Information reporting method and apparatus, device, and storage medium

Abstract

The present disclosure provides an information reporting method and apparatus, a device, and a storage medium. The method comprises: sending a performance monitoring report to a network device, wherein the performance monitoring report comprises a performance measurement result and / or a performance monitoring output result of an inference model or an inference function; the performance monitoring report is sent in association with a first inference report, and the first inference report is an inference report at at least one inference reporting moment associated with a monitoring reporting moment of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently, and the inference report comprises predicted channel state information (CSI) outputted by the inference model or the inference function.

Description

Information reporting methods, devices, equipment and storage media

[0001] This disclosure claims priority to Chinese Patent Application No. 202510133915.8, filed with the Chinese Patent Office on February 6, 2025, entitled "Information Reporting Method, Apparatus, Device and Storage Medium", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This disclosure relates to the field of communication technology, and in particular to an information reporting method, apparatus, device and storage medium. Background Technology

[0003] In the Channel State Information (CSI) reporting of related technologies, network devices configure reference signal (RS) resources for measurement and CSI reporting configuration for terminals. When CSI reporting is triggered, the terminal performs measurements and calculates the reporting amount based on the configured reference signal resources, and reports the reporting amount according to the CSI reporting configuration.

[0004] For CSI prediction and reporting, inference prediction of CSI at future time points is performed, and the predicted CSI is reported, which is also known as inference report reporting. However, in related technologies, the traditional CSI prediction method is used for CSI prediction reporting, without introducing inference models (such as artificial intelligence (AI) / machine learning (ML) models) or inference functions (such as AI functions / ML functions). As the inference model / inference function is applied to CSI reporting, the performance of the inference model / inference function directly affects the reporting results. Therefore, it is necessary to monitor the performance of the inference model / inference function. However, the current relevant protocols do not involve how to perform performance monitoring and reporting. That is to say, there is currently no relevant solution for how to perform performance monitoring and reporting. Summary of the Invention

[0005] This disclosure provides an information reporting method, apparatus, device, and storage medium to achieve performance monitoring and reporting of inference models / inference functions.

[0006] This disclosure provides an information reporting method applied to a terminal device, the method comprising:

[0007] Send a performance monitoring report to the network device, the performance monitoring report including the performance measurement results of the inference model or inference function and / or the performance monitoring output results;

[0008] The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0009] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0010] This disclosure provides an information reporting method applied to a network device, including:

[0011] Receive a performance monitoring report sent by the terminal device, the performance monitoring report including performance measurement results of the inference model or inference function and / or performance monitoring output results;

[0012] Wherein, the performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0013] The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function.

[0014] This disclosure provides a communication device, including: a memory, a transceiver, and a processor. The communication device may include a terminal device, wherein:

[0015] The memory is used to store computer programs; the transceiver is used to send and receive data under the control of the processor; the processor is used to read the computer programs in the memory and perform the following operations:

[0016] Send a performance monitoring report to the network device, the performance monitoring report including the performance measurement results of the inference model or inference function and / or the performance monitoring output results;

[0017] The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0018] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0019] This disclosure provides a communication device, including: a memory, a transceiver, and a processor. The communication device may include a network device, wherein:

[0020] The memory is used to store computer programs; the transceiver is used to send and receive data under the control of the processor; the processor is used to read the computer programs in the memory and perform the following operations:

[0021] Receive a performance monitoring report sent by the terminal device, the performance monitoring report including performance measurement results of the inference model or inference function and / or performance monitoring output results;

[0022] Wherein, the performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0023] The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function.

[0024] This disclosure provides a communication device, which includes a terminal device, comprising:

[0025] The first sending module is used to send a performance monitoring report to the network device. The performance monitoring report includes the performance measurement results of the inference model or inference function and / or the performance monitoring output results.

[0026] The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0027] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0028] This disclosure provides a communication device, which includes a network device, comprising:

[0029] The first receiving module is used to receive a performance monitoring report sent by the terminal device. The performance monitoring report includes performance measurement results of the inference model or inference function and / or performance monitoring output results.

[0030] Wherein, the performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0031] The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function.

[0032] This disclosure provides a processor-readable storage medium storing a computer program that causes the processor to execute the information reporting method provided in this disclosure.

[0033] This disclosure provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the information reporting method described above.

[0034] In the information reporting process of this embodiment, it is necessary to report not only the inference report including the prediction of CSI, but also the performance monitoring report of the inference model or inference function. The performance monitoring report can be reported in association with the first inference report, which is the inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. Alternatively, the performance monitoring report can be reported independently of the inference report. That is, this embodiment can realize the performance monitoring report being reported in association with the first inference report, or the performance monitoring report being reported independently of the inference report, thus realizing the performance monitoring reporting of the inference model / inference function. Attached Figure Description

[0035] Figure 1 is a schematic diagram of the network architecture applicable to the present disclosure.

[0036] Figure 2 is a timing diagram of a DCI-triggered CSI prediction reporting provided in an embodiment of this disclosure;

[0037] Figure 3 is a schematic diagram of a traditional CSI prediction;

[0038] Figure 4 is a schematic diagram of CSI prediction based on AI / ML provided in an embodiment of this disclosure;

[0039] Figure 5 is one of the performance monitoring schematic diagrams provided in the embodiments of this disclosure;

[0040] Figure 6 is a second schematic diagram of performance monitoring provided in an embodiment of this disclosure;

[0041] Figure 7 is a third schematic diagram of performance monitoring provided in an embodiment of this disclosure;

[0042] Figure 8 is a schematic diagram of CSI reporting in the case of a non-periodic / periodic / semi-persistent reference signal and a non-periodic PUSCH inference report provided by an embodiment of this disclosure;

[0043] Figure 9 is a schematic diagram of CSI reporting in the case of a periodic / semi-persistent reference signal and a semi-persistent PUSCH inference report provided in an embodiment of this disclosure;

[0044] Figure 10 is a schematic diagram of an information reporting method provided in an embodiment of this disclosure;

[0045] Figure 11 is a second schematic diagram of an information reporting method provided in an embodiment of this disclosure;

[0046] Figure 12 is a schematic diagram of the triggering / activation of non-periodic inference reports and semi-persistent monitoring reports and CSI reporting provided in an embodiment of this disclosure;

[0047] Figure 13 is a schematic diagram of the triggering / activation of a semi-persistent inference report and a non-periodic monitoring report and CSI reporting provided in an embodiment of this disclosure, wherein the non-periodic monitoring report is associated with one inference reporting moment;

[0048] Figure 14 is a schematic diagram of the triggering / activation of a semi-continuous inference report and a non-periodic monitoring report and CSI reporting provided in an embodiment of this disclosure, wherein the non-periodic monitoring report is associated with multiple inference reporting times;

[0049] Figure 15 is a schematic diagram of the activation and CSI reporting of a semi-persistent inference report and a semi-persistent monitoring report provided in an embodiment of this disclosure;

[0050] Figure 16 is a structural diagram of a communication device provided in an embodiment of this disclosure;

[0051] Figure 17 is a structural diagram of another communication device provided in an embodiment of this disclosure;

[0052] Figure 18 is a schematic diagram of a module of an information reporting device provided in an embodiment of this disclosure;

[0053] Figure 19 is a schematic diagram of another information reporting device provided in an embodiment of this disclosure. Detailed Implementation

[0054] To make the technical problems, technical solutions and advantages to be solved by this disclosure clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0055] In this disclosure, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0056] In this disclosure, the term "multiple" refers to two or more, and other quantifiers are similar.

[0057] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this disclosure.

[0058] This disclosure provides an information reporting method, apparatus, device, and storage medium to achieve performance monitoring and reporting of inference models / inference functions.

[0059] The method and apparatus are based on the same concept of the application. Since the methods and apparatus solve problems in similar ways, the implementation of the apparatus and methods can refer to each other, and the repeated parts will not be described again.

[0060] The technical solutions provided in this disclosure are applicable to a variety of systems, especially 6th generation mobile communication technology (6G) systems. For example, applicable systems include Global System for Mobile Communication (GSM), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA) General Packet Radio Service (GPRS), Long Term Evolution (LTE), LTE Frequency Division Duplex (FDD), LTE Time Division Duplex (TDD), Long Term Evolution Advanced (LTE-A), Universal Mobile Telecommunications System (UMTS), Worldwide Interoperability for Microwave Access (WiMAX), 5th Generation Mobile Communication Technology (5G) New Radio (NR), and 6G systems. All of these systems include terminal equipment and network equipment. The system may also include a core network component, such as the Evolved Packet System (EPS) or the 5G system (5GS).

[0061] Please refer to Figure 1, which is a schematic diagram of the network architecture applicable to the present disclosure. As shown in Figure 1, it includes a terminal 11 and a network device 12.

[0062] The terminal involved in this disclosure can be a device that provides voice and / or data connectivity to a user, a handheld device with wireless connectivity, or other processing devices connected to a wireless modem. The name of the terminal device may differ in different systems; for example, in a 5G system, the terminal device may be called User Equipment (UE). The wireless terminal device can communicate with one or more core networks (CNs) via a Radio Access Network (RAN). The wireless terminal device can be a mobile terminal device, such as a mobile phone (or "cellular" phone) or a computer with a mobile terminal device, for example, a portable, pocket-sized, handheld, computer-embedded, or vehicle-mounted mobile device, which exchanges voice and / or data with the radio access network. Examples include Personal Communication Service (PCS) phones, cordless phones, Session Initiated Protocol (SIP) phones, Wireless Local Loop (WLL) stations, Personal Digital Assistants (PDAs), Redcap terminals, and Low Power Wide Area (LPWA) terminals. Wireless terminal equipment can also be referred to as a system, subscriber unit, subscriber station, mobile station, mobile terminal, remote station, access point, remote terminal, access terminal, user terminal, user agent, or user device; however, this disclosure does not limit the terminology used in this embodiment.

[0063] The network device disclosed in this embodiment may be a base station, which may include multiple cells providing services to terminals. Depending on the specific application, the base station may also be called an access point, or a device in the access network that communicates with the wireless terminal device through one or more sectors on the air interface, or other names. The network device may be used to exchange received air frames with Internet Protocol (IP) packets, acting as a router between the wireless terminal device and the rest of the access network, where the rest of the access network may include an Internet Protocol (IP) communication network. The network device may also coordinate the attribute management of the air interface. For example, the network equipment involved in this disclosure can be a base transceiver station (BTS) in a Global System for Mobile communications (GSM) or Code Division Multiple Access (CDMA) system, a NodeB in a wide-band Code Division Multiple Access (WCDMA) system, an evolved Node B (eNB or e-NodeB) in a long term evolution (LTE) system, a 5G base station (NR Node B, gNB) in a next generation system, a base station in 6G, a Home evolved Node B (HeNB), a relay node, a femto, a pico, etc., and is not limited in this disclosure. In some network structures, the network equipment may include centralized unit (CU) nodes and distributed unit (DU) nodes, and the centralized unit and distributed unit may be geographically separated.

[0064] Network devices and terminals can each use one or more antennas for Multiple Input Multiple Output (MIMO) transmission. MIMO transmission can be Single User MIMO (SU-MIMO) or Multiple User MIMO (MU-MIMO). Depending on the configuration and number of antenna combinations, MIMO transmission can be 2D MIMO, 3D MIMO, Full Dimension MIMO (FD-MIMO), or Massive MIMO. It can also be diversity transmission, pre-coded transmission, or beamforming transmission, etc.

[0065] This disclosure includes the following technical descriptions:

[0066] In CSI reporting of related technologies, network devices (e.g., base stations) configure reference signal (RS) resources for measurement and CSI reporting configuration for terminal devices (i.e., UEs). When CSI reporting is triggered, the terminal performs measurements and calculates the reporting amount based on the configured reference signal, and reports the reporting amount according to the CSI reporting configuration. The timing diagram for CSI prediction-based CSI reporting is shown in Figure 2.

[0067] The base station configures the measurement reference signal resource (CSI-RS) for CSI prediction reporting and the CSI reporting configuration for the terminal. When CSI reporting is triggered, the terminal performs measurements and calculates the reported amount based on the configured reference signal, and reports the reported amount at the configured CSI reporting time. The CSI reporting includes CSI predictions for N4 future consecutive slot intervals. The j-th predicted CSI is the CSI associated with the slot interval [n+δ+(j-1)·d,n+δ+j·d-1], j=1,…,N4, where n is the uplink slot for CSI reporting, and the offset value δ ∈ {-n} related to the prediction window. ref ,0,1,2},N4*d is the total length of the prediction window.

[0068] As shown in Figure 3, for CSI prediction reporting (traditional CSI prediction reporting) in related technologies, CSI prediction uses traditional algorithms, such as autoregression (AR) or Wiener filtering. The input of the traditional algorithm is the CSI of N4 future time slots predicted based on the results of CSI-RS resource measurements, which is then compressed using the R18 eType II Doppler codebook before being reported.

[0069] After the introduction of AI algorithms in 5G NR New Radio, the terminal can use an algorithm based on Artificial Intelligence Markup Language (AIML) to calculate the reported amount based on the results of the configured reference signal measurement. At this time, the reported amount calculated based on AIML is the result of inference using the corresponding AIML model, that is, CSI reporting is inference reporting based on a specific model.

[0070] As shown in Figure 4, for AI / ML-based CSI prediction reporting, the CSI prediction uses an AI / ML model. The input to the AI / ML model is the result obtained from CSI-RS resource measurement, and the model output is the predicted CSI for N4 future time slots, i.e., the model inference result.

[0071] However, each AI model has its applicable scope, and AI model performance monitoring refers to the need to monitor the model's performance when the AI ​​model's usage scenario differs from its training scenario or when the inherent relationship between input and output changes, in order to determine whether the current model is applicable.

[0072] Performance monitoring for single-sided model-based use cases includes, but is not limited to, AI / ML-based UE-side CSI prediction use cases, UE-side beam management use cases, and UE-side positioning use cases. Single-sided model performance monitoring has different types, mainly divided into terminal-side performance monitoring and network-side performance monitoring. Terminal-side performance monitoring can be further divided into terminal-reported performance monitoring output and terminal-reported performance metric, based on the amount of CSI reporting.

[0073] Taking AI / ML-based CSI prediction as an example, the performance monitoring of the terminal side and the reporting of performance monitoring output by the terminal are shown in Figure 5. The UE side AI / ML performance monitoring process involves the UE side calculating the performance metric and reporting the performance monitoring output. This performance monitoring output helps the network side make lifecycle management decisions, such as rollback, inference model / function rollback, etc.

[0074] Taking AI / ML-based CSI prediction as an example, network-side performance monitoring is shown in Figure 6. This is the network (NW) side AI / ML performance monitoring process, where the UE side reports the predicted CSI and / or the corresponding true value, the network side calculates performance metrics, and the network side makes lifecycle management decisions.

[0075] Taking AI / ML-based CSI prediction as an example, the performance monitoring of terminal-side performance monitoring and terminal reporting of performance metrics is shown in Figure 7. The process includes AI / ML performance monitoring on the UE side, calculation of performance metrics on the UE side, reporting of performance metrics on the UE side, and lifecycle management decisions on the network side.

[0076] If it is a beam management use case, the CSI-related reference signals in the above diagram can be replaced with the beam management-related reference signals corresponding to SetA and SetB. If it is a location use case, it can be replaced with location-related reference signals.

[0077] The AI / ML model for beam prediction use cases performs spatial beam prediction, specifically predicting a large number of beams (SetA) from a small number of beams (SetB), or predicting a narrow beam (SetA) from a wide beam (SetB). Another AI / ML model for beam prediction use cases performs temporal beam prediction, specifically predicting the optimal beam for a future time slot (SetA) from beam measurement results of several historical time slots (SetB).

[0078] The AI / ML model in the location use case serves to perform direct AI / ML location, specifically predicting the UE's location based on the channel's time-domain response, frequency-domain response, power delay distribution, and delay distribution. Another location use case uses the AI / ML model to assist in location, specifically predicting the time of arrival (ToA) and the channel's line-of-sight (LOS) / non-line-of-sight (NLOS) states based on the channel's time-domain response, frequency-domain response, power delay distribution, and delay distribution.

[0079] The performance monitoring in this solution refers to the terminal-side performance monitoring of the aforementioned single-sided model, including performance monitoring of the terminal's reported performance monitoring output and performance monitoring of the terminal's reported performance metric. Without loss of generality, the following discussion on how performance monitoring reports and inference reports are correlated in CSI reporting uses AI / ML CSI prediction as an example, but it also applies to other single-sided model use cases such as AI / ML beam management and AI / ML localization.

[0080] In traditional CSI prediction, the prediction of CSI at N4 future time points, where N4 ∈ {1,2,4,8}, is configured by higher-layer Radio Resource Control (RRC) parameters. Considering the potentially large overhead of predicting multiple future CSI times and the significant overhead of the precoding matrix indicator (PMI) for a single time point in the eType II codebook, eType II doppler-based CSI reporting only supports aperiodic and semi-persistent reporting, and the reporting channel is the Physical Uplink Shared Channel (PUSCH). Given that AI CSI prediction will try to reuse the Rel-18 MIMO feedback method based on the "typeII-Doppler-r18" codebook and the existing CSI-RS configuration, AI CSI prediction CSI reporting should also only support aperiodic and semi-persistent PUSCH reporting.

[0081] If we consider aperiodic, PUSCH-based AI CSI prediction (inference report) CSI reporting, the supported reference signals can be aperiodic, semi-persistent, and periodic CSI-RS reference signals. Inference report CSI reporting can be activated by Downlink Control Information (DCI) or by Media Access Control (MAC) - Control Element (CE). As shown in Figure 8, when the inference report is activated by the DCI / MAC-CE sent by the network side, the UE obtains the time offset values ​​(offset value X) of the inference report reference signal CSI-RS transmission timing and the activation signal timing through the CSI-RS time offset configured by the higher-layer RRC parameters. The UE obtains the time offset values ​​(offset value Y) of the inference report reporting timing and the activation signal timing through the aperiodic CSI reporting offset configured by the higher-layer RRC parameters.

[0082] If we consider CSI reporting for semi-persistent, PUSCH-based AI CSI prediction (inference report), the supported reference signals can be semi-persistent and periodic CSI-RS reference signals, and the CSI reporting for the inference report can be activated by DCI. As shown in Figure 9, after the inference report is activated by DCI sent by the network side, the UE obtains the time offset values ​​(offset value X) of the CSI-RS transmission timing and activation signal timing of the inference report reference signal through the CSI-RS time offset configured by the higher-layer RRC parameters. The UE obtains the time offset values ​​(offset value Y) of the first inference report reporting timing and activation signal timing through the CSI reporting offset configured by the higher-layer RRC parameters. The UE obtains the reporting timing of subsequent semi-persistent CSI reports based on the first inference report reporting timing by using the reporting period (period T) of the semi-persistent CSI reporting configured by the higher-layer RRC parameters.

[0083] However, when CSI reports the introduction of AI models / AI functions / ML models / ML functions, the current relevant protocols do not address how to monitor and report the performance of AI models / AI functions / ML models / ML functions. In other words, there is currently no solution for how to monitor and report performance.

[0084] Please refer to Figure 10, which is a flowchart of an information reporting method provided in an embodiment of this disclosure, applied to a terminal device. As shown in Figure 10, it includes the following steps:

[0085] Step 1001: Send a performance monitoring report to the network device. The performance monitoring report includes the performance measurement results of the inference model or inference function and / or the performance monitoring output results.

[0086] The performance monitoring report is sent in conjunction with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0087] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0088] It should be understood that performance measurement results, also known as performance index results, characterize the gap between the predicted CSI of the inference model / inference function and the actual measured CSI. This disclosure does not impose specific limitations; for example, performance measurement results may include, but are not limited to, Generalized Cosine Similarity (GCS), Squared Generalized Cosine Similarity (SGCS), and Normalized Mean Square Error (NMSE). Performance monitoring output results characterize the results of performance monitoring of the inference model / inference function. For example, in one example, it may include a first performance monitoring output result or a second performance monitoring output result. The first performance monitoring output result may indicate that the performance monitoring of the inference model or inference function failed / was poor, while the second performance monitoring output result may indicate that the performance monitoring of the inference model or inference function passed / was good. For example, in another example, the performance monitoring output may include the number of times the first performance monitoring output or the second performance monitoring output occurs within a preset time range. For example, if the first performance monitoring output or the second performance monitoring output occurs k times within the preset time range, it indicates the number of times the performance monitoring of the inference model or inference function fails / has poor performance or passes / has good performance.

[0089] In the information reporting process of this embodiment, it is necessary to report not only the inference report including the prediction of CSI, but also the performance monitoring report of the inference model or inference function. The performance monitoring report can be reported in association with the first inference report, which is the inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. Alternatively, the performance monitoring report can be reported independently of the inference report. That is, this embodiment can realize the performance monitoring report being reported in association with the first inference report, or the performance monitoring report being reported independently of the inference report, thus realizing the performance monitoring reporting of the inference model / inference function.

[0090] In some embodiments, the method further includes:

[0091] Acquire the measured CSI and the predicted CSI obtained by the inference model or inference function. The measured CSI is the CSI measured at the prediction time corresponding to the predicted CSI, or the CSI measured at the first time. The first time is before or after the prediction time and within a preset time interval from the prediction time.

[0092] Based on the predicted CSI and the measured CSI, determine the performance metrics and / or performance monitoring outputs of the inference model or inference function.

[0093] It should be understood that the inference model or inference function can perform CSI prediction. For example, as an example, the CSI measured before the prediction time can be obtained first, and then input into the inference model or inference function. The inference model or inference function can then predict the CSI at the prediction time, thus obtaining the predicted CSI at the prediction time. Additionally, the measured CSI measured at the prediction time or a first time interval also needs to be obtained. Then, using the predicted CSI at the prediction time and the measured CSI measured at the prediction time or a first time interval before or after the prediction time (within a preset time interval), the performance measurement results and / or performance monitoring output results of the inference model or inference function are determined and reported to the network device, thus realizing the reporting of the performance measurement results and / or performance monitoring output results of the inference function. It can be understood that the first moment is before or after the predicted moment, and the time interval between the first moment and the predicted moment is within a preset duration, that is, the time interval is less than or equal to the preset duration. For example, if the predicted moment is ty and the preset duration is Ty, then the first moment can be any moment within the preset duration Ty before or after the predicted moment ty, that is, the first moment can be any moment within [ty-Ty, ty+Ty].

[0094] It should be noted that there are multiple ways to determine the performance measurement results and / or performance monitoring output results of the inference model or inference function. This disclosure does not impose specific limitations. As an example, the performance measurement results can be compared with corresponding preset thresholds, and the performance monitoring output results can be determined based on the comparison results. For instance, for SGCS, if SGCS is less than the corresponding preset threshold, the performance monitoring output result is the first performance monitoring output result (e.g., it can be, but is not limited to, 0); if SGCS is greater than or equal to the corresponding preset threshold, the performance monitoring output result is the second performance monitoring output result (e.g., it can be, but is not limited to, 1). Another example is that if the total number of times SGCS is less than the corresponding preset threshold within the timing range of the performance monitoring failure timer is greater than a preset number, the performance monitoring output result is the first performance monitoring output result; otherwise, the performance monitoring output result is the second performance monitoring output result. Yet another example is that the total number of times SGCS is less than the corresponding preset threshold is counted within the timing range of the performance monitoring failure timer; or, the total number of times SGCS is greater than or equal to the corresponding preset threshold is counted within the timing range of the performance monitoring failure timer, and the performance monitoring output result is the total number of times SGCS is greater than or equal to the corresponding preset threshold within the timing range of the performance monitoring failure timer.

[0095] In some embodiments, when the performance monitoring report is sent in conjunction with the first inference report, the predicted CSI is the CSI in the first inference report, and the measured CSI is obtained by measuring the first monitoring reference signal resource based on the network device configuration; and / or

[0096] When the performance monitoring report and the inference report are sent independently, the measured CSI is obtained based on the second monitoring reference signal resource configured by the network device, and the predicted CSI is obtained by inference based on the measurement results by the inference model or inference function. The measurement results are obtained based on the third monitoring reference signal resource configured by the network device.

[0097] It should be understood that, in one example of this embodiment, the network device may configure a first monitoring reference signal resource for the terminal device. This first monitoring reference signal resource can be used solely for actual CSI measurement to obtain the measured CSI. In this example, the terminal device can obtain the predicted CSI from the first inference report, and the terminal device can report the performance monitoring report in association with the first inference report. In another example of this embodiment, the network device may also configure two sets of reference signal resources for the terminal device: a second monitoring reference signal resource and a third monitoring reference signal resource. The second monitoring reference signal resource can be used for actual CSI measurement to obtain the measured CSI, and the third monitoring reference signal resource can be used for actual CSI measurement to obtain the measurement result, which serves as input to the inference model or inference function. The predicted CSI is then obtained through inference prediction using the inference model or inference function. In this example, the terminal device can report the performance monitoring report and the first inference report independently.

[0098] In this embodiment, the terminal device can measure the first monitoring reference signal resource configured by the network device to obtain the measured CSI, obtain the predicted CSI from the associated first inference report, and use the predicted CSI and the measured CSI to determine the performance measurement result and / or performance monitoring output result. The terminal device can then report a performance monitoring report, including the performance measurement result and / or performance monitoring output result, to the network device. Alternatively, the terminal device can measure the second monitoring reference signal resource configured by the network device to obtain the measured CSI, and measure the third monitoring reference signal resource configured by the network device to obtain the measurement result. The measurement result is input into an inference model or inference function, and a predicted CSI is obtained through prediction using the inference model or inference function. The terminal device then uses the predicted CSI and the measured CSI to determine the performance measurement result and / or performance monitoring output result, and reports a performance monitoring report, including the performance measurement result and / or performance monitoring output result, to the network device. In other words, this embodiment can use either of the above two reporting methods to achieve performance monitoring report reporting, improving the flexibility of performance monitoring report reporting.

[0099] In some embodiments, the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources; or, the second monitoring reference signal resource and the third monitoring reference signal resource are the same reference signal resource at different times.

[0100] In other words, the second and third monitoring reference signal resources can be different reference signal resources, or they can be the same reference signal resource at different times. Thus, actual measurements can be performed on the second monitoring reference signal resource to obtain the measured CSI. Similarly, actual measurements can be performed on the third monitoring reference signal resource (different from the second) to obtain the CSI measurement result, or on the third monitoring reference signal resource (identical to the second but at a different time) to obtain the CSI measurement result. The measurement result is input into an inference model or inference function, which then performs inference based on the measurement result to obtain a predicted CSI. Subsequently, the predicted CSI and measured CSI can be used to determine the performance measurement results and / or performance monitoring output results of the inference model or inference function. A performance monitoring report, including the predicted CSI, measured CSI, and the performance measurement results and / or performance monitoring output results of the inference model or inference function, is then reported to the network device, thus achieving the reporting of the performance monitoring report for the inference model or inference function.

[0101] In some embodiments, when the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources, the second monitoring reference signal resource and the third monitoring reference signal resource are located in different sets of reference signal resources, or the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources in the same set of reference signal resources.

[0102] The second and third monitoring reference signal resources configured for the terminal equipment can be located in different sets of reference signal resources, or they can be different reference signal resources in the same set of reference signal resources, thereby improving the flexibility of the configuration of the second and third monitoring reference signal resources.

[0103] In some embodiments, where the performance monitoring report and the inference report are sent independently, the performance monitoring report may also include first timestamp information associated with the performance measurement results and / or the performance monitoring output results.

[0104] In other words, each performance metric and / or performance monitoring output in the performance monitoring report has its associated first timestamp information. The performance metric and / or performance monitoring output can be the performance metric and / or performance monitoring output within their associated first timestamp information. When performance monitoring reports and inference reports are reported independently, the performance monitoring report includes not only the performance metric and / or performance monitoring output results, but also the first timestamp information associated with each performance metric and / or performance monitoring output result. This means that not only are the performance metric and / or performance monitoring output results reported, but also the associated first timestamp information is reported.

[0105] In some embodiments, at least one inference reporting time associated with the monitoring reporting time of a performance monitoring report is determined based on at least one of the following:

[0106] Index information for inference reports associated with performance monitoring reports;

[0107] Time-domain information for inference report sending and performance monitoring report sending;

[0108] First association rule.

[0109] It should be understood that performance monitoring reports and inference reports are different CSI reports, each with its own index information and its own reporting time-domain behavior. The time-domain information of inference report sending and performance monitoring report sending can also be understood as the time-domain behavior of inference report reporting and performance monitoring report reporting. As an example, the first association rule can be predefined by the protocol, and / or configured by the network device, and / or reported by the terminal device. That is, the first association rule can include at least one of the following: a rule predefined by the protocol; a rule configured by the network device; or a rule reported by the terminal device.

[0110] In the case of CSI-related reporting, the performance monitoring report is sent in association with at least one inference report at a time associated with the monitoring reporting time of the performance monitoring report. In this embodiment, at least one inference report time associated with the monitoring reporting time of the performance monitoring report can be determined through at least one of the above index information, time domain information and first association rule, thereby determining the associated first inference report and improving the accuracy of the determined associated first inference report.

[0111] In some embodiments, the method further includes:

[0112] Configure to receive performance monitoring reports sent by network devices;

[0113] Based on the performance monitoring report configuration, obtain index information.

[0114] Inference reports and monitoring reports are separate CSI reports, each with its own reporting and resource configurations. In the monitoring report reporting configuration, a new higher-layer RRC parameter can be introduced to indicate the associated inference report. Network devices can use this RRC parameter to configure the inference report configuration index (CSI-ReportConfigID) associated with the monitoring report in the monitoring report reporting configuration. Terminal devices obtain the index information of the inference report associated with the monitoring report based on the monitoring report configuration.

[0115] In some embodiments, where both inference reports and performance monitoring reports are sent aperiodically, the first association rule includes:

[0116] When the DCI triggering of the inference report and the DCI triggering of the monitoring report are jointly triggered, the reporting time of the interrelated inference report and the reporting time of the performance monitoring report are determined based on the maximum value of the CSI reporting offset in the inference report configuration sent by the network device and the CSI reporting offset in the performance monitoring report configuration.

[0117] If the DCI trigger in the inference report and the DCI trigger in the monitoring report are triggered separately, then any of the following conditions must be met:

[0118] The last symbol of the inference report associated with the performance monitoring report is before the first symbol triggered by the DCI of the performance monitoring report;

[0119] The last symbol of the inference report associated with the performance monitoring report is before the first symbol of the CSI reference resource reported by the performance monitoring report.

[0120] In some embodiments, when inference reports are sent aperiodically and performance monitoring reports are sent semi-continuously, the first association rule includes any of the following conditions:

[0121] The last symbol of the inference report associated with the performance monitoring report precedes the first symbol of the DCI activation in the performance monitoring report;

[0122] The last symbol of the inference report associated with the performance monitoring report is before the first symbol of the CSI reference resource in the first performance monitoring report.

[0123] In some embodiments, where inference reports are sent semi-persistently and performance monitoring reports are sent aperiodically, the first association rule includes:

[0124] The reporting time of the non-periodic performance monitoring report is associated with the reporting times of the K most recent and consecutive semi-persistent inference reports, where K is a positive integer, and the K most recent and consecutive semi-persistent inference report reporting times satisfy the following rules:

[0125] The last symbol of the last inference report reporting time among the K most recent and consecutive semi-persistent inference report reporting times is before the first symbol triggered by the DCI at the reporting time of the non-periodic performance monitoring report;

[0126] The last symbol of the last inference report in the K most recent and consecutive semi-persistent inference report reporting times precedes the first symbol of the CSI reference resource of the non-periodic performance monitoring report.

[0127] In some embodiments, when both inference reports and performance monitoring reports are sent semi-persistently, the first association rule includes:

[0128] The reporting time of the i-th performance monitoring report and K i The most recent and consecutive semi-persistent inference reporting times are associated, 1≤i≤N monitor K i N is a positive integer. monitor K is a positive integer representing the number of performance monitoring reports. i The most recent and consecutive semi-persistent inference reporting times satisfy the following rule:

[0129] K i The last symbol of the last inference reporting moment in the most recent and consecutive semi-persistent inference reporting moments precedes the first symbol of the CSI reference resource in the i-th performance monitoring report.

[0130] It's understandable that after a terminal device obtains the index information of an inference report associated with a monitoring report, it cannot directly obtain the correlation between one or more monitoring reports and a specific inference report. This is because monitoring reports and inference reports have different temporal behaviors. For inference report CSI reporting, AI CSI prediction should only support aperiodic and semi-persistent PUSCH reporting. For monitoring report CSI reporting, since the goal is to obtain the measured CSI (i.e., the target true CSI) corresponding to the predicted CSI time, it should also be high-precision CSI reporting, i.e., based on the R16 eType II codebook, or the R18 eType Doppler codebook, or a high-precision codebook similar to the R16 eType II codebook, or a high-precision codebook similar to the R18 eType II Doppler codebook. Therefore, monitoring report CSI reporting should only support aperiodic and semi-persistent PUSCH reporting. When inference reports and / or monitoring reports are semi-continuous PUSCH reports, multiple CSI reports will be periodically generated after DCI is triggered. Therefore, it is necessary to further clarify the specific time of the inference report CSI report associated with each monitoring report CSI report.

[0131] Since inference reports can be aperiodic and semi-persistent, and monitoring reports can also be aperiodic and semi-persistent, there are four possible intersections between them:

[0132] Scenario A: Non-periodic inference report, non-periodic monitoring report: In this case, there is only 1 inference reporting time and 1 monitoring reporting time.

[0133] Scenario B: Non-periodic inference report, semi-continuous monitoring report: In this case, there is only one inference reporting moment and multiple monitoring reporting moments.

[0134] Case C: Semi-persistent inference report, non-periodic monitoring report: In this case, it is necessary to determine which of the multiple inference reporting times the single monitoring reporting time is associated with, or which of the multiple inference reporting times the single non-periodic monitoring reporting time is associated with.

[0135] Case D: Semi-persistent inference report, semi-persistent monitoring report: In this case, it is necessary to determine which of the multiple inference reporting times each of the multiple monitoring reporting times is associated with.

[0136] For scenario A, since the monitoring report and the inference report are associated, and each has only one reporting time, if the DCI trigger of the inference report and the DCI trigger of the monitoring report are joint triggers, although the two reports may have different high-level RRC parameter configurations for the CSI reporting offset value Y... inf Y mon(Similar to the offset value Y in Figure 8), but in order to ensure that all aperiodic CSIs are reported in the same PUSCH resource, the standard defines the maximum slot offset in the slot offset of each A-CSI report using the DCI indicator (i.e., Y = max(Y)). inf Y mon This serves as the slot offset corresponding to this trigger state. If the DCI trigger of the inference report and the DCI trigger of the monitoring report are separate triggers, there is no ambiguity regarding the association between the monitoring report time and the inference report time, since the two reports are associated and each has only one report time. However, because the calculation of the performance metric / performance monitoring output of the monitoring report requires obtaining the predicted CSI in the inference report, there are additional timing requirements for the associated inference report time. The last symbol of the inference report time associated with the monitoring report should precede the first symbol of the DCI trigger of the monitoring report, or the last symbol of the inference report time associated with the monitoring report should precede the first symbol of the CSI reference resource in the monitoring report.

[0137] For aperiodic inference reports and aperiodic monitoring reports, the first association rule may include:

[0138] If the DCI trigger of the inference report and the DCI trigger of the monitoring report are jointly triggered, and the inference report and the monitoring report are reported on the PUSCH resource in the same time slot, the interrelated inference reporting time and monitoring reporting time are determined by the maximum value of the CSI reporting offset configured in the high-level RRC parameters of the inference report and the monitoring report.

[0139] If the DCI trigger of the inference report and the DCI trigger of the monitoring report are triggered separately, either of the following two conditions must be met:

[0140] The last symbol of the inference report associated with the monitoring report should precede the first symbol triggered by the DCI of the monitoring report;

[0141] The last symbol of the inference report associated with the monitoring report should precede the first symbol of the CSI reference resource in the monitoring report.

[0142] For scenario B, since the inference report and monitoring report have different temporal behaviors, they can only be triggered / activated separately. The non-periodic inference report triggered by DCI has only one inference reporting time, while the semi-persistent monitoring report activated by DCI has multiple monitoring reporting times.

[0143] In scenario B above, each monitoring report time of the semi-persistent monitoring report from DCI activation to DCI deactivation is associated with the inference report time of the related aperiodic inference report. Since the calculation of the performance metric / performance monitoring output of the monitoring report requires obtaining the predicted CSI from the inference report, there are additional timing requirements for the associated inference report times. The last symbol of the inference report time of the inference report associated with the monitoring report should precede the first symbol of DCI activation in the monitoring report, or the last symbol of the inference report time of the inference report associated with the monitoring report should precede the first symbol of the CSI reference resource in the first monitoring report.

[0144] That is, for aperiodic inference reports and semi-persistent monitoring reports, each monitoring reporting time of the semi-persistent monitoring report from the time DCI is activated to the time before DCI is deactivated is associated with the inference reporting time of the associated aperiodic inference report. The first association rule includes either of the following two conditions:

[0145] The last symbol of the inference report associated with the monitoring report should precede the first symbol of the DCI activation of the monitoring report;

[0146] The last symbol of the inference report associated with the monitoring report should precede the first symbol of the CSI reference resource in the first monitoring report.

[0147] For scenario C, since the inference report and monitoring report exhibit different temporal behaviors, they can only be triggered / activated separately. The semi-persistent inference report after DCI activation has multiple periodic inference reporting times, while the non-periodic monitoring report after DCI triggering has only one monitoring reporting time. Therefore, scenario C has two possibilities:

[0148] Alt.1: One association between the monitoring reporting time and multiple inference reporting times.

[0149] In Case C Alt.1, it is necessary to determine which of the multiple semi-persistent inference reporting moments is associated with the single aperiodic monitoring reporting moment. Since the purpose of performance monitoring is to confirm whether the currently used inference model is suitable for the channel conditions / network configuration / terminal characteristics or configuration at the current time or time period, the aperiodic monitoring reporting moment should be associated with the latest inference reporting moment before the first symbol triggered by the DCI in the monitoring report, or the latest inference reporting moment before the first symbol of the CSI reference resource in the monitoring report. It should be noted that the channel conditions may include, but are not limited to, path loss, penetration loss characteristics, direct / non-direct path distribution (LOS / NLOS distribution), and blocking characteristics. The network configuration may include, but is not limited to, the TXRU virtualization configuration of network devices (e.g., base stations), antenna configuration, port configuration, measurement reference signal period configuration, and reporting period configuration. The terminal features or configurations may include, but are not limited to, terminal TXRU virtualization configuration, antenna configuration, port configuration, and terminal speed.

[0150] Alt.2: Multiple associations in monitoring reporting time and multiple inference reporting times.

[0151] In Case C Alt.2, it is necessary to determine which of the multiple semi-persistent inference reporting moments a single aperiodic monitoring reporting moment is associated with. This information should be understood consistently on both the terminal and network sides; therefore, a parameter K can be introduced to indicate that the aperiodic monitoring reporting moment is associated with the K most recent and consecutive semi-persistent inference reporting moments. This parameter K can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. Since the calculation of the performance metric / performance monitoring output of the monitoring report requires obtaining the predicted CSI in the inference report, there are additional timing requirements for the associated inference reporting moments. The last symbol of the latest inference reporting moment among the K inference reporting moments associated with the monitoring report should precede the first symbol of the DCI activation in the monitoring report, or the last symbol of the latest inference reporting moment among the K inference reporting moments associated with the monitoring report should precede the first symbol of the CSI reference resource in the monitoring report.

[0152] That is, for semi-persistent inference reports and aperiodic monitoring reports, the parameter K determines that the aperiodic monitoring reporting time is associated with the K most recent and consecutive semi-persistent inference reporting times. The parameter K can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. The first association rule includes: the K most recent semi-persistent inference reporting times are determined by one of the following two rules:

[0153] The non-periodic monitoring reporting time should be associated with the K most recent and consecutive semi-persistent inference reporting times, and the last symbol of the last inference reporting time among the K should be before the first symbol triggered by the DCI of the monitoring report.

[0154] The non-periodic monitoring reporting time should be associated with the K most recent and consecutive semi-persistent inference reporting times, and the last symbol of the last inference reporting among the K should precede the first symbol of the CSI reference resource in the monitoring reporting.

[0155] For scenario D, the semi-persistent inference / monitoring report after DCI activation has multiple periodic inference / monitoring reporting times. In this case, it is necessary to determine which semi-persistent monitoring reporting time(s) are associated with each semi-persistent monitoring reporting time(s). Furthermore, since each trigger state in the semi-persistent reporting can only be associated with one reporting configuration, the semi-persistent inference / monitoring report(s) are activated separately.

[0156] Furthermore, in similar cases to Alt.2 in C, a set of parameters can be introduced. Used to indicate the time of the i-th semi-continuous monitoring report and K i The most recent and consecutive semi-persistent inference reporting times are associated, i = 1, ..., N. monitor K i i = 1, ..., N monitor The values ​​of the parameters can be the same or different. It can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. Since the calculation of performance metrics / performance monitoring outputs in the monitoring report requires obtaining the predicted CSI from the inference report, there are additional timing requirements for the associated inference reporting time. However, since monitoring reporting is a semi-continuous reporting with DCI activation, it is inappropriate to use the DCI activation of the monitoring report as a timing reference for other monitoring reporting times besides the first monitoring reporting time, similar to cases A, B, and C. Instead, the CSI reference resource corresponding to the monitoring reporting time can be used as a timing reference.

[0157] For semi-continuous inference / monitoring reports, via parameters Determining the time of the i-th monitoring report is related to K.i The most recent and consecutive semi-persistent inference reporting times are associated, i = 1, ..., N. monitor .parameter It can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. The first association rule includes: K i The most recent semi-persistent inference reporting time is determined by the following rule:

[0158] The i-th semi-continuous monitoring report time should be the same as K. i The most recent and consecutive semi-persistent inference reporting times are associated, and K satisfies i The last symbol of the last inference report in each inference report precedes the first symbol of the CSI reference resource in the i-th monitoring report.

[0159] As can be seen from the above, the first association rule can be different for inference reports and performance monitoring reports under different circumstances, in order to improve the accuracy of at least one inference reporting time associated with the monitoring reporting time of the determined performance monitoring report.

[0160] In some embodiments, when a performance monitoring report is sent in association with a first inference report, the performance measurement results and / or performance monitoring output results are associated with at least one predicted CSI in the first inference report, the at least one predicted CSI being determined based on a second association rule.

[0161] It is understood that when the performance monitoring report is sent in conjunction with the first inference report, the predicted CSI can be at least one predicted CSI in the first inference report, the performance measurement results and / or performance monitoring output results are associated with the above-mentioned at least one predicted CSI, and at least one predicted CSI can be determined based on the second association rule in order to determine the performance measurement results and / or performance monitoring output results to be reported.

[0162] In some embodiments, the second association rule includes at least one of the following:

[0163] The rules are predefined in the protocol;

[0164] Network device configuration rules;

[0165] Rules reported by terminal devices.

[0166] It should be understood that the second association rule can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side. That is, the second association rule can include at least one of the following: rules predefined by the protocol, rules configured by the network device, and rules reported by the terminal device.

[0167] In some embodiments, the rules reported by the terminal device include:

[0168] The terminal device reports a second timestamp information, each second timestamp information is associated with a performance metric result and / or performance monitoring output result, wherein at least one predicted CSI associated with the performance metric result and / or performance monitoring output result is a predicted CSI within the corresponding second timestamp information.

[0169] In some embodiments, the second timestamp information includes absolute time, relative time, time point, or time period.

[0170] Since the association between the monitoring reporting time and the inference reporting time has been determined according to the first association rule, when determining which one or more predicted CSIs in the inference report are associated with the performance metric or performance monitoring output, it is to determine which one or more predicted CSIs in the inference reporting time are associated with each performance metric (performance metric result) or performance monitoring output (performance monitoring output result) at the monitoring reporting time.

[0171] The method for determining the association between performance metrics or performance monitoring outputs and predicted CSIs is a second association rule, which can be predefined by the protocol. For example, if one inference reporting time is associated with one monitoring reporting time, and the inference report at the inference reporting time predicts future CSIs for N4 consecutive future slot intervals, while the monitoring report at the monitoring reporting time contains N4 performance metrics or performance monitoring outputs, the predefined second association rule is that the monitoring report contains the same number of performance metrics or performance monitoring outputs as the N4 predicted future CSIs in the inference report, and the order of these performance metrics or performance monitoring outputs is consistent with the reporting order of the predicted future CSIs in the inference report.

[0172] The second association rule can also be configured on the network side. For example, one inference reporting time is associated with one monitoring reporting time. The inference report of the inference time predicts the future CSIs as the CSIs of N4 future consecutive slot intervals, while the monitoring report of the monitoring time contains one performance metric or performance monitoring output. The second association rule configured on the network side instructs the terminal via RRC signaling / DCI signaling / MAC-CE signaling which of the N4 predicted future CSIs this performance metric or performance monitoring output corresponds to.

[0173] The second association rule can also be reported by the terminal side. For example, the terminal side reports second timestamp information, and each second timestamp is associated with a performance metric or performance monitoring output. Since the CSI predicted in the inference report can be the CSI of N4 consecutive future slot intervals, the performance metric / performance monitoring output can be one value or multiple values. When the performance metric / performance monitoring output is a single value, corresponding to the performance monitoring of one second timestamp, the performance metric / performance monitoring output can be the performance monitoring value of a single predicted CSI within the second timestamp (e.g., if the second timestamp is [n+δ, n+δ+d-1], the performance monitoring is the performance metric / performance monitoring output of the first predicted CSI). Alternatively, the performance metric / performance monitoring output can be a single performance monitoring value calculated from multiple predicted CSIs within the second timestamp (e.g., if the second timestamp is [n+δ, n+δ+N4*d-1], the performance monitoring is the average of the performance metrics for N4 consecutive predicted CSIs; or, for example, if the second timestamp is [n+δ, n+δ+N4*d-1], the performance monitoring is the total number of times the SGCS performance metric is less than the threshold among N4 consecutive predicted CSIs). When the performance metric / performance monitoring output has multiple values, corresponding to the performance monitoring of multiple second timestamps, each performance metric / performance monitoring output value corresponds to one second timestamp, which will not be elaborated further here.

[0174] The second timestamp information can be an absolute time (similar to UTC-time in positioning) or a relative time (similar to [n+δ,n+δ+N4*d-1] above, which is the interval of the time slot offset relative to the reporting time); it can be a point in time or a time interval.

[0175] Please refer to Figure 11. Figure 11 is a flowchart of an information reporting method provided by an embodiment of this disclosure, applied to a network device. As shown in Figure 11, it includes the following steps:

[0176] Step 1101: Receive the performance monitoring report sent by the terminal device. The performance monitoring report includes the performance measurement results of the inference model or inference function and / or the performance monitoring output results.

[0177] Among them, the performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0178] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0179] In some embodiments, the method further includes:

[0180] The configuration information is sent to the terminal device. The configuration information is used to indicate the first monitoring reference signal resource, which is used to measure and obtain the measured CSI; or it is used to indicate the second and third monitoring reference signal resources used for measurement. The second monitoring reference signal resource is used to measure and obtain the measured CSI, and the third monitoring reference signal resource is used to measure and obtain the measurement result. The measurement result is used by the inference model or inference function to infer and obtain the predicted CSI.

[0181] In some embodiments, the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources; or, the second monitoring reference signal resource and the third monitoring reference signal resource are the same reference signal resource at different times.

[0182] In some embodiments, when the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources, the second monitoring reference signal resource and the third monitoring reference signal resource are located in different sets of reference signal resources, or the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources in the same set of reference signal resources.

[0183] In some embodiments, where the performance monitoring report and the inference report are sent independently, the performance monitoring report may also include first timestamp information associated with the performance measurement results and / or the performance monitoring output results.

[0184] In some embodiments, the method further includes at least one of the following:

[0185] Configure the sending of performance monitoring reports to terminal devices;

[0186] Configure the inference report to be sent to the terminal device.

[0187] The process of the above method will be specifically described below with some specific embodiments.

[0188] The embodiments disclosed herein can solve at least the following problems:

[0189] Question 1: In AIML CSI prediction, is it necessary to report CSI to the correlation inference report and monitoring report?

[0190] In AI / ML beam prediction, Set B of the spatial beam prediction is a wide beam, while Set A is a narrow beam. In this case, the reference signal for the inference report is the wide beam reference signal, while the reference signal for the monitoring report is the narrow beam reference signal; the measurement reference signals for inference and monitoring cannot be reused. To reduce the overhead of reference signals and terminal measurements, it is reasonable to configure periodic, semi-persistent inference reporting in AIML beam management, and to associate a small number of non-periodic or low-density periodic, semi-persistent monitoring reports with the aforementioned inference reports to monitor model performance.

[0191] For AI CSI prediction, one approach is to reuse the configuration of AIML beam management, correlate inference reports and monitoring reports for CSI reporting.

[0192] However, the spatial characteristics of the measurement reference signals used for inference and monitoring in AI CSI prediction are consistent, differing only in their temporal location. Therefore, the measurement reference signals for inference and monitoring can be reused. Thus, for AI CSI prediction, another method can be used: CSI reporting is independent of inference and monitoring reports. The terminal obtains both the predicted CSI and the target ground truth CSI from the monitoring report. In this case, the base station configures the monitoring reference signal for the terminal to include two parts (these two parts can be different measurement times of the same measurement reference signal): the first part is the measurement used to obtain the AI / ML model input for predicting the CSI, and the second part is the measurement used to predict the target ground truth CSI for the corresponding time slot interval. In this method, performance metrics / performance monitoring outputs can be obtained independently based on the monitoring report; therefore, the monitoring report does not need to be associated with the inference report.

[0193] Question 2: In AIML CSI prediction, if the CSI reports from inference and monitoring reports are correlated, how do we determine the correlation between the specific time points in the inference and monitoring reports, given that these reports can exhibit different temporal behaviors and combinations of these behaviors? Once the correlation between the specific time points in the inference and monitoring reports is determined, how do we determine which specific performance metric or performance monitoring output from the monitoring reports is associated with the predicted CSI for the N4 future CSI times included in the inference report time point?

[0194] For AI CSI prediction, if the AIML beam management configuration is reused, and the CSI reporting of the correlation inference report and monitoring report is correlated, further rules need to be defined for the correlation at the reporting time after correlation, and for the correlation of the performance metric / performance monitoring output of the monitoring report with which one or more inference results.

[0195] Regarding issue 2, this disclosure provides method one: an associated CSI reporting method for CSI reporting of inference reports based on air interface AI / ML and CSI reporting of performance monitoring reports, wherein:

[0196] The inference report is an inference report for a single model on the terminal side, while the performance monitoring report is a performance monitoring report for the terminal side.

[0197] The terminal obtains the index information of the inference report associated with the monitoring report based on the performance monitoring report configuration of the higher-layer RRC signaling;

[0198] The terminal determines at least one inference reporting time associated with each monitoring reporting time. The method for determining the reporting time association is based on at least one of the following information:

[0199] Index information for inference reports associated with monitoring reports;

[0200] The temporal behavior of CSI reporting in reasoning reports and CSI reporting in monitoring reports;

[0201] First association rule.

[0202] The first association rule can be predefined by the protocol, configured on the network side, reported by the terminal side, or a combination of the above methods.

[0203] The terminal calculates a monitoring report for each monitoring reporting time, containing at least one performance metric or performance monitoring output. Each performance metric or performance monitoring output is associated with at least one predicted CSI in the inference report. The inference report is the inference report of at least one inference reporting time associated with the monitoring reporting time. The method for determining the association between the performance metric or performance monitoring output and the predicted CSI is a second association rule, which may be predefined by the protocol, configured on the network side, reported by the terminal side, or a combination of the above.

[0204] The terminal reports a performance monitoring report.

[0205] Regarding issue 1, this disclosure provides a second method: a CSI reporting method for performance monitoring reports (reported independently of inference reports), specifically for CSI reporting of performance monitoring reports based on air interface AIML, wherein:

[0206] The performance monitoring report is a terminal-side performance monitoring report;

[0207] The performance monitoring report is a stand-alone CSI report and is not correlated with reports from other CSIs.

[0208] The terminal performs measurements based on the reference signal resource set / reference signal resources in the resource configuration. The measurement results can be used as input to the AIML model input part for predicting CSI to obtain the predicted CSI, or as a measurement of the target true CSI for the time slot interval corresponding to the predicted CSI.

[0209] The terminal calculates a monitoring report for each monitoring reporting time, which includes at least one performance metric or performance monitoring output and a first timestamp, with each first timestamp associated with a performance metric or performance monitoring output;

[0210] The terminal reports a performance monitoring report.

[0211] The following is a detailed description of this publicly disclosed plan.

[0212] The performance monitoring in this solution refers to the terminal-side performance monitoring of the aforementioned single-sided model, including performance monitoring of the terminal-reported performance monitoring output (Figure 5) and performance monitoring of the terminal-reported performance metric (Figure 7). Without loss of generality, the following discussion on how the performance monitoring report and inference report are related in CSI reporting uses AI / ML CSI prediction as an example, but it is also applicable to other single-sided model use cases such as AI / ML beam management and AI / ML localization.

[0213] According to the AI / ML terminal-side performance monitoring process in Figures 5 and 7, the terminal needs to predict both the CSI and the corresponding ground truth CSI to calculate the performance metric / performance monitoring output. Based on the different sources of the predicted CSI and the corresponding ground truth CSI, two methods can be used.

[0214] Method 1: CSI reporting method for associated performance monitoring reports and inference reports:

[0215] The terminal obtains the predicted CSI from the inference report and the target true CSI corresponding to the predicted CSI from the monitoring report. The terminal calculates at least one performance metric / performance monitoring output based on the predicted CSI and the target true CSI. Each performance metric / performance monitoring output is correlated with the predicted CSI. In this method, the monitoring reference signal configured by the base station for the terminal in the monitoring report CSI reporting is only used for measuring the target true CSI. After obtaining the target true CSI at the predicted CSI time, the terminal needs to obtain the predicted CSI from the inference report associated with the monitoring report before calculating the performance metric / performance monitoring output. In this method, the performance monitoring RS is only used for measurement reporting; the UE behavior is a legacy behavior. This saves some RS, reduces the UE implementation complexity, and lowers processing time requirements.

[0216] Method 2, CSI reporting method that does not associate with performance monitoring reports and inference reports:

[0217] The terminal obtains both the predicted CSI and the target ground truth CSI from the monitoring report. In this case, the base station configures a monitoring reference signal for the terminal consisting of two parts: the first part is the measurement of the AIML model input for obtaining the predicted CSI, and the second part is the measurement of the target ground truth CSI for the corresponding time slot interval of the predicted CSI. Since the AIML model is located on the terminal side, the terminal can use the measurement results of the first part of the monitoring reference signal as the input of the AIML model to obtain the predicted CSI. The first and second parts of the reference signal resources can be different reference signal resources or different times of the same reference signal resource. At the same time, the terminal can obtain the target ground truth CSI at the time of the predicted CSI through the measurement results of the second part of the monitoring reference signal, and then calculate the performance metric / performance monitoring output based on both. In the above method, the performance metric / performance monitoring output can be obtained independently based on the monitoring report, so the monitoring report does not need to be associated with the inference report. In this method, the performance metric / performance monitoring output can be obtained directly from the monitoring report without establishing an associated CSI reporting, and multiple AI CSI prediction models can be monitored simultaneously.

[0218] If we consider Method 1, given that the AI / ML model for beam prediction uses spatial beam prediction, specifically a wide beam (Set B) predicting a narrow beam (Set A), the reference signal in the inference report can be a wide beam reference signal, while the reference signal in the monitoring report can be a narrow beam reference signal. If we consider Method 2, the reference signal in the monitoring report can include two types of reference signals with different spatial characteristics, and these two cannot be reused.

[0219] In AI / ML CSI prediction, the reference signals in the monitoring report and the inference report are reference signals with the same spatial characteristics, so both Method 1 and Method 2 mentioned above can be considered.

[0220] The following section uses AI / ML CSI prediction as an example to describe the specific implementation details of the aforementioned related CSI reporting methods. Without loss of generality, the method is also applicable to terminal-side single-sided model inference report CSI reporting based on AI / ML beam management and AI / ML positioning, as well as terminal-side performance monitoring report CSI reporting.

[0221] The specific implementation steps and details of Method 1 are as follows:

[0222] Terminal-side methods:

[0223] This method is a correlated CSI reporting method, targeting CSI reporting for inference reports and performance monitoring reports based on air interface AIML, wherein:

[0224] (1.1) The inference report is the inference report of the terminal-side single-side model, and the performance monitoring report is the terminal-side performance monitoring report;

[0225] (1.2) The terminal obtains the index information of the inference report associated with the monitoring report based on the performance monitoring report configuration of the higher-layer RRC signaling;

[0226] (1.3) The terminal determines at least one inference reporting time associated with each monitoring reporting time. The method for determining the reporting time association is based on at least one of the following information:

[0227] Index information for inference reports associated with monitoring reports;

[0228] The temporal behavior of CSI reporting in reasoning reports and CSI reporting in monitoring reports;

[0229] First association rule;

[0230] The first association rule can be predefined by the protocol, configured on the network side, reported by the terminal side, or a combination of the above methods.

[0231] (1.4) The terminal calculates a monitoring report for each monitoring reporting time, which includes at least one performance metric or performance monitoring output. Each performance metric or performance monitoring output is associated with at least one predicted CSI in the inference report. The inference report is the inference report of at least one inference reporting time associated with the monitoring reporting time. The method for determining the association between the performance metric or performance monitoring output and the predicted CSI is a second association rule, which may be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above methods.

[0232] (1.5) The terminal reports a performance monitoring report;

[0233] The following section uses AI / ML CSI prediction as an example to describe the specific implementation details of the aforementioned related CSI reporting methods. Without loss of generality, the method is also applicable to terminal-side single-sided model inference report CSI reporting based on AI / ML beam management and AI / ML positioning, as well as terminal-side performance monitoring report CSI reporting.

[0234] (1.1) The inference report is the inference report of the terminal-side single-side model, and the performance monitoring report is the terminal-side performance monitoring report:

[0235] For AI / ML-based CSI prediction, the inference report is the inference report of the terminal-side AI / ML CSI prediction one-sided model (see Figures 2 and 4). The performance monitoring report is the terminal-side performance monitoring report (see Figures 5 and 7). The monitoring reference signal resources configured by the network side for the terminal's performance monitoring report are used to obtain the target true CSI corresponding to the predicted CSI time.

[0236] (1.2) The terminal obtains the index information of the inference report associated with the monitoring report based on the performance monitoring report configuration of the higher-layer RRC signaling:

[0237] Inference reports and monitoring reports are separate CSI reports, each with its own reporting and resource configurations. In the monitoring report reporting configuration, a new higher-layer RRC parameter is introduced to indicate the associated inference report. The network side can use this RRC parameter to configure the inference report configuration index (CSI-ReportConfigID) associated with the monitoring report in the monitoring report reporting configuration. The terminal side obtains the index information of the inference report associated with the monitoring report based on the monitoring report configuration.

[0238] (1.3) The terminal determines at least one inference reporting time associated with each monitoring reporting time:

[0239] After obtaining the index information of an inference report associated with a monitoring report, the terminal device cannot directly obtain the association between one or more monitoring reports and a specific inference report. This is because monitoring reports and inference reports have different temporal behaviors. For inference report CSI reporting, AI CSI prediction should only support aperiodic and semi-persistent PUSCH reporting. For monitoring report CSI reporting, since the goal is to obtain the measured CSI (i.e., the target true CSI) corresponding to the predicted CSI time, it should also be high-precision CSI reporting, i.e., based on the R16 eType II codebook, or the R18 eType Doppler codebook, or a high-precision codebook similar to the R16 eType II codebook, or a high-precision codebook similar to the R18 eType II Doppler codebook. Therefore, monitoring report CSI reporting should only support aperiodic and semi-persistent PUSCH reporting. When inference reports and / or monitoring reports are semi-continuous PUSCH reports, multiple CSI reports will be periodically generated after DCI is triggered. Therefore, it is necessary to further clarify the specific time of the inference report CSI report associated with each monitoring report CSI report.

[0240] Since inference reports can be aperiodic and semi-persistent, and monitoring reports can also be aperiodic and semi-persistent, there are four possible intersections between them:

[0241] Scenario A: Non-periodic inference report, non-periodic monitoring report: In this case, there is only 1 inference reporting time and 1 monitoring reporting time.

[0242] Scenario B: Non-periodic inference report, semi-continuous monitoring report: In this case, there is only one inference reporting moment and multiple monitoring reporting moments.

[0243] Case C: Semi-persistent inference report, non-periodic monitoring report: In this case, it is necessary to determine which of the multiple inference reporting times the single monitoring reporting time is associated with, or which of the multiple inference reporting times the single non-periodic monitoring reporting time is associated with.

[0244] Case D: Semi-persistent inference report, semi-persistent monitoring report: In this case, it is necessary to determine which of the multiple inference reporting times each of the multiple monitoring reporting times is associated with.

[0245] For scenario A, since the monitoring report and the inference report are associated, and each has only one reporting time, if the DCI trigger of the inference report and the DCI trigger of the monitoring report are joint triggers, although the two reports may have different high-level RRC parameter configurations for the CSI reporting offset value Y... inf Y mon (Similar to the offset value Y in Figure 8), but in order to ensure that all aperiodic CSIs are reported in the same PUSCH resource, the standard defines the maximum slot offset in the slot offset of each A-CSI report using the DCI indicator (i.e., Y = max(Y)). inf Y mon This serves as the slot offset corresponding to this trigger state. If the DCI trigger of the inference report and the DCI trigger of the monitoring report are separate triggers, there is no ambiguity regarding the association between the monitoring report time and the inference report time, since the two reports are associated and each has only one report time. However, because the calculation of the performance metric / performance monitoring output of the monitoring report requires obtaining the predicted CSI in the inference report, there are additional timing requirements for the associated inference report time. The last symbol of the inference report time associated with the monitoring report should precede the first symbol of the DCI trigger of the monitoring report, or the last symbol of the inference report time associated with the monitoring report should precede the first symbol of the CSI reference resource in the monitoring report.

[0246] For aperiodic inference reports and aperiodic monitoring reports, the first association rule may include:

[0247] If the DCI trigger of the inference report and the DCI trigger of the monitoring report are jointly triggered, and the inference report and the monitoring report are reported on the PUSCH resource in the same time slot, the interrelated inference reporting time and monitoring reporting time are determined by the maximum value of the CSI reporting offset configured in the high-level RRC parameters of the inference report and the monitoring report.

[0248] If the DCI trigger of the inference report and the DCI trigger of the monitoring report are triggered separately, either of the following two conditions must be met:

[0249] The last symbol of the inference report associated with the monitoring report should precede the first symbol triggered by the DCI of the monitoring report;

[0250] The last symbol of the inference report associated with the monitoring report should precede the first symbol of the CSI reference resource in the monitoring report.

[0251] As shown in Figure 12, for scenario B, since the inference report and monitoring report have different temporal behaviors, they can only be triggered / activated separately. The aperiodic inference report triggered by DCI has only one inference reporting time, while the semi-persistent monitoring report activated by DCI has multiple monitoring reporting times.

[0252] In scenario B above, each monitoring report time of the semi-persistent monitoring report from DCI activation to DCI deactivation is associated with the inference report time of the related aperiodic inference report. Since the calculation of the performance metric / performance monitoring output of the monitoring report requires obtaining the predicted CSI from the inference report, there are additional timing requirements for the associated inference report times. The last symbol of the inference report time of the inference report associated with the monitoring report should precede the first symbol of DCI activation in the monitoring report, or the last symbol of the inference report time of the inference report associated with the monitoring report should precede the first symbol of the CSI reference resource in the first monitoring report.

[0253] That is, for aperiodic inference reports and semi-persistent monitoring reports, each monitoring reporting time of the semi-persistent monitoring report from the time DCI is activated to the time before DCI is deactivated is associated with the inference reporting time of the associated aperiodic inference report. The first association rule includes either of the following two conditions:

[0254] The last symbol of the inference report associated with the monitoring report should precede the first symbol of the DCI activation of the monitoring report;

[0255] The last symbol of the inference report associated with the monitoring report should precede the first symbol of the CSI reference resource in the first monitoring report.

[0256] For scenario C, since the inference report and monitoring report exhibit different temporal behaviors, they can only be triggered / activated separately. The semi-persistent inference report after DCI activation has multiple periodic inference reporting times, while the non-periodic monitoring report after DCI triggering has only one monitoring reporting time. Therefore, scenario C has two possibilities:

[0257] As shown in Figure 13, Alt.1: One association between the monitoring reporting time and multiple inference reporting times.

[0258] In Case C Alt.1, it is necessary to determine which of the multiple semi-persistent inference reporting moments is associated with the single aperiodic monitoring reporting moment. Since the purpose of performance monitoring is to confirm whether the currently used inference model is suitable for the channel conditions / network configuration / terminal characteristics or configuration at the current time or time period, the aperiodic monitoring reporting moment should be associated with the latest inference reporting moment before the first symbol triggered by the DCI in the monitoring report, or the latest inference reporting moment before the first symbol of the CSI reference resource in the monitoring report. It should be noted that the channel conditions may include, but are not limited to, path loss, penetration loss characteristics, direct / non-direct path distribution (LOS / NLOS distribution), and blocking characteristics. The network configuration may include, but is not limited to, the TXRU virtualization configuration of network devices (e.g., base stations), antenna configuration, port configuration, measurement reference signal period configuration, and reporting period configuration. The terminal features or configurations may include, but are not limited to, terminal TXRU virtualization configuration, antenna configuration, port configuration, and terminal speed.

[0259] As shown in Figure 14, Alt.2: Multiple associations between monitoring reporting times and multiple inference reporting times. In Case C Alt.2, it is necessary to determine which of the multiple semi-persistent inference reporting times the unique one-time non-periodic monitoring reporting time is associated with. The above information should be understood consistently on both the terminal and network sides; therefore, a parameter K can be introduced to indicate that the non-periodic monitoring reporting time is associated with the K most recent and consecutive semi-persistent inference reporting times. This parameter K can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. Since the calculation of the performance metric / performance monitoring output of the monitoring report requires obtaining the predicted CSI in the inference report, there are additional timing requirements for associated inference reporting times. The last symbol of the latest inference reporting time among the K inference reporting times associated with the monitoring report should precede the first symbol of the DCI activation in the monitoring report, or the last symbol of the latest inference reporting time among the K inference reporting times associated with the monitoring report should precede the first symbol of the CSI reference resource in the monitoring report.

[0260] That is, for semi-persistent inference reports and aperiodic monitoring reports, the parameter K determines that the aperiodic monitoring reporting time is associated with the K most recent and consecutive semi-persistent inference reporting times. The parameter K can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. The first association rule includes: the K most recent semi-persistent inference reporting times are determined by one of the following two rules:

[0261] The non-periodic monitoring reporting time should be associated with the K most recent and consecutive semi-persistent inference reporting times, and the last symbol of the last inference reporting time among the K should be before the first symbol triggered by the DCI of the monitoring report.

[0262] The non-periodic monitoring reporting time should be associated with the K most recent and consecutive semi-persistent inference reporting times, and the last symbol of the last inference reporting among the K should precede the first symbol of the CSI reference resource in the monitoring reporting.

[0263] For scenario D, the semi-persistent inference / monitoring report after DCI activation has multiple periodic inference / monitoring reporting times. In this case, it is necessary to determine which semi-persistent monitoring reporting time(s) are associated with each semi-persistent monitoring reporting time(s). Furthermore, since each trigger state in the semi-persistent reporting can only be associated with one reporting configuration, the semi-persistent inference / monitoring report(s) are activated separately.

[0264] Furthermore, as shown in Figure 15, a set of parameters can be introduced for Alt.2 in a similar case C. Used to indicate the time of the i-th semi-continuous monitoring report and K t The most recent and consecutive semi-persistent inference reporting times are associated, i = 1, ..., N. monitor K i i = 1, ..., N monitor The values ​​of the parameters can be the same or different. It can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. Since the calculation of performance metrics / performance monitoring outputs in the monitoring report requires obtaining the predicted CSI from the inference report, there are additional timing requirements for the associated inference reporting time. However, since monitoring reporting is a semi-continuous reporting with DCI activation, it is inappropriate to use the DCI activation of the monitoring report as a timing reference for other monitoring reporting times besides the first monitoring reporting time, similar to cases A, B, and C. Instead, the CSI reference resource corresponding to the monitoring reporting time can be used as a timing reference.

[0265] For semi-continuous inference / monitoring reports, via parameters Determining the time of the i-th monitoring report is related to K. i The most recent and consecutive semi-persistent inference reporting times are associated, i = 1, ..., N. monitor .parameter It can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above. The first association rule includes: K i The most recent semi-persistent inference reporting time is determined by the following rule:

[0266] The i-th semi-continuous monitoring report time should be the same as K. i The most recent and consecutive semi-persistent inference reporting times are associated, and K satisfies i The last symbol of the last inference report in each inference report precedes the first symbol of the CSI reference resource in the i-th monitoring report.

[0267] (1.4) The terminal calculates a monitoring report for each monitoring reporting time, containing at least one performance metric or performance monitoring output, each performance metric or performance monitoring output being associated with at least one predicted CSI in the inference report. The inference report is the inference report of at least one inference reporting time associated with that monitoring reporting time. The method for determining the association between the performance metric or performance monitoring output and the predicted CSI is the second association rule:

[0268] The terminal calculates the amount of monitoring reports reported at each monitoring reporting time. If the performance monitoring is terminal-side performance monitoring where the terminal reports performance metrics (Figure 7), then the monitoring report at each monitoring reporting time includes performance metrics. The performance metrics are calculated based on the predicted CSI for future times in the inference report and the actual measured true CSI for the aforementioned future times in the monitoring report. These metrics can be calculated using methods such as GCS (Generalized Cosine Similarity, GCS), SGCS (Squared Generalized Cosine Similarity, SGCS), MSE (Normalized Mean Square Error, NMSE), etc.

[0269] If the performance monitoring is terminal-side performance monitoring (Figure 5), then the monitoring report at each monitoring reporting time includes the performance monitoring output. The performance monitoring output is the output obtained by the terminal-side performance metric through the terminal-side performance monitoring decision recommendation algorithm. In a simple example, the performance monitoring output is a 1-bit output with two possible values, 0 and 1. The performance monitoring decision recommendation algorithm compares the input SGCS performance metric with a threshold. When the input SGCS is greater than the threshold, the performance monitoring output is 1; when the input SGCS is less than the threshold, the performance monitoring output is 0. Another example is that the performance monitoring output is the number of times the input SGCS performance metric is less than the threshold within the failure timer's time limit. The terminal side is configured with a performance monitoring failure timer. During the timer, whenever an SGCS performance metric is input into the performance monitoring decision recommendation algorithm, it is compared with the threshold. When the input SGCS is less than the threshold, the number of times the SGCS performance metric is less than the threshold is incremented by 1. When the timer ends, the terminal reports the total number of times the SGCS performance metric is less than the threshold and resets the count and the performance monitoring failure timer.

[0270] Since the association between the monitoring reporting time and the inference reporting time has been determined according to the first association rule mentioned above, determining which one or more predicted CSIs in the inference report are associated with a performance metric or performance monitoring output is equivalent to determining which one or more predicted CSIs in the inference reporting time are associated with each performance metric or performance monitoring output at the monitoring reporting time.

[0271] The method for determining the association between performance metrics or performance monitoring outputs and predicted CSIs is a second association rule, which can be predefined by the protocol. For example, one inference reporting time is associated with one monitoring reporting time. The inference report at the inference reporting time predicts future CSIs of N4 consecutive future slot intervals (Figure 2), and the monitoring report at the monitoring reporting time contains N4 performance metrics or performance monitoring outputs. The predefined second association rule is that the monitoring report contains the same number of performance metrics or performance monitoring outputs as the number N4 predicted future CSIs in the inference report, and the order of these performance metrics or performance monitoring outputs is consistent with the reporting order of the predicted future CSIs in the inference report.

[0272] The second association rule can also be configured on the network side. For example, one inference reporting time is associated with one monitoring reporting time. The inference report of the inference time predicts the future CSIs as the CSIs of N4 future consecutive slot intervals, while the monitoring report of the monitoring time contains one performance metric or performance monitoring output. The second association rule configured on the network side instructs the terminal via RRC signaling / DCI signaling / MAC-CE signaling which of the N4 predicted future CSIs this performance metric or performance monitoring output corresponds to.

[0273] The second association rule can also be reported by the terminal side. For example, the terminal side reports second timestamp information, and each second timestamp is associated with a performance metric or performance monitoring output. Since the CSI predicted in the inference report can be the CSI of N4 future consecutive slot intervals (Figure 2), the performance metric / performance monitoring output can be one value or multiple values. When the performance metric / performance monitoring output is a single value, corresponding to the performance monitoring of one second timestamp, the performance metric / performance monitoring output can be the performance monitoring value of a single predicted CSI within the second timestamp (e.g., if the second timestamp is [n+δ, n+δ+d-1], the performance monitoring is the performance metric / performance monitoring output of the first predicted CSI). Alternatively, the performance metric / performance monitoring output can be a single performance monitoring value calculated from multiple predicted CSIs within the second timestamp (e.g., if the second timestamp is [n+δ, n+δ+N4*d-1], the performance monitoring is the average of the performance metrics of N4 consecutive predicted CSIs; or, for example, if the second timestamp is [n+δ, n+δ+N4*d-1], the performance monitoring is the total number of times the SGCS performance metric is less than the threshold among N4 consecutive predicted CSIs). When the performance metric / performance monitoring output has multiple values, corresponding to the performance monitoring of multiple second timestamps, each performance metric / performance monitoring output value corresponds to one second timestamp, which will not be elaborated further here.

[0274] The second timestamp information can be an absolute time (similar to UTC-time in positioning) or a relative time (similar to [n+δ,n+δ+N4*d-1] above, which is the interval of the time slot offset relative to the reporting time); it can be a point in time or a time interval.

[0275] (1.5) Terminal reports performance monitoring reports:

[0276] The terminal submits a performance monitoring report at the corresponding monitoring and reporting time, and the performance monitoring report includes the aforementioned reporting volume.

[0277] Network-side methods:

[0278] The network-side methods correspond to those on the terminal side. Specifically, the "performance monitoring report configuration of higher-layer RRC signaling" mentioned in (1.2) on the terminal side is configured on the network side. The terminal inference report / monitoring report reporting configuration, the measurement reference signal resource set / reference signal resource configuration for inference reports / monitoring reports, and the trigger status / activation / deactivation configuration are all configured on the network side through higher-layer RRC signaling.

[0279] The terminal side and the network side have the same interpretation of the "one or more inference reporting times associated with each monitoring reporting time" in (1.3) above, because the information in (1.3) above exists on both the network side and the terminal side on the UE side.

[0280] Corresponding to the above (1.5) on the terminal side, the network side receives the CSI performance monitoring report. For terminal-side performance monitoring reports, the network side, based on the performance monitoring output of a single or multiple performance monitoring reports and / or the corresponding time slot interval information (the network side can perform joint statistics, anomaly removal, etc., on the performance monitoring output according to the time slot interval information of multiple performance monitoring reports), obtains the LCM decision output (e.g., inference model / function rollback, etc.) according to the network side's LCM decision algorithm, and performs relevant LCM operations through downlink signaling control of the terminal when necessary. For terminal-side performance monitoring reports, based on the performance metrics of a single or multiple performance monitoring reports and / or the corresponding time slot interval information (the network side can perform averaging, time-weighted, anomaly removal, etc., on the performance metrics according to the time slot interval information of multiple performance monitoring reports), obtains the LCM decision output according to the network side's performance monitoring decision algorithm and LCM decision algorithm, and performs relevant LCM operations through downlink signaling control of the terminal when necessary.

[0281] Example 1 (Method 1):

[0282] The base station is configured with aperiodic inference report (CSI) report #1. CSI report #1 is associated with N1 sets of reference signal resources and / or reference signal resources. The configuration information of these N1 sets and / or reference signal resources includes their frequency domain and time domain information. CSI report #1 contains the CSIs of N4 future consecutive time slot intervals inferred by the AI / ML CSI prediction model on the terminal side. The CSIs of these N4 future consecutive time slot intervals are obtained by the terminal based on the measurement results of at least one set and / or reference signal resource associated with CSI report #1, after processing on the terminal side (including but not limited to quantization and domain transformation, or optional processing), and then used as input to the AI / ML CSI prediction model. The AI / ML CSI prediction model infers the output, which is then processed again on the terminal side (including but not limited to quantization and domain transformation, or optional processing).

[0283] The base station is configured with a non-periodic monitoring report (CSI) #2, which is associated with CSI #1. CSI #2 is associated with N2 sets of reference signal resources and / or reference signal resources. The configuration information for these N2 sets and / or reference signal resources includes their frequency and time domain information. CSI #2 contains N4 performance metrics.

[0284] The DCI triggering of CSI report #1 and the DCI triggering of CSI report #2 are joint triggering. CSI report #1 and CSI report #2 are reported on the PUSCH resource in the same time slot. The reporting time of CSI report #1 and CSI report #2 is determined by the maximum value of the CSI reporting offset configured by the higher-level RRC parameters of CSI report #1 and CSI report #2.

[0285] The CSI report #1 predicts future CSIs for N4 consecutive future slot intervals. The monitoring report at the monitoring reporting time contains N4 performance metrics. The second association rule predefined by the protocol is that the monitoring report contains the same number of performance metrics as the N4 predicted future CSIs in the inference report, and the order of these performance metrics is consistent with the reporting order of the predicted future CSIs in the inference report.

[0286] Example 2 (Method 1):

[0287] The base station is configured with aperiodic inference report (CSI) report #1. CSI report #1 is associated with N1 sets of reference signal resources and / or reference signal resources. The configuration information of these N1 sets and / or reference signal resources includes their frequency domain and time domain information. CSI report #1 contains the CSIs of N4 future consecutive time slot intervals inferred by the AI / ML CSI prediction model on the terminal side. The CSIs of these N4 future consecutive time slot intervals are obtained by the terminal based on the measurement results of at least one set and / or reference signal resource associated with CSI report #1, after processing on the terminal side (including but not limited to quantization and domain transformation, or optional processing), and then used as input to the AI / ML CSI prediction model. The AI / ML CSI prediction model infers the output, which is then processed again on the terminal side (including but not limited to quantization and domain transformation, or optional processing).

[0288] The base station is configured with a non-periodic monitoring report (CSI) #2, which is associated with CSI #1. CSI #2 is associated with N2 sets of reference signal resources and / or reference signal resources. The configuration information for these N2 sets and / or reference signal resources includes their frequency and time domain information. CSI #2 contains N3 performance metrics.

[0289] The DCI trigger for CSI report #1 and the DCI trigger for CSI report #2 are triggered separately. The last symbol of the reporting time of CSI report #1 should be before the first symbol of the DCI trigger for CSI report #2.

[0290] CSI report #1 predicts future CSIs for N4 consecutive future slot intervals, while CSI report #2 predicts N3 performance metrics. The network side uses RRC signaling / DCI signaling / MAC-CE signaling to indicate the correlation between the N3 performance metrics in CSI report #2 and the N4 predicted future CSIs in CSI report #1.

[0291] Example 3 (Method 1):

[0292] The base station is configured with aperiodic inference report (CSI) report #1. CSI report #1 is associated with N1 sets of reference signal resources and / or reference signal resources. The configuration information of these N1 sets and / or reference signal resources includes their frequency domain and time domain information. CSI report #1 contains the CSIs of N4 future consecutive time slot intervals inferred by the AI / ML CSI prediction model on the terminal side. The CSIs of these N4 future consecutive time slot intervals are obtained by the terminal based on the measurement results of at least one set and / or reference signal resource associated with CSI report #1, after processing on the terminal side (including but not limited to quantization and domain transformation, or optional processing), and then used as input to the AI / ML CSI prediction model. The AI / ML CSI prediction model infers the output, which is then processed again on the terminal side (including but not limited to quantization and domain transformation, or optional processing).

[0293] The base station is configured with a semi-persistent monitoring report (CSI) report #2, which is associated with CSI report #1. CSI report #2 is associated with one set of reference signal resources and / or one reference signal resource. The configuration information of this set of reference signal resources and / or the reference signal resource includes its frequency domain and time domain information. Each time CSI report #2 is submitted, it contains one performance metric.

[0294] Assume that there are a total of N5 reporting times in CSI report #2 from the time DCI is activated to the time DCI is deactivated. Each monitoring reporting time in CSI report #2 from the time DCI is activated to the time DCI is deactivated is associated with the inference reporting time of CSI report #1. The last symbol of CSI report #1 should be before the first symbol of DCI activation in CSI report #2. At the i-th reporting time of CSI report #2 (i = 1, ..., N5), the terminal reports one performance metric and one second timestamp [n + δ + (i-1)*d, n + δ + i*d-1]. This second timestamp indicates that this performance metric is associated with the predicted future CSI in the time slot interval [n + δ + (i-1)*d, n + δ + i*d-1] of CSI report #1.

[0295] Example 4 (Method 1):

[0296] The base station is configured with a semi-persistent inference report (CSI) report #1. CSI report #1 is associated with one set of reference signal resources and / or one reference signal resource. The configuration information of this set of reference signal resources and / or the reference signal resource includes its frequency domain and time domain information. Assuming that there are a total of N6 reporting times for CSI report #1 from DCI activation to DCI deactivation, each reporting time of CSI report #1 contains N4 future consecutive time slot intervals of CSI inferred by the AI / ML CSI prediction model on the terminal side. The N4 future consecutive time slot intervals of CSI are obtained by the terminal based on at least one measurement result from the one set of reference signal resources and / or the reference signal resource associated with CSI report #1, processed on the terminal side (including but not limited to quantization and domain transformation, or optional processing), used as input to the AI / ML CSI prediction model, inferred from the AI / ML CSI prediction model, and then processed again on the terminal side (including but not limited to quantization and domain transformation, or optional processing).

[0297] The base station is configured with a non-periodic monitoring report (CSI) #2, which is associated with CSI #1. CSI #2 is associated with N2 sets of reference signal resources and / or reference signal resources. The configuration information for these N2 sets and / or reference signal resources includes their frequency and time domain information. CSI #2 contains N3 performance metrics.

[0298] The terminal reported its AIML capability level. Based on the capability level reported by the terminal and the rules predefined in the protocol, the network side determined that CSI report #2 should be associated with the reporting times of the K most recent and consecutive CSI reports #1. Furthermore, the last symbol of the last CSI report #1 among the K reports must precede the first symbol of the CSI reference resource of CSI report #2.

[0299] Since CSI report #2 contains N3 performance metrics, the terminal reports N3 performance metrics and corresponding N3 second timestamp information. The j-th (j=1,…,N3) second timestamp information is [n+δ+(j-1)*d,n+δ+j*d-1]. Through this second timestamp information, the terminal and the network can have a consistent understanding of which one or more of the K*N4 consecutive predicted CSIs in the K reporting times of CSI report #1 associated with CSI report #2 are associated with the j-th performance metric. Specifically, assume that the l-th (l=1,…,N4) future time slot interval [n] in the k-th reporting time... k +δ+(l-1)*d,n kFor the CSI of +δ+l*d-1], if the j-th second timestamp information contains the l-th future time slot interval in the k-th reporting time, then this j-th performance metric is associated with the l-th predicted CSI in the k-th reporting time of CSI report #1.

[0300] Example 5 (Method 1):

[0301] The base station is configured with a semi-persistent inference report (CSI) report #1. CSI report #1 is associated with one set of reference signal resources and / or one reference signal resource. The configuration information of this set of reference signal resources and / or the reference signal resource includes its frequency domain and time domain information. Assuming that there are a total of N6 reporting times for CSI report #1 from DCI activation to DCI deactivation, each reporting time of CSI report #1 contains N4 future consecutive time slot intervals of CSI inferred by the AI / ML CSI prediction model on the terminal side. The N4 future consecutive time slot intervals of CSI are obtained by the terminal based on at least one measurement result from the one set of reference signal resources and / or the reference signal resource associated with CSI report #1, processed on the terminal side (including but not limited to quantization and domain transformation, or optional processing), used as input to the AI / ML CSI prediction model, inferred from the AI / ML CSI prediction model, and then processed again on the terminal side (including but not limited to quantization and domain transformation, or optional processing).

[0302] The base station is configured with a semi-persistent monitoring report (CSI) report #2, which is associated with CSI report #1. CSI report #2 is associated with one set of reference signal resources and / or one reference signal resource. The configuration information of this set of reference signal resources and / or the reference signal resource includes its frequency domain and time domain information. Each time CSI report #2 is submitted, it contains N3 performance metrics.

[0303] Assuming there are a total of N5 reporting times from DCI activation to DCI deactivation in CSI report #2, the base station indicates via RRC signaling / DCI signaling / MAC-CE signaling Determine the reporting time of the i-th (i=1,…,N5) CSI report #2 and K. i It is associated with the reporting time of the most recent and consecutive semi-persistent CSI report #1, and satisfies K. i The last symbol in the last CSI report #1 is before the first symbol in the CSI reference resource reported in the i-th CSI report #2.

[0304] Since each CSI report #2 contains N3 performance metrics, the i-th CSI report #2 contains N3 performance metrics and corresponding N3 second timestamp information. The j-th (j=1,…,N3) second timestamp information is [n i +δ+(j-1)*d,n i +δ+j*d-1]. Using this second timestamp information, the terminal and network can determine the K of CSI report #1 associated with CSI report #2. i K in each reporting time i There is a consistent understanding of which specific CSI(s) among the N4 consecutive predicted CSIs are associated with the j-th performance metric of the i-th CSI report #2. Specifically, assume the m-th (m = 1, ..., K) i The l-th (l=1,…,N4) future time slot interval in CSI report #1 [n m +δ+(l-1)*d,n m If the second timestamp information of the i-th reporting time of CSI report #2 contains the l-th future time slot interval in the m-th reporting time, then the j-th performance metric of the i-th reporting time of CSI report #2 is associated with the l-th predicted CSI in the m-th reporting time of CSI report #1.

[0305] The specific implementation steps and details of Method 2 are as follows:

[0306] Terminal-side methods:

[0307] This method is a CSI (Content Serving Index) reporting method for performance monitoring reports, specifically for CSI reporting based on air interface AIML, wherein:

[0308] (2.1) The performance monitoring report is the terminal-side performance monitoring report.

[0309] (2.2) The performance monitoring report is a stand-alone CSI report and is not linked to other CSI reports.

[0310] (2.3) The terminal performs measurements based on the reference signal resource set / reference signal resources in the resource configuration. The measurement results can be used as input to the AIML model input part of the predicted CSI to obtain the predicted CSI, or as the measurement of the target true CSI of the time slot interval corresponding to the predicted CSI.

[0311] (2.4) The terminal calculates a monitoring report for each monitoring reporting time, which includes at least one performance metric or performance monitoring output and a first timestamp, each first timestamp being associated with a performance metric or performance monitoring output.

[0312] (2.5) The terminal reports performance monitoring reports.

[0313] The following section uses AI / ML CSI prediction as an example to describe the specific implementation details of the CSI reporting method for the aforementioned phase-independent performance monitoring report. Without loss of generality, the method is also applicable to CSI reporting for inference reports of terminal-side single-sided models based on AI / ML beam management and AI / ML positioning, as well as CSI reporting for terminal-side performance monitoring reports.

[0314] Regarding (2.1) above:

[0315] For AI / ML-based CSI prediction, the performance monitoring report is the terminal-side performance monitoring report (see Figure 5 and Figure 7). The network side configures the monitoring reference signal resource set / resource for the terminal's performance monitoring report. Its measurement results can be used as input to the AIML model input part to obtain the predicted CSI, or as a measurement of the target true CSI for the corresponding time slot interval of the predicted CSI.

[0316] Regarding (2.2) above:

[0317] The performance monitoring report is a stand-alone CSI report, not linked to any other CSI reports. Based on the monitoring reference signal resource set / resources configured for the terminal by the network side, the terminal can directly calculate performance metrics / performance monitoring outputs without relying on any other inference report CSI reports.

[0318] Regarding (2.3) above:

[0319] The terminal performs measurements based on the reference signal resource set / reference signal resource in the resource configuration. The measurement results can be used as input to the AIML model input part for predicting CSI to obtain the predicted CSI, or as a measurement of the target true CSI for the time slot interval corresponding to the predicted CSI.

[0320] Alt.1: In the resource configuration (CSI-ResourceConfig), the reference signal resources used for inference (i.e., the measurement reference signal resources used as input to the AIML model for predicting CSI) and the reference signal resources used to obtain the target true CSI (i.e., the measurement reference signal resources used to obtain the target true CSI corresponding to the predicted CSI) are in different reference signal resource sets. For example, the resource configuration includes a CSI resource set table (csi-RS-ResourceSetList), which indicates S≥2 CSI resource sets, where S1≥1 CSI resource set is used for inference, and S2≥1 CSI resource set is used for obtaining the target true CSI, S1+S2=S. The division of the CSI resource set used for inference and the CSI resource set used for obtaining the target true CSI, as well as their relationship, can be predefined by the protocol, configured on the network side, reported by the terminal side, or a combination of the above methods.

[0321] Alt.2: In the resource configuration, the reference signal resources used for inference and the reference signal resources for obtaining the target truth CSI are in the same set of reference signal resources. For example, the resource configuration includes a CSI resource set table indicating S ≥ 1 CSI resource sets, where the i-th resource set is for performance monitoring of the i-th AIML CSI prediction model. The i-th resource set contains reference signal resources:

[0322] in, It is K inf The reference signal resource used for inference is the aforementioned third monitoring reference signal resource.

[0323] It is K mon A reference signal resource for obtaining the target true value CSI, namely the second monitoring reference signal resource mentioned above.

[0324] The allocation of reference signal resources used for inference and reference signal resources used for target truth CSI acquisition, as well as the relationship between the two, can be predefined by the protocol, and / or configured on the network side, and / or reported by the terminal side, and / or a combination of the above methods.

[0325] In the resource configuration, the reference signal resources used for inference and the reference signal resources used to obtain the target truth CSI can be in different sets of reference signal resources or in the same set of reference signal resources.

[0326] The set of reference signal resources used for inference, and / or the reference signal resources used to obtain the target truth CSI, can be determined by parameter L. Parameter L can be predefined by the protocol, and / or configured on the network side, and / or a combination of the aforementioned methods.

[0327] Here is an example of parameter L:

[0328] The parameter L can be predefined by the protocol. For example, the reference signal resource set used for inference and the reference signal resource set used to obtain the target truth CSI are two resource sets. The first resource set is used for inference, and the second resource set is used to obtain the target truth CSI.

[0329] The parameter L can be configured on the network side. For example, the reference signal resource set used for inference and the reference signal resource set used to obtain the target truth CSI are a resource set containing 6 aperiodic CSI-RS resources:

[0330] {CSIRS1,CSIRS2,CSIRS3,CSIRS4,CSIRS5,CSIRS6}

[0331] CSIRS1 to CSIRS4 are reference signal resources used for inference, and CSIRS5 to CSIRS6 are reference signal resources used to acquire the target truth CSI. The parameter L indicates the above information and can be the index set of reference signals used for inference and / or acquiring the target truth CSI ({1,2,3,4}, {5,6}); it can also be the number of reference signals used for inference and / or acquiring the target truth CSI (4 / 2 refers to the first 4 / last 2 resources in the resource set used for inference and / or acquiring the target truth CSI); or it can be a bitmap indication of the reference signals used for inference and / or acquiring the target truth CSI (e.g., a 6-bit bitmap, with 1 indicating the reference signal resource used for inference: 111100).

[0332] The parameter L can be a combination of the aforementioned methods. For example, the set of reference signal resources used for inference and the set of reference signal resources used to obtain the target true CSI can be a single resource set. When the monitoring report is for the performance monitoring of M AIML CSI prediction models, a predefined rule can be used to arrange the M CSI prediction models in ascending order of their model indices, with the i-th model associated with the i-th signal resource set. The reference signal resources used for inference and the reference signal resources used to obtain the target true CSI can be indicated using the aforementioned network-side configuration method, which will not be elaborated further.

[0333] Regarding (2.4) above:

[0334] The terminal calculates a monitoring report for each monitoring reporting time, containing at least one performance metric or performance monitoring output, and a first timestamp. Each first timestamp is associated with a performance metric or performance monitoring output. The first timestamp can be an absolute time (similar to UTC-time in location tracking) or a relative time (similar to [n+δ, n+δ+N4*d-1], which is the interval of time slot offset relative to the reporting time); it can be a point in time or a time interval. Feature 4 of Method 2 is similar to Feature 4 of Method 1 and will not be repeated here.

[0335] Regarding (2.5) above: It is similar to (1.5) of Method 1 above, and will not be repeated here.

[0336] Example 6 (Method 2):

[0337] The base station is configured with a monitoring report (CSI) report #2, which associates two reference signal resource sets. One CSI resource set is used for inference, and the other is used to obtain the target ground truth CSI. The configuration information of these two reference signal resource sets includes the frequency domain and time domain information of the reference signal resource sets and / or the reference signal resources. The terminal uses at least one measurement result from the reference signal resource set associated with CSI report #2 for inference, and after terminal-side processing (including but not limited to quantization and domain transformation, or optional processing), it is used as input to the AI / ML CSI prediction model. The AI / ML CSI prediction model then infers the output, and after further terminal-side processing (including but not limited to quantization and domain transformation, or optional processing), the predicted CSI is obtained. The terminal also uses at least one measurement result from the reference signal resource set associated with CSI report #2 for monitoring, and after terminal-side processing (including but not limited to quantization and domain transformation, or optional processing), it obtains the target ground truth CSI. The terminal calculates at least one performance metric or performance monitoring output based on the predicted CSI and the target true CSI, as well as first timestamp information, each of which is associated with a performance metric or performance monitoring output.

[0338] Example 7 (Method 2):

[0339] Base station configuration monitoring report CSI report #2. CSI report #2 is associated with S reference signal resource sets, denoted as reference signal resource sets M1,…,M S The p-th reference signal resource set is used for performance monitoring of the p-th AIML CSI prediction model / function. The p-th reference signal resource set M... p Includes K inf +K mon One reference signal resource:

[0340] in It is K inf A reference signal resource for inference.

[0341] It is K mon A reference signal resource used to obtain the target true value CSI.

[0342] The configuration information for these S reference signal resource sets includes the frequency domain and time domain information of the reference signal resource sets and / or the reference signal resources. The terminal uses the reference signal resource set M from the p-th AIML CSI prediction model in CSI report #2. p Based on the reference signal resources used for inference At least one measurement result, after terminal-side processing (including but not limited to quantization and domain transformation, or optional processing), is used as input to the AI / ML CSI prediction model. The AI / ML CSI prediction model's inference output is then processed again on the terminal side (including but not limited to quantization and domain transformation, or optional processing) to obtain the predicted CSI. The terminal uses the reference signal resource set M associated with the p-th AIML CSI prediction model based on CSI report #2. p According to the reference signal resources used for monitoring After processing on the terminal side (including but not limited to quantization and domain transformation, or no processing is required), the target ground truth CSI is obtained. The terminal calculates at least one performance metric or performance monitoring output of the p-th AIML CSI prediction model based on the predicted CSI and the target ground truth CSI, as well as first timestamp information, with each first timestamp information associated with a performance metric or performance monitoring output.

[0343] Network-side methods:

[0344] The methods and features on the network side correspond to those on the terminal side. The configuration for reporting terminal monitoring reports, the measurement reference signal resource set / reference signal resource configuration for monitoring reports, and the trigger status / activation / deactivation configuration are all configured on the network side through higher-layer RRC signaling.

[0345] For the terminal side (2.5), the network side receives the CSI performance monitoring report. For terminal-side performance monitoring reports, the network side, based on the performance monitoring output of a single or multiple performance monitoring reports and / or the corresponding time slot interval information (the network side can perform joint statistics, anomaly removal, etc., on the performance monitoring output according to the time slot interval information based on multiple performance monitoring reports), obtains the LCM decision output according to the network side's LCM decision algorithm. When necessary, the network side controls the terminal to perform relevant LCM operations through downlink signaling. For terminal-side performance monitoring reports, based on the performance metrics of a single or multiple performance monitoring reports and / or the corresponding time slot interval information (the network side can perform averaging, time-weighted, anomaly removal, etc., on the performance metrics according to the time slot interval information based on multiple performance monitoring reports), obtains the LCM decision output according to the network side's performance monitoring decision algorithm and LCM decision algorithm. When necessary, the network side controls the terminal to perform relevant LCM operations through downlink signaling.

[0346] This disclosure presents a design for associating different CSI reports in AI / ML NR-based CSI reporting. It addresses the question of whether and how performance monitoring reports and inference reports need to be associated in AI / ML NR-based CSI reporting. Traditional CSI reporting, lacking AI algorithms, does not require performance monitoring CSI reporting and therefore avoids the potential issue of associating performance monitoring and inference reports.

[0347] If we consider the correlation between performance monitoring reports and inference reports, the terminal obtains the predicted CSI from the inference report and the target true CSI corresponding to the predicted CSI from the monitoring report. The advantages are that the performance monitoring reference signal (RS) is only used for measurement reporting, the UE behavior is a legacy behavior, the terminal implementation complexity is low, and the processing time requirement is low; on the base station side, the monitoring reference signal configured by the base station for the terminal is only used for measuring the target true CSI, thus saving some RS resources in resource allocation.

[0348] If the performance monitoring report is not associated with the inference report, the terminal obtains both the predicted CSI and the target ground truth CSI from the monitoring report. The advantage is that performance metrics / performance monitoring output can be obtained directly from the monitoring report without needing to establish a related CSI reporting mechanism. Furthermore, CSI reporting does not require consideration of various temporal behaviors and complex time-series relationships. Additionally, this method allows for simultaneous monitoring of multiple AI CSI prediction models, facilitating model switching.

[0349] Please refer to Figure 16, which is a structural diagram of a communication device provided in an embodiment of the present invention. The communication device may include a terminal device, as shown in Figure 16, including a memory 1620, a transceiver 1600, and a processor 1610.

[0350] Memory is used to store computer programs; transceiver is used to send and receive data under the control of the processor; processor is used to read the computer programs from memory and perform the following operations:

[0351] Send performance monitoring reports to network devices, including performance measurement results and / or performance monitoring output results of inference models or inference functions;

[0352] The performance monitoring report is sent in conjunction with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0353] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0354] In some embodiments, the processor, for reading a computer program from memory, also performs the following operations:

[0355] Acquire the measured CSI and the predicted CSI obtained by the inference model or inference function. The measured CSI is the CSI measured at the prediction time corresponding to the predicted CSI, or the CSI measured at the first time. The first time is before or after the prediction time and within a preset time interval from the prediction time.

[0356] Based on the predicted CSI and the measured CSI, determine the performance metrics and / or performance monitoring outputs of the inference model or inference function.

[0357] In some embodiments, when the performance monitoring report is sent in conjunction with the first inference report, the predicted CSI is the CSI in the first inference report, and the measured CSI is obtained by measuring the first monitoring reference signal resource based on the network device configuration; and / or

[0358] When the performance monitoring report and the inference report are sent independently, the measured CSI is obtained based on the second monitoring reference signal resource configured by the network device, and the predicted CSI is obtained by inference based on the measurement results by the inference model or inference function. The measurement results are obtained based on the third monitoring reference signal resource configured by the network device.

[0359] In some embodiments, the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources; or, the second monitoring reference signal resource and the third monitoring reference signal resource are the same reference signal resource at different times.

[0360] In some embodiments, when the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources, the second monitoring reference signal resource and the third monitoring reference signal resource are located in different sets of reference signal resources, or the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources in the same set of reference signal resources.

[0361] In some embodiments, where the performance monitoring report and the inference report are sent independently, the performance monitoring report may also include first timestamp information associated with the performance measurement results and / or the performance monitoring output results.

[0362] In some embodiments, at least one inference reporting time associated with the monitoring reporting time of a performance monitoring report is determined based on at least one of the following:

[0363] Index information for inference reports associated with performance monitoring reports;

[0364] Time-domain information for inference report sending and performance monitoring report sending;

[0365] First association rule.

[0366] In some embodiments, the processor, for reading a computer program from memory, also performs the following operations:

[0367] Configure to receive performance monitoring reports sent by network devices;

[0368] Based on the performance monitoring report configuration, obtain index information.

[0369] In some embodiments, the first association rule includes at least one of the following:

[0370] The rules are predefined in the protocol;

[0371] Network device configuration rules;

[0372] Rules reported by terminal devices.

[0373] In some embodiments, where both inference reports and performance monitoring reports are sent aperiodically, the first association rule includes:

[0374] When the DCI triggering of the inference report and the DCI triggering of the monitoring report are jointly triggered, the reporting time of the interrelated inference report and the reporting time of the performance monitoring report are determined based on the maximum value of the CSI reporting offset in the inference report configuration sent by the network device and the CSI reporting offset in the performance monitoring report configuration.

[0375] If the DCI trigger in the inference report and the DCI trigger in the monitoring report are triggered separately, then any of the following conditions must be met:

[0376] The last symbol of the inference report associated with the performance monitoring report is before the first symbol triggered by the DCI of the performance monitoring report;

[0377] The last symbol of the inference report associated with the performance monitoring report is before the first symbol of the CSI reference resource reported by the performance monitoring report.

[0378] In some embodiments, when inference reports are sent aperiodically and performance monitoring reports are sent semi-continuously, the first association rule includes any of the following conditions:

[0379] The last symbol of the inference report associated with the performance monitoring report precedes the first symbol of the DCI activation in the performance monitoring report;

[0380] The last symbol of the inference report associated with the performance monitoring report is before the first symbol of the CSI reference resource in the first performance monitoring report.

[0381] In some embodiments, where inference reports are sent semi-persistently and performance monitoring reports are sent aperiodically, the first association rule includes:

[0382] The reporting time of the non-periodic performance monitoring report is associated with the reporting times of the K most recent and consecutive semi-persistent inference reports, where K is a positive integer, and the K most recent and consecutive semi-persistent inference report reporting times satisfy the following rules:

[0383] The last symbol of the last inference report reporting time among the K most recent and consecutive semi-persistent inference report reporting times is before the first symbol triggered by the DCI at the reporting time of the non-periodic performance monitoring report;

[0384] The last symbol of the last inference report in the K most recent and consecutive semi-persistent inference report reporting times precedes the first symbol of the CSI reference resource of the non-periodic performance monitoring report.

[0385] In some embodiments, when both inference reports and performance monitoring reports are sent semi-persistently, the first association rule includes:

[0386] The reporting time of the i-th performance monitoring report and K i The most recent and consecutive semi-persistent inference reporting times are associated, 1≤i≤N monitor K i N is a positive integer. monitor K is a positive integer representing the number of performance monitoring reports. i The most recent and consecutive semi-persistent inference reporting times satisfy the following rule:

[0387] K i The last symbol of the last inference reporting moment in the most recent and consecutive semi-persistent inference reporting moments precedes the first symbol of the CSI reference resource in the i-th performance monitoring report.

[0388] In some embodiments, when a performance monitoring report is sent in association with a first inference report, the performance measurement results and / or performance monitoring output results are associated with at least one predicted CSI in the first inference report, the at least one predicted CSI being determined based on a second association rule.

[0389] In some embodiments, the second association rule includes at least one of the following:

[0390] The rules are predefined in the protocol;

[0391] Network device configuration rules;

[0392] Rules reported by terminal devices.

[0393] In some embodiments, the rules reported by the terminal device include:

[0394] The terminal device reports a second timestamp information, each second timestamp information is associated with a performance metric result and / or performance monitoring output result, wherein at least one predicted CSI associated with the performance metric result and / or performance monitoring output result is a predicted CSI within the corresponding second timestamp information.

[0395] In some embodiments, the second timestamp information includes absolute time, relative time, time point, or time period.

[0396] In Figure 16, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 1610 and memory represented by memory 1620. The bus architecture may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 1600 may be multiple components, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium, including wireless channels, wired channels, optical fibers, etc. For different user equipment, the user interface 1630 may also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0397] The processor 1610 is responsible for managing the bus architecture and general processing, while the memory 1620 can store the data used by the processor 1610 during operation.

[0398] In some embodiments, the processor 1610 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a complex programmable logic device (CPLD), and the processor may also adopt a multi-core architecture.

[0399] The processor executes any of the methods provided in the embodiments of the present invention by invoking a computer program stored in memory, according to the obtained executable instructions. The processor and memory may also be physically separated.

[0400] It should be noted that the processor of the communication device provided in the embodiments of the present invention can implement the information reporting method steps applied to the terminal device implemented in the method embodiments, and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail.

[0401] Please refer to Figure 17, which is a structural diagram of a communication device provided in an embodiment of the present invention. The communication device may include a network device, as shown in Figure 17, including a memory 1720, a transceiver 1700, and a processor 1710.

[0402] Memory is used to store computer programs; transceiver is used to send and receive data under the control of the processor; processor is used to read the computer programs from memory and perform the following operations:

[0403] Receive a performance monitoring report sent by the terminal device. The performance monitoring report includes the performance measurement results of the inference model or inference function and / or the performance monitoring output results.

[0404] Among them, the performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0405] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0406] In some embodiments, the processor, for reading a computer program from memory, also performs the following operations:

[0407] The configuration information is sent to the terminal device. The configuration information is used to indicate the first monitoring reference signal resource, which is used to measure and obtain the measured CSI; or it is used to indicate the second and third monitoring reference signal resources used for measurement. The second monitoring reference signal resource is used to measure and obtain the measured CSI, and the third monitoring reference signal resource is used to measure and obtain the measurement result. The measurement result is used by the inference model or inference function to infer and obtain the predicted CSI.

[0408] In some embodiments, the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources; or, the second monitoring reference signal resource and the third monitoring reference signal resource are the same reference signal resource at different times.

[0409] In some embodiments, when the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources, the second monitoring reference signal resource and the third monitoring reference signal resource are located in different sets of reference signal resources, or the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources in the same set of reference signal resources.

[0410] In some embodiments, where the performance monitoring report and the inference report are sent independently, the performance monitoring report may also include first timestamp information associated with the performance measurement results and / or the performance monitoring output results.

[0411] In some embodiments, the processor, for reading a computer program from memory, also performs at least one of the following:

[0412] Configure the sending of performance monitoring reports to terminal devices;

[0413] Configure the inference report to be sent to the terminal device.

[0414] In Figure 17, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 1710 and memory represented by memory 1720. The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 1700 can be multiple components, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium, including wireless channels, wired channels, optical fibers, etc. For different user equipment, the user interface 1730 can also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0415] Processor 1710 is responsible for managing the bus architecture and general processing, while memory 1720 can store data used by processor 1710 during operation.

[0416] In some embodiments, the processor 1710 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a complex programmable logic device (CPLD), and the processor may also adopt a multi-core architecture.

[0417] The processor executes any of the methods provided in the embodiments of the present invention by invoking a computer program stored in memory, according to the obtained executable instructions. The processor and memory may also be physically separated.

[0418] It should be noted that the processor of the communication device provided in the embodiments of the present invention can implement the information reporting method steps applied to network devices implemented in the method embodiments, and can achieve the same technical effect. Therefore, the parts and beneficial effects that are the same as those in the method embodiments will not be described in detail here.

[0419] Please refer to Figure 18, which is a structural diagram of an information reporting device (applicable to terminal devices) provided in an embodiment of the present invention. As shown in Figure 18, the device 1800 includes:

[0420] The first sending module 1801 is used to send a performance monitoring report to the network device. The performance monitoring report includes the performance measurement results of the inference model or inference function and / or the performance monitoring output results.

[0421] The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0422] The inference report includes the predicted channel state information (CSI) output by the inference model or inference function.

[0423] In some embodiments, the device further includes:

[0424] The acquisition module is used to acquire the measured CSI and the predicted CSI obtained by the inference model or inference function. The measured CSI is the CSI measured at the prediction time corresponding to the predicted CSI, or the CSI measured at the first time. The first time is before or after the prediction time and within a preset time interval from the prediction time.

[0425] The determination module is used to determine the performance metrics and / or performance monitoring outputs of the inference model or inference function based on the predicted CSI and the measured CSI.

[0426] In some embodiments, when the performance monitoring report is sent in conjunction with the first inference report, the predicted CSI is the CSI in the first inference report, and the measured CSI is obtained by measuring the first monitoring reference signal resource based on the network device configuration; and / or

[0427] When the performance monitoring report and the inference report are sent independently, the measured CSI is obtained based on the second monitoring reference signal resource configured by the network device, and the predicted CSI is obtained by inference based on the measurement results by the inference model or inference function. The measurement results are obtained based on the third monitoring reference signal resource configured by the network device.

[0428] In some embodiments, the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources; or, the second monitoring reference signal resource and the third monitoring reference signal resource are the same reference signal resource at different times.

[0429] In some embodiments, when the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources, the second monitoring reference signal resource and the third monitoring reference signal resource are located in different sets of reference signal resources, or the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources in the same set of reference signal resources.

[0430] In some embodiments, where the performance monitoring report and the inference report are sent independently, the performance monitoring report may also include first timestamp information associated with the performance measurement results and / or the performance monitoring output results.

[0431] In some embodiments, at least one inference reporting time associated with the monitoring reporting time of a performance monitoring report is determined based on at least one of the following:

[0432] Index information for inference reports associated with performance monitoring reports;

[0433] Time-domain information for inference report sending and performance monitoring report sending;

[0434] First association rule.

[0435] In some embodiments, the device further includes:

[0436] The second receiving module is used to receive performance monitoring report configurations sent by network devices.

[0437] The index retrieval module is used to retrieve index information based on the configuration in the performance monitoring report.

[0438] In some embodiments, the first association rule includes at least one of the following:

[0439] The rules are predefined in the protocol;

[0440] Network device configuration rules;

[0441] Rules reported by terminal devices.

[0442] In some embodiments, where both inference reports and performance monitoring reports are sent aperiodically, the first association rule includes:

[0443] When the DCI triggering of the inference report and the DCI triggering of the monitoring report are jointly triggered, the reporting time of the interrelated inference report and the reporting time of the performance monitoring report are determined based on the maximum value of the CSI reporting offset in the inference report configuration sent by the network device and the CSI reporting offset in the performance monitoring report configuration.

[0444] If the DCI trigger in the inference report and the DCI trigger in the monitoring report are triggered separately, then any of the following conditions must be met:

[0445] The last symbol of the inference report associated with the performance monitoring report is before the first symbol triggered by the DCI of the performance monitoring report;

[0446] The last symbol of the inference report associated with the performance monitoring report is before the first symbol of the CSI reference resource reported by the performance monitoring report.

[0447] In some embodiments, when inference reports are sent aperiodically and performance monitoring reports are sent semi-continuously, the first association rule includes any of the following conditions:

[0448] The last symbol of the inference report associated with the performance monitoring report precedes the first symbol of the DCI activation in the performance monitoring report;

[0449] The last symbol of the inference report associated with the performance monitoring report is before the first symbol of the CSI reference resource in the first performance monitoring report.

[0450] In some embodiments, where inference reports are sent semi-persistently and performance monitoring reports are sent aperiodically, the first association rule includes:

[0451] The reporting time of the non-periodic performance monitoring report is associated with the reporting times of the K most recent and consecutive semi-persistent inference reports, where K is a positive integer, and the K most recent and consecutive semi-persistent inference report reporting times satisfy the following rules:

[0452] The last symbol of the last inference report reporting time among the K most recent and consecutive semi-persistent inference report reporting times is before the first symbol triggered by the DCI at the reporting time of the non-periodic performance monitoring report;

[0453] The last symbol of the last inference report in the K most recent and consecutive semi-persistent inference report reporting times precedes the first symbol of the CSI reference resource of the non-periodic performance monitoring report.

[0454] In some embodiments, when both inference reports and performance monitoring reports are sent semi-persistently, the first association rule includes:

[0455] The reporting time of the i-th performance monitoring report and K i The most recent and consecutive semi-persistent inference reporting times are associated, 1≤i≤N monitor K i N is a positive integer. monitor K is a positive integer representing the number of performance monitoring reports. i The most recent and consecutive semi-persistent inference reporting times satisfy the following rule:

[0456] K i The last symbol of the last inference reporting moment in the most recent and consecutive semi-persistent inference reporting moments precedes the first symbol of the CSI reference resource in the i-th performance monitoring report.

[0457] In some embodiments, when a performance monitoring report is sent in association with a first inference report, the performance measurement results and / or performance monitoring output results are associated with at least one predicted CSI in the first inference report, the at least one predicted CSI being determined based on a second association rule.

[0458] In some embodiments, the second association rule includes at least one of the following:

[0459] The rules are predefined in the protocol;

[0460] Network device configuration rules;

[0461] Rules reported by terminal devices.

[0462] In some embodiments, the rules reported by the terminal device include:

[0463] The terminal device reports a second timestamp information, each second timestamp information is associated with a performance metric result and / or performance monitoring output result, wherein at least one predicted CSI associated with the performance metric result and / or performance monitoring output result is a predicted CSI within the corresponding second timestamp information.

[0464] In some embodiments, the second timestamp information includes absolute time, relative time, time point, or time period.

[0465] It should be noted that the information reporting device provided in this embodiment of the invention can implement the method steps applied to the terminal device as implemented in the above method embodiment, and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0466] Please refer to Figure 19. Figure 19 is a structural diagram of another information reporting device (which can be applied to network devices) provided in an embodiment of the present invention. As shown in Figure 19, the device 1900 includes:

[0467] The first receiving module 1901 is used to receive a performance monitoring report sent by the terminal device. The performance monitoring report includes performance measurement results of the inference model or inference function and / or performance monitoring output results.

[0468] Wherein, the performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report; or, the performance monitoring report and the inference report are sent independently.

[0469] The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function.

[0470] In some embodiments, the device further includes:

[0471] The second sending module is used to send configuration information to the terminal device. The configuration information is used to indicate the first monitoring reference signal resource, which is used to measure and obtain the measured CSI; or it is used to indicate the second and third monitoring reference signal resources used for measurement, which are used to measure and obtain the measured CSI, and the third monitoring reference signal resource is used to measure and obtain the measurement result. The measurement result is used by the inference model or inference function to infer and obtain the predicted CSI.

[0472] In some embodiments, the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources; or, the second monitoring reference signal resource and the third monitoring reference signal resource are the same reference signal resource at different times.

[0473] In some embodiments, when the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources, the second monitoring reference signal resource and the third monitoring reference signal resource are located in different sets of reference signal resources, or the second monitoring reference signal resource and the third monitoring reference signal resource are different reference signal resources in the same set of reference signal resources.

[0474] In some embodiments, where the performance monitoring report and the inference report are sent independently, the performance monitoring report may also include first timestamp information associated with the performance measurement results and / or the performance monitoring output results.

[0475] In some embodiments, the device further includes at least one of the following:

[0476] The third sending module is used to send performance monitoring report configuration to the terminal device;

[0477] The fourth sending module is used to send inference report configurations to the terminal device.

[0478] It should be noted that the information reporting device provided in this embodiment of the invention can implement the method steps applied to network devices as implemented in the above method embodiment, and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0479] This disclosure provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the information reporting method described above. It achieves the same technical effects, and therefore, the parts and beneficial effects identical to those in the method embodiment will not be described in detail here.

[0480] It should be noted that the division of units in the embodiments of the present invention is illustrative and only represents one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units.

[0481] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to related technologies, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0482] This disclosure provides a processor-readable storage medium storing a computer program. The computer program is used to cause the processor to execute the information reporting method provided in this disclosure, or the computer program is used to cause the processor to execute the information reporting method provided in this disclosure.

[0483] The processor-readable storage medium can be any available medium or data storage device that the processor can access, including but not limited to magnetic memory (e.g., floppy disk, hard disk, magnetic tape, magneto-optical disk (MO)), optical memory (e.g., CD, DVD, BD, HVD), and semiconductor memory (e.g., ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state drive (SSD)).

[0484] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0485] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-executable instructions. These computer-executable instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0486] These processor-executable instructions may also be stored in a processor-readable memory that can instruct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the processor-readable memory produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0487] These processor-executable instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0488] Furthermore, it should be noted that in the apparatus and method of this disclosure, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of this disclosure. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of this disclosure can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof, which can be achieved by those skilled in the art using their basic programming skills after reading the description of this disclosure.

[0489] It should be noted that the above division of modules is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can be implemented entirely in software via processing element calls; they can be fully implemented in hardware; or some modules can be implemented by processing element calls to software, while others are implemented in hardware. For example, a module can be a separate processing element, or it can be integrated into a chip in the aforementioned device. Alternatively, it can be stored as program code in the memory of the aforementioned device, and its function can be called and executed by a processing element of the device. The implementation of other modules is similar. Moreover, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element mentioned here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed through integrated logic circuits in the hardware of the processor element or through software instructions.

[0490] For example, each module, unit, subunit, or submodule can be one or more integrated circuits configured to implement the above methods, such as one or more application-specific integrated circuits (ASICs), one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs). As another example, when a module is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor capable of calling program code. Furthermore, these modules can be integrated together to implement a system-on-a-chip (SOC).

[0491] The terms “first,” “second,” etc., used in this disclosure and in the claims are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this disclosure described herein may be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. Additionally, the use of “and / or” in the specification and claims indicates at least one of the connected objects, such as A and / or B and / or C, indicating seven possibilities: A alone, B alone, C alone, and both A and B, both B and C, both A and C, and A, B, and C. Similarly, the use of “at least one of A and B” in this specification and claims should be understood as “A alone, B alone, or both A and B.”

[0492] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, this disclosure is also intended to include such modifications and variations.

Claims

An information reporting method, applied to a terminal device, the method comprising: Send a performance monitoring report to the network device, the performance monitoring report including the performance measurement results of the inference model or inference function and / or the performance monitoring output results; The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. The inference report includes the predicted channel state information (CSI) output by the inference model or inference function. The method according to claim 1, further comprising: Acquire the measured CSI and the predicted CSI obtained by the inference model or the inference function. The measured CSI is the CSI measured at the prediction time corresponding to the predicted CSI, or the CSI measured at a first time. The first time is before or after the prediction time and is within a preset time interval from the prediction time. Based on the predicted CSI and the measured CSI, determine the performance metric results and / or the performance monitoring output results of the inference model or the inference function. The method according to claim 2, wherein, When the performance monitoring report is sent in conjunction with the first inference report, the predicted CSI is the CSI in the first inference report, and the measured CSI is obtained by measuring based on the first monitoring reference signal resources configured for the network device. According to the method of claim 1, wherein, The at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report is determined based on at least one of the following: Index information of the inference report associated with the performance monitoring report; Time-domain information for inference report sending and performance monitoring report sending; First association rule. The method according to claim 4, further comprising: Configuration for receiving performance monitoring reports sent by the network device; According to the performance monitoring report configuration, obtain the index information. The method according to claim 4, wherein, The first association rule includes at least one of the following: The rules are predefined in the protocol; The rules configured for the network devices; The rules reported by the terminal device. The method according to claim 3, wherein, When the performance monitoring report is sent in association with the first inference report, the performance measurement result and / or the performance monitoring output result are associated with at least one predicted CSI in the first inference report, the at least one predicted CSI being determined based on a second association rule. The method according to claim 7, wherein, The second association rule includes at least one of the following: The rules are predefined in the protocol; The rules configured for the network devices; The rules reported by the terminal device. The method according to claim 8, wherein, The rules reported by the terminal device include: The second timestamp information reported by the terminal device is associated with a performance metric result and / or a performance monitoring output result, wherein at least one predicted CSI associated with the performance metric result and / or the performance monitoring output result is a predicted CSI within the corresponding second timestamp information. The method according to claim 9, wherein, The second timestamp information includes absolute time, relative time, time point, or time period. An information reporting method, applied to network devices, includes: Receive a performance monitoring report sent by the terminal device, the performance monitoring report including performance measurement results of the inference model or inference function and / or performance monitoring output results; The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function. The method according to claim 11, further comprising: The terminal device is sent configuration information, which is used to indicate a first monitoring reference signal resource, and the first monitoring reference signal resource is used to measure and obtain the measurement CSI. The method according to claim 11, further comprising at least one of the following: Configure the terminal device to send performance monitoring reports; Send the inference report configuration to the terminal device. A communication device, comprising: Memory, transceiver, and processor, among which: The memory is used to store computer programs; the transceiver is used to send and receive data under the control of the processor; the processor is used to read the computer programs in the memory and perform the following operations: Send a performance monitoring report to the network device, the performance monitoring report including the performance measurement results of the inference model or inference function and / or the performance monitoring output results; The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. The inference report includes the predicted channel state information (CSI) output by the inference model or inference function. A communication device, comprising: Memory, transceiver, and processor, The memory is used to store computer programs; the transceiver is used to send and receive data under the control of the processor. The processor is configured to read the computer program in the memory and perform the following operations: Receive a performance monitoring report sent by the terminal device, the performance monitoring report including performance measurement results of the inference model or inference function and / or performance monitoring output results; The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function. An information reporting device, the device comprising: The first sending module is used to send a performance monitoring report to the network device. The performance monitoring report includes the performance measurement results of the inference model or inference function and / or the performance monitoring output results. The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. The inference report includes the predicted channel state information (CSI) output by the inference model or inference function. An information reporting device, the device comprising: The first receiving module is used to receive a performance monitoring report sent by the terminal device. The performance monitoring report includes performance measurement results of the inference model or inference function and / or performance monitoring output results. The performance monitoring report is sent in association with the first inference report, and the first inference report is an inference report of at least one inference reporting time associated with the monitoring reporting time of the performance monitoring report. The inference report includes the predicted channel state information (CSI) output by the inference model or the inference function. A processor-readable storage medium storing a computer program for causing the processor to perform the method according to any one of claims 1 to 10, or the computer program for causing the processor to perform the method according to any one of claims 11 to 13.

Images