Communication method and apparatus

Abstract

The present application relates to the field of communication technologies and provides a communication method and apparatus for improving network performance. A first network element acquires a first measurement to be input and acquires a first AI model. The first network element then inputs the first measurement to the first AI model to obtain a first policy output by the first AI model. The first network element then executes the first policy. When the first network element determines that an exception occurs when the first policy is executed, the first network element sends instruction information to a second network element indicating that the exception occurred when the first policy was executed. A network element may recognize the execution status of another network element's execution of a policy obtained based on the AI ​​model, thereby improving network performance to some extent. In particular, when the network element determines that an exception occurs when the first policy is executed, the network element notifies the other network element of this situation, allowing the other network element to respond based on the exception and thereby improve network performance.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to Chinese Patent Application No. 202211186062.7, entitled "Communication Method and Apparatus," filed with the State Intellectual Property Office of China on September 27, 2022, the entire contents of which are incorporated herein by reference. Technical Field TECHNICAL FIELD Embodiments of the present application relate to the field of communication technologies, and more particularly to communication methods and devices. [Background technology]

[0002] Artificial intelligence (AI) is a technology that simulates the human brain to perform complex computing. Applying AI to wireless communications can improve network performance and user experience by intelligently collecting and analyzing data. The AI ​​model training process can be performed on core network devices or access network devices. Access network devices can use the trained AI model to obtain energy saving policies, load balancing policies, and other policies and implement the corresponding policies. Currently, each access network device independently determines and implements policies based on the AI ​​model, which limits network performance improvements. Summary of the Invention [Problem to be solved by the invention]

[0003] SUMMARY OF THE INVENTION Embodiments of the present application provide a communication method and apparatus for improving network performance. [Means for solving the problem]

[0004] According to a first aspect, a communication method is provided. The method may be executed by a first network element or a component used in the first network element, such as a chip or a processor. The following provides an example in which the method is executed by the first network element. First, the first network element acquires a first measurement to be input and acquires a first AI model. Then, the first network element inputs the first measurement to the first AI model to obtain a first policy output by the first AI model. Next, the first network element executes the first policy. Then, when the first network element determines that an exception occurs when the first policy is executed, the first network element sends instruction information to a second network element, where the instruction information indicates that an exception occurs when the first policy is executed.

[0005] When executing the first policy determined based on the first AI model, the first network element may detect the execution status and notify the second network element of the execution status. In this way, the network element may be aware of the execution status of the policy obtained based on the AI ​​model by another network element, thereby improving network performance to some extent. In particular, when the network element determines that an exception occurs when the first policy is executed, the network element notifies the other network element of this situation, so that the other network element can respond based on the exception, thereby improving network performance.

[0006] The first network element is a first access network device and the second network element is a core network element or an operation, administration, and maintenance (OAM) network element; the first network element is a first terminal device and the second network element is an access network device; the first network element is a first access network device and the second network element is a second access network device; or the first network element is a first terminal device and the second network element is a second terminal device.

[0007] In one possible implementation, the type of the first policy includes any one of an energy saving policy, a load balancing policy, a mobility optimization policy, a channel state information-reference signal (CSI-RS) feedback improvement policy, a beam management improvement policy, or a positioning accuracy improvement policy.

[0008] In one possible implementation, the indication information indicating that an exception will occur when the first policy is executed includes one or more of: an indication of the execution exception, a time at which the execution exception will occur, a cause of the execution exception, a measurement parameter used to determine that an exception will occur with respect to the policy, a configuration parameter used to determine that an exception will occur with respect to the policy, a validity period of the configuration parameter, a corrective action that the first network element expects to use for the execution exception, information required for the corrective action that the first network element expects to use for the execution exception, an identifier of the first AI model, a parameter of the first AI model, an identifier of the first policy, the first policy, or a validity period of the first policy. The one or more parameter information is used so that the second network element can more clearly know the relevant case of the exception to take more appropriate response measures.

[0009] In one possible implementation, the first network element receives first information from the second network element, the first information indicating a corrective manner that is allowed to be used for the execution exception, and the first network element corrects the generated exception for the first policy based on the corrective manner. The second network element can act as a manager or decider to inform the first network element of the corrective manner that is allowed to be used by the first network element, and the first network element appropriately corrects the exception to further improve network performance.

[0010] In one possible implementation, the corrective manner that can be used includes executing a second policy, where the second policy is a policy obtained based on a non-AI, and the second policy and the first policy are of the same type. This can be understood as the corrective manner that can be used being executing a conventional policy instead of the policy obtained based on the AI ​​(the second policy is the conventional policy). Correcting the exception generated for the first policy by the first network element based on the corrective manner includes the first network element skipping the first policy and executing the second policy.

[0011] In one possible implementation, the corrective manner that can be used includes executing a third policy, where the third policy is determined based on a second AI model, and the third policy and the first policy are of the same type. This can be understood as the corrective manner that can be used being to determine a new AI policy (the third policy is a new AI policy) by using a new AI model (the second model is a new AI model). The third policy may be obtained in the following manner: a second measurement amount to be input is input to the second AI model to obtain a third policy that is output by the second AI model. Correcting the exception generated with respect to the first policy based on the corrective manner by the first network element includes the first network element skipping the first policy and executing the third policy.

[0012] In one possible implementation, the corrective manner that is allowed to be used includes executing a fourth policy, where the fourth policy is a policy that is executed before the first policy, and the fourth policy and the first policy are of the same type. This can be understood as the corrective manner that is allowed to be used being a fallback to a previously executed policy (where the fourth policy is the previously executed policy). Correcting the exception generated with respect to the first policy based on the corrective manner by the first network element includes the first network element skipping the first policy and executing the fourth policy.

[0013] In one possible implementation, the corrective manner that can be used includes skipping execution of a policy of the same type as the first policy. This can be understood as the corrective manner that can be used is terminating the first policy. The first network element correcting the exception generated for the first policy based on the corrective manner includes the first network element exiting the first policy.

[0014] In one possible implementation, when the corrective action that is allowed to be used is to execute a third policy, the first information includes at least one of an identifier of the third policy, the third policy, an identifier of the second AI model, parameters of the second AI model, or performance information of a third network element, and the performance information of the third network element is used to determine the third policy.

[0015] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes, when the first information includes an identifier of a third policy, the first network element executing a third policy indicated by the identifier of the third policy to correct the exception generated regarding the first policy.

[0016] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes, when the first information includes a third policy, the first network element executing the third policy to correct the exception generated regarding the first policy.

[0017] In one possible implementation, the first network element correcting the exception generated with respect to the first policy based on the correction manner includes: when the first information includes an identifier of a second AI model, the first network element obtains the second AI model based on the identifier, inputs the second measurement into the second AI model to obtain a third policy, and executes the third policy to correct the exception generated with respect to the first policy, wherein the second measurement is the same as the first measurement or the measurement time of the second measurement is not earlier than the measurement time of the first measurement.

[0018] In one possible implementation, the first network element correcting the exception generated with respect to the first policy based on the correction manner includes: when the first information includes a parameter of a second AI model, the first network element inputs a second measurement amount to the second AI model corresponding to the parameter to obtain a third policy, and executes the third policy to correct the exception generated with respect to the first policy, wherein the second measurement amount is the same as the first measurement amount or the measurement time of the second measurement amount is not earlier than the measurement time of the first measurement amount.

[0019] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes: when the first information includes performance information of a third network element, the first network element corrects the first AI model based on the performance information to obtain a second AI model, inputs a third measurement amount into the second AI model to obtain a third policy, and executes the third policy to correct the exception generated regarding the first policy, wherein the third measurement amount is the same as the first measurement amount or the measurement time of the third measurement amount is not earlier than the measurement time of the first measurement amount.

[0020] According to a second aspect, a communication method is provided. The method may be executed by a second network element or a component, such as a chip or processor, used in the second network element. The following provides an example in which the method is executed by the second network element. First, the second network element receives instruction information from a first network element, where the instruction information indicates that an exception occurs when the first network element executes a first policy, the first policy being a policy output by the first AI model after the first network element inputs a first measurement quantity to be input to the first AI model; and the second network element sends second information to a third network element, where the second information indicates that an exception occurs when the first policy is executed or that an exception occurs when a first type of policy output by the first AI model is executed, the type of the first policy being a first type.

[0021] A network element may recognize the execution status of another network element executing a policy obtained based on an AI model, thereby improving network performance to a certain extent. Specifically, when determining that an exception occurs when a first policy is executed, the first network element notifies a second network element of this case, and the second network element responds based on the exception. Specifically, the second network element notifies a third network element that an exception occurs when the first policy is executed or that an exception occurs when a first type of policy obtained based on AI is executed, and the third network element responds accordingly. In this way, network performance is improved.

[0022] The first network element is a first access network device, the second network element is a core network element or an Operation, Administration, and Maintenance (OAM) network element, and the third network element is a second access network device; the first network element is a first terminal device, the second network element is an access network device, and the third network element is a second terminal device; or the first network element is a terminal device, the second network element is a first access network device, and the third network element is a second access network device.

[0023] In one possible implementation, the type of the first policy includes any one of an energy saving policy, a load balancing policy, a mobility optimization policy, a channel state information-reference signal (CSI-RS) feedback improvement policy, a beam management improvement policy, or a positioning accuracy improvement policy.

[0024] In one possible implementation, the instruction information indicating that an exception will occur when the first policy is executed includes one or more of an instruction for the execution exception, a time at which the execution exception will occur, a cause of the execution exception, a measurement parameter used to determine that an exception will occur with respect to the policy, a configuration parameter used to determine that an exception will occur with respect to the policy, a validity period of the configuration parameter, a corrective action that the first network element expects to use for the execution exception, information required for the corrective action that the first network element expects to use for the execution exception, an identifier of the first AI model, a parameter of the first AI model, an identifier of the first policy, or a validity period of the first policy.

[0025] In one possible implementation, the third network element satisfies any one of the following conditions: the third network element is executing a policy output by the first AI model; the third network element is executing the first policy; or the third network element is executing a fifth policy, where the fifth policy and the first policy are of the same type.

[0026] In one possible implementation, the second network element determines a corrective action that is allowed to be used for the execution exception and sends first information to the first network element, the first information indicating the corrective action that is allowed to be used for the execution exception.

[0027] In one possible implementation, allowable corrective actions include: executing a second policy, where the second policy is a policy obtained based on a non-AI model and the second policy and the first policy are of the same type; executing a third policy, where the third policy is a policy determined based on a second AI model and the third policy and the first policy are of the same type; or executing a fourth policy, where the fourth policy is a policy executed before the first policy and the fourth policy and the first policy are of the same type.

[0028] In one possible implementation, when the corrective action that is allowed to be used is to execute a third policy, the first information includes at least one of an identifier of the third policy, the third policy, an identifier of the second AI model, parameters of the second AI model, or performance information of a third network element, and the performance information of the third network element is used to determine the third policy.

[0029] In one possible implementation, using the performance information of the third network element to determine the third policy includes: using the performance information of the third network element to correct the first AI model, and using the corrected first AI model to determine the third policy.

[0030] In one possible implementation, the second network element receives performance information from a third network element.

[0031] According to a third aspect, a communication method is provided. The method may be executed by a third network element or a component, such as a chip or a processor, used in the third network element. The following provides an example in which the method is executed by the third network element. First, the third network element receives second information from the second network element, the second information indicating that an exception occurs when a first policy is executed or that an exception occurs when a first type of policy output by a first AI model is executed, the first policy belonging to the first type and obtained based on the first AI model; the third network element transmits performance information of the third network element to the second network element and / or determines whether an exception occurs with respect to the policy executed by the third network element, and the performance information is used by the first network element to correct the first policy or the first AI model.

[0032] A network element may be aware of the execution status of another network element executing a policy obtained based on an AI model, thereby improving network performance to some extent. In particular, when an exception occurs when a first policy is executed, the network element notifies a third network element of this situation, so that the third network element can respond based on the exception, thereby improving network performance.

[0033] The second network element is a core network element or an Operation, Administration, and Maintenance (OAM) network element, and the third network element is a second access network device; the second network element is an access network device, and the third network element is a second terminal device; or the second network element is a first access network device, and the third network element is a second access network device.

[0034] According to a fourth aspect, a communication method is provided. The method may be executed by a first network element or a component used in the first network element, such as a chip or a processor. The following provides an example in which the method is executed by the first network element. First, the first network element acquires a first measurement to be input and acquires a first AI model. Then, the first network element inputs the first measurement to the first AI model to obtain a first policy output by the first AI model. Then, the first network element executes the first policy. Next, when the first network element determines that an exception occurs when the first policy is executed, the first network element sends instruction information to a third network element, where the instruction information indicates that an exception occurs when the first policy is executed.

[0035] When executing the first policy determined based on the AI ​​model, the first network element may detect the execution status and notify the second network element of the execution status. In this way, the network element may be aware of the execution status of the policy obtained based on the AI ​​model by another network element, thereby improving network performance to some extent. In particular, when determining that an exception occurs when the first policy is executed, the network element notifies the other network element of this situation, so that the other network element can respond based on the exception, thereby improving network performance.

[0036] The first network element is a first terminal and the second network element is a second terminal; or the third network element is a first access network device and the third network element is a second access network device.

[0037] In one possible implementation, the first network element receives performance information from the third network element and modifies the first policy or the first AI model based on the performance information of the third network element.

[0038] According to a fifth aspect, there is provided a communication method. The method may be executed by a third network element or a component, such as a chip or a processor, used in the third network element. The following provides an example in which the method is executed by the third network element. First, the third network element receives, from the first network element, instruction information indicating that an exception occurs when the first network element executes a first policy, the first policy being a policy output by the first AI model after the first network element inputs a first measurement quantity to be input to the first AI model. Then, the third network element determines whether an exception occurs with respect to the policy executed by the third network element, and / or the third network element transmits performance information of the third network element to the first network element, where the performance information is used by the first network element to correct the first policy or the first AI model.

[0039] A network element may be aware of the execution status of another network element's execution of a policy obtained based on an AI model, thereby improving network performance to some extent. In particular, when an exception occurs with respect to a policy, the network element notifies a third network element of this situation, so that the third network element can respond based on the exception, thereby improving network performance.

[0040] The first network element is a first terminal and the second network element is a second terminal; or the third network element is a first access network device and the third network element is a second access network device.

[0041] According to a sixth aspect, there is provided a communications device. The device has functionality to implement any one of the aforementioned aspects and possible implementations of the aforementioned aspects. The functionality may be implemented by hardware or by executing corresponding software. The hardware or software includes one or more functional modules corresponding to the aforementioned functionality.

[0042] According to a seventh aspect, there is provided a communications device, the communications device including a processor and optionally further including a memory. The processor is coupled to the memory. The memory is configured to store computer programs or instructions. The processor is configured to execute some or all of the computer programs or instructions in the memory, and when some or all of the computer programs or instructions are executed, to implement functions in a method according to any one of the preceding aspects and possible implementations of the preceding aspects.

[0043] In one possible implementation, the apparatus may further include a transceiver configured to transmit a signal processed by the processor or receive a signal input to the processor, and the transceiver may perform a transmitting or receiving action in any one of the aspects and possible implementations of the aspects.

[0044] According to an eighth aspect, the present application provides a chip system. The chip system includes one or more processors (sometimes referred to as processing circuits). The processors are electrically coupled to a memory (sometimes referred to as a storage medium). The memory may or may not be located within the chip system. The memory is configured to store computer programs or instructions. The processor is configured to execute some or all of the computer programs or instructions in the memory, and when some or all of the computer programs or instructions are executed, to implement functions in a method according to any one of the aforementioned aspects and possible implementations of the aforementioned aspects.

[0045] In one possible implementation, the chip system may further include an input / output interface (sometimes called a communication interface). The input / output interface is configured to output a signal processed by the processor or receive a signal input to the processor. The input / output interface may perform a sending action or a receiving action in any one of the aspects and possible implementations of the aspects. Specifically, the output interface performs a sending action, and the input interface performs a receiving action.

[0046] In one possible implementation, a chip system may include a chip, or may include a chip and another discrete device.

[0047] According to a ninth aspect, a computer-readable storage medium is provided and configured to store a computer program, the computer program including instructions for implementing functionality in any one of the aspects and possible implementations of the aspects.

[0048] Alternatively, a computer-readable storage medium is provided and configured to store a computer program that, when executed by a computer, enables the computer to perform a method according to any one of the aforementioned aspects and possible implementations thereof.

[0049] According to a tenth aspect, there is provided a computer program product, the computer program product comprising computer program code that, when executed on a computer, enables the computer to perform a method according to any one of the preceding aspects and possible implementations of the preceding aspects.

[0050] According to an eleventh aspect, there is provided a communication system including a first network element that performs a method according to any one of the first aspect and possible implementations thereof, and a second network element that performs a method according to any one of the second aspect and possible implementations thereof. Optionally, the communication system further includes a third network element that performs a method according to any one of the third aspect and possible implementations thereof.

[0051] According to a twelfth aspect, there is provided a communication system including a first network element that performs a method according to the fourth aspect and any one of possible implementations thereof, and a third network element that performs a method according to the fifth aspect and any one of possible implementations thereof.

[0052] For the technical effects of the sixth to twelfth aspects, please refer to the descriptions of the first to fifth aspects, and the details will not be repeated. [Brief explanation of the drawings]

[0053] [Figure 1] 1a and 1b are diagrams of the structure of a communication system according to an embodiment of the present application; [Figure 2] FIG. 1 is a diagram of an AI model application architecture according to an embodiment of the present application. [Figure 3] 1 is a diagram of a communication procedure according to an embodiment of the present application; [Figure 4] 1 is a diagram of a communication procedure according to an embodiment of the present application; [Figure 5] 1 is a diagram of the structure of a communication device according to an embodiment of the present application; [Figure 6] 1 is a diagram of the structure of a communication device according to an embodiment of the present application; DETAILED DESCRIPTION OF THE INVENTION

[0054] In order to facilitate understanding of the technical solutions in the embodiments of the present application, the following briefly describes the system architecture of the method provided in the embodiments of the present application. It can be understood that the system architecture described in the embodiments of the present application is intended to more clearly describe the technical solutions in the embodiments of the present application, and does not constitute limitations on the technical solutions provided in the embodiments of the present application.

[0055] The technical solutions in the embodiments of the present application may be applied to various communication systems, such as a satellite communication system and a conventional mobile communication system. The satellite communication system may be integrated with a conventional mobile communication system (i.e., a terrestrial communication system). The communication system may be, for example, a wireless local area network (WLAN) communication system, a wireless fidelity (Wi-Fi) system, a long term evolution (LTE) system, an LTE frequency division duplex (FDD) system, an LTE time division duplex (TDD) system, a fifth generation (5G) system or new radio (NR) system, a sixth generation (6G) system, or another future communication system. The communication system further supports communication systems that integrate multiple wireless technologies, and may further be used, for example, in systems that integrate non-terrestrial networks (NTNs) and terrestrial mobile communication networks, such as unmanned aerial vehicles, satellite communication systems, or high altitude platform station (HAPS) communications.

[0056] For example, Figure 1a is a diagram of the architecture of a 5G communication system to which the present application is applicable. The 5G communication system includes terminal devices, access network devices, and core network devices. The network devices may communicate and interact with the core network devices to provide communication services for the terminal devices. The core network devices are, for example, devices in a 5G core network (CN). As a bearer network, the core network provides an interface to a data network, provides communication connectivity, authentication, management, and policy control for user equipment (UE), and carries data services.

[0057] A terminal device may be a wireless terminal, a wired terminal, or may be referred to as user equipment (UE). A terminal device may be located on land, including an indoor or outdoor device, a handheld device, or a vehicle-mounted device; on water (e.g., on a ship); or in the air (e.g., on an aircraft, balloon, or satellite). A terminal device may refer to a device that provides voice and / or data connectivity to a user, a handheld device with wireless connectivity, or another processing device connected to a wireless modem. A terminal device may communicate with one or more core networks via a radio access network. The terminal device may be a mobile phone, a tablet computer, a computer with wireless transceiver capabilities, a virtual reality (VR) terminal, an augmented reality (AR) terminal, a wireless terminal in industrial control, a wireless terminal in self-driving, a wireless terminal in telemedicine, a wireless terminal in a smart grid, a wireless terminal in transportation safety, a wireless terminal in a smart city, a wireless terminal in a smart home, etc. The wireless terminal may be, for example, a mobile phone (also called a "cellular" phone) or a computer with a mobile terminal, such as a portable, pocket-sized, handheld, computer-integrated, or vehicle-mounted mobile device that exchanges language and / or data with a wireless access network.For example, a wireless terminal may be a device such as a personal communications service (PCS) phone, a cordless phone, a session initiation protocol (SIP) phone, a wireless local loop (WLL) station, or a personal digital assistant (PDA). A terminal may also be called a system, a subscriber unit (SU), a subscriber station (SS), a mobile station (MB), a remote station (RS), an access point (AP), a remote terminal (RT), an access terminal (AT), a user terminal (UT), or a user agent (UA).

[0058] In this application, an (R)AN device is a device that provides wireless communication functions for terminal devices, and is also referred to as an access network device. In this application, RAN devices include, but are not limited to, next-generation base stations (gNodeBs, gNBs) in 5G, evolved NodeBs (eNBs), radio network controllers (RNCs), NodeBs (NBs), base station controllers (BSCs), base transceiver stations (BTSs), home base stations (e.g., home evolved NodeBs or HNBs), baseband units (BBUs), transmission reception points (TRPs), transmission points (TPs), and mobile switching centers. In systems using different radio access technologies, devices with base station functionality may be given different names. For example, the device is called a RAN or gNB (5G NodeB) in a 5th generation (5G) system, an evolved NodeB (eNB, or eNodeB) in an LTE system, and a NodeB in a 3rd generation (3G) system.

[0059] As shown in FIG. 1b, a base station may have an architecture in which a central unit (CU) is separated from distributed units (DUs). The RAN may be connected to a core network (e.g., the core network may be a Long Term Evolution (LTE) core network, a 5G core network, etc.). It may be understood that a base station is divided into a CU and a DU from the perspective of logical functions. The CU and DU may be physically separated or co-located. Multiple DUs may share one CU. Alternatively, one DU may be connected to multiple CUs (not shown in FIG. 1b). The CU and DU may be connected through an interface, for example, an F1 interface. The CU and DU may be divided based on the protocol layer of the wireless network. For example, one possible division scheme is that the CU is configured to perform functions of a radio resource control (RRC) layer, a service data adaptation protocol (SDAP) layer, and a packet data convergence protocol (PDCP) layer, and the DU is configured to perform functions such as a radio link control (RLC) layer, a medium access control (MAC) layer, and a physical layer. It may be understood that the division of the CU and DU into processing functions based on protocol layers is merely an example, and other divisions may exist. For example, the CU or DU may have more protocol layer functions through division. For example, the CU or DU may alternatively have some processing functions of the protocol layers through division. In one design, some functions of the RLC layer and functions of protocol layers above the RLC layer are configured on the CU, and the remaining functions of the RLC layer and functions of protocol layers below the RLC layer are configured on the DU. In another design, the functions of the CU or DU may alternatively be obtained through division based on service type or another system requirement.For example, the division may be performed based on delay, with functions whose processing time must meet the delay requirement being placed on the DU and functions whose processing time does not need to meet the delay requirement being placed on the CU. In another design, the CU may alternatively have one or more functions of a core network. The one or more CUs may be deployed in a centralized or distributed manner. For example, the CU may be deployed on the network side to facilitate centralized management. The DU may have multiple radio frequency functions, or the radio frequency functions may be deployed remotely.

[0060] The functions of the CU may be implemented by one entity or different entities. For example, the functions of the CU may be further divided, for example, into a control plane (CP) and a user plane (UP), i.e., into a CU control plane (CU-CP) and a CU user plane (CU-UP). For example, the CU-CP and CU-UP may be implemented by different functional entities and connected through an E1 interface. The CU-CP and CU-UP are coupled to the DU to jointly complete the functions of the base station. The CU control plane CU-CP further includes a further division architecture. In other words, the existing CU-CP is further divided into CU-CP 1 and CU-CP 2. CU-CP 1 includes various radio resource management functions, and CU-CP 2 includes only RRC functions and PDCP-C functions (i.e., basic functions of control plane signaling in the PDCP layer).

[0061] The core network devices may include one or more of the following network elements:

[0062] An access management network element (sometimes referred to as a mobility management network element or an access and mobility management network element) is a control plane network element provided by an operator network and is responsible for access control and mobility management for terminal devices to access the operator network, including functions such as mobility state management, temporary user identity assignment, and user authentication. In a 5G communication system, the access management network element may be an access and mobility management function (AMF) network element. In future communication systems, the access management network element may still be an AMF network element or may have another name. This is not a limitation in this application.

[0063] A session management network element is primarily responsible for session management in a mobile network, such as session establishment, modification, and release. Specific functions include, for example, assigning an IP address to a user and selecting a user plane network element that provides packet forwarding functionality. In a 5G communication system, the session management network element may be a session management function (SMF) network element. In future communication systems, the session management network element may still be an SMF network element or may have a different name. This is not a limitation in this application.

[0064] A user plane network element is responsible for forwarding and receiving user data in a terminal device. The user plane network element may receive user data from a data network and transmit the user data to the terminal device through an access network device. In addition, the user plane network element may alternatively receive user data from a terminal device through an access network device and forward the user data to the data network. The transmission resources and scheduling functions in the user plane network element that provides services for the terminal device are managed and controlled by the SMF network element. In a 5G communication system, the user plane network element may be a user plane function (UPF) network element. In future communication systems, the user plane network element may still be a UPF network element or may have a different name. This is not a limitation in this application.

[0065] The data management network element is configured to generate authentication credentials, process subscriber identities (e.g., store and manage subscription persistent identifiers), control access, manage subscription data, etc. In a 5G communication system, the data management network element may be a unified data management (UDM) network element. In future communication systems, the unified data management network element may still be a UDM network element or may have another name. This is not limited in this application.

[0066] The policy control network element mainly supports providing a unified policy framework for controlling network behavior and providing policy rules for control layer network functions, and is responsible for obtaining user subscription information relevant to policy decisions. In a 4G communication system, the policy control network element may be a policy and charging rules function (PCRF) network element. In a 5G communication system, the policy control network element may be a policy control function (PCF) network element. In future communication systems, the policy control network element may still be a PCF network element or may have another name. This is not a limitation in this application.

[0067] The network repository network element may provide network element discovery functionality and may be configured to provide network element information corresponding to a network element type based on a request from another network element. The NRF further provides network element management services, such as network element registration, update, or deregistration, and network element status subscription and push. In a 5G communication system, the network repository network element may be a network repository function (NRF) network element. In future communication systems, the network repository network element may still be an NRF network element or may have a different name. This is not a limitation in this application.

[0068] A network-exposed network element is a control plane network element provided by an operator. The network-exposed network element may be configured to securely expose the external interface of the operator network to third parties and securely expose services, capabilities, and the like provided by 3rd Generation Partnership Project (3GPP) network function devices to the outside world. For example, when a session management network element needs to communicate with a third-party network element, the network-exposed network element may act as a relay for communication between the session management network element and the third-party network element. When acting as a relay, the network-exposed network element may translate between the subscriber's identity and the third-party network element's identity. For example, when transmitting a subscriber's SUPI from the operator network to a third party, the network-exposed network element may translate the SUPI into an external identity (ID) corresponding to the SUPI. Conversely, when transmitting an external ID (third-party network element ID) to the operator network, the network-exposed network element may translate the external ID into the SUPI. In a 5G communication system, the network exposure function network element may be a network exposure function (NEF) network element. In future communication systems, the network exposure function network element may still be an NEF network element or may have another name, which is not limited in this application.

[0069] The network slice selection network element may be used to select an appropriate network slice for a terminal service. In a 5G communication system, the network slice selection network element may be a network slice selection function (NSSF) network element. In future communication systems, the network exposure function network element may still be an NSSF network element or may have another name. This is not a limitation in this application.

[0070] The network data analytics network element may collect data (via a network capability exposure function network element) from each network function (NF), such as a policy control network element, a session management network element, a user plane network element, an access management network element, and an application function network element, and perform analysis and prediction. In a 5G communication system, the network data analytics network element may be a network data analytics function (NWDAF). In future communication systems, the network exposure function network element may still be an NWDAF network element or may have a different name. This is not a limitation in this application.

[0071] The unified data repository network element is responsible for storing structured data information, including subscription information, policy information, and network or service data defined in a standard format. In a 5G communication system, the unified data repository network element may be a unified data repository (UDR). In future communication systems, the network exposure function network element may still be a UDR network element or may have a different name. This is not a limitation in this application.

[0072] It may be understood that a network element may also be referred to as a "device," "entity," etc. The aforementioned network elements or functions may be network elements within a hardware device, software functions running on dedicated hardware, or virtualized functions instantiated on a platform (e.g., a cloud platform). One or more services may be obtained through the division into the aforementioned network elements or functions. Furthermore, services may occur that exist independently of a network function. In this application, an instance of a function, an instance of a service included in a function, or an instance of a service that exists independently of a network function may be referred to as a service instance.

[0073] To facilitate understanding of the embodiments of the present application, application scenarios of the present application are described below. The network architectures and service scenarios described in the embodiments of the present application are intended to more clearly explain the technical solutions in the embodiments of the present application, and do not constitute any limitations on the technical solutions provided in the embodiments of the present application. Those skilled in the art may recognize that as new service scenarios emerge, the technical solutions provided in the embodiments of the present application can also be applied to similar technical problems.

[0074] The technical solutions in the embodiments of the present application may be applied to scenarios in which AI policies are executed, and are particularly applicable to scenarios in which performance degradation occurs when AI policies are executed, and are not limited to communication systems.

[0075] Figure 2 illustrates the AI ​​model application architecture. The data source is a terminal, access network device, gNB-CU, gNB-DU, operation, administration, and maintenance (OAM), or another management entity that provides data (e.g., network performance data detected by OAM, e.g., network load and channel quality). The model training unit analyzes the training data provided by the data source (e.g., by performing model establishment, training approximation, and reinforcement learning on it) to obtain a trained AI model. The model inference unit uses the trained AI model to obtain reasonable predictions (e.g., access network performance prediction, load prediction, and UE trajectory prediction) based on the inference data provided by the data source, performs policy adjustment, and guides the network to obtain reasonable and efficient policies (e.g., conservation policies or mobility optimization policies). The policy adjustment is uniformly planned by the actor entity, and the adjusted policies are sent to multiple network entities for execution. In addition, after multiple network entities execute the coordinated policy, the relevant data of the network entities is collected again in a data source.

[0076] In a wireless communication system, the AI ​​model training process may be executed on the OAM, the access network device, the gNB-CU, the terminal, or an independent network element entity (e.g., a RAN intelligent controller (RIC)). The process of performing inference based on the AI ​​model may be executed on the access network device, the gNB-CU, the terminal, or an independent network element entity (AIC). For example, the OAM and the access network device may exchange information, e.g., model information and data used for model training, through a current northbound interface. The access network device and the gNB-CU may exchange information with another network element by reusing a current interface such as F1, Xn, or Uu. The network element entity AIC may establish a communication link, e.g., a wired link or a wireless link, with another network element (e.g., the OAM or the access network device) for communication. When the CP and UP of a CU are separated, the CP is typically responsible for receiving AI models and subsequent AI inference and policy generation. When the CU-CP is further divided into CU-CP 1 and CU-CP 2, CU-CP 1 is typically responsible for receiving models, subsequent AI inference, and generating specific interaction signaling, while CU-CP 2 interacts with other network elements.

[0077] To facilitate understanding of the embodiments of the present application, the following describes some terms in the embodiments of the present application to help those skilled in the art to understand better.

[0078] 1. Energy Saving Policy: Specific parameters of the energy saving policy include, but are not limited to, one or more of the following: which cell(s) to deactivate, which carrier(s) to shut down (or turn on), which channel(s) to shut down (or turn on), which slot(s) to shut down (or turn on), the transmit power adjustment value (the adjustment can be a decrease or an increase), and the adjusted transmit power value. The energy saving policy can be applied to an access network device. For example, the access network device collects load, energy consumption, and energy efficiency information of the access network device and neighboring cells, UE trajectory information and measurement results, etc., predicts the load trend of the access network device, and takes energy saving measures in a timely and appropriate manner without affecting network coverage and user access, based on cell usage, KPI requirements, etc. The simplest energy-saving policy involves directly deactivating a cell. Other energy-saving policies include carrier shutdown, channel shutdown, slot shutdown, and transmit power adjustment. Complex energy-saving policies further include a combination of the aforementioned energy-saving measures. The AI ​​model is used to predict the load of each cell and formulate an appropriate energy-saving policy to reduce base station energy consumption or improve base station energy efficiency without affecting user services and cell coverage. If network coverage is affected or UE access or service requirements cannot be met, the current energy-saving policy can be modified, or the non-energy-saving state can be directly restored to a normal working state. Re-predicting the load or modifying the AI ​​model used and performing re-inference can be considered.

[0079] 2. Load Balancing Policy: Specific parameters of the load balancing policy include, but are not limited to, one or more of the following: which UEs accessing an access network device are selected by the access network device to be handed over to a neighboring access network device or a neighboring cell; the handover threshold or the adjustment value for the handover between the cell of the current access network device (or the current cell) and a neighboring access network device (or a neighboring cell); and the load trend of the access network device. The load balancing policy can be applied to an access network device. For example, the access network device collects load, energy consumption, and energy efficiency information of the access network device and neighboring cells, UE trajectory information and measurement results, etc., predicts the load trend of the access network device, and appropriately selects some UEs to be handed over to neighboring cells or receives UEs from the neighboring cells based on cell usage, KPI requirements, etc. In this way, the load levels of access network devices across the entire network area are close, and cases where heavy loads on some access network devices affect normal services and idle resources on some access network devices are reduced. However, prediction accuracy is not 100%. As a result, UE selection may be inappropriate or the target cell for handover may be inappropriate, resulting in handover failure or UE service impact. The load balancing effect may be poor due to inaccurate load prediction, or the original load balancing policy may no longer be applicable due to temporary exception changes to the load. In this case, the current load balancing policy can be terminated or modified, and consideration may be given to re-predicting the load or changing the AI ​​model used and performing re-inference.

[0080] 3. Mobility Optimization Policy: Specific parameters of the mobility optimization policy include, but are not limited to, one or more of the UE's predicted trajectory, handover time, and target cell for handover. The mobility optimization policy can be applied to an access network device. For example, the access network device collects UE historical trajectory information and predicts the UE's future trajectory by referring to the UE's measurement information. Based on the predicted trajectory, it determines in advance whether the UE will be handed over, delivers a handover configuration in advance, and instructs the target cell to prepare access resources. This reduces delays in the UE handover process and the probability of handover or access failure. However, trajectory prediction accuracy is not 100%. As a result, if the predicted trajectory is incorrect, the UE handover will fail and service will be interrupted. In this case, to prevent a similar exception from occurring again on the UE, it may be considered to perform model retraining and re-inference by referring to the exception, or to modify the model.

[0081] 4. CSI-RS Feedback Enhancement Policy: Specific parameters of the CSI-RS feedback enhancement policy include, but are not limited to, one or more of the following: downlink channel matrix. The main steps are as follows: First, the access network device and the UE exchange dictionaries, where the dictionaries are typically models pre-trained by the access network device based on the UE capabilities and the requirements of the access network device. Then, the access network device sends encoder and quantizer tools to the UE. Then, based on the downlink channel matrix and the existing dictionary, the UE can compress and quantize the matrix to be fed back and transfer the result to the access network device. Next, the access network device obtains the original channel matrix through inverse reconstruction based on the dictionary and data reported by the UE.

[0082] 5. Beam Management Enhancement Policy: Specific parameters of the beam management enhancement policy include, but are not limited to, one or more of the following: sparse matrix for sweeping. The main steps are as follows: The access network device instructs the UE to perform beam sweeping based on a sparse matrix for sweeping, the UE performs beam sweeping based on the matrix, and feeds back the sweeping result; the access network device obtains the optimal CSI-RS beam through inference based on the UE's sparse sweeping result, and the UE starts to perform the next phase of sweeping; and the UE feeds back the ID of the optimal CSI-RS beam.

[0083] 6. Positioning Accuracy Enhancement Policy: Specific parameters of the positioning accuracy enhancement policy include, but are not limited to, one or more of the following: UE location. Compared with the UE location determined in the prior art, this technology predicts a more refined location. For example, the UE is located in an urban area or outdoors.

[0084] The embodiments of the present application provide multiple communication methods. Network elements can exchange information about the execution status of policy execution, and any network element can create a response policy based on the execution status of policy execution by another network element. In this way, multiple network elements can implement better policies, improving network performance.

[0085] The following describes the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application, and cross-references may be made between the embodiments / examples. [Example]

[0086] Embodiment 1 3 is a diagram illustrating a communication procedure, which is applicable to, but not limited to, any one of the following communication scenarios: The first network element is a first access network device, the second network element is a core network device, and the third network element is a second access network device. The first network element is a first access network device, the second network element is an OAM, and the third network element is a second access network device. The first network element is a first terminal, the second network element is a first access network device, and the third network element is a second terminal. The first network element is a first terminal, the second network element is a first access network device, and the third network element is a second access network device.

[0087] Access network devices communicate with core network devices or OAM through the northbound or NG interface. Access network devices may communicate with OAM directly or through a core network device (e.g., AMF). Access network devices communicate with each other through the Xn interface. Terminals communicate with access network devices through the Uu interface.

[0088] The first terminal or the second terminal may be a terminal device described in Figure 1a or 1b. The first access network device or the second access network device may be the access network device described in Figure 1a or a gNB-CU, gNB-CU-CP, gNB-CU-CP 1, gNB-CU-CP 2, or gNB-DU described in Figure 1b. The core network device may be any network element described in Figure 1a or 1b, such as an AMF, UDM, SMF, UPF, PCF, NEF, or UDR.

[0089] The communication procedure includes the following steps:

[0090] Step 301: A first network element obtains a first measurement quantity to be input and obtains a first AI model.

[0091] An AI model is a model obtained based on artificial intelligence (AI).

[0092] The model training subject is not limited in this application. The first network element may train the first AI model based on AI, or a device / network element other than the first network element may train the first AI model based on AI and transmit information about the first AI model (e.g., model parameters; here, the model parameters may include, but are not limited to, the following parameters related to the model: the number of neurons in each layer, such as the input layer, output layer, or hidden layer, the connection relationships, the weight values ​​of each layer, etc.) to the first network element. For example, the second network element may train a model based on AI and transmit information about the trained model to the first network element. The second network element may further transmit information about the trained model to a third network element and / or another network element. For example, the first network element and the second network element train their respective models based on AI. For example, the first network element may train a model based on AI and transmit the trained model to the second network element.

[0093] For the process of training a model based on AI, please refer to the description of Figure 2. The details will not be described again. The trained model may have different applications, such as energy saving, load balancing, mobility optimization, channel state information-reference signal (CSI-RS) feedback enhancement, beam management enhancement, and positioning accuracy enhancement.

[0094] When the first network element is a first access network device, the first measurement may include, but is not limited to, one or more of: information measured by the first access network device, information measured by a neighboring access network device or neighboring cell (in the vicinity of the first access network device), information measured by a UE accessing the first access network device, or information measured by a UE accessing a neighboring access network device or neighboring cell. Note that the UE, the neighboring access network device, and the neighboring cell may transmit the information measured by the UE, the neighboring access network device, and the neighboring cell to the first access network device.

[0095] When the first network element is a first terminal, the first measurement may include, but is not limited to, one or more of the following: information measured by an access network device accessed by the first terminal, information measured by a neighboring access network device or neighboring cell (in the vicinity of the first access network device) of the access network device accessed by the first terminal, information measured by another UE accessing the first access network device, or information measured by a UE accessing a neighboring access network device or neighboring cell. Note that the neighboring access network device, neighboring cell, and the other UE may transmit the information measured by the neighboring access network device, neighboring cell, and the other UE to the first terminal.

[0096] The measured information is, for example, service information or load information.

[0097] Step 302: The first network element inputs a first measurement into a first AI model to obtain a first policy output by the first AI model.

[0098] The first policy type includes any one of the following: an energy saving policy, a load balancing policy, a mobility optimization policy, a channel state information-reference signal (CSI-RS) feedback improvement policy, a beam management improvement policy, or a positioning accuracy improvement policy. For the specific content (or usage or principle) of each type of policy, please refer to the above description. The details will not be described again.

[0099] Step 303: The first network element executes the first policy.

[0100] It can be understood that the first network element can execute one or more policies at a particular time, and the first policy is one of the one or more policies. For any policy, the first network element may send information about the execution status of the policy to the second network element. In this application, the first policy is used as an example for explanation.

[0101] Step 304: The first network element sends information about the execution status of the execution of the first policy to the second network element.

[0102] Correspondingly, the second network element receives, from the first network element, information regarding an execution status of the execution of the first policy by the first network element.

[0103] The information about the execution status indicates whether an exception occurs or does not occur when the first network element executes the first policy.

[0104] The first network element notifies the second network element of the execution status of the first network element's execution of the first policy, so that the second network element is aware of the execution status of the first network element's execution of the policy. Any network element may perform self-checking or adjustment regarding the policy executed by the network element based on the execution status of the policy executed by another network element. In this way, multiple network elements can execute better policies, improving network performance.

[0105] The first network element may periodically determine the execution status of the first policy execution (i.e., whether the execution is normal or not), and the period may be 1 minute, 5 minutes, 10 minutes, etc. Alternatively, the first network element may determine the execution status of the first policy execution (i.e., whether the execution is normal or not) based on an event trigger (e.g., a measurement quantity exceeds a specified threshold).

[0106] After determining the running status, the first network element may send information about the running status to the second network element. Alternatively, if the first network element periodically sends information about the running status, after determining the running status, the first network element may send information about the running status to the second network element when the next latest reporting period instant arrives.

[0107] For example, conditions (or criteria) used to determine whether an execution exception occurs for each type of policy may be pre-configured in the first network element, and the first network element may determine whether an exception occurs when each type of policy is executed based on the conditions (or criteria).

[0108] An execution exception is determined to occur with respect to the energy saving policy when one or more of the following cases occur (which may be understood as one or more of the following conditions being met), and the following cases are typically applicable to a scenario in which the first network element is a first access network device:

[0109] The load of a neighboring cell or neighboring access network device is equal to or greater than a specified load threshold. For example, the specified threshold may be 80% or 85% of maximum load. For example, the neighboring cell or neighboring access network device may transmit information indicating that the load is excessively high to the first access network device. The first access network device receives information indicating that the load is excessively high from the neighboring access network device or neighboring cell, and the first access network device needs to modify its current energy saving policy to shift load from the neighboring access network device or neighboring cell to the first access network device, thereby reducing the load level of the neighboring access network device or neighboring cell.

[0110] The number of accessing UEs and service requirements exceed the tolerable range in the current energy-saving state. As a result, the UEs do not have abundant resources to access the first access network device, or the time to access the first access network device becomes excessively long. For example, the duration value for the terminal device to access the first access network device is equal to or greater than a specified duration threshold. For example, a certain number of terminal devices fail to access the first access network device, and the number is equal to or greater than a specified quantity threshold. The specified quantity threshold may be a specific value, such as 100 or 1000. Alternatively, the specified quantity threshold may be a ratio value, for example, a ratio to the maximum number of UEs that can access the first access network device.

[0111] A service of a UE in a cell of a first access network device cannot be performed normally. For example, the uplink rate of a terminal device accessing the first access network device is equal to or lower than a specified first rate threshold. For example, the downlink rate of a terminal device accessing the first access network device is equal to or lower than a specified second rate threshold. The rate threshold may be set for each service.

[0112] The effectiveness of the AI-based energy saving policy is poor. For example, after the AI-based energy saving policy is implemented, the energy consumption of the first access network device is not reduced to the expected value, where, for example, the expected value is less than 50% of the energy consumption in a non-energy saving state; the energy efficiency (system load / overall energy consumption) does not reach the expected value, where, for example, the expected value is greater than 150% of the energy efficiency (system load / overall energy consumption) in a non-energy saving state; or the effectiveness of the AI-based energy saving policy is not as good as the effectiveness of a conventional energy saving policy. For example, when the first access network device terminates the AI-based energy saving policy after a specific period (e.g., 30 minutes) and then implements the conventional energy saving policy, the first access network device finds that its energy consumption is low or that its energy efficiency is high. For example, the energy efficiency of the first access network device is below a specified efficiency threshold. In another example, the energy efficiency of the execution of a first policy by a first access network device is less than or equal to the energy efficiency of the execution of a second policy by the first access network device or the energy efficiency of not executing an energy saving policy by the first access network device, where the second policy is an energy saving policy obtained based on a non-AI and the second policy and the first policy are of the same type. The second policy may be understood as a conventional energy saving policy rather than an energy saving policy obtained based on an AI.

[0113] An execution exception is determined to occur with respect to the load balancing policy when one or more of the following cases occur (which may be understood as one or more of the following conditions being met), and the following cases are typically applicable to a scenario in which the first network element is a first access network device:

[0114] The difference between the actual load and the predicted load of the first access network device is excessively large, for example, the load difference between the actual load of the first access network device and the load predicted by the first access network device according to the first policy is equal to or greater than a specified first load difference threshold.

[0115] The difference between the actual load and the predicted load of a neighboring access network device (or neighboring cell) is excessively large. For example, the load difference between the actual load of the neighboring cell or neighboring access network device and the load predicted by the neighboring cell or neighboring access network device is equal to or greater than a specified second load difference threshold. It may be understood that the neighboring access network device or neighboring cell may transmit information to the first access network device indicating that the difference between the actual load and the predicted load is excessively large.

[0116] The load difference between the actual load of the first access network device and the actual load of the neighboring access network device is equal to or greater than a specified third load difference threshold. For example, the load ratio between the actual load of the first access network device and the actual load of the neighboring access network device continuously exceeds a specified ratio (e.g., 3:1). It may be understood that the neighboring access network device or neighboring cell may transmit its actual load to the first access network device.

[0117] Currently, the effect of the AI-based load balancing policy is poor or not as good as that of the conventional load balancing policy. For example, in the process of implementing the AI-based load balancing policy, the conventional load balancing policy solution may be periodically fallen back or may be fallen back through an event trigger (e.g., when the load difference between the first access network device and the neighboring access network device (or neighboring cell) is greater than a certain threshold). If the load balancing effect of the conventional load balancing policy solution is better than that of the AI-based load balancing policy solution (e.g., when the load ratio between the first access network device and the neighboring access network device (or neighboring cell) is closer to 1:1), an exception may be considered to have occurred with respect to the AI-based load balancing policy. For example, the first load difference is greater than the second load difference. The first load difference is a load difference between an actual load of the first access network device and an actual load of a neighboring access network device when the first access network device executes a first policy, and the second load difference is a load difference between an actual load of the first access network device and an actual load of a neighboring access network device when the first access network device executes a second policy or does not execute a load balancing policy, and the second policy is a load balancing policy obtained based on a non-AI, and the second policy and the first policy are of the same type. The second policy may be understood as a conventional load balancing policy that is not a load balancing policy obtained based on an AI.

[0118] Problems such as RLF or potential failure (SHR) occur on the UE. In the process of implementing a load balancing policy, one of the main steps is to hand over the UE from a heavily loaded access network device to a lightly loaded access network device. However, it must also be ensured that the UE's service is not affected and that link problems such as RLF do not occur. An exception to the AI-based load balancing policy can be considered to have occurred if, in the process of implementing an AI-based load balancing policy, more UE call drops occur, more potential failures occur, or the UE's service capability (e.g., throughput) decreases compared to a process without the AI-based load balancing policy. For example, the first throughput is smaller than the second throughput. The first throughput is the throughput of a terminal device accessing the first access network device when the first access network device executes the first policy. The second throughput is the throughput of a terminal device accessing the first access network device when the first access network device executes the second policy or does not execute a load balancing policy. The second policy is a load balancing policy obtained based on a non-AI, and the second policy and the first policy are of the same type. The second policy may be understood as a conventional load balancing policy that is not a load balancing policy obtained based on an AI. For example, the first quantity is greater than the second quantity.The first quantity is a first quantity of terminal devices that experience a radio link failure (RLF), a radio link potential failure, or a call disconnection in a terminal device handover process when the first access network device executes a first policy, and the second quantity is a quantity of terminal devices that experience a radio link failure (RLF), a radio link potential failure, or a call disconnection in a terminal device handover process when the first access network device executes a second policy or does not execute a load balancing policy, where the second policy is a load balancing policy obtained based on a non-AI, and the second policy and the first policy are of the same type. The second policy may be understood as a conventional load balancing policy that is not a load balancing policy obtained based on an AI.

[0119] An execution exception is determined to occur with respect to the mobility optimization policy when one or more of the following cases occur (which may be understood as one or more of the following conditions being met), and the following cases are typically applicable to a scenario in which the first network element is a first access network device:

[0120] A mobility report received by the first access network device, e.g., from a SON, indicates that the call drop rate does not decrease, the handover / access delay does not decrease, the access success rate does not improve, etc. If, based on the report reported by the UE, the AI-based mobility optimization policy has been executed for a period of time (e.g., 5 minutes), the report reported by the UE indicates that the call drop rate does not decrease to a preset value (e.g., decreases to 50% of the call drop rate when the mobility optimization policy is not executed), the handover / access delay does not decrease (e.g., decreases to 50% of the handover / access delay when the mobility optimization policy is not executed), or the access success rate does not increase to a predicted value (e.g., 150% of the access success rate when the mobility optimization policy is not executed), an exception may be considered to have occurred with respect to the current AI-based mobility optimization policy.

[0121] For example, the call drop rate of a terminal device accessing the first access network device is equal to or greater than a specified call drop rate threshold. For example, the threshold is 50% of the call drop rate when a mobility optimization policy is not implemented. For example, the delay of handover of a terminal device from another access network device to the first access network device is equal to or greater than a specified delay threshold. For example, the threshold is 50% of the handover delay when a mobility optimization policy is not implemented. For example, the delay of access by a terminal device to the first access network device is equal to or greater than a specified delay threshold. For example, the threshold is 50% of the access delay when a mobility optimization policy is not implemented. For example, the access success rate of access by a terminal device to the first access network device is equal to or less than a specified threshold. For example, the threshold is 150% of the access success rate when a mobility optimization policy is not implemented.

[0122] The efficiency of the mobility optimization policy solution is not as good as that of the traditional mobility optimization solution. For example, in the process of executing an AI-based mobility optimization policy, the traditional mobility optimization policy solution may fall back periodically or through an event trigger (e.g., when the call drop rate is greater than a certain threshold). If the handover effect of the traditional mobility optimization policy solution is better than that of the AI-based mobility optimization policy (e.g., the call drop rate is lower), it may be considered that an exception occurs with respect to the AI-based mobility optimization policy.

[0123] For example, the first call drop rate is greater than or equal to the second call drop rate. The first call drop rate is the call drop rate of terminal devices accessing the first access network device when the first access network device executes the first policy, and the second call drop rate is the call drop rate of terminal devices accessing the first access network device when the first access network device executes the second policy or does not execute a mobility optimization policy. The second policy is a mobility optimization policy obtained based on a non-AI, and the second policy and the first policy are of the same type. The second policy can be understood as a conventional mobility optimization policy that is not a mobility optimization policy obtained based on an AI.

[0124] For example, the first handover delay is greater than or equal to the second handover delay. The first handover delay is the delay of handover of the terminal device from another access network device to the first access network device when the first access network device executes a first policy, and the second handover delay is the delay of handover of the terminal device from the other access network device to the first access network device when the first access network device executes a second policy or does not execute a mobility optimization policy. The second policy is a mobility optimization policy obtained based on a non-AI, and the second policy and the first policy are of the same type. The second policy may be understood as a conventional mobility optimization policy that is not a mobility optimization policy obtained based on an AI.

[0125] For example, the first access latency is greater than or equal to the second access latency. The first access latency is the latency of a terminal device accessing the first access network device when the first access network device executes the first policy, and the second access latency is the latency of a terminal device accessing the first access network device when the first access network device executes the second policy or does not execute a mobility optimization policy. The second policy is a mobility optimization policy obtained based on a non-AI, and the second policy and the first policy are of the same type. The second policy may be understood as a conventional mobility optimization policy that is not a mobility optimization policy obtained based on an AI.

[0126] For example, the first access success rate is less than or equal to the second access success rate. The first access success rate is the access success rate of the terminal device to the first access network device when the first access network device executes the first policy, and the second access success rate is the access success rate of the terminal device to the first access network device when the first access network device executes the second policy or does not execute a mobility optimization policy. The second policy is a mobility optimization policy not obtained based on AI, and the second policy and the first policy are of the same type. The second policy may be understood as a conventional mobility optimization policy that is not obtained based on AI.

[0127] Based on the measurement information of the UE, it is determined that the actual trajectory of the UE does not match the predicted trajectory. For example, the movement trajectory of the terminal device predicted by the first access network device when the first access network device executes the first policy differs from the actual movement trajectory of the terminal device. For example, the trajectory of the UE is predicted to be from cell 1 to cell 2, but based on the measurement results reported by the UE, the UE actually moves from cell 1 to cell 3. In this case, the prediction of the trajectory of the UE is inaccurate.

[0128] An execution exception is determined to have occurred with respect to the CSI-RS feedback improvement policy when one or more of the following cases occur (which may be understood as one or more of the following conditions being met), and the following cases are typically applicable to scenarios where the first network element is a first terminal or a first access network device:

[0129] The downlink throughput of the UE does not reach an expected value, where, for example, the expected value is 120% of the downlink throughput of the UE in a conventional CSI-RS feedback improvement policy solution. Alternatively, the downlink throughput of the UE is reduced, for example, to 80% of the downlink throughput of the UE in a conventional CSI-RS feedback improvement policy solution. For example, the throughput of a terminal device accessing a first access network device is below a specified throughput threshold.

[0130] If the first network element is a first access network device, the UE may transmit the throughput to the first access network device, so that the access network device performs the determination.

[0131] An execution exception is determined to have occurred with respect to the CSI-RS feedback improvement policy when one or more of the following cases occur (which may be understood as one or more of the following conditions being met), and the following cases are typically applicable to scenarios in which the first network element is a first terminal or a first access network device:

[0132] The beam sweep duration of the UE does not decrease to an expected value, where, for example, the expected value is 80% of the beam sweep duration of the UE in a conventional beam sweep solution. For example, the beam sweep duration value of a terminal device accessing a first access network device is equal to or greater than a specified sweep duration threshold.

[0133] The beam sweep duration of the UE does not exceed that in a conventional beam sweep solution. For example, the first duration is equal to or greater than the second duration. The first duration is the beam sweep duration of the terminal device accessing the first access network device when the first access network device executes the first policy, the second duration is the beam sweep duration of the terminal device accessing the first access network device when the first access network device executes the second policy or does not execute the beam management improvement policy, and the second policy is a beam management improvement policy obtained based on a non-AI.

[0134] The access success rate of the UE decreases. For example, the access success rate of the terminal device's access to the first access network device is equal to or lower than a specified success rate threshold. For example, the threshold is 80%, which is the access success rate in a conventional beam sweeping solution. For example, the first access success rate is equal to or lower than a second access success rate. The first access success rate is the access success rate of the terminal device's access to the first access network device when the first access network device executes a first policy, and the second access success rate is the access success rate of the terminal device's access to the first access network device when the first access network device executes a second policy or does not execute a beam management improvement policy, and the second policy is a beam management improvement policy obtained based on a non-AI.

[0135] The UE throughput is reduced. For example, the UE does not select an optimal beam due to an inappropriate target beam for sparse sweeping. As a result, the throughput is reduced compared to conventional beam sweeping solutions. For example, the throughput of a terminal device accessing a first access network device is equal to or less than a specified throughput threshold. For example, the first throughput is smaller than the second throughput. The first throughput is the throughput of a terminal device accessing the first access network device when the first access network device executes a first policy, and the second throughput is the throughput of a terminal device accessing the first access network device when the first access network device executes a second policy or does not execute a beam management improvement policy, and the second policy is a beam management improvement policy obtained based on a non-AI.

[0136] If the first network element is a first access network device, the UE may send one or more of the beam sweep duration value, the access success rate, and the throughput to the first access network device, thereby causing the access network device to perform the determination.

[0137] When one or more of the following cases occur (which may be understood as one or more of the following conditions being met), it is determined that an execution exception has occurred with respect to the positioning accuracy improvement policy, and the following cases are typically applicable to scenarios in which the first network element is a first terminal or a first access network device:

[0138] A certain percentage (e.g., more than 80%) of the UE's LOS / NLOS determination results are inaccurate, for example, the quantity of inaccurate line-of-sight transmission LOS or non-line-of-sight transmission NLOS results performed by the first access network device to the terminal device is greater than a specified quantity threshold.

[0139] Note that, for each type of policy, the conditions used to determine whether an exception occurs for that policy have been described above. These conditions are merely illustrative examples and should not constitute a limitation on determining whether an exception occurs for that policy. In this application, the conditions used to determine whether an exception occurs may be used interchangeably among multiple types of policies. In addition, it may be understood that when a first network element cannot directly measure a parameter in the above conditions used to determine whether an exception occurs for a policy, the parameter may be measured by a network element that can directly measure the parameter and then transmitted to the first network element.

[0140] In a specific example, when determining that the first policy has been successfully executed, the first network element transmits, to the second network element, indication information indicating that the first policy has been successfully executed (i.e., the information regarding the execution status is indication information indicating that the first policy has been successfully executed). The indication information indicating that the first policy has been successfully executed may include, but is not limited to, one or more of an indication of successful execution, a measurement parameter used to determine that the policy is successful, a configuration parameter used to determine that the policy is successful, an effective time of the configuration parameter, an identifier of the first AI model (e.g., a number or index used to identify the first AI model), a parameter of the first AI model, an identifier of the first policy (e.g., a number or index used to identify the first policy), the first policy, or an effective time of the first policy.

[0141] For example, the value of one or more bits indicates normal execution or an execution exception. For example, when the value of a bit is 0, it indicates normal execution. Or, when the value of a bit is 1, it indicates an execution exception.

[0142] The measurement parameters used to determine that the policy is normal may be understood as performance parameters of each network element (the first network element and / or another network element communicating with the first network element) when the first network element executes the first policy, and include one or more pieces of information in the first measurement quantity acquired in step 301 and / or one or more parameters (excluding the specified threshold) in the above case in which it is determined that an execution exception occurs (or the condition that is met when an execution exception occurs). For example, performance parameters related to normal execution include, but are not limited to, one or more of the following: current load of each cell, predicted load of each cell, handover success rate, access success rate, etc.

[0143] The configuration parameters used to determine whether a policy is healthy may be understood as various thresholds / threshold values ​​that are set to determine whether an exception occurs, and include, but are not limited to, one or more specified thresholds in the aforementioned conditions used to determine whether an execution exception occurs for a policy. For example, the configuration parameters used to determine whether a policy is healthy include, but are not limited to, one or more of the following: a specified load threshold, a specified duration threshold, a specified value threshold, a specified first rate threshold, a specified second rate threshold, a specified efficiency threshold, a specified first load difference threshold, a specified second load difference threshold, a specified third load difference threshold, a specified throughput threshold, a specified sweep duration threshold, a specified success rate threshold, or the effective time of various thresholds / threshold values.

[0144] In a specific example, when the first network element determines that an exception occurred when the first policy was executed, the first network element transmits instruction information to the second network element indicating that an exception occurred when the first policy was executed (i.e., the information related to the execution status is instruction information indicating that an exception occurred when the first policy was executed). The instruction information indicating that an exception occurred when the first policy was executed may include, but is not limited to, one or more of the following: an instruction for the execution exception, a time when the execution exception occurred, a cause of the execution exception, a measurement parameter used to determine that an exception occurred for the policy, a configuration parameter used to determine that an exception occurred for the policy, a validity period of the configuration parameter, a corrective action that the first network element expects to use for the execution exception, information required for the corrective action that the first network element expects to use for the execution exception, an identifier of the first AI model (e.g., a number or index used to identify the first AI model), a parameter of the first AI model, an identifier of the first policy (e.g., a number or index used to identify the first policy), the first policy, or a validity period of the first policy.

[0145] The time at which the execution exception occurs is, for example, 19:00 or 7:00 PM.

[0146] The cause of the execution exception may be a specific type of policy exception, such as an energy saving policy exception, a load balancing policy exception, a mobility optimization policy exception, a channel state information-reference signal (CSI-RS) feedback improvement policy exception, a beam management improvement policy exception, or a positioning accuracy improvement policy exception. The cause of the execution exception may be narrowed down to a specific cause. A specific cause may be understood as the aforementioned case in which it is determined that an execution exception has occurred (or the condition for the execution exception to occur is met).

[0147] The measurement parameters used to determine that an exception has occurred regarding the policy may be understood as performance parameters of each network element (the first network element and / or another network element communicating with the first network element) when the first network element executes the first policy, and include one or more pieces of information of the first measurement amount acquired in step 301 and / or one or more parameters (excluding the specified threshold) in the above case when it is determined that an execution exception has occurred (or the condition when an execution exception occurs is met). For details, please refer to the above-mentioned configuration parameters used to determine that the policy is normal. The details will not be described again.

[0148] The configuration parameters used to determine whether an exception has occurred for a policy may be understood as various thresholds / threshold values ​​that are set to determine whether an exception has occurred, including, but not limited to, one or more specified thresholds in the aforementioned conditions that are used to determine whether an execution exception has occurred for a policy. For details, please refer to the aforementioned configuration parameters used to determine whether a policy is normal. The details will not be described again.

[0149] The remedial measures that the first network element expects to use for the execution exception include, but are not limited to, any one of the following:

[0150] Aspect 1: Execute a second policy. Here, the second policy is a policy obtained based on a non-AI, and the second policy and the first policy are of the same type. This can be understood as a corrective aspect that is expected to be used being to execute a conventional policy that is not a policy obtained based on AI (where the second policy is a conventional policy).

[0151] Aspect 2: Execute a third policy. Here, the third policy is a policy determined based on a second AI model, and the third policy and the first policy are of the same type. This can be understood as a corrective aspect expected to be used being to determine a new AI policy by using a new AI model (where the third policy is a new AI policy). The third policy may be obtained in the following manner: a second measurement value to be input is input to a new second AI model based on the new second AI model, and a third policy output by the second AI model is obtained. In this way, the first network element can further report information about the second AI model, such as the number or index of the second AI model, or parameters of the second AI model, to the second network element. The second AI model may be obtained by correcting the first AI model or may be a newly trained AI model.

[0152] Aspect 3: Execute a fourth policy, where the fourth policy is a policy that was executed before the first policy, and the fourth policy and the first policy are of the same type. This can be understood as the expected corrective aspect being to fall back to the previously executed policy (where the fourth policy is the previously executed policy).

[0153] Aspect 4: Skip execution of any policy of the same type as the first policy This can be understood as the corrective aspect expected to be used is to terminate the first policy.

[0154] The information required for the correction manner that the first network element expects to use for the execution exception may be information used to correct the first AI model or information used to correct the first policy. The corrected first AI model may be the second AI model and may be used to determine a third policy, and the corrected first policy may be understood to be the third policy (new AI policy). The information required for the correction manner that the first network element expects to use for the execution exception may include, but is not limited to, one or more of the following: an identifier (e.g., an index or number) of the new AI model, parameters of the new AI model, an identifier (e.g., a policy number or index) of the new AI policy (i.e., the third policy), the new AI policy, performance information of another network element, historical trajectory of the UE, load information of neighboring cells, network configuration (e.g., network energy saving configuration), etc.

[0155] Optionally, step 305: The second network element sends the first information to the first network element.

[0156] In response, the first network element receives the first information from the second network element.

[0157] Step 305 is an optional step and may not be performed or may be performed together with the subsequent step 309 .

[0158] The first information may indicate one or more of the following: that the first network element is allowed to correct the execution exception, that the first network element is not allowed to correct the execution exception, the correction manner allowed to be used for the execution exception, or information that needs to be retransmitted by the first network element (e.g., information about the first measurement quantity or model-related information, where in this case the first network element may retransmit corresponding information to the second network element).

[0159] The value of one or more bits can indicate whether a correction is allowed or not allowed to be performed. For example, when the value of a bit is 0, it indicates that a correction is allowed to be performed, or when the value of a bit is 1, it indicates that a correction is not allowed to be performed.

[0160] If the first network element reports a corrective manner that it expects the first network element to use, the corrective manner that it allows the second network element to use for the execution exception may be the corrective manner that the first network element expects to use, or of course may not be the corrective manner that the first network element expects to use. The second network element may determine the corrective manner that it is allowed to use based on the corrective manner that the first network element expects to use, performance parameters of another network element, and / or the like.

[0161] Corrective measures that may be used include, but are not limited to, any of the following:

[0162] Aspect 1: Execute a second policy. Here, the second policy is a policy obtained based on a non-AI, and the second policy and the first policy are of the same type. This can be understood as an acceptable corrective aspect being to execute a conventional policy (the second policy is a conventional policy) rather than a policy obtained based on AI.

[0163] The first network element correcting the exception generated regarding the first policy based on the corrective manner includes the first network element skipping the first policy and executing the second policy.

[0164] Mode 2: Execute a third policy. Here, the third policy is a policy determined based on a second AI model, and the third policy and the first policy are of the same type. This can be understood as an allowable correction mode being used to determine a new AI policy (the third policy is the new AI policy) by using a new AI model (the second model is the new AI model). The third policy may be obtained in the following manner: a second measurement amount to be input is input to the second AI model to obtain a third policy that is output by the second AI model. The second AI model may be an AI model obtained by correcting (training) the first AI model, or may be a retrained AI model.

[0165] The first network element correcting the exception generated regarding the first policy based on the corrective manner includes the first network element skipping the first policy and executing a third policy.

[0166] Aspect 3: Execute a fourth policy, where the fourth policy is a policy that is executed before the first policy, and the fourth policy and the first policy are of the same type. This can be understood as an acceptable corrective action being to fall back to a previously executed policy (where the fourth policy is the previously executed policy).

[0167] The first network element correcting the exception generated regarding the first policy based on the correction manner includes the first network element skipping the first policy and executing the fourth policy.

[0168] Aspect 4: Skip the execution of any policy of the same type as the first policy. This can be understood as an acceptable corrective aspect to be used is to terminate the first policy.

[0169] The first network element remediating the exception generated for the first policy based on the remediation manner includes the first network element skipping the first policy.

[0170] A value of two or more bits can indicate the corrective action that is allowed to be used.

[0171] When the corrective action allowed to be taken is to execute a third policy (a new AI policy), the first information includes, but is not limited to, at least one of an index of the third policy, the third policy, an index of the second AI model, parameters of the second AI model, or performance information of the third network element, where the performance information of the third network element is used to determine the third policy.

[0172] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes, when the first information includes an identifier of a third policy, the first network element executing a third policy indicated by the identifier of the third policy to correct the exception generated regarding the first policy.

[0173] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes, when the first information includes a third policy, the first network element executing the third policy to correct the exception generated regarding the first policy.

[0174] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes, when the first information includes an identifier of a second AI model, inputting the second measurement into the second AI model to obtain a third policy and executing the third policy to correct the exception generated regarding the first policy. The second measurement may be the same as or different from the first measurement. The difference between the second measurement and the first measurement may mean that the measurement information included in the measurement is different. For example, the first measurement may include measurement information of a first access network device and measurement information of a UE, and the second measurement may include measurement information of a UE but not measurement information of the first access network device. The difference between the second measurement and the first measurement may mean that the values ​​of the measurement are different. For example, with respect to load information, the load information corresponding to the first measurement is 80%, and the load information corresponding to the second measurement is 85%. In general, the measurement time of the second measurement is not earlier than the measurement time of the first measurement.

[0175] In one possible implementation, the first network element correcting the exception generated regarding the first policy based on the correction manner includes, when the first information includes a parameter of a second AI model, the first network element inputting a second measurement quantity into the second AI model corresponding to the parameter to obtain a third policy, and executing the third policy to correct the exception generated regarding the first policy.

[0176] In one possible implementation, the first network element correcting the exception generated with respect to the first policy based on the correction manner includes, when the first information includes performance information of a third network element, correcting the first AI model based on the performance information to obtain a second AI model, inputting a third measurement into the second AI model to obtain a third policy, and executing the third policy to correct the exception generated with respect to the first policy, where the third measurement may be the same as or different from the first measurement. The third measurement being different from the first measurement may mean that the measurement information included in the measurement is different and / or the value of the measurement is different. For details, please refer to the description of the first measurement being different from the second measurement. The principle is similar, and the details will not be described again. In general, the measurement time of the third measurement is not earlier than the measurement time of the first measurement.

[0177] In one possible implementation, correcting the first AI model based on the performance information includes retraining the first AI model using the performance information as training data to obtain a second AI model.

[0178] When the first network element determines that an exception will occur when the first policy is executed, the first network element may wait for the second network element to notify it of a corrective action that is allowed to be used. After receiving the corrective action that is allowed to be used from the second network element, the first network element performs the corrective action based on the corrective action that is allowed to be used. Alternatively, when the first network element determines that an exception will occur when the first policy is executed, the first network element does not need to wait for the second network element to notify it of a corrective action that is allowed to be used, but instead determines a corrective action to be used and performs the corrective action based on the corrective action that is allowed to be used determined by the first network element. Alternatively, when the first network element determines that an exception will occur when the first policy is executed, the first network element may first determine a corrective action to be used and perform the corrective action based on the corrective action that is determined by the first network element. After receiving the corrective action that is allowed to be used from the second network element, the first network element performs the corrective action based on the corrective action that is allowed to be used.

[0179] The order of steps 305 and 306 is not limited.

[0180] Step 306: When the second network element determines, based on the information about the execution status, that an exception occurs when the first network element executes the first policy, the second network element sends second information to the third network element.

[0181] In response, the third network element receives the second information from the second network element.

[0182] When it is specified that the information on the execution status is to be transmitted only when an exception occurs when the policy is executed, the information on the execution status is indication information indicating that an exception occurs when the first policy is executed. In this case, after receiving the indication information indicating that an exception occurs when the first policy is executed, the second network element can transmit the second information to the third network element, and the process of determining whether an exception occurs when the first network element executes the first policy based on the information on the execution status may be omitted.

[0183] The second information indicates one or more of the following: the occurrence of the exception when the first policy is executed, the occurrence of an exception when a first type policy output by the first AI model is executed (the type of the first policy is a first type), identification information of the first network element (e.g., a cell ID), one or more pieces of information sent by the first network element to the second network element, one or more pieces of information sent by the second network element to the first network element, a self-check performed by the third network element regarding the currently executed policy, a policy guidance solution for the third network element, or information that needs to be sent by the third network element.

[0184] The policy guidance solutions for the third network element are similar to the corrective measures that may be used and sent by the second network element to the first network element, and may include, for example, any one of the following: executing the policy obtained based on the non-AI (which may be understood as a conventional policy), executing the new AI policy, executing the previous policy, or terminating the current policy.

[0185] The information that needs to be transmitted by the third network element may be understood as information requested by the second network element, and includes, but is not limited to, performance information of the third network element and information requested by the first network element from the second network element. The information requested by the first network element from the second network element includes, but is not limited to, information required for corrective measures that the first network element expects to use for the execution exception.

[0186] In one example, the third network element is any network element managed or serviced by the second network element. In one example, the third network element is executing a policy output by the first AI model (in other words, the first network element and the second network element are network elements using the same model). In one example, the third network element is executing the first policy. In one example, the third network element is executing a fifth policy. The fifth policy and the first policy are of the same type. The fifth policy may be a policy obtained based on AI or may be a policy obtained based on non-AI (i.e., a traditional policy). In one example, the third network element is a network element that has a potential AI policy execution exception when the second network element is associated with the first network element.

[0187] The sequence of steps 307 and 308 is not limited.

[0188] Optionally, step 307: After receiving the second information from the second network element, the third network element determines whether an exception occurs with respect to the policy executed by the third network element.

[0189] The second information may be understood as warning information. The third network element performs a self-check based on the warning information of the second network element to eliminate the risk.

[0190] Optionally, step 308: after the third network element receives the second information from the second network element, the third network element sends the information requested by the second network element (i.e., the information that needs to be sent by the third network element) to the second network element.

[0191] Correspondingly, the second network element receives from the third network element the information requested by the second network element (ie, the information that needs to be sent by the third network element).

[0192] Optionally, step 309: The second network element sends information required by the first network element to the first network element.

[0193] In response, the first network element receives from the second network element information required by the first network element.

[0194] The second network element can send information required by the first network element to the first network element based on the indication information indicating that an exception occurs when the first policy is executed in step 304 and / or the information requested by the second network element and sent by the third network element to the second network element (i.e., the information that needs to be sent by the third network element) in step 308. For the information required by the first network element, please refer to the content of the first information in step 305.

[0195] In a specific example, the second network element can determine (or re-determine) whether the first network element is allowed to correct the execution exception, whether the first network element is not allowed to correct the execution exception, or the correction manner that is allowed to be used for the execution exception, based on the information requested by the second network element and transmitted by the third network element (i.e., the information that needs to be transmitted by the third network element). In addition, the second network element indicates to the first network element that the first network element is allowed to correct the execution exception, whether the first network element is not allowed to correct the execution exception, or the correction manner that is allowed to be used for the execution exception. For details, please refer to the above description of the first information in step 305. The details will not be described again. [Example]

[0196] Embodiment 2: The differences from embodiment 1 include: the first network element communicates directly with the third network element without the participation of the second network element.

[0197] 4 is a diagram of a communication procedure that is applicable to, but not limited to, any one of the following communication scenarios:

[0198] The first network element is a first terminal, and the third network element is a second terminal.

[0199] The first network element is a first access network device, and the third network element is a second access network device.

[0200] Access network devices communicate with each other through an Xn interface, and terminals communicate with each other through a sidelink interface, such as a PC5 interface.

[0201] The first terminal or the second terminal may be the terminal device described in Figure 1a or Figure 1b. The first access network device or the second access network device may be the access network device described in Figure 1a or the gNB-CU, gNB-CU-CP, gNB-CU-CP 1, gNB-CU-CP 2, or gNB-DU described in Figure 1b.

[0202] The communication procedure includes the following steps:

[0203] Step 401: Obtain a first measurement to be input, and obtain a first AI model. An AI model is a model obtained based on artificial intelligence (AI).

[0204] Step 402: Input a first measurement amount to be input into a first AI model, and obtain a first policy output by the first AI model.

[0205] Step 403: Execute the first policy.

[0206] Step 404: The first network element sends information about the execution status of the execution of the first policy to the third network element.

[0207] Correspondingly, the third network element receives, from the first network element, information regarding the execution status of the execution of the first policy by the first network element.

[0208] The information about the execution status indicates that an exception occurs or does not occur when the first network element executes the first policy.

[0209] For the process from step 401 to step 404, please refer to the process from step 301 to step 304. The details will not be described again.

[0210] It should be noted that differences between step 404 and step 304 include the following: the information about the running status sent by the first network element to the third network element may include one or more pieces of information in the information about the running status sent by the first network element to the second network element in step 304 and / or one or more pieces of information in the second information sent by the second network element to the third network element in step 306.

[0211] In a specific example, the information about the execution status sent by the first network element to the third network element includes, but is not limited to, one or more of the following: an indication of the execution exception, a time at which the execution exception occurs, a cause of the execution exception, a measurement parameter used to determine that the exception occurs with respect to the policy, a configuration parameter used to determine that the exception occurs with respect to the policy, a validity period of the configuration parameter, a corrective action that the first network element expects to use for the execution exception, information required for the corrective action that the first network element expects to use for the execution exception, an identifier of the first AI model, a parameter of the first AI model, an identifier of the first policy, the first policy, a validity period of the first policy, identification information of the first network element, a self-check performed by the third network element with respect to a currently executed policy, or a policy guidance solution for the third network element.

[0212] Optionally, step 405: After receiving the information about the execution status from the first network element, the third network element determines whether an exception occurs for the policy executed by the third network element.

[0213] Further, optionally, when the information regarding the execution status is determined to be indicative information indicating that an exception occurs when the first policy is executed, the third network element determines whether an exception occurs with respect to the policy executed by the third network element.

[0214] Optionally, step 406: after the third network element receives the information regarding the execution status from the first network element, the third network element sends to the first network element the information requested by the first network element in the information regarding the execution status (i.e., the information that needs to be sent by the third network element).

[0215] Further, optionally, when the third network element determines that the information regarding the execution status is indicative information indicating that an exception occurs when the first policy is executed, the third network element transmits to the first network element information requested by the first network element in the information regarding the execution status (i.e., information that needs to be transmitted by the third network element).

[0216] For example, the information that needs to be transmitted by the third network element may include, but is not limited to, one or more of the following: The information required for the correction mode that the first network element expects to use for the execution exception may include, but is not limited to, one or more of the following: an identifier (e.g., an index or number) of a new AI model, parameters of a new AI model, an identifier (e.g., a policy number or index) of a new AI policy (i.e., a third policy), a new AI policy, performance information of another network element, a historical trajectory of a UE, load information of neighboring cells, a network configuration (e.g., a network energy saving configuration), etc.

[0217] Optionally, the first network element may further modify the first policy or the first AI model based on information requested by the first network element and transmitted by the third network element.

[0218] After the first network element modifies the first policy, the second policy is obtained, and the first network element may skip the first policy and execute the second policy.

[0219] The first network element may correct the first AI model to obtain a second AI model and input the measurement into the second AI model to obtain a third policy. The first network element may skip the first policy and execute the third policy. An aspect of correcting the first AI model based on performance information is to use the performance information as training data and retrain the first AI to obtain the second AI model.

[0220] Optionally, the third network element or the first network element may further transmit the information exchanged between the first network element and the third network element to another network element (e.g., a core network device or an OAM). [Example]

[0221] Embodiment 3 When communicating with the second network element, the first network element can also communicate with a third network element. In other words, the first network element can simultaneously perform the procedures of FIG. 3 and FIG. 4. In one possible implementation, when the first network element simultaneously performs step 304 and step 404, step 304 is used only for notification, and the second network element does not need to provide feedback to the first network element. For example, step 305 or step 309 do not need to be performed. In another possible implementation, when the first network element simultaneously performs step 304 and step 404, step 404 is used only for notification, and the third network element does not need to provide feedback to the first network element. For example, step 406 does not need to be performed. In another possible implementation, when the first network element simultaneously performs step 304 and step 404, the first network element can make a decision based on feedback from the second network element and feedback from the third network element.

[0222] The above describes a method in an embodiment of the present application, and the following describes an apparatus in an embodiment of the present application. The method and the apparatus are based on the same technical idea. The method and the apparatus have similar principles for solving problems. Therefore, the implementation of the apparatus and the method can be referred to each other. Details will not be repeated here.

[0223] In the embodiment of the present application, the device may be divided into functional modules based on the above-mentioned method example. For example, the device may be divided into functional modules corresponding to functions, or two or more functions may be integrated into one module. These modules may be implemented in the form of hardware or software functional modules. It should be noted that in this embodiment of the present application, the module division is an example and is merely a logical functional division. In a specific implementation, other division modes may be used.

[0224] Based on the same technical concept as the above methods, Figure 5 is a diagram of the structure of a communication device 500. The communication device 500 may include a processing module 510, and may optionally further include a receiving module 520a, a transmitting module 520b, and a storage module 530. The processing module 510 may be separately connected to the storage module 530, the receiving module 520a, and the transmitting module 520b, and the storage module 530 may be connected to the receiving module 520a and the transmitting module 520b.

[0225] In one example, the receiving module 520a and the transmitting module 520b may alternatively be integrated together and defined as a transceiver module.

[0226] In one example, the communication device 500 may be a first network element, or may be a chip or functional unit used in the first network element. The communication device 500 has any function of the first network element in the above-mentioned method. For example, the communication device 500 may perform the steps performed by the first network element in the method of FIG. 3 and FIG. 4.

[0227] The receiving module 520a may perform the receiving actions performed by the first network element in the above-described method embodiments.

[0228] The sending module 520b may perform the sending action performed by the first network element in the above method embodiments.

[0229] The processing module 510 may perform actions other than the sending and receiving actions performed by the first network element in the above method embodiments.

[0230] In one example, the processing module 510 is configured to: obtain a first measurement to be input, obtain a first artificial intelligence (AI) model; input the first measurement to the first AI model, obtain a first policy output by the first AI model; and execute the first policy. When determining that an exception occurs when the first policy is executed, the sending module 520b is configured to send instruction information to the second network element indicating that the exception occurs when the first policy is executed.

[0231] In one example, the type of the first policy includes any one of the following: an energy saving policy, a load balancing policy, a mobility optimization policy, a channel state information-reference signal (CSI-RS) feedback improvement policy, a beam management improvement policy, or a positioning accuracy improvement policy.

[0232] In one example, the instruction information indicating that an exception will occur when the first policy is executed includes one or more of the following: an instruction for the execution exception, a time at which the execution exception will occur, a cause of the execution exception, a measurement parameter used to determine that an exception will occur for the policy, a configuration parameter used to determine that an exception will occur for the policy, a validity period of the configuration parameter, a corrective action that the first network element expects to use for the execution exception, information required for the corrective action that the first network element expects to use for the execution exception, an identifier of the first AI model, a parameter of the first AI model, an identifier of the first policy, the first policy, or a validity period of the first policy.

[0233] In one example, the receiving module 520a is configured to receive first information from the second network element, the first information indicating a remediation manner that is allowed to be used for the execution exception, and the processing module 510 is configured to remediate the generated exception for the first policy based on the remediation manner.

[0234] In one example, corrective measures that may be used include any one of the following: executing a second policy, the second policy being a policy obtained based on non-AI, and the second policy and the first policy being of the same type; executing a third policy, the third policy being a policy determined based on a second AI model, and the third policy and the first policy being of the same type; or executing a fourth policy, the fourth policy being a policy executed before the first policy, and the fourth policy and the first policy being of the same type.

[0235] In one example, when the corrective manner allowed to be used is to execute a third policy, the first information includes at least one of the following: an identifier of the third policy, the third policy, an identifier of the second AI model, parameters of the second AI model, or performance information of a third network element, where the performance information of the third network element is used to determine the third policy.

[0236] In one example, the processing module 510 specifically: if the first information includes an identifier of a third policy, execute the third policy indicated by the identifier of the third policy to correct an exception generated with respect to the first policy; if the first information includes the third policy, execute the third policy to correct an exception generated with respect to the first policy; if the first information includes an identifier of a second AI model, input the second measurement amount into the second AI model to obtain the third policy, and execute the third policy to correct an exception generated with respect to the first policy. configured, wherein the second measurement quantity is the same as the first measurement quantity or the measurement time of the second measurement quantity is not earlier than the measurement time of the first measurement quantity; or, if the first information includes performance information of a third network element, correcting the first AI model based on the performance information to obtain a second AI model, inputting the third measurement quantity into the second AI model to obtain a third policy, and executing the third policy to correct the exception generated with respect to the first policy, wherein the third measurement quantity is the same as the first measurement quantity or the measurement time of the third measurement quantity is not earlier than the measurement time of the first measurement quantity.

[0237] In one example, the apparatus is a first access network device and the second network element is a core network element or an Operation, Administration, and Maintenance (OAM) network element; the apparatus is a first terminal device and the second network element is an access network device; the apparatus is a first access network device and the second network element is a second access network device; or the apparatus is a first terminal device and the second network element is a second terminal device.

[0238] In one example, when it is determined that an exception occurred when the first policy was executed, the sending module 520b is configured to send instruction information to a third network element indicating that an exception occurred when the first policy was executed. The receiving module 520a is configured to receive performance information from the third network element. The processing module 510 is configured to modify the first policy or the first AI model based on the performance information of the third network element.

[0239] In one example, the storage module 530 may store computer-executable instructions for a method performed by a first network element to enable the processing module 510, the receiving module 520a, and the transmitting module 520b to perform the method performed by the first network element in the aforementioned example.

[0240] In one example, the communication device 500 may be a second network element, or may be a chip or functional unit used in the second network element. The communication device 500 has any function of the second network element in the above-mentioned method. For example, the communication device 500 may perform the steps performed by the second network element in the method of FIG. 3.

[0241] The receiving module 520a may perform the receiving actions performed by the second network element in the method embodiments described above.

[0242] The sending module 520b may perform the sending action performed by the second network element in the method embodiments described above.

[0243] The processing module 510 may perform actions other than the sending and receiving actions performed by the second network element in the above method embodiments.

[0244] In one example, the receiving module 520a is configured to receive, from a first network element, instruction information indicating that an exception occurs when the first network element executes a first policy, where the first policy is a policy output by the first AI model after the first network element inputs a first measurement quantity to be input to the first AI model. The sending module 520b is configured to send, to a third network element, second information indicating that an exception occurs when the first policy is executed or that an exception occurs when a first type policy output by the first AI model is executed, where the type of the first policy is a first type.

[0245] In one example, the third network element satisfies any one of the following conditions: the third network element is executing the policy output by the first AI model; the third network element is executing the first policy; or the third network element is executing a fifth policy, where the fifth policy and the first policy are of the same type.

[0246] In one example, the processing module 510 is configured to determine a corrective manner that is allowed to be used for the execution exception. The sending module 520b is further configured to send first information to the first network element, where the first information indicates the corrective manner that is allowed to be used for the execution exception.

[0247] In one example, acceptable corrective actions include: executing a second policy, where the second policy is a policy obtained based on a non-AI model and the second policy and the first policy are of the same type; executing a third policy, where the third policy is a policy determined based on a second AI model and the third policy and the first policy are of the same type; executing a fourth policy, where the fourth policy is a policy executed before the first policy and the fourth policy and the first policy are of the same type; or skipping the execution of a policy.

[0248] In one example, when the corrective action permitted to be taken is to execute a third policy, the first information includes at least one of an identifier of the third policy, the third policy, an identifier of the second AI model, parameters of the second AI model, or performance information of a third network element, and the performance information of the third network element is used to determine the third policy.

[0249] In one example, the receiving module 520a is configured to receive performance information from a third network element. The processing module 510 is configured to: determine the performance information of the third network element as information used to correct a first AI model, determine the performance information of the third network element as information used to correct a first policy, determine information regarding a second AI model based on the performance information of the third network element and the AI, or determine information regarding a third policy based on the performance information of the third network element and the AI.

[0250] In one example, the receiving module 520a is further configured to receive performance information of a third network element.

[0251] In one example, the storage module 530 may store computer-executable instructions for a method performed by a second network element to enable the processing module 510, the receiving module 520a, and the transmitting module 520b to perform the method performed by the second network element in the aforementioned example.

[0252] In one example, the communication device 500 may be a third network element, or may be a chip or functional unit used in the third network element. The communication device 500 has any function of the third network element in the above-mentioned method. For example, the communication device 500 may perform the steps performed by the third network element in the method of FIG. 3 and FIG. 4.

[0253] The receiving module 520a may perform the receiving actions performed by the third network element in the above method embodiments.

[0254] The sending module 520b may perform the sending action performed by the third network element in the above method embodiment.

[0255] The processing module 510 may perform actions other than the sending and receiving actions performed by the third network element in the above method embodiments.

[0256] In one example, the receiving module 520a is configured to receive second information from the second network element, the second information indicating that an exception occurs when the first policy is executed or that an exception occurs when a first type of policy output by the first AI model is executed, the first policy belonging to the first type. The sending module 520b is configured to send performance information of the third network element to the second network element. The processing module 510 is configured to determine whether an exception occurs with respect to the policy executed by the third network element.

[0257] In one example, the receiving module 520a is configured to receive, from the first network element, instruction information indicating that an exception occurs when the first network element executes a first policy, the first policy being a policy output by the first AI model after the first network element inputs a first measurement quantity to be input to the first AI model. The sending module 520b is configured to send performance information of a third network element to the first network element. The processing module 510 is configured to determine whether an exception occurs with respect to the policy executed by the third network element.

[0258] In one example, the storage module 530 may store computer-executable instructions for a method performed by a third network element to enable the processing module 510, the receiving module 520a, and the transmitting module 520b to perform the method performed by the third network element in the above example.

[0259] For example, a storage module may include one or more memories. A memory may be one or more devices or components in a circuit configured to store programs or data. A storage module may be a register, a cache, a RAM, etc. A storage module may be integrated with a processing module. A storage module may be a ROM or another type of static storage device capable of storing static information and instructions. A storage module may be separate from a processing module.

[0260] The transceiver module may be an input / output interface, pins, circuitry, etc.

[0261] In one possible product form, the device may be implemented using a general-purpose bus architecture.

[0262] FIG. 6 is a block diagram of a communication device 600.

[0263] The communication device 600 may include a processor 610 and, optionally, further includes a transceiver 620 and a memory 630. The transceiver 620 may be configured to receive programs or instructions and transmit the programs or instructions to the processor 610. Alternatively, the transceiver 620 may be configured to perform a communication interaction between the communication device 600 and another communication device, for example, to exchange control signaling and / or service data. The transceiver 620 may be a code and / or data read / write transceiver, or the transceiver 620 may be a signal transmission transceiver between the processor and the transceiver machine. The processor 610 and the memory 630 are electrically coupled.

[0264] In one example, the communication device 600 may be a first network element or a chip used in the first network element. It should be understood that the device has any functionality of the first network element in the aforementioned methods. For example, the communication device 600 may perform the steps performed by the first network element in the methods of FIGS. 3 and 4. For example, the memory 630 may be configured to store a computer program. The processor 610 may be configured to invoke the computer program or instructions stored in the memory 630 to perform the method performed by the first network element in the aforementioned examples, or to perform the method performed by the first network element in the aforementioned examples by using the transceiver 620.

[0265] In one example, the communication device 600 may be a second network element or a chip used in a second network element. It should be understood that the device has any functionality of the second network element in the aforementioned method. For example, the communication device 600 may perform the steps performed by the second network element in the method of FIG. 3. For example, the memory 630 may be configured to store a computer program. The processor 610 may be configured to invoke the computer program or instructions stored in the memory 630 to perform the method performed by the second network element in the aforementioned example, or to perform the method performed by the second network element in the aforementioned example by using the transceiver 620.

[0266] In one example, the communication device 600 may be a third network element or a chip used in a third network element. It should be understood that the device has any functionality of the third network element in the aforementioned methods. For example, the communication device 600 may perform the steps performed by the third network element in the methods of FIGS. 3 and 4. For example, the memory 630 may be configured to store a computer program. The processor 610 may be configured to invoke the computer program or instructions stored in the memory 630 to perform the method performed by the third network element in the aforementioned examples, or to perform the method performed by the third network element in the aforementioned examples by using the transceiver 620.

[0267] The processing module 510 of FIG. 5 may be implemented using a processor 610 .

[0268] 5 may be implemented by using the transceiver 620. Alternatively, the transceiver 620 includes a receiver and a transmitter. The receiver performs the functions of the receiving module, and the transmitter performs the functions of the transmitting module.

[0269] The storage module 530 of FIG. 5 may be implemented by using the memory 630 .

[0270] As a possible product form, the apparatus may be implemented using a general purpose processor (which may also be called a chip or system on a chip).

[0271] In one possible implementation, a general-purpose processor implementing the device used in the first network element, or the device used in the second network element or the third network element includes a processing circuit (the processing circuit may also be referred to as a processor). Optionally, the general-purpose processor further includes a storage medium (the storage medium may also be referred to as a memory) and an input / output interface for internal connection to and communication with the processing circuit, the storage medium being configured to store instructions executed by the processing circuit to perform the method performed by the first network element, the second network element, or the third network element in the aforementioned examples.

[0272] The processing module 510 of FIG. 5 may be implemented using a processing circuit.

[0273] 5 may be implemented using an input / output interface. Alternatively, the input / output interface includes an input interface and an output interface. The input interface performs the function of the receiving module, and the output interface performs the function of the transmitting module.

[0274] The storage module 530 of FIG. 5 may be implemented by using a storage medium.

[0275] As a possible product form, the apparatus in the embodiments of the present application may be further implemented using one or more FPGAs (Field Programmable Gate Arrays), PLDs (Programmable Logic Devices), controllers, state machines, gate logic, discrete hardware components, any other suitable circuitry, or any combination of circuitry capable of performing the various functions described herein.

[0276] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, which, when executed by a computer, enables the computer to perform the aforementioned communication method. In other words, the computer program includes instructions for implementing the aforementioned communication method.

[0277] An embodiment of the present application further provides a computer program product, which includes computer program code that, when executed on a computer, enables the computer to perform the communication method described above.

[0278] An embodiment of the present application further provides a communication system, including one or more of a first network element, a second network element, and a third network element that perform the aforementioned communication method.

[0279] In addition, the processor referred to in the embodiments of this application may be a central processing unit (CPU) or a baseband processor. The baseband processor and the CPU may be integrated or separate, or may be a network processor (NP) or a combination of a CPU and an NP. The processor may further include a hardware chip or another general-purpose processor. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), another programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, or any combination thereof. The general-purpose processor may be a microprocessor, or the processor may be any conventional processor, etc.

[0280] The memory referred to in the embodiments of the present application may be volatile memory or nonvolatile memory, or may include volatile memory and nonvolatile memory. Nonvolatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may be random access memory (RAM) used as an external cache. By way of example and not limitation, many forms of RAM may be used, such as static random access memory (Static RAM, SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (Synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), Synchlink dynamic random access memory (Synchlink DRAM, SLDRAM), and direct Rambus random access memory (Direct Rambus RAM, DR RAM). Note that memory as described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0281] The transceiver referred to in the embodiments of this application may include a separate transmitter and / or a separate receiver, or the transmitter and receiver may be integrated. The transceiver may operate according to instructions of a corresponding processor. Optionally, the transmitter may correspond to a transmitter machine in a physical device, and the receiver may correspond to a receiver machine in a physical device.

[0282] Those skilled in the art may recognize that the method steps and units described with reference to the embodiments disclosed herein may be implemented by electronic hardware, computer software, or a combination thereof. To clearly explain the compatibility between hardware and software, the above generally describes the steps and configurations of each embodiment according to their functions. Whether a function is performed by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementations should not be considered to go beyond the scope of the present application.

[0283] In some embodiments provided herein, it should be understood that the disclosed systems, devices, and methods may be implemented in other ways. For example, the described device embodiments are merely examples. For example, the division into units is merely a logical division of function, and other divisions may be used in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not implemented. In addition, the shown or discussed mutual couplings or direct couplings or communication connections may be implemented through some interfaces, indirect couplings or communication connections between devices or units, or electrical, mechanical, or other forms of connection.

[0284] The units described as separate parts may or may not be physically separate, and the parts shown as units may or may not be physical units, and may be located in one location or distributed over multiple network units. Some or all of the units may be selected according to actual requirements to achieve the objectives of the solutions in the embodiments of the present application.

[0285] In addition, the functional units in the embodiments of the present application may be integrated into one processing unit, and each unit may exist physically alone, or two or more units may be integrated into one unit. The integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0286] When an integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, the integrated unit may be stored in a computer-readable storage medium. Based on this understanding, the technical solution in this application may essentially, or a portion that contributes to the prior art, or all or part of the technical solution may be expressed in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for instructing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method described in the embodiments of this application. The aforementioned storage medium includes any medium capable of storing program code, such as a USB flash drive, a removable hard disk drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0287] In this application, "and / or" describes an association relationship for describing associated objects and indicates that three relationships may exist. For example, A and / or B can represent three cases: only A is present, both A and B are present, or only B is present. The symbol " / " generally indicates an "or" relationship between associated objects. In this application, "plurality" means two or more. It should also be understood that in the description of this application, terms such as "first," "second," etc. are used merely for distinction and explanation, and should not be understood as indicating or implying relative importance or as indicating or implying a sequence.

[0288] Although the preferred embodiments of the present application have been described, other changes and modifications can be made to the embodiments once those skilled in the art are made aware of the basic inventive concepts. It is therefore intended that the following claims be interpreted to encompass the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0289] Obviously, those skilled in the art can make various modifications and variations to the embodiments of the present application without departing from the scope of the embodiments of the present application. Thus, the present application intends to cover these modifications and variations to the embodiments of the present application as long as they fall within the scope of protection defined by the following claims of the present application and their equivalent technologies.

Claims

1. A communication method applied to a first network element, comprising: Obtaining a first measurement to be input and obtaining a first artificial intelligence (AI) model; inputting the first measurement into the first AI model to obtain a first policy output by the first AI model; executing the first policy; and when determining that an exception occurs when the first policy is executed, sending indication information to a second network element, the indication information indicating that the exception occurs when the first policy is executed. method.

2. The first policy type is: Energy saving policy, load balancing policy, mobility optimization policy, channel state information-reference signal (CSI-RS) feedback improvement policy, beam management improvement policy, or positioning accuracy improvement policy The method of claim 1 , comprising any one of:

3. The instruction information indicating that the exception will occur when the first policy is executed includes: the execution exception includes one or more of an indication of the execution exception, a time when the execution exception occurs, a cause of the execution exception, a measurement parameter used to determine that the exception occurs with respect to the policy, a configuration parameter used to determine that the exception occurs with respect to the policy, a validity period of the configuration parameter, a corrective manner that the first network element is expected to use for the execution exception, information necessary for the corrective manner that the first network element is expected to use for the execution exception, an identifier of the first AI model, a parameter of the first AI model, an identifier of the first policy, the first policy, or a validity period of the first policy; The method according to claim 1 or 2.

4. receiving first information from the second network element, the first information indicating a corrective action that is allowed to be used for the execution exception; remediating the exception generated with respect to the first policy based on the remediation manner; The method of any one of claims 1 to 3, further comprising:

5. The corrective measures that may be used are: Executing a second policy, the second policy being a policy obtained based on a non-AI, and the second policy and the first policy being of the same type; Executing a third policy, the third policy being a policy determined based on a second AI model, and the third policy and the first policy being of the same type; or executing a fourth policy, the fourth policy being a policy that is executed before the first policy, and the fourth policy and the first policy being of the same type; 5. The method of claim 4, comprising any one of:

6. When the corrective manner that is allowed to be used is to implement the third policy, the first information is: the third policy identifier, the third policy identifier, the second AI model identifier, parameters of the second AI model, or performance information of a third network element, wherein the performance information of the third network element is used to determine the third policy; The method of claim 5.

7. Remediating the exception generated with respect to the first policy based on the remediation manner includes: when the first information includes the identifier of the third policy, executing the third policy indicated by the identifier of the third policy to remedy the exception generated with respect to the first policy; when the first information includes the third policy, executing the third policy to remedy the exception generated with respect to the first policy; When the first information includes the identifier of the second AI model, inputting a second measurement into the second AI model to obtain the third policy, and executing the third policy to correct the exception generated with respect to the first policy, wherein the second measurement is the same as the first measurement or a measurement time of the second measurement is not earlier than a measurement time of the first measurement; When the first information includes the parameter of the second AI model, inputting a second measurement into the second AI model corresponding to the parameter to obtain the third policy, and executing the third policy to correct the exception generated with respect to the first policy, wherein the second measurement is the same as the first measurement or a measurement time of the second measurement is not earlier than a measurement time of the first measurement; or When the first information includes the performance information of the third network element, correcting the first AI model based on the performance information to obtain the second AI model, inputting a third measurement amount into the second AI model to obtain the third policy, and executing the third policy to correct the exception generated with respect to the first policy, wherein the third measurement amount is the same as the first measurement amount or the measurement time of the third measurement amount is not earlier than the measurement time of the first measurement amount; The method of claim 6.

8. The first network element is a first access network device, and the second network element is a core network element or an operation, administration, and maintenance (OAM) network element; The first network element is a first terminal device and the second network element is an access network device; the first network element is a first access network device and the second network element is a second access network device; or the first network element is a first terminal device and the second network element is a second terminal device; 8. The method according to any one of claims 1 to 7.

9. A communication method applied to a second network element, comprising: receiving instruction information from a first network element, the instruction information indicating that an exception occurs when the first network element executes a first policy, the first policy being a policy output by a first artificial intelligence (AI) model after the first network element inputs a first measurement quantity to be input to the first AI model; sending second information to a third network element, the second information indicating that the exception occurs when the first policy is executed or that an exception occurs when a first type policy output by the first AI model is executed, the type of the first policy being a first type; A method comprising:

10. The first policy type is:

10. The method of claim 9, comprising any one of an energy saving policy, a load balancing policy, a mobility optimization policy, a channel state information-reference signal (CSI-RS) feedback enhancement policy, a beam management enhancement policy, or a positioning accuracy enhancement policy.

11. The instruction information indicating that the exception will occur when the first policy is executed includes: the execution exception includes one or more of an indication of the execution exception, a time at which the execution exception occurs, a cause of the execution exception, a measurement parameter used to determine that the exception occurs with respect to the policy, a configuration parameter used to determine that the exception occurs with respect to the policy, a validity period of the configuration parameter, a corrective manner that the first network element is expected to use for the execution exception, information necessary for the corrective manner that the first network element is expected to use for the execution exception, an identifier of the first AI model, a parameter of the first AI model, an identifier of the first policy, or a validity period of the first policy; 11. The method according to claim 9 or 10.

12. The third network element satisfies the following conditions: the third network element executing the policy output by the first AI model; the third network element is implementing the first policy; or the third network element is executing a fifth policy, wherein the fifth policy and the first policy are of the same type; 12. The method according to claim 9, wherein any one of the following conditions is satisfied:

13. determining an allowable remedial action to be taken for the execution exception; transmitting first information to the first network element, the first information indicating the corrective manner that is allowed to be used for the execution exception; 13. The method of any one of claims 9 to 12, further comprising:

14. The corrective measures that may be used are: Executing a second policy, wherein the second policy is a policy obtained based on a non-AI, and the second policy and the first policy are of the same type; Executing a third policy, the third policy being a policy determined based on a second AI model, and the third policy and the first policy being of the same type; or Executing a fourth policy, the fourth policy being a policy that is executed before the first policy, and the fourth policy and the first policy being of the same type.

14. The method of claim 13, comprising any one of:

15. When the corrective manner that is allowed to be used is to implement the third policy, the first information is: an identifier of the third policy, the third policy, an identifier of the second AI model, parameters of the second AI model, or performance information of the third network element; wherein the performance information of the third network element is used to determine the third policy.

15. The method of claim 14.

16. The performance information of the third network element is used to determine a third policy, the performance information of the third network element is used to correct the first AI model, and the corrected first AI model is used to determine the third policy.

16. The method of claim 15.

17. receiving the performance information from the third network element.

17. The method of claim 15 or 16.

18. The first network element is a first access network device, the second network element is a core network element or an Operation, Administration, and Maintenance (OAM) network element, and the third network element is a second access network device; the first network element is a first terminal device, the second network element is an access network device, and the third network element is a second terminal device; or the first network element is a terminal device, the second network element is a first access network device, and the third network element is a second access network device; 18. The method according to any one of claims 9 to 17.

19. A communication method applied to a third network element, comprising: receiving second information from a second network element, the second information indicating that an exception occurs when a first policy is executed or that an exception occurs when a policy of a first type output by a first AI model is executed, the first policy belonging to a first type, and the first policy being obtained based on the first AI model; transmitting performance information of the third network element to the second network element and / or determining whether an exception occurs with respect to a policy implemented by the third network element; A method comprising:

20. A communication method applied to a first network element, comprising: obtaining a first measurement to be input and obtaining a first AI model; inputting the first measurement into the first AI model to obtain a first policy output by the first AI model; executing the first policy; when determining that an exception occurs when the first policy is executed, sending indication information to a third network element, the indication information indicating that the exception occurs when the first policy is executed; receiving performance information from the third network element; correcting the first policy or the first AI model based on the performance information of the third network element; A method comprising:

21. A communication method applied to a third network element, comprising: receiving instruction information from a first network element, the instruction information indicating that an exception occurs when the first network element executes a first policy, the first policy being a policy output by the first AI model after the first network element inputs a first measurement quantity to be input to the first AI model; transmitting performance information of the third network element to the first network element and / or determining whether an exception occurs with respect to a policy implemented by the third network element; A method comprising:

22. A communication device comprising a functional module for implementing the method according to any one of claims 1 to 21.

23. 1. A communications device having a processor, the processor coupled to a memory; the memory is configured to store computer programs or instructions; The processor executes some or all of the computer program or instructions in the memory, and when some or all of the computer program or instructions are executed, performs the method of any one of claims 1 to 21. Communication equipment.

24. 1. A communications device having a processor and a memory, the memory is configured to store computer programs or instructions; The processor executes some or all of the computer program or instructions in the memory, and when some or all of the computer program or instructions are executed, performs the method of any one of claims 1 to 21. Communication equipment.

25. a chip system, the chip system having a processing circuit, the processing circuit coupled to a storage medium; The processing circuitry executes some or all of the computer programs or instructions on the storage medium, and when some or all of the computer programs or instructions are executed, performs the method of any one of claims 1 to 21. Chip system.

26. 22. A computer readable storage medium adapted to store a computer program, the computer program comprising instructions for carrying out the method of any one of claims 1 to 21.

27. 22. A computer program product comprising computer program code which, when run on a computer, enables the computer to carry out a method according to any one of claims 1 to 21.

Images