Monitoring methods for artificial intelligence (AI) model, terminals and devices

Abstract

The present disclosure relates to monitoring methods for an artificial intelligence (AI) model, terminals and devices. A method comprises: a terminal first receives a first reference signal, then uses a current AI model to process the first reference signal so as to obtain an estimation result, and finally sends the estimation result to a first device. Thus, the problem of how to define terminal behaviors and testing parameters in artificial intelligence (AI) model monitoring methods is solved to a certain extent.

Description

Monitoring method, terminal and device for artificial intelligence (AI) models Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a monitoring method, terminal, and device for an artificial intelligence (AI) model. Background Art

[0002] In related technologies, artificial intelligence (AI) models can improve positioning accuracy. Currently, AI models can be monitored throughout the lifecycle of each use case through monitoring solutions based on inference accuracy, system performance, data distribution, and more.

[0003] Summary of the Invention

[0004] The embodiments of the present disclosure provide a monitoring method, terminal, and device for an artificial intelligence (AI) model, which to a certain extent solve the problem of how to define terminal behavior and test parameters in the artificial intelligence (AI) model monitoring method.

[0005] According to a first aspect of an embodiment of the present disclosure, a monitoring method for an artificial intelligence (AI) model is proposed, the method being executed by a terminal and comprising:

[0006] receiving a first reference signal;

[0007] Processing the first reference signal using the current AI model to obtain an estimation result;

[0008] The estimation result is sent to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

[0009] According to a second aspect of an embodiment of the present disclosure, a monitoring method for an artificial intelligence (AI) model is proposed. The method is performed by a first device and includes:

[0010] Receiving an estimation result sent by a terminal, wherein the estimation result is obtained by the terminal by processing the first reference signal using a current AI model;

[0011] Based on the estimation result, determine whether to update the AI ​​model.

[0012] According to a third aspect of an embodiment of the present disclosure, a monitoring method for an artificial intelligence (AI) model is proposed, the method being executed by a terminal and comprising:

[0013] receiving a first reference signal;

[0014] Processing the first reference signal using the current AI model to obtain an estimation result;

[0015] Based on the estimation result, determine whether to update the AI ​​model.

[0016] According to a fourth aspect of an embodiment of the present disclosure, a monitoring method for an artificial intelligence (AI) model is provided. The method is performed by a first device and includes:

[0017] Send a first reference signal, where the first reference signal is used to assist the terminal in determining whether to update the current AI model.

[0018] According to a fifth aspect of an embodiment of the present disclosure, a terminal is provided, including:

[0019] a transceiver module, configured to receive a first reference signal;

[0020] a processing module, configured to process the first reference signal using a current AI model to obtain an estimation result;

[0021] The transceiver module is further used to send the estimation result to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

[0022] According to a sixth aspect of an embodiment of the present disclosure, a first device is provided, including:

[0023] a transceiver module, configured to receive an estimation result sent by a terminal, wherein the estimation result is obtained by the terminal by processing the first reference signal using a current AI model;

[0024] A processing module is used to determine whether to update the AI ​​model based on the estimation result.

[0025] According to a seventh aspect of an embodiment of the present disclosure, a terminal is provided, including:

[0026] a transceiver module, configured to receive a first reference signal;

[0027] a processing module, configured to process the first reference signal using a current AI model to obtain an estimation result;

[0028] The processing module is further used to determine whether to update the AI ​​model based on the estimation result.

[0029] According to an eighth aspect of the embodiments of the present disclosure, a first device is provided, including:

[0030] The transceiver module is used to send a first reference signal, wherein the first reference signal is used to assist the terminal in determining whether to update the current AI model.

[0031] According to a ninth aspect of an embodiment of the present disclosure, a terminal is provided, including:

[0032] one or more processors;

[0033] In which, the processor is used to call instructions to enable the terminal to execute the monitoring method for the artificial intelligence AI model described in any one of the first aspect and the third aspect.

[0034] According to a tenth aspect of the embodiments of the present disclosure, a first device is provided, including:

[0035] one or more processors;

[0036] In which, the processor is used to call instructions to enable the first device to execute the monitoring method for the artificial intelligence AI model described in any one of the second aspect and the fourth aspect.

[0037] According to the eleventh aspect of the embodiments of the present disclosure, a communication system is proposed, characterized in that it includes a terminal and a first device, wherein the terminal is configured to implement the monitoring method for the artificial intelligence AI model described in the first and third aspects, and the first device is configured to implement the monitoring method for the artificial intelligence AI model described in the second and fourth aspects.

[0038] According to the twelfth aspect of the embodiment of the present disclosure, a storage medium is proposed, which stores instructions, and is characterized in that when the instructions are executed on a communication device, the communication device executes the monitoring method for an artificial intelligence AI model as described in any one of the first aspect, the second aspect, the third aspect, and the fourth aspect.

[0039] According to the thirteenth aspect of the embodiment of the present disclosure, a program product is proposed, which includes a computer program, and is characterized in that when the computer program is run on a communication device, the communication device executes the monitoring method for the artificial intelligence AI model as described in any one of the first aspect, the second aspect, the third aspect, and the fourth aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following drawings required for describing the embodiments are introduced. The following drawings are merely some embodiments of the present disclosure and do not impose specific limitations on the protection scope of the present disclosure.

[0041] FIG1A is a schematic diagram illustrating an architecture of a communication system according to an embodiment of the present disclosure;

[0042] FIG1B is a schematic diagram of a framework of a monitoring method for an artificial intelligence (AI) model according to an embodiment of the present disclosure;

[0043] 2A-2C are interactive schematic diagrams illustrating a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure;

[0044] 3A-3E are flowcharts illustrating a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure;

[0045] 4A-4E are flowcharts illustrating a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure;

[0046] FIG5A is a schematic diagram illustrating interaction between a UE and a first device jointly monitoring an AI model according to an embodiment of the present disclosure;

[0047] FIG5B is a schematic diagram of interaction of a UE autonomous monitoring AI model according to an embodiment of the present disclosure;

[0048] FIG6A is a schematic structural diagram of a terminal proposed in an embodiment of the present disclosure;

[0049] FIG6B is a schematic structural diagram of a first device proposed in an embodiment of the present disclosure;

[0050] FIG6C is a schematic diagram of the structure of a terminal proposed in an embodiment of the present disclosure;

[0051] FIG6D is a schematic structural diagram of a first device proposed in an embodiment of the present disclosure;

[0052] FIG7A is a schematic structural diagram of a communication device proposed in an embodiment of the present disclosure;

[0053] FIG7B is a schematic diagram of the structure of the chip proposed in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0054] The embodiments of the present disclosure provide a monitoring method, terminal, and device for an artificial intelligence (AI) model.

[0055] In a first aspect, an embodiment of the present disclosure provides a monitoring method for an artificial intelligence (AI) model, the method being executed by a terminal and comprising:

[0056] receiving a first reference signal;

[0057] Processing the first reference signal using the current AI model to obtain an estimation result;

[0058] The estimation result is sent to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

[0059] In the above embodiment, after receiving the first reference signal, the terminal uses the current AI model to process the first reference signal to obtain an estimation result, and then sends the estimation result to the first device, thereby providing conditions for the terminal and the first device to jointly monitor the AI ​​model.

[0060] In conjunction with some embodiments of the first aspect, in some embodiments, the first reference signal includes at least one of the following:

[0061] Positioning reference signal PRS;

[0062] Synchronous signal block;

[0063] Channel State Information Reference Signal CSI-RS.

[0064] In the above embodiment, the terminal can obtain an estimation result by measuring at least one reference signal, thereby providing conditions for improving the reliability of assisting the first device in determining whether to update the AI ​​model through the estimation result.

[0065] In conjunction with some embodiments of the first aspect, in some embodiments, the first device is selected from one of the following:

[0066] Access network equipment;

[0067] Core network equipment;

[0068] Network server.

[0069] In the above embodiment, the terminal can send estimation results to multiple first devices to assist them in determining whether to update the AI ​​model, thereby improving the efficiency of the terminal and the first device jointly monitoring the artificial intelligence AI model.

[0070] In conjunction with some embodiments of the first aspect, in some embodiments, the method further includes:

[0071] When an updated AI model is received, the current AI model is replaced with the updated AI model.

[0072] In the above embodiment, when the terminal receives the updated AI model, it can replace the current AI model with the updated AI model to process the reference signal, thereby improving the accuracy of the AI ​​model.

[0073] In a second aspect, an embodiment of the present disclosure provides a monitoring method for an artificial intelligence (AI) model, the method being performed by a first device and comprising:

[0074] Receiving an estimation result sent by a terminal, wherein the estimation result is obtained by the terminal by processing the first reference signal using a current AI model;

[0075] Based on the estimation result, determine whether to update the AI ​​model.

[0076] In the above embodiment, the first device determines whether to update the AI ​​model based on the estimation result received from the terminal, thereby providing conditions for the terminal and the first device to jointly monitor the AI ​​model and improving the accuracy of the AI ​​model.

[0077] In conjunction with some embodiments of the second aspect, in some embodiments, the first device is an access network device, and the method further includes:

[0078] Sending the first reference signal to the terminal.

[0079] In the above embodiment, the first device sends the first reference signal to the terminal, so that the terminal measures the first reference signal, thereby providing conditions for the terminal and the first device to jointly monitor the AI ​​model.

[0080] In conjunction with some embodiments of the second aspect, in some embodiments, the first device is a core network device or a server, and the method further includes:

[0081] Sending first information to an access network device, where the first information is used to trigger the access network device to send a first reference signal to the terminal.

[0082] In the above embodiment, when the first device is a core network device or a server, the first information can be sent to the access network device to trigger the access network device to send a first reference signal to the terminal, thereby improving the reliability of the terminal and the first device monitoring AI model.

[0083] In conjunction with some embodiments of the second aspect, in some embodiments, determining whether to update the AI ​​model based on the estimation result includes:

[0084] determining an error of the estimation result according to a reference result associated with the first reference signal;

[0085] When the error of the estimation result is greater than a threshold, the AI ​​model is updated.

[0086] In the above embodiment, the first device first determines the error of the estimation result based on the reference result associated with the first reference signal, and then updates the AI ​​model when the error of the estimation result is greater than the threshold, thereby realizing joint monitoring of the AI ​​model by the terminal and the first device and improving the accuracy of the AI ​​model.

[0087] In conjunction with some embodiments of the second aspect, in some embodiments, the first reference signal is used for positioning, and the threshold is any one of the following:

[0088] a first value associated with a ground actual channel response error;

[0089] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0090] a third value associated with a time-of-arrival (TOA) error;

[0091] a fourth value associated with a time difference of arrival, TDOA, error;

[0092] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0093] In the above embodiment, when the first reference signal is used for positioning, the first device may determine the error of the estimation result based on any associated threshold, thereby improving the accuracy of the determined error of the estimation result.

[0094] In conjunction with some embodiments of the second aspect, in some embodiments, it is characterized in that,

[0095] The first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0096] The first reference signal is used for channel state information (CSI) measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0097] In the above embodiment, the first device determines the error of the estimation result by using a threshold value associated with different types of first reference signals, thereby improving the reliability of the determined error of the estimation result.

[0098] In conjunction with some embodiments of the second aspect, in some embodiments, the method further includes:

[0099] Sending the updated AI model to the terminal.

[0100] In the above embodiment, after updating the AI ​​model, the first device sends the updated AI model to the terminal, thereby improving the accuracy of the AI ​​model.

[0101] In conjunction with some embodiments of the second aspect, in some embodiments, the first reference signal includes at least one of the following:

[0102] Positioning reference signal PRS;

[0103] Synchronous signal block;

[0104] Channel State Information Reference Signal CSI-RS.

[0105] In a third aspect, an embodiment of the present disclosure provides a monitoring method for an artificial intelligence (AI) model, the method being executed by a terminal and comprising:

[0106] receiving a first reference signal;

[0107] Processing the first reference signal using the current AI model to obtain an estimation result;

[0108] Based on the estimation result, determine whether to update the AI ​​model.

[0109] In the above embodiment, after receiving the first reference signal, the terminal uses the current AI model to process the first reference signal to obtain an estimation result, and determines whether to update the AI ​​model based on the estimation result, thereby enabling the terminal to monitor the AI ​​model.

[0110] In conjunction with some embodiments of the third aspect, in some embodiments, determining whether to update the AI ​​model based on the estimation result includes:

[0111] determining an error of the estimation result according to a reference result associated with the first reference signal;

[0112] When the error of the estimation result is greater than a threshold, it is determined to update the AI ​​model.

[0113] In conjunction with some embodiments of the third aspect, in some embodiments, the first reference signal is used for positioning, and the threshold is any one of the following:

[0114] a first value associated with a ground actual channel response error;

[0115] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0116] a third value associated with a time-of-arrival (TOA) error;

[0117] a fourth value associated with a time difference of arrival, TDOA, error;

[0118] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0119] In conjunction with some embodiments of the third aspect, in some embodiments, it is characterized in that,

[0120] The first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0121] The first reference signal is used for channel state information (CSI) measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0122] In conjunction with some embodiments of the third aspect, in some embodiments, at least one of the following is further included:

[0123] receiving the threshold value;

[0124] Determine the threshold value according to the agreement;

[0125] The threshold is determined according to the configuration information.

[0126] In the above embodiment, the terminal determines the threshold of the first reference signal through at least one method, thereby providing a condition for improving the efficiency of the terminal in determining the error of the estimation result.

[0127] In conjunction with some embodiments of the third aspect, in some embodiments, the method further includes:

[0128] Determine to update the AI ​​model and send second information, wherein the second information is used to trigger the first device to update the AI ​​model, wherein the first device is any one of the following: an access network device, a core network device, and a network server.

[0129] In the above embodiment, after determining to update the AI ​​model, the terminal sends the second information to the first device to trigger the first device to update the AI ​​model, thereby enabling the terminal to monitor the AI ​​model and improving the accuracy of the AI ​​model.

[0130] In conjunction with some embodiments of the third aspect, in some embodiments, the method further includes:

[0131] Determine not to update the AI ​​model, and send the estimation result to the first device.

[0132] In the above embodiment, when determining not to update the AI ​​model, the terminal sends the estimation result to the first device, thereby improving the reliability of the terminal monitoring the AI ​​model.

[0133] In combination with some embodiments of the third aspect, in some embodiments, the first reference signal includes at least one of the following: a positioning reference signal PRS, a synchronization signal block, and a channel state information reference signal CSI-RS.

[0134] In a fourth aspect, an embodiment of the present disclosure provides a monitoring method for an artificial intelligence (AI) model, the method being executed by a first device and comprising:

[0135] Send a first reference signal, where the first reference signal is used to assist the terminal in determining whether to update the current AI model.

[0136] In conjunction with some embodiments of the fourth aspect, in some embodiments, the method further includes:

[0137] receiving second information, wherein the second information is used to trigger an update of the AI ​​model;

[0138] Updating the AI ​​model to obtain an updated AI model;

[0139] Sending the updated AI model to the terminal.

[0140] In conjunction with some embodiments of the fourth aspect, in some embodiments, the method further includes:

[0141] An estimation result is received without initiating an operation to update the AI ​​model.

[0142] In a fifth aspect, an embodiment of the present disclosure provides a terminal, comprising:

[0143] a transceiver module, configured to receive a first reference signal;

[0144] a processing module, configured to process the first reference signal using a current AI model to obtain an estimation result;

[0145] The transceiver module is further used to send the estimation result to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

[0146] With reference to some embodiments of the fifth aspect, in some embodiments, the first reference signal includes at least one of the following:

[0147] Positioning reference signal PRS;

[0148] Synchronous signal block;

[0149] Channel State Information Reference Signal CSI-RS.

[0150] In conjunction with some embodiments of the fifth aspect, in some embodiments, the first device is selected from one of the following:

[0151] Access network equipment;

[0152] Core network equipment;

[0153] Network server.

[0154] In conjunction with some embodiments of the fifth aspect, in some embodiments, the processing module is further configured to:

[0155] When an updated AI model is received, the current AI model is replaced with the updated AI model.

[0156] In a sixth aspect, an embodiment of the present disclosure provides a first device, the first device including:

[0157] a transceiver module, configured to receive an estimation result sent by a terminal, wherein the estimation result is obtained by the terminal by processing the first reference signal using a current AI model;

[0158] A processing module is used to determine whether to update the AI ​​model based on the estimation result.

[0159] In conjunction with some embodiments of the sixth aspect, in some embodiments, the first device is an access network device, and the transceiver module is further configured to:

[0160] Sending the first reference signal to the terminal.

[0161] In conjunction with some embodiments of the sixth aspect, in some embodiments, the first device is a core network device or a server, and the transceiver module is further used to:

[0162] Sending first information to an access network device, where the first information is used to trigger the access network device to send a first reference signal to the terminal.

[0163] In conjunction with some embodiments of the sixth aspect, in some embodiments, the processing module is further configured to:

[0164] determining an error of the estimation result according to a reference result associated with the first reference signal;

[0165] When the error of the estimation result is greater than a threshold, the AI ​​model is updated.

[0166] In conjunction with some embodiments of the sixth aspect, in some embodiments, the first reference signal is used for positioning, and the threshold is any one of the following:

[0167] a first value associated with a ground actual channel response error;

[0168] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0169] a third value associated with a time-of-arrival (TOA) error;

[0170] a fourth value associated with a time difference of arrival, TDOA, error;

[0171] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0172] With reference to some embodiments of the sixth aspect, in some embodiments, the first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0173] The first reference signal is used for channel state information (CSI) measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0174] In conjunction with some embodiments of the sixth aspect, in some embodiments, the transceiver module is further configured to:

[0175] Sending the updated AI model to the terminal.

[0176] With reference to some embodiments of the sixth aspect, in some embodiments, the first reference signal includes at least one of the following:

[0177] Positioning reference signal PRS;

[0178] Synchronous signal block;

[0179] Channel State Information Reference Signal CSI-RS.

[0180] In a seventh aspect, an embodiment of the present disclosure provides a terminal, comprising:

[0181] a transceiver module, configured to receive a first reference signal;

[0182] a processing module, configured to process the first reference signal using a current AI model to obtain an estimation result;

[0183] The processing module is further used to determine whether to update the AI ​​model based on the estimation result.

[0184] In conjunction with some embodiments of the seventh aspect, in some embodiments, the processing module is further configured to:

[0185] determining an error of the estimation result according to a reference result associated with the first reference signal;

[0186] When the error of the estimation result is greater than a threshold, it is determined to update the AI ​​model.

[0187] In conjunction with some embodiments of the seventh aspect, in some embodiments, the first reference signal is used for positioning, and the threshold is any one of the following:

[0188] a first value associated with a ground actual channel response error;

[0189] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0190] a third value associated with a time-of-arrival (TOA) error;

[0191] a fourth value associated with a time difference of arrival, TDOA, error;

[0192] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0193] In conjunction with some embodiments of the seventh aspect, in some embodiments, the first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0194] The first reference signal is used for channel state information (CSI) measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0195] In conjunction with some embodiments of the seventh aspect, in some embodiments, the method further includes at least one of the following:

[0196] The transceiver module is further configured to receive the threshold;

[0197] The processing module is further configured to determine the threshold value according to the protocol;

[0198] The processing module is further configured to determine the threshold value according to configuration information.

[0199] In conjunction with some embodiments of the seventh aspect, in some embodiments, the transceiver module is further configured to:

[0200] Determine to update the AI ​​model and send second information, wherein the second information is used to trigger the first device to update the AI ​​model, wherein the first device is any one of the following: an access network device, a core network device, and a network server.

[0201] In conjunction with some embodiments of the seventh aspect, in some embodiments, the transceiver module is further configured to:

[0202] Determine not to update the AI ​​model, and send the estimation result to the first device.

[0203] In conjunction with some embodiments of the seventh aspect, in some embodiments, the first reference signal includes at least one of the following:

[0204] Positioning reference signal PRS;

[0205] Synchronous signal block;

[0206] Channel State Information Reference Signal CSI-RS.

[0207] In an eighth aspect, an embodiment of the present disclosure provides a first device, the first device including:

[0208] The transceiver module is used to send a first reference signal, wherein the first reference signal is used to assist the terminal in determining whether to update the current AI model.

[0209] In conjunction with some embodiments of the eighth aspect, in some embodiments, the method further includes:

[0210] The transceiver module is further configured to receive second information, wherein the second information is used to trigger an update of the AI ​​model;

[0211] A processing module, configured to update the AI ​​model to obtain an updated AI model;

[0212] The transceiver module is also used to send the updated AI model to the terminal.

[0213] In conjunction with some embodiments of the eighth aspect, in some embodiments, the transceiver module is further configured to:

[0214] An estimation result is received without initiating an operation to update the AI ​​model.

[0215] In the ninth aspect, an embodiment of the present disclosure proposes a terminal, which includes: one or more processors; wherein the processors are used to execute the optional implementation method of the monitoring method for the artificial intelligence AI model proposed in the first and third aspects.

[0216] In the tenth aspect, an embodiment of the present disclosure proposes a first device, which includes: one or more processors; wherein the processors are used to execute the optional implementation method of the monitoring method for the artificial intelligence AI model proposed in the second and fourth aspects.

[0217] In the eleventh aspect, an embodiment of the present disclosure proposes a communication system, which includes: a terminal and a first device; wherein the terminal is configured to execute the method described in the optional implementation of the first and third aspects, and the first device is configured to execute the method described in the optional implementation of the second and fourth aspects.

[0218] In the twelfth aspect, an embodiment of the present disclosure proposes a storage medium, which stores instructions. When the instructions are executed on a communication device, the communication device executes the method described in the optional implementation of the first, second, third and fourth aspects.

[0219] In the thirteenth aspect, an embodiment of the present disclosure proposes a program product. When the program product is executed by a communication device, the communication device executes the method described in the optional implementation of the first aspect, the second aspect, the third aspect, and the fourth aspect.

[0220] In the fourteenth aspect, an embodiment of the present disclosure proposes a computer program, which, when executed on a computer, enables the computer to execute the method described in the optional implementation of the first, second, third and fourth aspects.

[0221] In a fifteenth aspect, an embodiment of the present disclosure provides a chip or a chip system, which includes a processing circuit configured to execute the method described in the optional implementation of the first, second, third, and fourth aspects above.

[0222] It is understandable that the above-mentioned terminal, first device, communication system, storage medium, program product, computer program, chip or chip system are all used to perform the method proposed in the embodiment of the present disclosure. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects of the corresponding method and will not be repeated here.

[0223] The present disclosure provides a monitoring method for an artificial intelligence (AI) model. In some embodiments, the terms "monitoring method," "measurement and configuration method," "configuration method," and "communication method" for an artificial intelligence (AI) model are interchangeable; the terms "monitoring device," "measurement and configuration device," "configuration device," and "communication device" for an artificial intelligence (AI) model are interchangeable; and the terms "monitoring system," "measurement and configuration system," and "communication system" for an artificial intelligence (AI) model are interchangeable.

[0224] The embodiments of the present disclosure are not exhaustive and are merely illustrative of some embodiments, and are not intended to be a specific limitation on the scope of protection of the present disclosure. In the absence of contradiction, each step in a certain embodiment can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a certain embodiment can also be implemented as an independent embodiment, and the order of the steps in a certain embodiment can be arbitrarily exchanged. In addition, the optional implementation methods in a certain embodiment can be arbitrarily combined; in addition, the embodiments can be arbitrarily combined. For example, some or all steps of different embodiments can be arbitrarily combined, and a certain embodiment can be arbitrarily combined with the optional implementation methods of other embodiments.

[0225] In each embodiment of the present disclosure, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between the embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form a new embodiment based on their inherent logical relationships.

[0226] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure.

[0227] In the embodiments of the present disclosure, unless otherwise specified, elements expressed in the singular, such as "a", "an", "the", "above", "said", "the", "the", etc., may mean "one and only one", or "one or more", "at least one", etc. For example, when using articles such as "a", "an", "the" in English in translation, the noun following the article may be understood as a singular expression or a plural expression.

[0228] In the embodiments of the present disclosure, “plurality” refers to two or more.

[0229] In some embodiments, the terms "at least one," "one or more," "a plurality of," "multiple," etc. may be used interchangeably.

[0230] In some embodiments, descriptions such as "at least one of A and B," "A and / or B," "A in one case, B in another case," or "in response to one case A, in response to another case B" may include the following technical solutions depending on the situation: in some embodiments, A (A is executed independently of B); in some embodiments, B (B is executed independently of A); in some embodiments, execution is selected from A and B (A and B are selectively executed); and in some embodiments, A and B (both A and B are executed). The above is also applicable when there are more branches such as A, B, and C.

[0231] In some embodiments, "A or B" and other descriptions may include the following technical solutions depending on the situation: in some embodiments, A (A is executed independently of B); in some embodiments, B (B is executed independently of A); in some embodiments, execution is selected from A and B (A and B are selectively executed). The above is also applicable when there are more branches such as A, B, C, etc.

[0232] The prefixes such as "first" and "second" in the embodiments of the present disclosure are only used to distinguish different description objects and do not constitute any restriction on the position, order, priority, quantity or content of the description objects. For the statement of the description object, please refer to the description in the context of the claims or embodiments, and no unnecessary restriction should be constituted due to the use of prefixes. For example, if the description object is a "field", the ordinal number before the "field" in the "first field" and the "second field" does not limit the position or order between the "fields". "First" and "second" do not limit whether the "fields" they modify are in the same message, nor do they limit the order of the "first field" and the "second field". For another example, if the description object is a "level", the ordinal number before the "level" in the "first level" and the "second level" does not limit the priority between the "levels". For another example, the number of description objects is not limited by the ordinal number and can be one or more. Taking "first device" as an example, the number of "devices" can be one or more. In addition, the objects modified by different prefixes can be the same or different. For example, if the description object is "device", then the "first device" and the "second device" can be the same device or different devices, and their types can be the same or different; for another example, if the description object is "information", then the "first information" and the "second information" can be the same information or different information, and their contents can be the same or different.

[0233] In some embodiments, “including A,” “comprising A,” “used to indicate A,” and “carrying A” can be interpreted as directly carrying A or indirectly indicating A.

[0234] In some embodiments, terms such as "in response to...", "in response to determining...", "in the case of...", "at the time of...", "when...", "if...", "if...", etc. can be used interchangeably.

[0235] In some embodiments, terms such as "greater than", "greater than or equal to", "not less than", "more than", "more than or equal to", "not less than", "higher than", "higher than or equal to", "not less than", and "above" can be replaced with each other, and terms such as "less than", "less than or equal to", "not greater than", "less than", "less than or equal to", "not more than", "lower than", "lower than or equal to", "not higher than", and "below" can be replaced with each other.

[0236] In some embodiments, devices and equipment can be interpreted as physical or virtual, and their names are not limited to the names recorded in the embodiments. In some cases, they can also be understood as "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "chip", "chip system", "entity", "subject", etc.

[0237] In some embodiments, "network" can be interpreted as devices included in the network, such as access network equipment, core network equipment, etc.

[0238] In some embodiments, "access network device (AN device)" may also be referred to as "radio access network device (RAN device)", "base station (BS)", "radio base station", "fixed station", and in some embodiments may also be understood as "node", "access point", "transmission point (TP)", "reception point (RP)", "transmission and / or reception point (TRP)", "panel", "antenna panel", "antenna array", "cell", "macrocell", "smallcell", "femtocell", "picocell", "sector", "cell group", "serving cell", "carrier", "component carrier", "bandwidth part (BWP)", etc.

[0239] In some embodiments, "terminal" or "terminal device" may be referred to as "user equipment (UE)", "user terminal" "mobile station (MS)", "mobile terminal (MT)", subscriber station (subscriber station), mobile unit (mobile unit), subscriber unit (subscribe runit), wireless unit (wireless unit), remote unit (remote unit), mobile device (mobile device), wireless device (wireless device), wireless communication device (wireless communication device), remote device (remoted device), mobile subscriber station (mobile subscriber station), access terminal (access terminal), mobile terminal (mobile terminal), wireless terminal (wireless terminal), remote terminal (remote terminal), handset (handset), user agent (user agent), mobile client (mobile client), client (client), etc.

[0240] In some embodiments, obtaining data, information, etc. may comply with the laws and regulations of the country where the data is obtained.

[0241] In some embodiments, data, information, etc. may be obtained with the user's consent.

[0242] FIG1A is a schematic diagram showing the architecture of a communication system according to an embodiment of the present disclosure.

[0243] As shown in FIG1A , a communication system 1100 includes a terminal 1101 and a first device 1102 .

[0244] In some embodiments, the terminal 1101 includes, for example, a mobile phone, a wearable device, an Internet of Things device, a car with communication function, a smart car, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical surgery, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, and at least one of a wireless terminal device in a smart home, but is not limited thereto.

[0245] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0246] In some embodiments, the access network device is, for example, a node or device that accesses a terminal to a wireless network. The access network device may include an evolved NodeB (eNB), a next generation evolved NodeB (ng-eNB), a next generation NodeB (gNB), a node B (NB), a home node B (HNB), a home evolved nodeB (HeNB), a wireless backhaul device, a radio network controller (RNC), a base station controller (BSC), a base transceiver station (BTS), a base band unit (BBU), a mobile switching center, a base station in a 6G communication system, an open base station (Open RAN), a cloud base station (Cloud RAN), a base station in other communication systems, and at least one of an access node in a Wi-Fi system, but is not limited thereto.

[0247] In some embodiments, the technical solution of the present disclosure can be applied to the Open RAN architecture. In this case, the interfaces between or within the access network devices involved in the embodiments of the present disclosure can be transformed into internal interfaces of the Open RAN, and the processes and information interactions between these internal interfaces can be implemented through software or programs.

[0248] In some embodiments, the access network device can be composed of a centralized unit (CU) and a distributed unit (DU), where the CU can also be called a control unit. The CU-DU structure can be used to split the protocol layer of the access network device, with the functions of some protocol layers centrally controlled by the CU, and the functions of the remaining part or all of the protocol layers distributed in the DU, which is centrally controlled by the CU, but is not limited to this.

[0249] In some embodiments, a core network device may be a device including one or more network elements, or may be multiple devices or device groups, each including all or part of the one or more network elements. The network element may be virtual or physical. The core network may include, for example, at least one of an Evolved Packet Core (EPC), a 5G Core Network (5GCN), and a Next Generation Core (NGC).

[0250] In some embodiments, the core network device may be an existing core network element, or may be a new core network element, for example, a server located on the core network side, etc., which is not limited in this disclosure.

[0251] In some embodiments, the network server can be any type of processor connected to the network and used for model training or updating, such as a computer, etc., which is not limited in this disclosure.

[0252] It can be understood that the communication system described in the embodiment of the present disclosure is for the purpose of more clearly illustrating the technical solution of the embodiment of the present disclosure, and does not constitute a limitation on the technical solution proposed in the embodiment of the present disclosure. Ordinary technicians in this field can know that with the evolution of the system architecture and the emergence of new business scenarios, the technical solution proposed in the embodiment of the present disclosure is also applicable to similar technical problems.

[0253] The following embodiments of the present disclosure may be applied to the communication system 1100 shown in FIG1A , or a portion thereof, but are not limited thereto. The entities shown in FIG1A are illustrative only. The communication system may include all or part of the entities shown in FIG1A , or may include other entities other than those shown in FIG1A . The number and form of the entities may be arbitrary. The entities may be physical or virtual. The connection relationships between the entities are illustrative only. The entities may be connected or disconnected, and the connection may be in any manner, including direct or indirect, wired or wireless.

[0254] The embodiments of the present disclosure can be applied to Long Term Evolution (LTE), LTE-Advanced (LTE-A), LTE-Beyond (LTE-B), SUPER3G, IMT-Advanced, 4th generation mobile communication system (4G), 5th generation mobile communication system (5G), 5G New Radio (NR), Future Radio Access (FRA), New Radio Access Technology (RAT), New Radio (NR), New Radio Access (NX), Future Generation Radio Access (FX), Global System for Mobile Communications (GSM (registered trademark)), CDMA2000, Ultra Mobile Broadband (Ultra Mobile Broadband), and other technologies. Broadband (UMB), IEEE802.11 (Wi-Fi (registered trademark)), IEEE802.16 (WiMAX (registered trademark)), IEEE802.20, Ultra-WideBand (UWB), Bluetooth (registered trademark), Public Land Mobile Network (PLMN) network, Device-to-Device (D2D) system, Machine-to-Machine (M2M) system, Internet of Things (IoT) system, Vehicle-to-Everything (V2X), systems using other communication methods, and next-generation systems based on them. In addition, multiple systems can also be combined (for example, a combination of LTE or LTE-A with 5G, etc.) for application.

[0255] In 5G and the sixth generation mobile communication system (6G), artificial intelligence-based positioning can use a framework as shown in FIG1B to improve positioning accuracy. FIG1B is a schematic diagram of a framework of a monitoring method for an artificial intelligence AI model shown in an embodiment of the present disclosure.

[0256] The following metrics / methods can be used to monitor Artificial Intelligence (AI) / Machine Learning (ML) models in the lifecycle management of each use case:

[0257] Monitoring based on inference accuracy can include metrics related to intermediate key performance indicators (KPIs);

[0258] System performance-based monitoring can include indicators related to system performance KPIs;

[0259] Other monitoring solutions can include at least the following two options:

[0260] Monitoring based on data distribution;

[0261] Monitor according to applicable conditions.

[0262] Currently, in Release 18 (Rel-18), the terminal behavior monitored by the AI ​​model has not been determined.

[0263] FIG2A is an interactive diagram of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG2A , an embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in a terminal 1101 and a first device 1102. The method includes:

[0264] Step S2101 : The first device 1102 sends a first reference signal to the terminal 1101 .

[0265] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0266] In some embodiments, when the first device 1102 is an access network device, the first device 1102 may directly send the first reference signal to the terminal 1101 .

[0267] In some embodiments, when the first device 1102 is a core network device or a server, the first device 1102 may send the first information to the access network device, and then send the first reference signal to the terminal 1101 through the access network device.

[0268] In some embodiments, the first information is used to trigger the access network device to send a first reference signal to the terminal 1101.

[0269] In some embodiments, the first reference signal may include at least one of the following: a positioning reference signal PRS, a synchronization signal block, and a channel state information reference signal CSI-RS, which is not limited in the present disclosure.

[0270] In some embodiments, the terms "PRS", "Positioning Reference Signal", "Positioning Reference Signal" and the like may be used interchangeably.

[0271] In some embodiments, terms such as "synchronization signal block", "SSB", "Synchronization Signal Block" and the like can be used interchangeably.

[0272] In some embodiments, terms such as "CSI-RS", "Channel State Information Reference Signal", and "Channel State Information Reference Signal" can be used interchangeably.

[0273] In some embodiments, the terminal 1101 receives a first reference signal sent by the first device 1102 .

[0274] In step S2102 , the terminal 1101 processes the first reference signal using the current AI model to obtain an estimation result.

[0275] In some embodiments, terms such as "AI", "artificial intelligence", and "Artificial Intelligence" can be used interchangeably.

[0276] In some embodiments, the current AI model may be an AI positioning model, or may be an AI model for beam prediction, or may be an AI model for CSI prediction, etc., which is not limited in this disclosure.

[0277] In some embodiments, the current AI model used by the terminal 1101 may be an AI model stored in the terminal 1101, or may be an AI model sent by the first device 1102, etc. This disclosure does not limit this.

[0278] In some embodiments, when the current AI model used by the terminal 1101 is valid, the terminal 1101 may directly use the current AI model to process the first reference signal.

[0279] In some embodiments, when the current AI model used by terminal 1101 fails, in order to ensure the accuracy of the AI ​​model, the current AI model needs to be updated so that terminal 1101 can use the updated AI model to process the first reference signal.

[0280] In some embodiments, after receiving the first reference signal, the terminal 1101 may process the measurement result of the first reference signal based on the current AI model to obtain an estimation result.

[0281] In some embodiments, when the terminal 1101 processes the first reference signal using the current AI model, it can obtain at least one physical layer estimation result, such as CIR, TDoA, RSTD, PRS RSRP, etc., which is not limited in this disclosure.

[0282] In some embodiments, the terms "CIR", "channel impulse response", "Channel Impulse Response", "channel impulse response" and the like can be used interchangeably.

[0283] In some embodiments, the terms "TDoA", "Time Difference Of Arrival", "Time Difference Of Arrival" and the like may be used interchangeably.

[0284] In some embodiments, terms such as “RSTD”, “Reference Signal Time Difference”, and “Reference Signal Time Difference” may be used interchangeably.

[0285] In some embodiments, the terms "RSRP", "reference signal received power", "reference signal received power" and the like can be used interchangeably.

[0286] Step S2103 : The terminal 1101 sends the estimation result to the first device 1102 .

[0287] In some embodiments, the estimation result can be used to assist the first device 1102 in determining whether to update the AI ​​model.

[0288] In some embodiments, if the current AI model is updated, the terminal 1101 may replace the current AI model with the updated AI model to process the reference signal, which is not limited in the present disclosure.

[0289] In some embodiments, if the current AI model is not updated, the terminal 1101 may retain the current AI model and use the current AI model to process the reference signal, which is not limited in this disclosure.

[0290] In some embodiments, the first device 1102 receives the estimation result sent by the terminal 1101 .

[0291] Step S2104 : The first device 1102 determines an error of the estimation result according to the reference result associated with the first reference signal.

[0292] In some embodiments, the reference result associated with the first reference signal may be pre-configured, which is not limited in this disclosure.

[0293] In some embodiments, after receiving the estimation result, the first device 1102 may determine an error of the estimation result based on a reference result associated with the first reference signal.

[0294] Step S2105: When the error of the estimation result is greater than the threshold, the first device 1102 updates the AI ​​model.

[0295] In some embodiments, the threshold value may be a pre-configured error critical value of the estimation result used to assist the first device 1102 in determining whether to update the AI ​​model, and this disclosure does not limit this.

[0296] In some embodiments, when the AI ​​model is an AI positioning model, the first reference signal may be a reference signal used for positioning. In this case, when the first reference signal is used for positioning, the threshold may be any one of the following: a first value associated with the actual ground channel response error; a second value associated with the physical reference signal received power (RSRP) measurement error; a third value associated with the time of arrival (TOA) error; a fourth value associated with the time difference of arrival (TDOA) error; or a fifth value associated with the reference signal time difference (RSTD) error, although this disclosure does not limit this.

[0297] In some embodiments, the terms "TOA", "Time of Arrival", "Time Of Arrival" and the like can be used interchangeably.

[0298] In some embodiments, when the AI ​​model is an AI model for beam prediction, the first reference signal may be a reference signal used for beam prediction. In this case, when the first reference signal is used for beam prediction, the threshold may be a value used to indicate the beam prediction result. For example, this may be a value associated with an actual beam index prediction error, a value associated with an actual ground channel response error, a value associated with an RSRP measurement error, and so on, although this disclosure is not limited thereto.

[0299] In some embodiments, when the AI ​​model is an AI model for CSI prediction, the first reference signal may be a reference signal for CSI measurement. In this case, when the first reference signal is used for channel state information (CSI) measurement, the threshold may be a value used to indicate the channel state information estimation result. For example, it may be a value associated with the actual beam index prediction error; a value associated with the actual ground channel response error; a value associated with model complexity, etc., which is not limited in this disclosure.

[0300] In some embodiments, when the error of the estimation result is greater than a threshold, it can be considered that the accuracy of the current AI model is low. At this time, in order to improve the accuracy of the AI ​​model, the first device 1102 can update the AI ​​model.

[0301] Step S2106: The first device 1102 sends the updated AI model to the terminal 1101.

[0302] In some embodiments, after updating the AI ​​model, the first device 1102 may send the updated AI model to the terminal 1101 so that the terminal 1101 uses the updated AI model to process the reference signal.

[0303] In some embodiments, after the terminal 1101 receives the updated AI model sent by the first device 1102, it can use the updated AI model to replace the current AI model to process the reference signal, thereby improving the accuracy of the AI ​​model.

[0304] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S2101 to S2106. For example, steps S2101 and S2102 may be implemented as independent embodiments, and step S2103 may be implemented as an independent embodiment, etc., but the present disclosure is not limited thereto.

[0305] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0306] In this embodiment, the first device first sends a first reference signal to the terminal, and then the terminal processes the first reference information to obtain an estimation result, and sends the estimation result to the first device. Thereafter, the first device determines the error of the estimation result based on the reference result associated with the first reference signal. When the error of the estimation result is greater than the threshold, the first device updates the AI ​​model, and finally sends the updated AI model to the terminal, thereby realizing joint monitoring of the artificial intelligence AI model by the terminal and the first device, and improving the accuracy of the AI ​​model.

[0307] FIG2B is an interactive diagram of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG2B , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in a terminal 1101 and a first device 1102, and includes:

[0308] Step S2201 : The first device 1102 sends first reference information to the terminal 1101 .

[0309] In some embodiments, the first reference signal is used to assist terminal 1101 in determining whether to update the current AI model.

[0310] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0311] In step S2202 , the terminal 1101 processes the first reference signal using the current AI model to obtain an estimation result.

[0312] For a detailed description of steps S2201 - S2202 , please refer to steps S2101 - S2102 in the embodiment shown in FIG2A , which will not be repeated here.

[0313] In step S2203 , the terminal 1101 determines an error in the estimation result according to the reference result associated with the first reference signal.

[0314] In some embodiments, the specific implementation manner in which the terminal 1101 determines the error of the estimation result according to the reference result associated with the first reference signal is referred to the description of the optional implementation manner of step S2104 in Figure 2A, which is not repeated here.

[0315] Step S2204: When the error of the estimation result is greater than the threshold, the terminal 1101 determines to update the AI ​​model.

[0316] In some embodiments, the specific implementation of the threshold value refers to the relevant description of the optional implementation of step S2105 in Figure 2A, which will not be repeated here.

[0317] In some embodiments, the terminal 1101 may determine the threshold by receiving a threshold. For example, the terminal 1101 may receive a threshold sent by a core network device, or may receive a threshold sent by a network server to determine the threshold, which is not limited in this disclosure.

[0318] In some embodiments, the terminal 1101 may also determine the threshold value according to a protocol agreement, which is not limited in this disclosure.

[0319] In some embodiments, the terminal 1101 may also determine a threshold value based on configuration information, which is not limited in this disclosure.

[0320] In some embodiments, when the error of the estimation result is greater than a threshold, it can be considered that the accuracy of the current AI model is low. At this time, in order to improve the accuracy of the AI ​​model, the terminal 1101 can determine to update the AI ​​model.

[0321] In some embodiments, when the terminal 1101 determines to update the current AI model, the terminal 1101 can send second information to the first device 1102 to trigger the first device 1102 to update the AI ​​model.

[0322] In some embodiments, when the terminal 1101 determines not to update the current AI model, the terminal 1101 can directly send the estimation result to the first device 1102, so that the first device 1102 does not start the AI ​​model update operation.

[0323] Step S2205 , the terminal 1101 sends second information to the first device 1102 .

[0324] In some embodiments, the second information is used to trigger the first device 1102 to update the AI ​​model.

[0325] In some embodiments, after determining to update the AI ​​model, the terminal 1101 may send second information to the first device 1102 .

[0326] In some embodiments, the first device 1102 receives the second information sent by the terminal 1101 .

[0327] In step S2206, the first device 1102 updates the AI ​​model to obtain an updated AI model.

[0328] In some embodiments, after receiving the second information, the first device 1102 may update the AI ​​model to obtain an updated AI model.

[0329] Step S2207: The first device 1102 sends the updated AI model to the terminal 1101.

[0330] In some embodiments, after receiving the updated AI model sent by the first device 1102 , the terminal 1101 may replace the current AI model with the updated AI model.

[0331] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S2201 to S2207. For example, step S2201 may be implemented as an independent embodiment, step S2202 may be implemented as an independent embodiment, and steps S2201+S2202 may be implemented as independent embodiments, but are not limited thereto.

[0332] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0333] In this embodiment, the first device first sends a first reference signal to the terminal, and then the terminal processes the first reference signal to obtain an estimation result, and determines the error of the estimation result based on the reference result associated with the first reference signal. When the error of the estimation result is greater than the threshold, the terminal can determine to update the AI ​​model and send second information to the first device to request the first device to update the AI ​​model. After that, the first device updates the AI ​​model and sends the updated AI model to the terminal, thereby realizing the terminal's monitoring of the artificial intelligence AI model and improving the accuracy of the AI ​​model.

[0334] FIG2C is an interactive diagram of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG2C , an embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in a terminal 1101 and a first device 1102. The method includes:

[0335] Step S2301 : The first device 1102 sends first reference information to the terminal 1101 .

[0336] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0337] In step S2302 , the terminal 1101 processes the first reference signal using the current AI model to obtain an estimation result.

[0338] In step S2303 , the terminal 1101 determines an error in the estimation result based on the reference result associated with the first reference signal.

[0339] For a detailed description of steps S2301 to S2303, reference may be made to steps S2201 to S2203 in the embodiment shown in FIG2B , which will not be repeated here.

[0340] Step S2304: When the error of the estimation result is less than or equal to the threshold, the terminal 1101 determines not to update the AI ​​model.

[0341] In some embodiments, when the error of the estimation result is less than or equal to a threshold, it can be considered that the accuracy of the current AI model is high. At this time, the terminal 1101 can determine not to update the AI ​​model.

[0342] In some embodiments, when the terminal 1101 determines not to update the current AI model, the terminal 1101 may retain the current AI model.

[0343] Step S2305 : The terminal 1101 sends the estimation result to the first device 1102 .

[0344] In some embodiments, if it is determined that the AI ​​model is not to be updated, the terminal 1101 may send the estimation result to the first device 1102 .

[0345] In some embodiments, the first device 1102 receives the estimation result sent by the terminal 1101 .

[0346] In some embodiments, when the first device 1102 receives the estimation result, it may be considered that the accuracy of the current AI model is high. At this time, the first device 1102 may not start the operation of updating the AI ​​model.

[0347] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S2301 to S2305. For example, steps S2301 and S2302 may be implemented as independent embodiments, and step S2303 may be implemented as an independent embodiment, etc., but the present disclosure is not limited thereto.

[0348] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0349] In this embodiment, the first device first sends a first reference signal to the terminal, and then the terminal processes the first reference signal to obtain an estimation result, and determines the error of the estimation result based on the reference result associated with the first reference signal. When the error of the estimation result is less than or equal to the threshold, the terminal can determine not to update the AI ​​model and send the estimation result to the first device. After receiving the estimation result, the first device does not update the AI ​​model, thereby realizing the terminal's monitoring of the artificial intelligence AI model and improving the accuracy of the AI ​​model.

[0350] FIG3A is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG3A , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in a terminal 1101 and includes:

[0351] Step S3101: Receive a first reference signal sent by the first device 1102.

[0352] In some embodiments, the terminal 1101 receives a first reference signal sent by the first device 1102 .

[0353] Step S3102: Use the current AI model to process the first reference signal to obtain an estimation result.

[0354] Step S3103 : Send the estimation result to the first device 1102 .

[0355] For a detailed description of steps S3101 - S3103 , please refer to steps S2101 - S2103 in the embodiment shown in FIG2A , which will not be repeated here.

[0356] Step S3104: Receive the updated AI model sent by the first device 1102.

[0357] In some embodiments, the terminal 1101 receives the updated AI model sent by the first device 1102.

[0358] For a detailed description of step S3104, please refer to step S2106 in the embodiment shown in FIG2A , which will not be repeated here.

[0359] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S3101 to S3104. For example, step S3101 may be implemented as an independent embodiment, step S3102 may be implemented as an independent embodiment, and steps S3101+S3102 may be implemented as independent embodiments, but the present invention is not limited thereto.

[0360] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0361] In this embodiment, after receiving the first reference signal sent by the first device, the terminal processes the first reference signal to obtain an estimation result, then sends the estimation result to the first device, and finally receives the updated AI model sent by the first device, thereby improving the accuracy of the AI ​​model.

[0362] FIG3B is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG3B , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in terminal 1101 and includes:

[0363] Step S3201: Receive a first reference signal sent by the first device 1102.

[0364] In some embodiments, the terminal 1101 receives a first reference signal sent by the first device 1102 .

[0365] Step S3202: Use the current AI model to process the first reference signal to obtain an estimation result.

[0366] Step S3203: Determine an error in the estimation result according to the reference result associated with the first reference signal.

[0367] Step S3204: When the error of the estimation result is greater than the threshold, determine to update the AI ​​model.

[0368] Step S3205: Send second information to the first device 1102.

[0369] For a detailed description of steps S3201 to S3205, reference may be made to steps S2201 to S2205 in the embodiment shown in FIG2B , which will not be repeated here.

[0370] Step S3206: Receive the updated AI model sent by the first device 1102.

[0371] For a detailed description of step S3206, please refer to step S2207 in the embodiment shown in FIG2B , which will not be repeated here.

[0372] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S3201 to S3206. For example, step S3201 may be implemented as an independent embodiment, step S3202 may be implemented as an independent embodiment, and steps S3201+S3202 may be implemented as independent embodiments, but are not limited thereto.

[0373] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0374] In this embodiment, after receiving the first reference signal sent by the first device, the terminal processes the first reference signal to obtain an estimation result, and then determines the error of the estimation result based on the reference result associated with the first reference signal. When the error of the estimation result is greater than the threshold, it determines to update the AI ​​model, and then receives the updated AI model sent by the first device, thereby enabling the terminal to monitor the AI ​​model and improving the accuracy of the AI ​​model.

[0375] FIG3C is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG3C , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in terminal 1101 and includes:

[0376] Step S3301: Receive a first reference signal sent by the first device 1102.

[0377] In some embodiments, the terminal 1101 receives a first reference signal sent by the first device 1102 .

[0378] Step S3302: Use the current AI model to process the first reference signal to obtain an estimation result.

[0379] Step S3303: Determine an error in the estimation result according to the reference result associated with the first reference signal.

[0380] Step S3304: When the error of the estimation result is less than or equal to the threshold, determine not to update the AI ​​model.

[0381] Step S3305: Send the estimation result to the first device 1102.

[0382] In some embodiments, the terminal 1101 sends the estimation result to the first device 1102 .

[0383] For a detailed description of steps S3301 to S3305, please refer to steps S2301 to S2305 in the embodiment shown in FIG2C , which will not be repeated here.

[0384] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S3301 to S3305. For example, step S3301 may be implemented as an independent embodiment, step S3302 may be implemented as an independent embodiment, and steps S3301+S3302 may be implemented as independent embodiments, but are not limited thereto.

[0385] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0386] In this embodiment, after receiving the first reference signal sent by the first device, the terminal processes the first reference signal to obtain an estimation result, and then determines the error of the estimation result based on the reference result associated with the first reference signal. When the error of the estimation result is less than or equal to the threshold, it is determined not to update the AI ​​model, and the estimation result is sent to the first device, thereby realizing the terminal monitoring of the AI ​​model.

[0387] FIG3D is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG3D , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in terminal 1101 and includes:

[0388] Step S3401: Receive a first reference signal.

[0389] In some embodiments, the first reference signal includes at least one of the following:

[0390] Positioning reference signal PRS;

[0391] Synchronous signal block;

[0392] Channel State Information Reference Signal CSI-RS.

[0393] Step S3402: Use the current AI model to process the first reference signal to obtain an estimation result.

[0394] Step S3403: Send the estimation result to the first device.

[0395] In some embodiments, the estimation result is used to assist the first device 1102 in determining whether to update the AI ​​model.

[0396] In some embodiments, the first device is selected from one of the following:

[0397] Access network equipment;

[0398] Core network equipment;

[0399] Network server.

[0400] In some embodiments, the method further comprises:

[0401] When an updated AI model is received, the current AI model is replaced with the updated AI model.

[0402] For a detailed description of steps S3401 - S3403 , please refer to the above embodiment description.

[0403] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S3401 to S3403. For example, step S3401 may be implemented as an independent embodiment, step S3402 may be implemented as an independent embodiment, and steps S3401+S3402 may be implemented as independent embodiments, but are not limited thereto.

[0404] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0405] In this embodiment, after receiving the first reference signal, the terminal uses the current AI model to process the first reference signal, obtains an estimation result, and then sends the estimation result to the first device, thereby providing conditions for the terminal and the first device to jointly monitor the AI ​​model.

[0406] FIG3E is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG3E , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used in terminal 1101 and includes:

[0407] Step S3501: Receive a first reference signal.

[0408] In some embodiments, the first reference signal is used for positioning, and the threshold is any one of the following:

[0409] a first value associated with a ground actual channel response error;

[0410] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0411] a third value associated with a time-of-arrival (TOA) error;

[0412] a fourth value associated with a time difference of arrival, TDOA, error;

[0413] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0414] In some embodiments, the first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0415] The first reference signal is used for channel state information CSI measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0416] In some embodiments, at least one of the following is further included:

[0417] Receiving threshold;

[0418] Determine the threshold value according to the agreement;

[0419] Determine the threshold based on the configuration information.

[0420] In some embodiments, the first reference signal includes at least one of the following:

[0421] Positioning reference signal PRS;

[0422] Synchronous signal;

[0423] Channel State Information Reference Signal CSI-RS.

[0424] Step S3502: Use the current AI model to process the first reference signal to obtain an estimation result.

[0425] Step S3503: Determine whether to update the AI ​​model based on the estimation result.

[0426] In some embodiments, determining whether to update the AI ​​model based on the estimation result includes:

[0427] determining an error in the estimation result based on a reference result associated with the first reference signal;

[0428] When the error of the estimation result is greater than a threshold, it is determined to update the AI ​​model.

[0429] In some embodiments, the method further comprises:

[0430] Determine to update the AI ​​model and send second information, wherein the second information is used to trigger the first device to update the AI ​​model, wherein the first device is any one of the following: an access network device, a core network device, and a network server.

[0431] In some embodiments, the method further comprises:

[0432] Determine not to update the AI ​​model, and send the estimation result to the first device.

[0433] For a detailed description of steps S3501-S3503, please refer to the above embodiment description.

[0434] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S3501 to S3503. For example, step S3501 may be implemented as an independent embodiment, step S3502 may be implemented as an independent embodiment, and steps S3501+S3502 may be implemented as independent embodiments, but are not limited thereto.

[0435] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0436] In this embodiment, after receiving the first reference signal, the terminal uses the current AI model to process the first reference signal to obtain an estimation result, and determines whether to update the AI ​​model based on the estimation result, thereby enabling the terminal to monitor the AI ​​model.

[0437] FIG4A is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG4A , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used on a first device 1102 and includes:

[0438] Step S4101: Send a first reference signal to terminal 1101.

[0439] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0440] In some embodiments, the first device 1102 sends a first reference signal to the terminal 1101 .

[0441] For a detailed description of step S4101, please refer to step S2101 in the embodiment shown in FIG2A , which will not be repeated here.

[0442] Step S4102 , receiving the estimation result sent by terminal 1101 .

[0443] In some embodiments, the first device 1102 receives the estimation result sent by the terminal 1101 .

[0444] Step S4103: Determine an error in the estimation result according to the reference result associated with the first reference signal.

[0445] Step S4104: When the error of the estimation result is greater than the threshold, update the AI ​​model.

[0446] Step S4105: Send the updated AI model to terminal 1101.

[0447] In some embodiments, the first device 1102 sends the updated AI model to the terminal 1101.

[0448] For a detailed description of steps S4102 to S4105 , please refer to steps S2103 to S2106 in the embodiment shown in FIG2A , which will not be repeated here.

[0449] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S4101 to S4105. For example, step S4101 may be implemented as an independent embodiment, step S4102 may be implemented as an independent embodiment, and steps S4101+S4102 may be implemented as independent embodiments, but are not limited thereto.

[0450] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0451] In this embodiment, after the first device sends a first reference signal to the terminal, it receives the estimation result sent by the terminal, and then determines the error of the estimation result based on the reference result associated with the first reference signal. When the error of the estimation result is greater than the threshold, the AI ​​model is updated and the updated AI model is sent to the terminal, thereby realizing joint monitoring of the AI ​​model by the terminal and the first device and improving the accuracy of the AI ​​model.

[0452] FIG4B is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG4B , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used on a first device 1102 and includes:

[0453] Step S4201: Send a first reference signal to terminal 1101.

[0454] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0455] In some embodiments, the first device 1102 sends a first reference signal to the terminal 1101 .

[0456] For a detailed description of step S4201, please refer to step S2201 in the embodiment shown in FIG2B , which will not be repeated here.

[0457] Step S4202: Receive the second information sent by terminal 1101.

[0458] In some embodiments, the first device 1102 receives the second information sent by the terminal 1101 .

[0459] Step S4203: Update the AI ​​model to obtain an updated AI model.

[0460] Step S4204: Send the updated AI model to terminal 1101.

[0461] In some embodiments, the first device 1102 sends the updated AI model to the terminal 1101.

[0462] For a detailed description of steps S4202 to S4204, please refer to steps S2205 to S2207 in the embodiment shown in FIG2B , which will not be repeated here.

[0463] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S4201 to S4204. For example, step S4201 may be implemented as an independent embodiment, step S4202 may be implemented as an independent embodiment, and steps S4201+S4202 may be implemented as independent embodiments, but are not limited thereto.

[0464] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0465] In this embodiment, after sending the first reference signal to the terminal, the first device receives the second information sent by the terminal, updates the AI ​​model, and sends the updated AI model to the terminal, thereby providing conditions for implementing the terminal monitoring AI model and improving the accuracy of the AI ​​model.

[0466] FIG4C is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG4C , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used on a first device 1102 and includes:

[0467] Step S4301: Send a first reference signal to terminal 1101.

[0468] In some embodiments, the first device 1102 may be an access network device, or may also be a core network device, or may also be a network server, which is not limited in the present disclosure.

[0469] In some embodiments, the first device 1102 sends a first reference signal to the terminal 1101 .

[0470] For a detailed description of step S4301, please refer to step S2301 in the embodiment shown in FIG2C , which will not be repeated here.

[0471] Step S4302: Receive the estimation result sent by terminal 1101.

[0472] In some embodiments, the first device 1102 receives the estimation result sent by the terminal 1101 .

[0473] For a detailed description of steps S4302-S4302, please refer to steps S2305-S2305 in the embodiment shown in FIG2C , which will not be repeated here.

[0474] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S4301 and S4302. For example, step S4301 may be implemented as an independent embodiment, step S4302 may be implemented as an independent embodiment, and steps S4301+S4302 may be implemented as independent embodiments, but the present invention is not limited thereto.

[0475] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0476] In this embodiment, the first device sends a first reference signal to the terminal, and then after receiving the estimation result sent by the terminal, the first device does not update the AI ​​model, thereby improving the accuracy of the AI ​​model.

[0477] FIG4D is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG4D , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used on a first device 1102 and includes:

[0478] Step S4401: receiving the estimation result sent by the terminal.

[0479] In some embodiments, the estimation result is obtained by the terminal processing the first reference signal using the current AI model.

[0480] In some embodiments, the first device is an access network device, and the method further includes:

[0481] A first reference signal is sent to the terminal.

[0482] In some embodiments, the first device is a core network device or a server, and the method further includes:

[0483] First information is sent to the access network device, where the first information is used to trigger the access network device to send a first reference signal to the terminal.

[0484] In some embodiments, the first reference signal includes at least one of the following:

[0485] Positioning reference signal PRS;

[0486] Synchronous signal block;

[0487] Channel State Information Reference Signal CSI-RS.

[0488] Step S4402: Determine whether to update the AI ​​model based on the estimation results.

[0489] In some embodiments, determining whether to update the AI ​​model based on the estimation result includes:

[0490] determining an error in the estimation result based on a reference result associated with the first reference signal;

[0491] When the error of the estimated result is greater than the threshold, the AI ​​model is updated.

[0492] In some embodiments, the first reference signal is used for positioning, and the threshold is any one of the following:

[0493] a first value associated with a ground actual channel response error;

[0494] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0495] a third value associated with a time-of-arrival (TOA) error;

[0496] a fourth value associated with a time difference of arrival, TDOA, error;

[0497] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0498] In some embodiments, the first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0499] The first reference signal is used for channel state information CSI measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0500] In some embodiments, the method further comprises:

[0501] Send the updated AI model to the terminal.

[0502] For a detailed description of steps S4401 - S4402 , please refer to the above embodiment description.

[0503] The monitoring method for an artificial intelligence (AI) model according to the embodiments of the present disclosure may include at least one of steps S4401 and S4402. For example, step S4401 may be implemented as an independent embodiment, step S4402 may be implemented as an independent embodiment, and steps S4401+S4402 may be implemented as independent embodiments, but the present invention is not limited thereto.

[0504] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0505] In this embodiment, the first device determines whether to update the AI ​​model based on the estimation result received from the terminal, thereby providing conditions for the terminal and the first device to jointly monitor the AI ​​model and improving the accuracy of the AI ​​model.

[0506] FIG4E is a flow chart of a method for monitoring an artificial intelligence (AI) model according to an embodiment of the present disclosure. As shown in FIG4E , the embodiment of the present disclosure relates to a method for monitoring an artificial intelligence (AI) model, which is used on a first device 1102 and includes:

[0507] Step S4501: Send a first reference signal.

[0508] In some embodiments, the first reference signal is used to assist the terminal in determining whether to update the current AI model.

[0509] In some embodiments, the method further comprises:

[0510] receiving second information, wherein the second information is used to trigger an update of the AI ​​model;

[0511] Update the AI ​​model to obtain an updated AI model;

[0512] Send the updated AI model to the terminal.

[0513] In some embodiments, the method further comprises:

[0514] Receive the estimation results without initiating any update of the AI ​​model.

[0515] For a detailed description of step S4501, please refer to the above embodiment description.

[0516] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0517] In this embodiment, the first device sends a first reference signal to the terminal, thereby providing conditions for the terminal to monitor the AI ​​model.

[0518] The following is an exemplary introduction to the above method.

[0519] This disclosure addresses, to a certain extent, the issue of how to define terminal behaviors and test parameters in artificial intelligence (AI) model monitoring methods. Optional implementation solutions are as follows:

[0520] The present disclosure relates to a monitoring method for an artificial intelligence (AI) model. Taking a first device as an access network device, a core network device, or a network server as an example, the method includes:

[0521] Embodiment 1: Jointly monitoring an AI model through a first device and a user equipment (UE).

[0522] A method for jointly monitoring an AI model by a first device and a UE is shown in FIG5A . FIG5A is an interactive schematic diagram of the UE and the first device jointly monitoring the AI ​​model according to an embodiment of the present disclosure.

[0523] Step 1): The UE reports the measurement results to the first device for AI model training.

[0524] Step 2) When the model passes the verification conditions, the first device delivers the AI ​​model to the UE.

[0525] Step 3) The UE can measure the reference signals (such as PRS, CSI-RS) to obtain at least one physical layer estimation result (such as CIR, TDoA, RSTD, PRS RSRP). PRS is the abbreviation of Positioning Reference Signal (PRS); CSI-RS is the abbreviation of Channel State Information Reference Signal (CSI-RS); CIR is the abbreviation of Channel Impulse Response (CIR); TDoA is the abbreviation of Time Difference Of Arrival (TDoA); RSTD is the abbreviation of Reference Signal Time Difference (RSTD); RSRP is the abbreviation of Reference Signal Received Power (RSRP).

[0526] Step 4): The UE sends the estimation result obtained based on step 3) to the first device, and then the location management function (LMF) can estimate the positioning result.

[0527] Step 5) At the same time, the first device can trigger the monitoring program of the AI ​​model periodically or based on an event.

[0528] Step 5-1) The first device may indicate assumed reference signal (e.g., PRS, CSI-RS) parameters to a next generation NodeB (gNB).

[0529] Step 5-2) The gNB under test may forward the assumed reference signal to the UE.

[0530] Step 6) The UE may process the assumed reference signal based on the current AI model and obtain an estimation result.

[0531] Step 7) The UE sends the estimation result obtained by processing based on the current AI model and the assumed reference signal to the first device.

[0532] Step 8) The first device may check whether the performance of the current AI model under the assumed reference signal exceeds a threshold.

[0533] Step 9) If the performance of the current AI model under the assumed reference signal exceeds the threshold, the first device may trigger an update process of the AI ​​model and deliver the updated AI model to the UE.

[0534] Embodiment 2: UE autonomously monitors the AI ​​model.

[0535] A method for UE to autonomously monitor an AI model is shown in FIG5B , which is an interactive schematic diagram of UE autonomously monitoring an AI model according to an embodiment of the present disclosure.

[0536] Step 1): The UE sends the measurement results to the first device for AI model training.

[0537] Step 2): When the AI ​​model passes the verification conditions, the first device can deliver the AI ​​model to the UE.

[0538] Step 3) The UE may measure reference signals (such as PRS, CSI-RS) to obtain at least one physical layer estimation result (such as CIR, TDoA, RSTD, PRS RSRP).

[0539] Step 4), the UE sends the estimation result obtained based on step 3) to the first device, and then the LMF can estimate the positioning result.

[0540] Step 5) The first device may trigger a monitoring program of the AI ​​model periodically or based on an event.

[0541] Step 5-1) The first device may indicate the assumed reference signal (e.g., PRS, CSI-RS) parameters to the gNB.

[0542] Step 5-2) The gNB under test may forward the assumed reference signal to the UE.

[0543] Step 6) The UE may process the assumed reference signal based on the current AI model and obtain an estimation result.

[0544] Step 7) The UE can autonomously detect whether the performance of the current AI model under the assumed reference signal exceeds the threshold.

[0545] Step 8) If the performance of the current AI model under the assumed reference signal exceeds a threshold, the UE may request the first device to update the AI ​​model.

[0546] Step 9) After receiving the UE request, the first device triggers the AI ​​model update program and delivers the updated AI model to the UE.

[0547] Therefore, the following normalized effects can be produced:

[0548] 1. AI positioning model monitoring.

[0549] The first device and the UE may monitor the quality of the AI ​​model based on a reference signal configured as a hypothetical reference resource. The configured reference resource may be all PRSs, SSBs, or all CSI-RSs, or a mixture of SSBs and CSI-RSs.

[0550] On each reference resource, the first device can estimate the current AI model quality by comparing it with the threshold given in the following table to monitor the AI ​​model quality.

[0551] Table 1: AI model quality comparison thresholds1

[0552] The threshold Qmodel_th1 is defined as the level at which the output of the AI ​​model cannot reliably estimate the channel response. As shown in Table 1, it can be defined by the standard deviation of the ground-truncated Gaussian distribution of the actual channel response error. Where m is the symbol of the threshold in meters.

[0553] In some embodiments, different configuration numbers may correspond to different or identical comparison thresholds. For example, when three thresholds are configured, and the corresponding configuration numbers are 0, 1, and 2, respectively, when configuration number 0 is configured, the corresponding comparison threshold may be [x]m, when configuration number 1 is configured, the corresponding threshold may be [y]m, when configuration number 2 is configured, the corresponding threshold may be [z]m, and so on. This disclosure does not limit this.

[0554] In some embodiments, any item comparison threshold can be configured based on the actual performance of the AI ​​model, and this disclosure does not limit this.

[0555] Table 2: AI model quality comparison thresholds2

[0556] The threshold Est_error_th2 is defined as the level at which the output of the AI ​​model cannot reliably estimate the RSRP measurement value. As shown in Table 2, it can be defined by the standard deviation of the estimated RSRP measurement error. db is the abbreviation for the threshold unit decibel (Decibel).

[0557] In some embodiments, the specific implementation of Table 2 can refer to the relevant description of Table 1 and will not be repeated here.

[0558] Table 3: AI model quality comparison thresholds3

[0559] The threshold Est_error_th3 is defined as the level at which the output of the AI ​​model cannot reliably estimate the physical ToA / TDoA / RSTD measurement value, which can be defined by the standard deviation of the estimated ToA / TDoA / RSTD measurement error. c It is the time unit of the threshold, which can be the minimum sampling code chip time, and this disclosure does not limit this.

[0560] In some embodiments, the specific implementation of Table 3 can refer to the relevant description of Table 1 and will not be repeated here.

[0561] 2. AI model monitoring for beam prediction.

[0562] The first device and the UE can monitor the quality of the AI ​​model based on the reference signal configured as the hypothetical reference resource. The configured reference resource can be all SSBs, all CSI-RSs, or a mixture of SSBs and CSI-RSs.

[0563] On each reference resource, the first device can estimate the current AI model quality by comparing it with the threshold given in the following table to monitor the AI ​​model quality.

[0564] Table 4: AI model quality comparison thresholds4

[0565] The threshold Est_error_th4 is defined as the level at which the output of the AI ​​model cannot reliably estimate the beam index, as shown in Table 4, which can be defined by the standard deviation of the actual beam index prediction error.

[0566] In some embodiments, the specific implementation of Table 4 can refer to the relevant description of Table 1 and will not be repeated here.

[0567] Table 5: AI model quality comparison thresholds5

[0568] Among them, the threshold Est_error_th5 is defined as the level at which the output of the AI ​​model cannot reliably estimate the channel response, as shown in Table 5, which can be defined by the standard deviation of the ground actual channel response error of the truncated Gaussian distribution.

[0569] In some embodiments, the specific implementation of Table 5 can refer to the relevant description of Table 1 and will not be repeated here.

[0570] Table 6: AI model quality comparison thresholds6

[0571] Among them, the threshold Est_error_th6 is defined as the level at which the output of the AI ​​model cannot reliably estimate the physical L1-RSRP measurement value defined in Table 6, which can be defined by the standard deviation of the estimated RSRP measurement error.

[0572] In some embodiments, the specific implementation of Table 6 can refer to the relevant description of Table 1 and will not be repeated here.

[0573] 3. AI model monitoring for CSI prediction.

[0574] The first device and the UE can monitor the quality of the AI ​​model based on the reference signal configured as the hypothetical reference resource. The configured reference resource can be all SSBs, all CSI-RSs, or a mixture of SSBs and CSI-RSs.

[0575] On each reference resource, the first device can estimate the current AI model quality by comparing it with the threshold given in the following table to monitor the AI ​​model quality.

[0576] Table 7: AI model quality comparison thresholds7

[0577] Among them, the threshold Est_error_th7 is defined as the level at which the output of the AI ​​model cannot reliably estimate the CSI prediction index, as shown in Table 7, which can be defined by the standard deviation of the actual beam index prediction error.

[0578] In some embodiments, the specific implementation of Table 7 can refer to the relevant description of Table 1 and will not be repeated here.

[0579] Table 8: AI model quality comparison thresholds8

[0580] Among them, the threshold Est_error_th8 is defined as the level at which the output of the AI ​​model cannot reliably estimate the squared generalized cosine similarity, as shown in Table 8, which can be defined by the standard deviation of the ground actual channel response error of the truncated Gaussian distribution.

[0581] In some embodiments, the specific implementation of Table 8 can refer to the relevant description of Table 1 and will not be repeated here.

[0582] Table 9: AI model quality comparison thresholds9

[0583] The threshold Est_error_th9 is defined as the level of excessive AI model complexity, as shown in Table 9. It can be defined as the standard deviation of the estimated complexity above the assumed model. FLOPs is the abbreviation for the number of floating point operations (FLOPs) per unit of complexity.

[0584] In some embodiments, the specific implementation of Table 9 can refer to the relevant description of Table 1 and will not be repeated here.

[0585] The embodiments of the present disclosure further provide an apparatus for implementing any of the above methods. For example, an apparatus is provided, comprising units or modules for implementing each step performed by a terminal in any of the above methods. For another example, another apparatus is provided, comprising units or modules for implementing each step performed by a first device (e.g., a RAN) in any of the above methods.

[0586] It should be understood that the division of the various units or modules in the above device is merely a division of logical functions. In actual implementation, they may be fully or partially integrated into a physical entity, or they may be physically separated. In addition, the units or modules in the device may be implemented in the form of a processor calling software: for example, the device includes a processor, the processor is connected to a memory, and the memory stores instructions. The processor calls the instructions stored in the memory to implement any of the above methods or implement the functions of the various units or modules of the above device, wherein the processor is, for example, a general-purpose processor, such as a central processing unit (CPU) or a microprocessor, and the memory is a memory within the device or a memory outside the device. Alternatively, the units or modules in the device can be implemented in the form of hardware circuits, and the functions of some or all of the units or modules can be realized by designing the hardware circuits. The above-mentioned hardware circuits can be understood as one or more processors; for example, in one implementation, the above-mentioned hardware circuit is an application-specific integrated circuit (ASIC), which realizes the functions of some or all of the above units or modules by designing the logical relationship of the components in the circuit; for example, in another implementation, the above-mentioned hardware circuit can be realized by a programmable logic device (PLD). Taking a field programmable gate array (FPGA) as an example, it can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by configuring the configuration file, thereby realizing the functions of some or all of the above units or modules. All units or modules of the above devices can be realized in the form of software called by the processor, or in the form of hardware circuits, or in part by the form of software called by the processor, and the rest by hardware circuits.

[0587] In the embodiments of the present disclosure, the processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), or a digital signal processor (DSP). In another implementation, the processor can implement certain functions through the logical relationship of the hardware circuit. The logical relationship of the above-mentioned hardware circuit is fixed or reconfigurable. For example, the processor is a hardware circuit implemented by an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and implementing the hardware circuit configuration can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units or modules. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), etc.

[0588] FIG6A is a schematic diagram of the structure of a terminal proposed in an embodiment of the present disclosure. As shown in FIG6A , the terminal 6100 may include at least one of a transceiver module 6101 and a processing module 6102. The terminal 6100 may include:

[0589] The transceiver module 6101 is configured to receive a first reference signal;

[0590] A processing module 6102 is configured to process the first reference signal using the current AI model to obtain an estimation result;

[0591] The transceiver module 6101 is further used to send the estimation result to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

[0592] Optionally, the first reference signal includes at least one of the following:

[0593] Positioning reference signal PRS;

[0594] Synchronous signal block;

[0595] Channel State Information Reference Signal CSI-RS.

[0596] Optionally, the first device is selected from one of the following:

[0597] Access network equipment;

[0598] Core network equipment;

[0599] Network server.

[0600] Optionally, the processing module 6102 is further configured to:

[0601] When an updated AI model is received, the current AI model is replaced with the updated AI model.

[0602] FIG6B is a schematic diagram of the structure of the first device proposed in an embodiment of the present disclosure. As shown in FIG6B , the first device 6200 may include: at least one of a transceiver module 6201 and a processing module 6202. The first device 6200 may include:

[0603] The transceiver module 6201 is configured to receive an estimation result sent by a terminal, where the estimation result is obtained by the terminal by processing the first reference signal using the current AI model;

[0604] The processing module 6202 is used to determine whether to update the AI ​​model based on the estimation results.

[0605] Optionally, the first device is an access network device, and the transceiver module 6201 is further configured to:

[0606] A first reference signal is sent to the terminal.

[0607] Optionally, the first device is a core network device or a server, and the transceiver module 6201 is further configured to:

[0608] First information is sent to the access network device, where the first information is used to trigger the access network device to send a first reference signal to the terminal.

[0609] Optionally, the processing module 6202 is further configured to:

[0610] determining an error in the estimation result based on a reference result associated with the first reference signal;

[0611] When the error of the estimated result is greater than the threshold, the AI ​​model is updated.

[0612] Optionally, the first reference signal is used for positioning, and the threshold is any one of the following:

[0613] a first value associated with a ground actual channel response error;

[0614] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0615] a third value associated with a time-of-arrival (TOA) error;

[0616] a fourth value associated with a time difference of arrival, TDOA, error;

[0617] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0618] Optionally, the first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0619] The first reference signal is used for channel state information CSI measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0620] Optionally, the transceiver module 6201 is further configured to:

[0621] Send the updated AI model to the terminal.

[0622] Optionally, the first reference signal includes at least one of the following:

[0623] Positioning reference signal PRS;

[0624] Synchronous signal block;

[0625] Channel State Information Reference Signal CSI-RS.

[0626] FIG6C is a schematic diagram of the structure of a terminal proposed in an embodiment of the present disclosure. As shown in FIG6C , the terminal 6300 may include at least one of a transceiver module 6301 and a processing module 6302. The terminal 6300 may include:

[0627] The transceiver module 6301 is configured to receive a first reference signal;

[0628] A processing module 6302 is configured to process the first reference signal using the current AI model to obtain an estimation result;

[0629] The processing module 6302 is further used to determine whether to update the AI ​​model based on the estimation results.

[0630] Optionally, the processing module 6302 is further configured to:

[0631] determining an error in the estimation result based on a reference result associated with the first reference signal;

[0632] When the error of the estimation result is greater than a threshold, it is determined to update the AI ​​model.

[0633] Optionally, the first reference signal is used for positioning, and the threshold is any one of the following:

[0634] a first value associated with a ground actual channel response error;

[0635] a second value associated with a physical reference signal received power (RSRP) measurement error;

[0636] a third value associated with a time-of-arrival (TOA) error;

[0637] a fourth value associated with a time difference of arrival, TDOA, error;

[0638] A fifth value associated with a Reference Signal Time Difference RSTD error.

[0639] Optionally, the first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or,

[0640] The first reference signal is used for channel state information CSI measurement, and the threshold is a value used to indicate a channel state information estimation result.

[0641] Optionally, it also includes:

[0642] The transceiver module 6301 is further configured to receive a threshold value;

[0643] The processing module 6302 is further configured to determine a threshold value according to the protocol;

[0644] The processing module 6302 is further configured to determine a threshold value based on configuration information.

[0645] Optionally, the transceiver module 6301 is further configured to:

[0646] Determine to update the AI ​​model and send second information, wherein the second information is used to trigger the first device to update the AI ​​model, wherein the first device is any one of the following: an access network device, a core network device, and a network server.

[0647] Optionally, the transceiver module 6301 is further configured to:

[0648] Determine not to update the AI ​​model, and send the estimation result to the first device.

[0649] Optionally, the first reference signal includes at least one of the following:

[0650] Positioning reference signal PRS;

[0651] Synchronous signal block;

[0652] Channel State Information Reference Signal CSI-RS.

[0653] FIG6D is a schematic diagram of the structure of the first device proposed in an embodiment of the present disclosure. As shown in FIG6D , the first device 6400 may include: at least one of a transceiver module 6401 and a processing module 6402. The first device 6400 may include:

[0654] The transceiver module 6401 is used to send a first reference signal, where the first reference signal is used to assist the terminal in determining whether to update the current AI model.

[0655] Optionally, it also includes:

[0656] The transceiver module 6401 is further configured to receive second information, wherein the second information is configured to trigger an update of the AI ​​model;

[0657] A processing module 6402 is configured to update the AI ​​model and obtain an updated AI model;

[0658] The transceiver module 6401 is also used to send the updated AI model to the terminal.

[0659] Optionally, the transceiver module 6401 is further configured to:

[0660] Receive the estimation results without initiating any update of the AI ​​model.

[0661] In some embodiments, the transceiver module may include a transmitting module and / or a receiving module, and the transmitting module and the receiving module may be separate or integrated. Optionally, the transceiver module may be interchangeable with the transceiver.

[0662] In some embodiments, the processing module can be a single module or can include multiple submodules. Optionally, the multiple submodules each execute all or part of the steps required to be executed by the processing module. Optionally, the processing module can be interchangeable with the processor.

[0663] Figure 7A is a schematic diagram of the structure of a communication device 7100 proposed in an embodiment of the present disclosure. Communication device 7100 can be a terminal, a first device, a chip, a chip system, or a processor that supports a terminal in implementing any of the above methods, or a chip, a chip system, or a processor that supports a first device in implementing any of the above methods. Communication device 7100 can be used to implement the methods described in the above method embodiments. For details, please refer to the description of the above method embodiments.

[0664] As shown in FIG7A , the communication device 7100 includes one or more processors 7101. The processor 7101 may be a general-purpose processor or a dedicated processor, for example, a baseband processor or a central processing unit. The baseband processor may be used to process communication protocols and communication data, and the central processing unit may be used to control a communication device (e.g., a base station, a baseband chip, a terminal device, a terminal device chip, a DU or a CU, etc.), execute programs, and process program data. The communication device 7100 is used to perform any of the above methods.

[0665] In some embodiments, the communication device 7100 further includes one or more memories 7102 for storing instructions. Optionally, all or part of the memories 7102 may be located outside the communication device 7100.

[0666] In some embodiments, the communication device 7100 further includes one or more transceivers 7103. When the communication device 7100 includes one or more transceivers 7103, the transceiver 7103 performs at least one of the communication steps such as sending and / or receiving in the above method (for example, step S2101, step S2103, step S2106, step S2201, step S2205, step S2207, step S2301, step S2305, but not limited thereto), and the processor 7101 performs the other steps (for example, step S2102, step S2104, step S2105, step S2202, step S2203, step S2204, step S2206, step S2302, step S2303, step S2304, step S2306).

[0667] In some embodiments, a transceiver may include a receiver and / or a transmitter. The receiver and transmitter may be separate or integrated. Optionally, the terms transceiver, transceiver unit, transceiver, and transceiver circuit may be used interchangeably; the terms transmitter, transmitting unit, transmitter, and transmitting circuit may be used interchangeably; and the terms receiver, receiving unit, receiver, and receiving circuit may be used interchangeably.

[0668] In some embodiments, the communication device 7100 may include one or more interface circuits 7104. Optionally, the interface circuit 7104 is connected to the memory 7102. The interface circuit 7104 may be configured to receive signals from the memory 7102 or other devices, and may be configured to send signals to the memory 7102 or other devices. For example, the interface circuit 7104 may read instructions stored in the memory 7102 and send the instructions to the processor 7101.

[0669] The communication device 7100 described in the above embodiment may be a terminal or a first device, but the scope of the communication device 7100 described in the present disclosure is not limited thereto, and the structure of the communication device 7100 may not be limited by FIG. 7A. The communication device may be an independent device or may be part of a larger device. For example, the communication device may be: 1) an independent integrated circuit IC, or a chip, or a chip system or subsystem; (2) a collection of one or more ICs, optionally, the above IC collection may also include a storage component for storing data or programs; (3) an ASIC, such as a modem; (4) a module that can be embedded in other devices; (5) a receiver, a terminal device, an intelligent terminal device, a cellular phone, a wireless device, a handheld device, a mobile unit, an in-vehicle device, a network device, a cloud device, an artificial intelligence device, etc.; (6) others, etc.

[0670] 7B is a schematic diagram of the structure of a chip 7200 proposed in an embodiment of the present disclosure. If the communication device 7100 can be a chip or a chip system, please refer to the schematic diagram of the structure of the chip 7200 shown in FIG7B , but the present disclosure is not limited thereto.

[0671] The chip 7200 includes one or more processors 7201 , and the chip 7200 is configured to execute any of the above methods.

[0672] In some embodiments, the chip 7200 further includes one or more interface circuits 7202. Optionally, the interface circuit 7202 is connected to the memory 7203. The interface circuit 7202 can be used to receive signals from the memory 7203 or other devices, and can be used to send signals to the memory 7203 or other devices. For example, the interface circuit 7202 can read instructions stored in the memory 7203 and send the instructions to the processor 7201.

[0673] In some embodiments, the interface circuit 7202 executes at least one of the communication steps such as sending and / or receiving in the above method (for example, step S2101, step S2103, step S2106, step S2201, step S2205, step S2207, step S2301, step S2305, but not limited to these), and the processor 7201 executes other steps (for example, step S2102, step S2104, step S2105, step S2202, step S2203, step S2204, step S2206, step S2302, step S2303, step S2304, step S2306).

[0674] In some embodiments, terms such as interface circuit, interface, transceiver pin, and transceiver may be used interchangeably.

[0675] In some embodiments, the chip 7200 further includes one or more memories 7203 for storing instructions. Alternatively, all or part of the memories 7203 may be located outside the chip 7200.

[0676] The present disclosure also proposes a storage medium having instructions stored thereon. When the instructions are executed on the communication device 7100, the communication device 7100 executes any of the above methods. Optionally, the storage medium is an electronic storage medium. Optionally, the storage medium is a computer-readable storage medium, but is not limited thereto and may also be a storage medium readable by other devices. Optionally, the storage medium may be a non-transitory storage medium, but is not limited thereto and may also be a temporary storage medium.

[0677] The present disclosure also provides a program product, which, when executed by the communication device 7100, enables the communication device 7100 to perform any of the above methods. Optionally, the program product is a computer program product.

[0678] The present disclosure also proposes a computer program, which, when executed on a computer, causes the computer to perform any one of the above methods.

Claims

1. A monitoring method for an artificial intelligence (AI) model, the method being executed by a terminal, characterized in that: The method comprises: receiving a first reference signal; Processing the first reference signal using the current AI model to obtain an estimation result; The estimation result is sent to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

2. The method according to claim 1, wherein The first reference signal includes at least one of the following: Positioning reference signal PRS; Synchronous signal block; Channel State Information Reference Signal CSI-RS.

3. The method according to claim 1, wherein The first device is selected from one of the following: Access network equipment; Core network equipment; Network server.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: When an updated AI model is received, the current AI model is replaced with the updated AI model.

5. A monitoring method for an artificial intelligence (AI) model, the method being executed by a first device, characterized in that: The method comprises: Receiving an estimation result sent by a terminal, wherein the estimation result is obtained by the terminal by processing the first reference signal using a current AI model; Based on the estimation result, determine whether to update the AI ​​model.

6. The method according to claim 5, wherein The first device is an access network device, and the method further includes: Sending the first reference signal to the terminal.

7. The method according to claim 5, wherein The first device is a core network device or a server, and the method further includes: Sending first information to an access network device, where the first information is used to trigger the access network device to send a first reference signal to the terminal.

8. The method according to any one of claims 5 to 7, wherein: Determining whether to update the AI ​​model based on the estimation result includes: determining an error of the estimation result according to a reference result associated with the first reference signal; When the error of the estimation result is greater than a threshold, the AI ​​model is updated.

9. The method according to claim 8, wherein The first reference signal is used for positioning, and the threshold is any one of the following: a first value associated with a ground actual channel response error; a second value associated with a physical reference signal received power (RSRP) measurement error; a third value associated with a time-of-arrival (TOA) error; a fourth value associated with a time difference of arrival, TDOA, error; A fifth value associated with a Reference Signal Time Difference RSTD error.

10. The method according to claim 8, wherein The first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or, The first reference signal is used for channel state information (CSI) measurement, and the threshold is a value used to indicate a channel state information estimation result.

11. The method according to any one of claims 5 to 10, wherein: The method further comprises: Sending the updated AI model to the terminal.

12. The method according to any one of claims 5 to 11, wherein: The first reference signal includes at least one of the following: Positioning reference signal PRS; Synchronous signal block; Channel State Information Reference Signal CSI-RS.

13. A monitoring method for an artificial intelligence (AI) model, the method being executed by a terminal, characterized in that: The method comprises: receiving a first reference signal; Processing the first reference signal using the current AI model to obtain an estimation result; Based on the estimation result, determine whether to update the AI ​​model.

14. The method according to claim 13, wherein Determining whether to update the AI ​​model based on the estimation result includes: determining an error of the estimation result according to a reference result associated with the first reference signal; When the error of the estimation result is greater than a threshold, it is determined to update the AI ​​model.

15. The method according to claim 14, wherein The first reference signal is used for positioning, and the threshold is any one of the following: a first value associated with a ground actual channel response error; a second value associated with a physical reference signal received power (RSRP) measurement error; a third value associated with a time-of-arrival (TOA) error; a fourth value associated with a time difference of arrival, TDOA, error; A fifth value associated with a Reference Signal Time Difference RSTD error.

16. The method according to claim 14, wherein The first reference signal is used for beam prediction, and the threshold is a value used to indicate a beam prediction result; or, The first reference signal is used for channel state information (CSI) measurement, and the threshold is a value used to indicate a channel state information estimation result.

17. The method according to any one of claims 14 to 16, wherein: Also include at least one of the following: receiving the threshold value; Determine the threshold value according to the agreement; The threshold is determined according to the configuration information.

18. The method according to any one of claims 13 to 17, wherein: The method further comprises: Determine to update the AI ​​model and send second information, wherein the second information is used to trigger the first device to update the AI ​​model, wherein the first device is any one of the following: an access network device, a core network device, and a network server.

19. The method according to any one of claims 13 to 18, wherein: The method further comprises: Determine not to update the AI ​​model, and send the estimation result to the first device.

20. The method according to any one of claims 13 to 19, wherein: The first reference signal includes at least one of the following: Positioning reference signal PRS; Synchronous signal block; Channel State Information Reference Signal CSI-RS.

21. A monitoring method for an artificial intelligence (AI) model, the method being executed by a first device, characterized in that: The method comprises: Send a first reference signal, where the first reference signal is used to assist the terminal in determining whether to update the current AI model.

22. The method according to claim 21, wherein The method further comprises: receiving second information, wherein the second information is used to trigger an update of the AI ​​model; Updating the AI ​​model to obtain an updated AI model; Sending the updated AI model to the terminal.

23. The method according to claim 21 or 22, wherein: The method further comprises: An estimation result is received without initiating an operation to update the AI ​​model.

24. A terminal, characterized in that: The terminal includes: a transceiver module, configured to receive a first reference signal; a processing module, configured to process the first reference signal using a current AI model to obtain an estimation result; The transceiver module is further used to send the estimation result to the first device, wherein the estimation result is used to assist the first device in determining whether to update the AI ​​model.

25. A first device, characterized in that: The first device includes: a transceiver module, configured to receive an estimation result sent by a terminal, wherein the estimation result is obtained by the terminal by processing the first reference signal using a current AI model; A processing module is used to determine whether to update the AI ​​model based on the estimation result.

26. A terminal, characterized in that: The terminal includes: a transceiver module, configured to receive a first reference signal; a processing module, configured to process the first reference signal using a current AI model to obtain an estimation result; The processing module is further used to determine whether to update the AI ​​model based on the estimation result.

27. A first device, characterized in that: The first device includes: The transceiver module is used to send a first reference signal, wherein the first reference signal is used to assist the terminal in determining whether to update the current AI model.

28. A terminal, characterized in that: include: one or more processors; Wherein, the terminal is used to execute the monitoring method for an artificial intelligence AI model described in any one of claims 1-4.

29. A first device, characterized in that: include: one or more processors; Wherein, the first device is used to execute the monitoring method for an artificial intelligence AI model described in any one of claims 5-12.

30. A terminal, characterized in that: include: one or more processors; Wherein, the terminal is used to execute the monitoring method for an artificial intelligence AI model described in any one of claims 13-20.

31. A first device, characterized in that: include: one or more processors; Wherein, the first device is used to execute the monitoring method for an artificial intelligence AI model described in any one of claims 21-23.

32. A communication system, characterized in that: It includes a terminal and a first device, wherein the terminal is configured to implement the monitoring method for an artificial intelligence AI model described in any one of claims 1-4 and 13-20, and the first device is configured to implement the monitoring method for an artificial intelligence AI model described in any one of claims 5-12 and 21-23.

33. A storage medium storing instructions, characterized in that: When the instruction is executed on a communication device, the communication device executes the monitoring method for an artificial intelligence (AI) model according to any one of claims 1 to 23.

34. A program product comprising a computer program, characterized in that When the computer program runs on a communication device, the communication device executes the monitoring method for an artificial intelligence (AI) model according to any one of claims 1 to 23.

Images