Communication method and communication device

Abstract

Provided are a communication method and a communication device, which can be applied to the technical field of wireless communications, the method comprising: a terminal device sending a first message to a network device when a measurement result based on a first beam set satisfies an event for indicating an artificial intelligence (AI) model to perform performance monitoring, the first message is used for requesting to monitor the performance of the AI model, and / or the first message is used for indicating that the event is satisfied; and after receiving the first message, the network equipment sends a second message to the terminal equipment, and the second message indicates the terminal equipment to monitor the performance of the AI model. By implementing the application, the network device can indicate the terminal device to monitor the performance of the AI model only after receiving the first message, so that the terminal device does not need to monitor the performance of the AI model according to a periodic monitoring mode, and signaling and resource overhead caused by frequent interaction between the terminal device and the network device can be effectively reduced.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a communication method and a communication device. Background Technology

[0002] Communication networks can employ artificial intelligence (AI) technology to reduce the resource overhead of beam management, thereby improving user experience. For example, in a traditional beam management framework, base stations need to configure terminals to measure a large number of beams. However, in AI-assisted beam management, base stations can configure terminals to measure a smaller number of beams. The terminals then input the measurement results of these beams into an AI model to predict the measurement results of a larger number of beams.

[0003] Currently, the aforementioned AI models are trained based on statistical data collected in specific time, location, and / or network scenarios. Due to the limited generalization ability of AI models, their performance degrades when the application scenario differs significantly from the data collection scenario. To ensure the effectiveness of AI-assisted beam management, base stations can configure terminals to monitor the performance of the AI ​​models. A common performance monitoring method is to configure the terminals for periodic monitoring. However, when the configured monitoring period is short, frequent interactions between the base station and the terminal are required, resulting in significant signaling and resource overhead. Summary of the Invention

[0004] This application provides a communication method and a communication device that can effectively reduce signaling and resource overhead.

[0005] In a first aspect, embodiments of this application provide a communication method applied to a terminal device, a chip in the terminal device, a device used in conjunction with the terminal device, or a device for implementing the functions of the terminal device, etc. The method includes: sending a first message to a network device based on the measurement results of a first beam set satisfying an event used to instruct an artificial intelligence (AI) model to perform performance monitoring; the first message being used to request performance monitoring of the AI ​​model; and / or, the first message being used to indicate that the measurement results of the first beam set satisfy the event; the AI ​​model being used to predict the measurement results of a third beam set based on the measurement results of a second beam set, the second beam set being the same as or different from the third beam set; the first beam set including K predicted beams in the third beam set; the K predicted beams dynamically changing with the measurement results of the second beam set; and receiving a second message sent by the network device, the second message being used to instruct performance monitoring of the AI ​​model.

[0006] In the method described in the first aspect, when the measurement results based on the first beam set satisfy an event that instructs the AI ​​model to perform performance monitoring, the terminal device sends a first message to the network device. Upon receiving the first message, the network device sends a second message to the terminal device, thereby triggering the terminal device to monitor the performance of the AI ​​model. In this way, the terminal device does not need to monitor the AI ​​model's performance periodically, effectively reducing the signaling and resource overhead caused by frequent interactions between the terminal device and the network device.

[0007] In one possible implementation, the measurement results of the second beam set include measured values ​​of the signal quality of the beams in the second beam set; the measurement results of the predicted third beam set include predicted values ​​of the signal quality of the beams in the third beam set, and / or, the predicted probability that the third beam set is the optimal beam.

[0008] In one possible implementation, the K predicted beams include the K largest predicted beams in the third beam set in terms of signal quality, or the K predicted beams include the K largest predicted beams in the third beam set in terms of the best predicted probability.

[0009] In one possible implementation, the measurement results of the first beam set include measured values ​​of the signal quality of K predicted beams; the method of sending a first message to the network device based on the measurement results of the first beam set satisfying an event used to instruct the AI ​​model to perform performance monitoring specifically includes: sending a first message to the network device based on the maximum measured value of the signal quality of the K predicted beams satisfying an event used to instruct the AI ​​model to perform performance monitoring.

[0010] In this method, the terminal device can determine the maximum measured value of the signal quality of the K predicted beams based on the measured value of the signal quality of each predicted beam, and determine the event that satisfies the requirement to instruct the artificial intelligence (AI) model to perform performance monitoring based on this maximum measured value, and then send the first message to the network device.

[0011] In one possible implementation, the above event is that the M1 maximum measured values ​​of the signal quality obtained from M1 consecutive measurements of K predicted beams are all less than a first threshold, where M1 is a positive integer.

[0012] In this method, the terminal device can execute the following process M1 times consecutively: predict the measurement results of the third beam set based on the measurement results of the second beam set; select K predicted beams from the third beam set based on the predicted measurement results of the third beam set; measure the selected K predicted beams to obtain the measured signal quality values ​​of the K predicted beams; and obtain the maximum measured signal quality value of the K predicted beams based on the measured signal quality values ​​of the K predicted beams. Since the measurement results of the third beam set obtained each time can be the same or different, the selected K predicted beams each time can be the same or different, which leads to the maximum measured signal quality value of the K predicted beams each time being the same or different.

[0013] When the event in this method is met, it indicates that the performance of the current AI model may be poor, resulting in inaccurate AI model predictions. Consequently, the beam with the best signal quality in the third beam concentration for M1 consecutive times is not selected into the K predicted beams, thus causing the maximum measured value of the K predicted beams obtained for M1 consecutive times to be less than the first threshold. In this case, sending the first message to the network device can promptly trigger the network device to instruct the terminal device to monitor the performance of the AI ​​model, thereby ensuring the effectiveness of AI-assisted beam management.

[0014] In one possible implementation, the above event is that the first proportion is greater than the second threshold; wherein, the first proportion is the proportion of the largest measured value less than the third threshold among the M2 largest measured values ​​of signal quality obtained from M2 consecutive measurements of K predicted beams, and M2 is a positive integer.

[0015] In this method, the terminal device can execute the following process M2 times consecutively: predict the measurement results of the third beam set based on the measurement results of the second beam set; select K predicted beams from the third beam set based on the predicted measurement results of the third beam set; measure the selected K predicted beams to obtain the measured signal quality values ​​of the K predicted beams; and obtain the maximum measured signal quality value of the K predicted beams based on the measured signal quality values ​​of the K predicted beams. Since the measurement results of the third beam set obtained each time can be the same or different, the selected K predicted beams can be the same or different each time, which in turn leads to the maximum measured signal quality value of the K predicted beams obtained each time being the same or different.

[0016] When the event in this method is met, it indicates that the current AI model's performance may be poor, leading to inaccurate AI model predictions. Consequently, the beam with the best signal quality in the third beam concentration M2 consecutive times is not selected into the K predicted beams. As a result, the proportion of the maximum measured values ​​less than the third threshold among the M2 maximum measured values ​​of the K predicted beams obtained in M2 consecutive times exceeds the second threshold. For example, if the second threshold is 50%, the number of maximum measured values ​​less than the third threshold among the M2 maximum measured values ​​exceeds M2 / 2. In this case, sending the first message to the network device can promptly trigger the network device to instruct the terminal device to monitor the performance of the AI ​​model, thereby ensuring the effectiveness of AI-assisted beam management.

[0017] In one possible implementation, the measurement results of the first beam set include the measured values ​​of the signal quality of K predicted beams; the method of sending a first message to the network device based on the measurement results of the first beam set satisfying an event for instructing the AI ​​model to perform performance monitoring specifically includes: determining an event for instructing the AI ​​model to perform performance monitoring based on the first beam and the second beam among the K predicted beams, wherein the second beam is the beam with the largest measured signal quality value among the K predicted beams, and the first beam is the beam with the largest predicted signal quality value in the third beam set, or the first beam is the beam with the highest predicted probability of the optimal beam in the third beam set; and sending the first message to the network device.

[0018] In this method, the terminal device can determine the first beam among the K predicted beams based on the measurement results of the predicted third beam set, and determine the second beam among the K predicted beams based on the measured values ​​of the signal quality of the K predicted beams. Based on the first beam and the second beam, the terminal device can determine the event that satisfies the requirement for performance monitoring of the artificial intelligence (AI) model, and then send the first message to the network device.

[0019] In one possible implementation, the above event is a second proportion less than a fourth threshold; wherein the second proportion is the proportion of events in which the first beam and the second beam are the same for M3 consecutive determinations, and M3 is a positive integer.

[0020] In this method, the terminal device can continuously execute the following process M3 times: predict the measurement results of the third beam set based on the measurement results of the second beam set; based on the predicted measurement results of the third beam set, select K predicted beams from the third beam set and determine the first beam among the K predicted beams; measure the selected K predicted beams to obtain the measured values ​​of the signal quality of the K predicted beams; and determine the second beam among the K predicted beams based on the measured values ​​of the signal quality of the K predicted beams. Since the measurement results of the third beam set obtained each time can be the same or different, the first beam determined each time can be the same or different, and the second beam determined each time can be the same or different.

[0021] When the first beam and the second beam are identical, it indicates that the beam with the best signal quality or the optimal beam among the K predicted beams has the highest prediction probability. In actual measurement, this also indicates that the beam with the best signal quality or the optimal beam among the K predicted beams, thus demonstrating the accuracy of the AI ​​model's prediction. If the proportion of events where the first beam and the second beam are identical in M3 trials does not exceed the third threshold, it indicates that the AI ​​model's predictions are less accurate. For example, if M3 is 10 trials, and the first beam and the second beam are identical in 3 trials and different in 7 trials, the second proportion is 0.3. When the fourth threshold is 0.5, it indicates that the AI ​​model's predictions are less accurate and the AI ​​model's performance is poor. In this case, sending the first message to the network device can promptly trigger the network device to instruct the terminal device to monitor the AI ​​model's performance, thereby ensuring the effectiveness of AI-assisted beam management.

[0022] In one possible implementation, the aforementioned event is that the average of the M4 first prediction errors determined consecutively M4 times is greater than a fifth threshold; where M4 is a positive integer, and the first prediction error is determined based on the measured values ​​of the signal quality of the first beam and the measured values ​​of the signal quality of the second beam.

[0023] In this method, the terminal device can continuously execute the following process M4 times: predict the measurement results of the third beam set based on the measurement results of the second beam set; based on the predicted measurement results of the third beam set, select K predicted beams from the third beam set and determine the first beam among the K predicted beams; measure the selected K predicted beams to obtain the measured values ​​of the signal quality of the K predicted beams; determine the second beam among the K predicted beams based on the measured values ​​of the signal quality of the K predicted beams; and determine the first prediction error based on the measured values ​​of the signal quality of the first beam and the measured values ​​of the signal quality of the second beam. Since the measurement results of the third beam set obtained each time can be the same or different, the first beam determined each time can be the same or different, and the second beam determined each time can be the same or different.

[0024] When the average of the M4 first prediction errors is greater than the fifth threshold in M4 consecutive determinations, it indicates that the AI ​​model's prediction is inaccurate and the AI ​​model's performance is poor. In this case, sending the first message to the network device can promptly trigger the network device to instruct the terminal device to monitor the performance of the AI ​​model, thereby ensuring the effectiveness of AI-assisted beam management.

[0025] In one possible implementation, the measurement results of the first beam set include the measured values ​​of the signal quality of the beams in the first beam set; when the measurement results of the first beam set satisfy the performance monitoring event of the artificial intelligence (AI) model, a first message is sent to the network device, including: determining a second prediction error based on the measured and predicted values ​​of the signal quality of K predicted beams; when the second prediction error satisfies the performance monitoring event of the AI ​​model, the first message is sent to the network device.

[0026] In this method, for example, the second prediction error can be the sum of the absolute values ​​of the differences between the measured values ​​and the predicted values ​​of each of the K prediction beams; or, the second prediction error can be the average of the absolute values ​​of the differences between the measured values ​​and the predicted values ​​of each of the K prediction beams; or, the second prediction error can be the sum of the squares of the differences between the measured values ​​and the predicted values ​​of each of the K prediction beams; or, the second prediction error can be the average of the squares of the differences between the measured values ​​and the predicted values ​​of each of the K prediction beams, but is not limited to these.

[0027] In one possible implementation, the above event is that the average of the second prediction errors determined M5 times consecutively is greater than the sixth threshold, where M5 is a positive integer.

[0028] In this method, the terminal device can continuously execute the following process M5 times: predict the measurement results of the third beam set based on the measurement results of the second beam set; select K predicted beams from the third beam set based on the predicted measurement results of the third beam set; measure the selected K predicted beams to obtain the measured values ​​of the signal quality of the K predicted beams; and determine the second prediction error based on the measured values ​​of the signal quality of the K predicted beams and the predicted values.

[0029] When the average of the M5 second prediction errors determined in M5 consecutive trials is greater than the sixth threshold, it indicates that the AI ​​model's prediction is inaccurate and the AI ​​model's performance is poor. In this case, sending the first message to the network device can promptly trigger the network device to instruct the terminal device to monitor the performance of the AI ​​model, thereby ensuring the effectiveness of AI-assisted beam management.

[0030] Secondly, embodiments of this application provide a communication method applied to a network device, a chip in a network device, a device used in conjunction with a network device, or a device for implementing the functions of a network device, etc. The method includes: receiving a first message sent by a terminal device, the first message being used to request performance monitoring of an AI model, and / or, the first message being used to indicate that the measurement results of a first beam set meet an event used to instruct the AI ​​model to perform performance monitoring, the AI ​​model being used to predict the measurement results of a third beam set based on the measurement results of a second beam set, the second beam set being the same as or different from the third beam set, the first beam set including K predicted beams in the third beam set, the K predicted beams dynamically changing with the measurement results of the second beam set, K being a positive integer; and sending a second message to the terminal device, the second message being used to instruct performance monitoring of the AI ​​model.

[0031] In the method described in the second aspect, the network device can send a second message to the terminal device after receiving the first message, thereby triggering the terminal device to monitor the performance of the AI ​​model. This eliminates the need for the terminal device to periodically monitor the AI ​​model's performance, effectively reducing signaling and resource overhead caused by frequent interactions between the terminal device and the network device.

[0032] Thirdly, embodiments of this application provide a communication device, which includes one or more processors and one or more memories; wherein the one or more memories are coupled to one or more processors, and the one or more memories are used to store computer instructions, which, when the one or more processors execute the computer instructions, cause the communication device to perform the methods described in the first or second aspect above, and any possible implementation thereof.

[0033] Fourthly, embodiments of this application provide a chip system including at least one processor and an interface, the interface being used to receive computer instructions and transmit them to the at least one processor; the at least one processor executes the computer instructions to cause a communication device to perform the methods described in the first or second aspect above, and any possible implementation thereof.

[0034] Fifthly, embodiments of this application provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer instructions, which, when executed by a processor, cause a communication device to perform the methods described in the first or second aspect above, and any possible implementation thereof.

[0035] Sixthly, embodiments of this application provide a computer program product including computer instructions that, when executed by a processor, cause a communication device to perform the methods described in the first or second aspect above, or any possible implementation thereof. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the architecture of a communication system provided in an embodiment of this application;

[0037] Figure 2 This is a schematic diagram of the input and output of an AI model provided in an embodiment of this application;

[0038] Figure 3 This is a schematic diagram of the input and output of another AI model provided in an embodiment of this application;

[0039] Figure 4 This is a schematic diagram illustrating the input and output of another AI model provided in an embodiment of this application;

[0040] Figure 5 This is a schematic diagram of an AI-assisted beam management process provided in an embodiment of this application;

[0041] Figure 6 This is a schematic diagram of another AI-assisted beam management process provided in an embodiment of this application;

[0042] Figure 7 This is a flowchart illustrating a communication method provided in an embodiment of this application;

[0043] Figure 8 This is a schematic diagram of the software structure of a communication device provided in an embodiment of this application;

[0044] Figure 9 This is a schematic diagram of the hardware structure of a communication device provided in an embodiment of this application. Detailed Implementation

[0045] In the description of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. "And / or" in this application is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Furthermore, "at least one" can refer to one or more, and "multiple" can refer to two or more. "First," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., are not necessarily different.

[0046] It should be noted that in this application, the terms "example" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "example" or "for example" in this application should not be construed as being preferred or superior to other embodiments or designs. Specifically, the use of terms such as "example" or "for example" is intended to present the relevant concepts in a concrete manner.

[0047] The following section introduces the communication systems applicable to the embodiments of this application:

[0048] The embodiments of this application can be applied to various communication systems, such as 5th generation (5G) or new radio (NR) systems, long term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, and LTE time division duplex (TDD) systems. The embodiments of this application can also be applied to future communication systems, such as 6th generation (6G) mobile communication systems. The technical solutions provided in this application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), and Internet of Things (IoT) communication systems or other communication systems.

[0049] Figure 1 A schematic diagram of the architecture of a communication system provided in an embodiment of this application is shown. Figure 1 As shown, the communication system includes a network device 110 and a terminal device 120. A communication connection is established between the network device 110 and the terminal device 120, and the network device 110 and the terminal device 120 transmit uplink / downlink data or information based on this communication connection. It should be noted that the communication system may also include other devices, but this embodiment does not limit this.

[0050] Network device 110 is a device that connects terminal device 120 to a wireless network. Network device 110 can be a node in a wireless access network, also known as a base station, or a radio access network (RAN) node (or device). The base station can be a distributed antenna system, and the device communicating with a terminal device can be a radio frequency head of the base station. For example, network equipment 110 may include evolved base stations (NodeBs or eNBs or eNodeBs) in LTE systems or evolved LTE systems (LTE-Advanced, LTE-A), such as traditional macro base stations (eNBs) and micro base stations (eNBs) in heterogeneous network scenarios; or it may include next-generation node Bs (gNBs) in 5G systems, or it may include transmission reception points (TRPs), home base stations (e.g., home evolved NodeBs, or home Node Bs, HNBs), base band units (BBUs), base band pools (BBU pools), or wireless fidelity (Wi-Fi) access points (APs); or it may be base station equipment in 5G or network equipment in future evolved public land mobile networks (PLMNs); or it may be wearable devices or vehicle-mounted devices; or it may include centralized units (CUs) and distributed units (DUs); or it may include non-terrestrial networks. Network devices in NTN (Network Network, NTN) can be deployed on high-altitude platforms or satellites, but this application does not limit the scope of the embodiments.

[0051] Terminal equipment 120, also known as user equipment (UE), mobile station (MS), mobile terminal (MT), etc., is a device that includes wireless communication capabilities (providing voice / data connectivity to users). Examples include handheld devices with wireless connectivity or in-vehicle devices. For instance, terminal equipment 120 can be a mobile phone, tablet computer, laptop computer, PDA, mobile internet device (MID), wearable device, virtual reality (VR) device, augmented reality (AR) device, wireless terminal in industrial control, wireless terminal in vehicle networking, wireless terminal in self-driving, wireless terminal in remote medical surgery, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, or wireless terminal in smart home, etc.

[0052] based on Figure 1 The following describes the relevant concepts involved in the embodiments of this application:

[0053] I. The Principle of AI-Assisted Beam Management

[0054] Artificial intelligence (AI) technology refers to the technology of performing complex calculations by simulating the human brain. With the improvement of data storage and computing power, AI technology is increasingly being applied in the field of communications to improve network performance and user experience. For example, AI technology can be applied to scenarios such as beam management, mobility management, and load balancing.

[0055] In beam management scenarios, terminal devices can configure and activate AI models with beam prediction capabilities. The terminal devices use these AI models to predict beam measurement results, and network devices assist in beam management based on the predicted beam measurement results.

[0056] Among them, the beam prediction function of the AI ​​model includes spatial beam prediction function and temporal beam prediction function.

[0057] Suppose there are two beam sets, Set 1 and Set 2, each of which can include one or more beams.

[0058] (1) Spatial beam prediction function: The terminal device predicts the measurement results of Set 2 based on the measurement results of Set 1. Specifically, the terminal device measures the measurement results of Set 1, which includes the measured signal quality values ​​of each beam in Set 1. Then, the terminal device inputs the measured signal quality values ​​of the beams in Set 1 into the AI ​​model, and the AI ​​model outputs the predicted measurement results of Set 2.

[0059] The measurement results of Set 1 and the predicted measurement results of Set 2 are measured at the same time. Signal quality may include, but is not limited to, one or more of the following: reference signal receiving power (RSRP), reference signal receiving quality (RSRQ), or signal to interference plus noise ratio (SINR).

[0060] Optionally, if the AI ​​model is a regression model, the predicted measurement results for Set 2 may include the predicted signal quality of each beam in Set 2.

[0061] Optionally, if the AI ​​model is a classification model, the predicted measurement results for Set 2 may include the predicted probability that each beam in Set 2 is the optimal beam. Here, the optimal beam can also be called the best beam, etc., and the optimal beam refers to the set Set 2. For example, the optimal beam can be the beam with the best signal quality in Set 2, then the predicted measurement results for Set 2 include the predicted probability that each beam in Set 2 is the beam with the best signal quality in Set 2.

[0062] Optionally, if the AI ​​model includes both a regression model and a classification model, the predicted measurement results of Set 2 can include both the predicted signal quality of each beam in Set 2 and the predicted probability that each beam in Set 2 is the optimal beam.

[0063] Optionally, the beams in Set 1 are different from those in Set 2, and the number of beams in Set 1 is less than the number of beams in Set 2.

[0064] For example, the beam in Set 1 is the synchronization signal / PBCH block (SSB) beam, which can also be called the beam that transmits the SSB or the beam that carries the SSB. The beam in Set 2 is the channel state information reference signal (CSI-RS) beam, which can also be called the beam that transmits the CSI-RS or the beam that carries the CSI-RS. Then, the input and output of the AI ​​model are as follows: Figure 2 As shown: The input of the AI ​​model is the measured value of the RSRP of the SSB beam, and the output is the predicted value of the RSRP of the CSI-RS beam and / or the predicted probability that the CSI-RS beam is the optimal beam.

[0065] Optionally, Set 1 is a subset of Set 2.

[0066] For example, Set 1 includes CSI-RS1 beam, CSI-RS2 beam, and CSI-RS3 beam, and Set 2 includes CSI-RS1 beam, CSI-RS2 beam, CSI-RS3 beam, CSI-RS4 beam, CSI-RS5 beam, CSI-RS6 beam, and CSI-RS7 beam. Then the input and output of the AI ​​model are as follows: Figure 3 As shown, the input to the AI ​​model is the measured RSRP of the CSI-RS1 beam to the CSI-RS3 beam, and the output is the predicted RSRP of the CSI-RS1 beam to the CSI-RS7 beam and / or the predicted probability of the optimal beam.

[0067] (2) Timing beam prediction function: The terminal device predicts the measurement results of Set 2 in the future T2 time by using the measurement results of Set 1 in the past T1 time.

[0068] Where T1 and T2 are positive integers, the past T1 moments are the T1 moments before the current moment, and the future T2 moments are the T2 moments after the current moment.

[0069] Specifically, the terminal device measures the measurement results of Set 1 over the past T1 time points. The measurement results of Set 1 over the past T1 time points include the measured values ​​of the signal quality of each beam in Set 1 over the past T1 time points. Then, the terminal device inputs the measured values ​​of the signal quality of the beams in Set 1 over the past T1 time points into the AI ​​model. The AI ​​model outputs the predicted measurement results of Set 2 over the next T2 time points.

[0070] Optionally, if the AI ​​model is a regression model, the predicted measurement results of Set 2 at future T2 times may include the predicted signal quality of each beam in Set 2 at future T2 times.

[0071] Optionally, if the AI ​​model is a classification model, the predicted measurement results of Set 2 at the next T2 times may include the predicted probability that each beam in Set 2 is the optimal beam at the next T2 times.

[0072] Optionally, if the AI ​​model includes both a regression model and a classification model, the predicted measurement results of Set 2 at the next T2 times can include both the predicted signal quality of each beam in Set 2 at the next T2 times and the predicted probability that each beam in Set 2 is the optimal beam at the next T2 times.

[0073] Optionally, referring to the relationship between Set 1 and Set 2 in the above-mentioned spatial beam prediction function, the beams in Set 1 are not the same as the beams in Set 2, and the number of beams in Set 1 is less than the number of beams in Set 2; or, Set 1 is a subset of Set 2.

[0074] Optionally, Set 1 and Set 2 can also be the same set. If Set 1 and Set 2 are the same set, the timing beam prediction function can also be described as the terminal device predicting the measurement results of Set 1 at the next T2 times based on the measurement results of Set 1 at the past T1 times.

[0075] Taking Set 1 as a subset of Set 2 as an example, Figure 4 As shown, time t is the current time, time t-T1 to time t-1 are the past T1 times, and time t+1 to time t+T2 are the future T2 times. Then, the input of the AI ​​model is the measured RSRP of the CSI-RS1 beam to the CSI-RS3 beam from time t-T1 to time t-1, and the output is the predicted RSRP of the CSI-RS1 beam to the CSI-RS7 beam from time t+1 to time t+T2 and / or the predicted probability of the optimal beam from time t+1 to time t+T2.

[0076] Optionally, the input to the AI ​​model may also include auxiliary information such as the beam index information in Set 1, but is not limited to this.

[0077] II. Steps of AI-Assisted Beam Management

[0078] Based on the principles of the aforementioned spatial beam prediction or temporal beam prediction functions, network devices can select appropriate beams for terminal devices, thereby transmitting downlink data or information to the terminal devices on those beams. For example, a network device can select an appropriate SSB beam for transmitting SSB, or an appropriate CSI-RS beam for transmitting CSI-RS.

[0079] The specific steps of this process are described below:

[0080] (1) If the beam prediction function of the AI ​​model configured in the terminal device is the spatial beam prediction function, then the terminal device and the network device shall perform the following steps 501 to 509:

[0081] 501. The network device performs a Set 1 beam scan on the terminal device.

[0082] Among them, the network device performing Set 1 beam scanning on the terminal device refers to the network device scanning each beam in Set 1.

[0083] Optionally, before performing a Set 1 beam scan, the network device may send the configuration information of the beams in Set 1 to the terminal device.

[0084] For example, if the beam in Set 1 is an SSB beam, then the configuration information of the beam in Set 1 includes, but is not limited to, at least one of the following: SSB index information, SSB frequency position, subcarrier spacing, and period.

[0085] 502. The terminal device measures the measurement results of Set 1, and makes a prediction based on the measurement results of Set 1 to obtain the predicted measurement results of Set 2.

[0086] When the network device scans each beam in Set 1, the terminal device can measure each beam in Set 1 to obtain the measurement results of Set 1. The terminal device inputs the measurement results of Set 1 into the AI ​​model, and the AI ​​model outputs the predicted measurement results of Set 2 based on the principle of the above-mentioned spatial beam prediction function.

[0087] 503. The terminal equipment determines the Top K beam set based on the predicted measurement results of Set 2.

[0088] The terminal device can select beams from Set 2 based on the predicted measurement results of Set 2 to obtain the Top K beam set, which includes K predicted beams.

[0089] Optionally, if the predicted measurement results of Set 2 include the predicted signal quality of the beams in Set 2, then the K predicted beams can be the K largest beams in terms of predicted signal quality; or, if the predicted measurement results of Set 2 include the predicted probability that the beams in Set 2 are the optimal beams, then the K predicted beams can be the K largest beams in terms of predicted probability.

[0090] For example, in Set 2, the beams are beam 1 to beam 7. The beams are sorted from largest to smallest according to the predicted value or predicted probability of the signal quality, resulting in beam 1, beam 3, beam 5, beam 2, beam 6, beam 7, and beam 4. If K is 3, then the K predicted beams in the Top K beam set are beam 1, beam 3, and beam 5.

[0091] 504. The terminal device sends the Top K beam set indication information to the network device.

[0092] For example, the indication information for the Top K beam set includes the index information for each of the K predicted beams.

[0093] Optionally, the indication information for the Top K beam set may also include the predicted signal quality of each of the K predicted beams and / or the predicted probability that each predicted beam is the optimal beam.

[0094] 505. The network equipment performs Top K beamforming scans on the terminal equipment.

[0095] Among them, the Top K beamset scanning of terminal devices by network devices can refer to the network devices scanning the K predicted beams in the Top K beamset.

[0096] Optionally, before performing Top K-beam set beam scanning, the network device can first send the configuration information of each of the K predicted beams to the terminal device. The configuration information of the predicted beams can be referred to the beam configuration information in Set 1, which will not be elaborated here.

[0097] 506. The measurement results of the Top K-wave beam set are obtained by the terminal equipment.

[0098] When the network device scans the K predicted beams, the terminal device can measure the K predicted beams to obtain the measurement results of the Top K beam set. The measurement results of the Top K beam set include the measured value of the signal quality of each of the K predicted beams.

[0099] 507. The terminal device sends the measurement results of the Top K beam set to the network device.

[0100] 508. The network device determines the target beam based on the measurement results of the Top K beam set.

[0101] For example, a network device can determine a target beam based on the measured signal quality of each of the K predicted beams. The number of target beams can be one or more, such as the beam with the largest measured signal quality among the K predicted beams.

[0102] 509. The network device sends the target beam indication information to the terminal device.

[0103] For example, the indication information of the target beam may include the index information of the target beam.

[0104] (2) When the AI ​​model configured on the terminal device has a spatial prediction function, the terminal device and the network device shall execute the following steps 601 to 605:

[0105] 601. The network device performed a Set 1 beam scan on the terminal device at the past T1 time point.

[0106] In this context, "the network device performs Set 1 beam scanning on the terminal device in the past T1 time periods" means that the network device scans each beam in Set 1 at each time period in the past T1 time periods.

[0107] Optionally, before performing a Set 1 beam scan, the network device may send the configuration information of the beams in Set 1 to the terminal device.

[0108] 602. The terminal device measures the measurement results of Set 1 at the past T1 time points, and makes a prediction based on the measurement results of Set 1 at the past T1 time points to obtain the predicted measurement results of Set 2 at the future T2 time points.

[0109] Specifically, the terminal device can measure each beam in Set 1 at each of the past T1 time points, obtaining the measurement results of Set 1 at each of the past T1 time points. The terminal device inputs the measurement results of Set 1 at each of the past T1 time points into the AI ​​model, and the AI ​​model outputs the predicted measurement results of Set 2 at the next T2 time points, based on the principle of the aforementioned time series prediction function.

[0110] 603. The terminal equipment determines the Top K beam set corresponding to the next T2 time points based on the predicted measurement results of Set 2.

[0111] Taking one of the future T2 time points as an example, the terminal device determines the Top K beam set corresponding to this time point based on the predicted measurement results of Set 2 at this time point. The Top K beam set corresponding to this time point includes the K predicted beams corresponding to this time point.

[0112] Optionally, if the predicted measurement result of Set 2 at this moment includes the predicted signal quality of the beams in Set 2 at this moment, then the K predicted beams corresponding to this moment can be the K largest beams in terms of predicted signal quality; or, if the predicted measurement result of Set 2 at this moment includes the predicted probability that the beams in Set 2 are the best beams at this moment, then the K predicted beams corresponding to this moment can be the K largest beams in terms of predicted probability.

[0113] Optionally, the K predicted beams corresponding to different times in the future T2 times may be the same or different.

[0114] For example, in Set 2, the beams are beam 1 to beam 7, the current time is time t, and the next T2 times include time t+1 and time t+2. If beam 1 to beam 7 are sorted from largest to smallest according to the predicted value or predicted probability of the signal quality at time t+1, we get beam 1, beam 3, beam 5, beam 2, beam 6, beam 7, and beam 4, and K is 3. Then the Top K beam set at time t+1 includes three predicted beams: beam 1, beam 3, and beam 5. If beams 1 to 7 are sorted from largest to smallest according to the predicted signal quality or predicted probability at time t+2, we get beam1, beam 5, beam 6, beam 7, beam 3, beam 2, and beam 4, and K is 3. Then the Top K beam set at time t+1 includes three predicted beams: beam 1, beam 5, and beam 6.

[0115] 604. The terminal device sends the indication information of the Top K beam set corresponding to the next T2 time points to the network device.

[0116] Optionally, the terminal device can also report the indication information of the Top K beam set corresponding to the next T2 time points to the network device.

[0117] Alternatively, the terminal device may report the indication information of the Top K beamset corresponding to one of the next T2 time points. For example, if the current time is time t, and the next T2 time points include time t+1 and time t+2, the terminal device may report the indication information of the Top K beamset corresponding to time t+1 at time t+1 and report the indication information of the Top K beamset corresponding to time t+2 at time t+2.

[0118] For example, the indication information for the Top K beam set corresponding to one of the next T2 time points includes the index information of each of the K predicted beams corresponding to that time point. Optionally, it may also include the measurement results of each predicted beam.

[0119] 605. The network device selects the target beam corresponding to the next T2 time points for the terminal device.

[0120] Taking one of the next T2 time points as an example, the network device can scan the K predicted beams corresponding to that time point. The terminal device measures the results of the K predicted beams at that time point, such as measured signal quality values, and then reports these results. Based on these results, the network device selects the target beam for that time point. Furthermore, the network device can send indication information of the target beam to the terminal device. Following this process, the network device can determine the corresponding target beam for each of the next T2 time points.

[0121] In the above Figure 5 or Figure 6 If the performance of the AI ​​model is poor, the measurement results predicted by the AI ​​model may be inaccurate, which may lead to the network equipment selecting an unsuitable target beam for the terminal device.

[0122] For example, if the signal strength of the target beam remains below a set threshold when the terminal device receives downlink data / information according to the target beam indication information, the terminal device determines that the target beam has failed. In this case, in order not to affect the downlink transmission performance of the terminal device, it is necessary to perform timely management operations such as switching, activating, deactivating, or rolling back the AI ​​model, or roll back to the traditional beam management method to select a suitable beam for the terminal device.

[0123] III. Performance Monitoring Methods for AI Models

[0124] The AI ​​model in the aforementioned AI-assisted beam management is trained based on statistical data collected in specific time, location, and / or network scenarios. Furthermore, the generalization ability of the AI ​​model is limited. Therefore, when the application scenario of the AI ​​model differs significantly from the data collection scenario, the performance of the AI ​​model will decline.

[0125] In order to monitor the performance changes of the AI ​​model in a timely manner, so as to perform the above-mentioned management operations or fall back to the traditional beam management method when the AI ​​model performs poorly, the network device can be configured to periodically monitor the performance of the AI ​​model. For example, the network device can be configured with the monitoring period and the reporting method of the monitoring results.

[0126] Specifically, the network device first sends configuration information regarding the monitoring cycle and the reporting method of the monitoring results to the terminal device, and then... Figure 5 For example, the network device sends the beam configuration information of Set 2 to the terminal device. Then, while performing step 501, the network device can also perform a Set 2 beam scan, allowing the terminal device to obtain the measurement results of Set 2. In this way, the terminal device can determine the monitoring results of the AI ​​model performance based on the measured results of Set 2 and the predicted Set 2 measurement results from step 502, and report the monitoring results to the network device. For instance, based on the measured results of Set 2 and the predicted Set 2 measurement results from step 502, the terminal device obtains the average prediction error of each beam in Set 2 and reports it to the network device.

[0127] When the monitoring cycle configured for network devices is long, terminal devices may not be able to determine the monitoring results in a timely manner, causing network devices to be unable to obtain the performance of the AI ​​model in a timely manner, thus affecting the effectiveness of AI-assisted beam management. However, when the monitoring cycle configured for network devices is short, network devices need to frequently perform Set 2 beam scans according to the configured monitoring cycle, and terminal devices need to frequently measure the Set 2 measurement results and report the monitoring results, resulting in more interaction between network devices and terminal devices, and greater signaling and resource overhead.

[0128] In order to reduce the interaction between network devices and terminal devices, thereby effectively reducing signaling and resource overhead, embodiments of this application provide a communication method and related equipment.

[0129] The communication method will be described in detail below:

[0130] Figure 7The illustration shows a flowchart of a communication method provided in an embodiment of this application. The main body executing this method can be a terminal device and a network device, or the main body can be a chip in the terminal device and a chip in the network device, or the main body can be other types of products. Those skilled in the art can make further extensions based on the content disclosed in the specification. Figure 7 The methods shown and the following embodiments are implemented using terminal devices and network devices as examples. For instance, the terminal device can be terminal device 120 as described above, and the network device can be network device 110 as described above. Figure 7 As shown, the method includes steps 701 to 702. Wherein:

[0131] 701. The terminal device sends a first message to the network device based on the measurement results of the first beam set, which satisfies the event used to instruct the artificial intelligence (AI) model to perform performance monitoring.

[0132] In this embodiment, the AI ​​model is used to predict the measurement results of the third beam set based on the measurement results of the second beam set. The terminal device can first configure and activate the AI ​​model. Then, the network device can perform beam scanning of the second beam set. While the network device is performing beam scanning of the second beam set, the terminal device measures the measurement results of the second beam set and inputs these results into the AI ​​model. The AI ​​model then outputs the predicted measurement results of the third beam set. The specific implementation of this process can refer to steps 501 to 502 above, or the description of steps 601 to 602 above.

[0133] The second beam set can be referred to in the description of Set 1 above, and the third beam set can be referred to in the description of Set 2 above. The second beam set and the third beam set may be the same or different.

[0134] For example, if the AI ​​model supports the above-mentioned spatial beam prediction function, the second beam set is different from the third beam set; if the AI ​​model supports the above-mentioned temporal beam prediction function, the second beam set is different from or the same as the third beam set.

[0135] For example, the measurement results of the second beam set include the measured values ​​of the signal quality of the beams in the second beam set. The signal quality may include, but is not limited to, one or more of RSRP, RSRQ, SINR, etc., and RSRP may be layer-1 RSRP (L1-RSRP).

[0136] For example, the predicted measurement results of the third beam set include predicted values ​​of the signal quality of the beams in the third beam set, and / or the predicted probability that a beam in the third beam set is the optimal beam. Here, the optimal beam refers to the third beam set, and the optimal beam can be the beam with the best signal quality in the third beam set. For example, a prediction probability of 0.8 means that this beam has a 0.8 probability of being the beam with the best signal quality in the third beam set.

[0137] After obtaining the measurement results of the predicted third beam set, the terminal device can select beams from the third beam set to determine the first beam set. The first beam set includes K predicted beams, which are beams in the third beam set, where K is a positive integer. This process can be referred to the description of steps 503 or 603 above.

[0138] In one example, refer to the above. Figure 5 and Figure 6 The K predicted beams are the K largest predicted beams in terms of signal quality in the third beam set, or the K predicted beams are the K largest predicted beams in terms of the best predicted probability in the third beam set.

[0139] For example, after the terminal device obtains the predicted signal quality values ​​of the beams in the third beam set, it can sort the beams in the third beam set in descending order of the predicted values ​​to obtain a sorting result. The first K beams in the sorting result are the K predicted beams. Alternatively, it can sort the beams in the third beam set in ascending order of the predicted values ​​to obtain a sorting result. The last K beams in the sorting result are the K predicted beams.

[0140] For example, after obtaining the predicted probability that the beam in the third beam set is the optimal beam, the terminal device can sort the beams in the third beam set in descending order of predicted probability to obtain a sorting result, and the first K beams in the sorting result are the K predicted beams; or, it can sort the beams in the third beam set in ascending order of predicted probability to obtain a sorting result, and the last K beams in the sorting result are the K predicted beams.

[0141] In another example, the K predicted beams include the K beams in the third beam set whose predicted signal quality is greater than a prediction threshold, or the K predicted beams include the K beams in the third beam set whose predicted probability is greater than a prediction probability threshold.

[0142] For example, after obtaining the predicted signal quality of the third beam set, the terminal device compares the predicted signal quality of each beam with a prediction threshold to determine the number of beams whose predicted signal quality is greater than the prediction threshold as L, where L is a positive integer. Based on L beams, K predicted beams are determined. Alternatively, after obtaining the predicted probability that the third beam set is the optimal beam, the terminal device compares the predicted probability of each beam with a prediction probability threshold to determine the number of beams whose predicted probability is greater than the prediction probability threshold as L. Based on L beams, K predicted beams are determined.

[0143] Optionally, if L is greater than K, the terminal device can randomly select K beams from these L beams, and the selected K beams are the K predicted beams; if L is less than K, the value of K can be updated to L, thereby determining these L beams as the K predicted beams.

[0144] Optionally, the K predicted beams in the first beam set change dynamically with the measurement results of the second beam set.

[0145] For example, the network device can perform multiple beam scans of the second beam set, thereby enabling the terminal device to perform the following operations multiple times: when performing beam scans of the second beam set, the terminal device measures the measurement results of the second beam set, inputs the measurement results of the second beam set into the AI ​​model, and the AI ​​model outputs the predicted measurement results of the third beam set; the terminal device determines K predicted beams in the first beam set based on the predicted measurement results of the third beam set.

[0146] Because factors such as the network environment and the location of the terminal device change each time the terminal device performs these operations, the measurement results of the second beam set obtained by the terminal device will change each time. Consequently, the measurement results of the predicted third beam set may be the same or different, and the K predicted beams in the first beam set may be the same or different.

[0147] Furthermore, the terminal equipment can measure the K predicted beams in the first beam set to obtain the measurement results of the first beam set.

[0148] The measurement results of the first beam set include the measured values ​​of the signal quality of the K predicted beams in the first beam set.

[0149] For example, after the terminal device determines K predicted beams, it can send indication information of the K predicted beams to the network device in step 504 or step 604 above. The indication information of the K predicted beams includes at least the index information of each of the K predicted beams. Then, the network device can perform beam scanning on the K predicted beams in step 505 above. While the network device is performing beam scanning on the K predicted beams, the terminal device measures the K predicted beams in step 506 above to obtain the measured values ​​of the signal quality of the K predicted beams.

[0150] Furthermore, the terminal device can send the first message to the network device based on the measured values ​​of the signal quality of these K predicted beams to meet the performance monitoring events of the AI ​​model.

[0151] In one example, the first message is used to request monitoring of the AI ​​model's performance.

[0152] In another example, the first message is used to indicate that the measurement results of the first beam set satisfy the event used to instruct the AI ​​model to perform performance monitoring, or the first message is used to indicate that the measured values ​​of the signal quality of the K predicted beams satisfy the event used to instruct the AI ​​model to perform performance monitoring, or the first message is used to indicate that the event used to instruct the AI ​​model to perform performance monitoring has occurred / been satisfied / be triggered.

[0153] In yet another example, the first message is used to request performance monitoring of the AI ​​model and to indicate that the measurement results of the first beam set satisfy an event for instructing the AI ​​model to perform performance monitoring; or, the first message is used to request performance monitoring of the AI ​​model and to indicate that the measured values ​​of the signal quality of the K predicted beams satisfy an event for instructing the AI ​​model to perform performance monitoring; or, the first message is used to request performance monitoring of the AI ​​model and to indicate that the event for instructing the AI ​​model to perform performance monitoring has occurred / been satisfied / triggered.

[0154] Optionally, after obtaining the measured signal quality values ​​of the K predicted beams, the terminal device can report the measured signal quality values ​​of the K predicted beams to the network device, referring to step 507. The first message may be the same message as the message carrying the measured signal quality values ​​of the K predicted beams, or it may be a different message from the message carrying the measured signal quality values ​​of the K predicted beams; this embodiment of the application does not limit this.

[0155] The following describes three possible ways in which the terminal device sends a first message to the network device based on the measured signal quality values ​​of these K predicted beams, which satisfy the event used to instruct the AI ​​model to perform performance monitoring:

[0156] Method 1: The terminal device sends a first message to the network device based on the maximum measured value of the signal quality of these K predicted beams satisfying the event used to instruct the artificial intelligence (AI) model to perform performance monitoring.

[0157] After obtaining the measured values ​​of the signal quality of these K predicted beams, the terminal device can determine the maximum measured value from these measured values. Then, based on the maximum measured value of the signal quality of these K predicted beams, the terminal device determines whether the event used to instruct the artificial intelligence (AI) model to perform performance monitoring is met.

[0158] For example, the K predicted beams are beam 1, beam 3 and beam 5 determined in step 503 above, and their corresponding measured RSRP values ​​are -80dBm, -60.7dBm and -57.6dBm, respectively. Then the maximum measured value of the signal quality of the K predicted beams is -57.6dBm.

[0159] In the first example: the event is that the M1 largest measured values ​​of the signal quality obtained from M1 consecutive measurements of K predicted beams are all less than the first threshold, where M1 is a positive integer.

[0160] Specifically, the terminal device performs the following operations M1 times consecutively to obtain M1 maximum measured values: When the second beam set is scanned, the terminal device measures the measurement results of the second beam set, inputs the measurement results of the second beam set into the AI ​​model, and the AI ​​model outputs the predicted measurement results of the third beam set; based on the predicted measurement results of the third beam set, the terminal device determines K predicted beams in the first beam set; when the K predicted beams are scanned, the terminal device measures the measured values ​​of the signal quality of the K predicted beams; based on the measured values ​​of the signal quality of the K predicted beams, the terminal device determines the maximum measured value of the signal quality of the K predicted beams.

[0161] If all M1 maximum measured values ​​obtained by the terminal device are less than the first threshold, it indicates that the performance of the current AI model may be poor, leading to inaccurate AI model predictions. Consequently, the beam with the best signal quality in the third beam set is not selected as one of the K predicted beams for M1 consecutive times. Therefore, the terminal device can send a first message to the network device, thereby promptly triggering the network device to instruct the terminal device to monitor the performance of the AI ​​model.

[0162] For example, a terminal device can determine whether an event has been satisfied by maintaining a counter:

[0163] Taking RSRP (Signal Quality Ratio) as an example, the initial value of counter is 0. If the maximum measured value of RSRP for the K predicted beams determined by the terminal device is less than the first threshold, counter is incremented by 1; if the maximum measured value of RSRP for the K predicted beams determined by the terminal device is greater than or equal to the first threshold, counter is reset to 0. Optionally, the terminal device can reset counter to 0 after sending the first message to the network device.

[0164] For example, M1 is 3, and the first threshold is -75dBm. If the maximum measured signal quality of the K predicted beams determined by the terminal device in the first execution of the above operation is -80dBm, then counter = 1. If the maximum measured signal quality of the K predicted beams determined by the terminal device in the second execution of the above operation is -70dBm, then counter = 0. If the maximum measured signal quality of the K predicted beams determined by the terminal device in the third execution of the above operation is -78dBm, then counter = 1. If the maximum measured signal quality of the K predicted beams determined by the terminal device in the fourth execution of the above operation is -81dBm, then counter = 2. If the maximum measured signal quality of the K predicted beams determined by the terminal device in the fifth execution of the above operation is -100dBm, then counter = 3. At this time, the value of counter is the same as the value of M1, and the terminal device sends the first message to the network device.

[0165] In the second example: the event is that the first proportion is greater than the second threshold. Here, the first proportion is the percentage of the largest measured value that is less than the third threshold among the M2 largest measured values ​​of signal quality obtained from M2 consecutive measurements of K predicted beams, where M2 is a positive integer.

[0166] Specifically, the terminal device performs the following operations M2 times consecutively to obtain M2 maximum measured values: When the second beam set is scanned, the terminal device measures the measurement results of the second beam set, inputs the measurement results of the second beam set into the AI ​​model, and the AI ​​model outputs the predicted measurement results of the third beam set; based on the predicted measurement results of the third beam set, the terminal device determines K predicted beams in the first beam set; when the K predicted beams are scanned, the terminal device measures the measured values ​​of the signal quality of the K predicted beams; based on the measured values ​​of the signal quality of the K predicted beams, the terminal device determines the maximum measured value of the signal quality of the K predicted beams.

[0167] If the proportion of the maximum measured values ​​less than the third threshold among the M2 maximum measured values ​​obtained by the terminal device is greater than the second threshold, it indicates that the performance of the current AI model may be poor, leading to inaccurate AI model predictions. Consequently, the beam with the best signal quality in the third beam set may not be selected as one of the K predicted beams for M2 consecutive times. Therefore, the terminal device can send a first message to the network device, thereby promptly triggering the network device to instruct the terminal device to monitor the performance of the AI ​​model.

[0168] For example, M2 is 5, the third threshold is -75dBm, and the second threshold is 60%. If the terminal device performs the above operation 5 times consecutively and obtains 5 maximum measured values ​​of the signal quality of the K predicted beams as -75dBm, -80dBm, -82dBm, -78dBm, and -81dBm respectively, the proportion of the maximum measured value less than -75dBm is 80%, and 80% is greater than 60%, then the terminal device sends the first message to the network device.

[0169] Optionally, the two examples in Method 1 above determine whether the event is satisfied based on the maximum measured value of the signal quality of the K predicted beams. However, in specific implementation, the above two examples can also be used to determine whether the event is satisfied based on the top A largest measured values ​​of the signal quality of the K predicted beams, where A is a positive integer greater than or equal to 2. This application does not limit this.

[0170] Method 2: The terminal device determines the events that satisfy the requirements for instructing the artificial intelligence (AI) model to perform performance monitoring based on the first beam and the second beam among the K predicted beams.

[0171] In this context, the second beam is the beam with the highest measured signal quality among the K predicted beams, and the first beam is the beam with the highest predicted signal quality in the third beam set, or the beam with the highest predicted probability among the optimal beams in the third beam set. It can be understood that the first beam is also the beam with the highest predicted signal quality among the K predicted beams, or the beam with the highest predicted probability among the optimal beams in the K predicted beams.

[0172] For example, the third beam set consists of beams 1 to 7 from step 503 above. Beams 1 to 7 are sorted from largest to smallest according to the predicted signal quality value or predicted probability, resulting in beam 1, beam 3, beam 5, beam 2, beam 6, beam 7, and beam 4. If K is 3, then the three predicted beams are beam 1, beam 3, and beam 5. If the measured RSRP values ​​of beam 1, beam 3, and beam 5 are -80dBm, -60.7dBm, and -57.6dBm respectively, then the beam with the largest measured signal quality value among the K predicted beams is beam 5.

[0173] In the first example: the event is that the second proportion is less than the fourth threshold; where the second proportion is the proportion of events where the first beam and the second beam are the same for M3 consecutive determinations, and M3 is a positive integer.

[0174] Specifically, the terminal device performs the following operations M3 times consecutively to determine the first and second beams corresponding to each of the M3 operations: When scanning the second beam set, the terminal device measures the measurement results of the second beam set, inputs the measurement results of the second beam set into the AI ​​model, and the AI ​​model outputs the predicted measurement results of the third beam set; based on the predicted measurement results of the third beam set, the terminal device determines K predicted beams in the first beam set and the first beam; when scanning the K predicted beams, the terminal device measures the measured values ​​of the signal quality of the K predicted beams; based on the measured values ​​of the signal quality of the K predicted beams, the terminal device determines the second beam among the K predicted beams.

[0175] Then, the terminal device can determine the ratio of the number of events where the first and second beams are the same to M3 (i.e., the second ratio) based on the first and second beams corresponding to each of the M3 events. If the second ratio is less than the fourth threshold, it means that the beam with the best signal quality or the optimal beam predicted by the AI ​​model has a small probability of being the beam with the best signal quality or the optimal beam in the actual test. In other words, the AI ​​model's predictions are less accurate, indicating that the AI ​​model's performance may be poor. In this case, the terminal device can send a first message to the network device, thereby triggering the network device to instruct the terminal device to monitor the performance of the AI ​​model in a timely manner.

[0176] Optionally, the aforementioned consecutive M3 times can be the most recent consecutive M3 times.

[0177] For example, after determining the first beam and the second beam for any given instance, the terminal device can determine whether the event of the first beam and the second beam being the same is satisfied, and store the result of any given instance in a buffer queue. This buffer queue can store at most M3 consecutive judgment results obtained most recently, in a first-in-first-out order; when the buffer queue stores M3 consecutive judgment results each time, the terminal device can determine a second ratio and determine whether the second ratio is less than a fourth threshold.

[0178] For example, M3 is 5, and the fourth threshold is 60%. If the terminal device determines the first and second beams the same in the first instance, the first and second beams are different in the second instance, the first and second beams are the same in the third instance, the first and second beams are different in the fourth instance, and the first and second beams are the same in the fifth instance, then the buffer queue stores the five judgment results from the first to the fifth instance. The number of events where the first and second beams are the same in these five judgment results is 3, and the second proportion is 3 / 5 = 60%. Since 60% is not less than 60%, the terminal device continues to obtain the situation of the first and second beams determined in the sixth instance. If the first and second beams are different in the sixth instance, then the buffer queue stores the five judgment results from the second to the sixth instance. The number of events where the first and second beams are the same in these five judgment results is 2, and the second proportion is 2 / 5 = 40%. Since 40% is less than 60%, the terminal device sends the first message to the network device.

[0179] Optionally, the event can also be described as the proportion of events where the first beam and the second beam are different in M3 consecutive determinations is greater than the seventh threshold.

[0180] For example, the seventh threshold can be 40%. Based on the most recent five consecutive determinations of the first and second beams in the example above, the number of events where the first beam and the second beam are different can be determined to be three, and 3 / 5 = 60%. Since 60% exceeds 40%, the terminal device sends the first message to the network device.

[0181] In the second example: the event is that the average of the M4 first prediction errors determined consecutively M4 times is greater than the fifth threshold; where M4 is a positive integer, and the first prediction error is determined based on the measured values ​​of the signal quality of the first beam and the measured values ​​of the signal quality of the second beam.

[0182] Specifically, the terminal device performs the following operations M4 times consecutively to obtain M4 first prediction errors: When the second beam set is scanned, the terminal device measures the measurement results of the second beam set, inputs the measurement results of the second beam set into the AI ​​model, and the AI ​​model outputs the predicted measurement results of the third beam set; based on the predicted measurement results of the third beam set, the terminal device determines K predicted beams in the first beam set and the first beam; when the K predicted beams are scanned, the terminal device measures the measured values ​​of the signal quality of the K predicted beams; based on the measured values ​​of the signal quality of the K predicted beams, the terminal device determines the second beam among the K predicted beams; based on the measured values ​​of the signal quality of the first beam and the measured values ​​of the signal quality of the second beam, the first prediction error is determined.

[0183] Then, the terminal device can determine the average of the M4 first prediction errors. If the average is greater than the fifth threshold, it indicates that the AI ​​model's prediction error is large and its performance may be poor. In this case, the terminal device can send a first message to the network device. This facilitates timely triggering of the network device to instruct the terminal device to monitor the AI ​​model's performance, thereby ensuring the effectiveness of AI-assisted beam management.

[0184] Optionally, the consecutive M4 times can be the most recent consecutive M4 times. For details, please refer to the description of the most recent consecutive M3 times above.

[0185] For example, the first prediction error can be the absolute value of the difference between the measured signal quality of the first beam and the measured signal quality of the second beam. Alternatively, the first prediction error can be the square of the difference between the measured signal quality of the first beam and the measured signal quality of the second beam, but is not limited to this.

[0186] For example, if the measured signal quality of the first beam is x, and the measured signal quality of the second beam is y, the first prediction error can be |xy| or (xy). 2 .

[0187] Optionally, the two examples in Method 2 above are based on the beam with the largest measured signal quality among the K predicted beams, and the beam with the largest predicted signal quality, to determine whether the event is satisfied. However, in specific implementation, the method in the two examples above can also be used to determine the event based on the beam with the largest measured signal quality among the K predicted beams, and the beam with the largest predicted signal quality among the K predicted beams, where A is a positive integer greater than or equal to 2. This application does not limit this.

[0188] Method 3: The terminal device determines the second prediction error based on the measured and predicted values ​​of the signal quality of the K predicted beams; if the second prediction error satisfies the event used to instruct the artificial intelligence (AI) model to perform performance monitoring, it sends the first message to the network device.

[0189] For example, the second prediction error can be the average of the absolute values ​​of the differences between the measured and predicted signal quality values ​​of the K predicted beams. Alternatively, the second prediction error can be the sum of the absolute values ​​of the differences between the measured and predicted signal quality values ​​of the K predicted beams. Alternatively, the second prediction error can be the average of the squares of the differences between the measured and predicted signal quality values ​​of the K predicted beams. Or, the second prediction error can be the average of the squares of the differences between the measured and predicted signal quality values ​​of the K predicted beams. However, it is not limited to these examples.

[0190] Taking RSRP as an example, the predicted signal quality of K prediction beams is... The measured signal quality values ​​of the K predicted beams are The second prediction error can then be determined using any one of the formulas 1 to 4 below.

[0191]

[0192] In the first example: the event is that the average of the M5 second prediction errors determined M5 times consecutively is greater than the sixth threshold, where M5 is a positive integer.

[0193] Specifically, the terminal device performs the following operations M5 times consecutively to obtain M5 second prediction errors: When the second beam set is scanned, the terminal device measures the measurement results of the second beam set, inputs the measurement results of the second beam set into the AI ​​model, and the AI ​​model outputs the predicted measurement results of the third beam set; the terminal device determines K predicted beams in the first beam set based on the predicted measurement results of the third beam set; when the K predicted beams are scanned, the terminal device measures the measured values ​​of the signal quality of the K predicted beams; the terminal device determines the second prediction error based on the measured values ​​of the signal quality of the K predicted beams and the predicted values.

[0194] If the average of the M5 second prediction errors obtained by the terminal device is greater than the sixth threshold, the prediction error of the AI ​​model is relatively large, and the performance of the AI ​​model may be poor. In this case, the terminal device can send a first message to the network device. This can help to promptly trigger the network device to instruct the terminal device to monitor the performance of the AI ​​model, thereby ensuring the effectiveness of AI-assisted beam management.

[0195] Optionally, the M5 consecutive times can be the most recent M5 consecutive times. For details, please refer to the description of the most recent M3 consecutive times above.

[0196] Based on the above description, the terminal device can determine whether the aforementioned event is met based on the measurement results of K predicted beams in the first beam set. This process is equivalent to the terminal device performing pre-monitoring of the AI ​​model's performance. When the measurement results of the K predicted beams in the first beam set meet the aforementioned event, it is equivalent to the terminal device pre-monitoring that the AI ​​model's performance is poor, and the terminal device sends a first message to the network device. Conversely, when the measurement results of the K predicted beams in the first beam set do not meet the aforementioned event, it is equivalent to the terminal device pre-monitoring that the AI ​​model's performance is good, and the terminal device does not need to send a first message to the network device. Furthermore, the pre-monitoring is based on K predicted beams, while the monitoring in the aforementioned "AI model performance monitoring method" is based on a larger number of beams, such as the third beam set in this application embodiment. Therefore, it can also reduce the data processing volume of the terminal device and improve its processing efficiency.

[0197] 702. The network device sends a second message to the terminal device, which is used to instruct the performance of the AI ​​model to be monitored.

[0198] In this embodiment, after receiving the first message in step 701, the network device can send a second message to the terminal device to instruct the terminal device to monitor the performance of the AI ​​model. Conversely, if the network device does not receive the first message in step 701, the network device does not need to instruct the terminal device to monitor the performance of the AI ​​model. In this way, the network device can determine whether it needs to instruct the terminal device to monitor the performance of the AI ​​model based on whether it has received the first message, rather than directly instructing the terminal device to periodically monitor the performance of the AI ​​model, thereby reducing signaling and resource overhead during periodic monitoring.

[0199] For example, the second message may include, but is not limited to, one or more of the following: monitoring duration, monitoring time offset, and reporting method of monitoring results. The monitoring time offset indicates the time when the terminal device begins monitoring, and the reporting method of monitoring results may be event-triggered reporting.

[0200] Furthermore, after receiving the second message, the terminal device can monitor the performance of the AI ​​model by referring to the content corresponding to the "performance monitoring method of the AI ​​model" mentioned above. For example, after receiving the second message, the terminal device, while measuring the measurement result of the second beam set, also measures the measurement result of the third beam set; then, the terminal device inputs the measurement result of the second beam set into the AI ​​model to predict the measurement result of the third beam set; based on the predicted measurement result of the third beam set and the measured measurement result of the third beam set, the monitoring result is obtained and reported to the network device.

[0201] Furthermore, if the monitoring results obtained by the terminal device indicate that the AI ​​model's performance is poor, the network device can send a management command for the AI ​​model to the terminal device, and / or, a command to fall back to traditional beam management. For example, the management command for the AI ​​model might be a deactivation command. After receiving the deactivation command and the command to fall back to traditional beam management, the terminal device obtains the beam measurement results using traditional beam management.

[0202] exist Figure 7 In the described method, when the measurement results based on the first beam set meet the performance monitoring event of the AI ​​model, the terminal device sends a first message to the network device. After receiving the first message, the network device sends a second message to the terminal device, thereby triggering the terminal device to monitor the performance of the AI ​​model. In this way, the terminal device does not need to monitor the performance of the AI ​​model in a periodic manner, which can effectively reduce the signaling and resource overhead caused by frequent interactions between the terminal device and the network device.

[0203] Figure 8 A schematic diagram of the software structure of a communication device according to an embodiment of this application is shown. Figure 8 The communication device 800 shown may include a transceiver unit 801 and a processing unit 802.

[0204] In one example, Figure 8 The communication device 800 shown can be used to perform some or all of the functions of the terminal device in the above embodiments. The communication device 800 can be the terminal device itself, a device within the terminal device, or a device compatible with the terminal device. Furthermore, the communication device 800 can also be a chip system.

[0205] The transceiver unit 801 is configured to: send a first message to the network device based on the measurement results of the first beam set satisfying an event used to instruct the AI ​​model to perform performance monitoring; the first message is used to request performance monitoring of the AI ​​model; and / or, the first message is used to indicate that the measurement results of the first beam set satisfy an event, and the AI ​​model is used to predict the measurement results of the third beam set based on the measurement results of the second beam set, the second beam set being the same as or different from the third beam set, the first beam set including K predicted beams from the third beam set, the K predicted beams dynamically changing with the measurement results of the second beam set, where K is a positive integer; and receive a second message sent by the network device, the second message being used to instruct performance monitoring of the AI ​​model.

[0206] In one possible implementation, the measurement results of the second beam set include measured values ​​of the signal quality of the beams in the second beam set; the measurement results of the predicted third beam set include predicted values ​​of the signal quality of the beams in the third beam set, and / or, the predicted probability that the third beam set is the optimal beam.

[0207] In one possible implementation, the K predicted beams include the K largest predicted beams in the third beam set in terms of signal quality, or the K predicted beams include the K largest predicted beams in the third beam set in terms of the best predicted probability.

[0208] In one possible implementation, the measurement results of the first beam set include the measured values ​​of the signal quality of K predicted beams; the transceiver unit 801 sends a first message to the network device based on the measurement results of the first beam set satisfying an event used to instruct the artificial intelligence (AI) model to perform performance monitoring, specifically including: sending a first message to the network device based on the maximum measured value of the signal quality of the K predicted beams satisfying an event used to instruct the artificial intelligence (AI) model to perform performance monitoring.

[0209] In one possible implementation, the above event is that the M1 maximum measured values ​​of the signal quality obtained from M1 consecutive measurements of K predicted beams are all less than a first threshold, where M1 is a positive integer.

[0210] In one possible implementation, the above event is that the first proportion is greater than the second threshold; wherein, the first proportion is the proportion of the largest measured value less than the third threshold among the M2 largest measured values ​​of signal quality obtained from M2 consecutive measurements of K predicted beams, and M2 is a positive integer.

[0211] In one possible implementation, the measurement results of the first beam set include the measured values ​​of the signal quality of K predicted beams; the transceiver unit 801 sends a first message to the network device based on the measurement results of the first beam set satisfying an event for instructing the AI ​​model to perform performance monitoring, specifically including: determining an event for instructing the AI ​​model to perform performance monitoring based on the first beam and the second beam among the K predicted beams, wherein the second beam is the beam with the largest measured signal quality value among the K predicted beams, and the first beam is the beam with the largest predicted signal quality value in the third beam set, or the first beam is the beam with the highest predicted probability of the optimal beam in the third beam set; and sending the first message to the network device.

[0212] In one possible implementation, the above event is a second proportion less than a fourth threshold; wherein the second proportion is the proportion of events in which the first beam and the second beam are the same for M3 consecutive determinations, and M3 is a positive integer.

[0213] In one possible implementation, the aforementioned event is that the average of the M4 first prediction errors determined consecutively M4 times is greater than a fifth threshold; where M4 is a positive integer, and the first prediction error is determined based on the measured values ​​of the signal quality of the first beam and the measured values ​​of the signal quality of the second beam.

[0214] In one possible implementation, the measurement results of the first beam set include the measured values ​​of the signal quality of the beams in the first beam set; the transceiver unit 801 sends a first message to the network device based on the measurement results of the first beam set satisfying the performance monitoring event of the artificial intelligence (AI) model, including: calling the processing unit 802 to determine a second prediction error based on the measured and predicted values ​​of the signal quality of K predicted beams; and sending the first message to the network device based on the second prediction error satisfying the performance monitoring event of the AI ​​model.

[0215] In one possible implementation, the above event is that the average of the second prediction errors determined M5 times consecutively is greater than the sixth threshold, where M5 is a positive integer.

[0216] In another example, Figure 8 The communication device 800 shown can be used to perform some or all of the functions of the network device in the above embodiments. The communication device 800 can be a network device, a device within a network device, or a device compatible with a network device. Furthermore, the communication device 800 can also be a chip system.

[0217] The transceiver unit 801 is configured to: receive a first message sent by the terminal device, the first message being used to request performance monitoring of the AI ​​model, and / or, the first message being used to indicate that the measurement results of the first beam set meet the event used to instruct the AI ​​model to perform performance monitoring, the AI ​​model being used to predict the measurement results of the third beam set based on the measurement results of the second beam set, the second beam set being the same as or different from the third beam set, the first beam set including K predicted beams in the third beam set, the K predicted beams dynamically changing with the measurement results of the second beam set, K being a positive integer; and send a second message to the terminal device, the second message being used to instruct performance monitoring of the AI ​​model.

[0218] It should be noted that the specific implementation method and beneficial effects of the operation performed by the communication device 800 can be found in the corresponding description in the above method embodiments, and will not be repeated here.

[0219] Figure 9A schematic diagram of the hardware structure of a communication device is shown. The communication device 900 can be a terminal device or a network device as described above. The communication device 900 may include: a processor, an external memory interface, internal memory, a Universal Serial Bus (USB) interface, a charging management module, a power management module, a battery, antenna 1, antenna 2, a mobile communication module, a wireless communication module, a sensor module, buttons, a motor, an indicator, a camera, a display screen, and a SIM card slot, etc. The audio module may include a speaker, a receiver, a microphone, an earphone jack, etc. The sensor module may include a pressure sensor, a gyroscope sensor, a barometric pressure sensor, a magnetic sensor, an accelerometer, a distance sensor, a proximity sensor, a fingerprint sensor, a temperature sensor, a touch sensor, an ambient light sensor, a bone conduction sensor, etc.

[0220] It is understood that the structure illustrated in the embodiments of this application does not constitute a specific limitation on the communication device 900. In other embodiments, the communication device 900 may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0221] The processor may include one or more processing units, such as an application processor (AP), a modem (also known as a baseband processor), a graphics processing unit (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors. The processor is the nerve center and command center of the communication device 900. The controller generates operation control signals based on instruction opcodes and timing signals to control instruction fetching and execution.

[0222] The wireless communication function of the communication device 900 can be implemented through antenna 1, antenna 2, a mobile communication module, a wireless communication module, and a modem, etc. In some embodiments, antenna 1 of the communication device 900 is coupled to the mobile communication module, and antenna 2 is coupled to the wireless communication module, so that the communication device 900 can communicate with other communication devices through wireless communication technology.

[0223] In addition, an operating system runs on top of the aforementioned components. Examples include the iOS operating system, the Android open-source operating system, and the Windows operating system.

[0224] The operating system of the communication device 900 can adopt a layered architecture, event-driven architecture, microkernel architecture, microservice architecture, or cloud architecture. This application embodiment uses the layered architecture Android system as an example to exemplify the hardware and software structure of the communication device 900. It should be noted that although this application embodiment uses the Android system as an example, its basic principles are equally applicable to terminals based on operating systems such as iOS or Windows.

[0225] Furthermore, some embodiments of this application provide a communication device, including: one or more processors and a memory; the memory is used to store computer programs / instructions, which, when executed by one or more processors, cause the communication device to perform the above-described method.

[0226] Some embodiments of this application provide a chip system applied to a communication device. The chip system includes at least one processor and an interface for receiving computer programs / instructions and transmitting them to the at least one processor. The at least one processor executes instructions to cause the communication device to perform the above-described method. The chip system may be a modem, or a system-on-a-chip (SoC) including a modem, and the above-described method may be implemented by a modem.

[0227] This application also provides a computer storage medium storing a computer program / instruction, which, when run on a processor, enables the implementation of the method flow described in the above method embodiments.

[0228] This application also provides a computer program product, which, when run on a computer, enables the implementation of the method flow described in the above method embodiments.

[0229] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some operations can be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0230] The descriptions of the various embodiments provided in this application can be referenced mutually. Each embodiment has its own emphasis, and parts not described in detail in a certain embodiment can be referred to the relevant descriptions of other embodiments. For the sake of convenience and brevity, for example, the functions and operations of the various devices and equipment provided in the embodiments of this application can be referred to the relevant descriptions of the method embodiments of this application. The method embodiments and the device embodiments can also be referenced, combined or cited from each other.

[0231] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer programs / instructions. When the computer program / instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer program / instructions can be stored in a computer storage medium or transferred from one computer storage medium to another. For example, a computer program / instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive, SSD), etc.

[0232] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A communication method, characterized in that, The method includes: Based on the measurement results of the first beam set satisfying an event used to instruct the AI ​​model to perform performance monitoring, a first message is sent to the network device. The first message is used to request performance monitoring of the AI ​​model, and / or, the first message is used to indicate that the measurement results of the first beam set satisfy the event. The AI ​​model is used to predict the measurement results of a third beam set based on the measurement results of a second beam set. The second beam set may be the same as or different from the third beam set. The first beam set includes K predicted beams from the third beam set. The K predicted beams change dynamically with the measurement results of the second beam set, where K is a positive integer. The system receives a second message from the network device, the second message being used to instruct the monitoring of the performance of the AI ​​model.

2. The method according to claim 1, characterized in that, The measurement results of the second beam set include the measured values ​​of the signal quality of the beams in the second beam set; The predicted measurement results of the third beam set include the predicted signal quality of the beams in the third beam set, and / or the predicted probability that the beams in the third beam set are the optimal beams.

3. The method according to claim 1 or 2, characterized in that, The K predicted beams include the K largest predicted values ​​of signal quality in the third beam set, or the K predicted beams include the K largest predicted probabilities of the optimal beam in the third beam set.

4. The method according to any one of claims 1-3, characterized in that, The measurement results of the first beam set include the measured values ​​of the signal quality of the K predicted beams; The measurement results based on the first beam set satisfy an event used to instruct the AI ​​model to perform performance monitoring, and a first message is sent to the network device, including: Based on the maximum measured value of the signal quality of the K predicted beams satisfying an event used to instruct the artificial intelligence (AI) model to perform performance monitoring, the first message is sent to the network device.

5. The method according to claim 4, characterized in that, The event is defined as the M1 maximum measured values ​​of the signal quality obtained from M1 consecutive measurements of the K predicted beams being less than a first threshold, where M1 is a positive integer.

6. The method according to claim 4, characterized in that, The event is defined as a first proportion being greater than a second threshold; Wherein, the first proportion is the proportion of the largest measured value less than the third threshold among the M2 largest measured values ​​of signal quality obtained from M2 consecutive measurements of the K predicted beams, and M2 is a positive integer.

7. The method according to claim 2 or 3, characterized in that, The measurement results of the first beam set include the measured values ​​of the signal quality of the K predicted beams; The measurement results based on the first beam set satisfy an event used to instruct the AI ​​model to perform performance monitoring, and a first message is sent to the network device, including: Based on the first beam and the second beam among the K predicted beams, an event is determined that satisfies the requirement for instructing the AI ​​model to perform performance monitoring. The second beam is the beam with the largest measured signal quality among the K predicted beams, and the first beam is the beam with the largest predicted signal quality in the third beam set. Alternatively, the first beam is the beam with the highest predicted probability among the optimal beams in the third beam set. The first message is sent to the network device.

8. The method according to claim 7, characterized in that, The event is when the second proportion is less than the fourth threshold; Wherein, the second ratio is the proportion of events in which the first beam and the second beam are the same for M3 consecutive determinations, and M3 is a positive integer.

9. The method according to claim 7, characterized in that, The event is that the average of the M4 first prediction errors determined M4 times consecutively is greater than the fifth threshold. Where M4 is a positive integer, and the first prediction error is determined based on the measured values ​​of the signal quality of the first beam and the measured values ​​of the signal quality of the second beam.

10. The method according to claim 2 or 3, characterized in that, The measurement results of the first beam set include the measured values ​​of the signal quality of the beams in the first beam set; The measurement results based on the first beamset satisfy the performance monitoring event of the artificial intelligence (AI) model, and a first message is sent to the network device, including: The second prediction error is determined based on the measured and predicted values ​​of the signal quality of the K predicted beams. Based on the event that the second prediction error satisfies the performance monitoring of the AI ​​model, the first message is sent to the network device.

11. The method according to claim 10, characterized in that, The event is defined as the average of the M5 second prediction errors determined consecutively M5 times being greater than the sixth threshold, where M5 is a positive integer.

12. A communication method, characterized in that, The method includes: The system receives a first message from a terminal device. The first message is used to request performance monitoring of the AI ​​model, and / or the first message is used to indicate that the measurement results of a first beam set meet an event used to instruct the AI ​​model to perform performance monitoring. The AI ​​model is used to predict the measurement results of a third beam set based on the measurement results of a second beam set. The second beam set may be the same as or different from the third beam set. The first beam set includes K predicted beams from the third beam set. The K predicted beams change dynamically with the measurement results of the second beam set, where K is a positive integer. A second message is sent to the terminal device, the second message being used to instruct the performance of the AI ​​model to be monitored.

13. A communication device, characterized in that, include: One or more processors, one or more memories; wherein the one or more memories are coupled to one or more processors, the one or more memories being used to store computer instructions that, when the one or more processors execute the computer instructions, cause the communication device to perform the method as described in any one of claims 1-11 or 12.

14. A chip system, characterized in that, The chip system includes at least one processor and an interface for receiving computer instructions and transmitting them to the at least one processor; the at least one processor executes the computer instructions to cause the communication device to perform the method as described in any one of claims 1-11 or 12.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, cause the communication device to perform the method as described in any one of claims 1-11 or 12.

16. A computer program product, characterized in that, The computer program product includes computer instructions that, when executed by a processor, cause the communication device to perform the method as described in any one of claims 1-11 or 12.

Images