Communication methods and communication devices

Abstract

Embodiments of the present invention provide a communication method and a communication device. The method comprises the step of sending a report indicating a switch in an artificial intelligence (AI) model or a switch in a mode when an entry or exit condition for an event is met. The method provided herein defines a plurality of events and corresponding entry and exit conditions for such events, thereby enabling an active model or mode switch when such conditions are met.

Description

[Technical Field]

[0001] This application relates to U.S. Provisional Patent Application No. 63 / 507,790, entitled “A METHOD OF AI MODEL MONITOR,” filed on 13 June 2023, and claims priority based on said U.S. Provisional Patent Application.

[0002] The application disclosures mentioned above are incorporated herein by reference in their entirety.

[0003] Embodiments of this application relate to the field of communications, and more specifically, to communication methods and communication devices. [Background technology]

[0004] AI-based algorithms are being introduced into modern wireless communications to solve several wireless problems, such as channel estimation, scheduling, channel status information (CSI) compression (from user equipment to base stations), multiple-input multiple-output (MIMO) beamforming, and positioning. As a data-driven method, AI-based algorithms inevitably suffer from low generalizability. The performance of artificial intelligence (AI) models is only as good as the data on which they are trained. Even if an AI model is trained on a large dataset, it may not possess the knowledge necessary to perform effectively in other environments, especially in wireless communications where channel information changes rapidly.

[0005] Due to generalization issues, user equipment (UE) or base stations (BS) need to detect their generalization performance and then switch to the appropriate AI model. For example, multiple AI models exist for multiple scenarios, including indoor cities, outdoor cities, rural areas, and high-speed rail. If the surrounding environment changes, the BS or UE needs to switch to a different model. In conventional technology, if inference performance deteriorates, the UE or BS either switches its AI model or falls back to a non-AI mode. This is a kind of reactive solution.

[0006] Therefore, finding a way to perform AI model switching in advance is an urgent technical problem that needs to be solved. [Overview of the Initiative]

[0007] Embodiments of the present invention provide a communication method and a communication device. In the technical solution of the present invention, a plurality of events, and entry and exit conditions corresponding to said events, are defined such that an active model or mode switching can be achieved when any one of these conditions is met.

[0008] According to a first aspect, one embodiment of the present invention provides a communication method comprising the step of sending a report indicating a switch in an artificial intelligence (AI) model or a switch in a mode when an entry or exit condition for an event is met.

[0009] The method provided in this application defines a plurality of events and corresponding entry and exit conditions, thereby enabling an active model or mode switch when the conditions are met.

[0010] The event can be any of the following: switching from an AI model to a non-AI model, switching from the first AI model to the second AI model, switching from a non-AI model to an AI model, or switching from the third AI model to the fourth AI model. Of these, the third AI model can be the same as the second AI model, and the fourth AI model can be the same as the first AI model.

[0011] In a possible implementation, the event is an action that switches from AI mode to non-AI mode, where the entry condition is determined by Diff11 and Thresh11, or the exit condition is determined by Diff11 and Thresh12, where Diff11 is a first difference value, Thresh11 is a first threshold, and Thresh12 is a second threshold, where the first difference value is the difference between a first data and a first anchor, where the first anchor contains one or more pieces of reference data, and the first and second thresholds are predefined or configured thresholds corresponding to the event.

[0012] The first data includes monitoring or measurement data from user equipment or network devices. Furthermore, the first data is monitoring or measurement data related to the AI ​​model. The first data may include any one or more of the following: detection data, measurement data, channel data, neuron data of the AI ​​model, and latent output data of the AI ​​model. The network device may be a BS.

[0013] The first anchor can be indicated by the BS, for example, the first anchor is the data anchor of the current AI / machine learning (ML) model. Alternatively, the first anchor is the data anchor closest to the first data, i.e., the first anchor is the data anchor with the smallest difference from the first data among multiple anchors.

[0014] Events related to the operation of switching from AI mode to non - AI mode may be referred to as event K1. Event K1 means that the current or best AI / ML model may be obsolete and the UE or BS should be switched to non - AI mode.

[0015] The method provided by this application defines event K1 and the entry and exit conditions corresponding to event K1, and can realize an active switch of the mode when any of the conditions is met.

[0016] In a possible implementation, the entry condition is Diff11 - Hys11>Thresh11, or the exit condition is Diff11 + Hys12<Thresh12.

[0017] Hys11 is the first hysteresis parameter, Hys12 is the second hysteresis parameter, and the first hysteresis parameter and the second hysteresis parameter are pre - defined or configured parameters corresponding to event K1.

[0018] The units of Thresh11, Thresh12, Hys11, and Hys12 are the same as the unit of Diff11.

[0019] Hys11 and Hys12 can be the same or different. For example, they can be similar. Thresh11 and Thresh12 can be the same or different. For example, they can be similar.

[0020] In a possible implementation, the entry condition is Diff11+offset11 - Hys11>Thresh11, or the exit condition is Diff11+offset12 + Hys12<Thresh12, where offset11 and offset12 are pre - defined or configured offsets.

[0021] offset11 and offset12 can be the same or different. The values of offset11 and offset12 can be positive or negative numbers.

[0022] The method provided by the present application defines an event K1 and entry and exit conditions corresponding to the event K1, and can realize an active switching of the mode when any of the conditions is satisfied.

[0023] In a possible implementation, the event is an operation of switching from a first AI model to a second AI model, where the entry condition is determined by Diff21, Diff22, Thresh21, and Thresh22, or the exit condition is determined by Diff21, Diff22, Thresh23, and Thresh24, where Diff21 is a second difference value, Diff22 is a third difference value, Thresh21 is a third threshold value, Thresh22 is a fourth threshold value, Thresh23 is a fifth threshold value, Thresh24 is a sixth threshold value, the second difference value is a difference value between the first data and the second anchor, the second anchor is a data anchor associated with the first AI model, the third difference value is a difference value between the first data and the third anchor, the third anchor is a data anchor associated with the second AI model, the third threshold value and the fifth threshold value are predefined or configured threshold values corresponding to the first AI model, and the fourth threshold value and the sixth threshold value are predefined or configured threshold values corresponding to the second AI model.

[0024] The event related to the operation of switching from the first AI model to the second AI model can be referred to as event K2. Event K2 means that the current AI / ML model (referred to as model 1) can be obsolete and the UE should be switched to another model (model 2).

[0025] In event K2, the current model becomes worse than Thresh21, and another model becomes better than Thresh22. Being worse than Thresh21 means that the second difference value between the first data and the second anchor is greater than Thresh21. Being better than Thresh22 means that the third difference value between the first data and the third anchor is less than Thresh22. Thresh21 and Thresh22 can be configured to be the same or different.

[0026] The method provided by the present application can define event K2 and the corresponding entry and exit conditions for the event K2, and realize the active switching of the model when any of the conditions is satisfied.

[0027] In a possible implementation, the entry condition is Diff21 - Hys21 > Thresh21 and Diff22 + Hys22 < Thresh22, or the exit condition is Diff21 + Hys23 < Thresh23 or Diff22 - Hys24 > Thresh24.

[0028] Hys21 is the third hysteresis parameter, Hys22 is the fourth hysteresis parameter, Hys23 is the fifth hysteresis parameter, Hys24 is the sixth hysteresis parameter, and the third hysteresis parameter, the fourth hysteresis parameter, the fifth hysteresis parameter and the sixth hysteresis parameter are predefined or configured parameters corresponding to event K2.

[0029] Hys21 and Hys23 can be the same or different. For example, they can be similar. Hys22 and Hys24 can be the same or different. For example, they can be similar. Thresh21 and Thresh23 can be the same or different. For example, they can be similar. Thresh22 and Thresh24 can be the same or different. For example, they can be similar.

[0030] In a possible implementation, the entry conditions are Diff21 + offset21 - Hys21 > Thresh21 and Diff22 + offset22 + Hys22 < Thresh22, and the exit conditions are Diff21 + offset23 + Hys23 < Thresh23 or Diff22 + offset24 - Hys24 > Thresh24, where offset21, offset22, offset23, and offset24 are predefined or configured offsets.

[0031] offset21 and offset23 can be the same or different. For example, they can be similar. offset22 and offset24 can be the same or different. For example, they can be similar. The values of offset21, offset22, offset23, and offset24 can be positive or negative numbers.

[0032] The method provided by this application can define event K2 and the entry and exit conditions corresponding to the event K2, and can realize the active switching of the model when any of the conditions is satisfied.

[0033] In a possible implementation, the event is an operation of switching from a non-AI mode to an AI mode, where the entry condition is determined by Diff31 and Thresh31, or the exit condition is determined by Diff31 and Thresh32, where Diff31 is the fourth difference value, Thresh31 is the seventh threshold value, Thresh32 is the eighth threshold value, the fourth difference value is the difference value between the first data and the fourth anchor, the fourth anchor includes one or more pieces of reference data, and the seventh threshold value and the eighth threshold value are predefined or configured threshold values corresponding to the event.

[0034] The event related to the operation of switching from a non-AI mode to an AI mode may be referred to as event K3. Event K3 means that the AI / ML model becomes better than the threshold value and the UE or BS should be switched from the non-AI mode to the AI mode.

[0035] In event K3, the fourth difference value between the first data and the fourth anchor is smaller than the threshold value. The fourth anchor can be indicated by the BS, or the fourth anchor is the data anchor closest to the first data, that is, the fourth anchor is the data anchor with the smallest difference from the first data among a plurality of anchors.

[0036] The method provided by this application defines event K3 and the entry and exit conditions corresponding to event K3, and can realize an active switch of the mode when any of the conditions is satisfied.

[0037] In a possible implementation, the entry condition is Diff31 + Hys31 < Thresh31, or the exit condition is Diff31 - Hys32 > Thresh32.

[0038] Hys31 is the seventh hysteresis parameter, Hys32 is the eighth hysteresis parameter, and the seventh and eighth hysteresis parameters are predefined or configured parameters corresponding to event K3.

[0039] The units of Thresh31, Thresh32, Hys31, and Hys32 are the same as those of Diff31.

[0040] Hys31 and Hys32 can be the same or different, for example, they can be similar. Thresh31 and Thresh32 can be the same or different, for example, they can be similar.

[0041] In a possible implementation, the entry condition is Diff31 + offset31 + Hys31 < Thresh31, and the exit condition is Diff31 + offset32 - Hys32 > Thresh32, where offset31 and offset32 are predefined or configured offsets.

[0042] offset31 and offset32 can be the same or different. The values of offset31 and offset32 can be positive or negative numbers.

[0043] The method provided by this application can define event K3 and the entry and exit conditions corresponding to the event K3, and realize the active switching of the mode when any of the conditions is satisfied.

[0044] In a possible implementation, the event is an operation of switching from a third AI model to a fourth AI model, where the entry condition is determined by Diff41 and Diff42, or the exit condition is determined by Diff41 and Diff42, Diff41 is a fifth difference value, Diff42 is a sixth difference value, the fifth difference value is the difference value between the first data and the fifth anchor, the fifth anchor is a data anchor associated with the third AI model, the sixth difference value is the difference value between the first data and the sixth anchor, and the sixth anchor is a data anchor associated with the fourth AI model.

[0045] The event related to the operation of switching from the third AI model to the fourth AI model can be referred to as event K4. In event K4, another model (model 2) is better than the current model (model 1), and the UE or BS should switch to model 2.

[0046] The method provided by the present application defines event K4 and the corresponding entry and exit conditions, and can achieve an active switch of the model when any of the conditions is satisfied.

[0047] In a possible implementation, the entry condition is Diff42 + Hys41 < Diff41, and the exit condition is Diff42 - Hys42 > Diff41.

[0048] Hys41 is a ninth hysteresis parameter, Hys42 is a tenth hysteresis parameter, and the ninth hysteresis parameter and the tenth hysteresis parameter are pre-defined or configured parameters corresponding to event K4.

[0049] Hys41 and Hys42 can be the same or different. For example, they can be similar.

[0050] In a possible implementation, the entry condition is Diff42 + offset42 + Hys41 < Diff41 + offset41, and the exit condition is Diff42 + offset42 - Hys42 > Diff41 + offset41, where offset41 and offset42 are pre-defined or configured offsets.

[0051] offset41 and offset42 can be the same or different. The values of offset41 and offset42 can be positive or negative numbers.

[0052] The method provided by this application defines event K4 and the corresponding entry and exit conditions for event K4, and can realize the active switching of the model when any of these conditions is satisfied.

[0053] In a possible implementation, the first data includes monitoring data or measurement data of a user device or a network device.

[0054] In a possible implementation, the first data includes monitoring data or measurement data related to an AI model of a user device or a network device.

[0055] In a possible implementation, the first data includes any one or more of detection data, measurement data, channel data, neuron data of an AI model, and potential output data of an AI model.

[0056] In a possible implementation, the method further includes the step of performing communication based on a report.

[0057] In a possible implementation, the method is executed by a user device or a network device.

[0058] According to a second aspect, the present application provides a communication device comprising a sending module configured to send a report indicating a switching of an artificial intelligence (AI) model or a mode switching when an entry condition or an exit condition of an event is satisfied.

[0059] In a possible implementation, the event is an operation of switching from an AI mode to a non-AI mode, where the entry condition is determined by Diff11 and Thresh11, or the exit condition is determined by Diff11 and Thresh12, where Diff11 is a first difference value, Thresh11 is a first threshold value, Thresh12 is a second threshold value, the first difference value is a difference value between first data and a first anchor, the first anchor includes one or more pieces of reference data, and the first threshold value and the second threshold value are pre-defined or configured threshold values corresponding to the event.

[0060] In a possible implementation, the entry condition is Diff11 - Hys11 > Thresh11, or the exit condition is Diff11 + Hys12 < Thresh12.

[0061] In a possible implementation, the entry condition is Diff11 + offset11 - Hys11 > Thresh11, or the exit condition is Diff11 + offset12 + Hys12 < Thresh12, where offset11 and offset12 are pre-defined or configured offsets.

[0062] In a possible implementation, Hys11 is a first hysteresis parameter, Hys12 is a second hysteresis parameter, and the first hysteresis parameter and the second hysteresis parameter are pre-defined or configured parameters corresponding to the event.

[0063] In a possible implementation, the event is an operation of switching from a first AI model to a second AI model, where the entry condition is determined by Diff21, Diff22, Thresh21, and Thresh22, or the exit condition is determined by Diff21, Diff22, Thresh23, and Thresh24, where Diff21 is a second difference value, Diff22 is a third difference value, Thresh21 is a third threshold value, Thresh22 is a fourth threshold value, Thresh23 is a fifth threshold value, Thresh24 is a sixth threshold value, the second difference value is the difference value between the first data and the second anchor, the second anchor is a data anchor associated with the first AI model, the third difference value is the difference value between the first data and the third anchor, the third anchor is a data anchor associated with the second AI model, the third threshold value and the fifth threshold value are predefined or configured threshold values corresponding to the first AI model, and the fourth threshold value and the sixth threshold value are predefined or configured threshold values corresponding to the second AI model.

[0064] In a possible implementation, the entry condition is Diff21 - Hys21 > Thresh21 and Diff22 + Hys22 < Thresh22, or the exit condition is Diff21 + Hys23 < Thresh23 or Diff22 - Hys24 > Thresh24.

[0065] In a possible implementation, the entry condition is Diff21 + offset21 - Hys21 > Thresh21 and Diff22 + offset22 + Hys22 < Thresh22, and the exit condition is Diff21 + offset23 + Hys23 < Thresh23 or Diff22 + offset24 - Hys24 > Thresh24, where offset21, offset22, offset23, and offset24 are predefined or configured offsets.

[0066] In a possible implementation, Hys21 is the third hysteresis parameter, Hys22 is the fourth hysteresis parameter, Hys23 is the fifth hysteresis parameter, Hys24 is the sixth hysteresis parameter, and the third, fourth, fifth, and sixth hysteresis parameters are predefined or configured parameters corresponding to an event.

[0067] In a possible implementation, the event is an operation of switching from a non-AI mode to an AI mode, where the entry condition is determined by Diff31 and Thresh31, or the exit condition is determined by Diff31 and Thresh32, where Diff31 is the fourth difference value, Thresh31 is the seventh threshold value, Thresh32 is the eighth threshold value, the fourth difference value is the difference value between the first data and the fourth anchor, the fourth anchor includes one or more pieces of reference data, and the seventh and eighth threshold values are predefined or configured threshold values corresponding to the event.

[0068] In a possible implementation, the entry condition is Diff31 + Hys31 < Thresh31, or the exit condition is Diff31 - Hys32 > Thresh32.

[0069] In a possible implementation, the entry condition is Diff31 + offset31 + Hys31 < Thresh31, and the exit condition is Diff31 + offset32 - Hys32 > Thresh32, where offset31 and offset32 are predefined or configured offsets.

[0070] In a possible implementation, Hys31 is the seventh hysteresis parameter, Hys32 is the eighth hysteresis parameter, and the seventh and eighth hysteresis parameters are predefined or configured parameters corresponding to the event.

[0071] In a possible implementation, the event is an operation of switching from a third AI model to a fourth AI model, where the entry condition is determined by Diff41 and Diff42, or the exit condition is determined by Diff41 and Diff42, Diff41 is a fifth difference value, Diff42 is a sixth difference value, the fifth difference value is the difference value between the first data and the fifth anchor, the fifth anchor is a data anchor associated with the third AI model, the sixth difference value is the difference value between the first data and the sixth anchor, and the sixth anchor is a data anchor associated with the fourth AI model.

[0072] In a possible implementation, the entry condition is Diff42 + Hys41 < Diff41, and the exit condition is Diff42 - Hys42 > Diff41.

[0073] In a possible implementation, the entry condition is Diff42 + offset42 + Hys41 < Diff41 + offset41, and the exit condition is Diff42 + offset42 - Hys42 > Diff41 + offset41, where offset41 and offset42 are pre - defined or configured offsets.

[0074] In a possible implementation, Hys41 is a ninth hysteresis parameter, Hys42 is a tenth hysteresis parameter, and the ninth hysteresis parameter and the tenth hysteresis parameter are pre - defined or configured parameters corresponding to the event.

[0075] In a possible implementation, the first data includes monitoring data or measurement data of a user device or a network device.

[0076] In a possible implementation, the first data includes monitoring data or measurement data related to an AI model of a user device or a network device.

[0077] In a possible implementation, the first data includes any one or more of the following: detection data, measurement data, channel data, neuron data of the AI ​​model, and latent output data of the AI ​​model.

[0078] In possible implementations, the device further includes a processing module configured to perform communication based on the report.

[0079] In possible implementations, the device is located on user equipment or network devices.

[0080] According to a third aspect, a communication device is provided comprising a processor and memory. The processor is connected to the memory. The memory is configured to store instructions, and the processor is configured to execute instructions. When the processor executes an instruction stored in the memory, the processor is enabled to perform a method in any possible implementation of the first aspect.

[0081] According to a fourth aspect, the present invention provides a computer-readable storage medium comprising instructions. When the instructions are executed on a processor, the processor is enabled to perform the method in any possible implementation of the first aspect.

[0082] According to a fifth aspect, the present application provides a computer program product comprising computer program code. When the computer program code is executed on a computer, the computer is enabled to perform the methods in any possible implementation of the first aspect.

[0083] It should be noted that all or part of the above computer program code can be stored on a first storage medium. The first storage medium can be packaged together with the processor or separately from the processor.

[0084] According to a sixth aspect, the present invention provides a chip system comprising memory and a processor. The memory is configured to store computer programs, and the processor is configured to call computer programs from the memory and execute computer programs, thereby enabling the electronic device on which the chip system is installed to perform the method in any possible implementation of the first aspect. [Brief explanation of the drawing]

[0085] [Figure 1] This is a schematic diagram of a communication system 100 according to one embodiment of the present invention.

[0086] [Figure 2] This is a schematic diagram of a communication system 100 according to one embodiment of the present invention.

[0087] [Figure 3] This is a schematic diagram of an electronic device (ED) 110 and base stations 170a, 170b and / or 170c according to one embodiment of the present application.

[0088] [Figure 4] This is a schematic diagram of a unit or module in a device according to one embodiment of the present application.

[0089] [Figure 5] This is a schematic diagram of an AI-based communication device.

[0090] [Figure 6] This is a schematic diagram of a device 500 that receives a reference data sample from a device 600 according to one embodiment of the present invention.

[0091] [Figure 7] This is a schematic diagram of a reference data sample consisting of multiple groups according to one embodiment of the present application.

[0092] [Figure 8]This is a schematic representation of a DNN-based approximation according to one embodiment of the present application.

[0093] [Figure 9] This is a flowchart of one embodiment of a communication method according to one embodiment of the present application.

[0094] [Figure 10] This is a flowchart of one embodiment of a communication method according to one embodiment of the present application.

[0095] [Figure 11] This is a schematic diagram illustrating the projection of a high-dimensional signal onto a low-dimensional signal according to one embodiment of the present invention.

[0096] [Figure 12] This is a flowchart of a communication method according to one embodiment of the present invention.

[0097] [Figure 13] This is a schematic diagram of matrix U determined according to one embodiment of the present invention.

[0098] [Figure 14] This is a schematic diagram of the first sampling matrix P1 according to one embodiment of the present application.

[0099] [Figure 15] This is a schematic diagram of the sampling matrix compression matrix U according to one embodiment of the present invention.

[0100] [Figure 16] This is a schematic diagram of a scoring distance in low-spectrum space according to one embodiment of the present invention.

[0101] [Figure 17] This is a flowchart of a communication method according to one embodiment of the present invention.

[0102] [Figure 18] This is a schematic block diagram of a communication device according to one embodiment of the present invention.

[0103] [Figure 19] This is a schematic block diagram of another communication device according to one embodiment of the present invention. [Modes for carrying out the invention]

[0104] The following description explains the technical solution in this application with reference to the attached drawings.

[0105] The following description illustrates the technical solutions in this application with reference to the accompanying drawings. It will be clear that the embodiments described are not all, but only a selection, of the embodiments of this application. All other embodiments that can be obtained by those skilled in the art without creative effort based on the embodiments of this application are within the scope of protection.

[0106] This application presents aspects, embodiments, or features relating to a system comprising multiple devices, components, modules, etc. It should be understood and recognized that individual systems may include additional devices, components, modules, etc., and / or may not include all of the devices, components, modules, etc. discussed in relation to the accompanying drawings. Furthermore, combinations of these options may be used.

[0107] In addition, in the embodiments of this application, the words “exemplary” and the phrase “as an example” are used, for example, to indicate an illustration or explanation. No embodiment or design solution described “exemplary” in this application should be construed as being superior or advantageous to other embodiments or design solutions. Rather, the use of the word “example” is intended to present a concept in a concrete manner.

[0108] In various parts of this specification, phrases such as "in some possible embodiments" and "in some possible application scenarios" do not necessarily refer to the same embodiment, but rather, unless specifically emphasized, mean "one or more embodiments, but not all." Unless specifically emphasized, the terms "includes," "equipment," and "possess," and their variations, all mean "includes, but not limited to."

[0109] In this application, “at least one” refers to one or more, and “plural” refers to two or more. “And / or” describing the association of related objects indicates that three relationships may exist. For example, A and / or B can mean A alone, both A and B, and B alone, where A and B can be singular or plural. The letter “ / ” generally indicates that the preceding and succeeding related objects are in an “or” relationship.

[0110] The application scenarios described in the embodiments of this application are intended to more clearly illustrate the technical solutions of the embodiments of this application and do not constitute a limitation of the technical solutions provided by the embodiments of this application. It will be known to those skilled in the art that the technical solutions provided by the embodiments of this application are equally applicable to similar technical problems as system architectures evolve and new application scenarios emerge.

[0111] The technical solutions in the embodiments of this application may be applied to various communication systems such as Global Systems for Mobile Communications (GSM®), Code Division Multiple Access (CDMA) systems, Wideband Code Division Multiple Access (WCDMA®) systems, General-Purpose Packet Radio Service (GPRS) systems, Long-Term Evolution (LTE) systems, LTE Frequency Division Duplex (FDD) systems, LTE Time Division Duplex (TDD) systems, Universal Mobile Communications System (UMTS), Worldwide Interoperability for Microwave Access (WiMAX®) communication systems, Wireless Local Area Networks (WLANs), 5th generation (5G) wireless communication systems, New Radio (NR) wireless communication systems, 6th generation (6G) wireless communication systems, or other evolving communication systems.

[0112] To better explain the solutions of the embodiments in this application, concepts and terms that may be relevant to this application are explained below.

[0113] (1) Data collection

[0114] Data is a crucial component of artificial intelligence (AI) / machine learning (ML) techniques. Data collection is the process of gathering data by network nodes, management entities, or UEs for the purposes of AI / ML model training, data analysis, and inference.

[0115] (2) AI / ML model training

[0116] AI / ML model training is the process of training an AI / ML model by learning input / output relationships in a data-driven manner, and obtaining a trained AI / ML model for inference.

[0117] (3) AI / ML model inference

[0118] This is the process of using a trained AI / ML model to generate a set of outputs based on a set of inputs.

[0119] (4) AI / ML Model Validation

[0120] As a subprocess of training, validation is used to evaluate the quality of an AI / ML model using a different dataset than the one used for model training. Validation can help select model parameters that generalize beyond the dataset used for model training. Post-training model parameters can be further refined through the validation process.

[0121] (5) AI / ML model testing

[0122] Similar to validation, testing is a subprocess of training, used to evaluate the performance of the final AI / ML model using a different dataset than the one used for model training and validation. Unlike AI / ML model validation, testing does not anticipate subsequent tuning of the model.

[0123] (6) Online training

[0124] Online training refers to an AI / ML training process in which the model used for inference is trained continuously, typically in (near) real-time, as new training samples arrive.

[0125] (7) Offline training

[0126] Offline training is an AI / ML training process in which a model is trained based on a collected dataset, and the trained model is later used or delivered for inference.

[0127] (8) AI / ML model delivery / transfer

[0128] AI / ML model delivery / transfer is a general term referring to the delivery of an AI / ML model from one entity to another in any manner. Delivery of an AI / ML model over an AI interface involves either parameters of a known model structure at the receiving end, or a new model with parameters. Delivery may include a full model or a partial model.

[0129] (9) Life cycle management (LCM)

[0130] When an AI / ML model is trained and / or inferred on a single device, it is necessary to monitor and manage the entire AI / ML process to ensure the performance gains obtained by the AI / ML technology. For example, due to the randomness of wireless channels and the mobility of UEs, the wireless signal propagation environment changes frequently. Nevertheless, it is difficult for an AI / ML model to maintain optimal performance at all times in all scenarios, and performance can degrade rapidly in several scenarios. Therefore, lifecycle management (LCM) of the AI / ML model is essential for the sustainable operation of AI / ML on NR air interfaces. Lifecycle management covers the entire procedure of the AI / ML technology applied to one or more nodes. In particular, it includes at least one of the following subprocesses: data acquisition, model training, model identification, model registration, model deployment, model configuration, model inference, model selection, model activation, deactivation, model switching, model fallback, model monitoring, model update, model transfer / delivery, and UE capability reporting. Model monitoring can be based on inference accuracy, including metrics related to intermediate key performance indicators (KPIs), or it can be based on system performance, including metrics related to system performance KPIs, such as accuracy and relevance, overhead, complexity (computation and memory costs), latency (timeliness of monitoring results from model failure to action), and power consumption. Furthermore, data distributions may shift after deployment due to environmental changes; therefore, models based on input or output data distributions should also be considered.

[0131] (10) Supervised learning

[0132] The goal of supervised learning algorithms is to train a model that maps feature vectors (inputs) to labels (outputs) based on training data containing example feature-label pairs. Supervised learning can analyze training data and generate an inferred function, which can then be used for mapping inferred data. Supervised learning can be further divided into two types: classification and regression. Classification is used when the output of the AI / ML model has categories, i.e., two or more classes. Regression is used when the output of the AI / ML model has real values ​​or continuous values.

[0133] (11) Unsupervised learning

[0134] In contrast to supervised learning, where an AI / ML model learns to map inputs to target outputs, unsupervised methods learn a concise representation of input data without using labeled data, which can then be used for data exploration or to analyze or generate new data. One typical example of unsupervised learning is clustering, which explores hidden structures in input data and provides classification results for the data.

[0135] (12) Reinforcement Learning

[0136] Reinforcement learning is used to solve sequential decision-making problems. It is a process of training an intelligent agent's actions in an environment based on inputs (states) and feedback signals (rewards). In reinforcement learning, the intelligent agent interacts with the environment by taking actions to maximize cumulative rewards. Whenever the intelligent agent takes an action, the current state in the environment may transition to a new state, and the resulting new state yields an associated reward. The intelligent agent can then take its next action based on the rewards received and the new state in the environment. During the training phase, the agent interacts with the environment to collect experience. Since directly interacting with real systems is expensive, the environment is often simulated by a simulator. In the inference phase, the agent can use the optimal decision-making rules learned from the training phase to achieve the maximum accumulated reward.

[0137] (13) Associative Learning

[0138] Federated learning (FL) is a machine learning technique used to train AI / ML models by a central node (e.g., a server) and multiple decentralized edge nodes (e.g., UEs, next-generation NodeBs, or "gNBs"). According to the wireless FL technique, the server may provide edge nodes with a set of model parameters (e.g., weights, biases, gradients) that describe a global AI / ML model. The edge nodes may initialize local AI / ML models using the received global AI / ML model parameters. The edge nodes may then train local AI / ML models using local data samples, thereby producing trained local AI / ML models. The edge nodes may then provide the server with a set of AI / ML model parameters that describe their local AI / ML models. Upon receiving multiple sets of AI / ML model parameters from multiple edge nodes, each describing a local AI / ML model at those multiple edge nodes, the server may aggregate the local AI / ML model parameters reported from the multiple UEs and update the global AI / ML model based on such aggregation. Subsequent iterations proceed very similarly to the first iteration. The server can send the aggregated global model to multiple edge nodes. The above procedure is repeated multiple times until the global AI / ML model is considered finalized, for example, until the AI / ML model converges or the training stop condition is met. Notably, the wireless FL technique does not involve the exchange of local data samples. In fact, local data samples remain at each edge node.

[0139] AI-based algorithms are being introduced into modern wireless communications to solve several wireless problems, such as channel estimation, scheduling, channel state information (CSI) compression (from user equipment to base stations), multiple-input multiple-output (MIMO) beamforming, and positioning. AI algorithms are data-driven methods that tune several predefined architectures using a set of data samples called a training dataset. In recent years, AI has trained DNN architectures (including CNNs, RNNs, transformers, etc.) by setting up neurons using the SGD algorithm.

[0140] AI techniques in communications (including ML techniques) include AI-based communications in the physical layer and / or in the MAC layer. For the physical layer, AI communications may aim to optimize component design and / or improve algorithmic performance. For the MAC layer, AI / ML-based communications may aim to utilize AI / ML capabilities for learning, prediction, and / or decision-making to optimize MAC layer functionality, such as intelligent TRP management, intelligent beam management, intelligent channel resource allocation, intelligent power control, intelligent spectrum utilization, intelligent modulation and coding scheme (MCS), intelligent hybrid automatic retransmission request (HARQ) strategies, and intelligent transmit / receive (Tx / Rx) mode adaptation, in order to solve complex optimization problems using the best possible strategies and / or optimal solutions.

[0141] AI architectures may involve multiple nodes, which may be organized in one of two modes: centralized or distributed, and both may be deployed in an access network, core network, edge computing system, or third-party network. Centralized training and computing architectures are sometimes constrained by high communication overhead and strict user data privacy. Distributed training and computing architectures may include several frameworks, e.g., distributed machine learning and federative learning. In some embodiments, the AI ​​architecture may include an intelligent controller that can run as a single agent or multiple agents based on joint optimization or individual optimization. Novel protocols and signaling mechanisms are desired to minimize signaling overhead and maximize overall system spectrum efficiency through personalized AI techniques, while allowing corresponding interface links to be personalized with customized parameters to meet specific requirements.

[0142] New protocols and signaling mechanisms are provided for operation within different operating modes, including AI and non-AI modes, and for switching between them, and for measurement and feedback to accommodate different possible measurements and information that may need to be fed back depending on the implementation.

[0143] Currently, neural network models are becoming increasingly large and deep, which can easily require more computing resources than just one or two computers. Most neural network models will be trained on powerful computing clouds. Users with a desired neural network architecture, raw training dataset, and training goals may not have sufficient local computing resources to train their models locally. To access a powerful computing cloud, users must fully transmit all specifications of their neural network architecture, training dataset, and training goals to the network cloud. Users are required to trust the cloud and grant it full authority to manipulate their intellectual property (neural network architecture, training dataset, and training goals).

[0144] As a data-driven approach, AI-based algorithms inevitably suffer from low generalizability: if test data samples are outliers to the training dataset, the neural network will not perform good inferences on the test data samples. Even if an AI model is trained on a large number of datasets, it may not possess the knowledge necessary to perform effectively in other environments, particularly in wireless communications where channel information changes rapidly.

[0145] In this application, AI models are exemplified by DNNs, i.e., deep neural networks or networks. Specific AI models should not be interpreted as limitations of this application.

[0146] Figure 1 is a schematic diagram of a communication system according to one embodiment of the present invention.

[0147] Referring to Figure 1, a simplified schematic illustration of a communication system is provided as an illustrative example without limitation. The communication system 100 includes a radio access network 120. The radio access network 120 may be a next-generation (e.g., 6G or later) radio access network or a legacy (e.g., 5G, 4G, 3G, or 2G) radio access network. One or more communication electrical devices (EDs) 110a-120j (collectively referred to as 110) may be interconnected with each other or connected to one or more network nodes (170a, 170b; collectively referred to as 170) in the radio access network 120. The core network 130 may be part of the communication system and may depend on or be independent of the radio access technology used in the communication system 100. The communication system 100 also includes a public switched telephone network (PSTN) 140, the Internet 150, and other networks 160.

[0148] Figure 2 is a schematic diagram of a communication system 100 according to one embodiment of the present invention.

[0149] Figure 2 shows an example of a communication system 100. Generally, the communication system 100 enables multiple wireless or wired elements to communicate data and other content. The purpose of the communication system 100 may be to provide content such as voice, data, video, and / or text via broadcast, multicast, and unicast, etc. The communication system 100 may operate by sharing resources such as carrier spectrum bandwidth among its components. The communication system 100 may include a terrestrial communication system and / or a non-terrestrial communication system. The communication system 100 may provide a wide range of communication services and applications (such as earth monitoring, remote sensing, passive sensing and positioning, navigation and tracking, autonomous delivery and mobility, etc.). The communication system 100 may provide high availability and robustness through the joint operation of the terrestrial and non-terrestrial communication systems. For example, integrating a non-terrestrial communication system (or its components) into a terrestrial communication system may result in what can be considered a heterogeneous network containing multiple layers. Compared to conventional communication networks, heterogeneous networks can achieve better overall performance through efficient multilink collaboration, more flexible functionality sharing, and faster physical layer link switching between terrestrial and non-terrestrial networks.

[0150] Terrestrial and non-terrestrial communication systems can be considered subsystems of a communication system. In the example shown, communication system 100 includes electronic devices (EDs) 110a-110d (collectively referred to as ED110), radio access networks (RANs) 120a-120b, non-terrestrial communication networks 120c, core network 130, public switched telephone network (PSTN) 140, the Internet 150, and other networks 160. RANs 120a-120b include their respective base stations (BSs) 170a-170b, which may be collectively referred to as terrestrial transceiver points (T-TRPs) 170a-170b. Non-terrestrial communication networks 120c include access nodes 120c, which may be collectively referred to as non-terrestrial transceiver points (NT-TRPs) 172.

[0151] Any ED110 may be configured to interface with, access, or communicate with any other T-TRP 170a-170b and NT-TRP 172, the Internet 150, the core network 130, the PSTN 140, other networks 160, or any combination thereof. In some examples, ED110a may communicate with T-TRP 170a for uplink and / or downlink transmissions via interface 190a. In some examples, ED110a, 110b, and 110d may communicate directly with each other via one or more sidelink air interfaces 190b. In some examples, ED110d may communicate with NT-TRP 172 for uplink and / or downlink transmissions via interface 190c.

[0152] Air interfaces 190a and 190b may use any suitable radio access technology or similar communication technology. For example, communication system 100 may implement one or more channel access methods in air interfaces 190a and 190b, such as code division multiple access (CDMA), time division multiple access (TDMA), frequency division multiple access (FDMA), orthogonal FDMA (OFDMA), or single-carrier FDMA (SC-FDMA). Air interfaces 190a and 190b may utilize other higher-dimensional signal spaces, which may involve combinations of orthogonal and / or non-orthogonal dimensions. Air interface 190c can enable communication between ED 110d and one or more NT-TRP 172 over a wireless link or simply a link. In some examples, the link is a dedicated connection for unicast transmission, a connection for broadcast transmission, or a connection between a group of EDs and one or more NT-TRPs for multicast transmission.

[0153] RAN120a and 120b communicate with the core network 130 to provide ED110a, 110b, and 110c with various services, including voice, data, and other services. RAN120a and 120b and / or the core network 130 may communicate directly or indirectly with one or more other RANs (not shown), which may or may not be directly served by the core network 130, and may or may not employ the same radio access technology as RAN120a, RAN120b, or both. The core network 130 may also serve as gateway access between (i) RAN120a and 120b or ED110a, 110b, and 110c, or both, and (ii) other networks (such as PSTN140, the Internet 150, and other networks 160). In addition, some or all of ED110a, 110b, and 110c may include functionality for communicating with different wireless networks over different wireless links using different wireless technologies and / or protocols. Instead of (or in addition to) wireless communication, ED110a, 110b, and 110c may communicate with service providers or switches (not shown) and the Internet 150 via wired communication channels. PSTN 140 may include a circuit-switched telephone network for providing basic telephone services (POTS). The Internet 150 includes a network of computers and / or subnets (intranets) and may incorporate protocols such as Internet Protocol (IP), Transmit Control Protocol (TCP), and User Datagram Protocol (UDP). ED110a, 110b, and 110c are multimode devices capable of operating according to multiple wireless access technologies and may incorporate multiple transceivers necessary to support such.

[0154] Figure 3 is a schematic diagram of the ED110 and base stations 170a, 170b and / or 170c according to one embodiment of the present application.

[0155] Figure 3 shows another example of the ED110 and base stations 170a, 170b, and / or 170c. The ED110 is used to connect people, objects, machines, etc. The ED110 may be widely used in various scenarios, such as cellular communication, device-to-device (D2D), vehicle-to-everything (V2X), peer-to-peer (P2P), machine-to-machine (M2M), machine-type communication (MTC), Internet of Things (IoT), virtual reality (VR), augmented reality (AR), industrial control, autonomous driving, remote medical care, smart grids, smart furniture, smart offices, smart wearables, smart transportation, smart cities, drones, robots, remote sensing, passive sensing, positioning, navigation and tracking, autonomous delivery and mobility, etc.

[0156] Each ED110 represents any suitable end-user device for wireless operation, and may include (or may be referred to as) a user equipment / device (UE), wireless transmit / receive unit (WTRU), mobile station, fixed or mobile subscriber unit, cellular telephone, station (STA), machine-type communications (MTC) device, personal digital assistant (PDA®), smartphone, laptop, computer, tablet, wireless sensor, consumer electronic device, smartbook, vehicle, automobile, truck, bus, train, or IoT device, industrial device, or equipment in the aforementioned devices (e.g., communication module, modem, or chip). Future generations of ED110 may be referred to using other terms. Base stations 170a and 170b are T-TRPs and will hereafter be referred to as T-TRP170. Similarly, as shown in Figure 3, an NT-TRP will hereafter be referred to as NT-TRP172. Each ED110 connected to the T-TRP170 and / or NT-TRP172 may be dynamically or quasi-statically switched on (i.e., established, activated, or enabled), switched off (i.e., released, deactivated, or disabled), and / or configured, depending on one or more of the availability and / or need for the connection.

[0157] ED110 includes a transmitter 201 and a receiver 203 coupled to one or more antennas 204. Only one antenna 204 is shown. One, some, or all of the antennas may alternatively be panels. The transmitter 201 and receiver 203 may be integrated, for example, as a transceiver. The transceiver is configured to modulate data or other content for transmission by at least one antenna 204 or a network interface controller (NIC). The transceiver is also configured to demodulate data or other content received by at least one antenna 204. Each transceiver includes any suitable structure for generating signals for wireless or wired transmission and / or for processing signals received wirelessly or by wire. Each antenna 204 includes any suitable structure for transmitting and / or receiving wireless or wired signals.

[0158] The ED110 includes at least one memory 208. The memory 208 stores instructions and data used, generated, or collected by the ED110. For example, the memory 208 can store software instructions or modules that are configured to implement some or all of the functionalities and / or embodiments described herein and executed by the processing unit 210. Each memory 208 includes any suitable volatile and / or non-volatile storage and retrieval device. Any suitable type of memory may be used, such as random access memory (RAM), read-only memory (ROM), hard disk, optical disk, subscriber identification module (SIM) card, memory stick, secure digital (SD) memory card, on-processor cache, and the like.

[0159] The ED110 may further include one or more input / output devices (not shown) or interfaces (such as a wired interface to the Internet 150 in Figure 1). The input / output devices enable interaction with users or other devices on the network. Each input / output device includes any suitable structure for providing information to or receiving information from a user, such as a speaker, microphone, keypad, keyboard, display, or touchscreen, including network interface communication.

[0160] ED110 further includes a processor 210 for performing operations related to preparing transmissions for uplink transmissions to NT-TRP172 and / or T-TRP170, processing downlink transmissions received from NT-TRP172 and / or T-TRP170, and processing sidelink transmissions to another ED110. Processing operations related to preparing transmissions for uplink transmissions may include operations such as coding, modulation, transmit beamforming, and generation of symbols for transmission. Processing operations related to processing downlink transmissions may include operations such as receive beamforming, demodulation, and decoding of received symbols. Depending on the embodiment, downlink transmissions may be received by a receiver 203, optionally using receive beamforming, and the processor 210 may extract signaling from the downlink transmissions (e.g., by detecting and / or decoding the signaling). An example of signaling may be a reference signal transmitted by NT-TRP172 and / or T-TRP170. In some embodiments, the processor 276 implements transmit beamforming and / or receive beamforming based on beam direction indications received from the T-TRP 170, such as beam angle information (BAI). In some embodiments, the processor 210 may perform operations related to network access (e.g., initial access) and / or downlink synchronization, such as detecting the synchronization sequence, decoding and retrieving system information, etc. In some embodiments, the processor 210 may perform channel estimation by using reference signals received from, for example, the NT-TRP 172 and / or the T-TRP 170.

[0161] Although not shown, the processor 210 may form part of the transmitter 201 and / or receiver 203. Although not shown, the memory 208 may form part of the processor 210.

[0162] The processor 210 and the processing components of the transmitter 201 and receiver 203 may each be implemented by one or more identical or different processors configured to execute instructions stored in memory (for example, in memory 208). Alternatively, some or all of the processors 210 and the processing components of the transmitter 201 and receiver 203 may be implemented using dedicated circuits such as programmed field-programmable gate arrays (FPGAs), graphical processing units (GPUs), or application-specific integrated circuits (ASICs).

[0163] In some implementations, T-TRP170 may be known by other names, among other possibilities, including base station, base transceiver station (BTS), radio base station, network node, network device, network-side device, transmit / receive node, node B, advanced node B (eNodeB or eNB), home eNodeB, next-generation node B (gNB), transmit point (TP), site controller, access point (AP), or wireless router, relay station, remote radio head, ground node, ground network device, or ground base station, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), positioning node, etc. T-TRP170 may be a macro BS, pico BS, relay node, donor node, or similar, or a combination thereof. T-TRP170 may refer to the aforementioned device or equipment in the aforementioned device (e.g., communication module, modem, or chip).

[0164] In some embodiments, some of the T-TRP170 may be distributed. For example, some of the modules of the T-TRP170 may be located remotely from the equipment housing the T-TRP170 antenna and may be coupled to the equipment housing the antenna via a communication link (not shown) sometimes known as a fronthaul, such as a Common Public Radio Interface (CPRI). Thus, in some embodiments, the term T-TRP170 may also refer to network-side modules that perform processing operations such as determining the location of the ED110, resource allocation (scheduling), message generation, and coding / decoding, and are not necessarily part of the equipment housing the T-TRP170 antenna. These modules may also be coupled to other T-TRPs. In some embodiments, the T-TRP170 may actually be multiple T-TRPs working together to serve the ED110, for example, through coordinated multipoint transmission.

[0165] The T-TRP170 includes at least one transmitter 252 and at least one receiver 254 coupled to one or more antennas 256. Only one antenna 256 is shown. One, some, or all of the antennas may alternatively be panels. The transmitter 252 and receiver 254 may be integrated as a transceiver. The T-TRP170 further includes a processor 260 for performing operations including operations related to: preparing a transmission for downlink transmission to ED110; processing an uplink transmission received from ED110; preparing a transmission for backhaul transmission to NT-TRP172; and processing a transmission received on backhaul from NT-TRP172. Processing operations related to preparing a transmission for downlink or backhaul transmission may include operations such as coding, modulation, precoding (e.g., MIMO precoding), transmit beamforming, and symbol generation for transmission. Processing operations related to processing transmissions received on the uplink or backhaul may include operations such as receive beamforming and demodulation and decoding of received symbols. The processor 260 may also perform operations related to network access (e.g., initial access) and / or downlink synchronization, such as generating synchronous signal block (SSB) content and generating system information. In some embodiments, the processor 260 also generates beam direction indications, such as BAI, which may be scheduled for transmission by the scheduler 253. The processor 260 performs other network-side processing operations described herein, such as determining the location of the ED110 and determining where to deploy the NT-TRP172. In some embodiments, the processor 260 may generate signaling to constitute, for example, one or more parameters of the ED110 and / or one or more parameters of the NT-TRP172. Any signaling generated by the processor 260 is sent by the transmitter 252. It should be noted that, as used herein, "signaling" may be alternatively referred to as control signaling.Dynamic signaling may be transmitted on a control channel, such as a physical downlink control channel (PDCCH), while static or semi-static upper-layer signaling may be included in packets transmitted on a data channel, such as a physical downlink shared channel (PDSCH).

[0166] The scheduler 253 may be coupled to the processor 260. The scheduler 253 may be contained within or operate separately from the T-TRP 170, and may schedule uplink, downlink, and / or backhaul transmissions, including issuing scheduling permissions and / or configuring scheduling-free ("configured permissions") resources. The T-TRP 170 further includes memory 258 for storing information and data. Memory 258 stores instructions and data used, generated, or collected by the T-TRP 170. For example, memory 258 can store software instructions or modules that are configured to implement some or all of the functionalities and / or embodiments described herein and are executed by the processor 260.

[0167] Although not shown, the processor 260 may form part of the transmitter 252 and / or receiver 254. Also, although not shown, the processor 260 may implement a scheduler 253. Although not shown, memory 258 may form part of the processor 260.

[0168] The processing components of processor 260, scheduler 253, and transmitter 252 and receiver 254 may each be implemented by one or more identical or different processors configured to execute instructions stored in memory, for example, in memory 258. Alternatively, some or all of the processing components of processor 260, scheduler 253, and transmitter 252 and receiver 254 may be implemented using dedicated circuitry such as FPGAs, GPUs, or ASICs.

[0169] Although NT-TRP172 is shown merely as a drone as an example, NT-TRP172 may be implemented in any suitable non-terrestrial form. Furthermore, in some implementations, NT-TRP172 may be known by other names such as a non-terrestrial node, a non-terrestrial network device, or a non-terrestrial base station. NT-TRP172 includes a transmitter 272 and a receiver 274 coupled to one or more antennas 280. Only one antenna 280 is shown. One, some, or all of the antennas may alternatively be panels. The transmitter 272 and receiver 274 may be integrated as a transceiver. NT-TRP172 further includes a processor 276 for performing operations including operations related to: preparing a transmission for a downlink transmission to ED110; processing an uplink transmission received from ED110; preparing a transmission for a backhaul transmission to T-TRP170; and processing a transmission received on backhaul from T-TRP170. Processing operations related to preparing a transmission for downlink or backhaul transmission may include operations such as coding, modulation, precoding (e.g., MIMO precoding), transmit beamforming, and generation of symbols for transmission. Processing operations related to processing a transmission received on uplink or backhaul may include operations such as receive beamforming and demodulation and decoding of the received symbols. In some embodiments, the processor 276 implements transmit beamforming and / or receive beamforming based on beam direction information (e.g., BAI) received from the T-TRP 170. In some embodiments, the processor 276 may generate signaling for configuring one or more parameters of the ED 110, for example. In some embodiments, the NT-TRP 172 implements physical layer processing but does not implement higher layer functions such as functions in the medium access control (MAC) layer or radio link control (RLC) layer. This is merely an example, and more generally, the NT-TRP 172 may implement higher layer functions in addition to physical layer processing.

[0170] The NT-TRP172 further includes memory 278 for storing information and data. Although not shown, a processor 276 may form part of the transmitter 272 and / or receiver 274. Although not shown, memory 278 may form part of the processor 276.

[0171] The processing components of processor 276, transmitter 272, and receiver 274 may each be implemented by the same or different one or more processors configured to execute instructions stored in memory, for example, in memory 278. Alternatively, some or all of the processing components of processor 276, transmitter 272, and receiver 274 may be implemented using dedicated circuitry such as programmed FPGAs, GPUs, or ASICs. In some embodiments, NT-TRP 172 may actually be multiple NT-TRPs working together to serve ED110, for example, through coordinated multipoint transmission.

[0172] T-TRP170, NT-TRP172, and / or ED110 may include other components, but these are omitted for clarity.

[0173] Figure 4 is a schematic diagram of a unit or module in a device according to one embodiment of the present application.

[0174] One or more steps of the methods of the provided embodiments may be performed by corresponding units or modules according to Figure 4. Figure 4 shows units or modules in a device, for example, ED110, T-TRP170, or NT-TRP172. For example, a signal may be transmitted by a transmitting unit or transmitting module. A signal may be received by a receiving unit or receiving module. A signal may be processed by a processing unit or processing module. Other steps may be performed by artificial intelligence (AI) or machine learning (ML) modules. Each unit or module may be implemented using hardware, one or more components or devices that run software, or a combination thereof. For example, one or more of the units or modules may be integrated circuits such as programmed FPGAs, GPUs, or ASICs. If modules are implemented using software for execution by a processor, for example, they are recognized by the processor to be searched by the processor in one or more instances, individually or together, as needed, in whole or in part, for processing, and the modules themselves may contain instructions for further deployment and instantiation.

[0175] Further details regarding ED110, T-TRP170, and NT-TRP172 are known to those skilled in the art; therefore, these details are omitted here.

[0176] Figure 5 is a schematic diagram of an AI-based communication device.

[0177] The wireless system includes multiple connected devices. Device 500 is either a base station (BS) or a user equipment (UE). Device 500 may have three systems: a detection system 510, a communication system 520, and / or an AI system 530. The detection system 510 detects and collects signals and data, the communication system 520 transmits and receives signals and data, and the AI ​​system 530 trains and infers an AI implementation. An exemplary AI implementation is based on two cycles of deep learning: a training cycle and an inference cycle. In some possible application scenarios, the training cycle may also be called a learning cycle, and the inference cycle may also be called a logical thinking cycle.

[0178] Deep learning consists of two cycles: training (or learning) and inference (or logical thinking). In the training cycle, the coefficients of neurons are learned from training data to satisfy specific training goals or targets. In the inference or logical thinking cycle, input data samples are fed back into the trained neural network, which will output predictions.

[0179] During the training cycle, the AI ​​system 530 of device 500 may train a DNN or a group of DNNs that the sensing system 510 of device 500 can use to generate signals and / or data. The communication system 520 of device 500 may receive signals or data from another device or a group of other devices. While the AI ​​system 530 is completing training and / or afterward, the device's communication may transmit the training results to another device or a group of other devices.

[0180] During the inference cycle, the AI ​​system 530 of device 500 may perform one or more inferences using one or more DNNs to satisfy one or more tasks, wherein the sensing system 510 of device 500 may generate signals and / or data, and the communication system 520 of device 500 may receive signals or data from another device or other multiple devices. After the AI ​​system 530 of device 500 has finished the inference, the communication system 520 of device 500 may transmit the inference results to another device or other multiple devices.

[0181] The AI ​​implementation may either switch between the two cycles or remain in both cycles simultaneously. For example, the AI ​​system 530 of device 500 may train a second DNN but still perform inference on the first DNN.

[0182] During the training cycle, the AI ​​system 530 of device 500 can function in single-user mode. In this mode, the AI ​​system 530 trains a DNN or multiple DNNs using data provided by the sensing system 510 of device 500. Examples of data include local sensing data and local channel data. Local sensing data includes RGB data, light detection and ranging (LiDAR) data, temperature data, air pressure data, electrical fault data, etc. Local channel data includes channel status information (CSI), received signal strength index (RSSI), latency data, etc.

[0183] Alternatively, the AI ​​system 530 of device 500 may function in cooperative mode. In this mode, the AI ​​system 530 trains a DNN or multiple DNNs using data received by the communication system 520 of device 500. The example data includes detection data, channel data, neuron data, and latent output data. Detection data includes RGB data, LiDAR data, temperature data, air pressure data, electrical fault data, etc. Channel data includes CSI, RSSI, delay data, etc. Neuron data includes the number of neurons or the number of gradients. Latent output data includes several latent outputs.

[0184] Figure 6 is a schematic diagram of a device 500 that receives a reference data sample from a device 600 according to one embodiment of the present invention. In cooperative mode, the AI ​​system 530 of device 500 may use the data by: accumulating detection data received by the communication system 520 of device 500 into a single training dataset; accumulating channel data received by the communication system 520 of device 500 into a single training dataset; setting up local neurons by neurons received by the communication system 520 of device 500, which is a typical associative learning scheme; and inputting latent outputs received by the communication system 520 of device 500 into its DNN.

[0185] Alternatively, in cooperative mode, the AI ​​system 530 of device 500 may: mix local detection data provided by the detection system 510 of device 500 with detection data received by the communication system 520 of device 500 to create a single training dataset; mix local channel data provided by the detection system 510 of device 500 with channel data received by the communication system 520 of device 500 to create a single training dataset; average the local neurons possessed by the AI ​​system 530 of device 500 together with the neurons received by the communication system 520 of device 500, which is a typical associative learning scheme; or average the local latent outputs possessed by the AI ​​system 530 of device 500 and input them into its DNN, thus using the data received by the communication system 520 of device 500 together with its local data.

[0186] Figure 7 is a schematic diagram of a plurality of reference data samples according to one embodiment of the present application. During the training cycle, the communication system 520 of device 500 may receive several reference data samples in both single-user mode and cooperative mode. Some devices transmit reference data samples on broadcast, multicast, or unicast channels. Other devices transmit indices or indices relating to the layer or layers to which the reference data samples relate, where, for example, there are three groups of reference data samples: a first group of reference data samples is shown to relate to an input layer to a DNN; a second group of reference data samples is shown to relate to one latent layer output of the DNN; and a third group of reference data samples is shown to relate to a layer output from the DNN.

[0187] The AI ​​system 530 of device 500 may measure the distance between its local data samples and reference data samples for each group. The AI ​​system 530 of device 500 may sample its local layer inputs, local latent layer outputs, and / or layer outputs randomly, non-randomly, uniformly, or non-uniformly. The AI ​​system 530 of device 500 then measures the distance between the local samples and the reference samples received by the communication system 520 of device 500. If the average distance for all groups consistently falls below a predefined threshold or a set of thresholds, the AI ​​system 530 of device 500 may indicate that the current training procedure is functioning as expected; otherwise, the AI ​​system 530 may indicate that it is abnormal.

[0188] In cases where a device does not have an AI system but does have a detection and communication system, the device's detection system may still be able to measure the distance between its local data samples and reference data samples associated with the layer inputs to the DNN. If the average distance on the layer inputs falls below a predefined threshold, the device's detection system may consider the detection device to have captured "good" data; otherwise, it may consider the data to be bad. The device's communication system may transmit only good data to other devices and not transmit bad data, or the device's communication system may label the detection data by distance before transmitting it to other devices.

[0189] The UE can report information about its own data to the BS, which then determines whether that data differs significantly from the training data. If the difference is excessively large, the BS can switch the operating mode from AI mode to non-AI mode, or to a different AI model. However, direct reporting of raw data by the UE may be considered a violation of user privacy. Transmitting raw data wirelessly is inefficient or violates privacy policies. Therefore, a secure and efficient method for transmitting data status information is an urgent technical problem that needs to be solved.

[0190] To protect raw data and conserve bandwidth, a group of reference data samples is encoded or compressed into a lower-dimensional space than their original space. The encoder or compressor can be linear or nonlinear. A linear encoder can be implemented using some standard basis, such as a Fourier basis, DCT, or wavelets, or it can be implemented using some customized basis. These basis matrices may consist of unitary matrices (orthonormal). A nonlinear encoder can be implemented using several DNNs. Figure 8 shows a schematic representation of a DNN-based approximation according to one embodiment of the present invention.

[0191] Unlike conventional compression schemes built for reliable reconstruction, encoders intentionally avoid reliable reconstruction when data is compressed into a lower-dimensional space, but preserve the largest possible topological distance. That is, the relative distance between two data samples in their original signal space can be well preserved after encoding into a lower-dimensional space.

[0192] Figure 9 is a flowchart of a communication method according to one embodiment of the present invention.

[0193] 710, Send the first coefficient.

[0194] The first coefficient is determined based on the first data and the reference basis, and the dimension of the first coefficient is smaller than the dimension of the first data.

[0195] The first data includes monitoring or measurement data of user equipment or network devices. Furthermore, the first data is monitoring or measurement data related to the AI ​​model. The network device in this embodiment may be a BS. If the first data is data sent from the UE to the BS via an uplink, the data is monitoring or measurement data of the UE. If the first data is data sent from the BS to the UE via a downlink, the data is monitoring or measurement data of the BS.

[0196] One or more reference bases are predefined or constructed. A reference base is one of several predefined or constructed reference bases. For example, a reference base can be constructed by BS for UE. A reference base can be an orthogonal basis, where any two columns of the reference base are perfectly orthogonal to each other. One typical orthogonal basis is a DFT basis.

[0197] 720, communication is performed based on the first coefficient.

[0198] Figure 10 is a flowchart of a communication method according to one embodiment of the present invention. To protect raw data and save bandwidth, a group of reference data samples is encoded or compressed into a lower-dimensional space than their original space. The encoder or compressor can be linear or nonlinear. A linear encoder can be implemented using some standard basis, such as a Fourier basis, discrete cosine transform (DCT), or wavelets. Alternatively, a linear encoder can use some customized basis, which may consist of unitary matrices (orthonormal). A nonlinear encoder can be implemented using several DNNs.

[0199] In the embodiments given below, the UE transforms the high-dimensional signal into a low-dimensional signal (coefficient) by converting the high-dimensional signal into an orthogonal normal basis U.

number

[0200] 810, one or more reference bases are constructed or predefined.

[0201] The Reference Basis Coefficient Index (CRBI) is used to indicate the coefficients relative to a reference basis (e.g., an orthogonal basis). {u1, u2, ..., u r Let} be the orthogonal normal set of vectors in the subspace Rn. This set forms a basis U for the subspace Rn. The elements represented by basis U in the subspace Rn can be described as a finite weighted linear combination of the elements of the basis. The coefficients of this weighted linear combination are the components or coordinates of the vectors relative to basis U.

number

[0202] Figure 11 is a schematic diagram illustrating the projection of a high-dimensional signal onto a low-dimensional signal according to one embodiment of the present application. For example,

number

number

number

number

number

[0203]

number

number

number

number

number

number

number

number

number

number

[0204] In one possible implementation scenario, multiple reference bases (U A ,U B ,U C ,...) are configured or predefined. BS is, for example, U X Construct a reference base that uses U X Report CRBI based on the formula.

number

number

number

[0205] In one possible implementation scenario, one reference matrix U is constructed or predefined, and one or more pruned bases are shown or predefined as reference bases. The reference matrix Y is a matrix of size M rows and N columns. The pruned bases for the reference bases are K columns of Y, for example, the first K columns of Y, where K is constructed and K ≤ N. Optionally, it may be specified which K columns of Y are selected as the pruning basis.

[0206] 820, UE determines its own coefficients in the reference basis.

[0207] The reference basis (U) is constructed or predefined. The BS can consist of one or more reference signals, and the UE is obtained by measuring the reference signals to obtain raw data.

number

number

number

number

[0208] The UE may obtain one or more reported data from a single time slot. Based on the observation interval (or unconstrained) in time, the UE derives the CRBI value to be reported in the uplink slot. Exemplarily, the UE reports a CRBI value in uplink time slot n. The UE may obtain one or more corresponding CRBI values ​​by measuring data in a configured time window n-5 to n-1. The UE may choose to report multiple CRBI values, or it may report the mean / maximum / minimum of multiple CRBI values.

[0209] 830. UE reports the CRBI or an index of the CRBI.

[0210] For example, UE retrieves P report data points from a time window of n-5 to n-1, and P CRBI values ​​corresponding to the P report data points are:

Number

[0211] The UE can directly report the CRBI or report the index corresponding to the CRBI. The BS can configure the Physical Uplink Control Channel (PUCCH) or the Physical Uplink Shared Channel (PUSCH) for the UE to report the CRBI. CRBI reporting supports periodic, aperiodic, and semi - persistent.

[0212] In some possible application scenarios, the UE reports the index corresponding to the CRBI. In this scenario, one or more CRBI tables are predefined or configured. The reference base can be associated with one CRBI table or multiple CRBI tables. When the reference base is associated with multiple CRBI tables, the BS indicates the CRBI table to be used.

[0213] The CRBI index of the CRBI table is reported by the UE. As shown in Table 1, 4 bits are used to indicate the CRBI index. All the CRBI values in Table 1 are represented by the same {c0, c1,..., c r}, and each CRBI index corresponds to a different CRBI value. In some possible implementations, the value of r in {c0, c1,..., c r} is different in different rows of the CRBI table. For example, some are {c0, c1,..., c5}, and some are {c0, c1,..., c6}. Table 1

Table 1

[0214] In some possible implementations, one CRBI index may correspond to a CRBI range, and Table 1 should not be interpreted as a limitation of this application.

[0215] In the communication method provided in this embodiment, the UE can report its data information to the BS with minimal air interface overhead, and the BS then determines whether the data is significantly different from the training data, improving the efficiency of data reporting and protecting data privacy.

[0216] Figure 12 is a flowchart of a communication method according to one embodiment of the present invention. In this embodiment, differential CRBI index reporting can be used.

[0217] 910. Determine the reference CRBI index.

[0218] The reference CRBI index can be indicated by the BS, or it can be configured or predefined.

[0219] 920. Report the offset level to BS.

[0220] The UE reports the offset level to the BS. Based on the offset level and the reference CRBI index, the BS determines the current data CRBI index. Exemplarily, the differential CRBI can be obtained by equation (1).

[0221] Offset level = Current data CRBI index - Reference CRBI index (1)

[0222] In the communication method provided in this embodiment, the UE can report its data information to the BS with minimal air interface overhead, and the BS then determines whether the data is significantly different from the training data, improving the efficiency of data reporting and protecting data privacy.

[0223] In addition, the communication method provided in the present application can also be applied to downlink (DL) transmission in which the BS indicates the CRBI or the CRBI index to the UE in order to indicate data information on the BS side. For specific implementation, reference can be made to the descriptions in FIGS. 9 to 12, which will not be repeated in the present application.

[0224] FIG. 13 is a schematic diagram of the matrix U determined according to an embodiment of the present application.

[0225] Each column of the matrix U can be a standard basis such as a Fourier basis, a DCT basis, a wavelet basis, and the like. Alternatively, the r columns of the matrix U can be constructed on the distribution of the group of reference samples x.

Number

Number

[0226] A sufficient amount (M) of n×1 samples

Number

Number

Number

Number

Number

number

number

[0227] Since each group of reference data samples corresponds to one layer output, each group of reference data samples has its own matrix U. First group

number

number

number

number

[0228] The device's communication system consists of a first matrix U1 and a first group of (compressed) reference samples.

number

number

[0229] Figure 14 is a schematic diagram of the first sampling matrix P1 according to one embodiment of the present application.

[0230] The first matrix U1 is n1×r1, and the second matrix U2 is n2×r2. When n1 and / or n2 is a very large number, the first sampling matrix P1 can be applied to the first matrix U1, and the second sampling matrix P2 can be applied to the second matrix U2. The first sampling matrix P1 is m1×n1 (m1 << n1), and each row of it has only one "1" to indicate the position to be sampled

Number

Number

[0231] FIG. 15 is a schematic diagram of a sampling matrix compression matrix U according to an embodiment of the present application.

[0232] In one possible implementation, the communication system of the device receives the first compact matrix θ1, the first sampling matrix P1, and the first group of (compressed) reference samples

Number

Number

[0233] Alternatively, the device's communication system uses a first compact matrix.

number

number

number

number

[0234] Figure 16 is a schematic diagram of the scoring distance in low-spectrum space according to one embodiment of the present application.

[0235] The device's communication system consists of two samples from the first group.

number

number

number

number

number

number

number

number

number

number

[0236] Alternatively, the device's communication system is based on two distributions of the first group.

number

number

number

number

number

number

number

number

number

number

[0237] Due to generalization issues, the UE or BS needs to detect its generalization performance and then switch to the appropriate AI model. For example, there may be multiple AI models for multiple scenarios, including indoor cities, outdoor cities, rural areas, high-speed rail, etc. If the surrounding environment changes, the BS or UE needs to switch to a different model. In the prior art, if inference performance deteriorates, the UE or BS either switches its own AI model or falls back to a non-AI mode. This is a kind of reactive solution. Embodiments of the present invention provide multiple event trigger designs that enable active model switching.

[0238] The BS configures multiple candidate AI / ML models for the UE, and each model is configured with a model index. The configuration signals can be Radio Resource Control (RRC), Media Access Control Element (MAC-CE), or Downlink Control Information (DCI), and can be broadcast, multicast, or unicast.

[0239] The BS constructs a data anchor associated with each candidate AI / ML model. The anchor is a reference data, for example, a reference coefficient.

number

number

[0240] The association between a data anchor and a candidate AI / ML model can be determined implicitly or explicitly configured. For example, the association between a data anchor and a candidate AI / ML model is implicitly determined by ensuring the anchor index has the same value as the model index, i.e., {model index k, anchor index k}. Alternatively, the association between a data anchor and a candidate AI / ML model can be explicitly configured, where BS configures data anchor j to associate with candidate model k, i.e., {model index k, anchor index j}.

[0241] Reference data (for example, coefficients (

number

number

number

number

number

[0242] The reference basis (U) is constructed or predefined. For example, BS can constitute a reference signal relating to the reference basis. Optionally, this reference signal may be detected by UE. UE is

number

number

[0243] UE uses its own data (for example, coefficients (

number

number

number

number

number

number

number

number

number

[0244]

number

[0245]

number

[0246]

number

[0247] Equations (2) to (4) are merely examples, and UE uses its own data (e.g., coefficients (

number

[0248] An anchor is a set of reference data, and the UE calculates the difference between its own data and the anchor according to a method that may be shown by the BS, or may be predefined, such as equation (4) or (5).

number

number

number

number

number

number

[0249]

number

[0250]

number

[0251] Alternatively, the difference between the reported data and the anchor can also be obtained using mutual information, the Hilbert-Schmidt independence criterion (HSIC) metric, the Kullback-Libra (KL) divergence, graphical edit distance, Wasserstein distance, Jensen-Shannon divergence (JSD) distance, DNN-based algorithms, etc.

[0252] For AI / ML mode monitoring, BS configures the monitoring report as a triggered event. One of several AI model monitoring events can be configured as event K1, event K2, event K3, event K4, etc.

[0253] Figure 17 is a flowchart of a communication method according to one embodiment of the present invention.

[0254] 1510. When the entry or exit conditions for an event are met, a report is sent indicating a switch in the AI ​​model or mode.

[0255] This application defines events as follows: Event K1: the operation of switching from an AI model to a non-AI model; Event K2: the operation of switching from one AI / ML model (Model 1) to another AI / ML model (Model 2); Event K3: the operation of switching from non-AI mode to AI mode; and Event K4: the operation of switching from one AI / ML model (Model 2) to another AI / ML model (Model 1).

[0256] The method provided in this application defines a plurality of events and corresponding entry and exit conditions, thereby enabling an active model or mode switch when the conditions are met.

[0257] 1520, Execute communication based on the report.

[0258] In event K1, the entry or exit condition requires determining whether the difference between the first data and the first anchor is greater than a threshold. The first anchor can be indicated by BS, for example, the first anchor is the data anchor of the UE's current AI / ML model. Alternatively, the first anchor is the data anchor closest to the UE's first data. The first data includes the UE's monitoring or measurement data. In possible implementations, the first data includes any one or more of the following: detection data, measurement data, channel data, neuron data of the AI ​​model, and latent output data of the AI ​​model.

[0259] Event K1 means that the current or best AI / ML model may be outdated, and the UE should be switched to non-AI mode.

[0260] The entry condition for event K1 is considered met when condition K1-1, specified below, is satisfied.

[0261] Diff11-Hys11>Thresh11 (K1-1)

[0262] Diff11 is the difference between the first data point and the first anchor. Hys11 is the hysteresis parameter for event K1, which is either predefined or configured by BS. Thresh11 is the threshold parameter for event K1, which is either predefined or configured by BS. Thresh11 and Hys11 are represented in the same unit as Diff11.

[0263] The exit condition for event K1 is considered met when the condition K1-2 specified below is also met.

[0264] Diff11+Hys12 <Thresh12 (K1-2)

[0265] Hys12 is a hysteresis parameter for event K1, which is either predefined or composed of BS. Thresh12 is a threshold parameter for event K1, which is either predefined or composed of BS. The units of Hys12 and Thresh12 are the same as the units of Diff11. Hys11 and Hys12 may be the same or different, for example, they may be similar. Thresh11 and thresh12 may be the same or different, for example, they may be similar.

[0266] BS configures whether the parameter "reportOnLeave" is enabled or disabled. If disabled, no report is sent when the leave condition is met.

[0267] In addition, BS also configures a parameter called "Time to Trigger," which specifies a range of values ​​that will be used to trigger the parameter, relating to the time at which specific criteria for an event must be met in order to trigger the report.

[0268] Another possible entry or exit condition for K1 is the existence of an additional offset.

[0269] The entry conditions for event K1 are considered met when the conditions K1-3 specified below are satisfied.

[0270] Diff11+offset11-Hys11>Thresh11 (K1-3)

[0271] The exit condition for event K1 is considered met when the conditions K1-4 specified below are met.

[0272] Diff11+offset12+Hys12 <Thresh12 (K1-4)

[0273] offset11 and offset12 may be the same or different. The values ​​of offset11 and offset12 may be positive or negative numbers.

[0274] In some possible implementations, the values ​​of Hys11 or Hys12 can be set to 0, or they can be predefined or not configured at all.

[0275] The method provided by this application defines an event K1 and entry and exit conditions corresponding to the event K1, and enables active mode switching when any of these conditions are met.

[0276] Event K2 means that the current AI / ML model (referred to as Model 1) may be outdated, and the UE should switch to a different model (Model 2).

[0277] In event K2, the current model will perform worse than Thresh21, and another model will perform better than Thresh22. Performing worse than Thresh21 means that the second difference value between the first data and the second anchor is greater than that of Thresh21. Performing better than Thresh22 means that the third difference value between the first data and the third anchor is smaller than that of Thresh22. The second anchor is the data anchor associated with the first AI model (Model 1), and the third anchor is the data anchor associated with the second AI model (Model 2).

[0278] The entry conditions for event K2 are considered met when condition K2-1, specified below, is satisfied.

[0279] Diff21-Hys21>Thresh21 and Diff22+Hys22 <Thresh22 (K2-1)

[0280] Diff21 is the difference between the first data and the second anchor, where the second anchor can be, for example, a data anchor associated with Model 1. Diff22 is the difference between the first data and the third anchor, where the third anchor can be, for example, a data anchor associated with Model 2. Hys21 and Hys22 are predefined or configured parameters corresponding to event K2. Thresh21 is a predefined or configured threshold corresponding to the first AI model, and Thresh22 is a predefined or configured threshold corresponding to the second AI model. Thresh21 and Thresh22 can be configured to be the same or different.

[0281] The exit condition for event K2 is considered met when either condition K2-2 or K2-3, as specified below, is met.

[0282] Diff21+Hys23 <Thresh23 (K2-2)

[0283] Diff22-Hys24>Thresh24 (K2-3)

[0284] Condition K2-2 means that the first AI model is better than threshold 23. Condition K2-3 means that the second AI model is worse than threshold 24. If either condition K2-2 or K2-3 is met, the UE leaves event K2.

[0285] Hys23 and Hys24 are predefined or configured parameters corresponding to event K2. Thresh23 is a predefined or configured threshold corresponding to the first AI model, and Thresh24 is a predefined or configured threshold corresponding to the second AI model. Hys21 and Hys23 may be the same or different, for example, they may be similar. Hys22 and Hys24 may be the same or different, for example, they may be similar. Thresh21 and Thresh23 may be the same or different, for example, they may be similar. Thresh22 and Thresh24 may be the same or different, for example, they may be similar.

[0286] BS configures whether the parameter "reportOnLeave" is enabled or disabled. If disabled, no report is sent when the leave condition is met.

[0287] In addition, BS also configures a parameter called "Time to Trigger," which specifies a range of values ​​that will be used to trigger the parameter, relating to the time at which specific criteria for an event must be met in order to trigger the report.

[0288] Another possible entry or exit condition for K2 is the existence of an additional offset.

[0289] The entry conditions for event K2 are considered met when condition K2-4, specified below, is satisfied.

[0290] Diff21+offset21-Hys21>Thresh21 and Diff22+offset22+Hys21 <Thresh22 (K2-4)

[0291] The exit condition for event K2 is considered met when either condition K2-5 or K2-6, as specified below, is met.

[0292] Diff21+offset23+Hys23 <Thresh23 (K2-5)

[0293] Diff22+offset24-Hys24>Thresh24 (K2-6)

[0294] offset21 and offset23 may be the same or different, for example, they may be similar. offset22 and offset24 may be the same or different, for example, they may be similar. The values ​​of offset21, offset22, offset23, and offset24 may be positive or negative numbers.

[0295] In some possible implementations, the values ​​of Hys21, Hys22, Hys23, or Hys24 can be set to 0, or they can be predefined or not configured at all.

[0296] The method provided herein defines an event K2 and corresponding entry and exit conditions, and enables active switching of the model when any of these conditions are met.

[0297] Event K3 means that the AI / ML model is performing better than the threshold, and the UE should switch from non-AI mode to AI mode.

[0298] In event K3, the fourth difference value between the first data and the fourth anchor is less than the threshold. The fourth anchor can be indicated by BS, or the fourth anchor is the data anchor closest to the first data, i.e., the fourth anchor is the data anchor with the smallest difference from the first data among multiple anchors.

[0299] The entry conditions for event K3 are considered met when condition K3-1, specified below, is satisfied.

[0300] Diff31+Hys31 <Thresh31 (K3-1)

[0301] Diff31 is the difference between the first data point and the fourth anchor. Hys31 is a predefined or configured hysteresis parameter for event K3. Thresh31 is a predefined or configured threshold corresponding to event K3. Thresh31 and Hys31 are represented in the same unit as Diff31.

[0302] When the entry conditions for event K3 are met, the UE can report the index of the nearest anchor.

[0303] The exit condition for event K3 is considered met when condition K3-2, specified below, is satisfied.

[0304] Diff31-Hys32>Thresh32 (K3-2)

[0305] Hys32 is a predefined or configured hysteresis parameter for event K3. Thresh32 is a predefined or configured threshold corresponding to event K3. Hys31 and Hys32 may be the same or different, for example, they may be similar. Thresh31 and Thresh32 may be the same or different, for example, they may be similar.

[0306] BS configures whether the parameter "reportOnLeave" is enabled or disabled. If disabled, no report is sent when the leave condition is met.

[0307] In addition, BS also configures a parameter called "Time to Trigger," which specifies a range of values ​​that will be used to trigger the parameter, relating to the time at which specific criteria for an event must be met in order to trigger the report.

[0308] Another possible entry or exit condition for K3 is the existence of an additional offset.

[0309] The entry conditions for event K3 are considered met when condition K3-3, specified below, is satisfied.

[0310] Diff31+offset31+Hys31 <Thresh31 (K3-3)

[0311] The exit condition for event K3 is considered met when the conditions K3-4 specified below are met.

[0312] Diff31+offset32-Hys32>Thresh32 (K3-4)

[0313] offset31 and offset32 may be the same or different. The values ​​of offset31 and offset32 may be positive or negative numbers.

[0314] In some possible implementations, the values ​​of Hys31 or Hys32 can be set to 0, or they can be predefined or not configured at all.

[0315] The method provided herein defines an event K3 and corresponding entry and exit conditions, and enables active mode switching when any of these conditions are met.

[0316] In Event K4, another model (Model 1) performs better than the current model (Model 2), and the UE should switch to Model 1.

[0317] The entry condition for event K4 is considered met when condition K4-1, specified below, is satisfied.

[0318] Diff42+Hys41 <Diff41 (K4-1)

[0319] Diff41 is the difference between the first data and the fifth anchor, where the fifth anchor can be, for example, a data anchor associated with Model 2. Diff42 is the difference between the first data and the sixth anchor, where the sixth anchor can be, for example, a data anchor associated with Model 1. Hys41 is a predefined or configured parameter corresponding to event K4.

[0320] The exit condition for event K4 is considered met when condition K4-2, specified below, is satisfied.

[0321] Diff42-Hys42>Diff41 (K4-2)

[0322] Hys42 is a predefined or configured parameter corresponding to event K4. Hys41 and Hys42 may be the same or different, for example, they may be similar.

[0323] BS configures whether the parameter "reportOnLeave" is enabled or disabled. If disabled, no report is sent when the leave condition is met.

[0324] In addition, BS also configures a parameter called "Time to Trigger," which specifies a range of values ​​that will be used to trigger the parameter, relating to the time at which specific criteria for an event must be met in order to trigger the report.

[0325] Another possible entry or exit condition for K4 is the existence of an additional offset.

[0326] The entry condition for event K4 is considered met when condition K4-3, specified below, is satisfied.

[0327] Diff42+offset42+Hys41 <Diff41+offset41 (K4-3)

[0328] The exit condition for event K4 is considered met when condition K4-4, specified below, is satisfied.

[0329] Diff42+offset42-Hys42>Diff41+offset41 (K4-4)

[0330] offset41 and offset42 may be the same or different. The values ​​of offset41 and offset42 may be positive or negative numbers.

[0331] In some possible implementations, the values ​​of Hys41 or Hys42 can be set to 0, or they can be predefined or not configured at all.

[0332] The method provided herein defines an event K4 and corresponding entry and exit conditions, and enables active switching of the model when any of these conditions are met.

[0333] The above embodiment is an example in which the UE actively performs model or mode switching, and the method by which the BS actively performs model or mode switching is similar to the above embodiment, and a specific implementation can be found in the description of events K1 to K4 above, which is not repeated in this application.

[0334] Figure 18 is a schematic block diagram of a communication device 1700 according to one embodiment of the present invention. The communication device 1700 includes a sending module 1710 configured to send a report indicating an AI model switch or mode switch when an event entry condition or exit condition is met.

[0335] In a possible implementation, the event is an action that switches from AI mode to non-AI mode, where the entry condition is determined by Diff11 and Thresh11, or the exit condition is determined by Diff11 and Thresh12, where Diff11 is a first difference value, Thresh11 is a first threshold, and Thresh12 is a second threshold, where the first difference value is the difference between a first data and a first anchor, where the first anchor contains one or more pieces of reference data, and the first and second thresholds are predefined or configured thresholds corresponding to the event.

[0336] In possible implementations, the entry condition is Diff11-Hys11>Thresh11, or the exit condition is Diff11+Hys12 <Thresh12である。

[0337] In a possible implementation, the entry condition is Diff11 + offset11 - Hys11 > Thresh11, or the exit condition is Diff11 + offset12 + Hys12 < Thresh12, where offset11 and offset12 are pre-defined or configured offsets.

[0338] In a possible implementation, Hys11 is the first hysteresis parameter, Hys12 is the second hysteresis parameter, and the first hysteresis parameter and the second hysteresis parameter are pre-defined or configured parameters corresponding to an event.

[0339] In a possible implementation, the event is an operation of switching from the first AI model to the second AI model, where the entry condition is determined by Diff21, Diff22, Thresh21, and Thresh22, or the exit condition is determined by Diff21, Diff22, Thresh23, and Thresh24, where Diff21 is the second difference value, Diff22 is the third difference value, Thresh21 is the third threshold value, Thresh22 is the fourth threshold value, Thresh23 is the fifth threshold value, Thresh24 is the sixth threshold value, the second difference value is the difference value between the first data and the second anchor, the second anchor is a data anchor associated with the first AI model, the third difference value is the difference value between the first data and the third anchor, the third anchor is a data anchor associated with the second AI model, the third threshold value and the fifth threshold value are pre-defined or configured threshold values corresponding to the first AI model, and the fourth threshold value and the sixth threshold value are pre-defined or configured threshold values corresponding to the second AI model.

[0340] In a possible implementation, the entry condition is Diff21 - Hys21 > Thresh21 and Diff22 + Hys22 < Thresh22, or the exit condition is Diff21 + Hys23 < Thresh23 or Diff22 - Hys24 > Thresh24.

[0341] In a possible implementation, the entry conditions are Diff21 + offset21 - Hys21 > Thresh21 and Diff22 + offset22 + Hys22 < Thresh22, and the exit conditions are Diff21 + offset23 + Hys23 < Thresh23 or Diff22 + offset24 - Hys24 > Thresh24, where offset21, offset22, offset23, and offset24 are pre - defined or configured offsets.

[0342] In a possible implementation, Hys21 is the third hysteresis parameter, Hys22 is the fourth hysteresis parameter, Hys23 is the fifth hysteresis parameter, Hys24 is the sixth hysteresis parameter, and the third hysteresis parameter, the fourth hysteresis parameter, the fifth hysteresis parameter, and the sixth hysteresis parameter are pre - defined or configured parameters corresponding to events.

[0343] In a possible implementation, the event is an operation of switching from a non - AI mode to an AI mode, where the entry condition is determined by Diff31 and Thresh31, or the exit condition is determined by Diff31 and Thresh32, where Diff31 is the fourth difference value, Thresh31 is the seventh threshold value, Thresh32 is the eighth threshold value, the fourth difference value is the difference value between the first data and the fourth anchor, the fourth anchor includes one or more pieces of reference data, and the seventh threshold value and the eighth threshold value are pre - defined or configured threshold values corresponding to the event.

[0344] In a possible implementation, the entry condition is Diff31 + Hys31 < Thresh31, or the exit condition is Diff31 - Hys32 > Thresh32.

[0345] In a possible implementation, the entry condition is Diff31 + offset31 + Hys31 < Thresh31, and the exit condition is Diff31 + offset32 - Hys32 > Thresh32, where offset31 and offset32 are pre-defined or configured offsets.

[0346] In a possible implementation, Hys31 is the seventh hysteresis parameter, Hys32 is the eighth hysteresis parameter, and the seventh hysteresis parameter and the eighth hysteresis parameter are pre-defined or configured parameters corresponding to an event.

[0347] In a possible implementation, the event is an operation of switching from the third AI model to the fourth AI model, where the entry condition is determined by Diff41 and Diff42, or the exit condition is determined by Diff41 and Diff42, Diff41 is the fifth difference value, Diff42 is the sixth difference value, the fifth difference value is the difference value between the first data and the fifth anchor, the fifth anchor is a data anchor associated with the third AI model, the sixth difference value is the difference value between the first data and the sixth anchor, and the sixth anchor is a data anchor associated with the fourth AI model.

[0348] In a possible implementation, the entry condition is Diff42 + Hys41 < Diff41, and the exit condition is Diff42 - Hys42 > Diff41.

[0349] In a possible implementation, the entry condition is Diff42 + offset42 + Hys41 < Diff41 + offset41, and the exit condition is Diff42 + offset42 - Hys42 > Diff41 + offset41, where offset41 and offset42 are pre-defined or configured offsets.

[0350] In possible implementations, Hys41 is the ninth hysteresis parameter, Hys42 is the tenth hysteresis parameter, and the ninth and tenth hysteresis parameters are predefined or configured parameters corresponding to events.

[0351] In possible implementations, the first data includes monitoring or measurement data from user equipment or network devices.

[0352] In possible implementations, the first data includes monitoring or measurement data related to the AI ​​model of the user's equipment or network device.

[0353] In a possible implementation, the first data includes any one or more of the following: detection data, measurement data, channel data, neuron data of the AI ​​model, and latent output data of the AI ​​model.

[0354] In possible implementations, the device further includes a processing module 1720 configured to perform communication based on the report.

[0355] In possible implementations, the device is located on user equipment or network devices.

[0356] As shown in Figure 19, the communication device 2200 may include a processor 2210 and a transceiver 2220. Optionally, the communication device 2200 may further include a memory 2230. The memory 2230 may be configured to store indication information, or it may be configured to store code, instructions, and the like that will be executed by the processor 2210.

[0357] Memory 2230 may include random memory, flash memory, read-only memory, programmable read-only memory, non-volatile memory, registers, or similar. Processor 2210 may be a central processing unit (CPU).

[0358] For other functions and operations of the communication device 2200, please refer to the processes of the embodiments of the method shown in Figures 5 to 16, which are not described again herein to avoid repetition.

[0359] One embodiment of the present invention further provides a computer storage medium which may store program instructions for performing the steps in the above-described method.

[0360] The storage medium may, at the user's discretion, be memory 2230.

[0361] One embodiment of the present invention further provides a computer program product, which includes computer program code. When the computer program code is executed on a computer, the computer is enabled to perform the steps in the method described above.

[0362] Optionally, all or part of the computer program code can be stored on a first storage medium. The first storage medium can be packaged together with the processor or separately from the processor.

[0363] One embodiment of the present invention further provides a chip system comprising an input / output interface, at least one processor, at least one memory, and a bus. The at least one memory is configured to store instructions, and the at least one processor is configured to invoke instructions in the at least one memory to perform the operations in the method of the above-described embodiment.

[0364] Those skilled in the art will understand that all or part of the processes of the method in the embodiment may be implemented by a computer program that instructs the associated hardware. The program may be stored in a computer-readable storage medium. When the program is executed, the processes of the method in the embodiment are executed. The aforementioned storage medium may include: a magnetic disk, an optical disk, read-only memory (ROM), or random access memory (RAM).

[0365] It should be understood that in some embodiments provided herein, the disclosed systems, devices, and methods may be implemented in other ways. For example, the embodiments of the described devices are illustrative only. For example, the division of units is merely a division of logical functions, and other divisions may be in actual implementations. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not performed. In addition, mutual coupling, direct coupling, or communication connection shown or discussed may be implemented by using some interfaces. Indirect coupling or communication connection between devices or units may be implemented electronically, mechanically, or in other forms.

[0366] Units described as separate parts may or may not be physically separate, and parts shown as units may or may not be physical units, may be located in one location, or may be distributed across multiple network units. Some or all of the units may be selected according to the actual need to achieve the objectives of the solution of the embodiment.

[0367] In addition, the functional units in the embodiments of the present invention may be integrated into a single processing unit, or each of these units may exist physically independently, or two or more units may be integrated into a single unit.

[0368] The foregoing describes only exemplary embodiments of the present invention. Those skilled in the art may make various modifications and variations to the present invention without departing from its scope.

Claims

1. It is a method of communication: The stage in which a report is sent indicating a switch in the artificial intelligence (AI) model or mode when the entry or exit conditions for an event are met. A method that includes [a certain feature].

2. The method according to claim 1, wherein the event is an operation to switch from AI mode to non-AI mode, the entry condition is determined by Diff 11 and Thresh 11, or the exit condition is determined by Diff 11 and Thresh 12, where Diff 11 is a first difference value, Thresh 11 is a first threshold, Thresh 12 is a second threshold, the first difference value is the difference value between first data and a first anchor, the first anchor includes one or more pieces of reference data, and the first threshold and the second threshold are predefined or configured thresholds corresponding to the event.

3. The method according to claim 2, wherein the entry condition is Diff11 - Hys11 > Thresh11, or the exit condition is Diff11 + Hys12 < Thresh12.

4. The method according to claim 2, wherein the entry condition is Diff11 + offset11 - Hys11 > Thresh11, or the exit condition is Diff11 + offset12 + Hys12 < Thresh12, where offset11 and offset12 are predefined or configured offsets.

5. The method according to claim 3 or 4, wherein Hys11 is a first hysteresis parameter, Hys12 is a second hysteresis parameter, and the first and second hysteresis parameters are predefined or configured parameters corresponding to the event.

6. The event is an operation to switch from the first AI model to the second AI model, the entry conditions are determined by Diff21, Diff22, Thresh21 and Thresh22, or the exit conditions are determined by Diff21, Diff22, Thresh23 and Thresh24, where Diff21 is the second difference value, Diff22 is the third difference value, Thresh21 is the third threshold, Thresh22 is the fourth threshold, Thresh23 is the fifth threshold, Thresh24 is the sixth threshold, and The method according to claim 1, wherein the second difference value is the difference value between the first data and the second anchor, the second anchor being a data anchor associated with the first AI model, the third difference value is the difference value between the first data and the third anchor, the third anchor being a data anchor associated with the second AI model, the third threshold and the fifth threshold are predefined or configured thresholds corresponding to the first AI model, and the fourth threshold and the sixth threshold are predefined or configured thresholds corresponding to the second AI model.

7. The method according to claim 6, wherein the entry conditions are Diff21-Hys21>Thresh21 and Diff22+Hys22<Thresh22, or the exit conditions are Diff21+Hys23<Thresh23 or Diff22-Hys24>Thresh24.

8. The method according to claim 6, wherein the entry conditions are Diff21 + offset21 - Hys21 > Thresh21 and Diff22 + offset22 + Hys22 < Thresh22, the exit conditions are Diff21 + offset23 + Hys23 < Thresh23 or Diff22 + offset24 - Hys24 > Thresh24, and offset21, offset22, offset23 and offset24 are predefined or configured offsets.

9. The method according to claim 7 or 8, wherein Hys21 is a third hysteresis parameter, Hys22 is a fourth hysteresis parameter, Hys23 is a fifth hysteresis parameter, Hys24 is a sixth hysteresis parameter, and the third, fourth, fifth, and sixth hysteresis parameters are predefined or configured parameters corresponding to the event.

10. The method according to claim 1, wherein the event is an operation to switch from a non-AI mode to an AI mode, the entry condition is determined by Diff 31 and Thresh 31, or the exit condition is determined by Diff 31 and Thresh 32, where Diff 31 is a fourth difference value, Thresh 31 is a seventh threshold, Thresh 32 is an eighth threshold, the fourth difference value is the difference value between the first data and the fourth anchor, the fourth anchor includes one or more pieces of reference data, and the seventh and eighth thresholds are predefined or configured thresholds corresponding to the event.

11. The method according to claim 10, wherein the entry condition is Diff31 + Hys31 < Thresh31, or the exit condition is Diff31 - Hys32 > Thresh32.

12. The method according to claim 10, wherein the entry condition is Diff31 + offset31 + Hys31 < Thresh31, and the exit condition is Diff31 + offset32 - Hys32 > Thresh32, where offset31 and offset32 are predefined or configured offsets.

13. The method according to claim 11 or 12, wherein Hys31 is a seventh hysteresis parameter, Hys32 is an eighth hysteresis parameter, and the seventh and eighth hysteresis parameters are predefined or configured parameters corresponding to the event.

14. The method according to claim 1, wherein the event is an operation to switch from a third AI model to a fourth AI model, the entry condition is determined by Diff41 and Diff42, or the exit condition is determined by Diff41 and Diff42, where Diff41 is a fifth difference value, Diff42 is a sixth difference value, the fifth difference value is the difference value between a first data and a fifth anchor, the fifth anchor is a data anchor associated with the third AI model, the sixth difference value is the difference value between the first data and a sixth anchor, and the sixth anchor is a data anchor associated with the fourth AI model.

15. The method according to claim 14, wherein the entry condition is Diff42 + Hys41 < Diff41, and the exit condition is Diff42 - Hys42 > Diff41.

16. The method according to claim 14, wherein the entry condition is Diff42 + offset42 + Hys41 < Diff41 + offset41, the exit condition is Diff42 + offset42 - Hys42 > Diff41 + offset41, and offset41 and offset42 are predefined or configured offsets.

17. The method according to claim 15 or 16, wherein Hys41 is a ninth hysteresis parameter, Hys42 is a tenth hysteresis parameter, and the ninth and tenth hysteresis parameters are predefined or configured parameters corresponding to the event.

18. The method according to any one of claims 2 to 17, wherein the first data includes monitoring data or measurement data of user equipment or network devices.

19. The method according to claim 18, wherein the first data includes monitoring data or measurement data related to the AI ​​model of the user equipment or the network device.

20. The method according to any one of claims 2 to 19, wherein the first data includes any one or more of detection data, measurement data, channel data, neuron data of the AI ​​model, and latent output data of the AI ​​model.

21. The step of executing communication based on the aforementioned report. The method according to any one of claims 1 to 20, further comprising:

22. The method according to any one of claims 1 to 21, which is performed by a user device or a network device.

23. A sending module configured to send a report indicating a switch or mode change of the artificial intelligence (AI) model when the entry or exit conditions for an event are met. A communication device equipped with the following features.

24. The communication device according to claim 23, wherein the event is an operation to switch from AI mode to non-AI mode, the entry condition is determined by Diff 11 and Thresh 11, or the exit condition is determined by Diff 11 and Thresh 12, where Diff 11 is a first difference value, Thresh 11 is a first threshold, Thresh 12 is a second threshold, the first difference value is the difference value between first data and a first anchor, the first anchor includes one or more pieces of reference data, and the first threshold and the second threshold are predefined or configured thresholds corresponding to the event.

25. The communication device according to claim 24, wherein the entry condition is Diff11 - Hys11 > Thresh11, or the exit condition is Diff11 + Hys12 < Thresh12.

26. The communication device according to claim 24, wherein the entry condition is Diff11 + offset11 - Hys11 > Thresh11, or the exit condition is Diff11 + offset12 + Hys12 < Thresh12, where offset11 and offset12 are predefined or configured offsets.

27. The communication device according to claim 25 or 26, wherein Hys11 is a first hysteresis parameter, Hys12 is a second hysteresis parameter, and the first and second hysteresis parameters are predefined or configured parameters corresponding to the event.

28. The event is an operation to switch from the first AI model to the second AI model, the entry condition is determined by Diff21, Diff22, Thresh21 and Thresh22, or the exit condition is determined by Diff21, Diff22, Thresh23 and Thresh24, where Diff21 is the second difference value, Diff22 is the third difference value, Thresh21 is the third threshold, Thresh22 is the fourth threshold, Thresh23 is the fifth threshold, Thresh24 is the sixth threshold, and the second The communication device according to claim 23, wherein the difference value is the difference value between the first data and the second anchor, the second anchor is a data anchor associated with the first AI model, the third difference value is the difference value between the first data and the third anchor, the third anchor is a data anchor associated with the second AI model, the third threshold and the fifth threshold are predefined or configured thresholds corresponding to the first AI model, and the fourth threshold and the sixth threshold are predefined or configured thresholds corresponding to the second AI model.

29. The communication device according to claim 28, wherein the entry conditions are Diff21-Hys21>Thresh21 and Diff22+Hys22<Thresh22, or the exit conditions are Diff21+Hys23<Thresh23 or Diff22-Hys24>Thresh24.

30. The communication device according to claim 28, wherein the entry conditions are Diff21 + offset21 - Hys21 > Thresh21 and Diff22 + offset22 + Hys22 < Thresh22, the exit conditions are Diff21 + offset23 + Hys23 < Thresh23 or Diff22 + offset24 - Hys24 > Thresh24, and offset21, offset22, offset23 and offset24 are predefined or configured offsets.

31. The communication device according to claim 29 or 30, wherein Hys21 is a third hysteresis parameter, Hys22 is a fourth hysteresis parameter, Hys23 is a fifth hysteresis parameter, and Hys24 is a sixth hysteresis parameter, and the third, fourth, fifth, and sixth hysteresis parameters are predefined or configured parameters corresponding to the event.

32. The communication device according to claim 23, wherein the event is an operation to switch from a non-AI mode to an AI mode, the entry condition is determined by Diff 31 and Thresh 31, or the exit condition is determined by Diff 31 and Thresh 32, where Diff 31 is a fourth difference value, Thresh 31 is a seventh threshold, Thresh 32 is an eighth threshold, the fourth difference value is the difference value between the first data and the fourth anchor, the fourth anchor includes one or more pieces of reference data, and the seventh and eighth thresholds are predefined or configured thresholds corresponding to the event.

33. The communication device according to claim 32, wherein the entry condition is Diff31 + Hys31 < Thresh31, or the exit condition is Diff31 - Hys32 > Thresh32.

34. The communication device according to claim 32, wherein the entry condition is Diff31 + offset31 + Hys31 < Thresh31, the exit condition is Diff31 + offset32 - Hys32 > Thresh32, and offset31 and offset32 are predefined or configured offsets.

35. The communication device according to claim 33 or 34, wherein Hys31 is a seventh hysteresis parameter, Hys32 is an eighth hysteresis parameter, and the seventh and eighth hysteresis parameters are predefined or configured parameters corresponding to the event.

36. The communication device according to claim 23, wherein the event is an operation to switch from a third AI model to a fourth AI model, the entry condition is determined by Diff41 and Diff42, or the exit condition is determined by Diff41 and Diff42, where Diff41 is a fifth difference value and Diff42 is a sixth difference value, the fifth difference value is the difference value between a first data and a fifth anchor, the fifth anchor is a data anchor associated with the third AI model, the sixth difference value is the difference value between the first data and a sixth anchor, and the sixth anchor is a data anchor associated with the fourth AI model.

37. The communication device according to claim 36, wherein the entry condition is Diff42 + Hys41 < Diff41, and the exit condition is Diff42 - Hys42 > Diff41.

38. The communication device according to claim 36, wherein the entry condition is Diff42 + offset42 + Hys41 < Diff41 + offset41, the exit condition is Diff42 + offset42 - Hys42 > Diff41 + offset41, and offset41 and offset42 are predefined or configured offsets.

39. The communication device according to claim 37 or 38, wherein Hys41 is a ninth hysteresis parameter, Hys42 is a tenth hysteresis parameter, and the ninth and tenth hysteresis parameters are predefined or configured parameters corresponding to the event.

40. The communication device according to any one of claims 24 to 39, wherein the first data includes monitoring data or measurement data of user equipment or network devices.

41. The communication device according to claim 40, wherein the first data includes monitoring data or measurement data related to the AI ​​model of the user equipment or the network device.

42. The communication device according to any one of claims 24 to 41, wherein the first data includes any one or more of the detection data, measurement data, channel data, neuron data of the AI ​​model, and latent output data of the AI ​​model.

43. The communication device according to any one of claims 23 to 42, further comprising a processing module configured to perform communication based on the aforementioned report.

44. The device is a communication device according to any one of claims 23 to 43, located on user equipment or a network device.

45. A communication device comprising a processor and memory, wherein the processor is connected to the memory; the memory is configured to store instructions, and the processor is configured to execute the instructions; and when the processor executes the instructions stored in the memory, the processor is enabled to perform the method according to any one of claims 1 to 22.

46. A computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed on a processor, the processor is enabled to perform the method according to any one of claims 1 to 22.

47. A computer program product comprising computer program code, wherein when the computer program code is executed on a computer, the computer is enabled to perform the method according to any one of claims 1 to 22.

Images