Model identification method, terminal, network equipment, system and storage medium

Abstract

The invention provides a model identification method, a terminal, a network device, a system and a storage medium, and the method comprises the steps: receiving a first signaling transmitted by the network device, the first signaling being used for indicating the terminal to execute AI model identification associated with a first function; determining a model identification result in a plurality of AI models based on the first signaling; and sending a second signaling to the network device based on the model identification result. According to the invention, the reliability of the used AI model is improved, the reliability of AI and communication fusion is improved, the joint development of communication and AI is promoted, and the availability is high.

Description

Model identification method, terminal, network device, system and storage medium TECHNICAL FIELD

[0001] The present disclosure relates to the field of communication, and in particular, to a model identification method, a terminal, a network device, a system and a storage medium. BACKGROUND

[0002] In order to enable artificial intelligence (AI) and machine learning (ML) on the air interface, an AI model (or ML model) can be created, various functions can be implemented through the AI model, and the AI model can be managed in the AI model life cycle.

[0003] SUMMARY

[0004] In order to improve the reliability of AI and communication integration, the embodiments of the present disclosure provide a model identification method, a terminal, a network device, a system and a storage medium.

[0005] According to a first aspect of the embodiments of the present disclosure, a model identification method is provided, the method is executed by a terminal, and the method comprises:

[0006] receiving first signaling sent by a network device; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function;

[0007] based on the first signaling, determining a model identification result in a plurality of AI models; wherein the model identification result includes a first AI model, and the first AI model is an AI model associated with the first function and having the best performance in the plurality of AI models;

[0008] based on the model identification result, sending second signaling to the network device.

[0009] According to a second aspect of the embodiments of the present disclosure, a model identification method is provided, the method is executed by a network device, and the method comprises:

[0010] sending first signaling to a terminal; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function;

[0011] receiving second signaling sent by the terminal based on a model identification result; wherein the model identification result includes a first AI model, and the first AI model is an AI model associated with the first function and having the best performance in a plurality of AI models deployed on the terminal.

[0012] According to a third aspect of the embodiments of the present disclosure, a terminal is provided, comprising:

[0013] a transceiver configured to receive first signaling sent by a network device; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function;

[0014] a processing module configured to determine, based on the first signaling, a model identification result from a plurality of AI models; wherein the model identification result includes a first AI model, and the first AI model is an AI model associated with the first function and having the best performance among the plurality of AI models;

[0015] the transceiver is further configured to send, based on the model identification result, second signaling to the network device.

[0016] According to a fourth aspect of embodiments of the present disclosure, a network device is provided, comprising:

[0017] a transceiver configured to send first signaling to a terminal; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function;

[0018] the transceiver is further configured to receive second signaling sent by the terminal based on a model identification result; wherein the model identification result includes a first AI model, and the first AI model is an AI model associated with the first function and having the best performance among a plurality of AI models deployed on the terminal.

[0019] According to a fifth aspect of embodiments of the present disclosure, a terminal is provided, comprising:

[0020] one or more processors;

[0021] wherein the processor is configured to perform the model identification method of any one of the first aspect.

[0022] According to a sixth aspect of embodiments of the present disclosure, a network device is provided, comprising:

[0023] one or more processors;

[0024] wherein the processor is configured to perform the model identification method of any one of the second aspect.

[0025] According to a seventh aspect of embodiments of the present disclosure, a model identification system is provided, comprising:

[0026] a terminal configured to implement the model identification method of any one of the first aspect;

[0027] a network device configured to implement the model identification method of any one of the second aspect.

[0028] According to an eighth aspect of the embodiments of the present disclosure, a storage medium is provided, which stores instructions, when the instructions are executed on a model identification device, causing the model identification device to perform the model identification method according to any one of the first aspect or the second aspect.

[0029] According to a ninth aspect of the embodiments of the present disclosure, a computer program product is provided, which includes a computer program, when the computer program is executed by a processor, is used to implement the model identification method according to any one of the first aspect or the second aspect.

[0030] In the embodiments of the present disclosure, the terminal can determine the model identification result in the plurality of AI models based on the first signaling sent by the network device, and send the second signaling to the network device based on the model identification result. By sending the model identification result to the network device, the network device can be ensured to be clear about the AI model associated with the specified function and having the optimal performance, the reliability of the used AI model is improved, and the reliability of AI and communication integration is improved, which is conducive to the common development of communication and AI, and has high usability.

[0031] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0032] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0033] FIG. 1 is one exemplary schematic diagram of an architecture of a communication system according to an embodiment of the present disclosure.

[0034] FIG. 2A is one exemplary interactive schematic diagram of a model identification method according to an embodiment of the present disclosure.

[0035] FIG. 2B is one exemplary corresponding relationship schematic diagram of functions and models according to an embodiment of the present disclosure.

[0036] FIG. 2C is one exemplary scenario schematic diagram of a total latency of a model identification process according to an embodiment of the present disclosure.

[0037] FIG. 2D-1 is one exemplary scenario schematic diagram of parallel output of prediction results of a plurality of AI models according to an embodiment of the present disclosure.

[0038] FIG. 2D-2 is one exemplary scenario schematic diagram of sequential output of prediction results of a plurality of AI models according to an embodiment of the present disclosure.

[0039] FIG. 2D-3 is a schematic diagram of one example scenario of parallel output of prediction results of part of AI models in a plurality of AI models according to an embodiment of the present disclosure.

[0040] FIG. 3A is a schematic diagram of one example flow of a model identification method according to an embodiment of the present disclosure.

[0041] FIG. 3B is a schematic diagram of one example flow of a model identification method according to an embodiment of the present disclosure.

[0042] FIG. 3C is a schematic diagram of one example flow of a model identification method according to an embodiment of the present disclosure.

[0043] FIG. 3D is a schematic diagram of one example flow of a model identification method according to an embodiment of the present disclosure.

[0044] FIG. 4 is a schematic diagram of one example flow of a model identification method according to an embodiment of the present disclosure.

[0045] FIG. 5A is a schematic block diagram of a terminal according to an embodiment of the present disclosure.

[0046] FIG. 5B is a schematic block diagram of a network device according to an embodiment of the present disclosure.

[0047] FIG. 6A is a schematic diagram of interaction of a communication device according to an embodiment of the present disclosure.

[0048] FIG. 6B is a schematic diagram of interaction of a chip according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0049] The example embodiments will be described in detail herein with reference to the attached drawings. The following description is made with reference to the accompanying drawings in which like reference numerals refer to like elements in the several figures. The following description of example embodiments does not represent all aspects of the present application. Instead, they are merely examples of apparatus and methods in accordance with some aspects of the present application as detailed in the appended claims.

[0050] The embodiments of the present disclosure provide a model identification method, a terminal, a network device, a system and a storage medium.

[0051] In a first aspect, a model identification method is provided. The method is performed by a terminal and includes: receiving first signaling sent by a network device; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function; determining a model identification result in a plurality of AI models based on the first signaling; wherein the model identification result includes a first AI model, and the first AI model is an AI model associated with the first function and having optimal performance in the plurality of AI models; and sending second signaling to the network device based on the model identification result.

[0052] In the above embodiments, the terminal can determine a model identification result in a plurality of AI models based on first signaling sent by a network device, and send second signaling to the network device based on the model identification result. By sending the model identification result to the network device, the network device can be ensured to know the AI model associated with a specified function and having optimal performance, the reliability of the used AI model is improved, the reliability of AI and communication integration is improved, the common development of communication and AI is promoted, and the usability is high.

[0053] In some embodiments of the first aspect, the first signaling includes at least one of the following: function indication information, the function indication information being used to indicate the first function; and a first condition set, the first condition set including one or more network configuration conditions associated with the first function.

[0054] In the above embodiments, the first signaling can include but is not limited to at least one of the above, thereby triggering a model identification process, and the first condition set associated with the first function can be configured, which is simple and has high usability.

[0055] In some embodiments of the first aspect, the method further includes: receiving measurement configuration information sent by the network device; and performing a first measurement associated with the first function based on the measurement configuration information to obtain a first measurement result.

[0056] In the above embodiments, the terminal can perform a first measurement to obtain a first measurement result, and subsequently determine an AI model having optimal performance in a plurality of AI models based on the first measurement result, thereby improving the reliability of the used AI model.

[0057] In some embodiments of the first aspect, in some embodiments, the determining the model identification result from the plurality of AI models comprises: inputting the first measurement result into the plurality of AI models to obtain a prediction result associated with the first function output by each AI model; determining a first prediction result with a smallest difference from the true result from the plurality of prediction results; and determining, from the plurality of AI models, an AI model outputting the first prediction result as the first AI model to obtain the model identification result.

[0058] In the above embodiments, the model identification result can be determined in the above manner, which is simple and convenient to use.

[0059] In some embodiments of the first aspect, in some embodiments, the method further comprises any one of the following: determining the true result by measurement; and determining the true result based on third signaling sent by the network device.

[0060] In the above embodiments, the true result can be determined by the terminal or sent to the terminal by the network device, and the disclosure does not limit this, which improves the flexibility of determining the true result and the reliability of model identification.

[0061] In some embodiments of the first aspect, in some embodiments, the second signaling comprises at least one of the following: a model identifier of the first AI model; and a first prediction result, which is a prediction result associated with the first function output by the first AI model.

[0062] In the above embodiments, the terminal can send the model identifier of the identified first AI model and / or the first prediction result output by the first AI model to the network device through the second signaling, which improves the reliability of AI and communication integration, is conducive to the joint development of communication and AI, and is highly usable.

[0063] In some embodiments of the first aspect, in some embodiments, the first signaling comprises at least one of the following: radio resource control (RRC) signaling; downlink control information (DCI); and media access control (MAC) CE.

[0064] In the above embodiments, the model identification process can be triggered based on the above signaling, which is simple and convenient to use.

[0065] In some embodiments of the first aspect, in some embodiments, the total latency of the AI model identification is determined based on at least one of: a first latency, the first latency being a latency from receiving the first signaling to successfully decoding the first signaling; a second latency, the second latency being a latency from successfully decoding the first signaling to performing the first measurement associated with the first function; a third latency, the third latency being a latency from performing the first measurement associated with the first function to determining the prediction result output by each AI model; a fourth latency, the fourth latency being a latency from determining the prediction result output by each AI model to obtaining the real result; a fifth latency, the fifth latency being a latency from obtaining the real result to determining the model identification result; and a sixth latency, the sixth latency being a latency from determining the model identification result to sending the second signaling.

[0066] In the above embodiments, the total latency of the AI model identification can be determined based on the at least one latency, ensuring that the network device and the terminal have consistent understanding of the total latency, and the availability is high.

[0067] In some embodiments of the first aspect, in some embodiments, the third latency is associated with an execution order of the prediction result output by the plurality of AI models; and the execution order includes any one of: the plurality of AI models outputting the prediction result in parallel; the plurality of AI models outputting the prediction result sequentially; and part of the plurality of AI models outputting the prediction result in parallel.

[0068] In the above embodiments, the third latency is associated with the execution order of the prediction result output by the plurality of AI models, and the availability is high.

[0069] In a second aspect, the embodiments of the present disclosure provide a model identification method, the method being performed by a network device, and the method includes: sending first signaling to a terminal; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function; and receiving second signaling sent by the terminal based on a model identification result; wherein the model identification result includes a first AI model, and the first AI model is an AI model that is associated with the first function and has the best performance among a plurality of AI models deployed on the terminal.

[0070] In some embodiments of the second aspect, in some embodiments, the first signaling includes at least one of: function indication information, the function indication information being used to indicate the first function; and a first condition set, the first condition set including one or more network configuration conditions associated with the first function.

[0071] In some embodiments of the second aspect, in some embodiments, the method further comprises: sending, to the terminal, measurement configuration information; wherein the measurement configuration information is used to configure the terminal to perform the first measurement associated with the first function.

[0072] In some embodiments of the second aspect, in some embodiments, the method further comprises: sending, to the terminal, third signaling, wherein the third signaling comprises the real result associated with the first function.

[0073] In some embodiments of the second aspect, in some embodiments, the second signaling comprises at least one of: a model identifier of the first AI model; a first prediction result, wherein the first prediction result is a prediction result output by the first AI model and associated with the first function.

[0074] In some embodiments of the second aspect, in some embodiments, the first signaling comprises at least one of: radio resource control (RRC) signaling; downlink control information (DCI); and media access control (MAC) control element (CE).

[0075] In some embodiments of the second aspect, in some embodiments, the total latency identified by the AI model is determined based on at least one of: a first latency from receiving the first signaling from the terminal to successfully decoding the first signaling by the terminal; a second latency from successfully decoding the first signaling by the terminal to performing the first measurement associated with the first function by the terminal; a third latency from performing the first measurement associated with the first function by the terminal to determining the prediction result output by each AI model by the terminal; a fourth latency from determining the prediction result output by each AI model by the terminal to obtaining the real result by the terminal; a fifth latency from obtaining the real result by the terminal to determining the model identification result by the terminal; and a sixth latency from determining the model identification result by the terminal to sending the second signaling by the terminal.

[0076] In some embodiments of the second aspect, in some embodiments, the third latency is associated with an execution order of outputting the prediction result by the plurality of AI models; wherein the execution order comprises any one of: outputting the prediction result by the plurality of AI models in parallel; outputting the prediction result by the plurality of AI models sequentially; and outputting the prediction result by part of the plurality of AI models in parallel.

[0077] In some embodiments of the second aspect, in some embodiments, the method further comprises: establishing a correspondence between a first condition set and a model identifier of the first AI model; wherein the first condition set comprises one or more network configuration conditions associated with the first function.

[0078] In some embodiments of the second aspect, in some embodiments, the method further comprises: based on the first correspondence, configuring the model identifier of the first AI model in a case that the terminal is configured with the first condition set.

[0079] In the third aspect, the embodiments of the present disclosure provide a terminal, comprising: a transceiver module configured to receive first signaling sent by a network device; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function; a processing module configured to determine a model identification result in a plurality of AI models based on the first signaling; wherein the model identification result comprises a first AI model, and the first AI model is an AI model associated with the first function and having the best performance in the plurality of AI models; and the transceiver module is further configured to send second signaling to the network device based on the model identification result.

[0080] In the fourth aspect, the embodiments of the present disclosure provide a network device, comprising: a transceiver module configured to send first signaling to a terminal; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function; and the transceiver module is further configured to receive second signaling sent by the terminal based on a model identification result; wherein the model identification result comprises a first AI model, and the first AI model is an AI model associated with the first function and having the best performance in a plurality of AI models deployed on the terminal.

[0081] In the fifth aspect, the embodiments of the present disclosure provide a terminal, comprising: one or more processors; wherein the processor is configured to perform the model identification method of any one of the first aspect.

[0082] In the sixth aspect, the embodiments of the present disclosure provide a network device, comprising: one or more processors; wherein the processor is configured to perform the model identification method of any one of the second aspect.

[0083] In the seventh aspect, the embodiments of the present disclosure provide a model identification system, comprising: a terminal configured to implement the model identification method of any one of the first aspect; and a network device configured to implement the model identification method of any one of the second aspect.

[0084] In a ninth aspect, the embodiments of the present disclosure provide a storage medium, which stores instructions. When the instructions are executed on a model identification device, the model identification device is caused to perform the model identification method according to any one of the first aspect or the second aspect.

[0085] In a tenth aspect, the embodiments of the present disclosure provide a computer program product, which comprises a computer program. When the computer program is executed by a processor, the model identification method according to any one of the first aspect or the second aspect is implemented.

[0086] In an eleventh aspect, the embodiments of the present disclosure provide a chip or a chip system. The chip or the chip system comprises processing circuitry configured to perform the method according to the optional implementation of the first aspect or the second aspect.

[0087] It can be understood that the terminal, the network device, the communication system, the storage medium, the computer program product, the chip or the chip system are all used to execute the method proposed in the embodiments of the present disclosure. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be described here.

[0088] The embodiments of the present disclosure propose the invention name. In some embodiments, the terms of the model identification method and the information transmission method, the model determination method can be replaced with each other, the terms of the model identification device and the information transmission device, the model determination device can be replaced with each other, and the terms of the communication system, the information transmission system, the model identification system can be replaced with each other.

[0089] The embodiments of the present disclosure are not exhaustive, but only illustrate some embodiments, and are not specific limitations on the protection scope of the present disclosure. In the case of no contradiction, each step in an embodiment can be implemented as an independent embodiment, and the steps can be combined arbitrarily, for example, the scheme after removing some steps in an embodiment can also be implemented as an independent embodiment, and the order of the steps in an embodiment can be exchanged arbitrarily, in addition, the optional implementation in an embodiment can be combined arbitrarily; in addition, the embodiments can be combined arbitrarily, for example, part or all steps of different embodiments can be combined arbitrarily, an embodiment can be combined with the optional implementation of other embodiments.

[0090] In each embodiment of the present disclosure, the terms and / or descriptions between the embodiments are consistent if there is no special description and logical conflict, and can be referred to each other, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0091] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments, and not as a limitation on the present disclosure.

[0092] In the embodiments of the present disclosure, an element expressed in singular form, such as "a", "an", "the", "said", "the aforementioned", "the foregoing", "this", and the like, unless otherwise specified, can represent "one and only one", or can represent "one or more", "at least one", and the like. For example, in the case of using an article such as "a", "an", "the" in English, the noun after the article can be understood as a singular expression, or can be understood as a plural expression.

[0093] In the embodiments of the present disclosure, "plurality" refers to two or more.

[0094] In some embodiments, the terms "at least one of", "one or more", "a plurality of", "multiple", and the like can be replaced with each other.

[0095] In some embodiments, the description modes such as "at least one of A, B", "A and / or B", "A in one case and B in another case", "in response to a case A, in response to a case B", and the like can include the following technical solutions according to the case: in some embodiments, A is executed regardless of B; in some embodiments, B is executed regardless of A; in some embodiments, A and B are selectively executed from A and B; in some embodiments, A and B are executed (A and B are both executed). When there are more branches such as A, B, C, and the like, it is similar to the above.

[0096] In some embodiments, the description modes such as "A or B" and the like can include the following technical solutions according to the case: in some embodiments, A is executed regardless of B; in some embodiments, B is executed regardless of A; in some embodiments, A and B are selectively executed from A and B; when there are more branches such as A, B, C, and the like, it is similar to the above.

[0097] The prefix words of "first", "second" and the like in the embodiments of the present disclosure are merely used to distinguish different description objects, and do not constitute limitation on the position, order, priority, quantity or content of the description objects. The description objects are described in the claims or embodiments, and should not be construed as redundant limitation because of the use of the prefix words. For example, the description object is "field", and the ordinal words before "field" in "first field" and "second field" do not limit the position or order between "fields", and "first" and "second" do not limit whether the "fields" modified thereby are in the same message or not, nor limit the order of "first field" and "second field". For another example, the description object is "level", and the ordinal words before "level" in "first level" and "second level" do not limit the priority between "levels". For another example, the quantity of the description object is not limited by the ordinal words, and can be one or more. For example, "first device", wherein the quantity of "device" can be one or more. In addition, the objects modified by different prefix words can be the same or different, for example, the description object is "device", and "first device" and "second device" can be the same device or different devices, and the types thereof can be the same or different. For another example, the description object is "information", and "first information" and "second information" can be the same information or different information, and the contents thereof can be the same or different.

[0098] In some embodiments, "including A", "containing A", "for indicating A", "carrying A" can be interpreted as directly carrying A, or indirectly indicating A.

[0099] In some embodiments, the terms of "in response to", "in response to determining", "in the case of", "when", "when", "if", "if" and the like can be replaced with each other.

[0100] In some embodiments, the terms of "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 lower than", "above" and the like can be replaced with each other, and the terms of "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", "below" and the like can be replaced with each other.

[0101] In some embodiments, an apparatus or the like can be interpreted as an entity, and can also be interpreted as virtual, and the name thereof is not limited to the name described in the embodiments. The terms "apparatus", "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "chip", "chip system", "entity", "subject" and the like can be replaced with each other.

[0102] In some embodiments, a "network" can be interpreted as an apparatus (for example, an access network device, a core network device, and the like) included in the network.

[0103] In some embodiments, the terms "access network device (AN device)", "radio access network device (RAN device)", "base station (BS)", "radio base station", "fixed station", "node", "access point", "transmission point (TP)", "reception point (RP)", "transmission / reception point (TRP)", "panel", "antenna panel", "antenna array", "cell", "macro cell", "small cell", "femto cell", "pico cell", "sector", "cell group", "serving cell", "carrier", "component carrier", "bandwidth part (BWP)" and the like can be replaced with each other.

[0104] In some embodiments, the terms "terminal," "terminal device," "user equipment (UE)," "user terminal," "mobile station (MS)," "mobile terminal (MT)," "subscriber station," "mobile unit," "subscriber unit," "wireless unit," "remote unit," "mobile device," "wireless device," "wireless communication device," "remote device," "mobile subscriber station," "access terminal," "mobile terminal," "wireless terminal," "remote terminal," "handset," "user agent," "mobile client," "client," and so on can be replaced with each other.

[0105] In some embodiments, the access network device, the core network device, or the network device can be replaced with a terminal. For example, the embodiments of the present disclosure can also be applied to a structure in which communication between the access network device, the core network device, or the network device and the terminal is replaced with communication between a plurality of terminals (e.g., device-to-device (D2D), vehicle-to-everything (V2X), etc.). In this case, the terminal can also be configured to have all or part of the functions of the access network device. In addition, the terms "uplink," "downlink," and the like can also be replaced with terms corresponding to the inter-terminal communication (e.g., "side"). For example, the uplink channel, the downlink channel, and the like can be replaced with the side channel, and the uplink, the downlink, and the like can be replaced with the sidelink.

[0106] In some embodiments, the terminal can be replaced with the access network device, the core network device, or the network device. In this case, the access network device, the core network device, or the network device can also be configured to have all or part of the functions of the terminal.

[0107] In some embodiments, obtaining data, information, and the like can comply with laws and regulations of the country where the location is.

[0108] In some embodiments, data, information, and the like can be obtained after obtaining the consent of the user.

[0109] In addition, each element, each row, or each column in the table of the embodiments of the present disclosure can be implemented as an independent embodiment, and any combination of any element, any row, or any column can also be implemented as an independent embodiment.

[0110] FIG. 1 is an architecture schematic diagram of a communication system according to an embodiment of the present disclosure.

[0111] As shown in FIG. 1, the communication system 100 includes a terminal 101 and a network device 102.

[0112] In some embodiments, the terminal 101 includes at least one of a mobile phone, a wearable device, an Internet of Things device, a communication-capable automobile, a smart automobile, a Pad, 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, a wireless terminal device in a smart home, and the like, but is not limited thereto.

[0113] In some embodiments, the network device 102 includes at least one of an access network device 102-1 and a core network device 102-2, but is not limited thereto.

[0114] In some embodiments, the access network device is, for example, a node or device that accesses a terminal to a wireless network, and the access network device can include at least one of an evolved NodeB (eNB) in a 5G communication system, a next generation eNB (ng-eNB), a next generation NodeB (gNB), a node B (NB), a home node B (HNB), a home evolved node B (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, an access node in a Wi-Fi system, but is not limited thereto.

[0115] In some embodiments, the technical solutions of the present disclosure can be applied to an Open RAN architecture, at this time, the interfaces between or within the access network devices involved in the embodiments of the present disclosure can become internal interfaces of the Open RAN, and the processes and information interactions between these internal interfaces can be realized through software or programs.

[0116] In some embodiments, the access network device can be composed of a central unit (CU) and a distributed unit (DU), wherein the CU can also be referred to as a control unit (control unit). The CU-DU structure can split the protocol layers of the access network device, and part of the functions of the protocol layers are controlled by the CU, and the remaining part or all of the functions of the protocol layers are distributed in the DU and controlled by the CU, but the present disclosure is not limited thereto.

[0117] In some embodiments, the core network device can be one device including multiple network elements, etc., or can be multiple devices or device groups, each including all or part of multiple network elements. The network element can be virtual or physical. The core network includes, for example, at least one of an evolved packet core (EPC), a 5G core network (5GCN), and a next generation core (NGC).

[0118] In some embodiments, in terms of identifying the function corresponding to the AI model (which can also be referred to as the ML model, and hereinafter collectively referred to as the AI model), one AI model can support one or more functions. The terminal can indicate the functions it supports for a specific sub-use case. In addition, there can be several models for each function, but the AI model deployed on the terminal is transparent to the network device. The network device does not know which model is suitable for the specified function.

[0119] For example, assume that the network device has two different beam codebook configurations, codebook #1 and codebook #2. Without prior knowledge of these configurations, the terminal deploys multiple AI models, model #1, model #2, model #3, and model #4. Each model is established based on certain assumptions about the beam shape. Assume that model #1 is adapted to codebook #1, and model 2 is adapted to codebook 2. If the network device uses codebook #1 during inference, the terminal blindly tests all four models and observes their performance in the absence of information about the network device configuration. Assume that model #1 is found to perform better. At this point, the terminal continues to use model #1 for inference. Without an explicit model identification process, the network device is always unaware of the existence of the four models on the terminal and cannot determine which model the terminal uses.

[0120] The above process is inefficient, introduces latency, and increases terminal power consumption. In addition, it is possible that none of the beam shapes used as training samples during model training matches the actual codebook, resulting in the model failing to achieve the corresponding function during inference.

[0121] To improve the reliability of the AI model used and ensure that the network device is aware of the AI model deployed on the terminal, the present disclosure provides the following model identification method, terminal, network device, system, and storage medium.

[0122] FIG. 2A is an interaction diagram of a model identification method according to an embodiment of the present disclosure. As shown in FIG. 2A, the present embodiment relates to a model identification method, and the method comprises:

[0123] In step S2101, the network device 102 sends first signaling to the terminal 101.

[0124] In some embodiments, the first signaling is used to instruct the terminal 101 to perform artificial intelligence (AI) model identification associated with a first function.

[0125] In one example, the first function can refer to an AI feature or feature group (FG) enabled by network configuration.

[0126] Exemplarily, the AI model can be associated with at least one of the following:

[0127] a specific network configuration;

[0128] a condition associated with a terminal capability of an AI feature or an FG;

[0129] an additional condition.

[0130] The additional condition can include, but is not limited to, a scenario, a site, a data set, and the like.

[0131] Exemplarily, the first function can include, but is not limited to, at least one of the following: a Transmission Configuration Indication (TCI) state of a Physical Downlink Shared Channel (PDSCH); beam correspondence; beam reporting; uplink beam management; beam failure recovery, and the like. The beam correspondence can also be referred to as beam prediction, which can predict a pair of optimal beams (including one transmitting beam and one receiving beam).

[0132] The above is only an exemplary description, and the present disclosure does not limit the first function.

[0133] In some embodiments, the first signaling can include, but is not limited to, at least one of the following:

[0134] Radio Resource Control (RRC) signaling;

[0135] Downlink Control Information (DCI);

[0136] a Medium Access Control Element (MAC CE).

[0137] The above is only an exemplary description, and the present disclosure does not specifically limit the signaling type of the first signaling.

[0138] In some embodiments, the AI model identification can refer to that the terminal 101 needs to select an optimal AI model from the deployed multiple AI models for the first function corresponding to a specific network configuration.

[0139] In one example, as shown in FIG. 2B, the first function can correspond to one or more AI models or sub-models, that is, the one or more AI models or sub-models can all implement the first function. It is necessary to determine the AI model or sub-model with the best performance from the one or more AI models or sub-models capable of implementing the first function (hereinafter, the AI model or sub-model is described uniformly, and it can be understood that the AI model can also be replaced by the sub-model).

[0140] In some embodiments, in addition to the AI model identification triggered by the network device 102 through the first signaling, the terminal can also be triggered to perform AI model identification through a predefined manner, such as when initially accessing the network device 102, or reaching a cycle of AI model identification, etc. The terminal 101 can automatically perform AI model identification. The present disclosure does not limit this.

[0141] In some embodiments, the terminal 101 receives the first signaling.

[0142] In some embodiments, the name of the first signaling is not limited and can be interchangeable with the model identification command, the model activation command, etc.

[0143] In some embodiments, the first signaling can include but is not limited to at least one of the following:

[0144] Function indication information;

[0145] A first condition set.

[0146] In one example, the function indication information can be used to indicate the first function, such as indicating the identity, index, etc. of the first function.

[0147] In one example, the first condition set can include one or more network configuration conditions associated with the first function.

[0148] Taking beam prediction as an example of the first function, the first condition set can include but is not limited to at least one of the following: the number of antenna ports on the terminal in an active state; the maximum moving speed of the terminal, etc.

[0149] The above is only an exemplary description, and the conditions associated with the first function configured by the network device 102 for the terminal 101 should all belong to the protection scope of the present disclosure.

[0150] Step S2102, the network device 102 sends measurement configuration information to the terminal 101.

[0151] In some embodiments, the measurement configuration information can configure the terminal 101 to perform information related to the first measurement associated with the first function.

[0152] In some embodiments, the measurement configuration information can include but is not limited to at least one of the following:

[0153] Information configuration that needs to be measured; beam set that needs to be measured; result type that needs to be measured.

[0154] Taking the first function as an example of beam prediction, the measurement configuration information can include: the information configuration to be measured includes reference signal (RS) configuration, the beam set to be measured can be a beam subset, and the result type to be measured can include a reference signal receiving power (RSRP) value. That is, the terminal 101 can measure the RSRP value on each beam in the beam subset.

[0155] In some embodiments, step S2101 and step S2102 can be combined for execution, for example, the network device 102 provides the measurement configuration information to the terminal 101 through the first signaling.

[0156] In some embodiments, step S2101 and step S2102 can be executed separately, for example, the network device 102 sends the first signaling, and in addition, provides the measurement configuration information to the terminal 101 through a separate signaling, for example, the fourth signaling.

[0157] In one example, the fourth signaling can include, but is not limited to, at least one of RRC signaling, MAC CE, and DCI.

[0158] In some embodiments, the terminal 101 receives the measurement configuration information.

[0159] In some embodiments, the name of the measurement configuration information is not limited, and can be interchangeable with configuration information, measurement configuration, etc.

[0160] Step S2103, the terminal 101 performs the first measurement associated with the first function to obtain the first measurement result.

[0161] In some embodiments, the terminal 101 performs the first measurement based on the measurement configuration information, thereby obtaining the first measurement result.

[0162] Taking beam prediction as an example, the terminal 101 measures the RSRP value on each beam in the beam subset to obtain the first measurement result.

[0163] The above is only an exemplary description, and the disclosure does not limit the specific content of performing the first measurement.

[0164] Step S2104, the terminal 101 determines the model identification result.

[0165] In some embodiments, the model identification result can include a first AI model, which is an AI model associated with the first function and having the best performance among the plurality of AI models.

[0166] In some embodiments, the terminal 101 can input the first measurement result described above into the plurality of AI models, obtain a prediction result associated with the first function output by each AI model. Among the plurality of prediction results, a first prediction result with the smallest difference from the true result (also referred to as the groud truth value) is determined. Among the plurality of AI models, the AI model outputting the first prediction result is determined as the first AI model, and the model identification result is obtained.

[0167] In some embodiments, the terminal 101 can determine the true result in any of the following ways:

[0168] Way 1, the terminal 101 can determine the true result by measurement.

[0169] Taking beam prediction as an example, the terminal 101 determines a pair of beams with the best performance by measuring the RSRP values of all beams, and determines the result as the true result.

[0170] Way 2, the true result is determined based on the third signaling sent by the network device 102.

[0171] Illustratively, the network device 102 can determine the true result and send it to the terminal 101 through the third signaling. The disclosure does not limit the way in which the network device 102 determines the true result, which can be a sample result obtained by the network device 102 in advance, or a true result measured by the network device 102.

[0172] In some embodiments, the terminal 101 can determine a plurality of AI models associated with the first function and having better performance among the plurality of AI models, and determine the plurality of AI models as the model identification result.

[0173] Step S2105, the terminal 101 sends the second signaling to the network device 102.

[0174] In some embodiments, the terminal 101 sends the second signaling based on the model identification result.

[0175] In some embodiments, the second signaling can include but is not limited to at least one of the following:

[0176] The model identifier of the first AI model;

[0177] The first prediction result.

[0178] The first prediction result is the prediction result associated with the first function output by the first AI model.

[0179] For example, taking beam prediction as an example, the terminal 101 sends the indexes of the predicted pair of best beams to the network device 102 through second signaling.

[0180] In some embodiments, the network device 102 receives the second signaling.

[0181] In some embodiments, the second signaling can be at least one of the following:

[0182] RRC signaling; MAC CE; uplink control information (UCI).

[0183] In some embodiments, the name of the second signaling is not limited and can be interchangeable with the identification result reporting signaling, the reporting signaling, or the report information.

[0184] Step S2106, the network device 102 establishes a corresponding relationship between the first condition set and the model identifier of the first AI model.

[0185] In some embodiments, the network device 102 can establish and maintain the above-mentioned corresponding relationship.

[0186] Step S2107, the network device 102 configures the model identifier of the first AI model based on the first corresponding relationship when the terminal is configured with the first condition set.

[0187] In some embodiments, when the network device 102 configures the terminal 101 with the first condition set later, the model identifier of the first AI model can be determined based on the maintained corresponding relationship, and the model identifier of the first AI model is configured to the terminal 101 through signaling, such as fifth signaling. After the terminal 101 receives the fifth signaling, it performs the first measurement, obtains the first measurement result, and directly inputs the first measurement result into the first AI model to obtain the prediction result output by the first AI model, which effectively reduces the model inference delay on the terminal 101 and has high availability.

[0188] In some embodiments, the above-mentioned model identification process can be understood as initial AI model identification. For example, for the network device 102, since the network configuration condition and / or channel state are constantly changing, it is expected that the terminal 101 determines the AI model with the best initial performance. If the network configuration condition and / or channel state change later, the model identification process can be performed again, or the terminal 101 can always select the AI model with the best initial performance. The present disclosure does not limit this.

[0189] In some embodiments, the total delay T of AI model identification is related to the behavior of the terminal 101 and can be determined based on at least one of the following:

[0190] a first time delay T1; a second time delay T2; a third time delay T3; a fourth time delay T4; a fifth time delay T5; a sixth time delay T6.

[0191] For example, as shown in FIG. 2C, T = T1+T2+T3+T4+T5+T6.

[0192] In one example, the first time delay T1 is a time delay from receiving the first signaling at the terminal 101 to successfully decoding the first signaling at the terminal 101.

[0193] In one example, the second time delay T2 is a time delay from successfully decoding the first signaling at the terminal 101 to performing the first measurement associated with the first function at the terminal 101.

[0194] In one example, the third time delay T3 is a time delay from performing the first measurement associated with the first function at the terminal 101 to determining the prediction result output by each AI model at the terminal 101.

[0195] Exemplarily, the third time delay T3 can be associated with an execution order of outputting the prediction result by the plurality of AI models.

[0196] For example, as shown in FIGS. 2D-1 to 2D-3, the execution order includes but is not limited to any one of the following:

[0197] the plurality of AI models output the prediction result in parallel;

[0198] the plurality of AI models output the prediction result in sequence;

[0199] part of the plurality of AI models output the prediction result in parallel.

[0200] Correspondingly, if the plurality of AI models output the prediction result in parallel, the third time delay T3 can be equal to a maximum time length of outputting the prediction result by the plurality of AI models.

[0201] Correspondingly, if the plurality of AI models output the prediction result in sequence, the third time delay T3 can be equal to a sum of time lengths of outputting the prediction result by the plurality of AI models.

[0202] Correspondingly, if part of the plurality of AI models output the prediction result in parallel, the third time delay T3 can be greater than a maximum time length of outputting the prediction result by the plurality of AI models and less than a sum of time lengths of outputting the prediction result by the plurality of AI models.

[0203] In one example, the fourth time delay T4 is a time delay from determining the prediction result output by each AI model at the terminal 101 to obtaining the real result.

[0204] In one example, the fifth time delay T5 is a time delay from obtaining the real result by the terminal 101 to determining the model recognition result.

[0205] In one example, the sixth time delay T6 is a time delay from determining the model recognition result by the terminal 101 to sending the second signaling.

[0206] In one example, at least one of T1 to T6 can be 0 or shortened in an appropriate manner. For example, the network device 102 sends the real result to the terminal 101 in advance, and the fourth time delay T4 can be 0. For another example, after the terminal 101 determines the model recognition result, the terminal 101 generally needs to send the second signaling to the network device 102 through the last uplink resource. The network device 102 can shorten the sixth time delay T6 through scheduling. For another example, the terminal 101 has strong capability and can support measuring the RSRP value on multiple beams at the same time, so that the second time delay T2 can be shortened.

[0207] The above is only an example for illustration. The terminal 101 and the network device 102 need to understand the total time delay of AI model recognition consistently.

[0208] In one example, the network device 102 can set the timer length based on the total time delay, so as to avoid the influence of long-time waiting for the model recognition result on the normal execution of functions.

[0209] In one example, the terminal 101 can select the most suitable uplink resource to send the second signaling based on the uplink resource scheduling, so as to reduce the total time delay of the model recognition.

[0210] In some embodiments, the names of information and the like are not limited to the names described in the embodiments, and terms such as “information”, “message”, “signal”, “signaling”, “report”, “configuration”, “indication”, “instruction”, “command”, “channel”, “parameter”, “domain”, “field”, “symbol”, “symbol”, “codebook”, “codeword”, “code point”, “bit”, “data”, “program”, “chip”, and the like can be replaced with each other.

[0211] In some embodiments, the terms “downlink control information (DCI),” “downlink (DL) assignment,” “DL DCI,” “uplink (UL) grant,” “UL DCI,” and the like can be replaced with each other.

[0212] In some embodiments, “acquire,” “obtain,” “get,” “receive,” “transmit,” “bidirectional transmission,” “send and / or receive,” and the like can be replaced with each other, which can be interpreted as receiving from other subjects, acquiring from protocols, acquiring from higher layers, obtaining by oneself, implementing autonomously, and the like.

[0213] In some embodiments, the terms “send,” “transmit,” “report,” “issue,” “transmit,” “bidirectional transmission,” “send and / or receive,” and the like can be replaced with each other.

[0214] The model identification method related to the embodiments of the present disclosure can include at least one of steps S2101-S2107. For example, step S2101 can be implemented as an independent embodiment, step S2102 can be implemented as an independent embodiment, step S2103 can be implemented as an independent embodiment, steps S2101+S2102+S2103 can be implemented as an independent embodiment, step S2104 can be implemented as an independent embodiment, step S2105 can be implemented as an independent embodiment, steps S2106+S2107 can be implemented as an independent embodiment, steps S2101-S2107 can be implemented as an independent embodiment, but not limited thereto.

[0215] In some embodiments, step S2101 is optional, and one or more of these steps can be omitted or replaced in different embodiments. For example, the terminal 101 automatically triggers model identification, and step S2101 can not be performed.

[0216] In some embodiments, steps S2102-S2103 are optional, and one or more of these steps can be omitted or replaced in different embodiments. For example, the terminal 101 does not need to perform measurement, and steps S2102-S2103 can not be performed.

[0217] In some embodiments, step S2105 is optional, and one or more of these steps can be omitted or replaced in different embodiments. For example, the network device 102 obtains the second signaling from other execution subjects, and step S2105 can not be performed.

[0218] In some embodiments, steps S2106-S2107 are optional, and one or more of these steps can be omitted or replaced in different embodiments. For example, steps S2106-S2107 can not be performed each time model identification needs to be performed again.

[0219] In some embodiments, steps S2101-S2107 are optional, and one or more of these steps can be omitted or replaced in different embodiments.

[0220] In some embodiments, the order of execution of steps S2101-S2107 is not limited.

[0221] In the above embodiments, the terminal can determine a model identification result among a plurality of AI models based on the first signaling sent by the network device, and send second signaling to the network device based on the model identification result. By sending the model identification result to the network device, the network device is ensured to be clear about the AI model associated with the specified function and having the best performance, the reliability of the used AI model is improved, and the reliability of AI and communication integration is improved, which is conducive to the common development of communication and AI and has high usability.

[0222] FIG. 3A is an interaction schematic diagram of a model identification method according to an embodiment of the present disclosure. As shown in FIG. 3A, the present embodiment relates to a model identification method, which is performed by the terminal 101 and includes:

[0223] Step S3101: Obtain first signaling.

[0224] In some embodiments, the first signaling is used to instruct the terminal 101 to perform artificial intelligence (AI) model identification associated with a first function.

[0225] In some embodiments, the terminal 101 can obtain the first signaling from the network device 102, but is not limited thereto and can also receive first signaling sent by other subjects.

[0226] In some embodiments, the terminal 101 obtains the first signaling specified by a protocol.

[0227] In some embodiments, the terminal 101 obtains the first signaling from upper layer(s).

[0228] In some embodiments, the terminal 101 processes to obtain the first signaling.

[0229] In some embodiments, step S3101 is omitted, and the terminal 101 autonomously implements the function instructed by the first signaling, or the terminal 101 obtains the first signaling based on a predefined rule or protocol agreement, or the above function is default or default. In some embodiments, the terminal 101 obtains the first signaling from the network device 102, but is not limited thereto and can also receive first signaling sent by other subjects.

[0230] In some embodiments, the optional implementation of step S3101 can refer to the optional implementation of step S2101 in FIG. 2A and other associated parts in the embodiments related to FIG. 2A, which will not be repeated here.

[0231] Step S3102: Obtain measurement configuration information.

[0232] In some embodiments, the measurement configuration information can configure the terminal 101 to perform information related to the first measurement associated with the first function.

[0233] In some embodiments, the terminal 101 can obtain the measurement configuration information from the network device 102, but is not limited thereto, and can also receive the measurement configuration information sent by other subjects.

[0234] In some embodiments, the terminal 101 obtains the measurement configuration information specified by a protocol.

[0235] In some embodiments, the terminal 101 obtains the measurement configuration information from an upper layer.

[0236] In some embodiments, the terminal 101 processes to obtain the measurement configuration information.

[0237] In some embodiments, step S3102 is omitted, and the terminal 101 autonomously implements the function indicated by the measurement configuration information, or the terminal 101 obtains the measurement configuration information based on a pre-defined rule or protocol agreement, or the above function is default or default.

[0238] In some embodiments, the optional implementation of step S3102 can refer to the optional implementation of step S2102 in FIG. 2A and other associated parts in the embodiments related to FIG. 2A, which will not be repeated here.

[0239] Step S3103: Determine the first measurement result.

[0240] In some embodiments, the optional implementation of step S3103 can refer to the optional implementation of step S2103 in FIG. 2A and other associated parts in the embodiments related to FIG. 2A, which will not be repeated here.

[0241] Step S3104: Determine the model identification result.

[0242] In some embodiments, the optional implementation of step S3104 can refer to the optional implementation of step S2104 in FIG. 2A and other associated parts in the embodiments related to FIG. 2A, which will not be repeated here.

[0243] Step S3105: Send the second signaling.

[0244] In some embodiments, the second signaling is used to indicate the model recognition result.

[0245] In some embodiments, the terminal 101 sends the second signaling to the network device 102.

[0246] In some embodiments, the network device 102 receives the second signaling.

[0247] In some embodiments, the optional implementation of step S3105 can refer to the optional implementation of step S2105 in FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which are not described herein again.

[0248] In some embodiments, steps S3101 to S3105 are optional, and one or more of the steps can be omitted or replaced in different embodiments.

[0249] In some embodiments, the execution order of steps S3101 to S3105 is not limited.

[0250] In the above embodiments, the terminal determines the model recognition result among multiple AI models based on the first signaling sent by the network device, and sends the second signaling to the network device to report the model recognition result. This ensures that the network device is clear about the AI model associated with the specified function and the performance of the AI model is optimal, improves the reliability of the AI model used, and improves the reliability of AI and communication integration, which is conducive to the joint development of communication and AI, and has high usability.

[0251] FIG. 3B is an interaction schematic diagram of a model recognition method according to an embodiment of the present disclosure. As shown in FIG. 3B, the embodiment of the present disclosure relates to a model recognition method, and the above method is executed by the terminal 101, which includes:

[0252] Step S3201, obtaining first signaling.

[0253] In some embodiments, the first signaling is used to instruct the terminal 101 to perform artificial intelligence (AI) model recognition associated with a first function.

[0254] In some embodiments, the terminal 101 can obtain the first signaling from the network device 102, but is not limited thereto, and can also receive the first signaling sent by other subjects.

[0255] In some embodiments, the terminal 101 obtains the first signaling specified by a protocol.

[0256] In some embodiments, the terminal 101 obtains the first signaling from an upper layer.

[0257] In some embodiments, the terminal 101 processes to obtain the first signaling.

[0258] In some embodiments, step S3201 is omitted, and the terminal 101 autonomously implements the function indicated by the first signaling, or the terminal 101 acquires the first signaling based on a predefined rule or protocol agreement, or the above function is default or default.

[0259] In some embodiments, the optional implementation of step S3201 can refer to the optional implementation of step S2101 in FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0260] Step S3202, determining the model recognition result.

[0261] In some embodiments, the optional implementation of step S3202 can refer to the optional implementation of step S2104 in FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0262] Step S3203, sending second signaling.

[0263] In some embodiments, the second signaling is used to report the model recognition result.

[0264] In some embodiments, the terminal 101 sends the second signaling to the network device 102.

[0265] In some embodiments, the network device 102 receives the second signaling.

[0266] In some embodiments, the optional implementation of step S3203 can refer to the optional implementation of step S2105 in FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0267] In some embodiments, steps S3201 to S3203 are optional, and one or more of these steps can be omitted or replaced in different embodiments.

[0268] In some embodiments, the execution order of steps S3201 to S3203 is not limited.

[0269] In the above embodiments, the terminal determines the model recognition result based on the acquired first signaling, and sends the second signaling to the network device to report the model recognition result. This ensures that the network device is clear about the AI model associated with the specified function and the optimal performance, improves the reliability of the used AI model, and improves the reliability of AI and communication integration, which is conducive to the joint development of communication and AI, and has high usability.

[0270] FIG. 3C is an interaction schematic diagram of a model identification method according to an embodiment of the present disclosure. As shown in FIG. 3C, the embodiment of the present disclosure relates to a model identification method, which is performed by the network device 102 and includes the following steps.

[0271] In step S3301, the first signaling is sent.

[0272] In some embodiments, the first signaling is used to instruct the terminal 101 to perform artificial intelligence (AI) model identification associated with the first function.

[0273] In some embodiments, the network device 102 sends the first signaling to the terminal 101.

[0274] In some embodiments, the terminal 101 receives the first signaling.

[0275] In some embodiments, the optional implementation of step S3301 can refer to the optional implementation of step S2101 in FIG. 2A and other associated parts in the embodiments related to FIG. 2A, which will not be described here.

[0276] In step S3302, the measurement configuration information is sent.

[0277] In some embodiments, the measurement configuration information can configure the terminal 101 to perform information related to the first measurement associated with the first function.

[0278] In some embodiments, the network device 102 sends the measurement configuration information to the terminal 101.

[0279] In some embodiments, the terminal 101 receives the measurement configuration information.

[0280] In some embodiments, the optional implementation of step S3302 can refer to the optional implementation of step S2102 in FIG. 2A and other associated parts in the embodiments related to FIG. 2A, which will not be described here.

[0281] In step S3303, the second signaling is acquired.

[0282] In some embodiments, the second signaling is used to indicate the model identification result.

[0283] In some embodiments, the network device 102 can acquire the second signaling from the terminal 101, but is not limited thereto, and can also receive the second signaling sent by other subjects.

[0284] In some embodiments, the network device 102 acquires the second signaling specified by a protocol.

[0285] In some embodiments, the network device 102 acquires the second signaling from an upper layer.

[0286] In some embodiments, the network device 102 processes to obtain the second signaling.

[0287] In some embodiments, the step S3303 is omitted, the network device 102 autonomously implements the function indicated by the second signaling, or the network device 102 acquires the second signaling based on a predefined rule or protocol agreement, or the above function is default or default.

[0288] In some embodiments, the optional implementation of step S3303 can refer to the optional implementation of step S2105 of FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0289] Step S3304, establishing a corresponding relationship.

[0290] In some embodiments, the corresponding relationship is a corresponding relationship between the first condition set and the model identifier of the first AI model.

[0291] In some embodiments, the optional implementation of step S3304 can refer to the optional implementation of step S2106 of FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0292] Step S3305, configuring the first condition set and the model identifier of the first AI model.

[0293] In some embodiments, the optional implementation of step S3304 can refer to the optional implementation of step S2107 of FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0294] In some embodiments, steps S3301 to S3305 are optional, and one or more of these steps can be omitted or replaced in different embodiments.

[0295] In some embodiments, the execution order of steps S3301 to S3305 is not limited.

[0296] In the above embodiments, the network device can trigger the model identification process through the first signaling, determine the model identification result through the second signaling, establish a corresponding relationship, and subsequently directly configure the model identifier of the first AI model, ensuring that the network device is associated with the specified function and the optimal AI model, improving the reliability of the used AI model, and improving the reliability of AI and communication integration, which is conducive to the common development of communication and AI, and has high usability.

[0297] FIG. 3D is an interaction schematic diagram of a model identification method according to an embodiment of the present disclosure. As shown in FIG. 3D, the embodiment of the present disclosure relates to a model identification method, which is performed by the network device 102 and includes the following steps.

[0298] In step S3401, the first signaling is sent.

[0299] In some embodiments, the first signaling is used to instruct the terminal 101 to perform the artificial intelligence (AI) model identification associated with the first function.

[0300] In some embodiments, the network device 102 sends the first signaling to the terminal 101.

[0301] In some embodiments, the terminal 101 receives the first signaling.

[0302] In some embodiments, the optional implementation of step S3401 can refer to the optional implementation of step S2101 in FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0303] In step S3402, the second signaling is obtained.

[0304] In some embodiments, the second signaling is used to indicate the model identification result.

[0305] In some embodiments, the network device 102 can obtain the second signaling from the terminal 101, but is not limited thereto, and can also receive the second signaling sent by other subjects.

[0306] In some embodiments, the network device 102 obtains the second signaling specified by a protocol.

[0307] In some embodiments, the network device 102 obtains the second signaling from the upper layer(s).

[0308] In some embodiments, the network device 102 processes to obtain the second signaling.

[0309] In some embodiments, step S3402 is omitted, and the network device 102 autonomously implements the function indicated by the second signaling, or the network device 102 obtains the second signaling based on a pre-defined rule or protocol agreement, or the above function is default or default.

[0310] In some embodiments, the optional implementation of step S3402 can refer to the optional implementation of step S2105 in FIG. 2A and other associated parts in the embodiments involved in FIG. 2A, which will not be repeated here.

[0311] In some embodiments, steps S3402 to S3402 are optional, and one or more of these steps can be omitted or replaced in different embodiments.

[0312] In some embodiments, the order of execution of steps S3402 to S3402 is not limited.

[0313] In the above embodiments, the network device can trigger the model identification process through the first signaling and determine the model identification result through the second signaling, ensuring that the network device is associated with the specified function and the AI model with the optimal performance, improving the reliability of the used AI model and the reliability of AI and communication fusion, which is conducive to the joint development of communication and AI and has high usability.

[0314] The above process is further illustrated as follows.

[0315] Embodiment 1, initial AI model identification:

[0316] Referring to FIG. 2B, when a specific configuration is used to activate a function, the terminal needs to select the best model for the function. This process is also called model identification.

[0317] One potential method to reduce latency and power consumption overhead is to perform model identification after model performance monitoring. In this method, once the best model is identified through monitoring, the terminal can initiate model identification. And notify the network device that the terminal has a model that performs reasonably well under certain network device side conditions, and request to assign a model identifier. Then, the network device can establish a correlation between the assigned model identifier and the network device side conditions, so that it can configure the model identifier for the terminal in the future when the same conditions occur. Therefore, for a given network device side condition, the terminal only needs to perform monitoring once.

[0318] When a function is activated, for example as shown in FIG. 4, the terminal will select the best model by comparing the execution of multiple models:

[0319] Step S4101, the network device sends a model identification command (or a function command or a model activation command) to the terminal through at least one of RRC, MAC CE and DCI, and configures certain conditions.

[0320] Step S4102, the network device configures RS for the terminal to perform measurements for prediction, such as a subset of beams.

[0321] Step S4103, the terminal makes predictions through different models or sub-models according to the measurement results.

[0322] Step S4104, the terminal performs measurements to obtain ground truth, such as the full set of beams.

[0323] Step S4105, the terminal compares the result with the ground truth, and selects the best model with the most accurate result.

[0324] Step S4106, the terminal reports the model identifier or the predicted value to the network device according to the uplink signal. Due to the uncertainty of the uplink signal, there may be a delay.

[0325] Exemplarily, some of the steps can be skipped. For example, in step S4102, the network device can send the ground truth of the beam to the terminal, and then the terminal can skip step S4104.

[0326] After the network device and the terminal perform model identification, the network device can establish a connection between the allocated model identifier and certain additional conditions on the network device side. Therefore, whenever the network device encounters these conditions when communicating with the terminal, it can automatically configure the corresponding model identifier for the terminal. Therefore, for a specific set of network device side conditions, the terminal will only need to go through model identification once.

[0327] Embodiment 2, model identification delay:

[0328] The total delay of the function / model activation delay can include the following processes, as shown in FIG. 2C.

[0329] The total time delay T of model identification will be defined as:

[0330] T=T1+T2+T3+T4+T5+T6;

[0331] Wherein, T1: command decoding time; T2: UE measurement delay for inference; T3: UE prediction delay through AI model; T4: UE measurement delay of ground truth; T5: UE processing delay to obtain the best model; T6: UE reporting delay.

[0332] For T3, the terminal can make predictions through multiple AI models in parallel, partially in parallel, or sequentially. Then the delay is different. It needs to consider different options, as shown in FIG. 2D-1, FIG. 2D-2, and FIG. 2D-3.

[0333] In the embodiments of the present disclosure, part or all of the steps, and optional implementation manners thereof, can be combined with part or all of the steps in other embodiments, or can be combined with optional implementation manners in other embodiments.

[0334] The embodiments of the present disclosure also propose an apparatus for implementing any of the above methods, for example, an apparatus comprising units or modules for implementing each step performed by the terminal in any of the above methods. For another example, another apparatus is proposed, comprising units or modules for implementing each step performed by the network device in any of the above methods.

[0335] It should be understood that the division of each unit or module in the above apparatus is only a logical function division, and all or part of them can be integrated into a physical entity or physically separated in actual implementation. In addition, the units or modules in the apparatus can be implemented in the form of processor calling software: for example, the apparatus includes a processor connected with a memory, the memory stores instructions, and the processor calls the instructions stored in the memory to implement any of the above methods or realize the functions of the units or modules of the above apparatus, wherein the processor is, for example, a general processor such as a central processing unit (CPU) or a microprocessor, and the memory is a memory in the apparatus or a memory outside the apparatus. Alternatively, the units or modules in the apparatus can be implemented in the form of hardware circuit, and the functions of part or all of the units or modules can be realized by the design of the hardware circuit. The above hardware circuit can be understood as one or more processors; for example, in one implementation, the above hardware circuit is an application-specific integrated circuit (ASIC), and the functions of part or all of the units or modules are realized by the design of the logical relationship of the elements in the circuit; for another example, in another implementation, the above hardware circuit is a programmable logic device (PLD), and a field programmable gate array (FPGA) is taken as an example, which can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by a configuration file, so as to realize the functions of part or all of the units or modules. All units or modules of the above apparatus can be implemented in the form of processor calling software, or all units or modules can be implemented in the form of hardware circuit, or part of the units or modules are implemented in the form of processor calling software, and the remaining part is implemented in the form of hardware circuit.

[0336] In the embodiments of the present disclosure, the processor is a circuit with signal processing capability. In one implementation, the processor can be a circuit with instruction reading and running capability, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), a digital signal processor (DSP), or the like. In another implementation, the processor can implement certain functions through a logical relationship of hardware circuits, and the logical relationship of the hardware circuits is fixed or can be reconfigured. 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 the reconfigurable hardware circuit, the processor loads a configuration document to implement the configuration of the hardware circuit. It can be understood that the processor loads instructions to implement the functions of the above part or all units or modules. In addition, the hardware circuit can also be 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), or the like.

[0337] FIG. 5A is a structural schematic diagram of a terminal according to an embodiment of the present disclosure. As shown in FIG. 5A, the terminal 5100 can include a transceiver module 5101 and a processing module 5102.

[0338] In some embodiments, the transceiver module 5101 is configured to receive first signaling sent by a network device; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model identification associated with a first function.

[0339] In some embodiments, the processing module 5102 is configured to determine a model identification result in a plurality of AI models based on the first signaling; wherein the model identification result includes a first AI model, and the first AI model is an AI model in the plurality of AI models that is associated with the first function and has the best performance.

[0340] In some embodiments, the transceiver module 5101 is further configured to send second signaling to the network device based on the model identification result.

[0341] Optionally, the transceiver module 5101 is configured to perform at least one of the communication steps (for example, step S2101, step S2102, step S2105, but not limited to) of the terminal 5100 in any of the above methods. Details are not repeated here.

[0342] Optionally, the processing module 5102 is configured to perform at least one of the other steps (for example, step S2103, step S2104, but not limited to) of the terminal 5100 in any of the above methods. Details are not repeated here.

[0343] FIG. 5B is a structural schematic diagram of a network device according to an embodiment of the present disclosure. As shown in FIG. 5B, the network device 5200 can include a transceiver module 5201.

[0344] In some embodiments, the transceiver module 5201 is configured to send first signaling to the terminal; wherein the first signaling is used to instruct the terminal to perform artificial intelligence AI model identification associated with a first function; and receive second signaling sent by the terminal based on a model identification result; wherein the model identification result includes a first AI model, and the first AI model is an AI model associated with the first function and having the best performance among a plurality of AI models deployed on the terminal.

[0345] Optionally, the transceiver module 5201 is configured to perform at least one of the communication steps (for example, step S2101, step S2102, step S2105, but not limited to) of the network device 5200 in any of the above methods. Details are not repeated here.

[0346] In some embodiments, the sending module and / or the receiving module can be referred to as a transceiver module, and the sending module and the receiving module can be separate or integrated together. Optionally, the transceiver module can be mutually replaced with the transceiver.

[0347] In some embodiments, the processing module can be one module, or can include a plurality of sub-modules. Optionally, the plurality of sub-modules respectively perform all or part of the steps required to be performed by the processing module. Optionally, the processing module can be mutually replaced with the processor.

[0348] FIG. 6A is a structural schematic diagram of a communication device 6100 according to an embodiment of the present disclosure. The communication device 6100 can be a terminal (for example, a user equipment, a vehicle, an Internet of Things device, etc.) or a network device (for example, an access network device, a core network device, etc.), can be a chip, a chip system, or a processor supporting the terminal to implement any of the above methods, or can be a chip, a chip system, or a processor supporting the network device to implement any of the above methods. The communication device 6100 can be used to implement the methods described in the above method embodiments, and details can be referred to the descriptions in the above method embodiments.

[0349] As shown in FIG. 6A, the communication device 6100 includes one or more processors 6101. The processor 6101 can be a general processor or a special-purpose processor, etc., such as a baseband processor or a central processing unit. The baseband processor can be configured to process communication protocols and communication data, and the central processing unit can be configured to control a communication apparatus (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 data of the programs. Optionally, the communication device 6100 is configured to perform any of the above methods. Optionally, the one or more processors 6101 are configured to invoke instructions to cause the communication device 6100 to perform any of the above methods.

[0350] In some embodiments, the communication device 6100 further includes one or more transceivers 6102. When the communication device 6100 includes one or more transceivers 6102, the transceiver 6102 performs at least one of the communication steps (e.g., steps S2101, S2102, S2105, but not limited to) in the above methods, and the processor 7101 performs at least one of the other steps (e.g., steps S2103, S2104, S2106, S2107, but not limited to). In optional embodiments, the transceiver can include a receiver and / or a transmitter, which can be separate or integrated together. Optionally, the terms transceiver, transceiving unit, transceiver, transceiving circuit, interface circuit, interface, etc. can be replaced by each other, and the terms transmitter, transmitting unit, transmitter, transmitting circuit, etc. can be replaced by each other, and the terms receiver, receiving unit, receiver, receiving circuit, etc. can be replaced by each other.

[0351] In some embodiments, the communication device 6100 further includes one or more memories 6103 for storing data. Optionally, all or part of the memory 6103 can also be outside the communication device 6100. In optional embodiments, the communication device 6100 can include one or more interface circuits 6104. Optionally, the interface circuit 6104 is connected to the memory 6102, and the interface circuit 6104 can be configured to receive data from the memory 6102 or other devices, and can be configured to send data to the memory 6102 or other devices. For example, the interface circuit 6104 can read data stored in the memory 6102 and send the data to the processor 6101.

[0352] The communication device 6100 described in the above embodiments can be a network device or a terminal, but the scope of the communication device 6100 described in the present disclosure is not limited thereto, and the structure of the communication device 6100 can not be limited by FIG. 6A. The communication device can be a standalone device or can be part of a larger device. For example, the communication device can be: 1) a standalone integrated circuit (IC), or a chip, or a chip system or subsystem; (2) a set of one or more ICs, which can optionally also include storage components for storing data, programs; (3) an ASIC, such as a modem; (4) a module that can be embedded in other devices; (5) a receiver, a terminal device, a smart terminal device, a cellular phone, a wireless device, a handset, a mobile unit, a vehicle-mounted device, a network device, a cloud device, an artificial intelligence device, and the like; (6) other devices, and the like.

[0353] FIG. 6B is a structural schematic diagram of a chip 6200 according to an embodiment of the present disclosure. For the case where the communication device 6100 is a chip or a chip system, the structural schematic diagram of the chip 6200 shown in FIG. 6B can be referred to, but is not limited thereto.

[0354] The chip 6200 includes one or more processors 6201. The chip 6200 is configured to perform any of the above methods.

[0355] In some embodiments, the chip 6200 further includes one or more interface circuits 6202. Optionally, the terms interface circuit, interface, transceiver pin, and the like can be replaced with each other. In some embodiments, the chip 6200 further includes one or more memories 6203 for storing data. Optionally, all or part of the memory 6203 can be outside the chip 6200. Optionally, the interface circuit 6202 is connected to the memory 6203, and the interface circuit 6202 can be configured to receive data from the memory 6203 or other devices, and the interface circuit 6202 can be configured to send data to the memory 6203 or other devices. For example, the interface circuit 6202 can read data stored in the memory 6203 and send the data to the processor 6201.

[0356] In some embodiments, the interface circuit 6202 performs at least one of the communication steps (such as steps S2101, S2102, S2105, but not limited thereto) of transmitting and / or receiving in the above methods. The interface circuit 6202 performing the communication steps such as transmitting and / or receiving in the above methods means that the interface circuit 6202 performs data interaction between the processor 6201, the chip 6200, the memory 6203, or a transceiver device. In some embodiments, the processor 6201 performs at least one of the other steps (such as steps S2103, S2104, S2106, S2107, but not limited thereto).

[0357] The modules and / or devices described in each embodiment of the virtual device, physical device, chip, etc. can be combined or separated according to circumstances. Optionally, part or all of the steps can also be performed by multiple modules and / or devices in cooperation, which is not limited here.

[0358] The disclosure further proposes a storage medium, and the storage medium stores instructions, which, when executed on the communication device 6100, causes the communication device 6100 to perform 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 to this, and it can also be a storage medium readable by other devices. Optionally, the storage medium can be a non-transitory storage medium, but is not limited to this, and it can also be a transitory storage medium.

[0359] The disclosure further proposes a program product, which, when executed by the communication device 6100, causes the communication device 6100 to perform any of the above methods. Optionally, the program product is a computer program product.

[0360] The disclosure further proposes a computer program, which, when executed on a computer, causes the computer to perform any of the above methods.

[0361] Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the features of the disclosure disclosed herein. The disclosure is intended to cover any variations, uses or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure that come within known or customary practice in the art to which the disclosure pertains. The specification and examples are to be regarded as illustrative only, and the true scope and spirit of the disclosure are indicated by the following claims.

[0362] It should be understood that the present disclosure is not limited to the precise structures described and shown in the drawings, and that various modifications and changes can be made without departing from the scope of the present disclosure. The scope of the present disclosure is limited only by the claims that follow.

Claims

1. A model recognition method, characterized in that, The method is executed by a terminal, and the method includes: The terminal receives a first signaling message sent by a network device; wherein the first signaling message is used to instruct the terminal to perform artificial intelligence (AI) model recognition associated with a first function; Based on the first signaling, a model recognition result is determined among multiple AI models; wherein, the model recognition result includes a first AI model, which is the AI ​​model among the multiple AI models that is associated with the first function and has the best performance; Based on the model recognition results, a second signaling message is sent to the network device.

2. The method according to claim 1, characterized in that, The first signaling includes at least one of the following: Function indication information, wherein the function indication information is used to indicate the first function; A first set of conditions, which includes one or more network configuration conditions associated with the first function.

3. The method according to claim 1 or 2, characterized in that, The method further includes: Receive measurement configuration information sent by the network device; Based on the measurement configuration information, a first measurement associated with the first function is performed to obtain a first measurement result.

4. The method according to claim 3, characterized in that, Determining the model recognition result among multiple AI models includes: The first measurement result is input into the multiple AI models to obtain the prediction result of each AI model that is associated with the first function; Among multiple prediction results, determine the first prediction result that differs least from the actual result; Among the multiple AI models, the AI ​​model that outputs the first prediction result is identified as the first AI model, and the model recognition result is obtained.

5. The method according to claim 4, characterized in that, The method further includes any one of the following: The true result is determined by measurement; The actual result is determined based on the third signaling sent by the network device.

6. The method according to any one of claims 1-5, characterized in that, The second signaling includes at least one of the following: The model identifier of the first AI model; The first prediction result is the prediction result output by the first AI model that is associated with the first function.

7. The method according to any one of claims 1-6, characterized in that, The first signaling includes at least one of the following: Radio Resource Control (RRC) signaling; Downlink Control Information (DCI); Media Access Control Unit (MAC CE) 8. The method according to any one of claims 1-7, characterized in that, The total latency of the AI ​​model identification is determined based on at least one of the following: The first delay is the delay from receiving the first signaling to successfully decoding the first signaling; The second delay is the delay from successfully decoding the first signaling to completing the first measurement associated with the first function; The third delay is the delay from the completion of the first measurement associated with the first function to the determination of the prediction result output by each AI model; The fourth delay is the delay from determining the prediction result output by each AI model to obtaining the actual result; The fifth delay is the delay from obtaining the true result to determining the model's recognition result; The sixth delay is the delay from determining the model recognition result to sending the second signaling.

9. The method according to claim 8, characterized in that, The third time delay is associated with the execution order in which the prediction results are output by the multiple AI models; wherein the execution order includes any one of the following: The multiple AI models output the prediction results in parallel; The multiple AI models sequentially output the prediction results; Some of the multiple AI models output the prediction results in parallel.

10. A model recognition method, characterized in that, The method is performed by a network device, and the method includes: Send a first signaling message to the terminal; wherein the first signaling message is used to instruct the terminal to perform artificial intelligence (AI) model recognition associated with a first function; The terminal receives a second signaling message based on the model recognition result; wherein the model recognition result includes a first AI model, and the first AI model is the AI ​​model with the best performance among the multiple AI models deployed on the terminal that is associated with the first function.

11. The method according to claim 10, characterized in that, The first signaling includes at least one of the following: Function indication information, wherein the function indication information is used to indicate the first function; A first set of conditions, which includes one or more network configuration conditions associated with the first function.

12. The method according to claim 10 or 11, characterized in that, The method further includes: Send measurement configuration information to the terminal; wherein the measurement configuration information is used to configure the terminal to perform a first measurement associated with the first function.

13. The method according to any one of claims 10-12, characterized in that, The method further includes: A third signaling message is sent to the terminal, the third signaling message including the actual result associated with the first function.

14. The method according to any one of claims 10-13, characterized in that, The second signaling includes at least one of the following: The model identifier of the first AI model; The first prediction result is the prediction result output by the first AI model that is associated with the first function.

15. The method according to any one of claims 10-14, characterized in that, The first signaling includes at least one of the following: Radio Resource Control (RRC) signaling; Downlink Control Information (DCI); Media Access Control Unit (MAC CE) 16. The method according to any one of claims 10-15, characterized in that, The total latency of the AI ​​model identification is determined based on at least one of the following: The first delay is the delay from when the terminal receives the first signaling to when the terminal successfully decodes the first signaling; The second delay is the delay from when the terminal successfully decodes the first signaling to when the terminal completes the first measurement associated with the first function; The third delay is the delay from when the terminal completes the first measurement associated with the first function to when the terminal determines the prediction result output by each AI model; The fourth delay is the delay from when the terminal determines the prediction result output by each AI model to when the terminal obtains the actual result; The fifth delay is the delay from when the terminal obtains the real result to when the terminal determines the model recognition result; The sixth delay is the delay from when the terminal determines the model recognition result to when the terminal sends the second signaling.

17. The method according to claim 16, characterized in that, The third time delay is associated with the execution order in which the prediction results are output by the multiple AI models; wherein the execution order includes any one of the following: The multiple AI models output the prediction results in parallel; The multiple AI models sequentially output the prediction results; Some of the multiple AI models output the prediction results in parallel.

18. The method according to any one of claims 10-17, characterized in that, The method also includes: Establish a correspondence between a first set of conditions and the model identifier of the first AI model; wherein the first set of conditions includes one or more network configuration conditions associated with the first function.

19. The method according to claim 18, characterized in that, The method further includes: When the terminal is configured with the first set of conditions, the model identifier of the first AI model is configured based on the first correspondence.

20. A terminal, characterized in that, include: The transceiver module is configured to receive a first signaling sent by a network device; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model recognition associated with a first function; The processing module is configured to determine a model recognition result among multiple AI models based on the first signaling; wherein the model recognition result includes a first AI model, which is the AI ​​model among the multiple AI models that is associated with the first function and has the best performance; The transceiver module is also configured to send a second signaling to the network device based on the model recognition result.

21. A network device, characterized in that, include: The transceiver module is configured to send a first signaling to the terminal; wherein the first signaling is used to instruct the terminal to perform artificial intelligence (AI) model recognition associated with a first function; The transceiver module is further configured to receive a second signaling sent by the terminal based on the model recognition result; wherein the model recognition result includes a first AI model, the first AI model being one of multiple AI models deployed on the terminal that is similar to the first AI model. A functionally relevant and optimally performing AI model.

22. A terminal, characterized in that, include: One or more processors; The processor is used to execute the model recognition method according to any one of claims 1-9.

23. A network device, characterized in that, include: One or more processors; The processor is used to execute the model recognition method according to any one of claims 10-19.

24. A model recognition system, characterized in that, include: A terminal, configured to implement the model recognition method according to any one of claims 1-9; A network device configured to implement the model recognition method according to any one of claims 10-19.

25. A storage medium storing instructions, characterized in that, When the instructions are executed on the model recognition device, the model recognition device performs the model recognition method as described in any one of claims 1-9 or 10-19.

26. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program is used to implement the model recognition method according to any one of claims 1-9 or 10-19.

Images