Model monitoring method and communication apparatus

Abstract

The present application provides a model monitoring method and a communication apparatus. A network device configures an association relationship between beams, and when a portion of predicted beams (for example, a first beam) among a number K of predicted beams obtained by a terminal device falls outside a set of measured beams, K being an integer greater than or equal to 1, an associated second beam may be looked up for the first beam in the set of measured beams on the basis of the association relationship; further, a prediction error corresponding to the first beam is determined using a measurement value of the second beam, so that the problem of not being able to calculate a prediction error due to mismatch between the measured beam and the predicted beam for model monitoring is solved, and balancing between accuracy and real-time capability for prediction error calculation can be implemented. In addition, by means of a compensation value, the accuracy of prediction error calculation can be improved.

Description

Method and communication apparatus for model monitoring

[0001] The present application claims priority to the Chinese patent application No. 202411402554.4, filed on September 30, 2024, and entitled "Method and communication apparatus for model monitoring", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the technical field of wireless communication, and more particularly, to a method and communication apparatus for model monitoring. BACKGROUND

[0003] At present, artificial intelligence (AI) plays a great role in beam management, especially in reducing the overhead of beam sweeping. In the process of using AI model, it is usually necessary to monitor the performance of the model to adjust or optimize in time according to the demand. The process of model monitoring is usually that the network side issues monitoring resources, the terminal device measures the monitoring resources to obtain the measurement value of the monitoring resources, and then compares the prediction value of the prediction resources based on the model output with the measurement value of the monitoring resources to judge the prediction error of the model.

[0004] However, in the above method of monitoring the performance of the model, the monitoring resources issued by the network side may not match the prediction resources based on the model prediction output, which will result in that the prediction error of the model cannot be calculated. SUMMARY

[0005] The present application provides a method and communication apparatus for model monitoring, which can solve the problem that the prediction error cannot be calculated due to the lack of monitoring resources in the process of model monitoring.

[0006] In a first aspect, a method for model monitoring is provided, which is performed by a communication device or a module (e.g., a processor, a chip, a circuit, an AI entity, etc., which can also be a logical module, hardware and / or software, etc. capable of implementing all or part of the functions of the communication device) for a communication device, which can correspond to a terminal device in the method embodiment. The method comprises: obtaining measurement information, the measurement information comprising measurement values of beams in a first beam set; determining, based on association relationship information, a second beam associated with a first beam in K beams from the first beam set, the association relationship information indicating an association relationship between at least one beam in the first beam set and at least one beam in a second beam set, the first beam not being covered by the first beam set, the first beam set being a subset of a first full beam set, the second beam set comprising a complement of the first beam set relative to the first full beam set and at least one beam in the second beam set comprising the first beam, or the second beam set comprising part or all beams in a second full beam set and at least one beam in the second beam set corresponding to one or more first beams, the measurement value of the second beam and a predicted value of the first beam being used for determination of a prediction error corresponding to the first beam, the K being an integer greater than or equal to 1.

[0007] In the technical solution, by configuring the association relationship between beams, when part of the prediction beams (e.g., the first beam) in the K prediction beams obtained by the terminal device falls outside the set of measurement beams (corresponding to the first beam set), the associated second beam can be found for the first beam in the set of measurement beams based on the association relationship. The measurement value of the second beam is approximated as the measurement value of the first beam, and then the measurement value of the second beam can be used to determine the prediction error corresponding to the first beam, solving the problem that the prediction error cannot be calculated due to the lack of monitoring resources (e.g., the first beam is missing in the monitoring resources).

[0008] In addition, in one implementation, the first full beam set is a full beam set of narrow beams, the first beam set is a subset of the first full beam set, and the second beam set comprises a complement of the first beam set relative to the first full beam set, so the second beam set is also a set of narrow beams, and the first beam is included in the second beam set; or in another implementation, the second full beam set can be a set of wide beams, and the second beam set comprises part or all beams in the second full beam set, so any beam in the second beam set can correspond to multiple narrow beams, for example, one beam in the second beam set corresponds to one or more first beams.

[0009] Based on the above association relationship configured by the network side, when the monitoring resource set and the prediction resource cannot be completely matched in the model monitoring process, the second beam associated with the first beam in the monitoring resource set is determined based on the association relationship, and the prediction error corresponding to the first beam is calculated by using the measurement value of the second beam, which is equivalent to expanding the range of the monitoring resource set. Compared with the loss of real-time performance and the increase of air interface overhead caused by the secondary scanning when the monitoring resource set and the prediction resource cannot be completely matched, the present application can balance the accuracy and real-time performance of the prediction error.

[0010] With reference to the first aspect, in some implementations of the first aspect, the method further includes: obtaining prediction information, the prediction information indicating predicted values of the K beams.

[0011] With reference to the first aspect, in some implementations of the first aspect, the method further includes: sending the predicted value of the first beam and the measurement value of the second beam.

[0012] In this implementation, the calculation of the prediction error is performed by the network side. After the terminal device obtains the predicted value of the first beam and determines the second beam based on the association relationship and obtains the measurement value of the second beam, the terminal device provides the predicted value of the first beam and the measurement value of the second beam to the network side, so that the network side calculates the prediction error and obtains the model monitoring result.

[0013] With reference to the first aspect, in some implementations of the first aspect, the second full beam set is the first full beam set; or one beam in the second full beam set corresponds to multiple beams in the first full beam set.

[0014] In this implementation, the first full beam set can be a full beam set of narrow beams, and the second full beam set can also be a set of narrow beams. At this time, the second full beam set can be the same as the first full beam set, that is, a set of the same narrow beams. Alternatively, the second full beam set can be a full beam set of wide beams, and one wide beam in the second full beam set can cover (or correspond to) one or more narrow beams in the first full beam set.

[0015] With reference to the first aspect, in some implementations of the first aspect, the second beam set includes beams in a fourth beam set used for predicting the K beams.

[0016] In this implementation, the second beam set includes beams in the fourth beam set, and the beams in the fourth beam set can be used to determine the K predicted beams. For example, the terminal device measures the fourth beam set, uses the measurement result as a model input, and outputs top-K predicted beams through model inference.

[0017] With reference to the first aspect, in some implementations of the first aspect, the method further includes receiving the association information.

[0018] In this implementation, the terminal device obtains the association information from the network side, so that in the model monitoring process, when the monitoring resource and the prediction resource cannot be completely matched, the association relationship indicated by the association information is used to match an approximate beam for the prediction beam in the set of measurement beams, and then the measurement value of the approximate beam is obtained for the calculation of the prediction error of the prediction beam.

[0019] With reference to the first aspect, in some implementations of the first aspect, the association relationship between the first beam and the second beam includes that the first beam and the second beam are quasi co-located (QCL), and / or the distance between the positions of the first beam and the second beam in the first full beam set or the second full beam set is less than a first threshold.

[0020] In this implementation, when configuring the association relationship, if two beams are quasi co-located or the distance between the two beams (or the distance between the positions of the two beams) is less than a threshold, the network side configures the association of the two beams. The two beams with the association relationship are approximate beams of each other, so that when one of the two beams is a prediction beam and falls outside the set of measurement beams, the approximate beam thereof can be found from the set of measurement beams, and the measurement value of the approximate beam is used to calculate the prediction error of the prediction beam.

[0021] With reference to the first aspect, in some implementations of the first aspect, the method further includes receiving first information, the first information indicating a set of compensation values, the set of compensation values including at least one compensation value, the compensation value being related to a beam position or a distance between beam positions, the beam position indicating a beam in the first beam set or a beam in the first full beam set.

[0022] In one implementation, the beam position corresponds to a beam in the first beam set or a beam in the first full beam set.

[0023] The present application further considers that, for the problem of mismatch between the monitoring resource and the prediction beam, although the associated second beam for the first beam falling outside the monitoring resource can be found to calculate the prediction error through the association information, the measurement value of the approximate beam cannot truly reflect the actual measurement value of the prediction beam, and the measurement value of the approximate beam is generally smaller than the prediction value. A compensation mechanism for the measurement value is set. The terminal device compensates the measurement value of the approximate beam, and then calculates the prediction error using the compensated measurement value of the approximate beam, which can achieve more accurate error calculation.

[0024] With reference to the first aspect, in some implementations of the first aspect, the compensation values are related to beam positions, and each compensation value in the set of compensation values corresponds to one or more beam positions.

[0025] In this implementation, the compensation values are related to beam positions. Different beam positions correspond to different compensation values, so that the compensated measurement value of the approximated beam is closer to the actual measurement value of the predicted beam (in the case that the predicted beam is not covered by the set of measured beams), improving the accuracy of the prediction error calculation.

[0026] With reference to the first aspect, in some implementations of the first aspect, the measurement value of the second beam and the predicted value of the first beam are used for determination of the prediction error corresponding to the first beam, including that the compensated measurement value of the second beam and the predicted value of the first beam are used for determination of the prediction error corresponding to the first beam, wherein the compensated measurement value of the second beam is obtained by compensating the measurement value of the second beam based on the set of compensation values.

[0027] With reference to the first aspect, in some implementations of the first aspect, the compensation values are related to distances between beam positions, and each compensation value in the set of compensation values corresponds to a distance value or a distance range, the set of compensation values includes a first compensation value, the first compensation value corresponds to a first distance value or a first distance range, and the first compensation value is used for compensating the measurement value of the second beam to obtain the compensated measurement value corresponding to the second beam in the case that the distance between the position of the first beam and the position of the second beam is the first distance value or falls within the first distance range.

[0028] In this implementation, the compensation values are related to distances between beams (or distances between beam positions). When the distances between the two associated beams are different, the compensation values for compensating the measurement value of the approximated beam are different, so that the compensated measurement value of the approximated beam is closer to the actual measurement value of the predicted beam (in the case that the predicted beam is not covered by the set of measured beams), improving the accuracy of the prediction error calculation.

[0029] In a possible implementation of the first aspect, the method further includes: determining that a distance between the location of the first beam and the location of the second beam is the first distance value or falls into the first distance range; determining the compensated measurement value corresponding to the second beam based on the first distance value or the first distance range and the set of compensation values, the compensated measurement value corresponding to the second beam being a sum of the measurement value of the second beam and the first compensation value corresponding to the first distance value or the first distance range; and using the compensated measurement value corresponding to the second beam and the predicted value of the first beam for determination of the prediction error corresponding to the first beam.

[0030] In a possible implementation of the first aspect, K is an integer greater than 1, and the method further includes: determining whether a distance between the locations of the first beam and the second beam in the first full beam set or the second full beam set is greater than or equal to a second threshold value; and in a case where the distance between the locations of the first beam and the second beam in the first full beam set or the second full beam set is greater than or equal to the second threshold value, determining the compensated measurement value corresponding to the second beam.

[0031] In this implementation, in a case where the predicted K beams are top-K candidate beams, K is greater than 1, and in a case where a set condition is met, the measurement value of the approximate beam of the predicted beam (for example, the first beam) is compensated, otherwise it is not compensated. Different top-1 beams and other beams in top-K except top-1 are mainly considered to compensate for the characteristics of the actual measurement value of the top-K beam, so that when any one beam (for example, the first beam) in top-K is not covered by the set of measured beams, a value (that is, the measurement value of the second beam after compensation or without compensation) that is more approximate to the actual measurement value of the beam can be determined for error calculation, so as to improve the accuracy of error calculation.

[0032] Here, based on the characteristics of the actual measurement value of the top-K beam, it is determined whether to compensate the measurement value of the approximate beam, mainly considering that if the predicted beam is the top-1 beam, the measurement value of the approximate beam is probably less than the actual measurement value of the predicted beam if it is not the maximum value in the measurement set, so compensation is closer to the actual measurement value of the top-1 beam. However, if the predicted beam is other than the top-1 beam in top-K, the measurement value of the approximate beam itself also meets the characteristic that the predicted beam is not the top-1 beam, so based on a set condition, it is determined whether to compensate the measurement value of the approximate beam.

[0033] In some implementations of the first aspect, the method further includes: obtaining configuration information of the at least one third beam set and second information indicating a correspondence between the at least two prediction time units and the at least one third beam set; and obtaining the measurement information includes measuring, according to the correspondence, the third beam set corresponding to each of the at least two prediction time units to obtain the measurement information corresponding to each of the at least two prediction time units.

[0034] This implementation is mainly proposed to address the problem of large indication overhead and air interface resource overhead when the method of the first aspect or any of its implementations is applied to a time-domain prediction scenario. In this implementation, the network side configures a correspondence between at least two prediction time units and at least one third beam set (where the third beam set corresponds to a monitoring resource set), and issues the correspondence at one time when model monitoring starts. Subsequently, the terminal device determines the monitoring resource set corresponding to each prediction time unit based on the correspondence and performs measurement. This way of issuing monitoring resources can reduce the indication overhead and air interface resource overhead compared to separately indicating the monitoring resource set corresponding to each prediction time unit. Especially in the case where the monitoring resource sets corresponding to some prediction time units are the same, or multiple prediction time units correspond to the same monitoring resource set, the indication overhead and air interface resource overhead will be reduced to a greater extent.

[0035] In some implementations of the first aspect, the obtaining the configuration information of the at least one third beam set and the second information includes receiving radio resource control (RRC) signaling indicating the configuration information and the second information, and the method further includes receiving downlink control information (DCI) for triggering reporting of measurement information of the third beam set corresponding to one or more of the at least two prediction time units.

[0036] In this implementation, the network side configures the correspondence between the at least two prediction time units and the at least one third beam set through RRC signaling, and subsequently triggers reporting of measurement information corresponding to some or all prediction time units through one DCI. Compared to separately indicating the monitoring resource set corresponding to each prediction time through one DCI, the pre-configuration of the correspondence through RRC signaling and the triggering of reporting of measurement information of any one or more prediction times through one DCI in this application reduces the overhead caused by repeated indication.

[0037] In a second aspect, a method for model monitoring is provided to be performed by a communication apparatus or a module (e.g., a processor, a chip, a circuit, an AI entity, etc., which can also be a logical module, hardware and / or software, etc. capable of implementing all or part of the functions of the communication apparatus) or for a communication apparatus (which can correspond to the terminal device in the method embodiments). The method comprises: obtaining configuration information of at least one third beam set and second information, the second information indicating a correspondence between at least two prediction time units and the at least one third beam set; and according to the correspondence, measuring the at least one third beam set in the at least two prediction time units to obtain respective measurement information of the at least two prediction time units, respective prediction information of the at least two prediction time units, and respective measurement information of the at least two prediction time units for model monitoring.

[0038] In the technical solution, the problem of large overhead and air interface resource overhead in the model monitoring process when the network side issues monitoring resources is solved. In the implementation, the network side configures a correspondence between at least two prediction time units and at least one monitoring resource set (i.e., a third beam set), and issues the correspondence at one time when the model monitoring starts. Subsequently, the terminal device determines the corresponding monitoring resource set of each prediction time unit based on the correspondence and performs measurement. Compared with the mode of indicating the corresponding monitoring resource set of each prediction time unit by a DCI, the mode of issuing the monitoring resources can reduce the indication overhead and air interface resource overhead.

[0039] In combination with the second aspect, in some implementation manners of the second aspect, the obtaining the configuration information of the at least one third beam set and the second information comprises: receiving radio resource control (RRC) signaling, the RRC signaling indicating the configuration information and the second information; and the method further comprises: receiving a downlink control information (DCI), the DCI being used to trigger reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units.

[0040] In combination with the second aspect, in some implementation manners of the second aspect, a prediction time unit in the at least two prediction time units corresponds to one-to-one to a third beam set in the at least one third beam set; or a first prediction time unit and a second prediction time unit in the at least two prediction time units correspond to the same third beam set in the at least one third beam set.

[0041] The methods of the third aspect and the fourth aspect below are respectively methods of the network side corresponding to the first aspect and the second aspect, and the beneficial technical effects can be referred to the related description of the first aspect or the second aspect, which will not be described herein again.

[0042] In a third aspect, a method for model monitoring is provided, which can be performed by a communication apparatus or a module (e.g., a processor, a chip, a circuit, an AI entity, etc., which can also be a logical module, hardware and / or software, etc. capable of realizing all or part of the functions of the communication apparatus) for the communication apparatus, which can correspond to the network device in the method embodiments. The method comprises: configuring association relationship information, the association relationship information indicating an association relationship between at least one beam in a first beam set and at least one beam in a second beam set, the first beam set comprising a second beam associated with a first beam, the first beam not being covered by the first beam set, the first beam set being a subset of a first full beam set, the second beam set comprising a complement of the first beam set relative to the first full beam set and at least one beam in the second beam set comprising the first beam, or the second beam set comprising part or all beams in a second full beam set and one of the at least one beam in the second beam set corresponding to one or more first beams; and sending the association relationship information.

[0043] With reference to the third aspect, in some implementations of the third aspect, the method further comprises: receiving a predicted value of the first beam and a measured value of the second beam; and determining a prediction error corresponding to the first beam according to the predicted value of the first beam and the measured value of the second beam.

[0044] With reference to the third aspect, in some implementations of the third aspect, the second full beam set is the first full beam set; or one beam in the second full beam set corresponds to multiple beams in the first full beam set.

[0045] With reference to the third aspect, in some implementations of the third aspect, the second beam set comprises a beam in a fourth beam set for predicting the K beams.

[0046] With reference to the third aspect, in some implementations of the third aspect, the association relationship between the first beam and the second beam comprises: the first beam and the second beam being quasi co-located (QCL); and / or a distance between positions of the first beam and the second beam in the first full beam set or the second full beam set being less than a first threshold.

[0047] With reference to the third aspect, in some implementations of the third aspect, the method further comprises: sending first information, the first information indicating a set of compensation values, the set of compensation values comprising at least one compensation value, the compensation value being related to a beam position or a distance between beam positions, the beam position indicating a beam in the first beam set or a beam in the first full beam set.

[0048] With reference to the third aspect, in some implementations of the third aspect, the compensation value is related to a beam position, and each compensation value in the set of compensation values corresponds to one or more beam positions.

[0049] With reference to the third aspect, in some implementations of the third aspect, determining, according to the predicted value of the first beam and the measured value of the second beam, a prediction error corresponding to the first beam comprises: determining, based on the set of compensation values and the measured value of the second beam, a compensated measured value corresponding to the second beam; and determining, based on the compensated measured value corresponding to the second beam and the predicted value of the first beam, the prediction error corresponding to the first beam.

[0050] With reference to the third aspect, in some implementations of the third aspect, the compensation value is related to a distance between a beam position, and each compensation value in the set of compensation values corresponds to a distance value or a distance range, the set of compensation values comprises a first compensation value, the first compensation value corresponds to a first distance value or a first distance range, and the first compensation value is used to compensate the measured value of the second beam to obtain the compensated measured value corresponding to the second beam in a case where a distance between the position of the first beam and the position of the second beam is the first distance value or falls within the first distance range.

[0051] With reference to the third aspect, in some implementations of the third aspect, the compensated measured value corresponding to the second beam is a sum of the measured value of the second beam and the first compensation value corresponding to the first distance value or the first distance range.

[0052] With reference to the third aspect, in some implementations of the third aspect, the method further comprises: transmitting configuration information of at least one third beam set and second information, the second information indicating a correspondence between at least two prediction time units and the at least one third beam set.

[0053] With reference to the third aspect, in some implementations of the third aspect, the transmitting the configuration information of at least one third beam set and the second information comprises: transmitting radio resource control (RRC) signaling, the RRC signaling indicating the configuration information and the second information; and the method further comprises: transmitting downlink control information (DCI), the DCI being used to trigger reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units.

[0054] In a fourth aspect, a method for model monitoring is provided, which can be performed by a communication apparatus or a module (e.g., a processor, a chip, a circuit, an AI entity, etc., which can also be a logical module, hardware and / or software, etc. capable of realizing all or part of the functions of the communication apparatus) for the communication apparatus, which can correspond to the network device in the method embodiments. The method comprises: generating configuration information of at least one third beam set and second information, the second information indicating a correspondence between at least two prediction time units and the at least one third beam set; and transmitting the configuration information and the second information.

[0055] In combination with the fourth aspect, in some implementations of the fourth aspect, the transmitting the configuration information and the second information comprises: transmitting radio resource control (RRC) signaling, the RRC signaling indicating the configuration information and the second information; and the method further comprises: transmitting downlink control information (DCI), the DCI being used to trigger reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units.

[0056] In the fourth aspect or some implementations of the fourth aspect, a prediction time unit in the at least two prediction time units corresponds to one-to-one to a third beam set in the at least one third beam set; or a first prediction time unit and a second prediction time unit in the at least two prediction time units correspond to the same third beam set in the at least one third beam set.

[0057] In a fifth aspect, a communication apparatus is provided, which has the functions of implementing the method in the first aspect or the second aspect, or any possible implementation of these aspects. The functions can be implemented by hardware, or by software, or by hardware executing corresponding software. The hardware or software comprises one or more units corresponding to the above functions.

[0058] In a sixth aspect, a communication apparatus is provided, which has the functions of implementing the method in the third aspect or the fourth aspect, or any possible implementation of these aspects. The functions can be implemented by hardware, or by software, or by hardware executing corresponding software. The hardware or software comprises one or more units corresponding to the above functions.

[0059] In a seventh aspect, a communication apparatus is provided, which comprises at least one processor configured to cause the communication apparatus to perform the method in the first aspect or the second aspect, or any possible implementation of the aspects; or perform the method in the third aspect or the fourth aspect, or any possible implementation of the aspects. Optionally, the at least one processor is coupled with at least one memory for storing computer program or instructions, and the at least one processor is configured to invoke and run the computer program or instructions from the at least one memory, so as to cause the communication apparatus to perform the method in the first aspect or the second aspect, or any possible implementation of the aspects; or perform the method in the third aspect or the fourth aspect, or any possible implementation of the aspects. Optionally, the at least one processor can be included in the communication apparatus, or can be configured outside the communication apparatus. Optionally, the communication apparatus further comprises the at least one memory. Optionally, the communication apparatus further comprises a communication interface.

[0060] In an eighth aspect, a communication apparatus is provided, which comprises a communication interface and a circuit, the communication interface is configured to receive a signal to be processed, and transmit the signal to the circuit; the circuit is configured to process the signal, so as to perform the method in the first aspect or the second aspect, or any possible implementation of the aspects; or perform the method in the third aspect or the fourth aspect, or any possible implementation of the aspects. Optionally, the communication interface is further configured to output the signal processed by the circuit. As an example, the communication interface can be a transceiver, a hardware circuit, a bus, a module, a pin, or other types of communication interfaces. The signal comprises information and / or data. Optionally, the communication apparatus can be a chip.

[0061] In a ninth aspect, a computer readable storage medium is provided, which stores computer program codes or instructions, when the computer program codes or instructions are run on a computer, so as to cause the method in the first aspect or the second aspect, or any possible implementation of the aspects to be implemented; or the method in the third aspect or the fourth aspect, or any possible implementation of the aspects to be implemented.

[0062] In a tenth aspect, a computer program product is provided, which comprises computer program codes or instructions, when the computer program codes or instructions are run on a computer, so as to cause the method in the first aspect or the second aspect, or any possible implementation of the aspects to be implemented; or the method in the third aspect or the fourth aspect, or any possible implementation of the aspects to be implemented.

[0063] In an eleventh aspect, a wireless communication system is provided, comprising the communication device according to the fifth aspect and the communication device according to the sixth aspect. BRIEF DESCRIPTION OF DRAWINGS

[0064] FIG. 1 is a schematic diagram of a communication system suitable for embodiments of the application.

[0065] FIG. 2 is another schematic diagram of a communication system suitable for embodiments of the application.

[0066] FIG. 3 is a schematic diagram of a possible application framework in a communication system.

[0067] FIG. 4 is a schematic diagram of another possible application framework in a communication system.

[0068] FIG. 5 is a schematic diagram of a prediction procedure for BM case 1.

[0069] FIG. 6 is a schematic diagram of a prediction procedure for BM case 2.

[0070] FIG. 7 is an example of a scenario for a model monitoring method according to embodiments of the application.

[0071] FIG. 8 is an example of a scenario for a model monitoring method according to embodiments of the application.

[0072] FIG. 9 is a schematic flowchart of a method 200 for model monitoring according to embodiments of the application.

[0073] FIG. 10 is a schematic diagram of establishing an association relationship between beams in a method for model monitoring according to embodiments of the application.

[0074] FIG. 11 is an example of a method for model monitoring according to embodiments of the application.

[0075] FIG. 12 is another example of a method for model monitoring according to embodiments of the application.

[0076] FIG. 13 is a schematic diagram of a scheme for issuing monitoring resources in BM case 2.

[0077] FIG. 14 is a schematic diagram of a scheme for issuing monitoring resources in BM case 2 according to embodiments of the application.

[0078] FIG. 15 is a schematic flowchart of another method 600 for model monitoring according to embodiments of the application.

[0079] FIG. 16 is an example of a method for model monitoring according to embodiments of the application.

[0080] FIG. 17 is a schematic block diagram of a communication device 1000 according to embodiments of the application.

[0081] FIG. 18 is a schematic block diagram of another communication apparatus 1100 provided in the present application.

[0082] FIG. 19 is a schematic structural diagram of a chip provided in the present application.

[0083] FIG. 20 is a schematic diagram of a system architecture of a communication device provided in the present application. DETAILED DESCRIPTION

[0084] The technical solutions in the present application will be described below with reference to the accompanying drawings.

[0085] The technical solutions provided in the present application can be applied to various communication systems, such as a 5th generation (5G) or new radio (NR) system, a long term evolution (LTE) system, an LTE frequency division duplex (FDD) system, an LTE time division duplex (TDD) system, a wireless local area network (WLAN) system, a satellite communication system, etc. In addition, it can also be applied to device to device (D2D) communication, vehicle-to-everything (V2X) communication, machine to machine (M2M) communication, machine type communication (MTC), and an internet of things (IoT) communication system, a future communication system, or a converged system of multiple systems, etc.

[0086] A network element in a communication system can send a signal to another network element or receive a signal from another network element. The signal can include information, signaling, or data, etc. The network element can also be replaced by an entity, a network entity, a device, a communication device, a communication module, a node, a communication node, etc. The device is taken as an example for description in the embodiments of the present application. For example, the communication system can include at least one terminal device and at least one network device. The network device can send a downlink signal to the terminal device, and / or the terminal device can send an uplink signal to the network device.

[0087] FIG. 1 is a schematic diagram of a communication system applicable to embodiments of the present application. As shown in FIG. 1, the communication system 100 can include at least one network device, such as the network device 110 shown in FIG. 1, and at least one terminal device, such as the terminal device 120 and the terminal device 130 shown in FIG. 1. The network device 110 and the terminal devices (such as the terminal device 120 and the terminal device 130) can communicate with each other through wireless links. The communication devices in the communication system, for example, the network device 110 and the terminal device 120, can communicate with each other through multi-antenna technology.

[0088] Optionally, the communication system can further include at least one AI node.

[0089] FIG. 2 is another schematic diagram of a communication system applicable to embodiments of the present application. Compared with the communication system 100 shown in FIG. 1, the communication system 100 shown in FIG. 2 further includes an AI node 140. The AI node 140 is configured to perform AI-related operations, such as constructing a training data set, training an AI model, or inference of an AI model.

[0090] In an implementation manner, the network device 110 can send data related to AI model training to the AI node 140, and the AI node 140 can construct a training data set and train an AI model. As an example, the data related to AI model training can include data reported by a terminal device. The AI node 140 can send a result of an AI model-related operation to the network device 110 and forward the result to a terminal device through the network device 110. For example, the result of the AI model-related operation can include at least one of the following: a trained AI model, an evaluation result or a test result of the model, and the like. As an example, part of the trained AI model can be deployed on the network device 110, and another part can be deployed on the terminal device. Alternatively, the trained AI model can be deployed on the network device 110, or the trained AI model can be deployed on the terminal device.

[0091] It should be understood that FIG. 2 only illustrates an example in which the AI node 140 is directly connected to the network device 110, and in other scenarios, the AI node 140 can also be connected to a terminal device. Alternatively, the AI node 140 can be connected to both the network device 110 and the terminal device. Alternatively, the AI node 140 can also be connected to one or more of the network device 110 and the terminal device through a third-party network element. Embodiments of the present application do not limit the connection relationship between the AI network element and other network elements.

[0092] Optionally, in another implementation manner, the AI node 140 can also be arranged as a module in the network device and / or the terminal device, for example, in the network device 110 or the terminal device shown in FIG. 1.

[0093] It should be noted that FIG. 1 and FIG. 2 are merely schematic diagrams drawn for the purpose of understanding, and other devices can also be included in the communication system, for example, wireless relay devices and / or wireless backhaul devices, etc., which are not shown in FIG. 1 and FIG. 2. In addition, in actual applications, the communication system can include multiple network devices and / or multiple terminal devices. The number of network devices and terminal devices is not limited in the embodiments of the present application.

[0094] In the embodiments of the present application, the terminal device can also be referred to as a user equipment (UE), an access terminal, a user unit, a user station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a user terminal, a terminal, a wireless communication device, a user agent or a user apparatus. The terminal device can be a device providing voice / data, such as a handheld device with wireless connection function, a vehicle-mounted device, etc. At present, some examples of terminal devices are: a mobile phone, a tablet computer, a notebook computer, a palm computer, a mobile internet device (MID), a wearable device, a virtual reality (VR) device, an augmented reality (AR) device, a wireless terminal in industrial control, a wireless terminal in self driving, a wireless terminal in remote medical surgery, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, a cellular phone, a cordless phone, a session initiation protocol (SIP) phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), a handheld device with wireless communication function, a computing device or other processing device connected to a wireless modem, a wearable device, a terminal device in a 5G network or a terminal device in a future communication system, etc., and the embodiments of the present application are not limited thereto.

[0095] In the embodiments of the present application, the apparatus for implementing the function of the terminal device can be a terminal device, or an apparatus capable of supporting the terminal device to implement the corresponding function, for example, a processor, a circuit or a chip, etc., which can be configured in the terminal device or used in combination with the terminal device. In the embodiments of the present application, only the apparatus for implementing the function of the terminal device is taken as an example for description, and the scheme of the embodiments of the present application is not limited.

[0096] The network device in the embodiments of the present application can be a device for communicating with the terminal device, and can include a radio access network (RAN) node, for example, a base station, for accessing the terminal device to a wireless network. The base station can generally cover various names in the following or replace the following names: Node B (NodeB), evolved Node B (eNB), next generation Node B (gNB), relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), primary station, secondary station, motor slide retainer (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), radio unit (RU), etc. The base station can be a macro base station, a micro base station, a relay node, a donor node or the like, or a combination thereof. The base station can also refer to a communication module, modem or chip for being arranged in the foregoing devices or apparatuses. The base station can also be a mobile switching center and a device assuming the function of a base station in D2D, V2X, M2M communication, a device assuming the function of a base station in a future communication system, etc. The base station can support the same or different access technology networks. Optionally, the RAN node can also be a server, a wearable device, a vehicle or a vehicle-mounted device, etc. For example, the access network device in vehicle to everything (V2X) technology can be a road side unit (RSU). The embodiments of the present application do not limit the specific technology and specific device form of the network device.

[0097] A base station can be fixed, or mobile. For example, a helicopter or drone can be configured to act as a mobile base station, with one or more cells moving according to the location of the mobile base station. In other examples, a helicopter or drone can be configured to act as a device that communicates with another base station.

[0098] In some deployments, the network device mentioned in embodiments of the present application can be a device including a CU, or a DU, or a device including a CU and a DU, or a control plane CU node (central unit-control plane (CU-CP)) and a user plane CU node (central unit-user plane (CU-UP)) and a DU node. For example, the network device can include a gNB-CU-CP, a gNB-CU-UP and a gNB-DU.

[0099] In some deployments, a plurality of RAN nodes cooperate to assist a terminal to implement wireless access, and different RAN nodes respectively implement part of the functions of a base station. For example, the RAN node can be a CU, a DU, a CU-CP, a CU-UP, or an RU, etc. The CU and the DU can be separately arranged, or can also be included in the same network element, such as a BBU. The RU can be included in a radio frequency device or a radio frequency unit, such as an RRU, an AAU or an RRH.

[0100] In a possible design, a processing unit in a BBU for implementing baseband functions is referred to as a base band high (BBH) unit, and a processing unit in an RRU / AAU / RRH for implementing baseband functions is referred to as a base band low (BBL) unit.

[0101] In different systems, the CU (or CU-CP and CU-UP), DU or RU can also have different names, but those skilled in the art can understand their meanings. For example, in an open radio access network (open RAN, ORAN / O-RAN) system, the CU can also be referred to as an O-CU (open CU), the DU can also be referred to as an O-DU, the CU-CP can also be referred to as an O-CU-CP, the CU-UP can also be referred to as an O-CU-UP, and the RU can also be referred to as an O-RU. Any of the CU (or CU-CP, CU-UP), DU and RU in the present application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0102] In the embodiments of the present application, the apparatus for implementing the function of the network device can be a network device, or can be an apparatus capable of supporting the network device to implement the corresponding function, such as a processor, a circuit or a chip, etc. The apparatus can be configured in the network device or used in combination with the network device. In the embodiments of the present application, only the apparatus for implementing the function of the network device is taken as an example for description, and the scheme of the embodiments of the present application is not limited.

[0103] The network device and / or the terminal device can be deployed on land, including indoor or outdoor, handheld or vehicle-mounted; can also be deployed on water surface; and can also be deployed on aircraft, balloons and satellites in the air. The scenarios where the network device and the terminal device are located are not limited in the embodiments of the present application. In addition, the terminal device and the network device can be hardware devices, or software functions running on special hardware, software functions running on general hardware, such as virtualized functions instantiated on a platform (for example, a cloud platform), or entities including special or general hardware devices and software functions. The specific forms of the terminal device and the network device are not limited in the present application.

[0104] Optionally, the AI node can be deployed in one or more of the following positions in the communication system: an access network device, a terminal device, or a core network device, etc., or the AI node can also be deployed separately, for example, in a position other than any of the above devices, such as a host or a cloud server of an over the top (OTT) system. The AI node can communicate with other devices in the communication system, which can be one or more of the following: a network device, a terminal device, or a network element of a core network, etc.

[0105] The number of AI nodes is not limited in the present application. For example, when there are multiple AI nodes, the multiple AI nodes can be divided based on functions, such as different AI nodes responsible for different functions.

[0106] Optionally, the AI node can be a device independent of each other, or can be integrated in the same device to implement different functions, or can be a network element in a hardware device, or can be a software function running on special hardware, or can be a virtualized function instantiated on a platform (for example, a cloud platform), and the specific form of the AI node is not limited in the present application. The AI node can also be referred to as an AI network element or an AI module.

[0107] FIG. 3 is a schematic diagram of a possible application framework in a communication system. As shown in FIG. 3, network elements in the communication system are connected through interfaces (e.g., NG, Xn) or air interfaces. One or more AI modules are deployed in one or more of the network element nodes, such as a core network device, an access network node (RAN node), a terminal, or one or more devices in an operation administration and maintenance (OAM). The access network node can be a single RAN node or can include multiple RAN nodes, such as a CU and a DU. The CU and / or the DU can also be provided with one or more AI modules. Optionally, the CU can be further split into a CU-CP and a CU-UP. The CU-CP and / or the CU-UP can be provided with one or more AI models.

[0108] The AI module is used to implement a corresponding AI function. The AI modules deployed in different network elements can be the same or different. The AI module can implement different functions according to different parameter configurations of the model of the AI module. The model of the AI module can be configured based on one or more of the following parameters: a structural parameter (such as at least one of a number of neural network layers, a width of a neural network, a connection relationship between layers, a weight of a neuron, an activation function of a neuron, or a bias in the activation function), an input parameter (such as a type of the input parameter and / or a dimension of the input parameter), or an output parameter (such as a type of the output parameter and / or a dimension of the output parameter). The bias in the activation function can also be referred to as a bias of the neural network.

[0109] One AI module can have one or more models. One model can infer an output including one parameter or multiple parameters. The learning process, the training process, or the inference process of different models can be deployed in different nodes or devices, or can be deployed in the same node or device.

[0110] The network device can be a network device provided with one or more AI modules, such as one or more of the core network device, the access network node (RAN node), or the OAM shown in FIG. 3. The AI module can be a RAN intelligent controller (RIC), such as a near-real-time RIC or a non-real-time RIC, as shown in FIG. 4. For example, the near-real-time RIC is deployed in the RAN node (e.g., in the CU or the DU), and the non-real-time RIC is deployed in the OAM, in a cloud server, in the core network device, or in another network device.

[0111] Figure 4 is a schematic diagram of another possible application framework in a communication system. As shown in Figure 4, the communication system includes a RIC. For example, the RIC can be an AI module in the RAN node shown in Figure 1, for implementing AI related functions. The RIC includes a near-real time RIC (near-RT RIC) and a non-real time RIC (Non-RT RIC). The non-real time RIC mainly processes non-real time information, such as data that is not sensitive to latency, which can be in the order of seconds. The real time RIC mainly processes near-real time information, such as data that is relatively sensitive to latency, which can be in the order of tens of milliseconds.

[0112] The near-real time RIC is used for model training and inference. For example, for training an AI model, and using the AI model for inference. The near-real time RIC can obtain network side and / or terminal side information from the RAN node (e.g., CU, CU-CP, CU-UP, DU and / or RU) and / or the terminal. The information can be used as training data or data for inference.

[0113] Optionally, the near-real time RIC can deliver the inference result to the RAN node and / or the terminal.

[0114] Optionally, the inference result can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the near-real time RIC delivers the inference result to the DU, which sends it to the RU.

[0115] The non-real time RIC is also used for model training and inference. For example, for training an AI model, and using the AI model for inference. The non-real time RIC can obtain network side and / or terminal side information from the RAN node (e.g., CU, CU-CP, CU-UP, DU and / or RU) and / or the terminal. The information can be used as training data or data for inference, and the inference result can be delivered to the RAN node and / or the terminal.

[0116] Optionally, the inference result can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the near-real time RIC delivers the inference result to the DU, which sends it to the RU.

[0117] The near-real time RIC and the non-real time RIC can also be separately provided as a device. Optionally, the near-real time RIC and the non-real time RIC can also be part of other devices. For example, the near-real time RIC can be provided in the RAN node (e.g., CU, DU), while the non-real time RIC can be provided in the OAM, the cloud server, the core network device or other devices.

[0118] Optionally, the implementation of the AI model can be a hardware circuit, or software, or a combination of software and hardware, without limitation. Non-limiting examples of software include program code, programs, subprograms, instructions, instruction sets, codes, code segments, software modules, applications, or software applications, etc.

[0119] Some related technologies related to the technical solutions of the present application are introduced below.

[0120] 1、Beam

[0121] In the present application, a beam refers to a specific shape and direction of energy distribution formed by electromagnetic waves transmitted by an antenna in space. Therefore, K beams refer to K different shapes and / or different directions of energy distribution, where K is a positive integer. One implementation of a beam is to use a sensor array (such as an antenna array) to achieve the purpose of directional transmission and reception of signals. Specifically, by adjusting the phase and amplitude of each element (such as an antenna) in the sensor array, signals at certain angles are subjected to constructive interference (i.e., wave peaks add up, enhancing the signal), while signals at other angles are subjected to destructive interference (i.e., wave peaks and troughs cancel each other out, weakening the signal), thereby forming a beam with a specific shape and direction. A beam is a type of communication resource. A beam can be a wide beam, or a narrow beam, or other types of beams. The technology for forming a beam can be beamforming technology or other technical means. Beamforming technology can be digital beamforming technology, analog beamforming technology, or hybrid digital / analog beamforming technology. Different beams can be considered as different resources. The same information or different information can be transmitted through different beams. Alternatively, multiple beams with the same or similar communication characteristics can be considered as one beam. A beam can include one or more antenna ports for transmitting at least one of a data channel, a control channel, and a sounding signal. A beam, which can also be understood as a spatial resource, can refer to a transmission or reception precoding vector with energy transmission directivity. Energy transmission directivity can refer to receiving a signal processed by the precoding vector with good reception power in a certain spatial position, such as meeting the reception demodulation signal-to-noise ratio, etc. Energy transmission directivity can also refer to receiving the same signal from different spatial positions with different reception powers through the precoding vector. Different devices (e.g., network devices or terminal devices) can have different precoding vectors, and different devices can also have different precoding vectors, i.e., corresponding to different beams. A device can use one or more of multiple different precoding vectors at the same time, i.e., simultaneously form one beam or multiple beams, according to the configuration or capability of the device.

[0122] Optionally, beam can be replaced by signal, downlink beam, transmit beam, transmit beam, fine beam, narrow beam, spatial filter, spatial filter, spatial parameter, spatial transmission filter, port, etc.

[0123] In this application, the information used to indicate the beam used for transmission can be referred to as beam indication information. The beam indication information can be one or more of the following: beam number (or number, index, identity (ID), etc.), uplink signal resource number, downlink signal resource number, absolute index of beam, relative index of beam, logical index of beam, index of antenna port corresponding to beam, antenna port group index corresponding to beam, index of downlink signal corresponding to beam, time index of downlink synchronization signal block corresponding to beam, beam pair link (BPL) information, transmission parameter (Tx parameter) corresponding to beam, reception parameter (Rx parameter) corresponding to beam, transmission weight corresponding to beam, weight matrix corresponding to beam, weight vector corresponding to beam, reception weight corresponding to beam, index of transmission weight corresponding to beam, index of weight matrix corresponding to beam, index of weight vector corresponding to beam, index of reception weight corresponding to beam, reception codebook corresponding to beam, transmission codebook corresponding to beam, index of reception codebook corresponding to beam, index of transmission codebook corresponding to beam. The beam indication information can also be embodied as transmission configuration index (TCI) or TCI state. One TCI state includes one or more quasi co-location (QCL) information, and each QCL information includes the ID of one reference signal (such as synchronization signal block) and one QCL type. For example: the terminal device can need to determine the beam for receiving the physical downlink shared channel (PDSCH) according to the TCI state indicated by the network device (usually carried by the physical downlink control channel (PDCCH)). In this application, the index information of the beam is a typical example of the beam indication information, and the index of the beam can be replaced by other beam indication information that can indicate the beam.

[0124] In this application, the prediction information can be the prediction result directly output by the AI model, or the result obtained after processing the prediction result directly output by the AI model. One prediction information refers to the prediction result directly output by the AI model in one prediction process, or refers to the result obtained after processing the prediction result. Wherein, the AI model can be deployed on a terminal device, or deployed on an OTT device on the side of the terminal device, when the AI model is deployed on the OTT device, the terminal device can receive the prediction result output by the AI model from the OTT device.

[0125] In this application, the optimal beam can refer to the beam that maximizes the receiving or transmitting energy. For example, the receiving end uses different receiving beams to receive signals, and the optimal beam includes the beam with the largest measurement value of the corresponding signal among the multiple different receiving beams. For another example, the transmitting end uses different transmitting beams to transmit signals, and the optimal beam can include the beam with the largest measurement value of the corresponding signal measured by the receiving end when the signal transmitted by the multiple transmitting beams reaches the receiving end. Wherein, the measurement value of the signal is, for example, the measured reference signal received power (RSRP), reference signal received quality (RSRQ), signal to interference plus noise ratio (SINR) or other possible estimated values.

[0126] When the beam is identified by the identifier (ID) of the beam, for example, the ID of the beam can be a channel state information reference signal resource indicator (CSI-RS resource indicator, CRI), or the beam ID can be a bit corresponding to the beam in a bitmap. Exemplarily, the number of bits included in the bitmap is equal to the number of all beams associated with the network device in one beam management inference task.

[0127] The beam can be divided into wide beam and narrow beam. Wide beam refers to a beam with a relatively large radiation range of transmitting or receiving antennas when transmitting or receiving signals. Wide beam is usually used in application scenarios that require broadcasting signals to a larger area or wider coverage. It can provide a wider coverage area, but the signal strength is relatively weak. Narrow beam refers to a beam with a relatively small radiation range of transmitting or receiving antennas. Narrow beam is usually used in application scenarios that require focusing signals to a specific target or area. It can provide higher signal strength and higher directivity, but the coverage is relatively small.

[0128] 2、reference signal (RS)

[0129] The reference signal can also be referred to as a pilot signal or a pilot. The reference signal is a known signal, which can be a known signal provided by a sending end to a receiving end for channel estimation, channel sounding, data demodulation, etc. The reference signal mainly includes a synchronization signal block (SSB) (also referred to as an SS / PBCH block) and a channel state information-reference signal (CSI-RS). The SSB is a cell broadcast signal, which contains a primary synchronization signal (PSS), a secondary synchronization signal (SSS), a physical broadcast channel (PBCH), and a demodulation reference signal (DMRS). There are various reference signals, and with the continuous evolution of standards, the names of the above-mentioned reference signals can change, and more reference signals can also appear. No specific limitation is made.

[0130] The reference signal resource described in the embodiments of the present application has a one-to-one correspondence with a beam. The reference signal resource can include a spatial domain resource, and the spatial domain resource included in the reference signal resource is a beam corresponding to the reference signal resource. The reference signal resource can also not include a spatial domain resource. In the case where the reference signal resource corresponds to a beam, the reference signal transmitted on the reference signal resource is used to determine the signal quality of the beam corresponding to the reference signal resource.

[0131] To implement beam management, methods such as hierarchical scanning can be used to reduce the overhead of beam scanning, for example, first scanning a wide beam, and then scanning part of the narrow beam under the wide beam. The selection of the beam is mainly completed through the reference signal and the corresponding beam measurement. The SSB is periodically transmitted according to the cell configuration, and its function is not only used for beam management, but also used for initial access, time-frequency synchronization, etc. Simply, the SSB signal can be considered as a wide beam signal. Correspondingly, the CSI-RS signal is a UE-level signal, and the network side configures one or more groups of CSI-RS signals for the UE according to the actual situation. Similarly, the CSI-RS signal is not only used for beam management, but also used for channel quality measurement, etc. The CSI-RS signal can be understood as a narrow beam signal. The CSI includes at least one of the rank indication (RI) information, the channel quality indicator (CQI) information, the precoding matrix indicator (PMI), or the layer 1 reference signal receiver power (L1-RSRP).

[0132] The conventional beam management system performs two-step beam scanning in the service beam selection stage: the first stage scans the SSB (i.e., scans the wide beam), in which the UE measures and reports the reference signal received power (RSRP) of the SSB beam to the network side; the second stage is that the network side screens out the SSB beam with the largest RSRP according to the RSRP of the SSB beam reported by the UE, and configures a CSI-RS signal, which is used for the terminal device to scan the narrow beam under the SSB beam with the largest RSRP to determine the optimal beam.

[0133] In AI-based beam management, generally, the AI model takes the received power of the wide beam or the sparsely scanned narrow beam measured by the UE as input, and the AI model infers the K optimal narrow beams as the output, which can be called Top-K candidate beams. As an example, the AI model can output the RSRP value or the beam identifier (ID) of the K candidate beams. The network side performs scanning according to the Top-K candidate beams to finally determine the optimal beam, and K is a positive integer equal to or greater than 1. The AI model can be deployed on the UE side or the network side.

[0134] At present, the application of AI beam management (beam management, BM) mainly lies in two aspects, namely, spatial domain prediction and time domain prediction, which can be denoted as BM case 1 and BM case 2, respectively.

[0135] FIG. 5 is a schematic diagram of a prediction process of BM case 1. The input of the AI model is the beam information (usually the RSRP value) of a specific pattern scanned at a certain time, and the set of beam information is referred to as set B; after prediction by the AI model, the prediction information of each beam in the complete beam set (referred to as set A or set A) is output. The UE selects the top-K beams in set A according to the prediction information of each beam, and reports the top-K beams and their related information to the network side.

[0136] FIG. 6 is a schematic diagram of a prediction process of BM case 2. A sliding time window T1 is used to collect the input information of the AI model, such as the RSRP of set B from time (t-N+1) to time (t). The AI model processes the input information and outputs the prediction results of the beams in the future time window T2. The time window T2 is shown as time (t+1) to time (t+M) in FIG. 6. If the AI model is a regression model, the prediction result is the predicted RSRP of each beam in set A. The UE determines the optimal top-K beams by comparing the predicted RSRP. The UE reports the predicted RSRP and the beam ID of the top-K beams to the network side; if the AI model is a classification model, the prediction result is the probability of each beam in set A becoming the optimal beam, and the optimal top-K beams are determined by comparing the probabilities. The UE reports the beam ID of the top-K beams to the network side. In FIG. 6, top-K(t+1) represents the prediction result output by the AI model at time (t+1), and top-K(t+M) represents the prediction result output by the AI model at time (t+M).

[0137] In the following embodiments of the present application, the AI model is mainly taken as a regression model as an example to introduce the scheme of the present application, that is, the output information of model inference or model prediction generally refers to the predicted RSRP of the top-K beams.

[0138] In the use process of the AI model, it is usually necessary to monitor the performance of the AI model. The monitoring process of the AI model is realized by the network side issuing a monitoring resource, which can also be referred to as a monitoring reference signal (monitoring RS). The process of model monitoring will be described below in conjunction with FIG. 7.

[0139] FIG. 7 is an example of a scenario applicable to the model monitoring method of the embodiments of the present application. In FIG. 7, the monitoring resource set can be represented as set M. After the UE obtains the measurement results of set B based on the measurements of set B, the UE takes the measurement results of set B as the input of the AI model, and predicts the above-mentioned top-K candidate beams. The top-K candidate beams can also be referred to as predicted resources or predicted beams. Then, the UE can obtain the measurement values of the monitoring resources by measuring the monitoring resource set (such as set A in FIG. 7, i.e., the full beam set) issued by the network side. The predicted values and the measurement values of the top-K candidate beams are compared, and the prediction error of the AI model is calculated.

[0140] However, in the above-mentioned model monitoring process, the monitoring resource set may not cover the predicted resources, resulting in that the prediction error of the model cannot be calculated.

[0141] The present application analyzes the reasons for this situation, which are mainly due to the fact that, in order to save air interface overhead and ensure the real-time performance of the monitoring process, the monitoring resources are usually not set A, but a subset of set A. In other words, the above-mentioned set M is generally a subset of set A. As a result, there may be a situation in which some (or all) of the top-K predicted beams based on the AI model prediction output fall outside set M, as shown in FIG. 8.

[0142] FIG. 8 is an example of a scenario applicable to the model monitoring method of the embodiments of the present application. As shown in FIG. 8, in the full beam set, the monitoring resource set (i.e., set M) is a subset of set A. The top-K predicted beams determined based on the AI model prediction may not be completely covered by set M, for example, in the example of FIG. 8, among the top-K predicted beams (specifically, K = 4), one predicted resource is not covered by the monitoring resources. As can be seen, in the model monitoring process, there may be a situation in which the predicted resources are not covered by the monitoring resources. The predicted beams falling outside set M cannot be used to calculate the prediction error because there are no corresponding actual measurement values. If a second beam scanning is considered, it will result in a loss of real-time performance and an increase in air interface resource overhead.

[0143] Therefore, the present application provides a method for model monitoring, which aims to solve the above-mentioned problem of the prediction error being unable to be calculated in the model monitoring stage.

[0144] The technical solutions provided by the present application are described in detail below.

[0145] FIG. 9 is a schematic flowchart of a method 200 for model monitoring provided by the present application. The method 200 involves a network device and a terminal device, and the method 200 can be implemented by the network device and the terminal device each performing corresponding steps. Alternatively, the communication devices (e.g., the network device or the terminal device) involved in the method 200 can also be replaced by apparatuses for these communication devices, for example, the network device can be replaced by a first apparatus, which can be a chip, a processor, a circuit applied to the network device, or an AI entity serving the network device, etc. The AI entity can be deployed on the network device or outside the network device. As an example, the AI entity can be an over the top (OTT) server or a cloud server. The terminal device can also be replaced by a corresponding apparatus, which will not be described herein. In the following embodiments, the network device and the terminal device are taken as examples for description.

[0146] 210. The network device configures the association relationship information.

[0147] The association relationship information indicates an association relationship between at least one beam in the first beam set and a beam in the second beam set, and the first beam is not covered by the first beam set. The first beam set is a subset of the first full beam set. The second beam set includes a complement of the first beam set relative to the first full beam set and at least one beam in the second beam set includes the first beam; or the second beam set includes part or all of the beams in the second full beam set and one of the at least one beam in the second beam set corresponds to one or more first beams.

[0148] Alternatively, the first beam not being covered by the first beam set can also mean that the first beam is different from any beam in the first beam set, or the first beam does not overlap with any beam in the first beam set, etc.

[0149] In configuring the association between the beams in the first set of beams and the beams in the second set of beams, each beam in the first set of beams can be configured with an associated beam in the second set of beams, or some beams in the first set of beams can be configured with an associated beam in the second set of beams respectively. In other words, some beams in the first set of beams can have no associated beam in the second set of beams, or each beam in the first set of beams can have an associated beam in the second set of beams, without limitation. The configuration of the association can depend on the proximity between the beams in the full set of beams (the first full set of beams or the second full set of beams), which can be measured based on a quasi co-location (QCL) or a distance between the beams, for example. For example, if beam 1 in the first set of beams and beam 2 in the second set of beams satisfy a QCL relationship, beam 1 and beam 2 can be configured to be associated; or if the distance between beam 1 in the first set of beams and beam 2 in the second set of beams is less than a set threshold, beam 1 and beam 2 can be configured to be associated. For another example, if beam 1 in the first set of beams has no QCL relationship with any beam in the second set of beams, and the distance between beam 1 and any beam in the second set of beams is greater than a set threshold, beam 1 has no associated beam in the second set of beams.

[0150] Optionally, the first beam is associated with the second beam can include one or more of the following:

[0151] The first beam and the second beam are QCL; and / or

[0152] The distance between the first beam and the second beam in the first full set of beams or the second full set of beams is less than a first threshold.

[0153] Quasi co-location generally refers to two or more antenna ports, although located in different physical positions, but the large-scale channel properties experienced by the transmission signals of the two or more antenna ports can be inferred from each other or considered the same. In other words, if the channel large-scale properties experienced by the signal transmitted by one antenna port can be determined by the channel of the signal transmitted by another antenna port, the two antenna ports are considered to be quasi co-located. As an example, the large-scale channel properties can include Doppler shift, Doppler spread, average delay, delay spread, and spatial Rx parameters, etc.

[0154] In the embodiments of the present application, the quasi co-location relationship of the antenna ports is applied to the beams.

[0155] As an example, if there is a quasi co-location relationship between two or more antenna ports, there is a quasi co-location relationship between the beams transmitted by the two or more antenna ports.

[0156] In addition, in the embodiments of the present application, the beam position can refer to the position of the beam in the full beam set. As an example, the position of beam 1 can be the position of the quantized number of beam 1 in the beam number set; or, as another example, based on the quantized number of beam 1, a two-dimensional matrix (for example, 64 beams can be converted into an 8x8 two-dimensional matrix) is obtained, and the position of beam 1 can be the geometric position of the quantized number of beam 1 in the two-dimensional matrix.

[0157] Based on the above description of the beam position, the distance between the positions of two beams in the full beam set can refer to the difference between the quantized numbers corresponding to the two beams; or, in the latter example of the above description of the beam position (i.e., the example involving a two-dimensional matrix), the distance between the positions of the two beams can be the Euclidean distance or the block distance between the geometric positions of the two beams. The distance between the positions of the two beams can also be considered as the distance between the two beams.

[0158] The first beam set is a subset of the first full beam set. In an implementation, the first full beam set can be the full beam set of narrow beams. The second full beam set can be the set of wide beams or the set of narrow beams.

[0159] In an implementation, the second beam set includes the complement of the first beam set with respect to the first full beam set, and at least one beam in the second beam set includes a first beam. In this implementation, the second beam set is the set of narrow beams.

[0160] In another implementation, the second beam set includes part or all of the beams of the second full beam set, and one of the at least one beam in the second beam set corresponds to one or more first beams. In this implementation, the second beam set can be the set of wide beams. The first beam of the K beams can be covered by a certain wide beam in the second beam set. At this time, the first beam corresponds to the wide beam in the second beam set. Based on the association relationship, the wide beam can correspond to multiple narrow beams in the first beam set, and at this time, as an example, the predicted error corresponding to the first beam can be calculated by using the measurement value closest to the predicted value of the first beam among the measurement values corresponding to the multiple narrow beams, wherein the narrow beam corresponding to the measurement value closest to the predicted value of the first beam corresponds to the second beam in the above embodiment.

[0161] Optionally, when describing the position of a beam in a full beam set, the full beam set can be the first full beam set or the second full beam set, without limitation. In other words, the coordinate representation when calculating the position of a beam or the distance between beams can be different when taking different full beam sets as reference. For example, the first full beam set includes 64 narrow beams, beam 1 corresponds to beam number 1 in the first full beam set, and beam 2 corresponds to beam number 5 in the first full beam set. The distance between beam 1 and beam 2 is calculated based on beam number 1 and beam number 5, and the distance between beam 1 and beam 2 is 4. For another example, the second full beam set includes 32 wide beams, beam 1 corresponds to the wide beam with beam number 2 in the second full beam set (or beam 1 is covered by the wide beam with beam number 2), and beam 2 corresponds to the wide beam with beam number 3 in the second full beam set. At this time, the distance between beam 1 and beam 2 is calculated based on beam number 2 and beam number 3 in the second full beam set, and thus the distance between beam 1 and beam 2 is 1.

[0162] In the embodiments of the present application, the first full beam set can be set A described above, and the beam set corresponding to the input information of the AI model is referred to as set B, or in other words, the measurement result obtained by the terminal device by measuring set B is used as the input information for AI model prediction.

[0163] In one example, the second beam set can be the complement of the first beam set with respect to the first full beam set.

[0164] As an example, set M described above can be a set of narrow beams, and each beam in set M can correspond to a quantized number of narrow beams; set A can be a full beam set of narrow beams.

[0165] In the embodiments of the present application, since the first beam set is actually a set of measurement beams, each beam in the first beam set has a corresponding measurement value. By determining the second beam associated with the first beam from the first beam set, the measurement value of the second beam can be obtained. Each beam in the first beam set can be referred to as a measurement beam or a monitoring beam, and the first beam set can be referred to as a set of measurement beams or a monitoring resource set.

[0166] FIG. 10 is a schematic diagram of establishing an association relationship between beams in a method for model monitoring provided by the present application. Based on the above description of the association relationship, FIG. 10 shows a schematic process of determining a second beam associated with a first beam in a first beam set based on the association relationship information. The first beam is a beam falling outside the first beam set. The first beam set is a subset of the first full beam set. The second beam set can be referred to the above description and will not be repeated. In an example, the second beam set can be the complement of the first beam set with respect to the first full beam set. That is, the union of the first beam set and the second beam set is the first full beam set. In this case, the first beam belongs to the second beam set.

[0167] 220. The network device sends the association relationship information.

[0168] Correspondingly, the terminal device obtains the association relationship information from the network device.

[0169] As an example, the association relationship can be sent by the network device to the terminal device after the start of the model monitoring stage. In the subsequent model monitoring process, when the association relationship information needs to be updated, the network device indicates the updated association relationship to the terminal device.

[0170] 230. The terminal device obtains measurement information, the measurement information including measurement values of beams in the first beam set.

[0171] The terminal device measures the beams in the first beam set issued by the network device to obtain measurement information. The measurement information includes the measurement value, such as RSRP, of each beam in the first beam set. The first beam set can correspond to set M described above. The first beam set can also be understood as a monitoring resource set for model monitoring.

[0172] 240. The terminal device determines a second beam associated with a first beam in the K beams from the first beam set based on the association relationship information.

[0173] The present application mainly proposes a solution for the case where the K predicted beams do not completely match the monitoring resource set. Therefore, when the first beam in the K predicted beams is not covered by the monitoring resource set, or in other words, the first beam falls outside the first beam set, the terminal device determines a second beam associated with the first beam from the first beam set based on the association relationship information, thereby obtaining the measurement value of the second beam. Using the measurement value of the second beam, combined with the predicted value of the first beam, the predicted error corresponding to the first beam can be calculated.

[0174] It can be seen that in the technical solutions provided in the present application, when one or more beams, such as the first beam, in the K predicted beams obtained by the terminal device are not covered by the set of measured beams (i.e., the first beam set), the terminal device determines the second beam associated with the first beam from the first beam set based on the association relationship information. The first beam is associated with the second beam, which can also mean that the first beam and the second beam are approximate, and therefore the measurement value of the second beam will be approximately used as the measurement value of the first beam. Thus, even if the first beam is not covered by the first beam set, the second beam approximate to the first beam determined by the association relationship can be used to calculate the prediction error corresponding to the first beam by using the measurement value of the second beam. Specifically, by comparing the predicted value of the first beam and the measurement value of the second beam, the prediction error corresponding to the first beam can be determined.

[0175] The prediction error can measure the prediction performance of the AI model, and is used for performance monitoring of the model. The difference between the predicted value of each of the K predicted beams and the measurement value thereof reflects the prediction error of the model. In the embodiments of the present application, the prediction error corresponding to the first beam is the prediction error calculated by using the predicted value of the first beam and the measurement value of the second beam associated with the first beam. For example, the AI model is a regression model, the predicted value of the RSRP of the first beam is RSRP = -70 dBm, and the measurement value of the RSRP of the second beam is RSRP = -80 dBm. The prediction error of the first beam is 10 dBm.

[0176] In the present application, the network device configures the association relationship between the beams, so that when part of the K predicted beams (such as the first beam) obtained by the terminal device falls outside the set of measured beams, the second beam associated with the first beam can be found in the set of measured beams based on the association relationship, and the measurement value of the second beam is used to determine the prediction error corresponding to the first beam, thereby solving the problem that the prediction error cannot be calculated due to the lack of measured beams for model monitoring.

[0177] Optionally, before step 240, the method 200 comprises step 250.

[0178] 250. The terminal device obtains prediction information, and the prediction information indicates the predicted values of the K beams.

[0179] As described above, the calculation of the prediction error of the model is achieved by the respective prediction value of the K beams and the respective corresponding measurement value of the K beams. The terminal device can obtain the measurement information of set B by measuring set B issued by the network device. The measurement information is taken as the input information of the model prediction, and the prediction information is obtained by model inference (i.e., model prediction). Specifically, the prediction information can be the K prediction values output by the AI model, and the K prediction values correspond to the K beams one by one, and K is an integer greater than or equal to 1. The detailed process of obtaining the prediction information based on the AI model inference of the terminal device can refer to the description in FIG. 5 or FIG. 6 above, and will not be described again.

[0180] In one example, if the AI model is a classification model, the output information is the probability that each beam in set A becomes the optimal beam; if the AI model is a regression model, the output information is the predicted RSRP of each beam in set A. According to the output information of the AI model, the terminal device determines the top-K beams through comparison, and the prediction values of the top-K beams are the prediction information.

[0181] The first beam described above is one of the K beams. Therefore, it can also be said that the first beam is a predicted beam. After determining the second beam associated with the first beam from the first beam set, the measurement value of the second beam is obtained, and based on the prediction value of the first beam and the measurement value of the second beam, the prediction error corresponding to the first beam can be calculated.

[0182] Optionally, in step 250, taking the example that the AI model is deployed on the terminal device, and the model prediction is also taken as an example of the terminal device. In another possible implementation, the model prediction can be performed by other devices other than the terminal device, such as OTT. In this implementation, after the terminal device measures set B and obtains the measurement value of the beam in set B, the terminal device sends the measurement value to the OTT, the OTT performs model prediction based on the received measurement value of set B to obtain the prediction information, and then provides the prediction information to the terminal device.

[0183] On the basis of the above scheme, the present application further considers that, for the problems of monitoring resources and predicting beam mismatch, although the associated relationship information can be used to find the associated second beam for the first beam falling outside the monitoring resources to calculate the prediction error, the second beam determined by the associated relationship may not be the optimal beam in the set of measurement beams, and the measurement value of the second beam is generally less than the predicted value of the first beam, for example, the measurement RSRP of the second beam is less than the predicted RSRP of the first beam. For this case, the present application proposes to configure a set of compensation values for the terminal device by the network side, and based on the set of compensation values, the terminal device compensates the measurement value of the second beam. Using the compensated measurement value of the second beam to calculate the prediction error corresponding to the first beam can realize more accurate prediction error calculation.

[0184] The related implementation of the set of compensation values will be described in detail below.

[0185] Optionally, the method 200 can further include steps 260-270.

[0186] 260. The network device sends first information, and the first information indicates the set of compensation values.

[0187] The terminal device receives the first information from the network device.

[0188] The set of compensation values includes at least one compensation value. The position of the compensation value is related to the position of the beam (i.e. the position of the beam) or the distance between the positions of the beams. The position of the beam indicates the beam in the first beam set or the beam in the first full beam set.

[0189] As an example, the position of the beam indicating the beam in the first beam set can mean that one compensation value is configured for each beam in the first beam set. Since the positions of the beams in the first beam set are different from each other, one compensation value is configured for each beam, which is equivalent to that each compensation value corresponds to a beam position. As another example, the position of the beam indicating the beam in the first full beam set can mean that one compensation value is configured for each beam in the first full beam set. In this example, no matter how the K predicted beams determined by the terminal device and the measurement beams in the first beam set change, the compensation value corresponding to the second beam can be determined after determining the second beam associated with the first beam from the first beam set, so as to compensate the measurement value of the second beam.

[0190] In one implementation, the compensation values in the set of compensation values are related to beam positions. In this implementation, each compensation value in the set of compensation values corresponds to one or more beam positions. When one compensation value corresponds to multiple beam positions, it indicates that the compensation values for the measurements of the beams corresponding to the multiple beam positions are the same. Here, "multiple" means two or more. Since each beam position indicates one beam, each compensation value corresponds to one beam position, i.e., each compensation value corresponds to one beam. Each compensation value is used to compensate the measurement of the corresponding beam.

[0191] In this implementation, when calculating the prediction error corresponding to the first beam, the measurement of the second beam is first compensated using the compensation value corresponding to the second beam to obtain the compensated measurement of the second beam. The prediction error corresponding to the first beam is determined based on the compensated measurement of the second beam and the prediction value of the first beam.

[0192] In another implementation, the compensation values in the set of compensation values are related to the distances between the beam positions. In this implementation, each compensation value in the set of compensation values corresponds to one distance value or distance range. Taking an example in which the set of compensation values includes a first compensation value, the first compensation value corresponds to a first distance value or a first distance range. The first compensation value is used to compensate the measurement of the second beam to obtain the compensated measurement of the second beam when the distance between the position of the first beam and the position of the second beam is the first distance value or falls within the first distance range.

[0193] In this implementation, when calculating the prediction error corresponding to the first beam, the distance between the position of the first beam and the position of the second beam associated with the first beam is calculated first. Here, the distance between the position of the first beam and the position of the second beam is taken as a first distance value or falls within a first distance range, and a first compensation value corresponding to the first distance value or the first distance range is determined from the compensation value set. The measurement value of the second beam is compensated by using the first compensation value to obtain the compensated measurement value corresponding to the second beam. Then, based on the compensated measurement value corresponding to the second beam and the prediction value of the first beam, the prediction error corresponding to the first beam is determined. It should be understood that the first distance value or the first distance range herein refers to any one of the at least one distance value or the at least one distance range corresponding to the at least one compensation value in the compensation value set. As an example, the compensation value set includes 8 compensation values, which correspond to 8 distance values or distance ranges, for example, compensation value 1 corresponds to distance value 1 or distance range 1, compensation value 2 corresponds to distance value 2 or distance range 2, and so on, and compensation value 8 corresponds to distance value 8 or distance range 8. Wherein, the 8 compensation values are different from each other, and the 8 distance values or distance ranges are different from each other. When the distance between the position of the first beam and the position of the second beam is determined to be distance value 2 or falls within distance range 2, the measurement value of the second beam is compensated by using compensation value 2; when the distance between the position of the first beam and the position of the second beam is determined to be distance value 8 or falls within distance range 8, the measurement value of the second beam is compensated by using compensation value 8, and so on.

[0194] In a possible implementation, in the case where the network side configures the terminal device with the compensation value set, after the terminal device determines the second beam associated with the first beam, the terminal device compensates the measurement value of the second beam by default by using the compensation value. The selection of the compensation value can be based on any one of the above implementations, which is not limited.

[0195] In another possible implementation, the network side configures the terminal device with the compensation value set, but whether the terminal device compensates the measurement value of the second beam is related to the distance between the position of the first beam and the position of the second beam. For example, after the terminal device determines the second beam associated with the first beam, the terminal device further needs to calculate whether the distance between the position of the first beam and the position of the second beam is greater than or equal to a second threshold value. In the case where the distance between the positions of the first beam and the second beam is greater than or equal to the second threshold value, the measurement value of the second beam is compensated by using the compensation value; otherwise, the measurement value of the second beam is not compensated. In the case of compensation, the prediction error corresponding to the first beam is calculated based on the compensated measurement value corresponding to the second beam; in the case of no compensation, the prediction error corresponding to the first beam is calculated based on the measurement value of the second beam.

[0196] Further, optionally, K is an integer greater than or equal to 1. When K is equal to 1, it means that the terminal device determines one optimal candidate beam based on AI model inference, and at this time, the optimal candidate beam is the first beam. If the first beam is not covered by the first beam set, after determining the second beam associated with the first beam from the first beam set based on the association relationship information, if the measurement value of the second beam is not the beam with the largest measurement value in the first beam set, and the measurement value of the second beam is less than the predicted value of the first beam, the measurement value of the second beam is used as the measurement value of the first beam to calculate the prediction error corresponding to the first beam. The calculated prediction error is likely to be inaccurate. In this case, the measurement value of the second beam is compensated by the compensation value, which can realize more accurate calculation of the prediction error. As an example, if K is equal to 1, the measurement value of the second beam can be compensated by the compensation value by default. When K is greater than 1, it means that the terminal device determines multiple optimal candidate beams based on AI model inference. The multiple optimal candidate beams are the K beams in the embodiments of the present application. In this case, if the first beam in the K beams falls outside the first beam set, the terminal device can determine whether to compensate the measurement value of the second beam based on whether the distance between the position of the first beam and the position of the second beam is greater than the second threshold, to improve the accuracy of the prediction error calculation.

[0197] As an example, the compensation value can be determined in the following ways: 1) a default compensation value is set, which can be obtained by subtracting the average RSRP of the monitoring resource set from the maximum RSRP in the historical prediction task; 2) as described in one of the above implementation manners, the compensation value is related to the beam position, and the compensation value can be obtained by subtracting the measurement value of the position corresponding to the second beam from the maximum RSRP in the historical prediction task (hereinafter referred to as the maximum prediction value). The measurement value of the position corresponding to the second beam can be the average of multiple measurement values within a time window; 3) as described in another implementation manner, the compensation value is related to the distance between the beam positions, and in this implementation, taking the first compensation value corresponding to the first distance range as an example, the first compensation value can be obtained by averaging the difference results obtained by subtracting the measurement values of multiple second beams whose distances from the beam position corresponding to the maximum prediction value fall within the first distance range from the maximum prediction value. For example, in the full beam set, there are 4 beams whose distances from the beam position corresponding to the maximum prediction value fall within the distance range 1, and then the compensation value corresponding to the distance range 1 can be obtained by averaging the 4 difference results obtained by subtracting the prediction values of the 4 beams from the maximum prediction value respectively.

[0198] Optionally, the set of compensation values can be set based on the representation of the positions of the beams in the first full set of beams or the second full set of beams. For example, in the example described above, the distance between two beams can be calculated with reference to the positions in the first full set of beams respectively, or with reference to the positions in the second full set of beams respectively, and the setting of the compensation values is similar.

[0199] 270. The terminal device compensates the measurement value of the second beam based on the set of compensation values, and the compensated measurement value of the second beam is used for calculation of the prediction error corresponding to the first beam.

[0200] Optionally, the calculation of the prediction error can be performed at the terminal side or the network side. When performed by the terminal, the terminal device calculates the prediction error based on the prediction value of each of the K beams and the corresponding measurement value, and obtains the model monitoring result. Then, the terminal device reports the model monitoring result to the network side. When performed by the network side, the terminal device reports the prediction value of each of the K beams and the corresponding measurement value (which can be the compensated measurement value) to the network side, and the network device calculates the prediction error based on the information reported by the terminal device, and obtains the model monitoring result.

[0201] The above method steps 210-270 are only for describing the implementation process of the method, and do not limit the order between the steps.

[0202] In summary, in the embodiments of the present application, based on the association relationship configured by the network device, when the first beam in the K predicted beams determined based on model inference falls outside the first beam set, the terminal device can determine the second beam associated with the first beam from the first beam set based on the association relationship information, so as to obtain the measurement value of the second beam. The measurement value of the second beam will be used for calculation of the prediction error corresponding to the first beam, which can solve the problem that the prediction error cannot be calculated when the first beam set cannot cover all K predicted beams. In addition, through the compensation value mechanism, the accuracy of the prediction error calculation can be improved.

[0203] The method 200 will be described by examples 1 and 2.

[0204] Example 1

[0205] In the performance monitoring process of the AI model deployed on the UE side, the resources for model performance monitoring (corresponding to the beams in the first beam set or set M described above, which can also be referred to as monitoring resources) issued by the network side cannot completely match the K predicted beams determined based on the AI model. In this case, the network side issues the first beam set and also issues the association relationship between at least one beam in the first beam set and the beams in the second beam set. In this way, even if part of the K beams falls outside the first beam set, the measurement values of these beams cannot be directly obtained, but as long as these beams have associated beams, the measurement values of the associated beams can be used to calculate the prediction errors corresponding to these beams.

[0206] In the embodiments of the present application, the monitoring resource set does not completely match or adapt to the predicted beams, or the monitoring resource set does not completely match or adapt to the prediction results of the model, which means that the K beams determined based on the model inference do not all fall into the monitoring resource set, or in other words, cannot all be covered by the monitoring resource set.

[0207] FIG. 11 is an example of a model monitoring method provided by an embodiment of the present application.

[0208] 301. The network device configures a first beam set and association relationship information.

[0209] The association relationship information indicates the association relationship between at least one beam in the first beam set and the beams in the second beam set.

[0210] Taking FIG. 5 or FIG. 6 described above as an example, the first beam set can be set M, and the second beam set can be represented as set 2. As an example, set 2 is part of set A other than set M. In this example, the association relationship information indicates the association relationship between at least one beam in set M and the beams in set 2. Set A is an example of the first full beam set.

[0211] 302. The network device sends the configuration information of the first beam set and the association relationship information.

[0212] The UE receives the configuration information of the first beam set and the association relationship information from the network side.

[0213] 303. The network device sends the first beam set and the fourth beam set.

[0214] The terminal device measures the beams in the fourth beam set, and the measurement results obtained are input information of the AI model. That is, the measurement results of the beams in the fourth beam set are input information of the AI model when the terminal device performs AI model inference, and the output information of the AI model is the prediction information in the above embodiments, where the prediction information indicates predicted values of the K beams. The K beams can be the top-K beams in set A. For example, in FIG. 5 or FIG. 6, the fourth beam set can be set B.

[0215] In a possible implementation, the second beam set includes beams in the fourth beam set used for predicting the K beams. In other words, the beams in the fourth beam set are used to predict the K beams, and specifically, the measurement results of the beams in the fourth beam set are input information of the model prediction, as described above. As in the above embodiments, the second beam set can be a set of narrow beams, or can also be a set of wide beams. In the implementation where the second beam set is a set of wide beams, as an example, the second beam set can include one or more beams in the fourth beam set.

[0216] The terminal device measures the beams in the first beam set, and the measurement results obtained are the measurement information in the above embodiments, which indicates measurement values of the beams in the first beam set.

[0217] For example, in set M and set B, the UE measures the beams in set B to obtain input information for AI model prediction. In addition, the UE also measures the beams in set M to obtain measurement results of the beams in set M. As an example, the measurement results can be RSRP.

[0218] 304. The UE performs model inference to obtain prediction information.

[0219] Specifically, the measurement results of the beams in set B are used as input of the AI model to perform model inference to obtain prediction information. The prediction information can be predicted values of the K beams.

[0220] As described above, depending on the type of the AI model, the predicted values indicate different information. For example, if the AI model is a classification model, the predicted value of each of the K beams is the probability that the beam becomes the optimal beam; if the AI model is a regression model, the predicted value of each of the K beams is the predicted RSRP of the beam.

[0221] 305. The UE determines the measurement values corresponding to the K beams respectively.

[0222] For the beams in the K beams falling into the first beam set, the measurement value corresponding to the beam is the actual measurement value of the beam. For example, if beam A is included in the K beams, beam A is a beam in the first beam set, and the UE obtains the measurement value of each beam in the first beam set because the UE measures the beams in the first beam set. The measurement value corresponding to beam A is the actual measurement value of beam A. If beam B is included in the K beams, beam B does not belong to the first beam set. At this time, the UE determines beam C associated with beam B from the first beam set based on the association relationship information, and takes the measurement value of beam C as the measurement value corresponding to beam B. Thus, in subsequent error calculation, the prediction error corresponding to beam A is determined according to the predicted value of beam A and the measurement value of beam A; the prediction error corresponding to beam B is determined according to the predicted value of beam B and the measurement value of beam C.

[0223] It can be seen that if the top-K beams determined based on the model prediction fall in set M, the UE compares the predicted values and the true measurement values of the top-K beams to calculate the prediction error; if some of the predicted top-K beams fall outside set M, or in other words, the beams fall in set 2, the UE finds the beams associated with the beams in set M through the association relationship information, and calculates the prediction error based on the measurement values of the associated beams and the predicted values of the beams.

[0224] 306、The UE sends information for model monitoring to the network device.

[0225] As an example, in Example 1, the information for model monitoring can include the predicted values of the K beams and the measurement values corresponding to the K beams. Further, the network device calculates the prediction error corresponding to each of the K beams according to the predicted value of each of the K beams and the measurement value corresponding to each of the K beams. Based on the prediction error corresponding to each of the K beams, the performance of the AI model is monitored.

[0226] In this example, after the UE obtains the predicted values and the corresponding measurement values of the K beams, the UE sends the information to the network device, and the network device calculates the prediction errors. As another example, the prediction errors can also be calculated at the UE side, and the UE feeds back the prediction errors of the K beams to the network device after the calculation. Alternatively, the UE can further process the prediction errors of the K beams, for example, calculate an average prediction error, or calculate other parameters that can evaluate the performance of the AI model based on the prediction errors of the K beams, and feed back the processed results to the network device. Alternatively, the terminal device can also trigger the reporting of the prediction errors based on an event. As an example, the UE calculates the prediction errors of the K beams and calculates an average value. The average value is compared with a set threshold. If the average value is lower than the set threshold, the terminal device feeds back the event to the network device.

[0227] In example 1, in the case that the beam set M for model monitoring issued by the network device cannot be completely matched with the predicted values, the association relationship is designed based on the similarity between the beams in set M and the beams in set 2, which realizes the expansion of the monitoring resource set and the balance between the accuracy and real-time performance of the prediction errors.

[0228] Example 2

[0229] In example 1, the network device configures a set of compensation values. In the process of model performance monitoring, if the first beam of the K beams predicted by the UE does not fall into the first beam set (i.e., the set of monitoring resources), after the UE determines the second beam associated with the first beam from the first beam set based on the association relationship information, the UE adds the corresponding compensation value determined from the set of compensation values to the measurement value of the second beam to obtain the compensated measurement value of the second beam. The prediction error corresponding to the first beam is calculated using the compensated measurement value of the second beam, and the calculated prediction error is more accurate.

[0230] FIG. 12 is another example of the method for model monitoring provided by the embodiments of the present application.

[0231] 501. The network device configures a first beam set and association relationship information.

[0232] Step 501, refer to the description of step 301, which will not be repeated here.

[0233] 502. The network device configures a set of compensation values.

[0234] Step 502, refer to the various implementations in the foregoing step 260.

[0235] 503、The network device sends configuration information of the first beam set and the association relationship information.

[0236] The terminal device acquires the configuration information and the association relationship information of the first beam set, and determines the first beam set according to the configuration information.

[0237] 504、The network device sends the first information, and the first information indicates the compensation value set.

[0238] The terminal device receives the first information and acquires the compensation value set according to the first information.

[0239] 505、The network device sends the first beam set and the fourth beam set.

[0240] The terminal device measures the beams in the first beam set, and the obtained measurement result is the measurement information in the above embodiment, which indicates the measurement values of the beams in the first beam set.

[0241] The terminal device measures the beams in the fourth beam set, and the obtained measurement result is the input information of the AI model.

[0242] 506、The UE performs model inference to obtain prediction information.

[0243] The UE takes the measurement result of the beam in the fourth beam set as the input information of the AI model, performs model inference, and obtains the prediction information, i.e., determines the predicted values of the K beams.

[0244] 507、The UE determines the respective measurement values of the K beams.

[0245] For the beams in the K beams falling into the first beam set, the measurement value corresponding to the beam is the actual measurement value of the beam. For example, if beam A is included in the K beams, beam A belongs to the beams in the first beam set. Since the UE measures the beams in the first beam set, the measurement value of each beam in the first beam set is obtained, the measurement value corresponding to beam A is the actual measurement value of beam A. If beam B is included in the K beams, beam B does not belong to the first beam set. At this time, the UE determines beam C associated with beam B from the first beam set based on the association relationship information, and can obtain the measurement value of beam C. Different from example 1, in example 2, the UE determines a first compensation value from the compensation value set and adds the first compensation value to the measurement value of beam C to obtain the compensated measurement value of beam C, and takes the compensated measurement value of beam C as the measurement value corresponding to beam B. Thus, in subsequent error calculation, the prediction error corresponding to beam A is determined according to the predicted value of beam A and the measurement value of beam A; the prediction error corresponding to beam B is determined according to the predicted value of beam B and the compensated measurement value of beam C. It should be understood that after the UE determines beam C associated with beam B, how to select the compensation value for compensating the measurement value of beam C can refer to various implementations in step 260 described above, and will not be described here.

[0246] It can be seen that if the top-K beams determined based on the model prediction fall in set M, the UE compares the predicted value and the actual measurement value of each of the top-K beams to calculate the prediction error; if part of the predicted top-K beams fall outside set M, or in other words, the part of the beams fall in set 2, the UE finds the associated beams in set M for each of the part of the beams through the association relationship information, and adds the corresponding compensation value to the measurement value of the associated beams in set M. Subsequently, the prediction error of the beams falling outside set M is calculated based on the predicted value of each of the beams and the compensated measurement value of the associated beams in set M.

[0247] 508. The UE sends information for model monitoring to the network device.

[0248] As an example, in example 2, the information for model monitoring can include the predicted value of each of the K beams and the measurement value corresponding to each of the K beams. For the beams in the K beams falling outside set M, the measurement value corresponding to the beam is the compensated measurement value of the associated beam. In this implementation, the UE sends the information for model monitoring to the network device, and the network device calculates the prediction error corresponding to each of the K beams to determine the performance of the AI model, thereby achieving performance monitoring of the AI model.

[0249] Optionally, in another implementation, the UE calculates the prediction errors of the K beams respectively, and feeds back the results of the model monitoring to the network device.

[0250] In Example 2, for the first beam falling outside the first beam set among the predicted top-K beams, based on the association relationship information, the measurement value of the associated beam (or the approximate beam) determined in the first beam set cannot truly reflect the actual measurement value of the top-K beams respectively, and the measurement value of the associated beam is generally smaller than the predicted value of the top-K beams. In this case, the application sets a compensation value mechanism to further improve the accuracy of the prediction error calculation and improve the accuracy of the model monitoring.

[0251] Considering that when the above-mentioned scheme provided by the application is applied to the model monitoring task of BM case2, there may be problems of large monitoring resource indication overhead and large air interface resource overhead, the application further provides a corresponding solution to reduce the above-mentioned problems existing in the model monitoring process in BM case2.

[0252] FIG. 13 is a schematic diagram of a scheme for issuing monitoring resources in BM case2. As shown in FIG. 13, when issuing monitoring resources, the network device indicates a corresponding monitoring resource set for each prediction time through DCI. For example, assuming that there are three prediction times, prediction times T0, T1 and T2. The network device sends three DCIs to indicate the respective monitoring resource sets of the three prediction times, for example, set M1, set M2 and set M3. The UE reports the model monitoring metric obtained through the model monitoring process. As can be seen, for each prediction time, the network device will indicate the corresponding monitoring resource set through a DCI. When the monitoring process involves multiple prediction times, and the monitoring resource sets corresponding to different prediction times are different, the indication overhead and the air interface resource overhead are large.

[0253] In view of the shortcomings of the scheme of independently indicating the monitoring resource set set M for each prediction time in BM case2, the application proposes an improved scheme as shown in FIG. 14.

[0254] FIG. 14 is a schematic diagram of a scheme of issuing monitoring resources in the BM case 2 provided by the present application. As shown in FIG. 14, in the embodiments of the present application, the scheme of issuing monitoring resource sets by the network device is as follows: the network device configures the correspondence between at least one monitoring resource set (each monitoring resource set can be represented as set M) and at least two prediction time units. Each prediction time unit corresponds to a different monitoring resource set, or multiple (two or more) prediction time units correspond to the same monitoring resource set. Subsequently, the network device can trigger the terminal device to perform measurement on the corresponding monitoring resource set in different prediction time units and report the measurement information through one DCI. In FIG. 14, three prediction time units are taken as an example. Set M 1, set M 2 and set M 3 can be the same monitoring resource set, or three different monitoring resource sets, or some of the monitoring resource sets are the same, for example, set M 1 and set M 2 are the same monitoring resource set 1, and set M 3 is a monitoring resource set 2.

[0255] FIG. 15 is a schematic flowchart of another method 600 of model monitoring provided by the present application.

[0256] 610. The network device generates configuration information of at least one third beam set and second information, wherein the second information indicates the correspondence between at least two prediction time units and the at least one third beam set.

[0257] In the method 600, one prediction time unit can correspond to the prediction time in the above-mentioned embodiments, or can be a time unit of other granularity, for example, including but not limited to any one or more of the following: time slot, subframe, frame, orthogonal frequency division multiplexing (OFDM) symbol, or a time unit composed of these granularities, for example, one time unit is 2 time slots, or 4 subframes, etc., without limitation.

[0258] As an example, the time unit can also be represented as time instance. The monitoring resource can also be represented as resources for set M / set A for monitoring.

[0259] Optionally, each third beam set includes one or more beams. In the embodiments of the present application, the beam corresponds to the monitoring resource, and accordingly, the beam set corresponds to the monitoring resource set. One third beam set can correspond to one monitoring resource set, for example, one set M in FIG. 13 or FIG. 14.

[0260] When the second information indicates a correspondence between the at least two predicted time units and one third beam set, it indicates that the at least two predicted time units correspond to the same third beam set. At this time, the network device can indicate the correspondence between the one monitoring resource set or beam and the at least two predicted time units through one DCI, so as to realize the monitoring resource delivery.

[0261] When the second information indicates a correspondence between the at least two predicted time units and a plurality of third beam sets, it indicates that the at least two predicted time units correspond to a plurality of third beam sets, for example, each of the at least two predicted time units corresponds to a different third beam set, or part of the predicted time units correspond to the same third beam set, but the other predicted time units each correspond to a different third beam set, etc. As an example, a predicted time unit in the at least two predicted time units corresponds to one-to-one with a third beam set in the at least one third beam set, or a first predicted time unit and a second predicted time unit in the at least two predicted time units correspond to the same third beam set in the at least one third beam set.

[0262] In these implementations, the network device pre-configures the at least one third beam set and the correspondence between the at least one third beam set and the at least two predicted time units through RRC signaling; and subsequently triggers the terminal device to report the measurement information of the third beam set corresponding to each of one or more predicted time units in the plurality of predicted time units through DCI.

[0263] 620, the network device sends configuration information of the at least one third beam set and the second information.

[0264] Correspondingly, the terminal device receives the configuration information and the second information.

[0265] 620, the terminal device measures the third beam set corresponding to each of the at least two predicted time units according to the correspondence, to obtain the measurement information of the third beam set corresponding to each predicted time unit. The measurement information of the third beam set corresponding to one predicted time unit can be simply referred to as the measurement information corresponding to the predicted time unit. The prediction information corresponding to each predicted time unit and the measurement information corresponding to each predicted time unit are used for model monitoring.

[0266] When the at least two prediction time units correspond to the same third beam set, the network device reduces the indication overhead and air interface resource overhead of the set of monitoring resources (i.e., the third beam set for model monitoring) compared with the delivery mode shown in FIG. 13. When the at least two prediction time units correspond to multiple third beam sets, the network device pre-configures the above correspondence relationship through RRC signaling, and then triggers the reporting of the measurement information of the third beam set corresponding to one or more prediction time units through DCI signaling, which also reduces the indication overhead of the monitoring resource and the air interface resource overhead compared with the delivery mode shown in FIG. 13.

[0267] The method 600 is described below in conjunction with Example 3.

[0268] Example 3

[0269] FIG. 16 is an example of a method for model monitoring provided by the present application.

[0270] 701. The network device pre-configures a correspondence relationship between at least one third beam set and at least two prediction time units through RRC signaling.

[0271] Step 701 can refer to the description of step 610, and will not be repeated here.

[0272] 702. The network device sends DCI, which indicates the reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units.

[0273] For clarity of description, the one or more prediction time units indicated by the DCI to report measurement information are referred to as prediction time unit x below. The prediction time unit x indicates the prediction time unit for which the network device indicates the reporting of measurement information.

[0274] As an example, the DCI includes a reporting configuration identifier field, for example, a CSI-reportConfigID field. Since each reporting configuration is associated with a corresponding measurement configuration, the measurement configuration is the set of monitoring resources or the third beam set in the present application. Therefore, the terminal device measures the corresponding set of monitoring resources according to the reporting configuration identifier included in the DCI, and reports the measurement information. In other words, according to the reporting configuration identifier indicated in the DCI, the terminal device can know which prediction time units to report measurement information, thereby determining one or more prediction time units x for which measurement information needs to be reported. The prediction time unit x can include part or all of the at least two prediction time units, which is not limited.

[0275] 703. The network device sends at least one set M.

[0276] Here, at least one set M is in one-to-one correspondence with at least one third beam set, and each set M is a third beam set.

[0277] 704、The network device sends set B.

[0278] 705、The UE measures set B, and performs model prediction (i.e., model inference) by taking the obtained measurement result of set B as the input of the AI model, to obtain prediction information. The prediction information indicates the prediction value corresponding to each of the at least two prediction time units.

[0279] 706、The UE scans set M corresponding to each of the one or more prediction time units (i.e., one or more prediction time units x) indicated by the DCI according to the correspondence relationship, to obtain the measurement value corresponding to each of the one or more prediction time units.

[0280] 707、The terminal device calculates the prediction error based on the prediction value of each of the one or more prediction time units (i.e., one or more prediction time units x) and the corresponding measurement value, to obtain the model monitoring result.

[0281] The terminal device compares the prediction value of each prediction time unit x with the measurement value corresponding to the prediction time unit x, to calculate the prediction error of each prediction time unit x. As described in the above embodiment, the measurement value corresponding to each prediction time unit x can include two cases: 1) the actual measurement value obtained by scanning the corresponding set M on each prediction time unit x; or 2) the actual measurement value obtained by scanning the corresponding set M on each prediction time unit x plus the corresponding compensation value. For the description of the compensation value, please refer to the foregoing embodiment, which will not be repeated here.

[0282] 708、The terminal device sends the model monitoring result to the network device.

[0283] Optionally, step 707 takes the UE to perform the calculation of the prediction error as an example. Similar to the description in the above embodiment, the prediction error calculation can also be performed on the network side. In this implementation, the terminal device sends the prediction value of each prediction time unit x determined based on the model prediction, and the measurement value or the compensated measurement value of each prediction time unit x obtained by scanning the corresponding set M, to the network side. Alternatively, the terminal device can also send other information obtained based on the prediction information and the measurement information to the network device for the calculation of the model monitoring result, which is not limited.

[0284] In Example 3, in the time domain prediction of BM case 2, the network device preconfigures the correspondence between at least two prediction time units and at least one monitoring resource set (i.e., a third beam set) through RRC signaling when issuing the monitoring resource, and then triggers the reporting of the measurement information corresponding to one or more prediction time units through the issuance of one DCI, thereby reducing the additional overhead caused by repeated issuance of the monitoring resource set.

[0285] It should be noted that the method 600 or its examples can be used alone or in combination with the method 200 or its examples described above. When the method 600 is used alone, it can reduce the indication overhead and air interface resource overhead when the monitoring resource is issued in the current model monitoring process of BM case 2; when the method 600 and the method 200 are used in combination, it can not only ensure that the prediction error cannot be calculated due to the mismatch between the monitoring resource and the prediction resource in the model monitoring process, but also reduce the indication overhead and air interface resource overhead when the monitoring resource is issued in BM case 2.

[0286] The method of model monitoring provided in the present application is described in detail above, and the corresponding communication device is introduced below.

[0287] FIG. 17 is a schematic block diagram of a communication device 1000 provided in the present application. As shown in FIG. 17, the communication device 1000 can include a processing module 1001 and a communication module 1002. The communication device 1000 can be a terminal device, or a communication device applied to or matched with the terminal device and capable of realizing the corresponding functions of the terminal device, such as a processor, a chip, a circuit, or an AI entity. Alternatively, the communication device 1000 can be a network device, or a communication device applied to or matched with the network device and capable of realizing the corresponding functions of the network device, such as a processor, a chip, a circuit, or an AI entity.

[0288] The communication module can also be referred to as a transceiver module, a transceiver, a transceiver, or a transceiver device. The processing module can also be referred to as a processor, a processing board, a processing unit, or a processing device. Optionally, the communication module is used to perform the sending operation and the receiving operation of the terminal device side or the network device side in the above method, and the device in the communication module for realizing the receiving function can be regarded as a receiving unit, and the device in the communication module for realizing the sending function can be regarded as a sending unit, that is, the communication module includes a receiving unit and a sending unit. When the communication device 1000 is applied to a network device, the processing module 1001 can be used to realize the processing function of the network device in each of the embodiments of FIGS. 7-16, and the communication module 1002 can be used to realize the transceiving function of the network device in each of the embodiments of FIGS. 7-16.

[0289] It should be noted that the aforementioned communication module and / or processing module can be implemented by a virtual module, for example, the processing module can be implemented by a software function unit or a virtual device, and the communication module can be implemented by a software function or a virtual device. Alternatively, the processing module or the communication module can also be implemented by an entity device, for example, if the device is implemented by a chip / chip circuit, the communication module can be an input / output circuit and / or a communication interface, which performs an input operation (corresponding to the aforementioned receiving operation) and an output operation (corresponding to the aforementioned sending operation); the processing module is an integrated processor or a microprocessor or an integrated circuit.

[0290] The division of the modules in the present application is illustrative, and is only a logical functional division. In actual implementation, another division manner can be used. In addition, each functional module in each example in the present application can be integrated in one processor, or can be a separate physical existence, or two or more modules can be integrated in one module. The integrated module can be implemented in the form of hardware, or in the form of a software functional module, or in the form of a combination of hardware and software.

[0291] FIG. 18 is a schematic block diagram of another communication device 1100 provided by the present application. Optionally, the communication device 1100 can be a chip or a chip system. Optionally, in the present application, the chip system can be composed of a chip, or can include a chip and other discrete devices.

[0292] The communication device 1100 can be used to implement the functions of any one of the network elements (for example, a network device or a terminal device) described in the foregoing embodiments. The communication device 1100 can include at least one processor 1110. Optionally, the processor 1110 is coupled with a memory, which can be located in the communication device 1100, or the memory can be integrated with the processor, or the memory can also be located outside the communication device 1100. As an example, the communication device 1100 can also include at least one memory 1120. The memory 1120 stores necessary computer programs (or computer instructions) and / or data for implementing the corresponding functions of any one of the network elements in any one of the method embodiments described above; the processor 1110 can execute the computer programs stored in the memory 1120 to complete the method implemented by any one of the network elements in any one of the method embodiments described above.

[0293] The communication device 1100 can also include a communication interface 1130, through which the communication device 1100 can exchange information with other devices. For example, the communication interface 1130 can be a transceiver, a circuit, a bus, a module, a pin, or another type of communication interface. When the communication device 1100 is a chip-type device or a circuit, the communication interface 1130 in the communication device 1100 can also be an input / output circuit that can input (or receive) information and output (or send) information. The processor can be an integrated processor or a microprocessor or an integrated circuit or a logic circuit, and the processor can determine output information based on input information.

[0294] In the present application, coupling between devices, units, or modules can be indirect coupling or communication connection, which can be electrical, mechanical, or other forms, and is used for information exchange between devices, units, or modules. The processor 1110 can operate in cooperation with the memory 1120 and the communication interface 1130. In the present application, the specific connection medium between the processor 1110, the memory 1120, and the communication interface 1130 is not limited.

[0295] Optionally, as shown in FIG. 18, the processor 1110, the memory 1120, and the communication interface 1130 are connected to each other through a bus 1140. The bus 1140 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one bus 1140 is represented by a line in FIG. 18, but it does not mean that there is only one bus or only one type of bus.

[0296] In an implementation, the communication apparatus 1100 can be applied to a network side, for example, a network device in the embodiments of the present application, or a host or a cloud device in an OTT system. Specifically, the communication apparatus 1100 can be a network device, or an apparatus capable of supporting the network device to implement the corresponding functions of the network device in any of the above method embodiments. The memory 1120 stores computer programs (or computer instructions) and / or data for implementing the corresponding functions of the network device. The processor 1110 can execute the computer programs or instructions stored in the memory 1120 to complete the methods performed by the network device in any of the above method embodiments. The communication interface in the communication apparatus 1100 can be used to interact with a terminal device, for example, to send the terminal device the association relationship information, to receive the predicted value of the first beam and the measured value of the second beam from the terminal device, to send RRC signaling indicating the correspondence between the at least two prediction time units and the at least one third beam set, to send the terminal device DCI for triggering the reporting of the measurement information of the third beam set corresponding to one or more prediction time units of the at least two prediction time units, or to receive a channel state information (CSI) report from the terminal device, the CSI report indicating the measurement information of the third beam set corresponding to the one or more prediction time units, etc.

[0297] In another implementation, the communication apparatus 1100 can be applied to a terminal side. For example, the communication apparatus 1100 can be a terminal device, or an apparatus capable of supporting the terminal device to implement the corresponding functions of the terminal device in any of the above method embodiments. The memory 1120 stores computer programs (or computer instructions) and / or data for implementing the corresponding functions of the terminal device in any of the above method embodiments. The processor 1110 can execute the computer programs stored in the memory 1120 to complete the methods performed by the terminal device in any of the above method embodiments. The communication interface in the communication apparatus 1100 can be used to interact with a network device (for example, a base station), to send information to the network device or to receive information from the network device, for example, to receive the association relationship information from the network device, to send the network device the predicted value of the first beam and the measured value of the second beam, to receive RRC signaling from the network device, the RRC signaling indicating the correspondence between the at least two prediction time units and the at least one third beam set, to receive DCI from the network device, the DCI for triggering the reporting of the measurement information of the third beam set corresponding to one or more prediction time units of the at least two prediction time units, to send the network device a CSI report, etc.

[0298] FIG. 19 is a schematic structural diagram of a chip provided in the present application. The chip 30 includes a circuit 31 and a communication interface 32. The circuit 31 can be a logic circuit, an integrated circuit, etc., and the communication interface 32 can also be referred to as an input / output circuit, an input / output interface, an interface circuit, etc., and can input information (or receive information) or output information (or send information). The chip 30 can perform the method performed by the network device or the terminal device in the embodiments of the present application. The circuit 31 can be one or more processors, or all or part of the circuit for control or processing in the one or more processors.

[0299] FIG. 20 is a schematic diagram of a system architecture of a communication device provided in the present application. The input / output control is used to manage the input and output signals of the communication device (such as a network device or a terminal device), for example, the input / output control can be in the form of one or more of a modem, a keyboard, a mouse, a touch screen, etc. The input / output control can also be part of the processor. The receiver / transmitter is used to communicate with other devices, and the receiver / transmitter can include a modem for modulating information (transmitting side device) or demodulating modulated information (receiving side device). The antenna is used to transmit or receive signals. The storage can be used to save computer code, which can be executed by the processor to implement the corresponding functions of the communication device. The processor can include intelligent hardware devices such as general-purpose processors, digital signal processors (DSPs), central processing units (CPUs), field-programmable gate arrays (FPGAs), graphics processing units (GPUs), neural processing units (NPUs), etc. The communication device provided in FIG. 20 can be a network device or a terminal device in the embodiments of the present application.

[0300] In addition, the present application also provides a computer-readable storage medium, which stores computer instructions, when the computer instructions are run on a computer, the operations and / or processes performed by the terminal device or the network device in the method embodiments of the present application are performed.

[0301] The present application also provides a computer program product, which includes computer program code or instructions, when the computer program code or instructions are run on a computer, the operations and / or processes performed by the terminal device or the network device in the method embodiments of the present application are performed.

[0302] The application further provides a chip, which comprises a processor, a memory for storing a computer program, and a communication interface. The memory is arranged independently of the chip. The processor is configured to execute the computer program stored in the memory, so that the operations and / or processes performed by the terminal device or the network device in any one of the method embodiments are performed. Further, the chip can further comprise a memory.

[0303] The application further provides a chip, which can comprise a circuit and an input / output interface. The circuit can be a logic circuit, an integrated circuit, etc. Illustratively, the circuit can be one or more processors, or all or part of the circuit in one or more processors for implementing one or more of processing, control or calculation functions. The input / output interface can also be an input / output circuit, or an interface circuit, which can input (or receive) and / or output (or send) information. The chip can comprise a chip system. Alternatively, the chip system can be composed of the chip, or can comprise the chip and other discrete devices. The chip can be used to perform the method implemented by the terminal device or the network device in the embodiments of the application. Alternatively, the chip can be a baseband chip, also known as a modem.

[0304] In addition, the application provides a communication system comprising the terminal device and the network device in any one of the embodiments of the application. The communication system can implement the method for model monitoring provided in any one of the embodiments of FIGS. 7-16.

[0305] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the foregoing method embodiments, which will not be described herein.

[0306] The processor in the embodiments of the present application has signal processing capability, and can be a central processing unit (CPU), and can also be a general processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, and can implement or execute the disclosed methods, steps and logic block diagrams in the present application. The general processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the present application can be directly embodied as hardware processor execution, or executed by a combination of hardware and software modules in the processor. The software module can be located in a random memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory, and combines the hardware to complete the steps of the above method.

[0307] In the embodiments of the present application, the memory is any other medium that can be used to carry or store desired program codes in the form of instructions or data structures and can be accessed by a computer, but is not limited to this. The memory in the present application can also be a circuit or any other device capable of realizing a storage function, used to store computer programs and / or data; or can also be a circuit or any other device capable of realizing a storage function, used to store computer programs and / or data. As an example, the memory can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically EPROM (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM) and direct rambus RAM (DR RAM). It should be noted that the memory of the system and method described herein is intended to include but not limited to the above types or any other suitable types of memory.

[0308] The technical solutions provided in the present application can be realized by software, hardware, firmware or any combination thereof, in whole or in part. When realized by software, the technical solutions can be realized in the form of a computer program product in whole or in part. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a terminal device, an access network device or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (such as floppy disk, hard disk, magnetic tape), optical media (such as digital video disc (DVD)), or semiconductor media, etc.

[0309] At least one (item) involved in the embodiments of the present application means one (item) or more (items). More (items) means two (items) or more than two (items). "And / or" is used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone.

[0310] The term "comprising" mentioned in the embodiments of the present application and any variation thereof is intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but optionally includes other steps or units not listed, or optionally includes other steps or units inherent to the process, method, product or device.

[0311] In the present application, the methods and / or terms between the method embodiments can be mutually referred to each other without logical contradiction, for example, the functions and / or terms between the device embodiments can be mutually referred to each other, for example, the functions and / or terms between the device examples and the method examples can be mutually referred to each other.

[0312] Those skilled in the art can appreciate that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized in electronic hardware or in a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0313] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the units is only a logical function division. In actual implementation, there can be another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0314] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e. they can be located in one place or distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0315] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit.

[0316] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0317] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for model monitoring, characterized in that, The method comprises: obtaining measurement information, the measurement information comprising measurement values of beams in a first beam set; determining, based on association information, a second beam associated with a first beam of K beams from the first beam set, wherein the association information indicates an association relationship between at least one beam in the first beam set and at least one beam in a second beam set, the first beam is not covered by the first beam set, the first beam set is a subset of a first full beam set, the second beam set comprises a complement of the first beam set with respect to the first full beam set and at least one beam in the second beam set comprises the first beam, or the second beam set comprises part or all of a second full beam set and at least one of the beams in the second beam set corresponds to one or more of the first beams, the measurement value of the second beam and the predicted value of the first beam are used for determination of a prediction error corresponding to the first beam, and K is an integer greater than or equal to 1.

2. The method of claim 1, wherein, The method further comprises: obtaining prediction information, the prediction information indicating predicted values of the K beams.

3. The method of claim 2, wherein, The method further comprises: sending the predicted value of the first beam and the measurement value of the second beam.

4. The method according to any one of claims 1 to 3, characterized in that, The second full beam set is the first full beam set; or one beam in the second full beam set corresponds to multiple beams in the first full beam set.

5. The method according to any one of claims 1 to 4, characterized in that, The second beam set comprises beams in a fourth beam set used for predicting the K beams.

6. The method according to any one of claims 1 to 5, characterized in that, The method further comprises: receiving the association information.

7. The method according to any one of claims 1 to 6, characterized in that, The association relationship between the first beam and the second beam comprises: The first beam and the second beam are quasi co-located (QCL); and / or, The distance between the positions of the first beam and the second beam in the first full beam set or the second full beam set is less than a first threshold.

8. The method according to any one of claims 1 to 7, characterized in that, The method further comprises: receiving first information, the first information indicating a set of compensation values, the set of compensation values comprising at least one compensation value, the compensation value being related to a beam position or a distance between beam positions, the beam position indicating a beam in the first beam set or a beam in the first full beam set.

9. The method of claim 8, wherein, The compensation value is related to a beam position, and each compensation value in the set of compensation values corresponds to one or more beam positions.

10. The method according to claim 8 or 9, characterized in that, The measurement value of the second beam and the predicted value of the first beam are used for determination of a prediction error corresponding to the first beam, comprising: The compensated measurement value of the second beam and the predicted value of the first beam are used for determination of a prediction error corresponding to the first beam, wherein the compensated measurement value of the second beam is obtained by compensating the measurement value of the second beam based on the set of compensation values.

11. The method of claim 8, wherein, The compensation value is related to a distance between the beam positions, each compensation value in the compensation value set corresponds to a distance range, the compensation value set includes a first compensation value, the first compensation value corresponds to a first distance value or a first distance range, and the first compensation value is used to compensate the measurement value of the second beam to obtain a compensated measurement value corresponding to the second beam in a case that the distance between the position of the first beam and the position of the second beam is the first distance value or falls within the first distance range.

12. The method of claim 11, wherein, The method further includes: determining that the distance between the position of the first beam and the position of the second beam falls within the first distance range; based on the compensation value set and the first distance range, determining the compensated measurement value corresponding to the second beam, the compensated measurement value corresponding to the second beam being a sum of the measurement value of the second beam and the first compensation value corresponding to the first distance range; The compensated measurement value corresponding to the second beam and the predicted value of the first beam are used for determination of a prediction error corresponding to the first beam.

13. The method according to any one of claims 10 to 12, characterized in that, The K is an integer greater than 1, and the method further includes: determining whether the distance between the positions of the first beam and the second beam in the first full beam set or the second full beam set is greater than or equal to a second threshold value; and in a case that the distance between the positions of the first beam and the second beam in the first full beam set or the second full beam set is greater than or equal to the second threshold value, determining the compensated measurement value corresponding to the second beam.

14. The method according to any one of claims 1 to 13, characterized in that, The method further includes: obtaining configuration information of at least one third beam set and second information, the second information indicating a correspondence between at least two prediction time units and the at least one third beam set; The measurement information includes: According to the correspondence, measuring the third beam set corresponding to each of the at least two prediction time units to obtain measurement information corresponding to each of the at least two prediction time units.

15. The method of claim 14, wherein, The method further includes: receiving radio resource control (RRC) signaling, the RRC signaling indicating the configuration information and the second information; and The method further includes: receiving a downlink control information (DCI), the DCI being used to trigger reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units.

16. A method for model monitoring, characterized in that, including: The configuration association relationship information indicates an association relationship between at least one beam in the first beam set and at least one beam in the second beam set, the first beam set includes a second beam associated with a first beam, the first beam is not covered by the first beam set, the first beam set is a subset of a first full beam set, the second beam set includes a complement of the first beam set relative to the first full beam set and at least one beam in the second beam set includes the first beam, or the second beam set includes part or all of the second full beam set and at least one beam in the second beam set corresponds to one or more first beams. The association relationship information is sent.

17. The method of claim 16, wherein, The method further includes: The predicted value of the first beam and the measured value of the second beam are received; and a prediction error corresponding to the first beam is determined according to the predicted value of the first beam and the measured value of the second beam.

18. The method of claim 16 or 17, wherein, The second full beam set is the first full beam set; or one beam in the second full beam set corresponds to multiple beams in the first full beam set.

19. The method according to any one of claims 16-18, characterized by, The second beam set includes a beam in a fourth beam set used for predicting the K beams.

20. The method of any one of claims 16-19, wherein, The association relationship between the first beam and the second beam includes: The first beam and the second beam are quasi co-located (QCL); and / or the distance between the first beam and the second beam is less than a first threshold.

21. The method of any one of claims 16-20, wherein, The method further includes: First information is sent, the first information indicating a compensation value set, the compensation value set including at least one compensation value, the compensation value being related to a beam position or a distance between beam positions, the beam position indicating a beam in the first beam set or a beam in the first full beam set.

22. The method of claim 21, wherein, The compensation value is related to a beam position, and each compensation value in the compensation value set corresponds to one or more beam positions.

23. The method of claim 21 or 22, wherein, The determination of the prediction error corresponding to the first beam according to the predicted value of the first beam and the measured value of the second beam includes: Based on the compensation value set and the measured value of the second beam, a compensated measured value corresponding to the second beam is determined; and based on the compensated measured value corresponding to the second beam and the predicted value of the first beam, the prediction error corresponding to the first beam is determined.

24. The method of claim 21, wherein, The compensation value is related to a distance between the compensation value and a beam position, and each compensation value in the compensation value set corresponds to a distance value or a distance range, the compensation value set including a first compensation value, the first compensation value corresponding to a first distance value or a first distance range, the first compensation value being used to compensate the measured value of the second beam to obtain the compensated measured value corresponding to the second beam when the distance between the position of the first beam and the position of the second beam is the first distance value or falls within the first distance range.

25. The method of claim 24, wherein, The compensated measurement value corresponding to the second beam is a sum of a measurement value of the second beam and the first compensation value corresponding to the first distance value or the first distance range.

26. The method of any one of claims 16-25, wherein, The method further comprises: sending configuration information of at least one third beam set and second information, the second information indicating a correspondence relationship between at least two prediction time units and the at least one third beam set.

27. The method of claim 26, wherein, The sending of the configuration information of at least one third beam set and the second information comprises: sending RRC signaling, the RRC signaling indicating the configuration information and the second information; and the method further comprises:

28. A method for model monitoring, the method comprising: sending downlink control information DCI, the DCI being used for triggering reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units. comprises: obtaining configuration information of at least one third beam set and second information, the second information indicating a correspondence relationship between at least two prediction time units and the at least one third beam set; and 29. The method of claim 28, wherein, according to the correspondence relationship, measuring the at least one third beam set in the at least two prediction time units to obtain respective measurement information, respective prediction information and respective measurement information corresponding to the at least two prediction time units, the respective measurement information and the respective prediction information corresponding to the at least two prediction time units being used for model monitoring. The obtaining of the configuration information of at least one third beam set and the second information comprises:

30. The method of claim 28 or 29, wherein, receiving RRC signaling, the RRC signaling indicating the configuration information and the second information; and the method further comprises: receiving downlink control information DCI, the DCI being used for triggering reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units.

31. A method for model monitoring, the method comprising: prediction time units in the at least two prediction time units correspond to third beam sets in the at least one third beam set one by one; or a first prediction time unit and a second prediction time unit in the at least two prediction time units correspond to a same third beam set in the at least one third beam set. comprises: generating configuration information of at least one third beam set and second information, the second information indicating a correspondence relationship between at least two prediction time units and the at least one third beam set; 32. The method of claim 31, wherein, and sending the configuration information and the second information. The sending of the configuration information and the second information comprises: sending RRC signaling, the RRC signaling indicating the configuration information and the second information; 33. The method of claim 31 or 32, wherein, and the method further comprises: sending downlink control information DCI, the DCI being used for triggering reporting of measurement information of a third beam set corresponding to one or more prediction time units in the at least two prediction time units. prediction time units in the at least two prediction time units correspond to third beam sets in the at least one third beam set one by one; or a first prediction time unit and a second prediction time unit in the at least two prediction time units correspond to a same third beam set in the at least one third beam set.

34. A communications device, characterized by comprise means or units for performing the method according to any one of claims 1-15, or comprise means or units for performing the method according to any one of claims 16-27, or comprise means or units for performing the method according to any one of claims 28-30, or comprise means or units for performing the method according to any one of claims 31-33.

35. A communications device, characterized by comprise at least one processor configured to execute computer programs or instructions stored in a memory, so that the method according to any one of claims 1-15 is performed, or the method according to any one of claims 16-27 is performed, or the method according to any one of claims 28-30 is performed, or the method according to any one of claims 31-33 is performed.

36. A chip, comprising: comprise a circuit and a communication interface, the communication interface being configured to receive a signal to be processed and send the signal to be processed to the circuit, and the circuit being configured to process the received signal, so that the method according to any one of claims 1-15 is performed, or the method according to any one of claims 16-27 is performed, or the method according to any one of claims 28-30 is performed, or the method according to any one of claims 31-33 is performed.

37. A computer-readable storage medium, comprising: The computer readable storage medium stores computer programs or instructions, which, when executed on a communication device, cause the communication device to perform the method according to any one of claims 1-15, or perform the method according to any one of claims 16-27, or perform the method according to any one of claims 28-30, or perform the method according to any one of claims 31-33.

38. A computer program product, characterised in that, The computer program product comprises computer program codes or instructions for performing the method according to any one of claims 1-15, or comprises computer program codes or instructions for performing the method according to any one of claims 16-27, or comprises computer program codes or instructions for performing the method according to any one of claims 28-30, or comprises computer program codes or instructions for performing the method according to any one of claims 31-33.

Images