a process for determining a set of reference points according to an exemplary embodiment of the present application.

[0084] FIG. 14 illustrates one example of G=(V,E) constructed according to the method described above.

[0085] FIG. 15 is a schematic diagram illustrating a process for determining a set S of reference position points according to an exemplary embodiment of the present application.

[0086] FIG. 16A illustrates a schematic diagram of a process of a method of FIG. 12 performed by a network node, according to an exemplary embodiment of the present application.

[0087] FIG. 16B is a block diagram illustrating a network node according to an exemplary embodiment of the present application.

[0088] FIG. 17 is a schematic diagram of a structure illustrating of an electronic apparatus applicable to an embodiment of the present application.

[0089] The following description with reference to the accompanying drawings is provided to aid in a thorough understanding of various embodiments of the present disclosure as defined by claims and equivalents thereof. This description includes various specific details to aid in understanding but should only be considered exemplary. Accordingly, those ordinary skill in the art will recognize that various changes and modifications can be made to the various embodiments described herein without departing from the scope and spirit of the present disclosure. In addition, descriptions of well-known features and structures may be omitted for the sake of clarity and brevity.

[0090] The terms and phrases used in the claims and the following description are not limited to dictionary meaning thereof, but are used only by the inventor to enable a clear and consistent understanding of the present disclosure. Accordingly, it should be apparent to those skilled in the art that, the following description of the various embodiments of the present disclosure is provided for an illustrative purpose only and is not intended to a purpose of limiting the present disclosure as defined by the appended claims and equivalents thereof.

[0091] In various examples of the disclosure described below, a hardware approach will be described as an example. However, since various embodiments of the disclosure may include a technology that utilizes both the hardware-based and the software-based approaches, they are not intended to exclude the software-based approach.

[0092] It should be understood that, "a", "an" and "the" in a singular form may also include a plural reference, unless the context clearly indicates otherwise. Thus, for example, a reference to a "part surface" includes a reference to one or more such surfaces. When it refers to one element as being "connected" or "coupled" to another element, the one element may be directly connected or coupled to the other element, or it may refer to a connection relationship between the one element and the other element established through an intermediate element. In addition, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled.

[0093] The term "include" or "may include" refers to the presence of a function, operation, or component of the corresponding disclosure that may be used in the various embodiments of the present disclosure, and does not limit the presence of one or more additional functions, operations, or features. In addition, the terms "include" or "have" may be interpreted to denote certain features, figures, steps, operations, constituent elements, components, or combinations thereof, but should not be interpreted to exclude the possibility of the presence of one or more other features, figures, steps, operations, constituent elements, components, or combinations thereof.

[0094] The term "or" as used in the various embodiments of the present disclosure includes any of the listed terms and all combinations thereof. For example, "A or B" may include A, may include B, or may include both A and B. When describing a plurality of (two or more) items, the plurality of items may refer to one, more, or all of the plurality of items if a relationship among the plurality of items is not explicitly defined. For example, for the description "a parameter A comprises A1, A2, A3", it may be implemented as parameter A comprising A1, A2 or A3, or as parameter A comprising at least two of the three items of the parameter A1, A2, A3.

[0095] All terms (including technical or scientific terms) used in the present disclosure have the same meaning as understood by those skilled in the art to which the present disclosure belongs, unless defined differently. Common terms as defined in dictionaries are interpreted to have a meaning consistent with the context in the relevant technology art and should not be interpreted in an idealized or overly formalistic manner, unless expressly so defined in the present disclosure.

[0096] At least part of the functions in a device or electronic apparatus provided in the embodiments of the present disclosure may be implemented through an AI model, such as, at least one of a plurality of modules of the device or electronic apparatus may be implemented through the AI model. A function associated with AI may be performed through the non-volatile memory, the volatile memory, and the processor.

[0097] The processor may include one or more processors. At this time, the one or more processors may be a general purpose processor, such as a central processing unit (CPU), an application processor (AP), or the like, or may be a graphics-only processing unit such as a graphics processing unit (GPU), a visual processing unit (VPU), and / or an AI-dedicated processor such as a neural processing unit (NPU).

[0098] The one or more processors control processing of input data in accordance with a predefined operating rule or artificial intelligence (AI) model stored in the non-volatile memory and the volatile memory. The predefined operating rule or artificial intelligence model is provided through training or learning.

[0099] Here, being provided through learning means that, by applying a learning algorithm to a plurality of learning data, a predefined operating rule or an AI model of a desired characteristic is made. The learning may be performed in a device or electronic apparatus itself in which AI according to an embodiment is performed, and / or may be implemented through a separate server / system.

[0100] The AI model may consist of a plurality of neural network layers. Each layer has a plurality of weight values, and performs a neural network calculation by calculating between the input data of this layer (such as, a calculation result of the previous layer and / or the input data of the AI model) and the plurality of weight values of the current layer. Examples of neural networks include, but are not limited to, a convolutional neural network (CNN), a deep neural network (DNN), a recurrent neural network (RNN), a restricted Boltzmann Machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), a generative adversarial networks (GAN), and a deep Q-network.

[0101] The learning algorithm is a method for training a predetermined target device (for example, a robot) using a plurality of learning data to cause, allow, or control the target device to make a determination or prediction. Examples of the learning algorithm include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0102] According to the present disclosure, at least one step of the method executed in an electronic apparatus, may be implemented using an artificial intelligence model. Processors of the electronic apparatus may perform a pre-processing operation on the data to convert into a form appropriate for use as an input for the artificial intelligence model. The artificial intelligence model may be obtained by training. Here, "obtained by training" means that a predefined operation rule or artificial intelligence model configured to perform a desired feature (or purpose) is obtained by training a basic artificial intelligence model with multiple pieces of training data by a training algorithm.

[0103]

[0104] A total CSI feedback overhead increases as the number of antennas of the base station increases, and in particular, in a cell with a dense number of UEs, the total CSI feedback overhead increases more significantly as the number of antennas of the base station increases. Specifically, the total CSI feedback overhead may be related to the number N of activated UEs in a cell, the size S of a single piece of CSI feedback information, and a CSI feedback period P (in milliseconds). For example, the total CSI feedback overheadTotalOverheadmay be calculated by the following equation (1):

[0105]

[0106] The existing method devotes to address a problem of reducing the total CSI overhead by reducing individual CSI overhead or increasing the CSI feedback period, while maintaining channel accuracy, for example:

[0107] (1) The three methods described above continuously investigate a solution for overhead reduction by no-compression, compression, and further compression of CSI information. A CSI compression technique has some gain in overhead reduction, but the gain of CSI compression and an AI-based CSI compression method relies on an assumption of sparsity in a channel structure. Since an actual channel is not completely sparse in most scenarios, the overhead gain obtained by the CSI compression relies on environment. In this technique, in order to cope with a channel aging problem, the UE still needs to periodically report the compressed CSI information, and the overhead required for periodic CSI reporting still exists. In addition, different compression ratios cause different compression losses, the higher the compression ratio, the higher the compression loss, and the lower the predicted channel accuracy will be.

[0108] (2) The AI-based CSI prediction method predicts the CSI of the UE at the future time by monitoring the historical CSI information. Although a part of the overhead required for CSI reporting may be reduced, the CSI prediction adopts a single UE time-domain correlation, and a source of the prediction information is a single UE, available sample points are fewer, which affects the prediction accuracy. If the sample points are increased, the CSI overhead cannot be well ensured.

[0109] To summarize, the current CSI compression method devotes to reducing the size of a single CSI report, but need to face inevitable compression loss, and the current AI-based CSI prediction is only a single-user CSI prediction, the CSI prediction accuracy is low, so it is difficult for the current scheme to maintain a balance between overhead and accuracy, and a method that ensures the accuracy of channel information on a basis of effectively reducing the overhead required for the CSI reporting is needed. Unlike the current scheme, the present application devotes to reduce the number of UEs that report the CSI in a cell, i.e., only partial UEs are required to report CSI, and other UEs are not required to report CSI, but instead, prediction may be performed for the CSI of the other UEs according to the CSI of the UEs that report the CSI. Thus, in the present application, the relatively low compression ratio and smaller period may be used in performing CSI reporting, it may maintain the high CSI accuracy on a basis of reducing the overall CSI reporting overhead of the cell.

[0110] Below, the technical solutions of the embodiments of the disclosure and the technical effects produced by the technical solutions of the disclosure will be explained by describing several optional embodiments. It should be noted that, the following embodiments may be referred to, imitated or combined with each other, and the same term, similar features and similar implementation steps in different embodiments will not be described repeatedly.

[0111] FIG. 1 is a flowchart illustrating a method performed by a network node according to an exemplary embodiment of the present application. In the present application, the network node may be a base station, but the present application is not limited thereto. The network node may also be a network server that may receive various information from the base station and predict CSI of the UE according to the received various information, and transmit the predicted CSI to the base station. The network node may be composed of a single entity or may be composed of a plurality of sub-entities. When the network node is composed of the plurality of sub-entities, different functions may be achieved by the plurality of sub-entities, respectively, each sub-entity may have a corresponding name, and a connection between the sub-entities may be a wired connection or a wireless connection, the present application is not limited thereto. For ease of description, in the following description, the example of a network node being a base station is used for illustration.

[0112] As shown in FIG. 1, at step S110, a Predictable Region (PR) corresponding to a first UE is determined according to position information of the first UE, wherein in the present application, the prediction region may also be referred to as a predictable region. In the following description, the first UE is also referred to as a "UE to be predicted"

[0113] Specifically, the present application divides a coverage region of a network node to obtain at least one prediction region according to historical CSIs received by the network node in the coverage region of the network node, wherein each prediction region is determined by a plurality of reference position points, and there is a corresponding UE in a connected state at each reference position point previously or currently, this UE and this reference position point may be referred to as a "reference UE" "CSI reporting UE", "CSI feedback UE", "reporting UE" or "feedback UE" In the present application, a "UE in the connected state" may also be referred to as an "active UE"; historical CSI of a plurality of reference position points corresponding to one prediction region may be used to predict CSI of a non-reference position point that is located within this prediction region, and such non-reference position point may be referred to as a "predictable point" a "request point" or a "point to be predicted" and the UE, of which CSI may be predicted, located at this non-reference position point may be referred to as an "active UE" The UE located at the non-reference position point that may be predicted by CSI may be referred to as a "predictable UE" a "non-feedback UE" a "non-CSI reporting UE" a "non-CSI feedback UE" "request UE" or "UE to be predicted" In addition, a position point located outside the prediction region within the coverage region that is different from the reference position points may be referred to as a "non-predictable point" or a "non-request point" and a UE located at such position point may be referred to as a "non-predictable UE" or a "non-request point" In the present application, a UE located at a reference position point of the prediction region is configured to periodically report CSI information to the network node, and an "unpredictable UE" located outside the prediction region is configured to periodically report CSI information to the network node, and a predictable UE located within the prediction region is configured not to need to report the CSI information to the network node, as shown in FIG. 2. In one embodiment of the present application, firstly, according to CSI collected by a network node from respective UEs within coverage region thereof for a longer period of time, hotspot regions may be identified from the coverage region for each time period. Then, candidate position points (which may also be referred to as "typical position points") for dividing prediction regions may be determined from the hotspot regions for each time period. Then, for each hotspot region for a current time period, a plurality of candidate reference position points (i.e., there is an active UE at or near this candidate reference position point) are selected from among the respective typical position points thereof. Then, the prediction regions are determined according to channel consistency spaces of the plurality of candidate reference position points. In the following description, before determining the typical position points, a "position point" is used to collectively refer to a position of the UE at the time of CSI reporting, and after dividing to obtain the prediction region, a "reference position point" a "predictable point" and a "unpredictable point" are used to represent individual position points or UEs classified in accordance with the above classification. Before describing step S110 in detail, a process of dividing the coverage region of the network node to obtain at least one prediction region is described in detail.

[0114] FIG. 3A is a flowchart illustrating a process of dividing a coverage region of a network node to obtain at least one prediction region, according to an exemplary embodiment of the present application.

[0115] As shown in FIG. 3A, at step S310, channel spatial consistency between respective position points within the coverage region of the network node is determined according to historical CSI received within the coverage region.

[0116] In exemplary embodiments of the present application, step S310 may include: identifying at least one hotspot region within the coverage region of the network node for each time period, according to the historical CSI received within the coverage region; performing channel spatial consistency analysis between the respective position points in each hotspot region for each time period, according to the historical CSI received within each hotspot region for each time period. This will be described in detail below.

[0117] Specifically, firstly, over a long period of time (e.g., tens of days), the network node may collect CSI periodically reported by the UE, position information "Position (i.e., reporting position information) of the UE at the time of reporting the CSI, reporting time information "Time", and a corresponding UE ID, wherein this reporting position information may be obtained by the network node according to a positioning method, as shown in FIG. 3B. In addition, as shown in FIG. 3C, the network node may divide time of one day into different time periods, and whether it is a holiday or not may be taken into account in the division of the time periods. On this basis, the network node may divide data samples (Time, UE ID, CSI, Position) including the above collected information to different time periods. For example, data samples corresponding to Time being 6:30 on weekdays are divided to the time period of 6:00~9:00 on weekdays.

[0118] Thereafter, the network node may identify at least one hotspot region within the coverage region of the network node for each time period according to the collected historical CSI. Specifically, for each time period, the network node may calculate a sample density per hour within each unit coverage region. For example, assuming that area of each unit coverage region of the network node is Cover, and that the network node successively receives a total of K data samples from UEs within one certain unit coverage region within a current time period of 6:00~9:00 (i.e., 3 hours), the sample density per hour within this unit coverage region is K / (Cover*3). Then, the network node compares the sample density per hour in each unit coverage region with a predetermined threshold value, and if the sample density per hour in a certain unit coverage region is greater than this predetermined threshold value, this unit coverage region may be identified as a high-density region, and a plurality of adjacent high-density regions (i.e., patches of high-density regions) are connected together to form a hotspot region. Due to a mobility of the UE and distribution characteristics of crowd, the hotspot regions formed for different time periods may be different, for example, (a) and (b) of FIG. 3D illustrate hotspot regions formed for time periods 1 and 2, respectively.

[0119] After identifying at least one hotspot region from the coverage region of the network node, channel spatial consistency analysis may be performed for respective position points in each hotspot region for each time period. This is described below with reference to FIG. 3E.

[0120] FIG. 3E is a flowchart illustrating a process of performing a channel spatial consistency analysis for each position point in each hotspot region for each time period, according to an exemplary embodiment of the present application.

[0121] As shown in FIG. 3E, at step S3001, channel spatial consistency coefficients between each position point and other position points within each hotspot region for each time period is determined, according to the historical CSI received within each hotspot region for each time period.

[0122] Specifically, for one time period, a circle is drawn centered on each position point (i.e., a position at which the UE reports the CSI) within each hotspot region, with a radius of a preset length Thd, and then, for each position point, the channel spatial consistency coefficient between this position point and each other position point within a circle corresponding to this position point is calculated, as shown in FIG. 3F. How to calculate the channel spatial consistency coefficient between two position points (for example, a first position point and a second position point) is described below.

[0123] Firstly, a channel matrix H of each position point may be determined according to CSI of each position point, using the following equation (2):

[0124]

[0125]

[0126] Then, an autocorrelation matrix R of the channel matrix H of each position point may be determined, e.g., the autocorrelation matrix R may be determined according to the following equation (3):

[0127]

[0128]

[0129]

[0130]

[0131]

[0132] The channel spatial consistency coefficients between each position point and other position points within each hotspot region for each time period may be determined by the above operations.

[0133] At step S3002, a channel spatial consistency distance of the each position point in each direction is determined, according to the channel spatial consistency coefficients between the each position point and the other position points.

[0134]

[0135] Then, according to the above calculated channel spatial consistency coefficients of each position point in the respective directions and the respective distances, the channel spatial consistency distance of each position point in the each direction is obtained, this channel spatial consistency distance represents a maximum distance at which a channel spatial consistency coefficient is greater than a predetermined threshold, e.g., FIG. 3H shows a diagram of channel spatial consistency distances when this predetermined threshold is 0.8.

[0136] Furthermore, in the present application, after the channel spatial consistency distance is determined, a portion of the position points in each hotspot region for each time period may be selected as candidate position points (i.e., typical position points) for division of the prediction region. That is, a plurality of typical position points may be determined from among the respective position points in each hotspot region for each time period according to channel spatial consistency distances.

[0137] Specifically, due to a mobility of the UE, the position of the UE is not fixed, and therefore, in the present application, the channel spatial consistency coefficients and the channel spatial consistency distances (both are included in channel spatial consistency information) obtained through the above steps S3001 and S3002 actually belong to the corresponding position point. In order to reduce the amount of information storage, in the present application, the network node may select typical position points from among a large number of position points and store channel spatial consistency coefficients and channel spatial consistency distances of these typical position points to represent channel information of the hotspot regions at which these typical position points are located. A process of a method for determining (i.e., selecting) the typical position points is described below with reference to FIG. 3I.

[0138] FIG. 3I is a diagram illustrating an example of determining typical position points from respective position points in each hotspot region for one time period, according to an exemplary embodiment of the present application.

[0139] As shown in FIG. 3I, for each hotspot region, the following steps S1 to S4 are performed.

[0140] At step S1, starting from an edge of a hotspot region in a first direction (e.g., a left-to-right direction), first encountered position point (e.g., a leftmost position point) is selected as one typical position point (i.e., a first typical position point), and a channel consistency space of this typical position point is determined according to the channel spatial consistency coefficient of this position point being greater than a predetermined threshold, wherein, other position points within this channel consistency space are no longer selected as typical position points.

[0141] At step S2, a next first encountered position point (e.g., a next leftmost position point) is continued to be searched in the first direction, and is selected as one typical position point (i.e., a second typical position point), and a channel consistency space of this typical position point is determined in the same manner as in step S1.

[0142] At step S3, step S2 is repeated until the last position point, which is selected as a typical position point.

[0143] At step S4, channel spatial consistency information (including channel spatial consistency coefficients and channel spatial consistency distances) of the typical position points selected according to steps S1 to S3 is stored.

[0144] Although in the example described above with reference to FIG. 3I, the typical position points are selected in a left-to-right direction, the present application is not limited to this, and the typical position points may also be selected in a top-to-bottom direction, a bottom-to-up direction, a right-to-left direction, or a top-left to bottom-right direction, and so on. That is, the present application does not make a specific limitation on the first direction, which may be any direction.

[0145] By performing the operations described above with reference to FIG. 3I for each hotspot region for each time period, each typical position point in each hotspot region for each time period may be determined, and thus the determined channel spatial consistency information may of each typical position point be stored.

[0146] As mentioned above, for any position point in the coverage region of the network node, according to channel spatial consistency coefficients thereof, a channel consistency space may be obtained by dividing the coverage region, and in this channel consistency space, a channel at any position has a strong consistency with this position point, that is, CSI of this position point may be used for channel prediction on any other position point in this channel consistency space.

[0147] As described above, the network node may determine the channel spatial consistency distance of each position point in the each direction, as shown in FIG. 4A, wherein channel spatial consistency distances of a Non Line of Sight (NLOS) environment are much smaller than that of a Line of Sight (LOS) environment or a light NLOS environment. The channel spatial consistency information of each position point may include at least one of channel matrix H thereof, channel spatial consistency coefficients of this position point with neighboring position points in the respective directions and channel spatial consistency distances of this position point in the respective directions. In one exemplary embodiment, the network node may store the channel spatial consistency information of each position point. In addition, in another exemplary embodiment, due to complexity of the channel environment, the present application may store the channel spatial consistency information of each position point according to a non-uniform granularity method. Specifically, when the channel environment is poor, the present application may store the channel spatial consistency information of the position points in accordance with a higher density granularity, that is, for coverage regions of the same size, channel spatial consistency information of more position points is stored. When the channel environment is better, the present application may store the channel spatial consistency information of the position points in accordance with a lower density granularity, that is, for coverage regions of the same size, the channel spatial consistency information of fewer position points is stored. By this storage method, the storage overhead may be reduced while ensuring the accuracy in CSI prediction. In another exemplary embodiment, as described above, a network node may determine the plurality of typical position points from among the respective position points in each hotspot region for each time period according to the channel spatial consistency distances, and store the channel spatial consistency information of only these determined typical position points.

[0148] Referring back to FIG. 3A, at step S320, the coverage region is divided to obtain the at least one prediction region according to the channel spatial consistency.

[0149] Specifically, according to the channel spatial consistency analysis described above, CSI reported by other UEs may be used to predict CSI of a UE that is closer to the other UEs, and for this purpose, the present application generates one prediction region according to a plurality of UEs and / or position points in close proximity to each other, and the channels have consistency in this predication region and therefore may be used for mutual derivation.

[0150] Furthermore, in the present application, when dividing the coverage region to obtain prediction regions, the prediction regions may include, but is not limited to, a prediction map region, a prediction puzzle region, and a prediction dictionary region. In the present application, a non-AI method or an AI method may be used to divide the coverage region of the network node to obtain at least one prediction region. A process of dividing to obtain the prediction regions using the non-AI method is firstly described below.

[0151] In one exemplary embodiment of the present application, firstly, it is needed to determine a time period to which current time belongs (i.e., a current time period), and then, the prediction regions for each hotspot region of the current time period is obtained by division. The process of dividing to obtain the prediction regions is described in detail below taking one hotspot region as an example with reference to FIG. 4B.

[0152] FIG. 4B is a flowchart illustrating a process for determining a prediction region according to an exemplary embodiment of the present application.

[0153] As shown in FIG. 4B, at step S410, a plurality of candidate reference position points are determined according to respective candidate position points (i.e., respective typical position points) in a current hotspot region. For example, one active UE that exists within the channel consistency space of each typical position point in the current hotspot region and that is closest to this typical position point is determined as one candidate reference position point.

[0154] Specifically, according to the channel spatial consistency information of each typical position point, a corresponding channel consistency space (i.e., a coverage region of the typical position point) is determined, and channel spatial consistency coefficients within this channel consistency space are not lower than a predetermined threshold value, this predetermined threshold indicates that channels have strong consistency and mutually representable within the channel consistency space; and then, an active UE is selected from the channel consistency space, as a UE for CSI reporting (which may also be referred to as a candidate reference position point, reporting UE, etc.), wherein the active UE represents a UE in a connected state in the network. If there is one active UE in the channel consistency space, this one active UE is selected as the UE for CSI reporting, and if there are at least two active UEs in the channel consistency space, an active UE closest to the typical position point corresponding to this channel consistency space is selected as the UE for CSI reporting, and if there is no active UE in this channel consistency space, the UE for CSI reporting is not selected for this channel consistency space, as shown in FIG. 4C, wherein D represents a distance of a position where the channel space consistency coefficient is greater than the above-described predetermined threshold value, from this typical position point.

[0155] At step S420, a channel consistency space of each candidate reference position point is determined according to channel spatial consistency distances of each candidate reference position point and a third threshold value.

[0156] Specifically, for the CSI or CSI feature vectors received from the above-determined candidate reference position points (i.e., UEs performing CSI reporting) for a sampling window period, channel spatial consistency coefficients and channel spatial consistency distances are calculated for each candidate reference position point in accordance with the process described above with reference to FIG. 3E, and then a geographic region (i.e., the channel consistency space), in which a channel spatial consistency coefficient is greater than a third threshold value ｐ, for any candidate reference position point is determined, such as S1 shown in FIG. 4D.

[0157] At step S430, prediction regions in the current hotspot region are determined according to the channel consistency space of each candidate reference position point.

[0158] Specifically, in the first embodiment, for any n candidate reference position points among the plurality of candidate reference position points in the current hotspot region, if there is no intersection among intersection regions of respective channel consistency spaces of the any n candidate reference position points, each of the intersection regions is divided as one prediction region, respectively. That is, if there is no intersection among the intersection regions where channel spatial consistency coefficients are greater than the third threshold value, of the any n candidate reference position points, each of the intersection regions may be divided as one prediction region, respectively, wherein candidate reference position points that decide a coverage area of each of the intersection regions are set as reference position points.

[0159] In a second embodiment, for any n candidate reference position points among the plurality of candidate reference position points in the current hotspot region, if there is an intersection among intersection regions of respective channel consistency spaces of the any n candidate reference position points, an union region of the intersection regions is divided as one prediction region. That is, there is an intersection among the intersection regions where channel spatial consistency coefficients are greater than the third threshold value, of the any n candidate reference position points, the union region of the respective intersection regions may be divided as one prediction region, wherein n is a positive integer greater than or equal to the second threshold value, for example, a process of determining one prediction region may be represented by the following equations (5), (6) and (7):

[0160]

[0161]

[0162]

[0163]

[0164] After the prediction region is obtained by division in accordance with the above method, the candidate reference position points that decide this prediction region are set as the reference position points, in accordance with the reference position point screening strategy, e.g., FIG. 6A illustrates an example of two different prediction regions; furthermore, a candidate reference position point, that is not capable of forming a region satisfying the number n of position points, is deleted (as illustrated in FIG. 6B) and is set as an unpredictable position point (i.e., a unpredictable UE, which still requires periodic CSI reporting), for example, FIG. 6C illustrates an example of one prediction region and one unpredictable position point; furthermore, a candidate reference position point in the prediction region, that does not affect the coverage area of this prediction region, is deleted (as illustrated in FIG. 6B) and is set as a predictable position point (i.e., a predictable UE, which does not require periodic CSI reporting), for example, FIG. 6D illustrates an example where a candidate reference position point located in the very center does not affect a coverage area of this prediction region and it is set as a predictable position point. Finally, it may be ensured that, for any position in the formed prediction region, reference position points of which the number is at least a second threshold value are always be founded, to predict the CSI at that arbitrary position.

[0165]

[0166] In another exemplary embodiment of the present application, if a network node has pre-acquired environmental information, being capable of Line-of-Sight (LOS) monitoring and / or UE speed estimation, the network node may pre-classify each of a plurality of candidate reference position points (i.e., typical position points) in each hotspot region, according to the environmental information, monitored LOS information and / or estimated UE speed magnitude and direction information, wherein the environmental information, the monitored LOS information and / or the estimated UE speed magnitude and direction information may be referred to as optional auxiliary information. That is, which candidate reference position points having strong channel spatial consistency are preliminarily determined according to the environmental information, the monitored LOS information and / or the estimated UE speed magnitude and direction information, to preliminarily classify the candidate reference position points. For example, if　the estimated speed directions of two UEs are opposite and both moving at high speeds, they are determined to have weak channel spatial consistency, it is obvious that they are not be located in the same prediction region, and accordingly, they are classified into different classifications. The above pre-classification result may be used in a process of dividing a coverage region to obtain at least one prediction region according to the result of the channel spatial consistency analysis described in detail above, wherein the above any n position points are from a same classification in the result of the pre-classification, and this pre-classification operation can further reduce the overhead and the complexity of obtaining the prediction region by division, by considering typical position points of the same classification when constructing the prediction region.

[0167] The above describes a process of dividing a coverage region of a network node to obtain at least one prediction region according to the result of the channel spatial consistency analysis by the non-AI based method, however, the present application is not limited to this, and the present application may also divide a coverage region of the network node to obtain at least one prediction region according to an AI based method.

[0168] Specifically, in an exemplary embodiment of the present application, in the process of dividing the coverage region to obtain the at least one prediction region based on the result of the channel spatial consistency analysis, the coverage region may be divided to obtain the at least one prediction region through a second AI network, based on the channel spatial consistency and position information of the respective position points in the coverage region. This second AI network may be at least one of a RNN, a CNN, a DNN, a GAN, a Transformer network, etc., and the present application does not make any specific limitation thereto, as long as it is capable of dividing the coverage region of the network node to obtain the prediction region based on the result of the channel spatial consistency analysis and position information of the respective position points in the coverage region, it will fall within the protection scope of the present application.

[0169] In another exemplary embodiment of the present application, in the process of dividing the coverage region to obtain the at least one prediction region based on the result of the channel spatial consistency analysis, in response to the network node having pre-acquired environmental information, being capable of Line-of-Sight (LOS) monitoring and / or UE speed estimation, the coverage region may be divided to obtain the at least one prediction region through a third AI network, based on the channel spatial consistency, the position information of the respective position points in the coverage region, the environmental information, monitored LOS information and / or estimated UE speed magnitude and direction information. Similarly, the third AI network may be at least one of a RNN, a CNN, a DNN, a GAN, a Transformer network, etc., and the present application also does not make specific limitations thereto, as long as it is capable of dividing the coverage region of the network node to obtain the prediction region according to the above various information, it will fall within the protection scope of the present application.

[0170] In addition, the second AI network and the third AI network described above may be AI networks that are trained offline and deployed in the network node after training. Alternatively, they may be AI networks that are trained online in real time at the network node and thus are continuously updated.

[0171] A process of dividing the coverage region of the network node to obtain the at least one prediction region is described above with reference to FIGS. 3A to 5, wherein historical CSI information received by an active UE in the coverage region of the network node for a more recent time period is used in the division process, and the obtained prediction region by division may be referred to as a prediction puzzle region, and furthermore, the present application is not limited to this, wherein the process of obtaining the prediction region by division described above in the present application may not take into account whether there is an active UE recently in the coverage region of the network node, but rather uses historical CSI information previously received in this coverage region for a longer time period, and then performs the division of the prediction region according to the received historical CSI information in accordance with the process described above with reference to FIGS. 3A to 5, and after obtaining these prediction regions by division, it is necessary to determine which prediction regions are available according to active UEs present in the coverage region of the network node, wherein the prediction regions determined in this way may be referred to as prediction dictionary regions, such prediction dictionary regions may be used to predict the CSI of the UE to be predicted. If there is no such prediction dictionary region for the UE to be predicted, the current UE to be predicted may be configured to report the CSI. That is, in the case of the prediction dictionary regions, there is an active UE at each of the plurality of selected reference position points. Since how to obtain the prediction region by division has been specifically described above with reference to FIGS. 3A to 5, the division process for the prediction dictionary region is not described in detail herein.

[0172] On this basis, a prediction region corresponding to a UE to be predicted may be determined according to position information of the UE to be predicted at a step S110. Specifically, the step S110 may include: determining, from among at least one prediction region within the coverage region of the network node, the prediction region to which a position of the UE to be predicted belongs, according to the position information of the UE to be predicted, as shown in FIG. 6F, firstly, the position information of the UE to be predicted (i.e., the position information of the predictable UE) is determined, and then the prediction region to which the UE to be predicted belongs is determined according to the position information of the UE to be predicted. That is, when a UE to be predicted located at a non-reference position point needs to be scheduled, the network node may proactively on demand or according to the CSI prediction request from the UE, determine whether this UE to be predicted belongs to a certain prediction region by traversing according to the position information of this UE to be predicted. If the traversal reveals that the UE to be predicted is located in a certain prediction region, then the prediction region where the UE to be predicted is located may be selected as the prediction region to be used for predicting the CSI of the UE to be predicted.

[0173] Returning to refer to FIG. 1, at step S120, the CSI of the UE to be predicted is predicted according to historical CSI of a plurality of reference position points corresponding to the prediction region.

[0174] Specifically, step S120 may include selecting the plurality of reference position points from among reference position points of the prediction region, according to a CSI prediction accuracy requirement and / or processing capability of the network node. In the present application, considering the prediction overhead of the CSI and the computational complexity, there is a need to limit the number of the reference position points selected from among the reference position points of the prediction region. For this purpose, the present application may set a first threshold value, a second threshold value, and a third threshold value, wherein the first threshold value is used to limit the maximum number of the reference position points that may be used at the time of prediction, and the first threshold value is determined according to the processing capability of the network node. In other words, the upper limit of the number of the reference position points that may be used at the time of prediction is limited by the processing capability of the network node. The second threshold value is used to limit the minimum number of the reference position points that need to be used at the time of prediction, the first threshold value is greater than the second threshold value. The third threshold value is used to limit a size of the prediction region, wherein the second threshold value and the third threshold value may be determined through pre-training of the first AI network, according to a certain CSI prediction accuracy requirement. In other words, the first threshold value is used for high prediction accuracy with respect to CSI, and the second threshold value and the third threshold value are used for ensuring CSI prediction accuracy. In addition, the present application may obtain the number of the reference position points corresponding to each CSI prediction accuracy requirement and / or processing capability of the network node, by the result of the pre-training of the first AI network according to each CSI prediction accuracy requirement and / or processing capability of the network node, so that a plurality of numbers of reference positions may be per-obtained according to a plurality of CSI prediction accuracy requirements and / or processing capabilities of the network node. On this basis, the network node may select the number of reference position point positions corresponding the subsequent CSI prediction accuracy requirement from the UE and / or the processing capability of the network node from among the pre-obtained plurality of numbers of reference position points according to a subsequent CSI prediction accuracy requirement from the UE and / or the processing capability of the network node.

[0175] In one exemplary embodiment of the present application, in selecting the reference position points from among reference position points of the prediction region corresponding to the UE to be predicted, it is determined whether the number of the reference position points of the prediction region corresponding to the UE to be predicted is less than or equal to the first threshold value. In response to the number of the reference position points of the prediction region being less than or equal to the first threshold value, all reference position points of the prediction region are selected, that is, the CSI of the UE to be predicted is predicted using CSI of all reference position points corresponding to the prediction region.

[0176] In response to the number of the reference position points of the prediction region being greater than the first threshold value, the plurality of reference position points are selected from among the reference positions of the prediction region, according to the position information of the UE to be predicted, the CSI prediction accuracy requirement, and / or the processing capacity of the network node. In summary, considering the CSI prediction overhead and the computational complexity, the present application uses the first threshold value to limit the maximum number of the reference position points to be used for the CSI prediction operation, however it may be possible to predict CSI that satisfies the CSI prediction accuracy requirement and / or the processing capability of the network node by using fewer reference position points than the first threshold value. Therefore the present application further determines the number of reference position points from among the predetermined plurality number of reference position points according to the CSI prediction accuracy requirement of the UE to be predicted and / or the processing capability of the network node, and then, by comparing this number of reference position points with the first threshold value, determines a final selected number of the reference position points. Specifically, in response to the number of reference position points determined according to the CSI prediction accuracy requirement and / or the processing capacity of the network node being less than the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose positions are closest to the UE to be predicted and of which the number being this number of reference position points; in response to this number of first reference position point determined according to the CSI prediction accuracy requirement and / or the processing capacity of the network node being greater than or equal to the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose position are closest to the UE to be predicted and of which the number being the first threshold value.

[0177] The above describes a process of selecting reference position points from among the reference position points(i.e., the reference position points that decides coverage region of the prediction region) of the prediction region, and after selecting the plurality of reference position points, the CSI of the UE to be predicted is predicted according to historical CSI of the plurality of reference position points, this process may include: predicting the CSI of the UE to be predicted through a first AI network, according to the position information of the UE to be predicted, the historical CSI of the plurality of reference position points, and reporting position information of the historical CSI. The process of predicting the CSI of the UE to be predicted using the selected reference position points is described in detail below with reference to FIGS. 7A and 7B.

[0178] FIG. 7A is a flowchart illustrating a process for predicting a CSI of a UE to be predicted according to historical CSI of a plurality of reference position points corresponding to the prediction region, according to an exemplary embodiment of the present disclosure. FIG. 7B is a schematic diagram illustrating a CSI prediction process based on a first AI network according to an exemplary embodiment of the present application, as shown in FIG. 7B, the first AI network may include an input layer, a preprocessing layer, an embedding layer, and a recovery layer. FIG. 7C is a diagram illustrating a structure of an embedding layer according to an exemplary embodiment of the present application. FIG. 7D is a diagram illustrating a process of data processing by a first AI network according to an exemplary embodiment of the present application.

[0179] As shown in FIG. 7A, at step S710, at least one historical CSI reported within a preset time interval prior to the current time of each reference position point, reporting position information of the at least one historical CSI, the position information of the first UE is preprocessed through a preprocessing layer of the first AI network, to obtain channel information and position difference information corresponding to each reference position point, wherein the position difference information is a position difference determined according to the reporting position information of the at least one historical CSI of each reference position point and the position information of the first UE.

[0180] Specifically, for each reference position point, the input layer of the first AI network needs to obtain the latest historical CSI reported by this reference position point to the network node, the position where it is located at the time of reporting this CSI (i.e., the reporting position information). Meanwhile, it is also necessary to obtain the current position information of the UE to be predicted.

[0181]

[0182] At step S720, feature information is extracted from input information through the embedding layer of the first AI network, wherein the input information includes the channel information and the position difference information corresponding to each reference position point. This step may include: performing feature extraction on the input information according to the number of antennas, to obtain feature information for an antenna dimension. Selecting some of the CSI feedback channel information and the latest CSI as the AI network input makes the prediction process more effective and accurate.

[0183]

[0184] At step S730, the CSI of the first UE is predicted according to the feature information through a recovery layer of the first AI network.

[0185]

[0186] In the above example described with reference to FIGS. 7A and 7B, the information inputted to the first AI network refers only to the at least one historical CSI reported within a predetermined interval of time prior to the current time for each reference position point, the reporting position information of the at least one historical CSI, the position information of the UE to be predicted, however, the present application is not limited to this, and in another exemplary embodiment, the reporting time information of the at least one historical CSI and the time information of predicting CSI of the UE to be predicted may also be input to the first AI network, that is, the preprocessing layer of the first AI network also preprocesses reporting time information of the at least one historical CSI to obtain time difference information corresponding to each reference position point, this time difference information may be further included in the input information inputted into the embedding layer the first AI network, such that the embedding layer extracts the feature information from the channel information, the position difference information, and the time difference information corresponding to each reference position point, wherein the time difference information is a time difference determined according to the reporting time information of the at least one historical CSI and the time information of predicting CSI of the UE to be predicted of each reference position point. The exemplary embodiment is described in detail below with reference to FIGS. 8A and 8B.

[0187] FIG. 8A is a flowchart illustrating a process for predicting a CSI of the UE to be predicted according to a historical CSI of a plurality of reference position points corresponding to the prediction region, according to another exemplary embodiment of the present disclosure. FIG. 8B is a schematic diagram illustrating a process of CSI prediction based on a first AI network according to another exemplary embodiment of the present application. The first AI network may include an input layer, a preprocessing layer, an embedding layer, and a recovery layer, as shown in FIG. 8B.

[0188] As shown in FIG. 8A, at step S810, by preprocessing at least one historical CSI reported by each reference position point within a preset time interval prior to a current time, reporting position information and reporting time information of the at least one historical CSI, position information of the UE to be predicted, and time information of predicting CSI through the preprocessing layer of the first AI network, channel information, position difference information, and time difference information corresponding to each reference position point is obtained, wherein the position difference information corresponding to each reference position point is a position difference determined according to the reporting position information of the at least one historical CSI of each reference position point and the position information of the UE to be predicted, and the time difference information corresponding to each reference position point is a time difference determined according to the reporting time information of the at least one historical CSI of each reference position point and the time information of predicting CSI of the UE to be predicted.

[0189] Specifically, for each reference position point, the input layer of the first AI network needs to obtain the latest historical CSI reported by the reference position point to the network node, the position at which the CSI is reported (i.e., the reporting position information), and the time (i.e., the reporting time information). At the same time, it is also needed to obtain the current position information of the UE to be predicted and the time information of predicting CSI, wherein the time information of predicting CSI represents a time which the UE to be predicted requests the network node to predict the CSI for, for example, the current time, or a future time.

[0190]

[0191] At step S820, feature information is extracted from the input information through the embedding layer of the first AI network, wherein the input information includes the channel information, the position difference information and the time difference information corresponding to each reference position point. The step may include: performing feature extraction on the input information according to the number of antennas, to obtain feature information for an antenna dimension. Selecting some of the CSI feedback channel information and the latest CSI as the input of the AI network makes the prediction process more effective and accurate.

[0192]

[0193]

[0194] In the present application, considering factors such as the mobility of the UE, according to the prediction region update strategy, the network node may continuously update the prediction regions within its coverage region and types of respective position points (i.e., reference position points, predictable points, and unpredictable points), to reduce the overhead of the CSI reporting while ensuring the CSI prediction accuracy. Accordingly, the method illustrated in FIG. 1 may further include: updating the prediction regions within the coverage region of the network node.

[0195] In an exemplary embodiment of the present application, the updating the prediction regions within the coverage region of the network node may include: in response to a UE located outside the prediction regions within the coverage region of the network node moving into a first prediction region, classifying the UE as a reference position point, or a predictable position point of which CSI information is capable of being predicted.

[0196] Specifically, firstly, a second prediction region is generated according to all reference position points of the first prediction region and this UE, wherein the process of generating the second prediction region is similar to the process of generating the prediction region described above, and thus is not repeated herein. Then, whether a region range of the first prediction region is greater than or equal to a region range of the second prediction region is determined.

[0197] In response to the region range of the first prediction region being less than the region range of the second prediction region, this UE is classified as a reference position point of the second prediction region. As shown in FIG. 9A, after a UE outside a first prediction region PR1 is moved inside the first prediction region PR1, a region range of a newly generated second prediction region PR1’ is larger than a region range of the original first prediction region PR1, then, this UE (i.e., the current position of this UE) is classified as a reference position point of the second prediction region PR1'. Specifically, since this UE is an unpredictable point before it is classified as the reference position point, it also needs to report CSI periodically, so after this UE is classified as the reference position point, its CSI reporting configuration can be left unchanged, that is, it still reports CSI in accordance with the previous CSI reporting configuration, but the present application is not limited to this, and the network node may change the reporting period in this UE's CSI reporting configuration.

[0198] In response to the region range of the first prediction region being equal to the region range of the second prediction region, this UE is classified as a predictable position point, of which the CSI information is capable of being predicted, in the first prediction region, and this UE is configured not to report the CSI information to the network node. Specifically, since this UE is an unpredictable point before it is classified as the predictable point, it needs to periodically report CSI, and after this UE is classified as the predictable point, this UE will no longer need to reporting CSI, and therefore the network node needs to change this UE's CSI reporting configuration, i.e., set this UE to not report CSI.

[0199] In another exemplary embodiment of the present application, the updating the prediction regions within the coverage region of the network node may include: in response to expiration of a prediction region generation period, determining whether a degree of aggregation of a plurality of UEs located outside the prediction regions within the coverage region of the network node exceeds a fourth threshold value; in response to the degree of aggregation of the plurality of UEs exceeding the fourth threshold value, dividing the coverage region to obtain a new prediction region based on channel spatial consistency of the plurality of UEs, wherein the plurality of UEs are set as reference position points, unpredictable position points, and / or predictable position points of which the CSI information is capable of being predicted.

[0200] Specifically, the network node may set a prediction region generation period, and when the set prediction region generation period expires, as shown in FIG. 9A, it may attempt to generate a new prediction region using the unpredictable points, and in this process, the network node may determine whether the degree of aggregation of the unpredictable points (i.e., the unpredictable UEs) exceeds the fourth threshold value according to position information received from the respective unpredictable points, for example, determine whether the number of unpredictable points per square meter exceeds the fourth threshold value, and if the degree of aggregation of the unpredictable points exceeds the fourth threshold value, it may be considered that these unpredictable points aggregate to form a new hotspot region, and thus a new prediction region may be generated in accordance with the prediction region division process described above with reference to FIG. 3A, and accordingly, these original unpredictable points may be set as the reference position point and / or the predictable points of which the CSI is capable of being predicted, or maintain to be unpredictable points, wherein the CSI reporting configurations of the unpredictable points which are set as the reference position points may not be modified, and the unpredictable points which are set as the reference position points still report CSI in accordance with the original CSI reporting configuration and alternatively, their CSI reporting periods may be reconfigured. In addition, CSI reporting configurations of unpredictable points which are configured as the predictable points may be modified, that is, the unpredictable points which are configured as the predictable points are modified to not report CSI. Since the process of dividing the prediction region has been described in detail above with reference to FIG. 3A, it will not be repeated herein.

[0201] In yet another exemplary embodiment of the present application, the updating the prediction regions within the coverage region of the network node may include: for each prediction region within the coverage region of the network node, determining whether a CSI prediction accuracy of the each prediction region is lower than a fifth threshold value, according to CSI received from predictable position points within the each prediction region and CSI predicted for the predictable position points; in response to a CSI prediction accuracy of any prediction region is lower than the fifth threshold value, deleting this prediction region, and for reference position points and predictable position points of this prediction region, performing re-division for a prediction region.

[0202] Specifically, as shown in FIG. 9A, for an existing prediction region within the coverage region of the network node, the network node may periodically check the CSI prediction accuracy of predictable points, and to this end, the network node may configure predictable points within the prediction region to perform non-periodic CSI reporting, calculate the CSI prediction accuracy for the prediction region (for example, calculate evaluation metrics such as a Normalized Mean Square Error (NMSE) or a Cosine Similarity of correlation information between CSI received from the predictable points and CSI predicted for the predictable points), and then compare this calculated CSI prediction accuracy with the fifth threshold value, and if this calculated CSI prediction accuracy is lower than the fifth threshold value, then it indicates that the prediction region is no longer suitable for CSI prediction, so the prediction region is deleted, and for the reference position points and the predictable points of this prediction region, a re-division for a prediction region is performed, and their CSI reporting configurations are reconfigured. Since the division process of the prediction region is described above with reference to FIG. 3A, and therefore will not be further discussed herein.

[0203] In yet another exemplary embodiment of the present application, the updating the prediction regions within the coverage region of the network node may include: in response to a shift of a time period occurring, updating all prediction regions and all reference position points within the coverage region, as illustrated by S911 and S912 in FIG. 9B. Specifically, the updating all prediction regions and all reference position points within the coverage region may include: deleting all prediction regions and all reference position points in the coverage region; and re-determining prediction regions and reference position points in each hotspot region for a current time period according to the each hotspot region and candidate position points (i.e., typical position points) in the each hotspot region. That is, the prediction regions and reference position points in each hotspot region are re-determined for a current time period after the time period shifted.

[0204] Specifically, as shown in FIG. 9C, when the time period shifted (e.g., when it shifted from a time period of 6:00 to 9:00 on a weekday to 9:00 to 12:00 on a weekday), it is necessary to update all prediction regions and all reference position points (i.e., UEs for which CSI feedback is requested) in the coverage region of the network node. That is, all the original prediction regions and all the original reference position points in the coverage region of the network node are needed to be deleted, and the prediction regions and reference position points in the coverage region of the network node are re-determined for the current time period according to each hotspot region for the current time period and typical position points in the each hotspot region.

[0205] As mentioned above, describing the determination of hotspot regions for different time periods and typical position points corresponding to each hotspot region with reference to FIG. 3E, it is possible to determine whether a time period has shifted (which may also be referred to as whether a time state has shifted) by examining the time, and if the time period has shifted, the hotspot regions and the typical position points corresponding to the hotspot regions will change accordingly. In this case, since the existing prediction regions and the UEs (i.e., reference position points) for which CSI feedback is requested are determined according to the hotspot regions and the typical position points corresponding to the hotspot regions for the previous time period, all the prediction regions and all the reference position points in the coverage region of the network node need to be cleared, and prediction regions and reference position points are re-determined (i.e., updated) according to hotspot regions for the current time period and typical position points corresponding to the hotspot regions. By updating all the prediction regions in the coverage region of the network node according to the typical position points, a large amount of computational resources may be saved and high accuracy may be maintained.

[0206] In yet another exemplary embodiment of the present application, the updating the prediction regions within the coverage region of the network node may include: periodically performing the following operations: determining CSI prediction accuracy of each predictable position point within the coverage region; in response to a proportion of the predictable position points within the coverage region of which the CSI prediction accuracy does not satisfy a predetermined requirement, being greater than a sixth threshold value, updating all prediction regions and all reference position points within the coverage region, as shown in S921 and S922 in FIG. 9B; in response to the proportion of the predictable position points within the coverage region of which the CSI prediction accuracy does not satisfy the predetermined requirement, being less than or equal to the sixth threshold value, updating a prediction region corresponding to the predictable position points of which the CSI prediction accuracy does not satisfy the predetermined requirement, and reference position points of this prediction region, as shown in S931 and S932 in FIG. 9B.

[0207]

[0208]

[0209]

[0210] In yet another exemplary embodiment of the present application, the updating the prediction regions within the coverage region of the network node may include: in response to a new UE appears at an edge of a first prediction region within the coverage region of the network node, expanding the first prediction region and increasing reference position points of the expanded first prediction region (i.e., adding a UE for which the CSI feedback is requested), as shown in S941 and S942 shown in FIG. 9B.

[0211] Furthermore, in order to make the above-described technical solutions of the present application more clearly understood by those skilled in the art, the entire process is described below with reference to FIG. 9D.

[0212] FIG. 9D is a detailed flowchart illustrating a method performed by a network node according to an exemplary embodiment of the present application.

[0213] As shown in FIG. 9D, at step S901, historical CSI is collected. That is, the network node may collect CSI periodically reported by the UE for a longer time period, and may also collect position information "Position" of the UE at the time of reporting the CSI, reporting time information "Time" and / or a corresponding UE ID. In addition, the network node may compose the collected information into sample data (e.g. (Time, UE ID, CSI, Position)). Furthermore, the network node may divide these sample point data into different time periods, as described above with reference to step S310.

[0214] At step S902, the network node may identify at least one hotspot region within the coverage region of the network node for each time period according to the collected historical CSI, and determine a plurality of typical position points from among the respective position points in each hotspot region for each time period. As this has been detailed above with reference to FIG. 3H, it will not be repeated here.

[0215] At step S903, for the current time period, the network node may divide the coverage region to obtain at least one prediction region according to the result of the channel spatial consistency analysis, and, at the same time, determine reference position points for each prediction region. As this has been described in detail above with reference to step S320, it will not be repeated herein.

[0216] At step S904, CSI of a UE to be predicted in each prediction region is predicted according to the historical CSI of the plurality of reference position points corresponding to each prediction region. As this has been described in detail above with reference to step S120, it will not be repeated here.

[0217] At step S905, the prediction regions and the reference position points are updated according to shift of a time period and a prediction accuracy check, and at step S906, a decision is made to return to step S903 or S906 according to whether the CSI prediction accuracy is met, for example, since the process of updating the prediction regions and the reference position points within the coverage region of the network node has been described above with reference to FIG. 1, and is therefore will not be repeated.

[0218] FIG. 10 is a diagram illustrating an overall process of a method performed by a network node according to an exemplary embodiment of the present application. FIG. 11 is a diagram illustrating a process of a method performed by a network node according to another exemplary embodiment of the present application.

[0219] The method performed by the network node may be implemented by a channel prediction device as shown in FIG. 10, wherein the channel prediction device may include a wireless access network controller and a CSI prediction entity, and the CSI prediction entity may include a spatial consistency analysis module, a channel prediction module, a prediction region generation module, and a region update control module.

[0220] Specifically, the spatial consistency analysis module may perform a channel spatial consistency analysis between respective position points within the coverage region of the network node according to historical CSI received within the coverage region and CSI received from a UE during subsequent operation, wherein, by the channel spatial consistency analysis, channel spatial consistency coefficients between each position point and other position points within the coverage area are determined and channel spatial consistency distances of each position point within the coverage area in respective directions are determined. The operation performed by this spatial consistency analysis module may correspond to the database preparation operation as stage 1 in FIG. 11, this operation is mainly used to perform the channel spatial consistency analysis between respective position points within the coverage region of the network node and storage of channel spatial consistency information of each position point. Specifically, the first database S is used to store the channel spatial consistency information (i.e. channel spatial consistency coefficients, channel spatial consistency distances, and channel information (i.e. CSI) of respective points) of respective position points within the coverage area of the network node obtained based on long-term historical information, the second database N is used to store relevant information of active UEs within the coverage area of the network node. As this has been described in detail above with reference to the accompanying drawings, it will not be repeated herein.

[0221]

[0222] On this basis, the channel prediction module may proactively on demand or according to a CSI prediction request from the UE to be predicted, determine the prediction region corresponding to the UE to be predicted according to position information of the UE to be predicted, and then predict the CSI of the UE to be predicted according to the historical CSI of the plurality of reference position points corresponding to the prediction region. Specifically, the channel prediction module uses a prediction region selection strategy to perform filtering, according to the position information of the UE to be predicted, to obtain the prediction region at which the UE to be predicted is located, from among the at least one prediction region generated by the prediction region generation module, and then, selects a plurality of reference position points from among the reference position points of this prediction region according to a CSI prediction accuracy requirement and the position information of the UE to be predicted, and predicts the CSI of the UE to be predicted according to the historical CSI of the selected plurality of reference position points, wherein, in the prediction process, the channel prediction module may predict the CSI using the necessary parameters for prediction (e.g., all-channel information, reporting position information and / or reporting time information of the CSI) obtained from the radio access network controller, and the position information of the UE to be predicted and / or the time information of predicting CSI. The operations performed by the channel prediction module may correspond to the operations performed by a first AI network which is a stage 3 in FIG. 11, the first AI network may perform CSI prediction to obtain the predicted CSI of the UE to be predicted. As this has been described in detail above with reference to steps S110 and S120 of FIG. 1, it will not be repeated here.

[0223] In addition, the region update control module may update the prediction regions within the coverage region of the network node according to the prediction region update strategy, which is not repeated herein as it has been described in detail above.

[0224] By the above method performed by a network node proposed in the present application, non-reference position points in the prediction regions may be monitored according to a much larger period than that in the prior art, and they do not need to report the CSI for long time, and thus the overhead of CSI feedback may be significantly reduced; furthermore, the prediction region generation result may reflect the results of the actual distribution of UEs, and CSI prediction may be performed only in the region with a high density of UEs, which may minimize the computational overhead of channel prediction; furthermore, the above method for predicting CSI proposed in the present application may achieve high-precision CSI prediction, which may improve the system performance and effectively improve the system throughput.

[0225] FIG. 12 is a flowchart illustrating a method performed by a network node according to another exemplary embodiment of the present application.

[0226] As shown in FIG. 12, at step S1210, a plurality of reference position points from among a set of reference position points are determined according to position information of first UE.

[0227] In the present application, the set of reference position points is determined from among respective position points within a coverage region of the network node according to channel spatial consistency coefficients between the respective position points, CSI report frequencies and CSI report periods of the respective position points. This will be described in detail below with reference to FIG. 13.

[0228] FIG. 13 is a flowchart illustrating a process for determining a set of reference points according to an exemplary embodiment of the present application.

[0229] As shown in FIG. 13, at step S1310, channel spatial consistency coefficients between respective position points are determined according to historical CSI received at the respective position points. Since the process of determining the channel spatial consistency coefficients between the position points has been described above with reference to FIG. 3E, it will not be repeated here.

[0230] At step S1320, weights corresponding to CSI validity of the respective position points are determined according to the channel spatial consistency coefficients between the respective position points, the CSI reporting frequencies and the CSI reporting periods of the respective position points.

[0231]

[0232]

[0233]

[0234]

[0235]

[0236]

[0237] At step S1330, a plurality of position points are selected from the individual position points as the set of reference position points according to the weights corresponding to CSI validity of the individual position points.

[0238] Specifically, step S1330 may include: selecting, from among the respective position points, a plurality of position points having the largest sum of weights and the smallest total number of position points as the set of reference position points, wherein a channel spatial consistency coefficient between each of the respective position points other than the set of reference position points and reference position points of which the number is at least a second threshold value in the set of reference position points is greater than a third threshold value

[0239] Considering the CSI prediction accuracy, if CSI of one position point predicts the CSI using the historical CSI of a very small number of position points, the accuracy of the predicted CSI usually is poor, so as in the method described above with reference to FIG. 1, the second threshold is defined to limit the number of the reference position points for prediction of the CSI of the position points. In summary, the present application considers weight maximization when selecting the set of reference position points, i.e. positions with high CSI effectiveness perform reporting. Specifically, the set of reference position points may be determined by the following equations (11), (12) and (13):

[0240]

[0241]

[0242]

[0243]

[0244] The process of determining the set of reference position points S is exemplarily described below with reference to FIG. 15.

[0245] FIG. 15 is a diagram illustrating a process for determining a set of reference position points S according to an exemplary embodiment of the present application.

[0246]

[0247]

[0248]

[0249]

[0250]

[0251] At step S1220, CSI of the first UE is predicted according to historical CSI of the plurality of reference position points.

[0252] Specifically, the step of predicting the CSI of the first UE according to the historical CSI of the plurality of reference position points includes: predicting the CSI of the first UE through a first AI network, according to the position information of the first UE, the historical CSI of the plurality of reference position points, and reporting position information of the historical CSI. As the process of predicting the CSI of the first UE using the AI network has described in detail with reference to FIGS. 7A to 8B above, it is therefore not repeated herein.

[0253] FIG. 16A illustrates a diagram of a process of a method performed by a network node of FIG. 12, according to an exemplary embodiment of the present application.

[0254] As shown in FIG. 16A, a network node may perform a channel spatial consistency analysis on respective position points (i.e., positions at which respective UEs report CSI) within a coverage region according to channel information in a database, and then, determine a set of reference position points from among the respective position points according to the channel spatial consistency of the respective position points, CSI reporting frequencies at the respective position points, and CSI reporting periods (i.e., perform UE selection based on spatial consistency as shown in FIG. 16A, thereby obtaining the selected UE for CSI reporting), determines a plurality of reference position points from among the set of reference position points according to position information of a UE to be predicted, and then predicts CSI of the UE to be predicted according to historical CSI (i.e., CSI reported by the UEs) of the plurality of reference position points, e.g., by utilizing an AI network as shown in FIG. 16A for CSI prediction (i.e., obtaining the predicted CSI of the UE to be predicted through an AI-based inter-UE CSI prediction process).

[0255] FIG. 16B is a block diagram illustrating a network node according to an exemplary embodiment of the present application.

[0256] As shown in FIG. 16B, the network node 1600 includes a transceiver 1612 and a processor 1611, wherein the processor 1611 is coupled to the transceiver 1612 and configured to perform the method performed by the network node described above with reference to FIGS. 1 to 16A. Details regarding the operations of the above-described method performed by the network node may be found in the descriptions of FIGS. 1 to 16A, none of which will be repeated herein.

[0257] In embodiments of the present disclosure, there is also provided an electronic apparatus that includes at least one processor, and alternatively, further includes at least one transceiver and / or at least one memory coupled to the at least one processor, wherein, the at least one processor is configured to perform the steps of the method provided in any alternative embodiment of the present disclosure.

[0258] FIG. 17 illustrates a schematic diagram of a structure of an electronic apparatus applicable to an exemplary embodiment of the present application. As shown in FIG. 17, the electronic apparatus 4000 shown in FIG. 17 includes: a processor 4001 and a memory 4003. Wherein the processor 4001 and the memory 4003 are coupled, e.g., through a bus 4002. Alternatively, the electronic apparatus 4000 may further include a transceiver 4004 which may be used for data interaction between the electronic apparatus and other electronic apparatuses, such as transmitting of data and / or receiving of data. It should be noted that, each of the processor 4001, the memory 4003, and the transceiver 4004 is not limited to one in a practice application, and the structure of the electronic apparatus 4000 does not constitute a limitation of the embodiments of the present disclosure. Alternatively, the electronic apparatus may be the first network node, the second network node, or the third network node.

[0259] The processor 4001 may be a Central Processing Unit (CPU), general purpose processor, Digital Signal Processor (DSP), Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (FPGA) or other programmable logic device, transistor logic device, hardware part, or any combination thereof. It may implement or perform various exemplary logic boxes, modules, and circuits described in conjunction with the disclosed contents of the present disclosure. The processor 4001 may also be a combination that implements computing functions, such as a combination containing one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0260] The bus 4002 may include a pathway to transfer information between the above components. The bus 4002 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, and the like. The bus 4002 may be classed as an address bus, a data bus, a control bus, and the like. For ease of representation, only one bold line is shown in FIG. 17, but it does not mean that there is only one bus or one type of bus.

[0261] The memory 4003 may be a Read Only Memory (ROM) or other types of static storage apparatuses that can store static information and instructions, a Random Access Memory (RAM) or other types of dynamic storage apparatuses that can store information and instructions, may be an Electrically Erasable Programmable Read Only Memory (EEPROM), Compact Disc Read Only Memory (CD-ROM) or other optical disc storages, an optical disc storage (including a compressed disc, laser disc, optical disc, digital universal disc, Blu-ray disc, etc.), a disk storage medium, other magnetic storage apparatuses, or any other medium that can be used to carry or store computer programs and can be read by a computer, it is not limited herein.

[0262] The memory 4003 is used to store computer programs or executable instructions for performing the embodiments of the present disclosure, and is controlled for execution by the processor 4001. The processor 4001 is used to execute the computer programs or executable instructions stored in the memory 4003 to implement the steps shown in the preceding method of the embodiments.

[0263] According to embodiments in the disclosure, a method performed by a network node is provided. The method comprises determining a prediction region corresponding to a first UE, according to position information of the first UE; and predicting Channel State Information (CSI) of the first UE, according to historical CSI of a plurality of reference position points corresponding to the prediction region.

[0264] For example, the determining the prediction region corresponding to the first UE, according to the position information of the first UE, comprises determining, from among at least one prediction region within a coverage region of the network node, a prediction region to which a position of the first UE belongs, according to the position information of the first UE.

[0265] For example, the predicting the CSI of the first UE, according to the historical CSI of the plurality of reference position points corresponding to the prediction region, comprises selecting the plurality of reference position points from among reference position points of the prediction region, according to a CSI prediction accuracy requirement and / or processing capability of the network node.

[0266] For example, the selecting the plurality of reference position points from among the reference position points of the prediction region, according to the CSI prediction accuracy requirement and / or the processing capability of the network node, comprises in response to the number of the reference position points of the prediction region being less than or equal to a first threshold value, selecting all reference position points of the prediction region; and in response to the number of the reference position points of the prediction region being greater than the first threshold value, selecting the plurality of reference position points from among the reference positions of the prediction region, according to the position information of the first UE, the CSI prediction accuracy requirement, and / or the processing capacity of the network node.

[0267] For example, the selecting the plurality of reference position points from among the reference positions of the prediction region, according to the position information of the first UE, the CSI prediction accuracy requirement, and / or the processing capacity of the network node, comprises in response to a first number of reference position points determined according to the CSI prediction accuracy requirement and / or the processing capacity of the network node being less than the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose positions are closest to the first UE and of which the number being the first number of reference position points; and in response to the first number of first reference position point determined according to the CSI prediction accuracy requirement and / or the processing capacity of the network node being greater than or equal to the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose position are closest to the first UE and of which the number being the first threshold value.

[0268] For example, there is a UE in a connected state at each of the plurality of reference position points previously or currently.

[0269] For example, the number of the plurality of reference position points is greater than or equal to a second threshold value, wherein the second threshold value is determined according to a CSI prediction accuracy requirement, the first threshold value is determined according to a processing capacity of the network node, and the first threshold value is greater than the second threshold value.

[0270] For example, the predicting the CSI of the first UE, according to the historical CSI of the plurality of reference position points corresponding to the prediction region, comprises predicting the CSI of the first UE through a first Artificial Intelligence (AI) network, according to the position information of the first UE, the historical CSI of the plurality of reference position points, and reporting position information of the historical CSI.

[0271] For example, the predicting the CSI of the first UE through the first AI network, according to the position information of the first UE, the historical CSI of the plurality of reference position points, and the reporting position information of the historical CSI, comprises preprocessing at least one historical CSI reported at a preset time interval prior to the current time of each reference position point, reporting position information of the at least one historical CSI, the position information of the first UE, through a preprocessing layer of the first AI network, to obtain channel information and position difference information corresponding to each reference position point; extracting feature information from input information through a embedding layer of the first AI network, wherein the input information comprises the channel information and the position difference information corresponding to each reference position point; and predicting the CSI of the first UE according to the feature information through a recovery layer of the first AI network.

[0272] For example, the preprocessing layer further preprocesses reporting time information of the at least one historical CSI to obtain time difference information corresponding to each reference position point. The input information further comprises the time difference information.

[0273] For example, the extracting the feature information from the input information through the embedding layer of the first AI network comprises performing feature extraction on the input information according to the number of antennas, to obtain feature information for an antenna dimension.

[0274] For example, the position difference information is a position difference determined according to the reporting position information of the at least one historical CSI of each reference position point and the position information of the first UE, and / or the time difference information is a time difference determined according to the reporting time information of the at least one historical CSI of each reference position point and the time information of the predicted CSI of the first UE.

[0275] For example, the method comprises configuring a UE, that is not located at a reference position point, in the prediction region within a coverage region of the network node to not report CSI information to the network node.

[0276] For example, each prediction region is obtained by determining channel spatial consistency between respective position points within a coverage region of the network node, according to historical CSI received within the coverage region; and dividing the coverage region to obtain at least one prediction region according to the channel spatial consistency.

[0277] For example, the determining the channel spatial consistency between the respective position points within the coverage region of the network node, according to the historical CSI received within the coverage region, comprises identifying at least one hotspot region within the coverage region for each time period, according to the historical CSI received within the coverage region; and performing channel spatial consistency analysis between the respective position points in each hotspot region for each time period, according to the historical CSI received within each hotspot region for each time period.

[0278] For example, the performing the channel spatial consistency analysis between the respective position points in each hotspot region for each time period, according to the historical CSI received within each hotspot region for each time period, comprises determining channel spatial consistency coefficients between each position point and other position points within each hotspot region for each time period, according to the historical CSI received within each hotspot region for each time period; and determining a channel spatial consistency distance of the each position point in each direction, according to the channel spatial consistency coefficients between the each position point and the other position points, wherein the channel spatial consistency distance represents a maximum distance at which a channel spatial consistency coefficient is greater than a predetermined threshold. A portion of the position points in each hotspot region for each time period are selected as candidate position points for dividing of the prediction region.

[0279] For example, the dividing the coverage region to obtain the at least one prediction region according to the channel spatial consistency comprises performing the following operations for each hotspot region for a current time period in the coverage region: determining a plurality of candidate reference position points according to respective candidate position points in a current hotspot region; determining a channel consistency space of each candidate reference position point, according to channel spatial consistency distances of each candidate reference position point and a third threshold value; and determining a prediction region in the current hotspot region, according to the channel consistency space of each candidate reference position point.

[0280] According to embodiments in the disclosure a method performed by a network node is provided. The method comprises determining a plurality of reference position points from among a set of reference position points, according to position information of first UE; predicting CSI of the first UE, according to historical CSI of the plurality of reference position points. The set of reference position points is determined from among respective position points according to channel spatial consistency coefficients between the respective position points within a coverage region of the network node, CSI report frequencies and CSI report periods of the respective position points.

[0281] According to embodiments, a network node is provided. The network node comprises a transceiver for transmitting and receiving a signal; and a processor coupled to the transceiver and configured to determine a prediction region corresponding to a first UE, according to position information of the first UE; and predict Channel State Information (CSI) of the first UE, according to historical CSI of a plurality of reference position points corresponding to the prediction region.

[0282] According to embodiments, a computer-readable storage is provided. the computer-readable storage medium stores instructions, wherein the instructions, when being executed by at least one processor, cause the at least one processor to determine a prediction region corresponding to a first UE, according to position information of the first UE; and predict Channel State Information (CSI) of the first UE, according to historical CSI of a plurality of reference position points corresponding to the prediction region.

[0283] According to embodiments in the disclosure a method performed by a network node is provided. The method comprises determining a prediction region corresponding to a first UE, according to position information of the first UE; obtaining historical channel state information (CSI) from a plurality of reference position points corresponding to the prediction region; and determining CSI of the first UE, using a first artificial intelligence (AI) network, based on the obtained historical CSI.

[0284] For example, the obtaining the historical CSI from the plurality of the reference position points corresponding to the prediction region, comprises selecting the plurality of reference position points from among reference position points of the prediction region, according to at least one of a CSI prediction accuracy requirement or processing capability of the network node.

[0285] For example, the determining the CSI of the first UE, using the first AI network, based on the obtained historical CSI, comprises determining the CSI of the first UE, using the first AI network, based on the position information of the first UE, the obtained historical CSI, and reporting position information of the historical CSI.

[0286] For example, the determining the CSI of the first UE, using the first AI network, based on the position information of the first UE, the obtained historical CSI, and the reporting position information of the historical CSI, comprises preprocessing at least one historical CSI reported at a preset time interval prior to the current time of each reference position point, reporting position information of the at least one historical CSI, the position information of the first UE, through a preprocessing layer of the first AI network, to obtain channel information and position difference information corresponding to each reference position point; extracting feature information from input information through a embedding layer of the first AI network, wherein the input information comprises the channel information and the position difference information corresponding to each reference position point; and determining the CSI of the first UE according to the feature information through a recovery layer of the first AI network.

[0287] According to embodiments, a network node is provided. The network node comprises a transceiver for transmitting and receiving a signal; memory comprising one or more storage media configured to store instructions; and at least one processor comprising processing circuitry. The at least one processor is configured to determine a prediction region corresponding to a user equipment (UE), according to position information of the UE; obtain historical channel state information (CSI) from a plurality of reference position points corresponding to the prediction region; and determine CSI of the UE, using an artificial intelligence (AI) network, based on the obtained historical CSI.

[0288] a transceiver for transmitting and receiving a signal; and a processor coupled to the transceiver and configured to determine a prediction region corresponding to a first UE, according to position information of the first UE; and predict Channel State Information (CSI) of the first UE, according to historical CSI of a plurality of reference position points corresponding to the prediction region.

[0289] An embodiment of the present disclosure provides a computer readable storage medium storing computer programs or instructions, the computer programs or instructions, when being executed by at least one processor may perform or implement the steps in the preceding method of the embodiments and corresponding contents.

[0290] An embodiment of the present disclosure provides a computer program product including computer programs, the computer programs, when being executed by a processor, may implement the steps shown in the preceding method of the embodiments and corresponding contents.

[0291] For one or more embodiments, at least one of the components set forth in one or more of the preceding figures may be configured to perform one or more operations, techniques, processes, and / or methods as set forth herein. For example, a processor (e.g., baseband processor) as described herein in connection with one or more of the preceding figures may be configured to operate in accordance with one or more of the examples set forth herein. For another example, circuitry associated with a UE, base station, network element, etc. as described above in connection with one or more of the preceding figures may be configured to operate in accordance with one or more of the examples set forth herein.

[0292] Any of the above described embodiments may be combined with any other embodiment (or combination of embodiments), unless explicitly stated otherwise. The foregoing description of one or more implementations provides illustration and description, but is not intended to be exhaustive or to limit the scope of embodiments to the precise form disclosed. Modifications and variations are possible in light of the above teachings or may be acquired from practice of various embodiments.

[0293] The methods according to various embodiments described in the claims and / or the specification of the disclosure may be implemented in hardware, software, or a combination of hardware and software.

[0294] When implemented by software, a computer-readable storage medium storing one or more programs (software modules) may be provided. One or more programs stored in such a computer-readable storage medium (e.g., non-transitory storage medium) are configured for execution by one or more processors in an electronic device. The one or more programs include instructions that cause the electronic device to execute the methods according to embodiments described in the claims or specification of the disclosure.

[0295] Such a program (e.g., software module, software) may be stored in a random-access memory, a non-volatile memory including a flash memory, a read only memory (ROM), an electrically erasable programmable read only memory (EEPROM), a magnetic disc storage device, a compact disc-ROM (CD-ROM), digital versatile discs (DVDs), other types of optical storage devices, or magnetic cassettes. Alternatively, it may be stored in a memory configured with a combination of some or all of the above. In addition, respective constituent memories may be provided in a multiple number.

[0296] Further, the program may be stored in an attachable storage device that can be accessed via a communication network, such as e.g., Internet, Intranet, local area network (LAN), wide area network (WAN), or storage area network (SAN), or a communication network configured with a combination thereof. Such a storage device may access an apparatus performing an embodiment of the disclosure through an external port. Further, a separate storage device on the communication network may be accessed to an apparatus performing an embodiment of the disclosure.

[0297] In the above-described specific embodiments of the disclosure, a component included therein may be expressed in a singular or plural form according to a proposed specific embodiment. However, such a singular or plural expression may be selected appropriately for the presented context for the convenience of description, and the disclosure is not limited to the singular form or the plural elements. Therefore, either an element expressed in the plural form may be formed of a singular element, or an element expressed in the singular form may be formed of plural elements.

[0298] Meanwhile, specific embodiments have been described in the detailed description of the disclosure, but it goes without saying that various modifications are possible without departing from the scope of the disclosure.

[0299] The terms "first", "second", "third", "fourth", "1", "2" and the like (if exists) in the specification and claims of the present disclosure and the above drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequence. It should be understood that, data used as such may be interchanged in appropriate situations, so that the embodiments of the present disclosure described here may be implemented in an order other than the illustration or text description.

[0300] It should be understood that, although each operation step is indicated by an arrow in the flowcharts of the embodiments of the present disclosure, an implementation order of these steps is not limited to an order indicated by the arrows. Unless explicitly stated herein, in some implementation scenarios of the embodiments of the present disclosure, the implementation steps in the flowcharts may be executed in other orders according to requirements. In addition, some or all of the steps in each flowchart may include a plurality of sub steps or stages, based on an actual implementation scenario. Some or all of these sub steps or stages may be executed at the same time, and each sub step or stage in these sub steps or stages may also be executed at different times. In scenarios with different execution times, an execution order of these sub steps or stages may be flexibly configured according to a requirement, which is not limited by the embodiment of the present disclosure.

[0301] The above text and accompanying drawings are provided as examples only to assist readers in understanding the present disclosure. They are not intended and should not be interpreted as limiting the scope of the present disclosure in any way. Although certain embodiments and examples have been provided, based on the content disclosed herein, it is apparent to those skilled in the art that, changes can be made to the illustrated embodiments and examples without departing from the scope of the present disclosure, and other similar implementation methods based on the technical concepts of the present disclosure also belongs to a protection scope of the embodiments of the present disclosure.

Claims

1.A method performed by a network node, comprising:determining a prediction region corresponding to a user equipment (UE), according to position information of the UE;obtaining historical channel state information (CSI) from a plurality of reference position points corresponding to the prediction region; anddetermining CSI of the UE, using an artificial intelligence (AI) network, based on the obtained historical CSI.2.The method according to claim 1, wherein the determining the prediction region corresponding to the UE, according to the position information of the UE, comprises:determining, from among at least one prediction region within a coverage region of the network node, a prediction region to which a position of the UE belongs, according to the position information of the UE.3.The method according to claim 1, wherein the obtaining the historical CSI from the plurality of the reference position points corresponding to the prediction region, comprises:selecting the plurality of reference position points from among reference position points of the prediction region, according to at least one of a CSI prediction accuracy requirement or processing capability of the network node.4.The method according to claim 3, wherein the selecting the plurality of reference position points from among the reference position points of the prediction region, according to at least one of the CSI prediction accuracy requirement or the processing capability of the network node, comprises:in response to the number of the reference position points of the prediction region being less than or equal to a first threshold value, selecting all reference position points of the prediction region; andin response to the number of the reference position points of the prediction region being greater than the first threshold value, selecting the plurality of reference position points from among the reference positions of the prediction region, according to at least one of the position information of the UE, the CSI prediction accuracy requirement, or the processing capacity of the network node.5.The method according to claim 4, wherein the selecting the plurality of reference position points from among the reference positions of the prediction region, according to at least one of the position information of the UE, the CSI prediction accuracy requirement, or the processing capacity of the network node, comprises:in response to a number of reference position points determined according to at least one of the CSI prediction accuracy requirement or the processing capacity of the network node being less than the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose positions are closest to the UE and of which the number being the number of reference position points; andin response to the number of reference position point determined according to at least one of the CSI prediction accuracy requirement or the processing capacity of the network node being greater than or equal to the first threshold value, selecting, from among the reference position points of the prediction region, reference position points whose position are closest to the UE and of which the number being the first threshold value.6.The method according to claim 1, wherein there is a UE in a connected state at each of the plurality of reference position points previously or currently.7.The method according to claim 6, wherein the number of the plurality of reference position points is greater than or equal to a second threshold value, wherein the second threshold value is determined according to a CSI prediction accuracy requirement, the first threshold value is determined according to a processing capacity of the network node, and the first threshold value is greater than the second threshold value.8.The method according to claim 1, wherein the determining the CSI of the UE, using the AI network, based on the obtained historical CSI, comprises:determining the CSI of the UE, using the AI network, based on the position information of the UE, the obtained historical CSI, and reporting position information of the historical CSI.9.The method according to claim 8, wherein the determining the CSI of the UE, using the AI network, based on the position information of the UE, the obtained historical CSI, and the reporting position information of the historical CSI, comprises:preprocessing at least one historical CSI reported at a preset time interval prior to the current time of each reference position point, reporting position information of the at least one historical CSI, the position information of the UE, through a preprocessing layer of the AI network, to obtain channel information and position difference information corresponding to each reference position point;extracting feature information from input information through a embedding layer of the AI network, wherein the input information comprises the channel information and the position difference information corresponding to each reference position point; anddetermining the CSI of the UE according to the feature information through a recovery layer of the AI network.10.The method according to claim 9, wherein the preprocessing layer further preprocesses reporting time information of the at least one historical CSI to obtain time difference information corresponding to each reference position point,wherein the input information further comprises the time difference information.11.The method according to claim 9 or 10, wherein the extracting the feature information from the input information through the embedding layer of the AI network comprises:performing feature extraction on the input information according to the number of antennas, to obtain feature information for an antenna dimension.12.The method according to claim 9 or 10, wherein,the position difference information is a position difference determined according to the reporting position information of the at least one historical CSI of each reference position point and the position information of the UE, and / orthe time difference information is a time difference determined according to the reporting time information of the at least one historical CSI of each reference position point and the time information of the predicted CSI of the UE.13.The method according to claim 1, further comprising:configuring a UE, that is not located at a reference position point, in the prediction region within a coverage region of the network node to not report CSI information to the network node.14.The method according to claim 1, wherein each prediction region is obtained by:determining channel spatial consistency between respective position points within a coverage region of the network node, according to historical CSI received within the coverage region; anddividing the coverage region to obtain at least one prediction region according to the channel spatial consistency.15.A network node comprising:a transceiver for transmitting and receiving a signal;memory comprising one or more storage media configured to store instructions; andat least one processor comprising processing circuitry,wherein the at least one processor is configured to perform one of methods 1 to 14.

Images