Artificial intelligence / machine learning-based beam management method and wireless communication device

Abstract

Embodiments of the present application provide an artificial intelligence / machine learning (AI / ML)-based beam management method and a wireless communication device. The beam management method is executed on a user equipment (UE), and comprises: the UE initiates a switch between a first mode and a second mode to execute AI / ML-based beam prediction, wherein the first mode refers to the AI / ML-based beam prediction supporting beam prediction based on a UE-side model, and the second mode refers to the AI / ML-based beam prediction supporting beam prediction based on a network device-side model.

Description

Beam management method and wireless communication equipment based on artificial intelligence / machine learning Technical Field

[0001] The embodiments of the present application relate to the field of mobile communication technologies, and more particularly to a beam management method and wireless communication device based on artificial intelligence / machine learning (AI / ML). Background Art

[0002] In existing technologies, artificial intelligence / machine learning (AI / ML) is a system that can replace human labor through computational learning. AI / ML can be used to solve various problems, such as natural human language processing, computing, and graphics processing. In recent years, AI / ML has been applied in the field of communications. However, AI / ML-based beam management methods have unresolved issues in the field of wireless communications. Therefore, it is necessary to propose an AI / ML-based beam management method and wireless communication device to improve the problems of existing technologies and other issues.

[0003] Summary of the Invention

[0004] Embodiments of the present application provide a beam management method and wireless communication device based on artificial intelligence / machine learning (AI / ML).

[0005] An embodiment of the present application provides a beam management method based on artificial intelligence / machine learning AI / ML, which is executed on a user device, wherein the AI / ML-based beam management method includes: the user device initiating a conversion between a first mode and a second mode to perform AI / ML-based beam prediction, wherein the first mode means that the AI / ML-based beam prediction supports beam prediction based on a user device side model, and the second mode means that the AI / ML-based beam prediction supports beam prediction based on a network device side model.

[0006] Through the above technical solution, the user equipment initiates a transition between the first mode and the second mode to perform AI / ML-based beam prediction. This can reduce the UE's processing complexity, power consumption, and reporting overhead, while helping to improve network performance.

[0007] An embodiment of the present application provides a beam management method based on artificial intelligence / machine learning AI / ML, which is executed on a user device, wherein the beam management method includes: the user device outputs predicted beam information of multiple time periods to perform AI / ML-based beam prediction, wherein the AI / ML-based beam prediction supports beam prediction based on a user device side model; and the user device reports the predicted beam information of the multiple time periods to a network device in a channel state information (CSI) reporting instance, wherein the one CSI instance includes multiple sub-reporting instances, and the user device reports the multiple sub-reporting instances to the network device according to different reporting priorities.

[0008] With the above technical solution, the user equipment outputs predicted beam information for multiple time periods to perform AI / ML-based beam prediction, and the user equipment reports the predicted beam information for multiple time periods to the network device in a single instance of reporting channel state information (CSI). This reduces UE processing complexity, power consumption, and reporting overhead, while also helping to improve network performance.

[0009] An embodiment of the present application provides a beam management method based on artificial intelligence / machine learning AI / ML, which is executed on a network device, wherein the AI / ML-based beam management method includes: the network device initiates a conversion between a first mode and a second mode to perform AI / ML-based beam prediction, wherein the first mode means that the AI / ML-based beam prediction supports beam prediction based on a user device side model, and the second mode means that the AI / ML-based beam prediction supports beam prediction based on a network device side model.

[0010] Through the above technical solution, the network device initiates a switch between the first mode and the second mode to perform AI / ML-based beam prediction, thereby reducing the processing complexity, power consumption, and reporting overhead of the network device.

[0011] An embodiment of the present application provides a beam management method based on artificial intelligence / machine learning AI / ML, which is executed on a network device, wherein the beam management method includes: the network device receives predicted beam information of multiple time periods output by a user device to perform AI / ML-based beam prediction, wherein the AI / ML-based beam prediction supports beam prediction based on a user device side model; and the network device receives the predicted beam information of the multiple time periods reported by the user device in a channel state information (CSI) instance, wherein the one CSI instance includes multiple sub-reporting instances, and the network device receives the multiple sub-reporting instances reported by the user device according to different reporting priorities.

[0012] Through the above technical solution, a network device receives predicted beam information for multiple time periods output by a user device to perform AI / ML-based beam prediction, and the network device also receives the predicted beam information for multiple time periods reported by the user device in a single instance of reporting channel state information (CSI). This reduces the processing complexity, power consumption, and overhead of the network device.

[0013] A wireless communication device provided in an embodiment of the present application includes: a processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory to execute the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0014] The user equipment provided in an embodiment of the present application includes a processor and a memory. The memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to perform the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0015] The base station provided in an embodiment of the present application includes a processor and a memory. The memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to perform the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0016] The network element provided in an embodiment of the present application includes a processor and a memory. The memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to perform the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0017] The chip provided in the embodiment of the present application is used to implement the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0018] Specifically, the chip includes: a processor for calling and running a computer program from a memory, so that a device equipped with the chip executes the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0019] The computer-readable storage medium provided in an embodiment of the present application is used to store a computer program, which enables a computer to execute the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0020] The computer program product provided in an embodiment of the present application includes computer program instructions, which enable a computer to execute the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0021] The computer program provided in the embodiment of the present application, when running on a computer, enables the computer to execute the above-mentioned beam management method based on artificial intelligence / machine learning AI / ML.

[0022] In the above technical solution, the user equipment or network equipment initiates the conversion between the first mode and the second mode to perform AI / ML-based beam prediction. In this way, the processing complexity, power consumption, and reporting overhead of the UE or network equipment can be reduced. The user equipment outputs the predicted beam information of multiple time periods to perform AI / ML-based beam prediction, and the user equipment reports the predicted beam information of the multiple time periods to the network equipment in one channel state information CSI instance. In this way, the processing complexity, power consumption, and reporting overhead of the UE can be reduced, and at the same time, it helps to improve the performance of the network. The network equipment receives the predicted beam information of multiple time periods output by the user equipment to perform AI / ML-based beam prediction, and the network equipment receives the predicted beam information of the multiple time periods reported by the user equipment in one channel state information CSI instance. In this way, the processing complexity, power consumption, and overhead of the network equipment can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0024] FIG1 is a schematic diagram of a wireless communication system architecture provided in an embodiment of the present application;

[0025] FIG2A is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application; ...

[0026] FIG2B is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application;

[0027] FIG3A is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application;

[0028] FIG3B is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application;

[0029] FIG3C is a schematic diagram of beam management based on artificial intelligence / machine learning AI / ML according to an embodiment of the present application;

[0030] FIG4 is a schematic diagram of beam management based on artificial intelligence / machine learning AI / ML according to an embodiment of the present application;

[0031] FIG5 is a schematic structural diagram of a wireless communication device provided in an embodiment of the present application;

[0032] FIG6 is a schematic structural diagram of a chip according to an embodiment of the present application;

[0033] FIG7 is a schematic block diagram of a wireless communication system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] The following will describe the technical solutions in the embodiments of this application in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0035] The technical solutions of the embodiments of the present application can be applied to various wireless communication systems, such as: Long Term Evolution (LTE) system, LTE Frequency Division Duplex (FDD) system, LTE Time Division Duplex (TDD) system, 5G communication system or future wireless communication systems, etc.

[0036] Exemplarily, a wireless communication system 100 used in an embodiment of the present application is shown in Figure 1. The wireless communication system 100 may include a network device 110, which may be a device that communicates with a user equipment 120 (User Equipment, UE). The network device 110 can provide communication coverage for a specific geographical area and can communicate with user equipment located within the coverage area. Optionally, the network device 110 may be a base station or a location management function (LMF) for providing positioning services. Optionally, the base station may be an evolved base station (eNB or eNodeB) in an LTE system, or the base station may be a mobile switching center, a relay station, an access point, a vehicle-mounted device, a wearable device, a hub, a switch, a bridge, a router, a network device in a 5G network, or a base station in a future communication system, etc.

[0037] The wireless communication system 100 also includes at least one user device 120 located within the coverage area of ​​the network device 110. As used herein, "user device" includes, but is not limited to, a device configured to receive / send communication signals via a wired connection, such as a Public Switched Telephone Network (PSTN), a Digital Subscriber Line (DSL), a digital cable, a direct cable connection; and / or another data connection / network; and / or via a wireless interface, such as a cellular network, a Wireless Local Area Network (WLAN), a digital television network such as a DVB-H network, a satellite network, an AM-FM broadcast transmitter; and / or another user device; and / or an Internet of Things (IoT) device. A user device configured to communicate via a wireless interface may be referred to as a "wireless communication user device 120," a "wireless user device 120," or a "mobile user device 120." Examples of mobile user equipment 120 include, but are not limited to, satellite or cellular telephones; Personal Communications System (PCS) user equipment 120, which may combine a cellular radio telephone with data processing, fax, and data communication capabilities; a PDA, which may include a radiotelephone, a pager, Internet / Intranet access, a web browser, a notepad, a calendar, and / or a Global Positioning System (GPS) receiver; and conventional laptop and / or palmtop receivers or other electronic devices that include a radiotelephone transceiver. User equipment may be referred to as access user equipment 120, a subscriber unit, a subscriber station, a mobile station, a mobile station, a remote station, a remote user device, a mobile device, a wireless communication device, or a user agent. The access user device 120 can be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA), a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a wearable device, a user device in a 5G network, or a user device in a future evolved PLMN, etc.

[0038] Some embodiments of the present application mainly consider how to make the models or functions provided by artificial intelligence / machine learning (AI / ML) work better when using AI / ML technology to solve various problems in wireless communication systems.

[0039] In some embodiments of the present application, the user equipment 120 initiates a conversion between the first mode and the second mode to perform AI / ML-based beam prediction. In this way, the processing complexity, power consumption, and reporting overhead of the UE can be reduced, while helping to improve network performance. In some embodiments of the present application, the user equipment 120 outputs predicted beam information for multiple time periods to perform AI / ML-based beam prediction, and the user equipment 120 reports the predicted beam information for the multiple time periods to the network device 110 in one instance of reporting channel state information CSI. In this way, the processing complexity, power consumption, and reporting overhead of the UE can be reduced, while helping to improve network performance.

[0040] In some embodiments of the present application, the network device 110 initiates a transition between the first mode and the second mode to perform AI / ML-based beam prediction. This can reduce the processing complexity, power consumption, and reporting overhead of the network device. In some embodiments of the present application, the network device 110 receives predicted beam information for multiple time periods output by the user device 120 to perform AI / ML-based beam prediction, and the network device 110 receives the predicted beam information for the multiple time periods reported by the user device 120 in a channel state information (CSI) instance. This can reduce the processing complexity, power consumption, and overhead of the network device.

[0041] Optionally, the user equipments 120 may perform device-to-device (D2D) communication with each other.

[0042] Optionally, the 5G communication system or 5G network may also be referred to as a New Radio (NR) system or NR network.

[0043] The wireless communication system 100 also includes a network 130. Network 130 may be an IP mobile communication network operated by a mobile communication operator. For example, network 130 may be a core network used by a mobile communication operator that operates and manages the wireless communication system 100, or a core network used by a virtual mobile communication operator such as an MVNO (Mobile Virtual Network Operator).

[0044] The network 130 can be connected to the network device 110 as a relay device for transmitting user data. The user device 120 sends and receives user data via the network 130. It should be noted that the communication of user data is not limited to IP communication, and non-IP communication may also be used.

[0045] FIG1 exemplarily shows a network device 110 , two user devices 120 and a network 130 . Optionally, the wireless communication system 100 may include multiple network devices and each network device may include other numbers of user devices within its coverage area, which is not limited in the embodiments of the present application.

[0046] Optionally, the wireless communication system 100 may further include other network entities such as a network controller, a mobility management entity, and a network element, which is not limited in this embodiment of the present application. For example, the network 130 may include other network entities such as a network controller, a mobility management entity, and a network element, which is not limited in this embodiment of the present application.

[0047] It should be understood that in the embodiments of the present application, a device having wireless communication capabilities in a network / system may be referred to as a wireless communication device. Taking the wireless communication system 100 shown in Figure 1 as an example, the wireless communication device may include a network device 110 having communication capabilities, a user device 120, and a network 130. The network device 110 and the user device 120 may be the specific devices described above and will not be described in detail here. The wireless communication device may also include other devices (network 130) in the wireless communication system 100. For example, the network 130 may include other network entities such as a network controller and a mobility management entity, which is not limited in the embodiments of the present application.

[0048] It should be understood that the terms "system" and "network" are often used interchangeably in this document. The term "and / or" in this document is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this document generally indicates that the associated objects before and after are in an "or" relationship. In some embodiments, the term "configuration" may refer to "pre-configuration" and "network configuration". The terms "definition" or "pre-definition" in the embodiments of the present application may be implemented by pre-storing corresponding codes, tables or other methods of indicating relevant information in devices (for example, including UE and network devices). This application does not limit the specific implementation methods. For example, "definition" or "pre-definition" may refer to those defined in the protocol. It should also be understood that the "protocol" in the present invention may refer to a standard protocol in the field of communications, for example, it may include the Long Term Evolution (LTE) protocol, the New Radio (NR) protocol, and related protocols used in future communication systems. This application is not limited to this.

[0049] Some embodiments of the present application solve the following technical problems.

[0050] Technical Issue 1: AI-based beam prediction may support beam prediction based on the terminal-side model or the network-side model. If both are supported at the same time, switching between the two models may be necessary. Therefore, how to switch between the two models is one of the issues that needs to be addressed.

[0051] Technical Issue 2: AI-based beam prediction may support beam prediction based on terminal-side models. For terminal-side model time-domain beam prediction, predicted beam information for multiple time periods may be reported in a single CSI reporting instance. Due to the large amount of content reported by CSI, it may be necessary to prioritize CSI reports. How should CSI reporting be prioritized? Network-side models may also require the terminal to report measurement beam information for multiple time periods. If the measurement beam information for multiple time periods is reported in a single reporting instance, the reporting overhead required is high, which may lead to conflicts and may necessitate prioritization of CSI information reporting.

[0052] Technical Issue 3: AI-based beam prediction may support beam prediction based on terminal-side models. For terminal-side models, the terminal side may need to report confidence information of the predicted beam and may also report probability information of the predicted beam. The form of reporting confidence information and probability information needs to be further clarified.

[0053] Technical Issue 4: AI-based beam prediction may support beam prediction based on the terminal-side model. For the terminal-side model, the terminal side may need to report the predicted beam information for multiple time periods, and the number of beams contained in the beam prediction information for each time period may be greater than 4. If the indication of the predicted beam information indicates SSBRI or CRI in the existing way, the indication overhead will be relatively large. Therefore, it is necessary to consider the compression of the indication information of the predicted beam reported by the terminal side.

[0054] In response to the above-mentioned technical problems, some embodiments of the present application solve the above-mentioned technical problems from the following aspects:

[0055] (1) For AI / ML-supported beam prediction, the terminal may be capable of supporting both UE-side models and network-side models. This may involve switching from the UE-side model to the network-side model, or vice versa. This requires the terminal and the network to interact through air interface signaling to complete the above process.

[0056] (2) For beam prediction supported by AI / ML, the terminal-side model may need to report predicted beam information for multiple time periods simultaneously in a single reporting instance. This results in high terminal reporting overhead and a high probability of conflict. Therefore, the present invention prioritizes reported predicted beam information based on time periods, proposes a discarding scheme that prioritizes reports based on their content, and provides possible methods for prioritizing reported content.

[0057] (3) For beam prediction supported by AI / ML, the terminal-side model may need to report the predicted beam information of multiple time periods simultaneously in one reporting instance. Multiple predicted beam information may be output for each time period. In this case, the terminal reporting overhead is relatively large. The present invention considers optimizing the indication of the predicted beam to reduce the reporting overhead. At the same time, it also proposes that when different numbers of beams are reported in different time periods, the terminal can flexibly select the representation form of the beam indication information.

[0058] (4) For beam prediction supported by AI / ML, the terminal-side model may need to report predicted beam information for multiple time periods simultaneously in one reporting instance. Multiple predicted beam information may be output for each time period. At the same time, the terminal may also report probability information or confidence information of multiple beams. This embodiment provides a possible representation method of probability information and confidence information when the terminal provides feedback.

[0059] To facilitate understanding of the technical solutions of the embodiments of the present application, the following describes the technical solutions related to the embodiments of the present application. The following lists the technical solutions of the first embodiment, the second embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and the ninth embodiment for illustration, but the present application is not limited thereto. First embodiment:

[0060] In some embodiments of the present application, the solution of the first embodiment may be implemented in conjunction with the solution of the second embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and / or the ninth embodiment, or may be implemented independently of the solution of the second embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and the ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0061] FIG2A is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application. As shown in FIG2A , the beam management method based on artificial intelligence / machine learning AI / ML is executed on a user equipment (UE) and includes at least one of the following operations: Operation 201A: The user equipment initiates a conversion between a first mode and a second mode to perform AI / ML-based beam prediction. The first mode means that the AI / ML-based beam prediction supports beam prediction based on a user equipment side model, and the second mode means that the AI / ML-based beam prediction supports beam prediction based on a network device side model.

[0062] Through the above technical solution, the user equipment initiates a transition between the first mode and the second mode to perform AI / ML-based beam prediction. This can reduce the UE's processing complexity, power consumption, and reporting overhead, while helping to improve network performance.

[0063] FIG2B is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application. As shown in FIG2B , the beam management method based on artificial intelligence / machine learning AI / ML is executed on a network device and includes at least one of the following operations: Operation 201B: The network device initiates a conversion between a first mode and a second mode to perform AI / ML-based beam prediction. The first mode refers to that the AI / ML-based beam prediction supports beam prediction based on a user device side model, and the second mode refers to that the AI / ML-based beam prediction supports beam prediction based on a network device side model.

[0064] Through the above technical solution, the network device initiates a switch between the first mode and the second mode to perform AI / ML-based beam prediction, thereby reducing the processing complexity, power consumption, and reporting overhead of the network device.

[0065] In some embodiments, the first mode serves as a basic feature of the user device and the second mode serves as a capability of the user device; or the second mode serves as a basic feature of the user device and the first mode serves as a capability of the user device. In some embodiments, if the currently running AI / ML model applicable to the first function is located on the user device, the transition between the first mode and the second mode is initiated by the user device.

[0066] Specifically, in some embodiments, the UE is, for example, the user equipment 120 shown in Figure 1. The network device is, for example, the network device 110 shown in Figure 1. The network device 110 is, for example, a base station or a LMF.

[0067] Exemplarily, in some embodiments of the present application, a fusion design of the two modes is considered, or two schemes are supported at the same time, and the configurations of user equipment measurement and prediction are consistent. Specifically, in some embodiments, beam prediction based on AI / ML may support beam prediction based on the user equipment side model, and may also support beam prediction based on the network side model, and may also support beam prediction based on the user equipment side model and the network side model at the same time. It is assumed here that the user equipment can support beam prediction based on the user equipment side model, which is the first mode, and can also support beam prediction based on the network side model, which is the second mode. From the perspective of the user equipment, both modes may serve as basic features of the user equipment, while the other mode serves as a capability item of the user equipment. For example, the second mode serves as a basic feature of the user equipment, and the first mode serves as a capability item of the user equipment. Therefore, the first mode model requires the user equipment to have the ability to perform beam prediction based on the model on the user equipment side; in the actual system operation process, switching and selection of the two modes may be involved, and the conversion between the two modes may adopt at least one of the following schemes:

[0068] Option 1:

[0069] In some embodiments, if the user equipment detects that the output result of the currently running AI / ML model applicable to the first function is less than a predefined first threshold, the user equipment sends a first deactivation request to the network device to request the network device to deactivate the currently running AI / ML model applicable to the first function. The first function refers to beam management or an AI / ML-based function. AI / ML-based functions refer to positioning, CSI prediction, beam prediction, mobility management, etc. In some embodiments, the output result includes the confidence of the predicted beam or the reference signal received power RSRP value of the predicted beam. In some embodiments, the first deactivation request occupies 1 bit, and the 1 bit is used to indicate a request to deactivate the currently running AI / ML model applicable to the first function.

[0070] Assume that the currently running AI / ML model applicable to a certain function is located on the user equipment side, and the conversion between different modes is initiated by the user equipment side. Assume that the user equipment side monitors that the output of the model is not good, for example, the confidence of the predicted beam output by the user equipment side model is low, or the RSRP value is small. At this time, the user equipment side requests the network side to deactivate the currently activated / working model. The base station side sends a deactivation command to the user equipment side based on the request of the user equipment side. At the same time, the network side activates the AI / ML model on the network side and activates the measurement resources in the measurement set. It should be noted that the above-mentioned deactivation request instruction on the user equipment side and the response deactivation command on the network side do not need to specify the corresponding model. The deactivation request on the user equipment side may only require 1 bit to represent, for example, bit 1 / 0 represents a request to deactivate the currently working model; and the deactivation command on the network side can also be represented by 1 bit, for example, a bit of 0 indicates deactivation of the model on the user equipment side, and if it is 1, it indicates activation of the model on the user equipment side.

[0071] Option 2:

[0072] In some embodiments, if the currently running AI / ML model applicable to the first function is located on the user device, the transition between the first mode and the second mode is initiated by the network device.

[0073] In some embodiments, if the network device detects that the output result of the currently running AI / ML model applicable to the first function is less than a predefined first threshold, the network device sends a first deactivation command to the UE to deactivate the currently running AI / ML model applicable to the first function. The first function refers to beam management or an AI / ML-based function. AI / ML-based functions refer to positioning, CSI prediction, beam prediction, mobility management, etc. In some embodiments, the output result includes the confidence of the predicted beam or the reference signal received power RSRP value of the predicted beam. In some embodiments, the first deactivation command occupies 1 bit, and the 1 bit is used to indicate whether to deactivate the currently running AI / ML model applicable to the first function.

[0074] Assuming that the currently running AI / ML model applicable to a certain function is located on the user equipment side, and the conversion between different modes is initiated by the network side, assuming that the network side believes that the beam prediction result reported by the user equipment side is not ideal, for example, the network side believes that the confidence of the predicted beam output by the user equipment side model is low, or the RSRP value is small, then the network side sends a deactivation command to the user equipment side to deactivate the model currently working on the user equipment, and at the same time, the network side activates the AI / ML model on the network side and activates the measurement resource sending in the measurement set; the above-mentioned deactivation signaling can be carried in the measurement resource sending command in the activation measurement set, and the above-mentioned activation command can be for a certain model or not for a specific model, that is, the network does not need to know which specific model is working on the user equipment side, but only sends an instruction to activate the model on the user equipment side, for example, using 1 bit to indicate whether to activate the model on the user equipment side. If the bit is 0, it indicates that the model on the user equipment side is deactivated, and if it is 1, it indicates that the model on the user equipment side is activated.

[0075] Option 3:

[0076] In some embodiments, if the currently running AI / ML model applicable to the first function is located on a network device, the conversion between the first mode and the second mode is initiated by the user device. In some embodiments, if the user device detects that the reporting amount of the currently running AI / ML model applicable to the first function is greater than a predefined second threshold, the user device sends a second deactivation request to the network device to request the network device to deactivate the currently running AI / ML model applicable to the first function. The first function refers to beam management or an AI / ML-based function. AI / ML-based functions refer to positioning, CSI prediction, beam prediction, mobility management, etc. In some embodiments, the second deactivation request occupies 1 bit, and the 1 bit is used to indicate a request to deactivate the currently running AI / ML model applicable to the first function.

[0077] Assume that the currently running AI / ML model applicable to a certain function is located on the network side, and the conversion between different modes is initiated by the user equipment. For example, the user equipment side still requires a large amount of reporting for the network-side model, the measurement complexity of the reference signal to be processed is high, which may exceed the capability limit of the user equipment, and the reporting overhead is large. In this case, the user equipment side may initiate a request to the network side to deactivate the AI / ML model on the network side. At the same time, the user equipment side may also tell the network side the ID of the model to be activated and the input type required by the model, or directly tell the network side the required input type. If the network side knows the input type required by the user equipment side based on standard pre-definition / agreement, the user equipment side does not need to send the above information, thereby facilitating the network side to send measurement reference resources according to the request of the user equipment side. The network side decides whether to deactivate the current model or switch to another model based on the request of the user equipment side, and sends the network side's decision to the user equipment side. If the network side agrees to deactivate the current model, it will send the corresponding measurement resources according to the request of the user equipment side.

[0078] It should be noted that the above-mentioned request on the user device side is not for a specific model, that is, the user device does not need to know which model is specifically working on the network side. It only needs to send an instruction request to activate the model on the network side. For example, 1 bit is used to request whether to activate the model on the user device side. If the bit is 0 / 1, it means a request to deactivate the model on the user device side; and the response decision on the network side does not need to be targeted at a specific model. It only needs to inform the user device through 1 bit, for example, bit 0 / 1 agrees to the request on the user device side.

[0079] Option 4:

[0080] In some embodiments, if the currently running AI / ML model applicable to the first function is located on a network device, the conversion between the first mode and the second mode is initiated by the network device. In some embodiments, if the network device detects that the reporting amount of the currently running AI / ML model applicable to the first function is greater than a predefined second threshold, the network device sends a second deactivation command to the UE to deactivate the currently running AI / ML model applicable to the first function. The first function refers to beam management or an AI / ML-based function. AI / ML-based functions refer to positioning, CSI prediction, beam prediction, mobility management, etc. In some embodiments, the second deactivation command occupies 1 bit, and the 1 bit is used to indicate whether to deactivate the currently running AI / ML model applicable to the first function.

[0081] Assume that the currently running AI / ML model for a certain function is located on the network side, and the transition between different modes is initiated by the network side. For example, the network side believes that the confidence of the predicted beam output by the current working model is low, or the RSRP value is small, which is not conducive to the network side making further decisions, or believes that the reporting overhead on the user equipment side is large. In this case, the network side may send an activation command to the user equipment side to activate the model on the user equipment side and activate the corresponding reference signal measurement resources. The activation command does not need to be model-specific. Of course, if the network knows which models are loaded on the user equipment side, the activation command can be model-specific.

[0082] It should be noted that if the user equipment side may only support one model, the network side only needs to send the corresponding activation instruction, for example, through 1 bit to indicate the activation of the model on the user equipment side; if the user equipment side supports multiple models, and multiple models are based on the same type of input, the network side may also instruct the user equipment to activate that model, or directly send a 1-bit activation instruction. The specific activation of the model depends on the processing on the user equipment side. If the user equipment side contains multiple models that require different inputs, the network side may indicate the activation of a model on the user equipment side and activate the corresponding measurement reference resource.

[0083] Furthermore, in the case where the model on the network side is different from the model on the user device side, if the capabilities reported by the user device side indicate that the user device can support both the model on the user device side and the model on the network side, then when starting the AI / ML model to implement a certain function, it may involve the question of whether to activate the model on the network side or the model on the user device side. For this, the user device may be required to report the results of the beam measurement and the prediction results of the user device side model at the same time. In this way, the network side can compare the output results of the models on the user device side and the network side based on the results reported by the user device side, and thus select the model with better results; if the model on the network side is selected, it may be necessary to send a deactivation command to the user device to activate the model on the user device side; if the model on the user device side is selected, it is necessary to deactivate the reporting of the reference signal measurement results.

[0084] Second embodiment:

[0085] In some embodiments of the present application, the solution of the second embodiment may be implemented in conjunction with the solution of the first embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and / or the ninth embodiment, or may be implemented independently of the solution of the first embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and the ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0086] Figure 3A is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application. As shown in Figure 3A, the beam management method based on artificial intelligence / machine learning AI / ML is executed on a user equipment (UE) and includes at least one of the following operations: Operation 301A: The user equipment outputs predicted beam information of multiple time periods to perform beam prediction based on AI / ML. The AI / ML-based beam prediction supports beam prediction based on a user equipment side model. Operation 302A: The user equipment reports the predicted beam information of the multiple time periods to the network device in a channel state information (CSI) instance. The one CSI instance includes multiple sub-reporting instances, and the user equipment reports the multiple sub-reporting instances to the network device according to different reporting priorities.

[0087] With the above technical solution, the user equipment outputs predicted beam information for multiple time periods to perform AI / ML-based beam prediction, and the user equipment reports the predicted beam information for multiple time periods to the network device in a single instance of reporting channel state information (CSI). This reduces UE processing complexity, power consumption, and reporting overhead, while also helping to improve network performance.

[0088] Figure 3B is a flow chart of a beam management method based on artificial intelligence / machine learning AI / ML provided in an embodiment of the present application. As shown in Figure 3B, the beam management method based on artificial intelligence / machine learning AI / ML is executed on a network device and includes at least one of the following operations: Operation 301B: The network device receives predicted beam information for multiple time periods output by a user device to perform AI / ML-based beam prediction. The AI / ML-based beam prediction supports beam prediction based on a user device side model. Operation 302B: The network device receives the predicted beam information for the multiple time periods reported by the user device in a channel state information (CSI) instance. The CSI instance includes multiple sub-reporting instances, and the network device receives the multiple sub-reporting instances reported by the user device according to different reporting priorities.

[0089] Through the above technical solution, a network device receives predicted beam information for multiple time periods output by a user device to perform AI / ML-based beam prediction, and the network device also receives the predicted beam information for multiple time periods reported by the user device in a single instance of reporting channel state information (CSI). This reduces the processing complexity, power consumption, and overhead of the network device.

[0090] In some embodiments of the present application, the indication information of a CSI instance includes: taking the predicted beam corresponding to the maximum RSRP value of each time period as a reference point; or taking the predicted beam corresponding to the maximum probability value of each time period as a reference point; or taking the predicted beam corresponding to the maximum RSRP value or the maximum confidence value of each time period as a reference point.

[0091] For example, in some embodiments of the present application, for beam prediction based on AI / ML, for a scenario in which beam information at multiple future moments is predicted based on beam measurement information at historical moments, if the model is located on the user device side, the user device side may output predicted beam information for multiple future moments, which may include ID information or CRI of multiple beams predicted at multiple future moments, as well as RSRP information corresponding to the beam ID or CRI, and may also include reliability information corresponding to different beams, etc.; since the number of predicted beams output at each moment may be greater than 4, and the predicted beam information at multiple moments may be reported in the same CSI reporting instance, and since the reporting requires more uplink resources, this embodiment considers a simplified reporting scheme for the predicted beam information by the user device, thereby reducing the feedback overhead of the user device, and the simplification of the reported information by the user device may include at least one of the following schemes.

[0092] Option 1:

[0093] In some embodiments of the present application, if the number of beams in the beam set in each time period is M, M is a positive integer, and the number of predicted beams is K, then the positions of the K beams in the beam set are expressed as a combination number: That is, the K beams are selected from the M beams, and the number of combinations represents the occupied bits.

[0094] For example, in some embodiments of the present application, the user equipment feeds back the position information of the topK beams in the beam set in each time period. Assuming that the number of beams in the beam set is M and the predicted number of beams is K, the positions of these K beams in the beam set can be expressed as follows by the combination number: That is, K beams are selected from M beams, and the number of bits required is

[0095] Option 2:

[0096] In some embodiments of the present application, if the number of beams in the beam set in each time period is M, and the M beams are equally divided into N beam segments, M and N are positive integers, and the number of beams contained in each beam segment is M / N, it is indicated that each beam segment occupies M / N bits. In some embodiments of the present application, N bits are used to indicate whether each beam segment in each time period has a predicted beam. If L of the N bits are 0, it indicates that the beam set occupies (NL)M / N bits. If L of the N bits are 1, it indicates that the beam set occupies LM / N bits.

[0097] For example, in some embodiments of the present application, for beam prediction in the time domain, the positions of the beams that may be predicted in each time period in the beam set are similar, and the beams in the beam set may cover the entire area of ​​the cell in sequence according to the number, so that the top beams predicted by the AI / ML model in each time period are close to each other. The K beams are likely located in adjacent or adjacent beams. Therefore, the entire beam set can be indicated in segments. Assuming there are M beams, they can be divided into N segments, each containing M / N beams. Therefore, indicating each segment requires M / N bits. N bits are also used to indicate whether each segment contains a predicted beam. If a segment does not contain a predicted beam, no further bitmap information is required for that segment, thereby reducing feedback overhead. It should be noted that the segment N can be adjustable. For example, N has multiple candidate values, and the user equipment can select an appropriate N value from the candidate values ​​for feedback. The user equipment needs to feedback the N value to the user equipment. The feedback of the N value can be fed back in an indexed manner. For example, if N has four values ​​{4, 8, 16, 32}, the feedback of the N value requires 2 bits, and the four states 00, 01, 10, and 11 correspond to 4, 8, 16, and 32, respectively. Furthermore, if the N value is indicated to the user equipment by the base station, the user equipment does not need to feedback it.

[0098] Option 3:

[0099] For example, in some embodiments of the present application, for beam prediction in the time domain, the positions of the beams that may be predicted in each time period in the beam set are similar, and the beams in the beam set may cover the entire area of ​​the cell in sequence by number. In this way, the TopK beams predicted by the AI / ML model in each time period are very likely to be located in multiple adjacent or similar beams. In order to better indicate the position of the predicted beam in the beam set and reduce the feedback overhead of the user equipment, this solution provides a specific indication method of the predicted beam in the beam set based on the predicted beam information that may be fed back by the user equipment. The specific indication method is as follows:

[0100] Case 1:

[0101] In some embodiments of the present application, a reference beam is set in the beam set for each time period, and first length information is indicated for the reference beam. The first length information is relative to the position of the reference beam on both sides, and the number of bits occupied by indicating the position of the predicted beam in the beam set is determined based on the first length information. In some embodiments of the present application, if the beam set for each time period has multiple candidate reference beam positions, a candidate reference beam is selected from the multiple candidate reference beams based on the predicted beam, and second length information is indicated based on the candidate reference beam.

[0102] Exemplarily, in some embodiments of the present application, the user equipment feeds back the beam information of the predicted Top K beams in the beam set; assuming that the above beam information is mainly indicated by a bitmap, then based on the predicted beam may be multiple adjacent or similar beams, a reference beam can be set in the beam set, and then a length information can be indicated relative to the reference beam. The length information is relative to the position on both sides of the reference beam, based on which the number of bits required to indicate the position of the predicted beam in the beam set can be determined. For example, the reference beam can be the middle position of the beam index in the beam set. If there are M beams in the beam set, the reference beam can be the beam corresponding to the M / 2 index; in addition, it can also be assumed that there are multiple candidate reference beam positions, so that the user equipment can select a candidate reference beam based on the specific information of the predicted beam, and indicate a length information based on the reference beam, thereby achieving the purpose of reducing the indication overhead; for example, the beam set can be divided into N equal parts, and then through To indicate which reference beam is selected, a length information L is then indicated to determine the number of bits required to indicate the position of the predicted beam in the beam set, as shown in FIG3C .

[0103] In addition, the positions of the above-mentioned candidate reference points can be predefined by the standard, or indicated by the base station, or determined by the user equipment and fed back to the base station, or predefined by the standard, determined by the user equipment and fed back to the base station.

[0104] Case 2:

[0105] In some embodiments of the present application, the indication information of a CSI instance includes a predicted beam corresponding to the maximum RSRP value in each time period as a reference point.

[0106] Exemplarily, in some embodiments of the present application, the user equipment feeds back the beam information of the predicted Top K beams in the beam set, as well as the RSRP information of the predicted Top K beams in the beam set; assuming that the above beam information is mainly indicated by a bitmap, then based on the fact that the predicted beams may be multiple adjacent or similar beams, a reference beam can be set in the beam set, and then a length information can be indicated relative to the reference beam. The length information is relative to the position on both sides of the reference beam, based on which the number of bits required to indicate the position of the predicted beam in the beam set can be determined. Therefore, the same solution as in Case 1 can be adopted for Case 2. Secondly, since Case 2 may need to indicate the maximum RSRP information corresponding to the predicted beam in each time period, the predicted beam corresponding to the maximum RSRP in each time period can be used as a reference point.

[0107] Case 3:

[0108] In some embodiments of the present application, the indication information of a CSI instance includes a predicted beam corresponding to a maximum probability value in each time period as a reference point.

[0109] Exemplarily, in some embodiments of the present application, the user equipment feeds back the beam information of the predicted Top K beams in the beam set, as well as the probability information of the predicted Top K beams in the beam set; assuming that the above beam information is mainly indicated by a bitmap, then based on the fact that the predicted beam may be multiple adjacent or similar beams, a reference beam can be set in the beam set, and then a length information can be indicated relative to the reference beam. The length information is relative to the position on both sides of the reference beam, based on which the number of bits required to indicate the position of the predicted beam in the beam set can be determined. Therefore, the same solution as in Case 1 can be adopted for Case 3. Furthermore, since Case 3 may need to indicate the maximum probability information corresponding to the predicted beam in each time period, the predicted beam corresponding to the maximum probability value in each time period can be used as a reference point.

[0110] Case 4:

[0111] In some embodiments of the present application, the indication information of the CSI instance includes a predicted beam corresponding to the maximum RSRP value or the maximum confidence value in each time period as a reference point.

[0112] For example, in some embodiments of the present application, the user equipment feeds back beam information for the predicted top K beams in the beam set, RSRP information for the predicted top K beams in the beam set, and confidence information for the RSRP information. Assuming that the above beam information is primarily indicated via a bitmap, since the predicted beams may be multiple adjacent or similar beams, a reference beam can be set in the beam set, and then length information can be indicated relative to the reference beam. This length information is relative to the position on both sides of the reference beam. Based on this, the number of bits required to indicate the position of the predicted beam in the beam set can be determined. Therefore, for Case 4, the same solution as for Case 1 can be adopted. Furthermore, since Case 4 may require indicating the maximum RSRP and / or confidence information corresponding to the predicted beam in each time period, the predicted beam corresponding to the maximum RSRP value or maximum confidence value in each time period can be used as a reference point. Alternatively, one of the predicted beams corresponding to the maximum RSRP value and maximum confidence information can be selected and fed back to the base station via a one-bit indication.

[0113] Furthermore, when indicating the length relative to the reference point, the reference point may be included or not. In addition, the length information of the above scheme may also be indicated in an indexed manner. For example, the standard may predefine an index list of length information. The number of values ​​L of the length information in the list is F. Then the number of bits required to indicate the specific value of L is The specific value of L is related to the number of beams in the predicted beam set, where the value of L is less than or equal to the number of beams in the beam set minus 1. For example, if the predicted beam set contains 64 beams, the value of L can be {2, 4, 6, 8, 12, 16, 24, 32, 48, 52, 54, 56, 58, 60, 62, 64}, and the value of L requires 4 bits to indicate.

[0114] Option 4:

[0115] In some embodiments of the present application, when the number K of predicted beams is greater than or equal to a predefined threshold, the positions of the predicted K beams in the beam set are fed back in a bitmap manner, and the length of the bitmap is the number of beams in the beam set. In some embodiments of the present application, when the number K of predicted beams is less than or equal to a predefined threshold, the positions of the predicted K beams in the beam set are fed back in a beam identifier ID manner. In some embodiments of the present application, the feeding back of the positions of the predicted K beams in the beam set in a beam ID manner includes feeding back the positions of the predicted K beams in the beam set in a synchronization signal block resource indicator SSBRI or a channel state information reference signal resource indicator CRI.

[0116] Exemplarily, in some embodiments of the present application, for the number of predicted beams fed back by the user device, that is, the TopK beams K may have multiple values, or the value of K is dynamically changeable. In the above case, if the value of K is less than a certain threshold, the user device can use the beam ID method to indicate the TopK beams, that is, use SSBRI or CRI to indicate the position information of the specific beam in the beam set or indicate it by the combination number; however, when the number of predicted beams K is greater than a certain threshold, if the position information of the beam in the beam set is still fed back in the form of beam ID, the bit overhead may be large. At this time, the predicted positions of the K beams in the beam set can be fed back in the form of a bitmap, and the length of the bitmap is the number of beams in the beam set; the predicted position information of the beam in the beam set can also be fed back in the form of the combination number in Scheme 1. The method in scheme 2 can also be used to feed back the position information of the predicted beam in the beam set; for example, assuming that the number of beams in the beam set is 32, the predicted number of beams K may have multiple values ​​K = {1, 4, 6, 8, 12, 16}. Obviously, when the value of K is small, the ID information of the beam can be directly indicated by SSBRI or CRI. Specifically, when K = 1 or 4, the required bit overhead is 5 or 20 for indicating by beam ID. Obviously, the bit overhead at this time is less than 32; and if the value of K is greater than 6, if the beam ID indication method is still used, the user equipment needs to spend more bits, and at this time a bitmap can be used for indication, and the maximum length of the bitmap is 32. ; Furthermore, if the user equipment side knows that the distribution of the predicted beams in the beam set meets the scenario mentioned in Solution 2, that is, multiple predicted beams may be adjacent or similar beams, then the user equipment can use Solution 2 to provide feedback; for another example, the network side may instruct the user equipment to report the maximum number of predicted beams, and the user equipment determines how many predicted beams to report based on the information of the predicted beams. If the number of predicted beams reported by the user equipment is less than or equal to a certain threshold value, the user equipment feeds back the index information of the predicted beam in the traditional SSBRI or CRI manner; if the number of predicted beams reported by the user equipment is greater than or equal to a certain threshold value, the user equipment feeds back the position information of the predicted beam in the beam set in the form of a bitmap or the number of combinations in Solution 1. It is also possible to use the method in Solution 2 to feedback the position information of the predicted beam in the beam set, or use other forms of indication methods without specific restrictions here, but the number of bits required by this method is less than the traditional SSBRI or CRI indication method; and the determination of the above-mentioned threshold value can be configured on the network side, or predefined by the standard, or it can be a processing behavior of the user equipment itself.The determination of the threshold value is related to the number of beams included in the prediction beam set and the number of prediction beams that need to be reported. The basis for setting the threshold value is that the bit overhead required by the indication method of the prediction beam information used after the threshold is met is less than or equal to the bit overhead of the indication method of the prediction beam information used before the threshold is not met;.

[0117] It should be noted that there may be multiple threshold values ​​for the above-mentioned indication method. This embodiment does not impose specific constraints because the determination of the threshold value is related to the number of beams in the beam set and the number K of predicted beams. The core idea of ​​this embodiment is that it may be necessary to determine whether the user equipment adopts different predicted beam indication methods based on the number of beams in the beam set and the value of the number K of predicted beams; furthermore, if the user equipment indicates the position information of the predicted beam in the beam set in multiple ways, the user equipment side may also feedback to the base station which specific indication method is adopted. For example, the user equipment side may use four methods to indicate the position information of the predicted beam in the beam set. The user equipment can indicate to the base station which specific method is adopted through 2 bits of information.

[0118] Third embodiment:

[0119] In some embodiments of the present application, the solution of the third embodiment may be implemented in conjunction with the solution of the first embodiment, the second embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and / or the ninth embodiment, or may be implemented independently of the solution of the first embodiment, the second embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and the ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0120] In some embodiments of the present application, the reporting priority decreases as the numbers of the multiple sub-reporting instances increase.

[0121] For example, in some embodiments of the present application, for AI / ML-based beam prediction, for scenarios where beam information for multiple future time periods is predicted based on beam measurement information at historical moments, if the model is located on the user device side, the user device side may output predicted beam information for multiple future time periods. Since the number of predicted beams output for each time period may be greater than 4, and the predicted beam information for multiple time periods may be reported in the same CSI reporting instance, and since the reporting requires a large number of uplink resources, corresponding discarding rules or reporting priority rules need to be designed in consideration of possible conflicts in reported resources. The specific reporting priority rules may adopt the following scheme:

[0122] Assume that the predicted beam information of multiple time periods is reported in one CSI reporting instance, but the CSI reporting instance can be split into N sub-reporting instances, where Nn represents the n+1th sub-reporting instance of the split. The specific splitting latitude can be split according to the predicted beam information of different time periods, or multiple predicted beam information of different time periods. For example, N=2, the two sub-reporting instances are represented as N0 and N1 respectively, where N0 (or N1) may correspond to the predicted beam information of one or more time periods respectively; it should be noted that if a reporting sub-instance contains the predicted beam information of multiple time periods, then the predicted beam information of these multiple time periods The predicted beam information is continuous; for example, assuming that the above-mentioned CSI reporting instance contains predicted beam information for 8 time periods, the reporting instance can be split into two sub-reporting instances, the first reporting instance contains the predicted beam information for the first 5 continuous time periods, and the second reporting instance contains the predicted beam information for the last 3 continuous time periods; note that the number of time periods of the predicted beam information contained in the predicted beam information corresponding to the above-mentioned multiple sub-reporting instances may be the same or different; furthermore, when the above-mentioned CSI reporting instance is split into multiple sub-reporting instances, the reporting sub-instance with a smaller number contains the predicted beam information for an earlier time period. Therefore, this embodiment stipulates that the reporting sub-instance with a smaller number has a higher reporting priority. In addition, the above-mentioned splitting method and the predicted beam information for those time periods contained in each reporting sub-instance may be indicated by the network side, may be predefined by the standard, or may be determined by the user equipment itself;

[0123] In order to more intuitively represent the priorities corresponding to different reporting sub-instances, this embodiment continues to use the CSI reporting priority expression specified in the existing standard to represent the priorities of different sub-reporting instances, which can be specifically expressed as follows:

[0124] Pri iCSI (y,k,c,s)=2·N cells ·M s y+N cells ·M s k+M s c+s

[0125] The smaller the calculated parameter is, the higher the priority is. Other parameters are as follows:

[0126] 1) When the reported CSI is aperiodic CSI (A-CSI), y=0; when the reported CSI is semi-persistent CSI (SP-CSI) and is transmitted on PUSCH, y=1; when the reported CSI is semi-persistent CSI (SP-CSI) and is transmitted on PUCCH, y=2; when the reported CSI is periodic CSI (SP-CSI), y=3.

[0127] 2) The first sub-reporting instance N0 carrying L1-RSRP or L1-SINR corresponds to k=0; the n+1th sub-reporting instance Nn carrying L1-RSRP or L1-SINR corresponds to k=n; in other cases: k=N.

[0128] 3) c: cell index; N_{cells}: number of cells, configured through high-level parameters maxNrofServingCells.

[0129] 4)s:reportConfigID;M_{s}:maxNrofCSI-ReportConfigurations.

[0130] It should be noted that the above representation may not be the only one, and there may be multiple representations. The main idea of ​​this embodiment is to split a single CSI reporting instance according to the order of the predicted beam information time periods of the multiple time periods included. Among the multiple reporting sub-instances after splitting, the sub-instance containing the predicted beam information in the earlier time period has a higher reporting priority;

[0131] Furthermore, the reporting of predicted beam information for multiple time periods may also be based on multiple reporting instances. The k values ​​corresponding to multiple reporting instances are the same, all 0, and the reportConfigID of the reporting instance corresponding to the predicted beam information in the earlier time period is smaller. This can ensure that the reporting priority of the predicted beam information in the earlier time period is higher.

[0132] In addition, it should be noted here that for user equipment, if some information is applicable to the predicted beams of all time periods, then this part of the information is carried in the first sub-reporting instance, or the information of the two for different sub-reporting instances is independent; for example, for the user equipment side model, the user equipment may report the predicted beam information of multiple time periods, which may specifically include the indication information of the top K predicted beams for the next N time periods, and the RSRP information corresponding to the top K predicted beams, and may also include the confidence information corresponding to the top K predicted beams; and the reporting of RSRP information may be reported in the form of differentials. A maximum RSRP information can be selected from the predicted beam information of multiple time periods, and the RSRP values ​​of the predicted beams in other time periods are differentiated from the maximum value, and the differential results are quantified and reported. Then, the reporting of information related to the maximum RSRP can be placed in the first sub-reporting instance, which may specifically include the value of the maximum RSRP and the index information of the maximum RSRP.

[0133] In addition, it should be pointed out for this embodiment that the beam information of the predicted Top K beams can be expressed in multiple forms. For example, the position information of the predicted beam in the predicted beam set can be indicated by a bitmap, or by a combination number, or the position of each beam in the predicted beam set can be indicated separately. For example, if the beam set contains 32 beams, the indication of each beam can be indicated by 5 bits of information. Of course, there may be other representation methods, such as some of the methods mentioned in the second embodiment. However, no matter which representation method is used, it does not affect the reporting priority constraint in this embodiment.

[0134] A reporting instance mentioned in this example may specifically correspond to a reportConfigID; or a reporting instance refers to reporting through a CSI.

[0135] Fourth embodiment:

[0136] In some embodiments of the present application, the solution of the fourth embodiment may be implemented in conjunction with the solution of the first embodiment, the second embodiment, the third embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and / or the ninth embodiment, or may be implemented independently of the solution of the first embodiment, the second embodiment, the third embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and the ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0137] In some embodiments of the present application, the CSI instance includes: beam information of K beams predicted in the beam set, where K is a positive integer greater than or equal to 1; beam information of the K beams predicted in the beam set, and RSRP information of the K beams predicted in the beam set; or beam information of the K beams predicted in the beam set, and probability information of the K beams predicted in the beam set; or beam information of the K beams predicted in the beam set, RSRP information of the K beams predicted in the beam set, and confidence information of the RSRP information. In some embodiments of the present application, the value of K in different time periods is the same or different.

[0138] In some embodiments of the present application, the beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the RSRP information of the predicted K beams in the beam set is carried in a second sub-reporting instance. In some embodiments of the present application, the beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the probability information of the predicted K beams in the beam set is carried in a second sub-reporting instance. In some embodiments of the present application, the beam information of the predicted K beams in the beam set and the RSRP information of the predicted K beams in the beam set are carried in a first sub-reporting instance, and the confidence information of the RSRP information is carried in a second sub-reporting instance. In some embodiments of the present application, the beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the RSRP information of the predicted K beams in the beam set and the confidence information of the RSRP information are carried in a second sub-reporting instance. In some embodiments of the present application, the beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, the RSRP information of the predicted K beams in the beam set is carried in a second sub-reporting instance, and the confidence information of the RSRP information is carried in a third sub-reporting instance.

[0139] For example, in some embodiments of the present application, for AI / ML-based beam prediction, for scenarios where beam information for multiple future time periods is predicted based on beam measurement information at historical moments, if the model is located on the user device side, the user device side may output predicted beam information for multiple future time periods. Since the number of predicted beams output for each time period may be greater than 4, and the predicted beam information for multiple time periods may be reported in the same CSI reporting instance, and since the reporting requires a large number of uplink resources, corresponding discarding rules or reporting priority rules need to be designed in consideration of possible conflicts in reported resources. The specific reporting priority rules may adopt the following scheme:

[0140] Assume that the predicted beam information of multiple time periods is reported in one CSI reporting instance, but the CSI reporting instance can be split into N sub-reporting instances, for example, N = 2, where Nn represents the n+1th sub-reporting instance of the split. The specific splitting dimension can be split according to the content. Since the user equipment may report different content, it is specifically divided into the following cases:

[0141] Case 1: Beam information of the predicted Top K beams in the beam set.

[0142] Case 2: Beam information of the predicted top K beams in the beam set, and RSRP information of the predicted top K beams in the beam set.

[0143] Case 3: Beam information of the predicted top K beams in the beam set, and probability information of the predicted top K beams in the beam set.

[0144] Case 4: Beam information of the top K beams predicted in the beam set, RSRP information of the top K beams predicted in the beam set, and confidence information of the RSRP information.

[0145] Here we mainly consider the two sub-reporting instances and split cases 2 to 4 according to content. For example, for case 2, the beam information of the predicted beam is carried in sub-reporting instance 1, and the RSRP information of the predicted Top K beam is carried in sub-reporting instance 2; for case 3, the beam information of the predicted beam is carried in sub-reporting instance 1, and the probability information of the predicted Top K beam is carried in sub-reporting instance 2; for case 4, the beam information of the predicted beam is carried in sub-reporting instance 1, and the RSRP information of the predicted Top K beam and the corresponding confidence information are carried in sub-reporting instance 2; the reason for splitting the beam information of the predicted Top K beam, and the RSRP information, probability information, and confidence information is mainly because when resources conflict or are limited, the user equipment gives priority to reporting the beam information of the predicted Top K beam, and the base station only has the predicted Top K beam. Even if the beam information of K beams is not available, further processing can still be performed. For example, the base station can perform the P2 process in traditional beam measurement based on the reported beam information of the predicted beam. Therefore, when the above CSI reporting instance is split into multiple sub-reporting instances, the beam information of the predicted top K beams is given the highest priority. In addition, the above top K predicted beams only indicate that the number of predicted beams in a certain time period is K, and does not mean that the number of predicted beams in all time periods is K. The value of K may vary for different time periods.

[0146] In order to more intuitively represent the priorities corresponding to different reporting sub-instances, this embodiment continues to use the CSI reporting priority expression specified in the existing standard to represent the priorities of different sub-reporting instances, which can be specifically expressed as follows:

[0147] Pri iCSI (y,k,c,s)=2·N cells ·M s y+N cells ·M s k+M s c+s

[0148] The smaller the calculated parameter is, the higher the priority is. Other parameters are as follows:

[0149] 1) When the reported CSI is aperiodic CSI (A-CSI), y=0; when the reported CSI is semi-persistent CSI (SP-CSI) and is transmitted on PUSCH, y=1; when the reported CSI is semi-persistent CSI (SP-CSI) and is transmitted on PUCCH, y=2; when the reported CSI is periodic CSI (SP-CSI), y=3.

[0150] 2) The first sub-reporting instance N0 carrying L1-RSRP or L1-SINR corresponds to k=0; the n+1th sub-reporting instance Nn carrying L1-RSRP or L1-SINR corresponds to k=n; in other cases: k=N.

[0151] 3) c: cell index; N_{cells}: number of cells, configured through high-level parameters maxNrofServingCells.

[0152] 4)s:reportConfigID;M_{s}:maxNrofCSI-ReportConfigurations.

[0153] It should be noted that the above representation may not be the only one, and there may be multiple representations. The main idea of ​​this embodiment is to split a single CSI reporting instance according to the impact of the reported content in the CSI on the base station processing. Among the multiple reporting sub-instances after the split, the reporting priority of the beam information containing the predicted Top K beams is the highest.

[0154] In addition, it should be pointed out for this embodiment that the beam information of the predicted Top K beams can be expressed in multiple forms. For example, the position information of the predicted beam in the predicted beam set can be indicated by a bitmap, or by a combination number, or the position of each beam in the predicted beam set can be indicated separately. For example, if the beam set contains 32 beams, the indication of each beam can be indicated by 5 bits of information. Of course, there may be other representation methods, such as some of the methods mentioned in the second embodiment. However, no matter which representation method is used, it does not affect the reporting priority constraint in this embodiment.

[0155] Fifth embodiment:

[0156] In some embodiments of the present application, the solution of the fifth embodiment may be implemented in conjunction with the solution of the first embodiment, the second embodiment, the third embodiment, the fourth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and / or the ninth embodiment, or may be implemented independently of the solution of the first embodiment, the second embodiment, the third embodiment, the fourth embodiment, the sixth embodiment, the seventh embodiment, the eighth embodiment, and the ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0157] In some embodiments of the present application, the predicted beam information of the multiple time periods is the predicted beam information of N time periods, the predicted beam information of X time periods is carried in the first sub-reporting instance, and the predicted beam information of Nx time periods is carried in the second sub-reporting instance, where N and X are positive integers, and N is greater than or equal to X. In some embodiments of the present application, the beam information corresponding to the even beams in the beam set is carried in the first sub-reporting instance, and the beam information corresponding to the odd beams in the beam set is carried in the second sub-reporting instance; or the beam information corresponding to the odd beams in the beam set is carried in the first sub-reporting instance, and the beam information corresponding to the even beams in the beam set is carried in the second sub-reporting instance. In some embodiments of the present application, the index corresponding to the reference value of the probability information, the RSRP information, or the confidence information and the corresponding quantization value are carried in the first sub-reporting instance.

[0158] For example, in some embodiments of the present application, for beam prediction based on AI / ML, for a scenario in which beam information at multiple future moments is predicted based on beam measurement information at historical moments, if the model is on the user device side, the user device side may output predicted beam information for multiple future moments, which may include ID information or CRI of multiple beams predicted at multiple future moments, and RSRP information corresponding to the beam ID or CRI, and may also include reliability information corresponding to different beams, etc.; since the number of predicted beams output at each moment may be greater than 4, and the predicted beam information at multiple moments may be reported in the same CSI reporting instance, and since the reporting requires more uplink resources, this embodiment considers prioritizing the predicted beam content reported by the user device. The specific priority division can refer to the reporting priority of the content in CSI part 2 in the existing standard. First, this embodiment provides the content that the predicted beam information fed back by the user device may contain under different feedback content, and then prioritizes each part of the content. The possible feedback content of the user device includes the following cases:

[0159] Case 1: Beam information of the predicted Top K beams in the beam set.

[0160] Case 2: Beam information of the predicted top K beams in the beam set, and RSRP information of the predicted top K beams in the beam set.

[0161] Case 3: Beam information of the predicted top K beams in the beam set, and probability information of the predicted top K beams in the beam set.

[0162] Case 4: Beam information of the top K beams predicted in the beam set, RSRP information of the top K beams predicted in the beam set, and confidence information of the RSRP information.

[0163] For Case 1, the predicted beam information fed back by the user equipment corresponds to specific beam ID information, such as SSBRI or CRI information, and may also be bitmap information for the predicted beam set.

[0164] For Case 2, the user equipment feedback information specifically includes the beam index information of the TopK beams and the RSRP information of the TopK beams in the beam set. At the same time, in order to indicate the RSRP information of the predicted beam, it also includes the position indication information of the maximum RSRP of all predicted beams.

[0165] For Case 2, if the number of predicted beams for different time periods fed back by the user equipment is variable, the information fed back by the user equipment may also include indication information of the number of beams predicted for different time periods or indication information indicating the number of predicted beams; it should be noted here that the number of beams predicted for each time period may correspond to a minimum value and / or a maximum value, and the specific number of predicted beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period; in addition, the total number of predicted beams for multiple time periods may be indicated by the base station, and the specific number of beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period.

[0166] For Case 3, the user equipment feedback information specifically includes the beam index information of the TopK beams and the probability information of the TopK beams in the beam set. The representation of the probability information may be based on S values ​​uniformly quantized from 0 to 1, with a quantization accuracy of 1 / S, indicating the number of bits required for each predicted beam. It may also be a judgment value based on a certain threshold. For example, if the confidence level is greater than a certain threshold, the corresponding indication information is 1, and if it is lower than a certain threshold, the corresponding indication information is 0.

[0167] For Case 3, if the number of predicted beams for different time periods fed back by the user equipment is variable, the information fed back by the user equipment may also include information on the number of beams predicted for different time periods or indication information indicating the number of predicted beams; it should be noted here that the number of beams predicted for each time period may correspond to a minimum value and / or a maximum value, and the specific number of predicted beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period; in addition, the total number of predicted beams for multiple time periods may be indicated by the base station, and the specific number of beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period.

[0168] For Case 4, the information fed back by the user equipment may include the confidence information of each predicted beam in addition to the information mentioned in Case 2. The confidence information may be represented by S values ​​uniformly quantized from 0 to 1, with a quantization accuracy of 1 / S, indicating the number of bits required for each predicted beam. It may also be a judgment value based on a certain threshold. For example, if the confidence level is greater than a certain threshold, the corresponding indication information is 1, and if it is lower than a certain threshold, the corresponding indication information is 0.

[0169] Solution 1: Assume that the user equipment reports predicted beam information for N time periods. For the above different cases, assume that the information reported by the user equipment can be divided into two groups: Group 0 and Group 1. Group 0 mainly carries information for the first X time periods, and Group 1 carries beam prediction information for the remaining N / N time periods. The information carried by Group 0 has a higher priority than that carried by Group 1. Here, there is no specific constraint on the value of X: X∈{1,2,...,N-1}, and N can be N∈[2,...,16]. The length of the above time period can be a slot, 1ms, 5ms, 10ms, 20ms, etc. Note that the N time periods can be further divided into more groups, for example, N groups. The smaller the time period index, the smaller the corresponding group index, and the smaller the group index, the higher the reporting priority. The specific content that may be reported by the user device for each case in the above case is as described above and will not be further described here; furthermore, the predicted beam information for each time period in the above case is mainly the index information of the predicted beam in the beam set, and the representation of the index information may exist in many different forms. This embodiment does not impose specific restrictions. For example, it can be the SSBRI or CRI of each predicted beam, or the index information of the beam set indicated by a bitmap. Each bit indicates whether the corresponding numbered beam in the beam set is a predicted beam. If the bit is 1 / 0, it indicates that the corresponding beam is a predicted beam, and if the bit is 0 / 1, it indicates that the corresponding beam is not a predicted beam; or it may be represented by the number of combinations of scheme one in the second embodiment; or it may be represented by scheme two or scheme three in the second embodiment.

[0170] Solution 2: Assume that the user equipment feeds back predicted beam information for N time periods. For the above different cases, assume that the information reported by the user equipment can be divided into two groups, Group 0 and Group 1. For each case, the information carried by Group 0 and Group 1 will also be different. The specific information carried by Group 0 and Group 1 for different cases is given below.

[0171] First, assuming that the number of predicted beams that the user equipment needs to report in each time period is fixed, for the four cases described above, Group 0 may carry beam information corresponding to even (or odd) beams in the beam set. That is, if the indexes of some beams in the predicted beam set are even (or odd) numbers, the information for these predicted beams is carried in Group 0; if the indexes of some beams in the predicted beam set are odd (or even) numbers, the information for these predicted beams is carried in Group 1. The smaller the Group index, the greater the reporting priority. The specific content of the predicted beam information for each case is described in the first half of this embodiment. Furthermore, for the above case, the index information of the predicted beam in the beam set in the predicted beam information for each time period may exist in various forms, which are not specifically restricted in this embodiment. For example, it may be the SSBRI or CRI of each predicted beam, or the index information of the beam set indicated by a bitmap, where each bit indicates whether the corresponding numbered beam in the beam set is a predicted beam. If the bit is 1 / 0, it indicates that the corresponding beam is a predicted beam, and if the bit is 0 / 1, it indicates that the corresponding beam is not a predicted beam. Alternatively, it may be represented by the number of combinations in solution 1 of the second embodiment, or by solution 2 or solution 3 of the second embodiment.

[0172] Secondly, for the above cases 2 / 3 / 4, if the number of beams predicted in each time period is variable, then as mentioned above, the predicted beam information may also include information on the number of predicted beams in each time period or indication information indicating the number of predicted beams; this part of information is placed in Group 0; furthermore, the information carried in the above Group 0 and Group 1 can also be interchanged, that is, if the index of some beams in the predicted beam in the beam set is an even number, then the information of this part of the predicted beam is carried in Group 0; if the index of some beams in the predicted beam in the beam set is an odd number, then the information of this part of the predicted beam is carried in Group 1.

[0173] Solution 3: Assume that the user equipment feeds back predicted beam information for N time periods, and the indication of the predicted beam in each time period adopts the indication method of SSBRI or CRI, or adopts some indication methods mentioned in the second embodiment. For the above different cases, it is assumed that the information reported by the user equipment can be divided into multiple groups. The smaller the group number, the higher the corresponding priority. For each case, the information carried in each group will also be different. The specific information carried in different groups is given below for different cases.

[0174] Case 2: For Case 2, it is assumed that it can be divided into two groups. Group 0 can carry the index information of the predicted beam, and Group 1 can carry the RSRP information corresponding to the predicted beam.

[0175] Case 3: For Case 3, it is assumed that it can be divided into two groups. Group 0 can carry the index information of the predicted beam; while Group 1 can carry the probability confidence information corresponding to the predicted beam.

[0176] Case 4: For Case 4, assume that it can be divided into two groups, Group 0 can carry the index information of the predicted beam and the RSRP information corresponding to the predicted beam; and Group 1 can carry the confidence information corresponding to the predicted beam; or Group 0 carries the index information of the predicted beam, and Group 1 carries the RSRP information corresponding to the predicted beam, and the confidence information corresponding to the predicted beam; or it can be divided into 3 groups, Group 0 carries the index information of the predicted beam, Group 1 carries the RSRP information of the predicted beam, and Group 3 carries the confidence information of the predicted beam.

[0177] It should be noted here that, for the above three schemes, the user equipment's reporting of the RSRP information of the predicted beam, the reporting of the confidence information of the predicted beam, and the reporting of the probability information of the predicted beam may involve the selection of a reference value, and other values ​​are differentially quantized relative to the reference value; for example, the maximum / minimum RSRP / confidence value / probability value is selected as a reference, and the other RSRP / confidence value / probability values ​​are differentiated from the maximum / minimum RSRP / confidence value / probability value, then the index corresponding to the maximum / minimum RSRP / confidence value / probability value, and the corresponding quantized value (or other representations corresponding to the maximum / minimum RSRP / confidence value / probability value) are all placed in Group 0.

[0178] Sixth embodiment:

[0179] In some embodiments of the present application, the solution of the sixth embodiment may be implemented in conjunction with the solution of the first, second, third, fourth, fifth, seventh, eighth, and / or ninth embodiment, or may be implemented independently of the solution of the first, second, third, fourth, fifth, seventh, eighth, and ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0180] In some embodiments of the present application, the predicted beam in each time period has a higher priority in the beam set. or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

[0181] In some embodiments of the present application, the predicted beam in each time period has a higher priority in the beam set. or The RSRP information of the predicted beams and the predicted beams in each time period have higher priority in the beam set or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The RSRP information of the predicted beams and the predicted beams in each time period have lower priorities in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

[0182] In some embodiments of the present application, the predicted beam in each time period has a higher priority in the beam set. or The probability information of the predicted beams and the predicted beams in each time period have higher priority in the beam set or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The probability information of the predicted beams and the predicted beams of each time period with lower priority in the beam set or The beam index information of the predicted beam is carried in the second sub-reporting instance.

[0183] In some embodiments of the present application, the predicted beam in each time period has a higher priority in the beam set. or The RSRP information and confidence information of the predicted beams and the predicted beams in each time period have higher priority in the beam set. or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The RSRP information and confidence information of the predicted beams and the predicted beams in each time period have lower priorities in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

[0184] In some embodiments of the present application, the index and corresponding quantization value corresponding to the global and / or local maximum probability value, the global and / or local maximum RSRP value or the global and / or local maximum confidence value are carried in the first sub-reporting instance.

[0185] In some embodiments of the present application, the index corresponding to the maximum probability value, maximum RSRP value or maximum confidence value of the predicted beam in each time period is the starting point, and the priority value corresponding to the starting point is the first value. In some embodiments of the present application, the index of the beam on the first side of the starting point and the index of the beam on the second side of the starting point are traversed alternately, and the priority value is increased by 1 each time the index of a beam is traversed until all beams in the beam set are traversed. In some embodiments of the present application, the alternating traversal of the index of the beam on the first side of the starting point and the index of the beam on the second side of the starting point includes sequentially traversing the index of the first beam on the first side of the starting point and the index of the first beam on the second side of the starting point, and then sequentially traversing the index of the second beam on the first side of the starting point and the index of the second beam on the second side of the starting point.

[0186] For example, in some embodiments of the present application, for beam prediction based on AI / ML, for a scenario in which beam information at multiple future moments is predicted based on beam measurement information at historical moments, if the model is on the user device side, the user device side may output predicted beam information for multiple future moments, which may include ID information or CRI of multiple beams predicted at multiple future moments, and RSRP information corresponding to the beam ID or CRI, and may also include reliability information corresponding to different beams, etc.; since the number of predicted beams output at each moment may be greater than 4, and the predicted beam information at multiple moments may be reported in the same CSI reporting instance, and since the reporting requires more uplink resources, this embodiment considers prioritizing the predicted beam content reported by the user device. The specific priority division can refer to the reporting priority of the content in CSI part 2 in the existing standard. First, this embodiment provides the content that the predicted beam information fed back by the user device may contain under different feedback content, and then prioritizes each part of the content. The possible feedback content of the user device includes the following cases:

[0187] Case 1: Beam information of the predicted Top K beams in the beam set.

[0188] Case 2: Beam information of the predicted top K beams in the beam set, and RSRP information of the predicted top K beams in the beam set.

[0189] Case 3: Beam information of the predicted top K beams in the beam set, and probability information of the predicted top K beams in the beam set.

[0190] Case 4: Beam information of the top K beams predicted in the beam set, RSRP information of the top K beams predicted in the beam set, and confidence information of the RSRP information.

[0191] For Case 1, the predicted beam information fed back by the user equipment corresponds to specific beam ID information, such as SSBRI or CRI information, and may also be bitmap information for the predicted beam set.

[0192] For Case 2, the user equipment feedback information specifically includes the beam index information of the TopK beams and the RSRP information of the TopK beams in the beam set. At the same time, in order to indicate the RSRP information of the predicted beam, it also includes the position indication information of the maximum RSRP of all predicted beams.

[0193] For Case 2, if the number of predicted beams for different time periods fed back by the user equipment is variable, the information fed back by the user equipment may also include indication information of the number of beams predicted for different time periods or indication information indicating the number of predicted beams; it should be noted here that the number of beams predicted for each time period may correspond to a minimum value and / or a maximum value, and the specific number of predicted beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period; in addition, the total number of predicted beams for multiple time periods may be indicated by the base station, and the specific number of beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period.

[0194] For Case 3, the user equipment feedback information specifically includes the beam index information of the TopK beams and the probability information of the TopK beams in the beam set. The representation of the probability information may be based on S values ​​uniformly quantized from 0 to 1, with a quantization accuracy of 1 / S, indicating the number of bits required for each predicted beam. It may also be a judgment value based on a certain threshold. For example, if the confidence level is greater than a certain threshold, the corresponding indication information is 1, and if it is lower than a certain threshold, the corresponding indication information is 0.

[0195] For Case 3, if the number of predicted beams for different time periods fed back by the user equipment is variable, the information fed back by the user equipment may also include information on the number of beams predicted for different time periods or indication information indicating the number of predicted beams; it should be noted here that the number of beams predicted for each time period may correspond to a minimum value and / or a maximum value, and the specific number of predicted beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period; in addition, the total number of predicted beams for multiple time periods may be indicated by the base station, and the specific number of beams for each time period can be determined by the user equipment based on the results of the predicted beams for each time period.

[0196] For Case 4, the information fed back by the user equipment may include the confidence information of each predicted beam in addition to the information mentioned in Case 2. The confidence information may be represented by S values ​​uniformly quantized from 0 to 1, with a quantization accuracy of 1 / S, indicating the number of bits required for each predicted beam. It may also be a judgment value based on a certain threshold. For example, if the confidence level is greater than a certain threshold, the corresponding indication information is 1, and if it is lower than a certain threshold, the corresponding indication information is 0.

[0197] This embodiment considers dividing the beam prediction information reported by the user equipment into two groups. The content that each group may carry is also different for the above different cases. The following discusses different cases separately:

[0198] Case 1: For Case 1, this embodiment considers dividing the beam prediction information of multiple time periods reported by the user equipment into two groups, where Group 0 mainly carries the predicted beams of each time period with higher priority in the beam set. (or ) prediction beam index information; Group 1 mainly carries the prediction beam of each time period with lower priority in the beam set (or ) beam index information of the predicted beams.

[0199] Case 2: For Case 2, this embodiment considers dividing the beam prediction information of multiple time periods reported by the user equipment into two groups, where Group 0 mainly carries the predicted beams of each time period with higher priority in the beam set. (or ) predicted beams, and the beam set in each time period including the beams with higher priority (or ) beam index information corresponding to the prediction beams; Group 1 mainly carries the prediction beams with lower priority in the beam set for each time period (or ) predicted beams, and the beam set in each time period including the lower priority beams (or ) beam index information corresponding to the predicted beams.

[0200] Case 3: For Case 3, this embodiment considers dividing the beam prediction information of multiple time periods reported by the user equipment into two groups, where Group 0 mainly carries the predicted beams of each time period with higher priority in the beam set. (or ) prediction beam probability information, and the beam set in each time period includes the beams with higher priority (or ) beam index information corresponding to the prediction beams; Group 1 mainly carries the prediction beams with lower priority in the beam set for each time period (or ) prediction beam probability information, and the beam set in each time period includes the lower priority (or ) beam index information corresponding to the predicted beams.

[0201] Case 4: For Case 4, this embodiment considers dividing the beam prediction information of multiple time periods reported by the user equipment into two groups, where Group 0 mainly carries the predicted beam of each time period with a higher priority in the beam set. (or ) predicted beams’ RSRP information and confidence information, as well as the beam set with higher priority in each time period (or ) beam index information corresponding to the prediction beams; Group 1 mainly carries the prediction beams with lower priority in the beam set for each time period (or ) predicted beams’ RSRP information and confidence information, as well as the beam set in each time period including the beams with lower priority (or ) prediction beam index information in the beam set.

[0202] Furthermore, the RSRP information may be reported in a differential manner, and the specific differential method may be the following:

[0203] Method 1: In each time period, a predicted beam is selected whose corresponding RSRP is the largest. The maximum RSRP is used as a reference. The RSRPs of other predicted beams are differentiated from the maximum RSRP and quantized and reported. The maximum RSRP is quantized using an absolute value. For example, the maximum RSRP is quantized using 7 bits in the existing standard. For this method, the user equipment needs to report a maximum RSRP index for each time period.

[0204] Method 2: A maximum RSRP is selected for all time periods. This maximum RSRP is used as a reference. The RSRPs of all other time periods are differentially calculated from the maximum RSRP and quantized and reported. The maximum RSRP is quantized using an absolute value. For example, the existing standard quantizes the maximum RSRP using 7 bits. In this method, the user equipment needs to report a maximum RSRP index for all time periods.

[0205] Method 3: Select the maximum RSRP (global maximum RSRP) for all time periods and use it as a reference. Then, select the maximum RSRP (local maximum RSRP) for each time period except the time period containing the maximum RSRP. Subtract each local maximum RSRP from the global maximum RSRP, and report the difference in quantization. The difference can be quantized using fewer bits, such as 3 or 4 bits. Finally, in each time period, the RSRP corresponding to the predicted beam, excluding the local and global maximum RSRPs, is subtracted from the maximum RSRP in the time period. The difference is then reported in quantization, using 3 or 4 bits. For this method, the user equipment needs to report a maximum RSRP index information for all time periods; at the same time, it also needs to report a maximum RSRP index information for each time period except the time period where the global maximum RSRP is located; and the index information of the global maximum RSRP can be indicated in two ways. One is to select an indication in all beam sets in multiple time periods. For example, there are N time periods, and the number of beams contained in the beam set of each time period is M. Here, it is assumed that the beam set corresponding to each time period is the same, then the global maximum RSRP needs to be indicated. bits; the other is to first determine which time period is in the beam set, and then indicate the position in the beam set; for example, there are N time periods, and the number of beams contained in the beam set of each time period is M. Here, it is assumed that the beam set corresponding to each time period is the same, then the global maximum RSRP needs to be However, in some special cases, the time period of the global maximum RSRP may be indicated by default. If the global maximum RSRP is in a specific time period predefined by the standard, this information may not be indicated. For example, the global maximum RSRP is in the beginning or end of the time period. Furthermore, the above-mentioned method for indicating the global maximum RSRP information may also be applicable to scenarios where the user equipment is required to feedback the probability information and confidence information of the predicted beam.

[0206] Regardless of which of the above methods is used to report the RSRP information corresponding to the predicted beam in Case 2 and Case 3, the index information of the global and / or local maximum RSRP involved in the above method can be placed in Group 0; furthermore, the quantized value corresponding to the global and / or local maximum RSRP involved in the above method can also be placed in Group 0. In addition, if the reporting of confidence information and probability information also adopts a differential quantization form similar to RSRP reporting, it is also possible to include the above three methods. Therefore, the index information of the global and / or local maximum confidence or probability involved in the above method can be placed in Group 0; furthermore, the quantized value corresponding to the global and / or local maximum confidence or probability involved in the above method can also be placed in Group 0.

[0207] The priority information of the predicted beam for each time period can be divided as follows:

[0208] Taking the index of the predicted beam with the maximum RSRP value in the beam set as the starting point, assuming that the priority value corresponding to this point is 0, traverse to both sides of the beam set index respectively. The priority value is increased by 1 for each traversal of an index. If the traversal on one side is completed, continue to traverse to the single side until all beams in the beam set are traversed. The larger the priority value, the lower the priority. The priority traversal direction can be the side with the index greater than or less than the maximum RSRP, or when the difference between the maximum beam index and the index of the predicted beam with twice the maximum RSRP is negative, the left side is traversed first. If the above difference is positive, the right side is traversed first. If the difference is 0, both sides can be traversed. Specifically, as shown in Figure 4, for example, the maximum beam index of the beam set is 32, and the beam index corresponding to the maximum RSRP is 12. Obviously, 32-24=8 is greater than 0, so the right side is traversed first.

[0209] The above starting traversal position may also be based on the index position of the predicted beam corresponding to the maximum confidence in the beam set, or may be based on the index position of the predicted beam corresponding to the maximum probability in the beam set.

[0210] Furthermore, in addition to prioritizing the reporting of the prediction beam in each time period separately, the reporting priority information of the prediction beam can also be determined jointly for all prediction time periods. For example, on the basis of the method of prioritizing the reporting of the prediction beam in a single time period, if the prediction beam information of each time period includes RSRP information, then similar to the division of the prediction beam priority in a single time period, the position of the maximum RSRP corresponding to the reporting beam in different time periods can be determined, and then the position of the maximum RSRP corresponding to the reporting beam in different time periods can be traversed in chronological order, and the priority numbering is performed from small to large. After traversing the maximum RSRP corresponding to the reporting beam in all time periods, the position of the maximum RSRP can be determined. After the location of the largest RSRP, the second round of traversal begins, and the left or right beam of the location of the largest RSRP of the predicted beam in each time period is traversed in turn, and the priority number is numbered from small to large; after the RSRP location corresponding to the predicted beam of all time periods is traversed, the third round of traversal is started, and the right or left beam of the location of the largest RSRP of the predicted beam in each time period is traversed in turn, and the priority number is numbered from small to large; then, the second and third rounds of traversal are repeated until the predicted beams of all time periods are traversed; according to the above traversal rules, the smaller the number, the higher the priority, and then, the higher the priority. The prediction beams are placed in Group 0, and the remaining prediction beams are placed in Group 1, where K n Indicates the number of predicted beams in the nth time period; the above starting traversal position can also be the index position of the predicted beam corresponding to the maximum confidence in the beam set, or the index position of the predicted beam corresponding to the maximum probability in the beam set.

[0211] In addition, it should be noted that if the number of beams K predicted in different time periods is variable, for example, the number of beams predicted in different time periods is K1, K2, ..., K N , then the predicted beam in Group0 that will carry time period n∈{1,2,...,N} has a higher priority in the beam set (or ) prediction beam index information, or prediction beam index information and probability information, or prediction beam index information, confidence information and RSRP information; and Group 1 will carry the prediction beam with higher priority in the beam set in time period n∈{1,2,...,N} (or ) predicted beam index information, or the predicted beam index information and probability information, or the predicted beam index information, confidence information and RSRP information; furthermore, when the number K of beams predicted in different time periods is variable, for example, the number of beams predicted in different time periods is K1, K2, ..., K N , then the number of predicted beams in each time period needs to be placed in Group0; secondly, if the number of beams predicted in different time periods K is variable, for example, the number of beams predicted in different time periods is K1, K2, ..., K N , then the number of prediction beams in each time period needs to be placed in Group 0; and the higher priority of each prediction time period (or ) prediction beam index information, or prediction beam index information and probability information, or prediction beam index information, confidence information and RSRP information can be placed in Group 1, and Group 2 will carry the prediction beam with higher priority in the beam set in time period n∈{1,2,...,N} (or ) prediction beam index information, or the prediction beam index information and probability information, or the prediction beam index information, confidence information and RSRP information.

[0212] It should also be noted that if the number of beams K predicted in different time periods is variable, for example, the number of beams predicted in different time periods is K1, K2, ..., K N , the reported predicted beam information can be split into part 1 and part 2, where the number of predicted beams in each time period is carried in part 1, and other information is carried in part 2 in the manner mentioned in this embodiment. The priority of part 1 is higher than that of part 2, and the information in part 1 and part 2 follows the multiplexing rules of part 1 and part 2 of traditional CSI reporting information when performing UCI multiplexing. The priority and division of the content carried in part 2 are also the same as those in the previous part and will not be repeated here.

[0213] Seventh embodiment:

[0214] In some embodiments of the present application, the solution of the seventh embodiment may be implemented in conjunction with the solution of the first embodiment, the second embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the eighth embodiment, and / or the ninth embodiment, or may be implemented independently of the solution of the first embodiment, the second embodiment, the third embodiment, the fourth embodiment, the fifth embodiment, the sixth embodiment, the eighth embodiment, and the ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0215] Exemplarily, in some embodiments of the present application, for the prediction of spatial beams based on AI / ML, if the model is located on the terminal side, the terminal side model may output predicted 1 or multiple beam information, where the beam information may include the ID information of the beam, the RSRP information corresponding to the beam, and the confidence information or probability information that the predicted beam is a Top1 or TopK beam; similarly, for the scenario where the beam information at multiple future moments is predicted based on the beam measurement information at historical moments, the terminal side model may also output predicted beam information for multiple future moments, which may include the ID information or CRI of multiple beams at multiple predicted future moments, and the RSRP information corresponding to the beam ID or CRI, and may also include the confidence information or probability information that the predicted beam is a Top1 or TopK beam.

[0216] Beam confidence or probability information can serve as a basis for the base station to further obtain the optimal beam. It can also be used as a measure for the base station to measure the credibility of the predicted beam information fed back by the terminal. However, the confidence or probability information feedback needs to be quantified before being fed back to the base station. We provide the following possible quantification methods:

[0217] Solution 1: Equal interval quantization, the maximum and minimum probability values ​​or confidence values ​​can be subtracted and divided into N equal parts, through bits are used to represent the confidence or probability information of each predicted beam; for example, 1 or 100% is divided into 8 equal parts, and 3 bits are used to represent the confidence information of each beam.

[0218] Solution 2: Based on a threshold value, determine whether the confidence or probability information of the output beam is lower than the threshold value or higher than the threshold value, and represent it with 1 bit. If the corresponding bit value is 0, it means it is lower than the threshold value. If the corresponding bit value is 1, it means it is higher than the threshold value. The threshold value can be determined by the base station side, or it can be predefined by the standard, or it can be determined by the terminal side and informed to the base station side. In this way, the confidence or probability information corresponding to each beam only needs 1 bit to represent it, and the base station side may not need to know the specific confidence or probability information, and may only determine the selection of those predicted beams based on a larger range of values. Based on the above purpose, this method can effectively reduce the overhead of terminal indication.

[0219] Solution 3: Using a differential reporting method, with the maximum confidence or probability value as the reference value, the maximum confidence or probability value in all time periods is selected and differentiated from the reference value. Quantization is then performed to determine the maximum confidence or probability value in all time periods. The confidence or probability values ​​corresponding to the remaining prediction beams are then differentiated from the maximum confidence or probability values ​​in all time periods, and the differential results are quantitatively reported. For example, using 1 or 100% as the reference value, the maximum confidence or probability value in all time periods is selected and differentiated from 1 or 100%, and the differential results are quantitatively reported (upward quantization). The confidence or probability values ​​corresponding to the remaining prediction beams are then differentiated from the maximum confidence or probability values ​​in all time periods, and the differential results are quantitatively reported. The aforementioned differential quantization reporting can also directly select the maximum confidence or probability value across all time segments and align it for absolute quantization. For example, 1 or 100% is used as the reference value for the maximum value, and the segment is divided equally. Based on the equal division result, the maximum confidence or probability value across all time segments is quantized. The confidence or probability value corresponding to the remaining prediction beam is then differentiated from the maximum confidence or probability value across all time segments, and the difference result is quantized and reported. The bit overhead required for absolute quantization is related to the equal division performed, while relative differential quantization can use fewer bits to represent it.

[0220] Solution 4: Select a maximum value from the confidence values ​​or probability values ​​in the prediction beam information corresponding to multiple time periods as a reference value. Then, select a maximum confidence value or probability value from the confidence values ​​or probability values ​​corresponding to the prediction beam in each time period and differentiate it from the maximum confidence value or probability value of the prediction beam in multiple time periods. Align and perform quantitative feedback. At the same time, differentiate the confidence value or probability value in each time period from the maximum confidence value or probability value of its own prediction beam, and report the difference quantitatively. The maximum confidence value or probability value in the prediction beam information corresponding to all time periods in this solution can also be differentially quantified and reported using the method described in Solution 3. Subsequent processing continues to use the solution in Solution 4.

[0221] Eighth embodiment:

[0222] In some embodiments of the present application, the solution of the eighth embodiment may be implemented in conjunction with the solution of the first, second, third, fourth, fifth, sixth, seventh, and / or ninth embodiment, or may be implemented independently of the solution of the first, second, third, fourth, fifth, sixth, seventh, and ninth embodiment. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0223] For example, in some embodiments of the present application, for beam prediction based on AI / ML, for beam prediction of multiple future time periods, if the model is on the terminal side, the terminal may need to report the predicted beam information of multiple future moments at the same time. At the same time, the terminal may also report the beam information obtained by traditional measurements, including SSBRI or CRI and corresponding RSRP information; if the predicted beam information of multiple future moments conflicts with the beam information obtained by traditional measurements, the instance believes that the measured beam information should be reported first. The specific priority division can be reflected by the following formula:

[0224] Pri iCSI (y,k,c,s)=2·N cells ·M s y+N cells ·M s k+M s c+s

[0225] The smaller the calculated parameter is, the higher the priority is. Other parameters are as follows:

[0226] 1) When the reported CSI is aperiodic CSI (A-CSI), y=0; when the reported CSI is semi-persistent CSI (SP-CSI) and is transmitted on PUSCH, y=1; when the reported CSI is semi-persistent CSI (SP-CSI) and is transmitted on PUCCH, y=2; when the reported CSI is periodic CSI (SP-CSI), y=3.

[0227] 2) If L1-RSRP or L1-SINR is carried and obtained through measurement, the corresponding k=0; and if the predicted beam information of multiple future time periods carrying L1-RSRP or L1-SINR is reported, the corresponding K value is K=1; in other cases, K=2.

[0228] 3) c: cell index; N_{cells}: number of cells, configured through high-level parameters maxNrofServingCells.

[0229] 4)s:reportConfigID;M_{s}:maxNrofCSI-ReportConfigurations.

[0230] At the same time, the method in this embodiment is also applicable to the scenarios in which the predicted beam information of multiple future time periods in the third and fourth embodiments is divided into multiple sub-reporting instances or groups. It is only necessary to modify the value of K corresponding to different reporting instances or groups.

[0231] Ninth embodiment:

[0232] In some embodiments of the present application, the solution of the ninth embodiment may be implemented in conjunction with the solution of the first, second, third, fourth, fifth, sixth, seventh, and / or eighth embodiments, or may be implemented independently of the solution of the first, second, third, fourth, fifth, sixth, seventh, and eighth embodiments. In some embodiments of the present application, the solutions of multiple embodiments may be implemented in combination or independently.

[0233] Exemplarily, in some embodiments of the present application, for the prediction of spatial beams based on AI / ML, if the model is located on the terminal side, the terminal side model may output predicted 1 or multiple beam information, where the beam information may include the ID information of the beam, the RSRP information corresponding to the beam, and the confidence information or probability information that the predicted beam is a Top1 or TopK beam; similarly, for the scenario where the beam information at multiple future moments is predicted based on the beam measurement information at historical moments, the terminal side model may also output predicted beam information for multiple future moments, which may include the ID information or CRI of multiple beams at multiple predicted future moments, and the RSRP information corresponding to the beam ID or CRI, and may also include the confidence information or probability information that the predicted beam is a Top1 or TopK beam.

[0234] For spatial beam prediction, the predicted beam and the measured beam may be the same. In this case, only the measured beam needs to be reported. This may result in both predicted and measured beam information being included in one beam report. Assume that the possible feedback content of the terminal includes the following cases:

[0235] Case 1: Beam information of the predicted Top K beams in the beam set.

[0236] Case 2: Beam information of the predicted top K beams in the beam set, and RSRP information of the predicted top K beams in the beam set.

[0237] Case 3: Beam information of the predicted top K beams in the beam set, and probability information of the predicted top K beams in the beam set.

[0238] Case 4: Beam information of the top K beams predicted in the beam set, RSRP information of the top K beams predicted in the beam set, and confidence information of the RSRP information.

[0239] For Case 1, there is no difference between the predicted beams and the measured beams reported by the terminal. If the network side needs to be informed of which beams are measured beams, the network side can be informed implicitly. For example, if the terminal uses the traditional SSBRI or CRI reporting method, the terminal reports the number indication information to inform the network side that there are M measured beams, and then puts the information of the measured beams in the position of the first M beams to inform the network; the terminal can also use 1 bit to indicate whether each beam is a measured beam or a predicted beam. A bit of 1 or 0 indicates that the corresponding beam is a measured beam.

[0240] For Case 2, the reporting of beam information is the same as that for Case 1; however, the RSRP information corresponding to the measured beam needs to be used.

[0241] For Case 3, the reporting of beam information is the same as that of Case 1; however, the probability information can be omitted from reporting.

[0242] For Case 4, the reporting of beam information is the same as that for Case 1; however, the RSRP information corresponding to the measured beam needs to be used; and the confidence information can be defaulted to not be reported.

[0243] Figure 5 is a schematic structural diagram of a wireless communication device 700 provided in an embodiment of the present application. The wireless communication device can be a user equipment, a base station, or a network element. The wireless communication device 700 shown in Figure 5 includes a processor 710, which can call and execute a computer program from a memory to implement the method in the embodiment of the present application.

[0244] Optionally, as shown in FIG6 , the wireless communication device 700 may further include a memory 720. The processor 710 may call and execute a computer program from the memory 720 to implement the method in the embodiment of the present application. The memory 720 may be a separate device independent of the processor 710 or may be integrated into the processor 710.

[0245] Optionally, as shown in FIG5 , the wireless communication device 700 may further include a transceiver 730. The processor 710 may control the transceiver 730 to communicate with other devices. Specifically, the transceiver 730 may send information or data to other devices or receive information or data sent by other devices. The transceiver 730 may include a transmitter and a receiver. The transceiver 730 may further include one or more antennas.

[0246] Optionally, the wireless communication device 700 may specifically be the network device 110 of the embodiment of the present application, and the wireless communication device 700 may implement the corresponding processes implemented by the network device 110 in each method of the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0247] Optionally, the wireless communication device 700 may specifically be a user equipment of an embodiment of the present application, and the wireless communication device 700 may implement the corresponding processes implemented by the user equipment in each method of the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0248] Optionally, the wireless communication device 700 may specifically be a network element in an embodiment of the present application, and the wireless communication device 700 may implement the corresponding processes implemented by the network element in each method in the embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0249] Figure 6 is a schematic structural diagram of a chip according to an embodiment of the present application. The chip 800 shown in Figure 6 includes a processor 810, which can call and run a computer program from a memory to implement the method according to the embodiment of the present application.

[0250] Optionally, as shown in FIG6 , the chip 800 may further include a memory 820. The processor 810 may call and execute computer programs from the memory 820 to implement the methods in the embodiments of the present application. The memory 820 may be a separate device independent of the processor 810 or may be integrated into the processor 810.

[0251] Optionally, the chip 800 may further include an input interface 830. The processor 910 may control the input interface 830 to communicate with other devices or chips, and specifically, may obtain information or data sent by other devices or chips.

[0252] Optionally, the chip 800 may further include an output interface 840. The processor 810 may control the output interface 840 to communicate with other devices or chips, and specifically, may output information or data to other devices or chips.

[0253] Optionally, the chip can be applied to the network device 110 in the embodiment of the present application, and the chip can implement the corresponding processes implemented by the network device 110 in each method of the embodiment of the present application. For the sake of brevity, it will not be repeated here.

[0254] Optionally, the chip can be applied to the user equipment in the embodiments of the present application, and the chip can implement the corresponding processes implemented by the user equipment in the various methods of the embodiments of the present application. For the sake of brevity, they will not be repeated here.

[0255] Optionally, the chip can be applied to the network element in the embodiment of the present application, and the chip can implement the corresponding processes implemented by the mobile network element in each method of the embodiment of the present application. For the sake of brevity, it will not be repeated here.

[0256] Figure 7 is a schematic block diagram of a wireless communication system 100 provided in an embodiment of the present application. As shown in Figure 7, the communication system 100 includes a user device 120 and a network device 110. The user device 120 can be used to implement the corresponding functions implemented by the user device 120 in the above method, and the network device 110 can be used to implement the corresponding functions implemented by the network device 110 in the above method. For the sake of brevity, these functions are not further described here.

[0257] It should be understood that the processor of the embodiment of the present application may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiment may be completed by hardware integrated logic circuits in the processor or software instructions.

[0258] It is understood that the memory in the embodiments of the present application may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory. The embodiments of the present application also provide a computer-readable storage medium for storing a computer program.

[0259] Optionally, the computer-readable storage medium may be applied to a network device in an embodiment of the present application, and the computer program causes a computer to execute the corresponding processes implemented by the network device in each method in the embodiment of the present application. For the sake of brevity, no further description is given here. Optionally, the computer-readable storage medium may be applied to a user device in an embodiment of the present application, and the computer program causes a computer to execute the corresponding processes implemented by the user device in each method in the embodiment of the present application. For the sake of brevity, no further description is given here.

[0260] An embodiment of the present application also provides a computer program product, including computer program instructions.

[0261] Optionally, the computer program product may be applied to the network device in the embodiments of the present application, and the computer program instructions cause the computer to execute the corresponding processes implemented by the network device in the various methods of the embodiments of the present application. For the sake of brevity, no further description is given here. Optionally, the computer program product may be applied to the user equipment in the embodiments of the present application, and the computer program instructions cause the computer to execute the corresponding processes implemented by the user equipment in the various methods of the embodiments of the present application. For the sake of brevity, no further description is given here.

[0262] The embodiment of the present application also provides a computer program.

[0263] Optionally, the computer program may be applied to the network device in the embodiments of the present application. When the computer program is run on a computer, the computer executes the corresponding processes implemented by the network device in the various methods of the embodiments of the present application. For the sake of brevity, no further details are given here. Optionally, the computer program may be applied to the user equipment in the embodiments of the present application. When the computer program is run on a computer, the computer executes the corresponding processes implemented by the user equipment in the various methods of the embodiments of the present application. For the sake of brevity, no further details are given here.

[0264] Those skilled in the art will appreciate that the units and algorithm steps of each example 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 performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel 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.

[0265] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A beam management method based on artificial intelligence / machine learning AI / ML, executed on a user equipment, wherein: The AI / ML-based beam management method includes: The user device initiates a conversion between a first mode and a second mode to perform AI / ML-based beam prediction, wherein the first mode means that the AI / ML-based beam prediction supports beam prediction based on a user device side model, and the second mode means that the AI / ML-based beam prediction supports beam prediction based on a network device side model.

2. The AI / ML-based beam management method according to claim 1, wherein: The first mode is used as a basic feature of the user equipment, and the second mode is used as a capability item of the user equipment; or the second mode is used as a basic feature of the user equipment, and the first mode is used as a capability item of the user equipment.

3. The AI / ML-based beam management method according to claim 1, wherein: If the currently running AI / ML model applicable to the first function is located on the user device, the conversion between the first mode and the second mode is initiated by the user device.

4. The AI / ML-based beam management method according to claim 3, wherein: If the user device detects that the output result of the currently running AI / ML model applicable to the first function is less than a predefined first threshold, the user device sends a first deactivation request to the network device to request the network device to deactivate the currently running AI / ML model applicable to the first function.

5. The AI / ML-based beam management method according to claim 4, wherein: The output result includes the confidence of the predicted beam or the reference signal received power RSRP value of the predicted beam.

6. The AI / ML-based beam management method according to claim 4, wherein: The first deactivation request occupies 1 bit, and the 1 bit is used to indicate a request to deactivate the currently running AI / ML model applicable to the first function.

7. The AI / ML-based beam management method according to claim 1, wherein: If the currently running AI / ML model applicable to the first function is located on the network device, the conversion between the first mode and the second mode is initiated by the user device.

8. The AI / ML-based beam management method according to claim 7, wherein: If the user device detects that the reporting amount of the currently running AI / ML model applicable to the first function is greater than a predefined second threshold, the user device sends a second deactivation request to the network device to request the network device to deactivate the currently running AI / ML model applicable to the first function.

9. The AI / ML-based beam management method according to claim 8, wherein: The second deactivation request occupies 1 bit, and the 1 bit is used to indicate a request to deactivate the currently running AI / ML model applicable to the first function.

10. A beam management method based on artificial intelligence / machine learning AI / ML, executed on a user equipment, wherein: The beam management method comprises: The user equipment outputs predicted beam information for a plurality of time periods to perform AI / ML-based beam prediction, wherein the AI / ML-based beam prediction supports beam prediction based on a user equipment-side model; and The user equipment reports the predicted beam information of the multiple time periods to the network device in a channel state information (CSI) instance, wherein the one CSI instance includes multiple sub-reporting instances, and the user equipment reports the multiple sub-reporting instances to the network device according to different reporting priorities.

11. The AI / ML-based beam management method according to claim 10, wherein: The reporting priority decreases as the numbers of the multiple sub-reporting instances increase.

12. The AI / ML-based beam management method according to claim 11, wherein: The CSI instance includes: Beam information of the K beams predicted in the beam set, where K is a positive integer greater than or equal to 1; Beam information of the predicted K beams in the beam set, and RSRP information of the predicted K beams in the beam set; or Beam information of the predicted K beams in the beam set, and probability information of the predicted K beams in the beam set; or Beam information of the predicted K beams in the beam set, RSRP information of the predicted K beams in the beam set, and confidence information of the RSRP information.

13. The AI / ML-based beam management method according to claim 12, wherein: The K values ​​in different time periods are the same or different.

14. The AI / ML-based beam management method according to claim 12, wherein: The beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the RSRP information of the predicted K beams in the beam set is carried in a second sub-reporting instance.

15. The AI / ML-based beam management method according to claim 12, wherein: The beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the probability information of the predicted K beams in the beam set is carried in a second sub-reporting instance.

16. The AI / ML-based beam management method according to claim 12, wherein: The beam information of the predicted K beams in the beam set and the RSRP information of the predicted K beams in the beam set are carried in a first sub-reporting instance, and the confidence information of the RSRP information is carried in a second sub-reporting instance.

17. The AI / ML-based beam management method according to claim 12, wherein: The beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the RSRP information of the predicted K beams in the beam set and the confidence information of the RSRP information are carried in a second sub-reporting instance.

18. The AI / ML-based beam management method according to claim 12, wherein: The beam information of the predicted K beams in the beam set is carried in the first sub-reporting instance, the RSRP information of the predicted K beams in the beam set is carried in the second sub-reporting instance, and the confidence information of the RSRP information is carried in the third sub-reporting instance.

19. The AI / ML-based beam management method according to claim 12, wherein: The predicted beam information of the multiple time periods is the predicted beam information of N time periods, the predicted beam information of X time periods is carried in the first sub-reporting instance, and the predicted beam information of NX time periods is carried in the second sub-reporting instance, where N and X are positive integers, and N is greater than or equal to X.

20. The AI / ML-based beam management method according to claim 12, wherein: The beam information corresponding to the even beams in the beam set is carried in the first sub-reporting instance, and the beam information corresponding to the odd beams in the beam set is carried in the second sub-reporting instance; or the beam information corresponding to the odd beams in the beam set is carried in the first sub-reporting instance, and the beam information corresponding to the even beams in the beam set is carried in the second sub-reporting instance.

21. The AI / ML-based beam management method according to claim 12, wherein: The index corresponding to the reference value of the probability information, the RSRP information or the confidence information and the corresponding quantization value are carried in the first sub-reporting instance.

22. The AI / ML-based beam management method according to claim 12, wherein: The predicted beam in each time period has a higher priority in the beam set. or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

23. The AI / ML-based beam management method according to claim 12, wherein: The predicted beam in each time period has a higher priority in the beam set. or The RSRP information of the predicted beams and the predicted beams in each time period have higher priority in the beam set or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The RSRP information of the predicted beams and the predicted beams in each time period have lower priorities in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

24. The AI / ML-based beam management method according to claim 12, wherein: The predicted beam in each time period has a higher priority in the beam set. or The probability information of the predicted beams and the predicted beams in each time period have higher priority in the beam set or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The probability information of the predicted beams and the predicted beams of each time period with lower priority in the beam set or The beam index information of the predicted beam is carried in the second sub-reporting instance.

25. The AI / ML-based beam management method according to claim 12, wherein: The predicted beam in each time period has a higher priority in the beam set. or The RSRP information and confidence information of the predicted beams and the predicted beams in each time period have higher priority in the beam set. or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The RSRP information and confidence information of the predicted beams and the predicted beams in each time period have lower priorities in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

26. The AI / ML-based beam management method according to claim 12, wherein: The index and the corresponding quantization value corresponding to the global and / or local maximum probability value, the global and / or local maximum RSRP value or the global and / or local maximum confidence value are carried in the first sub-reporting instance.

27. The AI / ML-based beam management method according to claim 12, wherein: The index corresponding to the maximum probability value, maximum RSRP value or maximum confidence value of the predicted beam in each time period is the starting point, and the priority value corresponding to the starting point is the first value.

28. The AI / ML-based beam management method according to claim 27, wherein: The indexes of the beams on the first side of the starting point and the indexes of the beams on the second side of the starting point are traversed alternately, and the priority value is increased by 1 each time the index of a beam is traversed, until all beams in the beam set are traversed.

29. The AI / ML-based beam management method according to claim 28, wherein: Alternating traversal of the index of the beam toward the first side of the starting point and the index of the beam toward the second side of the starting point includes sequentially traversing the index of the first beam toward the first side of the starting point and the index of the first beam toward the second side of the starting point, and then sequentially traversing the index of the second beam toward the first side of the starting point and the index of the second beam toward the second side of the starting point.

30. The AI / ML-based beam management method according to claim 12, wherein: If the number of beams in the beam set in each time period is M, M is a positive integer, and the number of predicted beams is K, then the positions of the K beams in the beam set are expressed as the combination number: That is, the K beams are selected from the M beams, and the number of combinations represents the occupied bits.

31. The AI / ML-based beam management method according to claim 12, wherein: If the number of beams in the beam set in each time period is M, the M beams are equally divided into N beam segments, M and N are positive integers, and the number of beams contained in each beam segment is M / N, indicating that each beam segment occupies M / N bits.

32. The AI / ML-based beam management method according to claim 12 or 31, wherein: N bits are used to indicate whether each beam in each time period has a predicted beam. If L bits among the N bits are 0, it indicates that the beam set occupies (NL)M / N bits.

33. The AI / ML-based beam management method according to claim 12, wherein: A reference beam is set in the beam set in each time period, and first length information is indicated for the reference beam. The first length information is relative to the positions on both sides of the reference beam, and the number of bits occupied by the position of the predicted beam in the beam set is determined based on the first length information.

34. The AI / ML-based beam management method according to claim 33, wherein: If the beam set of each time period has a plurality of candidate reference beam positions, one candidate reference beam is selected from the plurality of candidate reference beams based on the predicted beam, and second length information is indicated based on the one candidate reference beam.

35. The AI / ML-based beam management method according to claim 12, wherein: The indication information of the CSI instance includes: The predicted beam corresponding to the maximum RSRP value in each time period is used as the reference point; or The predicted beam corresponding to the maximum probability value in each time period is used as the reference point; or The predicted beam corresponding to the maximum RSRP value or the maximum confidence value in each time period is used as the reference point.

36. The AI / ML-based beam management method according to claim 12, wherein: When the number K of predicted beams is greater than or equal to a predefined threshold, the positions of the predicted K beams in the beam set are fed back in a bitmap manner, and the length of the bitmap is the number of beams in the beam set.

37. The AI / ML-based beam management method according to claim 12, wherein: When the number K of predicted beams is less than or equal to a predefined threshold, the positions of the predicted K beams in the beam set are fed back in the form of beam identifiers IDs.

38. The AI / ML-based beam management method according to claim 12, wherein: The feedback of the predicted positions of the K beams in the beam set by means of beam ID includes feedback of the predicted positions of the K beams in the beam set by means of synchronization signal block resource indicator SSBRI or channel state information reference signal resource indicator CRI.

39. A beam management method based on artificial intelligence / machine learning AI / ML, executed on a network device, wherein: The AI / ML-based beam management method includes: The network device initiates a conversion between a first mode and a second mode to perform AI / ML-based beam prediction, wherein the first mode refers to the AI / ML-based beam prediction supporting beam prediction based on a user equipment side model, and the second mode refers to the AI / ML-based beam prediction supporting beam prediction based on a user equipment side model. AI / ML beam prediction supports beam prediction based on network device-side models.

40. The AI / ML-based beam management method according to claim 39, wherein: The first mode serves as a basic feature of the network device, and the second mode serves as a capability item of the network device; or the second mode serves as a basic feature of the network device, and the first mode serves as a capability item of the network device.

41. The AI / ML-based beam management method according to claim 39, wherein: If the currently running AI / ML model applicable to the first function is located on the user device, the transition between the first mode and the second mode is initiated by the network device.

42. The AI / ML-based beam management method according to claim 41, wherein: If the network device detects that the output result of the currently running AI / ML model applicable to the first function is less than a predefined first threshold, the network device sends a first deactivation command to the UE to deactivate the currently running AI / ML model applicable to the first function.

43. The AI / ML-based beam management method according to claim 42, wherein: The output result includes the confidence of the predicted beam or the reference signal received power RSRP value of the predicted beam.

44. The AI / ML-based beam management method according to claim 42, wherein: The first deactivation command occupies 1 bit, and the 1 bit is used to indicate whether to deactivate the currently running AI / ML model applicable to the first function.

45. The AI / ML based beam management method according to claim 39, wherein: If the currently running AI / ML model applicable to the first function is located on a network device, the conversion between the first mode and the second mode is initiated by the network device.

46. ​​The AI / ML-based beam management method according to claim 45, wherein: If the network device monitors that the reporting amount of the currently running AI / ML model applicable to the first function is greater than a predefined second threshold, the network device sends a second deactivation command to the UE to deactivate the currently running AI / ML model applicable to the first function.

47. The AI / ML-based beam management method according to claim 46, wherein: The second deactivation command occupies 1 bit, and the 1 bit is used to indicate whether to deactivate the currently running AI / ML model applicable to the first function.

48. A beam management method based on artificial intelligence / machine learning AI / ML, executed on a network device, wherein: The beam management method comprises: The network device receives predicted beam information for multiple time periods output by the user equipment to perform AI / ML-based beam prediction, wherein the AI / ML-based beam prediction supports beam prediction based on a user equipment side model; and The network device receives the predicted beam information of the multiple time periods reported by the user equipment in a channel state information (CSI) instance, wherein the one CSI instance includes multiple sub-reporting instances, and the network device receives the multiple sub-reporting instances reported by the user equipment according to different reporting priorities.

49. The AI / ML-based beam management method according to claim 48, wherein: The reporting priority decreases as the numbers of the multiple sub-reporting instances increase.

50. The AI / ML-based beam management method according to claim 49, wherein: The CSI instance includes: Beam information of the K beams predicted in the beam set, where K is a positive integer greater than or equal to 1; Beam information of the predicted K beams in the beam set, and RSRP information of the predicted K beams in the beam set; or Beam information of the predicted K beams in the beam set, and the probability of the predicted K beams in the beam set information; or Beam information of the predicted K beams in the beam set, RSRP information of the predicted K beams in the beam set, and confidence information of the RSRP information.

51. The AI / ML-based beam management method according to claim 50, wherein: The K values ​​in different time periods are the same or different.

52. The AI / ML-based beam management method of claim 50, wherein: The beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the RSRP information of the predicted K beams in the beam set is carried in a second sub-reporting instance.

53. The AI / ML-based beam management method according to claim 50, wherein: The beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the probability information of the predicted K beams in the beam set is carried in a second sub-reporting instance.

54. The AI / ML-based beam management method according to claim 50, wherein: The beam information of the predicted K beams in the beam set and the RSRP information of the predicted K beams in the beam set are carried in a first sub-reporting instance, and the confidence information of the RSRP information is carried in a second sub-reporting instance.

55. The AI / ML based beam management method according to claim 50, wherein: The beam information of the predicted K beams in the beam set is carried in a first sub-reporting instance, and the RSRP information of the predicted K beams in the beam set and the confidence information of the RSRP information are carried in a second sub-reporting instance.

56. The AI / ML based beam management method according to claim 50, wherein: The beam information of the predicted K beams in the beam set is carried in the first sub-reporting instance, the RSRP information of the predicted K beams in the beam set is carried in the second sub-reporting instance, and the confidence information of the RSRP information is carried in the third sub-reporting instance.

57. The AI / ML based beam management method according to claim 50, wherein: The predicted beam information of the multiple time periods is the predicted beam information of N time periods, the predicted beam information of X time periods is carried in the first sub-reporting instance, and the predicted beam information of NX time periods is carried in the second sub-reporting instance, where N and X are positive integers, and N is greater than or equal to X.

58. The AI / ML-based beam management method according to claim 50, wherein: The beam information corresponding to the even beams in the beam set is carried in the first sub-reporting instance, and the beam information corresponding to the odd beams in the beam set is carried in the second sub-reporting instance; or the beam information corresponding to the odd beams in the beam set is carried in the first sub-reporting instance, and the beam information corresponding to the even beams in the beam set is carried in the second sub-reporting instance.

59. The AI / ML-based beam management method according to claim 50, wherein: The index corresponding to the reference value of the probability information, the RSRP information or the confidence information and the corresponding quantization value are carried in the first sub-reporting instance.

60. The AI / ML based beam management method according to claim 50, wherein: The predicted beam in each time period has a higher priority in the beam set. or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

61. The AI / ML-based beam management method of claim 50, wherein: The predicted beam in each time period has a higher priority in the beam set. or The RSRP information of the predicted beams and the predicted beams in each time period have higher priority in the beam set or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The RSRP information of the predicted beams and the predicted beams in each time period have lower priorities in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

62. The AI / ML based beam management method according to claim 50, wherein: The predicted beam in each time period has a higher priority in the beam set. or The probability information of the predicted beams and the predicted beams in each time period have higher priority in the beam set or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The probability information of the predicted beams and the predicted beams of each time period with lower priority in the beam set or The beam index information of the predicted beam is carried in the second sub-reporting instance.

63. The AI / ML based beam management method according to claim 50, wherein: The predicted beam in each time period has a higher priority in the beam set. or The RSRP information and confidence information of the predicted beams and the predicted beams in each time period have higher priority in the beam set. or The index information of the predicted beams is carried in the first sub-reporting instance, and the predicted beams in each time period have a lower priority in the beam set. or The RSRP information and confidence information of the predicted beams and the predicted beams in each time period have lower priorities in the beam set. or The beam index information of the predicted beam is carried in the second sub-reporting instance.

64. The AI / ML based beam management method according to claim 50, wherein: The index and the corresponding quantization value corresponding to the global and / or local maximum probability value, the global and / or local maximum RSRP value or the global and / or local maximum confidence value are carried in the first sub-reporting instance.

65. The AI / ML based beam management method according to claim 50, wherein: The index corresponding to the maximum probability value, maximum RSRP value or maximum confidence value of the predicted beam in each time period is the starting point, and the priority value corresponding to the starting point is the first value.

66. The AI / ML based beam management method according to claim 65, wherein: The indexes of the beams on the first side of the starting point and the indexes of the beams on the second side of the starting point are traversed alternately, and the priority value is increased by 1 each time the index of a beam is traversed, until all beams in the beam set are traversed.

67. The AI / ML based beam management method according to claim 66, wherein: Alternating traversal of the index of the beam toward the first side of the starting point and the index of the beam toward the second side of the starting point includes sequentially traversing the index of the first beam toward the first side of the starting point and the index of the first beam toward the second side of the starting point, and then sequentially traversing the index of the second beam toward the first side of the starting point and the index of the second beam toward the second side of the starting point.

68. The AI / ML based beam management method according to claim 50, wherein: If the number of beams in the beam set in each time period is M, M is a positive integer, and the number of predicted beams is K, then the positions of the K beams in the beam set are expressed as the combination number: That is, the K beams are selected from the M beams, and the number of combinations represents the occupied bits.

69. The AI / ML based beam management method according to claim 50, wherein: If the number of beams in the beam set in each time period is M, the M beams are equally divided into N beam segments, M and N are positive integers, and the number of beams contained in each beam segment is M / N, indicating that each beam segment occupies M / N bits.

70. The AI / ML based beam management method according to claim 50 or 69, wherein: N bits are used to indicate whether each beam in each time period has a predicted beam. If L bits among the N bits are 0, it indicates that the beam set occupies (NL)M / N bits.

71. The AI / ML based beam management method according to claim 50, wherein: A reference beam is set in the beam set in each time period, and first length information is indicated for the reference beam. The first length information is relative to the positions on both sides of the reference beam, and the number of bits occupied by the position of the predicted beam in the beam set is determined based on the first length information.

72. The AI / ML based beam management method according to claim 71, wherein: If the beam set of each time period has a plurality of candidate reference beam positions, one candidate reference beam is selected from the plurality of candidate reference beams based on the predicted beam, and second length information is indicated based on the one candidate reference beam.

73. The AI / ML based beam management method according to claim 50, wherein: The indication information of the CSI instance includes: The predicted beam corresponding to the maximum RSRP value in each time period is used as the reference point; or The predicted beam corresponding to the maximum probability value in each time period is used as the reference point; or The predicted beam corresponding to the maximum RSRP value or the maximum confidence value in each time period is used as the reference point.

74. The AI / ML based beam management method according to claim 50, wherein: When the number K of predicted beams is greater than or equal to a predefined threshold, the positions of the predicted K beams in the beam set are fed back in a bitmap manner, and the length of the bitmap is the number of beams in the beam set.

75. The AI / ML based beam management method according to claim 50, wherein: When the number K of predicted beams is less than or equal to a predefined threshold, the positions of the predicted K beams in the beam set are fed back in the form of beam identifiers ID.

76. The AI / ML based beam management method of claim 50, wherein: The feedback of the predicted positions of the K beams in the beam set by means of beam ID includes feedback of the predicted positions of the K beams in the beam set by means of synchronization signal block resource indicator SSBRI or channel state information reference signal resource indicator CRI.

77. A wireless communication device comprising: A processor and a memory, the memory being used to store a computer program, the processor being used to call and run the computer program stored in the memory to execute the method as claimed in any one of claims 1 to 76.

Images