Communication method and communication apparatus

Abstract

The present application provides a communication method and a communication apparatus. In the method, a terminal device may determine a count of reported beams based on predictive uncertainty. This allows for flexible variation in the count of reported beams, preventing a plurality of beams output by an AI model from being directly reported, and thereby improving beam management efficiency. In addition, when predictive uncertainty is relatively low, the terminal device can report a relatively small count of beams. This prevents terminal devices and network devices from performing unnecessary measurements on a relatively large count of beams, and reduces signaling overhead.

Description

A communication method and communication device

[0001] This application claims priority to Chinese Patent Application No. 202510197102.5, filed on February 20, 2025, entitled "A Method and Device for Communication", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communications, and more specifically, to a beam management method and a communication apparatus. Background Technology

[0003] To reduce the overhead of beam training, artificial intelligence (AI) models can be used for beam prediction. For example, when an AI model is deployed on the terminal side, the terminal can output the predicted beam information through the AI ​​model. Furthermore, the base station and the terminal can measure the predicted beam and select the beam to be used for subsequent communication.

[0004] However, this method of beam management is inefficient. Summary of the Invention

[0005] This application provides a communication method and communication device that can improve the efficiency of beam management.

[0006] Firstly, a communication method is provided, which can be applied to the terminal device side, for example, to the terminal device or its components (such as chips, circuits, or chip systems). Specifically, the method can be applied to the terminal or the communication module in the terminal, or to the circuits or chips in the terminal responsible for communication functions (such as modem chips, also known as baseband chips, or system-on-chip (SoC) chips or system-in-package (SIP) chips containing modem cores).

[0007] Specifically, the method includes: receiving Q first signals from an access network device, where Q is a positive integer; determining N beams based on the measurement results of the Q first signals and an AI model, where the value of N is determined based on the prediction uncertainty, which is determined based on the AI ​​model and the measurement results of the Q first signals, wherein the input of the AI ​​model includes the measurement results of the Q first signals, and N is a positive integer; and sending indication information of the N beams to the access network device.

[0008] Based on the above scheme, the terminal device can determine the number of beams to be reported according to the prediction uncertainty. This allows the number of beams to be reported to be flexibly changed, avoiding the direct reporting of multiple beams output by the AI ​​model and improving the efficiency of beam management.

[0009] On the other hand, when the prediction uncertainty is low, the terminal device can report a smaller number of beams, avoiding unnecessary measurements of a large number of beams by the terminal device and network device, thus saving signaling overhead.

[0010] For example, prediction uncertainty is used to represent the uncertainty when predicting beam information based on the measurement results of Q first signals and the AI ​​model.

[0011] For example, the output of the AI ​​model includes beam prediction information. This beam prediction information is used to determine N beams. For instance, the beam prediction information includes information on M beams, which is used to determine the beams used for communication between the terminal device and the access network device. The M beams include N beams, where M is a positive integer. In other words, the beam prediction information also includes information on N beams, and the terminal device can select the N beams to report from the M beams included in the beam prediction information.

[0012] Optionally, in this application, sending indication information for N beams to the access network device can be understood as sending indication information to the access network device, which is used to indicate N beams.

[0013] In conjunction with the first aspect, in some implementations of the first aspect, the value of N is determined based on the prediction uncertainty, including: determining the value of N based on the relationship between the prediction uncertainty and the first uncertainty threshold.

[0014] Based on the above scheme, the terminal device can determine the number of beams to be reported according to the relationship between the prediction uncertainty and the first uncertainty threshold. This allows the number of beams to be reported to be adjusted flexibly, making it more flexible and suitable for different communication scenarios.

[0015] As one implementation method, the value of N is determined according to the relationship between the prediction uncertainty and the first uncertainty threshold, including: when the prediction uncertainty is greater than or equal to the first uncertainty threshold, N is a first value; or when the prediction uncertainty is less than the first uncertainty threshold, N is a second value, wherein both the first value and the second value are positive integers, and the first value is greater than the second value.

[0016] Thus, when prediction accuracy is high, fewer beams can be reported, saving signaling overhead. Furthermore, terminal and network devices can subsequently measure fewer beams, thereby not only reducing power consumption and measurement overhead for terminal and network devices, but also improving the efficiency of determining the beams to be used.

[0017] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: receiving first information from a first network element, the first information including a first uncertainty threshold.

[0018] Optionally, the first information may also include a first value and / or a second value.

[0019] In this application, the first information includes a first uncertainty threshold, which can be understood as the first information indicating the first uncertainty threshold. Similarly, the first information also includes a first value and / or a second value, which can be understood as the first information also indicating the first value and / or the second value.

[0020] Based on the above scheme, the network side can indicate the candidate value of the number of beams that need to be reported to the terminal device, so that the number of beams reported by the terminal device is more in line with the current actual prediction, thus improving the efficiency of beam management.

[0021] In conjunction with the first aspect, in some implementations of the first aspect, the value of N is determined based on the prediction uncertainty, including: determining the value of N based on the relationship between the prediction uncertainty and the uncertainty thresholds in the first list, wherein the first list includes at least two uncertainty thresholds.

[0022] The first list can also be called the uncertainty threshold list.

[0023] Based on the above scheme, the terminal device can determine the number of beams to be reported according to the prediction uncertainty and the uncertainty threshold list. This allows for multiple possibilities in the number of beams to be reported, providing greater flexibility and making it suitable for different communication scenarios.

[0024] As one implementation, the value of N is determined based on the relationship between the prediction uncertainty and the uncertainty thresholds in the first list, including: when the prediction uncertainty is greater than or equal to the (i-1)th uncertainty threshold in the first list, and / or when the prediction uncertainty is less than the ith uncertainty threshold in the first list, the value of N is N. i N i It is a positive integer.

[0025] For example, N i The value of is determined by i. For example, N i =i.

[0026] Based on the above scheme, the terminal device can determine the value of N according to the position of the prediction uncertainty in the first list, so that the value of N has multiple possibilities, avoiding the reporting of a fixed number of beams indefinitely and meeting the needs of various scenarios.

[0027] One implementation method involves determining the value of N based on the relationship between the predicted uncertainty and the uncertainty thresholds in the first list. This includes: determining multiple uncertainty intervals based on the first list; and determining the value of N based on the correspondence between the predicted uncertainty, the uncertainty intervals, and the value of N. The correspondence between the uncertainty intervals and the value of N includes the fact that different uncertainty intervals correspond to different values ​​of N. Therefore, the value of N is determined based on the predicted uncertainty, which can be understood as: determining the value of N based on the correspondence between the predicted uncertainty and the uncertainty intervals, where the uncertainty intervals are determined based on the uncertainty threshold list.

[0028] Specifically, the terminal device can obtain multiple uncertainty intervals from the uncertainty threshold list, then determine the uncertainty interval in which the prediction uncertainty is located, and then determine the N corresponding to the uncertainty interval in which the prediction uncertainty is located based on the correspondence between the uncertainty interval and the value of N, which is the value of N.

[0029] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: receiving first information from a first network element, the first information including a first list.

[0030] For example, the first information also includes a second list, which is used to indicate N i Or, in other words, based on the second list, the terminal device can determine N. i The correspondence between i and .

[0031] In this application, the first information includes a first list, which can be understood as indicating the first list. Similarly, the first information also includes a second list, which can be understood as also indicating the second list.

[0032] In conjunction with the first aspect, in some implementations of the first aspect, the output of the AI ​​model also includes prediction uncertainty, wherein the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals, including: inputting the measurement results of Q first signals into the AI ​​model to obtain the prediction uncertainty.

[0033] Based on the above scheme, the AI ​​model can output the prediction uncertainty, which allows the terminal device to directly obtain the prediction uncertainty, simplifying the computational complexity of the terminal device and saving power consumption.

[0034] In conjunction with the first aspect, in some implementations of the first aspect, the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals. This includes determining the prediction uncertainty based on a first gradient, where the first gradient is the gradient of the AI ​​model with respect to the measurement results of the Q first signals. In other words, the first gradient is the gradient of the AI ​​model when the independent variable takes the value of the measurement results of the Q first signals. In other words, the terminal device can first determine the structure of the AI ​​model, then, based on the measurement results of the Q first signals, obtain the value of the gradient of the AI ​​model when the independent variable takes the value of the measurement results of the Q first signals, i.e., the first gradient, and then determine the prediction uncertainty based on the first gradient.

[0035] Based on the above scheme, the terminal device can determine the prediction uncertainty on its own, which simplifies the design and training complexity of the AI ​​model and makes it easy to implement.

[0036] In conjunction with the first aspect, in some implementations of the first aspect, the N beams include a first beam and a second beam, and the spatial interval between the first beam and the second beam is greater than or equal to an interval threshold.

[0037] In one implementation, the first beam and the second beam can be any two of the N beams, or they can be two specific beams from the N beams.

[0038] It should be understood that the relationship between the first beam and the second beam can be referred to the second aspect below, which will not be elaborated here.

[0039] Based on the above scheme, the terminal device can determine the number of beams to be reported according to the prediction uncertainty, and determine which beams to report according to the interval threshold. This not only allows the number of reported beams to be flexibly changed, but also makes the spatial relationship between the reported beams more appropriate. Therefore, it effectively reduces or avoids the situation where the beam quality of multiple reported beams is poor, thereby improving the efficiency of beam management.

[0040] Secondly, a communication method is provided, which can be applied to the terminal device side, for example, to the terminal device or its components (such as chips, circuits, or chip systems). Specifically, the method can be applied to a terminal or a communication module in a terminal, or to a circuit or chip in the terminal that is responsible for communication functions (such as a modem, or a SoC chip or SIP chip containing a modem core).

[0041] Specifically, the method includes: receiving Q first signals from an access network device, where Q is a positive integer; determining N beams based on the measurement results of the Q first signals, an AI model, and an interval threshold, wherein the N beams include a first beam and a second beam, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold, wherein the input of the AI ​​model includes the measurement results of the Q first signals, where N is a positive integer; and sending indication information of the N beams to the access network device.

[0042] Based on the above scheme, the terminal device can determine the reported beams according to the interval threshold. This allows for a larger spatial distance between the reported beams, resulting in lower correlation of beam quality. This can effectively reduce or avoid the situation where multiple reported beams have poor beam quality, thereby improving the efficiency of beam management.

[0043] For example, the output of the AI ​​model includes beam prediction information, which is used to determine N beams. For instance, the beam prediction information includes information on M beams, which is used to determine the beams used for communication between the terminal device and the access network device. The M beams comprise N beams, where M is a positive integer. In other words, the beam prediction information also includes information on N beams, and the terminal device can select the N beams to be reported from the M beams included in the beam prediction information.

[0044] In one implementation, the first beam and the second beam can be any two of the N beams, or they can be two specific beams from the N beams.

[0045] In conjunction with the second aspect, in some implementations of the second aspect, the interval threshold includes a first interval value, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold, including: the spatial interval between the first beam and the second beam is greater than or equal to the first interval value.

[0046] Based on the above scheme, the interval threshold can be a single value, and the terminal device can directly determine the first beam and the second beam based on this interval threshold. This simplifies the computational complexity of the terminal device and saves its power consumption.

[0047] For example, the spatial spacing between the first beam and the second beam is the horizontal spacing between the first beam and the second beam in space; or, the spatial spacing between the first beam and the second beam is the vertical spacing between the first beam and the second beam in space; or, the spatial spacing between the first beam and the second beam is the minimum or maximum value between the horizontal spacing between the first beam and the second beam and the vertical spacing between the first beam and the second beam in space; or, the spatial spacing between the first beam and the second beam is the sum of the horizontal spacing between the first beam and the second beam and the vertical spacing between the first beam and the second beam in space; or, the spatial spacing between the first beam and the second beam is the square root of the sum of the squares of the horizontal spacing between the first beam and the second beam and the vertical spacing between the first beam and the second beam in space.

[0048] Based on the above scheme, the spatial spacing between the first beam and the second beam can be any of the above methods, which can be applied to a variety of application scenarios and has greater versatility.

[0049] In conjunction with the second aspect, in some implementations of the second aspect, the spacing threshold includes a first spacing value and a second spacing value, and the spatial spacing between the first beam and the second beam is greater than or equal to the spacing threshold, including: the horizontal spatial spacing between the first beam and the second beam is greater than or equal to the first spacing value; and / or, the vertical spatial spacing between the first beam and the second beam is greater than or equal to the second spacing value.

[0050] Based on the above scheme, the interval threshold may include a first interval value and a second interval value. The terminal device can determine the first beam and the second beam based on the first interval value and the second interval value, so that the spatial relationship between the determined N beams can be more appropriate.

[0051] In conjunction with the second aspect, in some implementations of the second aspect, the method further includes: receiving second information from the first network element, the second information including an interval threshold.

[0052] Based on the above scheme, the network side can indicate the interval threshold to the terminal device, so that the spatial relationship between the beams reported by the terminal device is more consistent with the current actual prediction situation, thus improving the efficiency of beam management.

[0053] In conjunction with the second aspect, in some implementations of the second aspect, determining N beams based on the measurement results of Q first signals, the AI ​​model, and the interval threshold includes: determining N beams based on the interval threshold when the prediction uncertainty is greater than or equal to a second uncertainty threshold, wherein the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals.

[0054] Based on the above scheme, when the prediction uncertainty is greater than or equal to the second uncertainty threshold, it indicates that the accuracy of the prediction information output by the AI ​​model is low, that is, the certainty of the prediction is not high. In this case, it is very likely that the selected optimal beam is far from the actual optimal beam. Therefore, the terminal device can determine the reported beam according to the interval threshold, which can reduce the spatial correlation between the reported beams, that is, reduce the correlation between the reported beams, thereby improving the robustness of the reported beams.

[0055] In conjunction with the second aspect, in some implementations of the second aspect, the method further includes: receiving third information from the first network element, the third information including a second uncertainty threshold.

[0056] The third information includes the second uncertainty threshold, which can be understood as the third information being used to indicate the second uncertainty threshold.

[0057] In conjunction with the second aspect, in some implementations of the second aspect, the output of the AI ​​model also includes prediction uncertainty. The method further includes: inputting the measurement results of Q first signals into the AI ​​model to obtain the prediction uncertainty.

[0058] Based on the above scheme, the AI ​​model can output the prediction uncertainty, which allows the terminal device to directly obtain the prediction uncertainty, simplifying the computational complexity of the terminal device and saving power consumption.

[0059] In conjunction with the second aspect, in some implementations of the second aspect, the method further includes: determining the prediction uncertainty based on a first gradient, wherein the first gradient is the gradient of the AI ​​model with respect to the measurement results of Q first signals.

[0060] Based on the above scheme, the terminal device can determine the prediction uncertainty on its own, which simplifies the design and training complexity of the AI ​​model and makes it easy to implement.

[0061] In conjunction with the second aspect, in some implementations of the second aspect, the first beam is any one of the N beams, and the second beam is any one of the N beams that is different from the first beam.

[0062] Based on the above scheme, any two beams among the N beams meet the interval threshold limit, which can make the correlation between the N beams low and effectively reduce or avoid the situation where the beam quality of multiple reported beams is poor.

[0063] In conjunction with the second aspect, in some implementations of the second aspect, the signal strengths of both the first beam and the second beam are greater than the first strength threshold.

[0064] Based on the above scheme, any two beams among the N beams meet the first intensity threshold constraint, which can make the intensity of the N beams relatively high and improve the reliability of the reported beams.

[0065] In conjunction with the second aspect, in some implementations of the second aspect, the method further includes: receiving fourth information from the first network element, the fourth information including a first intensity threshold.

[0066] The fourth information includes the first intensity threshold, which can be understood as the fourth information being used to indicate the first intensity threshold.

[0067] Based on the above scheme, the network side can indicate the first intensity threshold to the terminal device, so that the beam reported by the terminal device is more consistent with the current actual prediction situation, thus improving the efficiency of beam management.

[0068] In conjunction with the second aspect, in some implementations of the second aspect, the method further includes: receiving fifth information from the first network element, the fifth information including information about the AI ​​model.

[0069] The fifth piece of information includes information about the AI ​​model, which can be understood as information used to indicate the AI ​​model.

[0070] Thirdly, a communication method is provided, comprising: an access network device sending Q first signals to a terminal device, where Q is a positive integer; the terminal device determining N beams based on the measurement results of the Q first signals and an AI model, wherein the value of N is determined based on the prediction uncertainty, which is determined based on the AI ​​model and the measurement results of the Q first signals, wherein the input of the AI ​​model includes the measurement results of the Q first signals, and N is a positive integer; and the terminal device sending indication information of the N beams to the access network device.

[0071] In conjunction with the third aspect, in some implementations of the third aspect, the method further includes: the first network element sending first information to the terminal device, the first information including a first uncertainty threshold.

[0072] In conjunction with the third aspect, in some implementations of the third aspect, the method further includes: the first network element sending first information to the terminal device, the first information including a first list, the first list being used to determine the value of N.

[0073] Fourthly, a communication method is provided, comprising: an access network device sending Q first signals to a terminal device, where Q is a positive integer; the terminal device determining N beams based on the measurement results of the Q first signals, an AI model, and an interval threshold, wherein the N beams include a first beam and a second beam, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold, wherein the input of the AI ​​model includes the measurement results of the Q first signals, where N is a positive integer; and the terminal device sending indication information of the N beams to the access network device.

[0074] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the method further includes: the first network element sending second information to the terminal device, the second information including an interval threshold.

[0075] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the terminal device determines N beams based on the measurement results of Q first signals, the AI ​​model, and the interval threshold, including: when the prediction uncertainty is greater than or equal to a second uncertainty threshold, the terminal device determines N beams based on the interval threshold, wherein the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals.

[0076] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the method further includes: the first network element sending third information to the terminal device, the third information including a second uncertainty threshold.

[0077] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the output of the AI ​​model also includes prediction uncertainty. The method further includes: the terminal device inputs the measurement results of Q first signals into the AI ​​model to obtain the prediction uncertainty.

[0078] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the method further includes: the terminal device determining the prediction uncertainty based on a first gradient, wherein the first gradient is the gradient of the measurement results of the AI ​​model with respect to Q first signals.

[0079] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the method further includes: the first network element sending fourth information to the terminal device, the fourth information including a first intensity threshold, the first intensity threshold being used to determine N beams.

[0080] In conjunction with the fourth aspect, in some implementations of the fourth aspect, the method further includes: the first network element sending fifth information to the terminal device, the fifth information including information about the AI ​​model.

[0081] Fifthly, a communication device is provided, which can be applied to a terminal device. This communication device has the functions described in the first aspect above. For example, the communication device includes modules, units, or means corresponding to the operations involved in the first aspect. These modules, units, or means can be implemented through software, hardware, or a combination of both. For instance, the communication device can be a terminal device or a component of a terminal device (e.g., a chip, circuit, or chip system).

[0082] Specifically, the device may include: a transceiver unit for receiving Q first signals from an access network device, where Q is a positive integer; a processing unit for determining N beams based on the measurement results of the Q first signals and an AI model, where the value of N is determined based on the prediction uncertainty, which is determined based on the AI ​​model and the measurement results of the Q first signals, wherein the input of the AI ​​model includes the measurement results of the Q first signals, and N is a positive integer; the transceiver unit is also used to send indication information of the N beams to the access network device.

[0083] In conjunction with the fifth aspect, in some implementations of the fifth aspect, the transceiver unit is also used to: receive first information from the first network element, the first information including a first uncertainty threshold.

[0084] In conjunction with the fifth aspect, in some implementations of the fifth aspect, the transceiver unit is also used to: receive first information from the first network element, the first information including a first list, the first list being used to determine the value of N.

[0085] Sixthly, a communication device is provided, which can be applied to a terminal device. This communication device has the functions described in the first aspect above. For example, the communication device includes modules, units, or means corresponding to the operations involved in the first aspect. These modules, units, or means can be implemented through software, hardware, or a combination of both. For instance, the communication device can be a terminal device or a component of a terminal device (e.g., a chip, circuit, or chip system).

[0086] Specifically, the device may include: a transceiver unit for receiving Q first signals from an access network device, where Q is a positive integer; a processing unit for determining N beams based on the measurement results of the Q first signals, an AI model, and an interval threshold, wherein the N beams include a first beam and a second beam, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold, wherein the input of the AI ​​model includes the measurement results of the Q first signals, where N is a positive integer; the transceiver unit is also used to send indication information of the N beams to the access network device.

[0087] In conjunction with the sixth aspect, in some implementations of the sixth aspect, the transceiver unit is also used to: receive second information from the first network element, the second information including an interval threshold.

[0088] In conjunction with the sixth aspect, in some implementations of the sixth aspect, the transceiver unit is also used to: receive third information from the first network element, the third information including a second uncertainty threshold.

[0089] In conjunction with the sixth aspect, in some implementations of the sixth aspect, the output of the AI ​​model also includes prediction uncertainty. The processing unit is also used to: input the measurement results of Q first signals into the AI ​​model to obtain the prediction uncertainty.

[0090] In conjunction with the sixth aspect, in some implementations of the sixth aspect, the transceiver unit is also used to: determine the prediction uncertainty based on a first gradient, where the first gradient is the gradient of the AI ​​model with respect to the measurement results of Q first signals.

[0091] It should be understood that for any parts not described in detail in the third to sixth aspects, please refer to the first and second aspects, which will not be repeated here.

[0092] A seventh aspect provides a communication device comprising an interface circuit and one or more processors. The one or more processors are coupled to a memory. The memory stores part or all of a computer program or instructions necessary for implementing the functions involved in any of the possible implementations of the first to fourth aspects described above. The one or more processors are executable to carry out the computer program or instructions, which, when executed, cause the communication device to implement the methods in any of the possible implementations of the first to fourth aspects described above. The interface circuit is used to implement communication functions within the communication device and / or communication functions between the communication device and other devices or components.

[0093] In one possible design, the processor is used to communicate with other devices or components through the interface circuit.

[0094] In one possible design, the communication device may also include the memory.

[0095] The aforementioned communication device can also be a terminal device, or a module within a terminal device (e.g., a circuit, chip, or chip system), or a logical node, logical module, or software capable of implementing all or part of the functions of the terminal device. The aforementioned communication device can also be an access network device, or a module within an access network device (e.g., a circuit, chip, or chip system), or a logical node, logical module, or software capable of implementing all or part of the functions of the access network device. The aforementioned communication device can also be a first network element, or a module within a first network element (e.g., a circuit, chip, or chip system), or a logical node, logical module, or software capable of implementing all or part of the functions of the first network element.

[0096] Eighthly, a computer-readable storage medium is provided, which stores a computer-readable program or instructions that, when read and executed by a computer, cause the computer to perform any of the possible designs of the first to fourth aspects described above.

[0097] Ninth aspect, a computer program product is provided, comprising a computer program that, when read and executed by a computer, causes the computer to perform the method in any of the possible implementations of the first to fourth aspects described above.

[0098] In a tenth aspect, a chip or chip system is provided, comprising: a processor for executing computer programs or instructions in a memory to implement the methods in any of the possible implementations of the first to fourth aspects described above.

[0099] It should be understood that any beneficial effects not fully described in the third to tenth aspects above can be referred to the first to second aspects and any possible implementation thereof. Attached Figure Description

[0100] Figure 1 is a schematic diagram of a network architecture applicable to an embodiment of this application.

[0101] Figure 2 is a schematic diagram of a beam training method.

[0102] Figure 3 is a schematic flowchart of a communication method provided in this application.

[0103] Figure 4 is a schematic diagram of the structure of the AI ​​model provided in this application.

[0104] Figure 5 is a schematic diagram of the relationship between the first location and the first region provided in this application.

[0105] Figure 6 is a schematic diagram of the beam of the access network equipment provided in this application.

[0106] Figure 7 is a schematic diagram of the beam of the terminal device provided in this application.

[0107] Figure 8 is a schematic flowchart of a communication method provided in this application.

[0108] Figure 9 is a schematic diagram of the distance between the two beams on the loop provided in this application.

[0109] Figure 10 is a schematic diagram of the spacing between the two beams provided in this application.

[0110] Figure 11 is a schematic block diagram of a communication device 2000 provided in an embodiment of this application.

[0111] Figure 12 is a schematic block diagram of a communication device 3000 provided in an embodiment of this application. Detailed Implementation

[0112] The technical solutions in this application will now be described with reference to the accompanying drawings.

[0113] The technical solutions of this application can be applied to various communication systems, such as Long Term Evolution (LTE), 5th Generation (5G), New Radio (NR), Internet of Things (IoT), Wireless-Fidelity (WiFi), wireless communication related to the 3rd Generation Partnership Project (3GPP), or other wireless communication that may emerge in the future. This application does not limit these applications.

[0114] The technical solutions provided in this application can also be applied to machine-type communication (MTC), device-to-device (D2D) networks, machine-to-machine (M2M) networks, Internet of Things (IoT) networks, or other networks. IoT networks, for example, can include vehicle-to-everything (V2X) networks. The communication methods in V2X systems are collectively referred to as vehicle-to-other-device (V2X), where X can represent anything. For example, V2X can include vehicle-to-vehicle (V2V) communication, vehicle-to-infrastructure (V2I) communication, vehicle-to-pedestrian (V2P) communication, or vehicle-to-network (V2N) communication, etc.

[0115] Figure 1 is a schematic diagram of a network architecture applicable to an embodiment of this application. This network architecture may include, but is not limited to, user equipment (UE) and a radio access network (RAN). Optionally, the network architecture may further include one or more of the following: user plane function (UPF), data network (DN), access and mobility management function (AMF), session management function (SMF), and a first network element. The DN may be the Internet. The AMF, SMF, UPF, and first network element are network elements within the core network.

[0116] The following is a brief introduction to some of the network elements shown in Figure 1.

[0117] 1. User equipment (UE): can also be called terminal equipment, access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication equipment, user agent, or user device.

[0118] Terminal devices can be devices that provide voice / data to users, such as handheld devices with wireless connectivity, in-vehicle devices, etc. Currently, examples of terminals include: mobile phones, tablets, laptops, handheld computers, mobile internet devices (MIDs), virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, computing devices or other processing devices connected to wireless modems, terminal devices in 5G networks, or terminal devices in future evolved public land mobile networks (PLMNs), etc., and the embodiments of this application are not limited to these.

[0119] By way of example and not limitation, in this embodiment, the terminal device can also be a wearable device. Wearable devices, also known as wearable smart devices, are a general term for devices that utilize wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, and shoes. Wearable devices are portable devices that are worn directly on the body or integrated into the user's clothing or accessories. Wearable devices are not merely hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are feature-rich, large in size, and can achieve complete or partial functions without relying on a smartphone, such as smartwatches or smart glasses, as well as those that focus on a specific type of application function and require the use of other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0120] Furthermore, in this embodiment, the terminal device can also be a terminal device in an IoT system. IoT is an important component of the future development of information technology. Its main technical feature is to connect objects to the network through communication technology, thereby realizing an intelligent network of human-machine interconnection and object interconnection.

[0121] It should be noted that terminal devices and access network devices can communicate with each other using some air interface technology (such as New Radio (NR) or LTE). Terminal devices can also communicate with each other using some air interface technology (such as NR or LTE).

[0122] In this embodiment, the device for implementing the functions of the terminal device can be the terminal device itself, or it can be a device that supports the terminal device in implementing the functions, such as a chip system or a chip. This device can be installed in the terminal device. In this embodiment, the chip system can be composed of chips, or it can include chips and other discrete devices.

[0123] 2. Radio access network (RAN): This provides authorized users in a specific area with access to a communication network. Specifically, it can include wireless network equipment in the 3rd generation partnership project (3GPP) network or access points in non-3GPP networks.

[0124] The RAN manages radio resources, provides access services to user equipment, and forwards control signals and user equipment data between the user equipment and the core network. The RAN can also be exemplified by a base station in a traditional network.

[0125] For example, the access network device in this application embodiment can be any communication device with wireless transceiver function for communicating with user equipment. The access network equipment includes, but is not limited to: evolved Node B (eNB), baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), access point (AP), wireless relay node, wireless backhaul node, transmission point (TP), or transmission and reception point (TRP) in a wireless fidelity (WIFI) system. It can also be a gNB in ​​a 5G system, such as NR, or a transmission point (TRP or TP), one or a group of antenna panels (including multiple antenna panels) of a base station in a 5G system, or a network node constituting a gNB or transmission point, such as a baseband unit (BBU), or a distributed unit (DU), centralized unit (CU), or radio unit (RU).

[0126] In some deployments, the gNB may include a CU and a DU. Optionally, the gNB may also include an RU. The CU implements some of the gNB's functions, and the DU implements some of the gNB's functions. For example, the CU is responsible for handling non-real-time protocols and services, implementing the functions of the radio resource control (RRC) and packet data convergence protocol (PDCP) layers. The DU is responsible for handling physical layer protocols and real-time services, implementing the functions of the radio link control (RLC), medium access control (MAC), and physical (PHY) layers. The RU implements some physical layer processing functions, radio frequency processing, and related functions of active antennas. Since the information in the RRC layer eventually becomes the information in the PHY layer, or is transformed from the information in the PHY layer, in this architecture, higher-layer signaling, such as RRC layer signaling, can also be understood as being sent by the DU, or by the DU+RU. It is understood that access network equipment can be devices that include one or more of the following: CU nodes, DU nodes, and RU nodes. Furthermore, the CU can be classified as an access network device in the radio access network (RAN) or as an access network device in the core network (CN), and this application does not limit this. Optionally, the CU may include a central unit-control plane (CU-CP) and a central unit-user plane (CU-UP). The RU may be included in radio equipment or radio unit, such as in an RRU, AAU, or RRH.

[0127] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an open-radio access network (O-RAN) system, CU can also be called an open-central unit (O-CU) (open CU); DU can also be called an open-distributed unit (O-DU) (open DU); CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. For ease of description, this application uses CU, CU-CP, CU-UP, DU, and RU as examples. Any of the units CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented through software modules, hardware modules, or a combination of software and hardware modules. For ease of description, the following detailed description uses a base station as an example of an access network device.

[0128] The service area or coverage area of ​​an access network device can include one or more cells. Taking the access network device as a base station as an example, in an analog network using an omnidirectional antenna structure, one cell is the coverage area of ​​one base station; while in a digital cellular mobile network using a 120° angle antenna structure, a cell is one-third of the regular hexagonal area covered by each 120° antenna. Therefore, a base station area can contain one or more cells. Not all cells have a dedicated base station, but they must be covered by a specific base station. That is to say, typically one cell is associated with one access network device, and one access network device can be associated with multiple cells. Terminal devices usually reside in a cell or connect to the access network device in a cell. Both access network devices and cells have their own identifiers. For example, in the current 5G network, access network devices can be identified by the base station identifier (gNB ID), and cells can be identified by the physical cell identity (PCI), new radio cell identity (NCI), cell identity (CI), and new radio cell global identifier (NCGI).

[0129] 3. First network element: Used to support beam management assisted by artificial intelligence (AI) technology, for example, sending an AI model to the terminal device to determine the beam.

[0130] For example, the first network element can be a functional network element in the core network. It can be an independent core network element, or the first network element can be coupled with any of the aforementioned core network elements, and the core network element can implement the function of the first network element.

[0131] For example, the first network element can be an access network device, such as a base station. Optionally, in the O-RAN architecture, the function of the first network element can be performed by the DU, or by the CU, or some functions can be performed by the DU, some by the CU, and some by the RU, without limitation.

[0132] For example, the first network element can be a terminal cloud or a server, which can be a logical function or an independent device, without limitation.

[0133] It should be understood that the network architecture shown in Figure 1 above is only an example. The network architecture applicable to the embodiments of this application is not limited to this. Any network architecture that can realize the functions of the above-mentioned network elements is applicable to the embodiments of this application.

[0134] In the above network architecture, the N1 interface is the interface between the UE and the AMF; the N2 interface is the interface between the RAN and the AMF network elements, used for transmitting radio parameters and non-access stratum (NAS) signaling; the N3 interface is the interface between the RAN and the UPF network elements, used for transmitting user plane data; the N4 interface is the interface between the SMF and UPF network elements, used for transmitting information such as service policies, tunnel identification information for N3 connections, data buffer indication information, and downlink data notification messages; the N6 interface is the interface between the DN and UPF network elements, used for transmitting user plane data; and the N9 interface is the user plane interface between UPFs, used for transmitting uplink and downlink user data streams between different UPFs.

[0135] It should be understood that the network elements and communication interfaces between them shown in Figure 1 are simplified examples using names specified or anticipated in the current 5G system protocols, but this does not limit the embodiments of this application to only currently known communication systems. Therefore, the standard names appearing when describing using current protocols are functional descriptions. This application does not limit the specific names of network elements, interfaces, or signaling, but only indicates the functions of the network elements, interfaces, or signaling, which can be extended to other systems, such as 2G, 3G, 4G, or future communication systems.

[0136] It should be noted that the aforementioned network elements may also be referred to as entities, devices, apparatuses, or modules, etc., and this application does not specifically limit them.

[0137] To facilitate understanding of the embodiments of this application, some basic concepts involved in this application will be briefly explained.

[0138] 1. Millimeter wave communication

[0139] Millimeter wave bands are generally considered to be electromagnetic wave bands ranging from 30 GHz to 300 GHz. Compared to traditional sub-6 GHz bands, millimeter waves have wider spectrum resources, enabling high data rate transmission. Simultaneously, the shorter wavelength of millimeter waves allows for smaller antenna sizes and easier integration of multiple antennas, making millimeter wave communication a key technology for 5G systems and future communication systems. However, the higher the frequency, the greater the channel fading. Compared to traditional sub-6 GHz bands, millimeter wave bands experience significantly greater channel attenuation. Therefore, equipment communicating in the millimeter wave band needs to utilize beamforming or other techniques, employing specific spatial filtering parameters (or spatial filtering parameters, spatial filters, spatial parameters, spatial parameters) to concentrate signal energy in a specific direction, i.e., a specific beam direction, thereby improving the equivalent channel gain between transceiver devices and ensuring coverage performance and data transmission rates for millimeter wave communication. The beam on the transmitting end device side can be called the transmitting beam, and the beam on the receiving end device side can be called the receiving beam.

[0140] 2. Beam Training

[0141] Generally, we believe that there is a correspondence between beam and spatial filtering parameters under a given antenna structure. Therefore, in this application, the concepts of beam and spatial filtering parameters can be used interchangeably in most scenarios. For example, beam identifier can also be understood as spatial filtering parameter identifier. In the initial stage of establishing a connection between transceiver devices, since the location and channel information between the transceiver devices are usually unknown, the transceiver devices need to undergo a beam training process to find a suitable beam direction and its corresponding spatial filtering parameters. Currently, the mainstream beamforming technology in NR systems relies on precoding technology under a multi-antenna structure. Precoding is mainly divided into digital precoding and analog precoding. Digital precoding relies on multiple radio frequency (RF) channels, but due to the high cost of RF channels, the beamforming gain achieved by digital precoding using only a small number of RF channels is relatively small. Currently, the beamforming gain of devices mainly relies on analog precoding. Analog precoding refers to the process where a broadband signal passes through a phase shifter before being transmitted through each antenna array. The signal transmitted by each antenna has a different phase difference relative to the original signal. These different phase differences constitute the analog precoding vector, thereby enabling the signal to be concentrated in a specific direction in space. In this analog precoding framework, the analog precoding vector is its spatial filtering parameter. Because in analog precoding, the entire broadband signal on a single symbol can only be transmitted in one beam direction, searching all possible beam directions during beam training consumes a significant amount of time, resulting in high beam training overhead, low transmission efficiency, and poor mobility support.

[0142] Specifically, in the NR system, the beam training process between the base station and the terminal is completed through a channel state information (CSI) reporting process. The main process involves the base station first configuring multiple reference signals for the terminal, including the time-frequency location, reference signal index, number of ports, and port pattern for each reference signal. When transmitting each reference signal, the base station can use different spatial filtering parameters, i.e., transmit the reference signal in different beam directions. The terminal receives each reference signal configured by the base station and measures its reference signal received power (RSRP), then reports the reference signal indexes and corresponding RSRP quantization values ​​of several reference signals with higher RSRPs. After receiving the reported information from the terminal device, the base station, knowing the spatial filtering parameters used to transmit each reference signal, can ultimately determine which spatial filtering parameters and in which directions will enable the terminal device to receive higher-energy signals, thus completing the beam training process.

[0143] In NR, there are two types of reference signals used for beam training. The first type is the synchronization signal / physical broadcast channel block (SSB), and the second type is the non-zero power channel state information reference signal (NZP-CSI-RS). The SSB is a cell-specific periodic reference signal that each base station periodically transmits. It is mainly used for synchronization between terminals and base stations, as well as for base stations to broadcast basic configuration information within the cell. For robust transmission, base stations typically use a thicker beam when transmitting the SSB compared to when transmitting data. The NZP-CSI-RS can be configured for specific terminals, and the beam thickness and shape used for transmitting it are unrestricted.

[0144] As shown in Figure 2, in a practical system, the base station typically instructs the terminal to measure periodic SSBs (e.g., SSB 1, SSB 2, and SSB 3 in Figure 2) to determine a better wide beam (or coarse beam). Then, it configures several NZP-CSI-RS (e.g., NZP-CSI-RS1, NZP-CSI-RS2, and NZP-CSI-RS3 in Figure 2) for the terminal. Using the determined narrow beam direction within the wide beam direction, it continues to transmit NZP-CSI-RS. Finally, based on the NZP-CSI-RS measurement information reported by the terminal, it determines a better narrow beam direction, which is the beam direction for data transmission.

[0145] 3. AI-based beam training

[0146] To address the aforementioned beam training issues, one approach is to use AI models (such as neural network models) for beam prediction. The following section primarily introduces two types of AI models used for beam prediction.

[0147] The first type of AI model predicts the entire beam information by measuring the results of a portion of the beam, or it inputs the information of a wide beam into the AI ​​model to obtain the information of a narrow beam.

[0148] Specifically, the base station's beams can be divided into two sets, namely set 1 and set 2, where the number of beams in set 1 is generally less than that in set 2. The terminal can measure the beams in set 1, and then use an AI model and the measurement results of the beams in set 1 to predict the beam information in set 2. Set 1 and set 2 can be further divided into the following two cases:

[0149] In the first case, set 1 is a subset of set 2. In this case, the AI ​​model can be understood as being able to predict all beam information by using the measurement results of some beams in set 2.

[0150] The second scenario is that the beams in set 1 are different from the beams in set 2. For example, the beams in set 1 are wider beams, while the beams in set 2 are thinner beams (e.g., the beams needed for subsequent data communication). In this case, the AI ​​model can be understood as being able to predict the information of the thinner beams based on the measurement results of the wider beams.

[0151] For example, as shown in Figure 2, the beam used for transmitting and receiving SSB is typically a wide beam, while the beam used for transmitting and receiving NZP-CSI-RS is a narrow beam. Therefore, the measurement results of SSB can be input into an AI model, which can output information about the beam used for transmitting and receiving NZP-CSI-RS, i.e., the information about the narrow beam can be predicted from the measurement information of the wide beam.

[0152] The second type of AI model can use measurement information from low-frequency base stations to predict high-frequency beam information. When low-frequency and high-frequency base stations are co-located, the channel coefficient between the low-frequency base station and the terminal is highly correlated with the channel coefficient between the high-frequency base station and the terminal. Therefore, the high-frequency beam information can be predicted using measurement information from the low-frequency base station. When low-frequency and high-frequency base stations are not co-located, the measurement information between the low-frequency base station and the terminal can provide radio frequency fingerprint information of the channel. According to the principle of radio frequency fingerprint positioning, this radio frequency fingerprint information can usually be mapped to location information. At a fixed location, the beam information between the terminal and the high-frequency base station is generally determined. Therefore, the beam information of the high-frequency base station can also be predicted using the measurement information between the low-frequency base station and the terminal.

[0153] It should be understood that the first and second types of AI models mentioned above are merely examples. Other types of AI models may also exist, such as a mixture of the first and second types of AI models, or other AI models unrelated to the first and / or second types of AI models. As long as they can output beam prediction information, this application does not limit the input of specific AI models or the structure of AI models.

[0154] When an AI model is deployed on the terminal side, the terminal can input corresponding measurement information into the AI ​​model according to the specific AI model category to obtain prediction information for each beam. Then, the best K beams can be reported to the base station, i.e., the Top K beams. The base station and the terminal can further measure these K beams. If some or all of these K beams have good quality, the base station and the terminal will subsequently use the corresponding beams for communication. Typically, K is an integer greater than 1. Reporting multiple beams, compared to only reporting the single best-quality beam predicted by the AI ​​model, can improve the accuracy of beam prediction.

[0155] However, reporting Top K beams may have the following drawbacks:

[0156] On the one hand, not all terminals need to report K beams. For example, some terminals may have a higher prediction accuracy, in which case reporting only one beam (e.g., the beam with the best prediction quality) is sufficient. Reporting an additional K-1 beams requires unnecessary measurement and reporting between the base station and the terminal, increasing unnecessary overhead.

[0157] On the other hand, when a terminal performs beam prediction, the prediction results usually exhibit strong spatial correlation. That is, beams that are close together typically have similar predicted beam quality values. Therefore, if the terminal directly reports the K beams with the best beam quality, these K beams may be spatially close beams. However, the actual beam quality also exhibits a similar strong spatial correlation. In this case, if any one of the beams is far from the best beam (i.e., its accuracy is low), it will cause all K beams to be far from the best beam, resulting in potentially poor quality for all K beams, meaning insufficient prediction robustness.

[0158] In summary, directly reporting the Top K beams can lead to higher overhead or lower accuracy in subsequent beam reporting and measurement, resulting in lower efficiency in beam management.

[0159] In view of this, this application provides a communication method and communication device that can improve the efficiency of beam management.

[0160] Figure 3 is a schematic flowchart of a communication method provided in this application. As shown in Figure 3, the method 300 includes the following steps.

[0161] S310, the access network device sends Q first signals to the terminal device, and correspondingly, the terminal device measures Q first signals, where Q is a positive integer.

[0162] For example, the first signal is a signal used for beam prediction, which can be a reference signal (or pilot), such as an SSB, a channel state information reference signal (e.g., NZP-CSI-RS), etc. The specific first signal is related to the input of the subsequent AI model. For example, if the AI ​​model is the first type of model mentioned above, then the Q first signals can be reference signals corresponding to the beams in beam set 1, such as the first signal being a reference signal corresponding to a partial wide beam (e.g., an SSB). For example, if the AI ​​model is the second type of model mentioned above, then the Q first signals can be reference signals in a low-frequency base station or low-frequency cell.

[0163] Optionally, when the Q first signals are signals transmitted by a high-frequency base station or signals in a high-frequency cell, the terminal device can use Q different beams to receive and measure the Q first signals, thereby obtaining the measurement results of the Q first signals. Alternatively, the terminal device can use one or more receiving beams for each first signal, and the measurement result of each first signal can include the measurement results corresponding to multiple receiving beams, or a measurement result can be determined from the measurement results corresponding to multiple receiving beams (e.g., the measurement result corresponding to the receiving beam that makes the RSRP the highest).

[0164] Optionally, the terminal device can determine the measurement results of Q first signals. For example, the measurement results of the first signals can be RSRP or normalized RSRP. For instance, the AI ​​model is a first-type AI model, and the input to this AI model can be the RSRP or normalized RSRP corresponding to the beams in set 1. For example, the measurement results of the first signals can be time delay power spectrum, normalized time delay power spectrum, angle spectrum, normalized angle spectrum, time delay angle power spectrum, normalized time delay angle power spectrum, multipath information, etc., where the time delay in the spectrum or the time delay in the multipath information can be the directly measured time delay, or it can be the time delay with the first path delay shifted to 0. The measurement results of the Q first signals can be adapted according to the input of a specific AI model; this application embodiment does not limit the specific measurement results of the first signals.

[0165] S320, the terminal device determines N beams based on the measurement results of Q first signals and the AI ​​model, where N is a positive integer.

[0166] The value of N is determined based on the prediction uncertainty, which is determined based on the measurement results of the AI ​​model and the Q first signals. The input of the AI ​​model includes the measurement results of the Q first signals, and the output of the AI ​​model includes beam prediction information, which is used to determine the N beams. Generally speaking, the greater the uncertainty, the larger N is (or at least it does not decrease).

[0167] In this application, prediction uncertainty is used to represent the uncertainty in predicting beam information based on the measurement results of Q first signals and the AI ​​model. Alternatively, prediction uncertainty is used to represent the uncertainty of beam prediction information, or conversely, prediction uncertainty can also be used to represent the accuracy of beam prediction information.

[0168] In this context, prediction uncertainty can also be replaced with first uncertainty, first uncertainty value, or prediction uncertainty value, etc.

[0169] It should be understood that uncertainty refers to the degree of uncertainty about the predicted result due to the existence of errors. Uncertainty can be used to characterize the quality level of the predicted result. The smaller the uncertainty, the higher the quality and level of the predicted result, and the higher its use value; the larger the uncertainty, the lower the quality and level of the predicted result, and the lower its use value. Conversely, uncertainty also indicates the reliability or accuracy of the predicted result. Therefore, uncertainty can also be replaced by confidence, accuracy, certainty, prediction confidence, prediction accuracy, prediction certainty, etc., and can further be reflected in the variance or standard deviation of the predicted value. Uncertainty and confidence can be understood as having an inverse relationship: the higher the uncertainty, the lower the confidence; the higher the uncertainty, the lower the confidence. Confidence can be determined using methods similar to those used to determine prediction uncertainty (which will be explained in detail later), or the prediction uncertainty can be determined first, and then the confidence can be determined by taking the reciprocal, inversion, etc. Generally speaking, the larger the confidence, the smaller N is (or at least it does not increase). For ease of explanation, the following descriptions will use uncertainty; however, the scheme described below can also be extended to determine N beams based on confidence levels. When extended to determine N beams based on confidence levels, the following order will be reversed.

[0170] As one implementation, the output of the AI ​​model also includes prediction uncertainty. For example, as shown in Figure 4(a), the input of the AI ​​model includes the measurement results of Q first signals, and the output of the AI ​​model includes beam prediction information and prediction uncertainty.

[0171] In this implementation, the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals. Specifically, this means that the measurement results of the Q first signals are input into the AI ​​model to obtain the prediction uncertainty. In other words, the output of the AI ​​model not only directly includes the prediction information of beam quality, but also directly includes the uncertainty.

[0172] Specifically, in this application, the AI ​​model can be constructed and trained by the first network element and then distributed to the terminal device for use. When constructing the AI ​​model, the first network element can construct it to output two types of information: beam prediction information and prediction uncertainty. Furthermore, the first network element can train the AI ​​model so that the terminal device can use the AI ​​beam to perform beam prediction and obtain the prediction uncertainty.

[0173] For example, during the training phase, the prediction uncertainty for a specific measurement sample can be any of the following forms:

[0174] Form 1: The prediction uncertainty is determined by the first network element based on the difference between the beam quality measured in this measurement and the beam quality measured at other surrounding locations.

[0175] For example, as shown in Figure 5, the horizontal axis represents the east-west position coordinates, the vertical axis represents the north-south position coordinates, the points in the figure represent positions, and the gradient color bar on the right represents the RSRP value of a certain beam. Taking a data sample i at the location of the pentagram in Figure 5 (hereinafter referred to as the first position) as an example, if the AI ​​model is the aforementioned second type of AI model, the input of this AI model is the measurement information of the low-frequency base station measured by the UE at the first position. The output of the first AI network includes the beam information of the high-frequency base station measured by the UE at the first position (the RSRP value of each beam, or the normalized RSRP value of each beam, or the probability that each beam is the optimal beam, the predicted one or more beam indices, one or more beam pair indices, etc.). In addition, the output of the AI ​​model may also include the prediction uncertainty, the label value of which (i.e., the target value during the model training phase) is determined according to the following formula:

[0176] In the above formula, R i The beam information representing sample i at the first position of the high-frequency base station can be, for example, a tensor composed of RSRP measured on each beam or normalized RSRP, the shape of which is related to the number of base station beams and / or the number of UE beams. j This represents the beam information of the high-frequency base station for sample j within the first region. The first region is the area surrounding the first location, such as a circle (as shown in Figure 5), sphere, ellipse, ellipsoid, rectangle, square, cuboid, cube, etc., centered on the first location. The set of indices of the samples (i.e., the beam information of the high-frequency base station) located within the first region is... It can include i or not (that is, j can take the same value as i, or it can take only a value different from i). This represents the number of samples (or the remaining samples after removing sample i) within the first region. The UE (e.g., UEs participating in training sample collection, which may or may not include the terminal device executing step S320) can report the beam information of the high-frequency base station it measures to the first network element. The first network element then performs model training. In the actual training process, after collecting a large amount of data, the first network element can also perform certain measurement data filtering to construct... Optionally, the UE can also report its own location information for the first network element to determine the first area; optionally, when a UE reports multiple samples, it can also report the relative distance between the multiple samples for the first network element to determine the first area; |R i -R j | F R represents i -R j The F-norm, when R i -R j When it is a matrix, mathematically it is R. i -R j The square root of the sum of the squares of all the elements.

[0177] It should be understood that the difference in Form 1 can be appropriately processed beyond the F-norm to represent the prediction uncertainty. For example, the prediction uncertainty can be further divided by the number of beams (i.e., R) based on the square root of the sum of the above squares. i -R j (The number of elements in the middle), for example, the prediction uncertainty can be the sum of the absolute values ​​of the differences, or the prediction uncertainty can be the sum of the absolute values ​​of the differences divided by the number of beams, or the prediction uncertainty can be R i -R j The sum of the cube or higher powers of the absolute values ​​of each element, and the corresponding division by the number of beams, etc.

[0178] In Form 1, the uncertainty essentially reflects the degree of fluctuation in the quality of the beam around the first position. When the degree of fluctuation is large, it is generally more difficult to predict and more likely to be inaccurate. Conversely, when the degree of fluctuation is small, it is generally easier to predict and more likely to be accurate.

[0179] Form 2: The uncertainty is determined based on the difference in beam quality between the current measurement and other surrounding measurements, as well as the distance between other surrounding measurements and the current measurement. For example, as shown in Figure 5, the label value of this uncertainty is determined according to the following formula:

[0180] In the above formula, where d i,jThis represents the distance or square of the distance between the two measurements (or other distance-dependent variables), and the meanings of the other parameters are the same as in equation (1). In this form, the measurement closer to the first position will have a greater impact on the uncertainty. This form can also be modified in other ways, such as form 1, which will not be elaborated here.

[0181] It should be understood that forms 1 and 2 above are merely examples, and the prediction uncertainty can also take other forms. For example, the label value of the prediction uncertainty is the deviation between the best one or more beam indices of the current measurement in the first region and the best one or more beam indices of other surrounding measurements. Another example is that the label value of the prediction uncertainty is the high-frequency component after performing a spatial Fourier transform on the beam measurement values ​​in the first region. Yet another example is that the label value of the prediction uncertainty is the high-frequency component after performing a spatial Fourier transform on the optimal beam, etc.

[0182] Furthermore, in each of the specific forms mentioned above, the AI ​​model has a clear label value for the prediction uncertainty during the training phase. Below, we introduce another form of prediction uncertainty, where the AI ​​model's output includes the variance of the current prediction. In this form, the loss function during the training phase can be set as follows:

[0183] in R and R represent the true value and predicted value of the high-frequency beam information, respectively. This represents the variance of the prediction. In this form, during the training phase, the uncertainty output by the AI ​​model does not have a clear label value (i.e., the target ground truth value), but is trained together with the high-frequency beam information.

[0184] As another implementation, the output of the AI ​​model does not directly include prediction uncertainty. For example, as shown in Figure 4(b), the input of the AI ​​model includes the measurement results of Q first signals, and the output of the AI ​​model includes beam prediction information.

[0185] In this implementation, the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals. Specifically, it can mean that the prediction uncertainty is determined based on the first gradient, which is the gradient of the AI ​​model with respect to the measurement results of the Q first signals. Alternatively, the first gradient is the gradient of the AI ​​model when the independent variable takes the values ​​of the measurement results of the Q first signals.

[0186] Specifically, assuming the input to the AI ​​model is x (i.e., the measurement results of Q first signals), where x is typically a vector, matrix, or a higher-dimensional tensor, the AI ​​model can be represented by the function f(·). When the output of the AI ​​model is... At that time, there was action. The gradients of each output of the AI ​​model with respect to the input x are:

[0187] in, Equal to The elements in x, such as x0, x1, etc., are elements in x.

[0188] Furthermore, the prediction uncertainty can be determined based on w, for example, it can be the sum, average, weighted average, sum of squares, weighted sum of squares, square root of the sum of squares, square root of the average of squares, square root of the weighted average of squares, etc. of the moduli of each element of w.

[0189] It should be understood that the gradient refers to a vector or vector that represents the directional derivative of a function at a point in which it reaches its maximum value along that direction. The gradient can reflect the direction and rate of change of the output with respect to the input. When the magnitude of the gradient is large, it means that the output may be unstable for the current input, which also indicates that the uncertainty of the current prediction may be large.

[0190] The value of N can be determined based on the prediction uncertainty, or in other words, the number of beams reported by the terminal device is determined based on the prediction uncertainty. For example, beam prediction information can include information on M beams, which is used to determine the beams used for communication between the terminal device and the access network device. The M beams include N beams, where M is a positive integer. That is, the value of N is an integer less than or equal to M, and the terminal device can select N beams from the M beams based on the first uncertainty.

[0191] As one implementation method, the value of N is determined based on the prediction uncertainty, including: the terminal device determines the value of N based on the relationship between the prediction uncertainty and the first uncertainty threshold.

[0192] For example, when the prediction uncertainty is greater than or equal to the first uncertainty threshold, N is the first value; when the prediction uncertainty is less than the first uncertainty threshold, N is the second value, where both the first and second values ​​are positive integers, and the first value is greater than the second value.

[0193] In other words, when the uncertainty of beam prediction is high, N can take a larger value (i.e., the first value), meaning the terminal device can choose to report a larger number of beams. When the uncertainty of beam prediction is low, N can take a smaller value (i.e., the second value), meaning the terminal device can choose to report a smaller number of beams. Thus, when the prediction accuracy is high, fewer beams can be reported, saving signaling overhead. Furthermore, the terminal device and network device can subsequently measure the fewer beams (i.e., use this measurement to further determine the quality of the predicted beams), thereby not only reducing the power consumption and measurement overhead of the terminal device and network device, but also improving the efficiency of determining the beams to be used.

[0194] For example, the second value can be 1, that is, when the uncertainty of beam prediction is low, the terminal device can report the information of 1 beam.

[0195] For example, the first value can also be a value greater than 1 and less than M. That is, when the uncertainty of beam prediction is high, the terminal device can select information of multiple beams from the M beams output by the AI ​​model and report them, so that the terminal device and network device can measure fewer beams.

[0196] For example, the first value can also be M, that is, when the uncertainty of beam prediction is high, the terminal device can report the information of M beams obtained by the AI ​​model output, so that the terminal device and the network device can measure these beams and then determine the appropriate beam for communication.

[0197] It should be understood that in this application, "greater than or equal to" and "greater than" are interchangeable, as are "less than" and "less than or equal to". That is to say, "equal to" can be divided into the "greater than" branch or the "less than" branch.

[0198] Optionally, the first uncertainty threshold can be predefined by the protocol or configured by the network side to the terminal device, without restriction. For example, in this implementation, method 300 further includes: S301, the first network element sends first information to the terminal device, and correspondingly, the terminal device receives the first information, wherein the first information includes the first uncertainty threshold.

[0199] For example, when the first network element is an access network device, the first information can be carried in downlink control information (DCI), MAC control element (MAC CE), or RRC signaling, without limitation. When the first network element is a core network device, the first information can be carried in the signaling between the core network device and the terminal device.

[0200] Optionally, the first value and / or the second value can be predefined by the protocol or configured by the network side to the terminal device, without limitation. For example, in this implementation, method 300 further includes: the first network element sending sixth information to the terminal device, and correspondingly, the terminal device receiving the sixth information, wherein the sixth information includes the first value and / or the second value. Optionally, the sixth information and the first information can be sent simultaneously in the same message; in other words, the first information may include the first value and / or the second value.

[0201] As another implementation method, the value of N is determined based on the prediction uncertainty, including: the terminal device determines the value of N based on the relationship between the prediction uncertainty and the uncertainty threshold in the first list.

[0202] The first list can be used to describe the correspondence between the uncertainty threshold and the number of beams that need to be reported.

[0203] For example, the first list may include at least two uncertainty thresholds. When the prediction uncertainty is greater than or equal to the (i-1)th uncertainty threshold in the first list, and the prediction uncertainty is less than the ith uncertainty threshold in the first list, the value of N is N. i N i N is a positive integer. i The value of N is determined by i, where i is a positive integer. Alternatively, the case of equality can also be categorized under the condition of less than, that is, when the prediction uncertainty is greater than the (i-1)th uncertainty threshold in the first list, and the prediction uncertainty is less than or equal to the ith uncertainty threshold in the first list, the value of N is N. i In this application, similar explanations can be substituted accordingly, and will not be repeated here. That is to say, the terminal device can determine the value of N based on the position of the prediction uncertainty in the first list.

[0204] For example, the 0th uncertainty threshold can be understood as the lower bound of uncertainty (e.g., 0 or negative infinity), which may not exist directly in the first list. Therefore, in the above method, for i=1, it can also be understood that when the prediction uncertainty is less than the 1st (i.e., the i-th) uncertainty threshold in the first list, N takes the value N1. That is, when the prediction uncertainty is less than the smallest threshold in the first list, the prediction uncertainty no longer needs to satisfy the condition of being greater than the branch. Similarly, when the first list includes V thresholds, the (V+1)th threshold can be understood as the upper bound of uncertainty (e.g., 1 or positive infinity), which may also not exist directly in the first list. Therefore, in the above method, for i=V+1, it can also be understood that when the prediction uncertainty is greater than or equal to the Vth (i.e., the (i-1)-th) uncertainty threshold in the first list (i.e., when the prediction uncertainty is greater than or equal to the largest threshold in the first list, the condition of being less than the branch no longer needs to be satisfied), N takes the value N1. V+1 .

[0205] For example, when the first list includes a lower bound for the uncertainty threshold, the above method can be replaced by: when the prediction uncertainty is greater than or equal to the (i-1)th uncertainty threshold in the first list, and the prediction uncertainty is less than the ith uncertainty threshold in the first list, the value of N is N. i+1 .

[0206] For example, when the first list includes an upper bound for the uncertainty threshold, for instance, assuming the first list includes V thresholds, then the Vth threshold is the upper bound for the uncertainty (e.g., 1 or positive infinity). In this case, i = V can be understood as: when the prediction uncertainty is greater than or equal to the Vth (i-1th) uncertainty threshold in the first list, and the prediction uncertainty is less than the Vth (i.e., the i-th) uncertainty threshold in the first list, the value of N is N. v .

[0207] Optionally, the uncertainty thresholds in the first list are arranged in ascending order.

[0208] In one example of this implementation, N i =i, meaning that when the prediction uncertainty is greater than or equal to the (i-1)th uncertainty threshold in the first list, and the prediction uncertainty is less than the ith uncertainty threshold in the first list, the value of N is i.

[0209] Table 1 provides an example of the first list. As shown in Table 1, this first list includes two uncertainty thresholds, namely threshold #1 and threshold #2, where threshold #1 is less than threshold #2. When the prediction uncertainty is greater than or equal to threshold #1 and less than threshold #2, i.e., i = 2, then N = 2. When the prediction uncertainty is less than threshold #1, i.e., i = 1, then N = 1. When the prediction uncertainty is greater than or equal to threshold #2, i.e., i = 3, then N = 3.

[0210] Table 1

[0211] For example, multiple uncertainty intervals can be determined based on the first list, and the value of N can be determined based on the correspondence between the predicted uncertainty and the uncertainty intervals and the value of N. The correspondence between the uncertainty intervals and the value of N includes the fact that different uncertainty intervals correspond to different values ​​of N. Therefore, the value of N is determined based on the predicted uncertainty, which can be understood as: determining the value of N based on the correspondence between the predicted uncertainty and the uncertainty intervals, where the uncertainty intervals are determined based on the uncertainty threshold list. Specifically, the terminal device can obtain multiple uncertainty intervals from the uncertainty threshold list, then determine the uncertainty interval in which the predicted uncertainty lies, and then determine the N corresponding to the uncertainty interval in which the predicted uncertainty lies, which is the value of N, based on the correspondence between the uncertainty intervals and the value of N.

[0212] For example, when the first list includes threshold #1 and threshold #2, the terminal device can obtain the uncertain threshold intervals shown in the first column of Table 2. Different intervals correspond to different values ​​of N (for example, these can be determined through the second list below). For instance, the value of N corresponding to the i-th interval is N. i (Special, N) i =i), as shown in Table 2. When the prediction uncertainty falls into a certain interval, the value of N is the value of N corresponding to that interval.

[0213] Table 2

[0214] For example, N i The value of N is determined according to the second list, which is used to indicate N. i Or, indicating i and N i The correspondence is as follows: When the prediction uncertainty is greater than or equal to the (i-1)th uncertainty threshold in the first list, and the prediction uncertainty is less than the ith uncertainty threshold in the first list, N takes the value N. i N i The correspondence with i is determined by the second list.

[0215] For example, Table 3 provides an example of the second list. Combining Tables 1 and 3, we can see that when the prediction uncertainty is greater than or equal to threshold #1 and less than threshold #2 (i.e., i = 2), N = 5. When the prediction uncertainty is less than threshold #1 (i.e., i = 1), N = 4. When the prediction uncertainty is greater than or equal to threshold #2 (i.e., i = 3), N = 6. Optionally, Table 3 may only contain the second column, in which case i defaults to the row number in that list.

[0216] Table 3

[0217] Alternatively, the uncertainty thresholds in the first list can also be arranged in descending order. In this case, the value of N can be determined based on the following method: for example, when the prediction uncertainty is greater than or equal to the i-th uncertainty threshold in the first list, and the prediction uncertainty is less than the (i-1)-th uncertainty threshold in the first list, the value of N is N. i N i N is a positive integer. i The value of is determined by i, where i is a positive integer. The relationship between Ni and i can be found above. Similarly, the condition equal to can also be placed in the less than branch, which will not be elaborated here.

[0218] Optionally, the first list can be predefined by the protocol or configured by the network side to the terminal device, without restriction. For example, in this implementation, method 300 further includes: S301, the first network element sends first information to the terminal device, and correspondingly, the terminal device receives the first information, wherein the first information includes the first list.

[0219] In other words, network devices can configure multiple uncertainty thresholds for terminal devices, so that the terminal devices can determine the value of N based on these multiple uncertainty thresholds.

[0220] Optionally, the second list can be predefined by the protocol or configured by the network-side terminal device, and the second list can also be included in the first information.

[0221] It should be understood that when the confidence level is determined by the AI ​​model, the above scheme can be adjusted accordingly. For example, when the confidence level is less than or equal to the confidence threshold, N takes the first value; when the confidence level is greater than the confidence threshold, N takes the second value. Similarly, "equal to" can be placed in the "greater than" branch. The confidence threshold can be predefined or configured by the network side to the terminal. For another example, when the confidence level is less than or equal to the (i-1)th value in the descending confidence threshold list, and greater than the ith value in the ascending confidence threshold list, N takes the value N. i Other cases follow the same principle and will not be elaborated further.

[0222] In this application, the AI ​​model can be understood as a function with a specific structure, which can be replaced by an AI network. This AI model can be understood as an AI model used for beam prediction, and therefore can be called a beam prediction AI model. This AI model has corresponding inputs and outputs. Its inputs include the measurement results of Q first signals, and its outputs include beam prediction information. Specifically, the type of this AI model can be the first type of AI model mentioned above, or it can be the second type of AI model or other forms, without limitation.

[0223] For example, the AI ​​model may include one or more of the following: fully connected layers, convolutional neural network layers, transformer modules, etc. Each layer or module may also include an activation function as the output of that layer or module, or it may be directly output.

[0224] Optionally, method 300 further includes: S302, the first network element sends fifth information to the terminal device, and correspondingly, the terminal device receives the fifth information, wherein the fifth information includes information about the AI ​​model.

[0225] Specifically, the information of the AI ​​model may include information about the aforementioned structure and the parameters within that structure. For example, specific parameters may include the weights and biases of fully connected layers, and the convolutional kernels and biases of convolutional neural network layers. For instance, when the first network element is an access network device, this fifth information can be carried in DCI, MAC CE, or RRC signaling, without limitation. When the first network element is a core network device, this fifth information can be carried in the signaling between the core network device and the terminal device.

[0226] It should be understood that the first and fifth information can be sent through the same message, that is, S301 and S302 can be executed simultaneously, or the first and fifth information can be sent through different messages, that is, S301 and S302 can be executed sequentially. The order of execution is not limited in this application.

[0227] It should be understood that in this application, the first network element can train an AI model and send the information of the AI ​​model to the terminal device for use, that is, the AI ​​model is deployed on the terminal device side. The structure of the AI ​​model can be constructed by the first network element or defined by a protocol, and is not limited thereto.

[0228] For example, the following are examples of possible beam prediction information, which may include any one of the following situations or a combination of the following situations.

[0229] Scenario 1: Beam prediction information includes one or more beams from the access network device. Specifically, the output of the AI ​​model may include one or more beams from the access network device. These beams can also be referred to as base station-side beams, access network device-side beams, or beams on the access network device side. Optionally, beams can be represented by beam indices; therefore, in Scenario 1, the output of the AI ​​model can be one or more beam indices from the access network device.

[0230] For example, one or more beams of an access network device can be an index (or identifier) ​​of one or more beams, wherein the index can be a number, which can be an integer or can be converted (e.g., by rounding) to an integer. Alternatively, the index can be an array.

[0231] For example, as shown in Figure 6, when the access network device uses a discrete Fourier transform beamcodebook or a beamcodebook based on physical angles, it can have A1 beams in the horizontal direction and A2 beams in the vertical direction, where A1 and A2 are both positive integers. Each circle in Figure 6 represents a beam, therefore, the access network device has a total of A1*A2 beams. In this case, a beam of the access network device in the AI ​​model output can be a single number, with the value ranging from 1 to A1*A2 (or 0 to A1*A2-1). Each integer indicates a beam of the access network device, and the mapping relationship from the integer to the beam can be mapped in either a horizontal-then-vertical order or a vertical-then-horizontal order. Alternatively, a beam of the access network device in the output of the AI ​​model can be a binary array. The first number in the binary array represents the horizontal beam, with a value range of 1 to A1 (or 0 to A1-1), and the second number in the binary array represents the vertical beam, with a value range of 1 to A2 (or 0 to A2-1). Or, the second number in the binary array represents the horizontal beam, with a value range of 1 to A1 (or 0 to A1-1), and the first number in the binary array represents the vertical beam, with a value range of 1 to A2 (or 0 to A2-1). The binary array composed of these two numbers together indicates a beam of the access network device.

[0232] Scenario 2: The beam prediction information includes one or more beam combinations, each of which includes a beam from the access network device and a beam from the terminal device. Specifically, the output of the AI ​​model may include one or more beam combinations, each including a beam from the access network device and a beam from the terminal device.

[0233] In this application, the beam of the terminal device can also be referred to as the terminal-side beam or the beam on the terminal device side. Optionally, the beam can be represented by a beam index. Therefore, in case 2, the output of the AI ​​model can be one or more beam indices of the access network device and one or more beam indices of the terminal device, wherein the beam indices of the access network device and the beam indices of the terminal device are related.

[0234] For example, one or more beams of the access network device can be indices of one or more beams, specifically represented as in Case 1. For example, one or more beams of the terminal device can be indices of one or more beams, specifically represented in a similar manner to Case 1. That is, the index of a beam of the terminal device can be a number, which can be an integer, or can be converted (e.g., by rounding) to an integer. Alternatively, the index can be an array.

[0235] For example, as shown in Figure 7, when the terminal device uses a discrete Fourier transform beamcodebook or a beamcodebook based on physical angles, it can have B1 beams in the horizontal direction and B2 beams in the vertical direction, where B1 and B2 are both positive integers. Each circle in Figure 7 represents a beam, therefore, the terminal device has a total of B1*B2 beams. In this case, a beam of the terminal device in the AI ​​model output can be a single number, with the value ranging from 1 to B1*B2 (or 0 to B1*B2-1). Each integer indicates a beam of the terminal device, and the mapping relationship from the integer to the beam can be mapped in either the horizontal-then-vertical order or the vertical-then-horizontal order. Alternatively, a beam of the terminal device in the output of the AI ​​model can be a binary array. The first number in the binary array represents the horizontal beam, with a value range of 1 to B1 (or 0 to B1-1), and the second number in the binary array represents the vertical beam, with a value range of 1 to B2 (or 0 to B2-1). Or, the second number in the binary array represents the horizontal beam, with a value range of 1 to B1 (or 0 to B1-1), and the first number in the binary array represents the vertical beam, with a value range of 1 to B2 (or 0 to B2-1). The binary array composed of these two numbers together indicates a beam of the terminal device.

[0236] Optionally, each beam of the terminal device is associated with angle information. The first configuration information may include the association between each beam and angle information of the terminal device. For example, the angle information can be an angle in a global coordinate system or an angle in a local coordinate system. Optionally, the beam of the terminal device in the output of the AI ​​model can directly indicate the angle information of the terminal device's beam. For example, the beam of the terminal device in the output of the AI ​​model is an array of length 2, where one number indicates the angle of the horizontal beam of the terminal device, and the other number indicates the angle of the vertical beam of the terminal device.

[0237] Optionally, the beams of both the access network device and the terminal device in the beam combination can be indicated by beam indices. The type of the beam index for the access network device and the type of the beam index for the terminal device can be the same or different, without restriction. For example, if both the beam index of the access network device and the beam index of the terminal device are integers, then the beam prediction information can be represented by a binary array, where one number represents the beam index of the access network device and the other number represents the beam index of the terminal device. Alternatively, if both the beam index of the access network device and the beam index of the terminal device are binary arrays, then the beam prediction information can be represented by a quaternion array, where two numbers represent the beam index of the access network device and the other two numbers represent the beam index of the terminal device. Another example is where one index of the access network device and the beam index of the terminal device is an integer and the other index is a binary array.

[0238] Optionally, the beam of the access network device in the beam combination is indicated by the beam index, and the beam of the terminal device is indicated by the beam angle. The beam index of the access network device can be an integer or a binary array. For example, the beam prediction information can be represented by a quaternion array, where two numbers represent the beam index of the access network device, and the other two numbers represent the horizontal and vertical angles corresponding to the beam of the terminal device, respectively.

[0239] Scenario 3: Beam prediction information includes one or more beam pairs, each beam pair associated with a beam from an access network device and a beam from a terminal device. Specifically, the output of the AI ​​model may include one or more beam pairs, each beam pair associated with a beam from an access network device and a beam from a terminal device.

[0240] In one implementation, a beam pair consists of a transmit beam and a receive beam. When the two beams form a beam pair, the transmitter can use the transmit beam in the beam pair to send information #1 to the receiver, and the receiver can use the receive beam in the beam pair to receive information #1 from the transmitter. That is, the transmitter and receiver can use the beam pair to communicate.

[0241] For example, the output of the AI ​​model can be one or more beam pair indices, each beam pair index associating a beam from an access network device and a beam from a terminal device. This association can be mapped according to the beam priority order of the access network devices or the beam priority order of the terminal devices.

[0242] For example, as shown in Figure 6, the access network device has a total of A1*A2 beams, and as shown in Figure 7, the terminal device has a total of B1*B2 beams. Therefore, the beams of the access network device and the beams of the terminal device can form A1*A2*B1*B2 beam pairs. The index of each beam pair can be represented by z, and the value of z is an integer between 1 and A1*A2*B1*B2 (or 0 to A1*A2*B1*B2-1). Each integer is used to indicate one beam of the access network device and one beam of the terminal device. The mapping relationship from the integer to the beam pair can be mapped in the order of access network device beam first and terminal device beam second, or in the order of terminal device beam first and access network device beam second.

[0243] The difference between Case 2 and Case 3 is that in Case 2, the AI ​​model directly outputs the beam index of the access network device and the beam index or angle of the terminal device, while in Case 3, the AI ​​model outputs the index of the beam pair, which can be further converted into the beam of the access network device and the beam of the terminal device based on the index of the beam pair.

[0244] Scenario 4: The beam prediction information includes the first signal strength tensor of multiple beams of the access network device, where each position in the first signal strength tensor is associated with one beam of the access network device. Specifically, the output of the AI ​​model may include the first signal strength tensor of multiple beams of the access network device, where each position in the first signal strength tensor is associated with one beam of the access network device.

[0245] For example, each position in the first signal strength tensor corresponds to a beam of the access network device. Each element in the tensor represents the signal strength of the beam of the access network device corresponding to that position, such as the beam's RSRP value, the normalized RSRP value, the probability that the beam is the optimal beam, or a recommended value for the beam. Therefore, the number of elements in the first signal strength tensor can represent the number of beams of the access network device. Optionally, the first signal strength tensor can be a one-dimensional tensor, and the length of this one-dimensional tensor can represent the number of beams of the access network device. For example, as shown in Figure 6, the access network device has a total of A1*A2 beams, and the first signal strength tensor can have A1*A2 elements with a length of A1*A2. Optionally, the first signal strength tensor can be a two-dimensional tensor, and the shape of this two-dimensional tensor can represent the number of beams of the access network device. For example, as shown in Figure 6, the access network device has a total of A1*A2 beams. The elements in the first signal strength tensor can be A1*A2, and its shape can be [A1]*[A2]. The i-th row and j-th column in this two-dimensional tensor represent the signal strength of the beam of the access network device with beam index (i, j), where i and j are positive integers.

[0246] Scenario 5: The beam prediction information includes a second signal strength tensor for multiple beam pairs, where each location in the second signal strength tensor is associated with a beam from the access network device and a beam from the terminal device. Specifically, the output of the AI ​​model may include a second signal strength tensor for multiple beam pairs, where each location in the second signal strength tensor is associated with a beam pair, and each beam pair includes a beam from the access network device and a beam from the terminal device.

[0247] For example, each position in the second signal strength tensor corresponds to a beam pair, and each element in the tensor represents the signal strength of the beam pair corresponding to that position. This can identify the signal strength (or normalized strength) that the terminal device can receive when the access network device and the terminal device use this beam pair to transmit a signal. Examples include the RSRP value of the beam pair, the normalized RSRP value of the beam pair, the probability that the beam is the optimal beam, or a recommended value for the beam pair. Therefore, the number of elements in this second signal strength tensor can represent the number of beam pairs.

[0248] Optionally, the first signal strength tensor can be a one-dimensional tensor, the length of which can represent the number of beam pairs. For example, as shown in Figure 6, the access network device has a total of A1*A2 beams, and as shown in Figure 7, the terminal device has a total of B1*B2 beams. Then, the second signal strength tensor can have A1*A2*B1*B2 elements, and its length is A1*A2*B1*B2. In this case, the position of a certain element in the one-dimensional tensor can be understood as the beam pair index in Case 3. Therefore, the method of mapping the position of a certain element to the beams of the specific access network device and the terminal device can refer to the method of mapping the beam pair index to the beams of the access network device and the terminal device in Case 3.

[0249] Optionally, the first signal strength tensor can be a two-dimensional tensor, where the size of one dimension is equal to the number of beams in the access network device, and the size of the other dimension is equal to the number of beams in the terminal device. For example, as shown in Figure 6, the access network device has a total of A1*A2 beams, and as shown in Figure 7, the terminal device has a total of B1*B2 beams. Then, the second signal strength tensor can have A1*A2*B1*B2 elements, and its shape can be [C1]*[C2], where C1 = A1*A2 and C2 = B1*B2. In this case, the position of each element in the second signal strength tensor can be represented by (c1, c2), where c1 is an integer from 1 to C1 or an integer from 0 to C1-1, and c2 is an integer from 1 to C2 or an integer from 0 to C2-1. At this point, c1 can be understood as the beam index of the access network device, and c2 can be understood as the beam index of the terminal device. The element in the second signal strength tensor located at (c1, c2) can be understood as the signal strength (or normalized strength) that the access network device can receive when transmitting a signal using the beam of the access network device side with beam index c1, and the terminal device can receive a signal using the beam of the terminal device side with beam index c2. Furthermore, c1 can be associated with the horizontal and vertical beam indices of the access network device side, as detailed in Case 1, and c2 can be associated with the horizontal and vertical beam indices of the terminal device side, as detailed in Case 2.

[0250] Similarly, the first signal strength tensor can be a three-dimensional tensor. One dimension of this three-dimensional tensor has a size equal to the number of horizontal beams of the access network device, another dimension equals the number of vertical beams of the access network device, and yet another dimension equals the number of beams of the terminal device. For example, as shown in Figure 6, the access network device has a total of A1*A2 beams, and as shown in Figure 7, the terminal device has a total of B1*B2 beams. Therefore, the second signal strength tensor can have A1*A2*B1*B2 elements, and its shape can be [A1]*[A2]*[C2], where C2 = B1*B2. The position of each element in the second signal strength tensor can then be represented by (a1, a2, c2), where a1 is an integer from 1 to A1 or from 0 to A1-1, a2 is an integer from 1 to A2 or from 0 to A2-1, and c2 is an integer from 1 to C2 or from 0 to C2-1. At this point, a1 can be understood as the horizontal beam index of the access network device, a2 as the vertical beam index of the access network device, and c2 as the beam index on the terminal device side. The element in the second signal strength tensor located at (a1, a2, c2) can be understood as the signal strength (or normalized strength) that the access network device can receive when transmitting signals using the access network device side beam with horizontal beam index a1 and vertical beam index a2, and the terminal device can receive signals using the terminal device side beam with beam index c2. Furthermore, c2 can be associated with the horizontal and vertical beam indices on the terminal device side; for details, please refer to Case 2.

[0251] Alternatively, the first signal strength tensor can be a three-dimensional tensor. One dimension of this three-dimensional tensor has a size equal to the number of beams in the access network device, another dimension equals the number of horizontal beams in the terminal device, and yet another dimension equals the number of vertical beams in the terminal device. For example, as shown in Figure 6, the access network device has a total of A1*A2 beams, and as shown in Figure 7, the terminal device has a total of B1*B2 beams. Therefore, the second signal strength tensor can have A1*A2*B1*B2 elements, and its shape can be [C1]*[B1]*[B2], where C1 = A1*A2. In this case, the position of each element in the second signal strength tensor can be represented by (c1, b1, b2), where c1 is an integer from 1 to C1 or from 0 to C1-1, b1 is an integer from 1 to B1 or from 0 to B1-1, and b2 is an integer from 1 to B2 or from 0 to B2-1. At this point, b1 can be understood as the horizontal beam index of the terminal device, b2 as the vertical beam index of the terminal device, and c1 as the beam index on the access network device side. The element in the second signal strength tensor located at (c1, b1, b2) can be understood as the signal strength (or normalized strength) that the access network device can receive when transmitting signals using the access network device side beam with beam index c1, and the terminal device can receive signals using the terminal device side beam with horizontal beam index b1 and vertical beam index b2. Furthermore, c1 can be associated with the horizontal and vertical beam indices on the access network device side; for details, please refer to Case 2.

[0252] Similarly, the first signal strength tensor can be a four-dimensional tensor, where the beams of the access network device and the terminal device are both represented by binary arrays. Specifically, the size of one dimension of this four-dimensional tensor is equal to the number of horizontal beams of the access network device, the size of another dimension of this four-dimensional tensor is equal to the number of horizontal beams of the terminal device, the size of yet another dimension of this four-dimensional tensor is equal to the number of vertical beams of the access network device, and the size of yet another dimension of this four-dimensional tensor is equal to the number of vertical beams of the terminal device. For example, similar to the examples in Figures 6 and 7 above, its shape can be [A1]*[A2]*[B1]*[B2].

[0253] Optionally, the signal strength tensor output in Case 4 can be understood as a probability tensor that each beam is the optimal beam, or a tensor of recommended values ​​for each beam; the signal strength tensor output in Case 5 can be understood as a probability tensor that each beam pair is the optimal beam pair, or a tensor of recommended values ​​for each beam pair. When predicting beams only on the access network equipment side, the form of the first beam can adopt the above-described Case 1 and Case 4. When predicting beams on both the access network equipment side and the terminal equipment side simultaneously, the form of the first beam can adopt the above-described Case 2, Case 3, and Case 5.

[0254] For example, after determining N values, the terminal device can select N beams from the beam prediction information, such as selecting the N beams with the best prediction quality (e.g., the highest prediction intensity), or using other methods to select them, which is not limited in this application.

[0255] Optionally, as an alternative to method 320, when the output of the AI ​​model is the signal strength tensor of a beam or beam pair (or the probability that each beam or beam pair is the optimal beam or beam pair, the recommended value for each beam or beam pair, etc.), the terminal device can also determine N beams based on a first strength threshold. That is, in this alternative, the value of N can be determined based on the first strength threshold, rather than on the prediction uncertainty.

[0256] For example, the terminal device determines the intensity prediction values ​​of each beam or beam based on the output of the AI ​​model. It can report all beams whose intensity prediction values ​​are greater than a first intensity threshold. Alternatively, the N beams determined by the terminal device need to satisfy the following conditions: the sum of the intensity prediction values ​​of these N beams is greater than the first intensity threshold; or the sum of the intensity prediction values ​​of the corresponding beam pairs of these N beams is greater than the first intensity threshold. For instance, the terminal device can sort the intensity prediction values ​​of the beams or beam pairs output by the AI ​​model, and then select beams or beam pairs from high to low intensity until the sum of the intensity prediction values ​​of the selected N beams is greater than the first intensity threshold. This method can achieve a similar effect to the scheme of determining N beams based on prediction uncertainty. For example, when the predicted intensity of a certain beam is particularly high, it can indicate that the signal quality of that beam is particularly good, and it can also indirectly reflect that the prediction accuracy is relatively high. In this case, only these high-intensity beams need to be reported.

[0257] In this application, when a terminal device performs a step such as "determine based on the AI ​​model...", it typically needs to perform some related inference calculations. The specific implementation of these inference calculations can be performed locally on the terminal device, for example, by using the terminal device's own chip, CPU, graphics processing unit (GPU), or other hardware modules for calculation; or the specific implementation of these inference calculations can be that the terminal device sends the corresponding computational inference task to external devices such as terminal cloud, server, base station, or nodes or network elements with computing capabilities in the core network, and the external devices perform the calculations and send the inference results back to the terminal device; or it can be a combination of both, such as the terminal performing part of the calculation locally and the external devices performing part of the calculation. This application does not impose any restrictions on the specific inference calculation process of the AI ​​model.

[0258] S330: The terminal device sends N beam indication information to the access network device.

[0259] Specifically, the terminal device can report the identified N beams, that is, send indication information for the N beams.

[0260] Based on the above scheme, the terminal device can determine the number of beams to be reported according to the prediction uncertainty. This allows the number of beams to be reported to be flexibly changed, avoiding the direct reporting of multiple beams output by the AI ​​model and improving the efficiency of beam management.

[0261] On the other hand, when the prediction uncertainty is low, the terminal device can report a smaller number of beams, avoiding unnecessary measurements of a large number of beams by the terminal device and network device, thus saving signaling overhead.

[0262] Optionally, after S330, method 300 further includes: S340, where the terminal device and the access network device measure the N beams and determine the beam used for communication based on the measurement results of the N beams.

[0263] Specifically, if N is a number greater than 1, the terminal equipment and access network equipment can measure these N beams to determine the optimal communication beam.

[0264] Figure 8 is a schematic flowchart of a communication method provided in this application. As shown in Figure 8, the method 700 includes the following steps.

[0265] S710, the access network device sends Q first signals to the terminal device, and correspondingly, the terminal device measures Q first signals, where Q is a positive integer.

[0266] For details on the S710, please refer to the S310; it will not be elaborated upon here.

[0267] In S720, the terminal device determines N beams based on the measurement results of Q first signals, the AI ​​model, and the interval threshold, where N is a positive integer.

[0268] The N beams include a first beam and a second beam, with the spatial interval between the first and second beams being greater than or equal to an interval threshold. The AI ​​model's input includes the measurement results of Q first signals, and its output includes beam prediction information used to determine the N beams. In other words, the terminal device can select the N beams to be reported from the beam prediction information based on the interval threshold. The value of N can be determined according to the relevant method in S300, or it can be determined by other methods (e.g., configured as a fixed value). "The spatial interval between the first and second beams is greater than or equal to the interval threshold" can also be replaced with "The spatial interval between the first and second beams is greater than the interval threshold." The following explanation primarily uses the example of "greater than or equal to."

[0269] In this application, the spatial spacing between two beams can also be referred to as the distance between two beams or the spatial positional relationship between two beams, which indicates the spatial correlation between the two beams.

[0270] Similarly, the interval threshold can also be called the distance threshold, the positional relationship threshold, etc., without any restrictions.

[0271] The N beams are determined based on the interval threshold. This can be understood as referring to which N beams are determined according to the interval threshold, or the specific content of the N beams is determined according to the interval threshold.

[0272] As one implementation, the interval threshold includes a first interval value, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold, including: the spatial interval between the first beam and the second beam is greater than or equal to the first interval value.

[0273] Specifically, the spatial spacing between the first and second beams can be determined based on the horizontal spacing, the vertical spacing, or a combination of both. For example, the spatial spacing between the first and second beams can be: the horizontal spacing, or the vertical spacing, or the maximum value of the horizontal and vertical spacing, or the minimum value of the horizontal and vertical spacing, or the sum of the horizontal and vertical spacing, or the square root of the sum of the squares of the horizontal and vertical spacing, etc. In other words, the spatial spacing between the first and second beams being greater than or equal to the first spacing value can mean: the horizontal spacing between the first and second beams is greater than or equal to the first spacing value; or the vertical spacing between the first and second beams is greater than or equal to the first spacing value; or the maximum value of the horizontal and vertical spacing is greater than the first spacing value; or the minimum value of the horizontal and vertical spacing is greater than the first spacing value; or the sum of the horizontal and vertical spacing is greater than the first spacing value; or the square root of the sum of the squares of the horizontal and vertical spacing is greater than the first spacing value, etc.

[0274] Specifically, the horizontal spacing between beams (i1, j1) and (i2, j2) can be represented as h(i1, i2), and the vertical spacing can be represented as v(j1, j2). The horizontal spacing h(i1, i2) can be |i1-i2|, or it can be the distance between the two beams on the ring. The vertical spacing is similar.

[0275] For example, as shown in Figure 9, there are a total of O*P beams for the horizontal beam, where O is the oversampling factor of the discrete Fourier transform (DFT) beam and P is the number of antennas in the horizontal direction. These O*P beams are distributed on the ring, with indices of 0, 1, 2, ..., O*P-2, O*P-1, as shown in Figure 9. On the ring, the distance between beam 0 and beam O*P-1 is 1 instead of O*P-1, so h(i1,i2) can also be min(|i1-i2|, O*P-|i1-i2|).

[0276] Figure 10 is a schematic diagram of the spacing between two beams. As shown in Figure 10, each small square represents a beam. Assuming there are a total of 8*8=64 beams on the access network side, the horizontal axis represents the horizontal left index of the beam, and the vertical axis represents the vertical index of the beam. The index of beam 1 is (1,6), and the index of beam 2 is (4,2). If the distance between two beams is defined as the maximum value between the horizontal distance and the vertical distance of the two beams, then the distance between the shaded beams around beam 1 and beam 1 is 1. Similarly, the distance between the shaded beams around beam 2 and beam 2 is also 1.

[0277] In other words, when the interval threshold includes only one threshold, the terminal device can make the distance between the first and second beams in the N beams greater than the threshold when determining the N beams, thereby reducing the spatial correlation between the N beams.

[0278] As another implementation, the spacing threshold includes a first spacing value and a second spacing value. The spatial spacing between the first beam and the second beam is greater than or equal to the spacing threshold, including: the horizontal spatial spacing between the first beam and the second beam is greater than or equal to the first spacing value; and / or, the vertical spatial spacing between the first beam and the second beam is greater than or equal to the second spacing value. Similarly, "the horizontal spatial spacing between the first beam and the second beam is greater than or equal to the first spacing value" can also be replaced with "the horizontal spatial spacing between the first beam and the second beam is greater than the first spacing value," and "the vertical spatial spacing between the first beam and the second beam is greater than or equal to the second spacing value" can also be replaced with "the vertical spatial spacing between the first beam and the second beam is greater than the second spacing value." The following mainly uses examples where both are "greater than or equal to" for illustration.

[0279] In other words, when the interval threshold includes two thresholds, when determining N beams, the terminal device can ensure that the horizontal distance between the first beam and the second beam in the N beams is greater than one of the thresholds, and that the vertical distance between the first beam and the second beam is greater than the other threshold; or, it only needs to ensure that the horizontal distance between the first beam and the second beam is greater than one of the thresholds; or, it only needs to ensure that the vertical distance between the first beam and the second beam is greater than one of the thresholds, thereby reducing the spatial correlation between the N beams.

[0280] The interval threshold can be configured by the network side to the terminal device or predefined by the protocol, and is not limited. For example, the method 700 further includes: S701, the first network element sends second information to the terminal device, and the terminal device receives the second information accordingly, wherein the second information includes the aforementioned interval threshold.

[0281] For example, when the first network element is an access network device, the second information can be carried in DCI, MAC CE, or RRC signaling, without restriction. When the first network element is a core network device, the second information can be carried in the signaling between the core network device and the terminal device.

[0282] In another implementation, the N beams include a first beam and a second beam, which have spatial correlation (or beam correlation) with different SSBs. That is, the spacing between two beams can be represented by their spatial relationship with the SSBs.

[0283] Specifically, spatial correlation between beam A and SSB A indicates that beam A and the beam used by the access network device to transmit SSB A are the same or similar beams. For example, in a 5G NR system, this spatial correlation can be represented by a type D quasi-colocation (QCL) relationship. For instance, if beam A is represented by the NZP-CSI-RS A resource, and NZP-CSI-RS A and SSB A have a type D quasi-colocation relationship, it means that the access network device will use the same or similar beams to transmit NZP-CSI-RS A and SSB A. In actual networks, access network devices generally use wider beams to transmit SSBs. Therefore, when the first beam and the second beam have spatial correlation (or beam correlation) with different SSBs, it indicates that there is a certain beam spacing or spatial spacing between the first beam and the second beam, achieving the same effect as the above scheme.

[0284] For details regarding the AI ​​model and the beam prediction information output by the AI ​​mode, please refer to Method 300; it will not be elaborated upon here.

[0285] It should be understood that in method 700, the beam prediction information may include information on M beams. The information on the M beams is used to determine the beams used for communication between the terminal device and the access network device. The M beams include N beams, where M is a positive integer. That is, N is an integer less than or equal to M. The terminal device can select N beams from the M beams based on an interval threshold.

[0286] Furthermore, similar to S301 in method 300, method 700 may include S702, in which the first network element sends fifth information to the terminal device, and the terminal device receives the fifth information, wherein the fifth information includes information about the AI ​​model.

[0287] In one implementation scenario, the first beam is any one of the N beams, and the second beam is any one of the N beams that is different from the first beam. Specifically, the terminal device can sequentially select N beams such that the spatial interval between the N beams is greater than or equal to the interval threshold. That is, in this scenario, the terminal device can determine the N beams by default based on the interval threshold, and the spatial interval between any two beams in the N beams is greater than or equal to the interval threshold. For example, taking Figure 10 as an example, the interval threshold is 2. Assuming that beam 1 is the beam with the highest predicted intensity, beam 1 is first determined as the beam that needs to be reported. According to the previous example, the interval between the shaded beams around beam 1 and beam 1 is 1. Therefore, the interval between these beams and beam 1 is less than the threshold, so these beams can be removed. From the remaining beams, the beam with the highest predicted intensity is selected, for example, beam 2. Then beam 2 is also the beam that needs to be reported (naturally, the interval between beam 2 and beam 1 is also greater than or equal to the interval threshold). The beam spacing is equal to the interval threshold. When the beam spacing is defined as the maximum value of the horizontal and vertical intervals, the actual interval between the two beams in the figure is 3. Similarly, since the interval between the beams marked with shade around beam 2 and beam 2 is 1, the interval between this part of the beams and beam 2 is also less than the threshold. Therefore, this part of the beams can also be removed. Then, the largest beam is selected from the remaining beams (the beams removed in the previous step also need to be removed), for example, beam 3 (not shown in the figure). This process is repeated until N beams are selected.

[0288] As another implementation scenario, the terminal device can determine N beams based on the interval threshold when the prediction uncertainty is greater than or equal to the second uncertainty threshold. The phrase "when the prediction uncertainty is greater than or equal to the second uncertainty threshold" can also be replaced with "when the prediction uncertainty is greater than the second uncertainty threshold." The following explanation primarily uses "greater than or equal to" as an example.

[0289] Optionally, in one implementation, in S720, the terminal device determines N beams based on the measurement results of Q first signals, the AI ​​model, and the interval threshold, including: determining N beams based on the interval threshold when the prediction uncertainty is greater than or equal to a second uncertainty threshold, wherein the prediction uncertainty is determined based on the AI ​​model and the measurement results of Q first signals.

[0290] Specifically, the terminal device can determine N beams based on an interval threshold when the prediction uncertainty is greater than or equal to the second uncertainty threshold. When the prediction uncertainty is less than the second uncertainty threshold, N beams are not determined based on the interval threshold. For example, the N beams are the N best beams in the beam prediction information (i.e., the Top N beams), and they do not need to meet the condition of the interval threshold. That is, among the N beams actually selected, the beam interval between any two beams may be greater than or greater than the interval threshold, or it may be less than the interval threshold, without specific restrictions.

[0291] The prediction uncertainty can be output by the AI ​​model or determined by the terminal device itself, without restriction.

[0292] Alternatively, as an alternative to this implementation scenario, the aforementioned prediction uncertainty can be replaced with confidence level. That is, the terminal device can determine N beams based on the interval threshold if the prediction confidence level is less than or equal to the confidence level threshold.

[0293] For the meaning of prediction uncertainty, confidence level, etc., and the method for terminal equipment to obtain prediction uncertainty, please refer to Method 300.

[0294] Based on the above scheme, when the prediction uncertainty is greater than or equal to the second uncertainty threshold, it indicates that the accuracy of the prediction information output by the AI ​​model is low, that is, the certainty of the prediction is not high. In this case, it is very likely that the selected optimal beam is far from the actual optimal beam. Therefore, the terminal device can determine the reported beam according to the interval threshold, which can reduce the spatial correlation between the reported beams, that is, reduce the correlation between the reported beams, thereby improving the robustness of the reported beams.

[0295] As another implementation scenario, the signal strengths of both the first beam and the second beam are greater than the first strength threshold.

[0296] The signal strength of a beam typically refers to the intensity of the received radio waves in a specific beam direction. It reflects the strength of the electromagnetic wave reaching the receiver as it propagates through space, representing signal power and / or signal energy. Beam signal strength is usually expressed in decibel-milliwatts (dBm). dBm is a unit representing a relative power value, calculated with 1 milliwatt as a reference level. For example, if the signal strength of a beam is -60 dBm, it means that the signal power of that beam is 60 dB less than 1 milliwatt. The first strength threshold can be a numerical value representing the signal strength of a beam. The statement "The signal strengths of both the first and second beams are greater than the first strength threshold" can also be replaced with "The signal strengths of both the first and second beams are greater than or equal to the first strength threshold." The following explanation primarily uses "greater than" as an example.

[0297] Specifically, the terminal device can sequentially select N beams so that the signal strength of the N beams is greater than a first strength threshold.

[0298] In this method, the terminal device can first determine N beams based on an interval threshold. If the strength of all N beams is greater than a first strength threshold, the determination of N beams is complete. If the strength of some of the N beams is less than or equal to the first strength threshold, the beams with strength less than or equal to the first strength threshold can be removed. In one implementation method, the terminal can subsequently report only the beams with strength greater than the first strength threshold (the number is less than N). In another implementation method, the terminal device can further select several beams with the highest strength from the removed beams to make up the N beams, thereby completing the determination of N beams. In this implementation method, if the number of beams with signal strength greater than the first strength threshold among the M beams is less than N, for example, K, the terminal device can subsequently report only these K beams.

[0299] Alternatively, if the prediction uncertainty is greater than or equal to the second uncertainty threshold, the terminal device can first determine N beams based on the interval threshold. If the strength of all N beams is greater than the first strength threshold, the determination of N beams is complete. If the strength of some of the N beams is less than or equal to the first strength threshold, the beams with strength less than or equal to the first strength threshold can be removed. In one implementation, the terminal can subsequently report only the beams with strength greater than the first strength threshold. In another implementation, the terminal device can further select several beams with the highest strength from the removed beams to complete the determination of N beams. In this implementation, if the number of beams with signal strength greater than the first strength threshold among the M beams is less than N, for example, K, the terminal device can subsequently report only these K beams.

[0300] It should be understood that in this implementation, the signal strength of each of the N beams can be greater than or equal to the first strength threshold, or only some of the N beams can be greater than the first strength threshold.

[0301] The first strength threshold can be predefined by the protocol or configured by the network device to the terminal device, and is not limited thereto. For example, in this implementation scenario, the method further includes: S704, the first network element sends fourth information to the terminal device, and correspondingly, the terminal device receives the fourth information, wherein the fourth information includes the first strength threshold.

[0302] For example, when the first network element is an access network device, the fourth information can be carried in DCI, MAC CE, or RRC signaling, without restriction. When the first network element is a core network device, the fourth information can be carried in the signaling between the core network device and the terminal device.

[0303] It should be understood that the second, third, fourth, and fifth information can be sent through the same message, that is, S701, S702, S703, and S704 can be executed simultaneously. Alternatively, the second, third, fourth, and fifth information can also be sent through different messages, that is, S701, S702, S703, and S704 can be executed sequentially. The order in which they are executed is not restricted by this application.

[0304] S730: The terminal device sends N beam indication information to the access network device.

[0305] Specifically, the terminal device can report its determined N beams, that is, send indication information for N beams.

[0306] Based on the above scheme, the terminal device can determine the beams that need to be reported according to the interval threshold. This allows for a larger spatial distance between the reported beams, resulting in lower correlation of beam quality and effectively reducing or avoiding the occurrence of multiple reported beams having poor beam quality.

[0307] Optionally, after S730, method 700 further includes: S740, whereby the terminal equipment and the access network equipment measure the N beams and determine the beam used for communication based on the measurement results of the N beams.

[0308] Specifically, the terminal equipment and access network equipment can measure these N beams to determine the quality of these N beams and the optimal communication beam.

[0309] It should be understood that methods 300 and 700 described above can be implemented individually or in combination. For example, in method 300, after determining the value of N based on the prediction uncertainty threshold, a corresponding number of beams can be selected from the M beams in the beam prediction information based on the interval threshold in method 700, i.e., N beams can be selected. The N beams include the first beam and the second beam, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold. As another example, in method 300, after determining the value of N based on the prediction uncertainty threshold, it can be further determined based on the second uncertainty threshold in method 700 whether a corresponding number of beams needs to be selected from the M beams in the beam prediction information based on the interval threshold. For example, if the prediction uncertainty is greater than or equal to the second uncertainty threshold, N beams can be determined according to the interval threshold. In this case, the N beams include the first beam and the second beam, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold.

[0310] The communication method provided in the embodiments of this application has been described in detail above with reference to Figures 1 to 10. The above-described communication method is mainly introduced from the perspective of interaction between terminal devices and network devices. It is understood that, in order to achieve the above functions, the terminal devices and network devices include hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, based on the units and method steps of the various examples described in conjunction with the embodiments disclosed in this application, this application can be implemented in hardware or a combination of hardware and computer software. Whether a certain function is executed by hardware or by computer software driving hardware depends on the specific application scenario and design constraints of the technical solution.

[0311] Figures 11 and 12 are schematic block diagrams of communication devices provided in embodiments of this application. These communication devices can be used to implement the functions of the first terminal device or network device in the above method embodiments, and thus can also achieve the beneficial effects of the above method embodiments. In the embodiments of this application, the communication device can be the UE shown in Figure 1, the RAN shown in Figure 1, the first network element shown in Figure 1, or a module (such as a chip) applied to the UE, RAN, or the first network element.

[0312] As shown in Figure 11, the communication device 2000 includes a transceiver unit 2020 and a processing unit 2010, used to implement the functions of the first network element, terminal device, or access network device in the method embodiments shown in Figure 3 or Figure 8; wherein, the transceiver unit is used to implement the transceiver operations in the embodiments, and the processing unit is used to implement other operations besides transceiver, such as:

[0313] When the communication device 2000 is used to implement the functions of the terminal device (or UE) in the method embodiment shown in FIG3: the transceiver unit 2020 is used to: receive Q first signals from the access network device, where Q is a positive integer; the processing unit 2010 is used to: determine N beams based on the measurement results of the Q first signals and the AI ​​model, where the value of N is determined based on the prediction uncertainty, which is determined based on the AI ​​model and the measurement results of the Q first signals, wherein the input of the AI ​​model includes the measurement results of the Q first signals, and N is a positive integer; the transceiver unit 2020 is also used to: send indication information of the N beams to the access network device.

[0314] When the communication device 2000 is used to implement the function of the access network device in the method embodiment shown in FIG3: the transceiver unit 2020 is used to: send Q first signals to the terminal device, where Q is a positive integer; the transceiver unit 2020 is also used to: receive indication information of N beams from the terminal device. The value of N is determined based on the prediction uncertainty, which is determined based on the AI ​​model and the measurement results of the Q first signals. The input of the AI ​​model includes the measurement results of the Q first signals, where N is a positive integer.

[0315] When the communication device 2000 is used to implement the function of the first network element in the method embodiment shown in FIG3: the transceiver unit 2020 is used to send one or more of the first information and the fifth information to the terminal device.

[0316] When the communication device 2000 is used to implement the functions of the terminal device (or UE) in the method embodiment shown in FIG8: the transceiver unit 2020 is used to: receive Q first signals from the access network device, where Q is a positive integer; the processing unit 2010 is used to: determine N beams based on the measurement results of the Q first signals, the AI ​​model, and the interval threshold, wherein the N beams include the first beam and the second beam, and the spatial interval between the first beam and the second beam is greater than or equal to the interval threshold, wherein the input of the AI ​​model includes the measurement results of the Q first signals, where N is a positive integer; the transceiver unit 2020 is also used to: send indication information of the N beams to the access network device.

[0317] When the communication device 2000 is used to implement the function of the access network device in the method embodiment shown in FIG8: the transceiver unit 2020 is used to: send Q first signals to the terminal device, where Q is a positive integer; the transceiver unit 2020 is also used to: receive indication information of N beams from the terminal device, wherein the N beams include a first beam and a second beam, the spatial interval between the first beam and the second beam is greater than or equal to an interval threshold, and the input of the AI ​​model includes the measurement results of the Q first signals, where N is a positive integer.

[0318] When the communication device 2000 is used to implement the function of the first network element in the method embodiment shown in FIG8: the transceiver unit 2020 is used to send one or more of the second information, the third information, the fourth information and the fifth information to the terminal device.

[0319] Optionally, the communication device 2000 may further include a storage unit, which can be used to store program code, program instructions and / or data. The processing unit 2010 can read the instructions and / or data in the storage unit so that the communication device 2000 can implement the aforementioned method embodiments.

[0320] Optionally, the transceiver unit 2020 may include a sending unit and a receiving unit. The sending unit is used to implement the sending operation in the above method embodiment, that is, to execute the sending action of the communication device 2000. The receiving unit is used to implement the receiving operation in the above method embodiment, that is, to execute the receiving action of the communication device 2000.

[0321] It should be noted that the communication device 2000 may include a transmitting unit but not a receiving unit. Alternatively, the communication device 2000 may include a receiving unit but not a transmitting unit. Specifically, it depends on whether the above-described scheme executed by the communication device 2000 includes both transmitting and receiving actions. A more detailed description of the processing unit 2010 and the transceiver unit 2020 can be found in the relevant descriptions in the method embodiments shown in Figure 3 or Figure 8.

[0322] The transceiver unit can also be called a transceiver module, which includes a sending module and a receiving module. The storage unit can also be called a storage module.

[0323] Optionally, when the communication device 2000 is a terminal device or a communication module within a terminal device, the processing unit 2010 in the above embodiments can be implemented by at least one processor or processor-related circuitry. Specifically, the processor may include a modem chip, or a SoC chip or SIP chip containing a modem core. The transceiver unit 2020 can be implemented by a transceiver or transceiver-related circuitry. The transceiver unit 2020 may also be referred to as a communication unit or communication interface. The storage unit can be implemented by at least one memory.

[0324] Optionally, when the communication device 200 is a circuit or chip responsible for communication functions in a terminal device, such as a modem chip or a SoC chip or SIP chip containing a modem core, the function of the processing unit 2010 can be implemented by a circuit system in the aforementioned chip that includes one or more processors or processing cores. The function of the transceiver unit 2020 can be implemented by the interface circuit or data transceiver circuit on the aforementioned chip.

[0325] As shown in Figure 12, the communication device 3000 includes a processor 3010 and an interface circuit 3020. The processor 3010 and the interface circuit 3020 are coupled together. It is understood that the interface circuit 3020 can be a transceiver or an input / output interface. Optionally, the communication device 3000 may also include a memory 3030 for storing instructions executed by the processor 3010, or storing input data required by the processor 3010 to execute instructions, or storing data generated after the processor 3010 executes instructions. Sometimes, the interface circuit 3020 can also be understood as part of the processor 3010, in which case the communication device 3000 includes the processor 3010.

[0326] When the communication device 3000 is used to implement the method shown in FIG3 or FIG8, the processor 3010 is used to implement the function of the processing unit 2010, and the interface circuit 3020 is used to implement the function of the transceiver unit 2020. For example, when the interface circuit 3020 is a transceiver, it may include a transmitter and / or a receiver, respectively used to implement the functions of the transmitting unit and the receiving unit. When the interface circuit 3020 is an input / output interface, it may include an output interface and / or an input interface, respectively used to implement the functions of the transmitting unit and the receiving unit.

[0327] When the communication device 3000 is a chip, the chip includes a processor and a transceiver. The processor can be a processing module integrated on the chip, a microprocessor, or an integrated circuit. The transceiver can be an input / output circuit or a communication interface. The sending operation in the above method embodiments can be understood as the chip's output, and the receiving operation in the above method embodiments can be understood as the chip's input. Optionally, when the communication device 3000 is a chip, it may include a memory, such as the memory built into the chip; alternatively, the communication device 3000 may not include a memory, for example, although the chip is connected to a memory, the memory and the chip are independent of each other.

[0328] Furthermore, when the aforementioned communication device is a chip applied to a terminal, the terminal chip implements the functions of the terminal in the above method embodiments. The terminal chip receiving information can be understood as the information being first received by other modules in the terminal (such as an RF module or antenna), and then sent to the terminal chip by these modules. The terminal chip sending information can be understood as the information being first sent to other modules in the terminal (such as an RF module or antenna), and then sent by these modules.

[0329] When the aforementioned communication device is a chip applied to a base station, the base station chip implements the functions of the base station in the above method embodiments. The base station chip receiving information can be understood as the information being first received by other modules in the base station (such as an RF module or antenna), and then sent to the base station chip by these modules. The base station chip sending information can be understood as the information being sent down to other modules in the base station (such as an RF module or antenna), and then sent by these modules.

[0330] When the aforementioned communication device is a chip applied to a first network element, the first network element chip implements the functions of the first network element in the above method embodiments. The first network element chip receiving information can be understood as the information being first received by other modules (such as radio frequency modules or antennas) in the first network element, and then sent to the first network element chip by these modules. The first network element chip sending information can be understood as the information being sent down to other modules (such as radio frequency modules or antennas) in the first network element, and then sent by these modules.

[0331] In this application, entity A sends information to entity B, either directly or indirectly through other entities. Similarly, entity B receives information from entity A, either directly or indirectly through other entities. Entities A and B can be RAN nodes or terminals, or modules within RAN nodes or terminals. Information transmission and reception can be between RAN nodes and terminals, such as between a base station and a terminal; between two RAN nodes, such as between a CU and a DU; or between different modules within a single device, such as between a terminal chip and other modules of the terminal, or between a base station chip and other modules of the base station.

[0332] It is understood that the processor in the embodiments of this application can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0333] The method steps in the embodiments of this application can be implemented in hardware or in software instructions executable by a processor. The software instructions can consist of corresponding software modules, which can be stored in random access memory, flash memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. The storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. Alternatively, the ASIC can reside in a base station or terminal. The processor and storage medium can also exist as discrete components in a base station or terminal.

[0334] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this application are performed entirely or partially. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a network device, a user equipment, or other programmable device. The computer program or instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; it can also be an optical medium, such as a digital video optical disc; or it can be a semiconductor medium, such as a solid-state drive. The computer-readable storage medium may be a volatile or non-volatile storage medium, or may include both types of storage media.

[0335] In the above embodiments, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0336] In this document, "at least one" means one or more. "More than one" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, or B exists alone, where A and B can be singular or plural. In the textual description of this application, the character " / " generally indicates that the related objects before and after are in an "or" relationship; in the formulas of this application, the character " / " indicates that the related objects before and after are in a "division" relationship. "Including at least one of A, B, and C" can mean: including A; including B; including C; including A and B; including A and C; including B and C; including A, B, and C.

[0337] In this application, information A includes information B. This can be understood as information B directly existing within information A, or information A being used to indicate information B. "Used to indicate" can include both direct and indirect indication. When describing an indication information as being used to indicate A, it can include whether the indication information directly or indirectly indicates A, but does not necessarily mean that the indication information includes A. The information indicated by the indication information is called the information to be indicated. In specific implementations, there are many ways to indicate the information to be indicated. The information to be indicated can be sent as a whole, or it can be divided into multiple sub-information messages and sent separately. The sending period and / or timing of these sub-information messages can be the same or different. This application does not limit the specific sending method. The sending period and / or timing of these sub-information messages can be predefined, for example, predefined according to a protocol, or configured by the transmitting device by sending configuration information to the receiving device. This configuration information can be, but is not limited to, one or a combination of at least two of RRC signaling, MAC layer signaling, and physical layer signaling.

[0338] It should be understood that in the various embodiments of this application, the terms "first," "second," and various numerical designations are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application. The order of the sequence numbers of the above processes does not imply the order of execution; the execution order of each process should be determined by its function and internal logic.

[0339] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0340] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0341] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0342] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0343] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0344] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0345] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method of communication, comprising: The method comprises: receiving Q first signals from an access network device, Q being a positive integer; determining N beams according to measurement results of the Q first signals and an AI model, a value of N being determined according to a prediction uncertainty, the prediction uncertainty being determined according to the AI model and the measurement results of the Q first signals, wherein input of the AI model comprises the measurement results of the Q first signals, and N is a positive integer; sending indication information of the N beams to the access network device.

2. The method of claim 1, wherein, The value of N is determined according to a prediction uncertainty, comprising: determining the value of N according to a size relationship between the prediction uncertainty and a first uncertainty threshold.

3. The method of claim 2, wherein, The determination of the value of N according to the size relationship between the prediction uncertainty and the first uncertainty threshold comprises: in a case where the prediction uncertainty is greater than or equal to the first uncertainty threshold, N is a first value; or in a case where the prediction uncertainty is less than the first uncertainty threshold, N is a second value, wherein the first value and the second value are both positive integers, and the first value is greater than the second value.

4. The method according to claim 2 or 3, characterized in that, The method further comprises: receiving first information from the first network element, the first information comprising the first uncertainty threshold.

5. The method of claim 1, wherein, The value of N is determined according to a prediction uncertainty, comprising: determining the N value according to the size relationship between the prediction uncertainty and the uncertainty threshold in the first list, the first list comprising at least two uncertainty thresholds.

6. The method of claim 5, wherein, The determination of the value of N according to the size relationship between the prediction uncertainty and the uncertainty threshold in the first list comprises: In a case where the prediction uncertainty is greater than or equal to an (i-1)th uncertainty threshold in the first list and / or the prediction uncertainty is less than an ith uncertainty threshold in the first list, N takes a value of N i , N i is a positive integer.

7. The method according to claim 5 or 6, characterized in that, The method further comprises: receiving first information from the first network element, the first information comprising a first list.

8. The method according to any one of claims 1 to 7, characterized in that, The output of the AI model further comprises the prediction uncertainty, wherein the prediction uncertainty is determined according to the AI model and the measurement results of the Q first signals, comprising: inputting the measurement results of the Q first signals into the AI model to obtain the prediction uncertainty.

9. The method according to any one of claims 1 to 7, characterized in that, The prediction uncertainty is determined according to the AI model and the measurement results of the Q first signals, which comprises: determining the prediction uncertainty according to a first gradient, the first gradient being a gradient of the AI model with respect to the measurement results of the Q first signals.

10. The method according to any one of claims 1 to 9, characterized in that, The N beams comprise a first beam and a second beam, and a spatial interval between the first beam and the second beam is greater than or equal to an interval threshold.

11. A method of communication, comprising: The method comprises: receiving Q first signals from an access network device, Q being a positive integers; determining N beams according to measurement results of the Q first signals, an AI model and an interval threshold, the N beams comprising a first beam and a second beam, and a spatial interval between the first beam and the second bean being greater than or equal to an interval threshold, wherein input of the AI model comprises the measurement results of the Q first signals, and N being a positive integer; sending indication information of the N beams to the access network device.

12. The method according to claim 10 or 11, characterized in that, The interval threshold comprises a first interval value, and the interval between the first beam and the second beam in space is greater than or equal to the interval threshold, comprising: The interval between the first beam and the second beam in space is greater than or equal to the first interval value.

13. The method of claim 12, wherein, The interval between the first beam and the second beam in space is a horizontal interval between the first beam and the second beam in space; or, The interval between the first beam and the second beam in space is a vertical interval between the first beam and the second beam in space; Or, The interval between the first beam and the second beam in space is a minimum or maximum value between a horizontal interval between the first beam and the second beam in space and a vertical interval between the first beam and the second beam in space; or, The interval between the first and second beams in space is the sum of the horizontal interval between the first and second beams in space and the vertical interval between the first and second beams in space; or, The interval between the first and second beams in space is the square root of the square sum of the horizontal interval between the first and second beams in space and the vertical interval between the second and second beams in space.

14. The method of claim 10 or 11, wherein, The interval threshold comprises a first interval value and a second interval value, and the interval between the first beam and the second beam in space is greater than the interval threshold, comprising: The horizontal interval between the first beam and the second beam in space is greater than or equal to the first threshold value; and / or, The vertical interval between the first beam and the second beam in space is greater than or equal to the second interval value.

15. The method according to any one of claims 10 to 14, characterized in that, The method further comprises: Receiving second information from the first network element, the second information comprising the interval threshold.

16. The method according to any one of claims 11 to 15, characterized in that, The determination of the N beams according to the measurement results of the Q first signals, the AI model and the interval threshold comprises: In a case where the prediction uncertainty is greater than or equal to a second uncertainty threshold, determining the N beams according to the interval threshold, wherein the prediction uncertainty is determined according to the AI model and the measurement results of the Q first signals.

17. The method of claim 16, wherein, The method further comprises: Receiving third information from the first network element, the third information comprising the second uncertainty threshold.

18. The method according to claim 16 or 17, characterized in that The output of the AI model further comprises the prediction uncertainty, and the method further comprises: Inputting the measurement results of the Q first signals into the AI model to obtain the prediction uncertainty.

19. The method of claim 16 or 17, wherein, The method further comprises: Determining the prediction uncertainty according to a first gradient, the first gradient being a gradient of the AI model with respect to the measurement results of the Q first signals.

20. The method of any one of claims 10 to 19, wherein, The first beam is any one of the N beams, and the second beam is any one of the N beams different from the first beam.

21. The method according to any one of claims 10 to 19, characterized in that, The signal strength of the first beam and the second beam is greater than a first strength threshold.

22. The method of claim 21, wherein, The method further comprises: Receiving fourth information from the first network element, the fourth information comprising the first strength threshold.

23. The method of any one of claims 1 to 22, wherein, The method further comprises: receive fifth information from the first network element, the fifth information including information of the AI model.

24. A communications device, characterized by comprise a module or unit for performing the method of any one of claims 1 to 23.

25. A communications device, characterized by comprise one or more processors for executing computer programs or instructions stored in a memory, causing the apparatus to perform the method of any one of claims 1 to 23.

26. A computer-readable storage medium, characterized in that, The storage medium has stored therein computer programs or instructions, which, when executed by the communication apparatus, implement the method of any one of claims 1 to 23.

27. A computer program product, characterised in that, comprise a computer program, which, when executed by a processor, implement the method of any one of claims 1 to 23.

Images