Functionality conversion method and apparatus, model conversion method and apparatus, and device, storage medium and program product

Abstract

Disclosed in the present invention are a functionality conversion method and apparatus, a model conversion method and apparatus, and a device, a storage medium and a program product. The functionality conversion method comprises at least one of the following: performing functionality conversion on a first functionality; performing model conversion on a first model; acquiring functionality conversion information; and acquiring model conversion information, wherein the first functionality comprises at least one first model. Therefore, a terminal can autonomously decide whether to perform AI / ML functionality and / or model conversion, or determine, on the basis of conversion information decided by a network and indicated to the terminal, whether to perform the AI / ML functionality and / or model conversion, thereby improving the timeliness of the AI / ML functionality and / or model conversion.

Description

Function and model conversion method, device, equipment, storage medium and program product

[0001] Cross-reference to Related Applications

[0002] The present disclosure is based on and claims priority from Chinese Patent Application No. 202410878664.1 filed on July 2, 2024, the content of which is hereby incorporated by reference in its entirety into the present disclosure. TECHNICAL FIELD

[0003] The present application relates to the technical field of wireless communication, and particularly relates to a function and model conversion method, device, equipment, storage medium and program product. BACKGROUND

[0004] With the development of communication technology, artificial intelligence (AI) / machine learning (ML) models have gradually become an indispensable part of network architecture. For example, AI / ML models are applied in scenarios such as CSI (Channel State Information) compression, CSI prediction, beam management, positioning, cell handover, etc. At present, the conversion of AI / ML models is very flexible. For example, if the configured parameters or the scenario change, the model no longer matches the environment, and the performance of the AI / ML model will deteriorate. Therefore, how to timely and quickly convert the AI / ML model has become a problem that the field focuses on. SUMMARY

[0005] To solve the problems in the prior art, the embodiments of the present application provide a function and model conversion method, device, equipment, storage medium and program product, which can improve the timeliness of AI / ML model and / or function conversion.

[0006] In a first aspect, the embodiments of the present application provide a function conversion method, comprising:

[0007] Converting the first function.

[0008] In some embodiments, the method further comprises:

[0009] Obtaining function conversion information.

[0010] In some embodiments, the method comprises at least one of the following:

[0011] Converting the first model; wherein the first function comprises at least one first model;

[0012] Obtaining model conversion information.

[0013] In some embodiments, the method comprises:

[0014] The first function is converted to a second function.

[0015] In some embodiments, the method comprises at least one of:

[0016] When the function conversion information is not acquired within a first time duration, the second function is autonomously determined;

[0017] When the function conversion information is acquired within the first time duration, the second function is determined according to the function conversion information.

[0018] In some embodiments, the method comprises at least one of:

[0019] When the function conversion information is not acquired within a first time duration, the target function is autonomously determined to perform function conversion on the first function;

[0020] When the function conversion information is acquired within the first time duration, the first function is converted according to the function conversion information.

[0021] In some embodiments, the starting point of the first time duration comprises one of:

[0022] Sending monitoring information;

[0023] Sending measurement information;

[0024] Sending a function conversion request;

[0025] Sending a function activation request;

[0026] Sending a function deactivation request;

[0027] Sending a function fallback request.

[0028] In some embodiments, the function conversion information comprises at least one of the following information: indication information of performing function conversion, function related information, function group related information.

[0029] In some embodiments, the method comprises:

[0030] The first model is converted to a second model.

[0031] In some embodiments, the method comprises at least one of:

[0032] When the model conversion information is not acquired within a first time duration, the second model is autonomously determined;

[0033] When the model conversion information is acquired within the first time length, the second model is determined according to the model conversion information.

[0034] In some embodiments, the method comprises at least one of:

[0035] When the model conversion information is not acquired within the first time length, the target model is autonomously determined to perform model conversion on the first model.

[0036] When the model conversion information is acquired within the first time length, the first model is converted according to the model conversion information.

[0037] In some embodiments, the starting point of the first time length comprises one of:

[0038] The monitoring information is sent.

[0039] The measurement information is sent.

[0040] The model conversion request is sent.

[0041] The function conversion request is sent.

[0042] The model activation request is sent.

[0043] The model deactivation request is sent.

[0044] The function activation request is sent.

[0045] The function deactivation request is sent.

[0046] The model rollback request is sent.

[0047] The function rollback request is sent.

[0048] In some embodiments, the model conversion information comprises at least one of the following information: indication information of performing model conversion, model related information, model group related information.

[0049] In some embodiments, the method further comprises:

[0050] The first information is acquired; wherein the first information comprises at least one of the following information: function related information, function group related information, and correspondence information between functions and function groups.

[0051] In some embodiments, the first information comprises function related information and / or function group related information that allows the terminal to autonomously perform conversion.

[0052] In some embodiments, the first information further comprises at least one of the following information: model related information, model group related information, correspondence information between models and model groups, and correspondence information between models and functions.

[0053] In some embodiments, the first information comprises model-related information and / or model group-related information that allows the terminal to autonomously perform conversion.

[0054] In some embodiments, the method further comprises:

[0055] sending second information; wherein the second information comprises at least one of the following information: function-related information, function group-related information, and correspondence information between functions and function groups.

[0056] In some embodiments, the second information further comprises at least one of the following information: model-related information, model group-related information, correspondence information between models and model groups, and correspondence information between models and functions.

[0057] In some embodiments, the method further comprises:

[0058] acquiring third information; wherein the third information comprises at least one of the following information: model-related information, model group-related information, correspondence information between models and model groups, and correspondence information between models and functions.

[0059] In some embodiments, the third information comprises model-related information and / or model group-related information that allows the terminal to autonomously perform conversion.

[0060] In some embodiments, the method further comprises:

[0061] sending fourth information; wherein the fourth information further comprises at least one of the following information: model-related information, model group-related information, correspondence information between models and model groups, and correspondence information between models and functions.

[0062] Compared with the prior art, the functional conversion method of the embodiments of the present application has the following advantages:

[0063] The terminal can autonomously perform function conversion on the first function; or acquire function conversion information sent by the network side, and perform function conversion on the first function based on the acquired function conversion information; wherein the first function includes at least one first model, and the first model includes an AI / ML model, so that the terminal can autonomously decide whether to perform AI / ML function and / or model conversion, or determine whether to perform AI / ML function and / or model conversion according to the conversion information decided by the network and indicated to the terminal, thereby improving the timeliness of AI / ML function and / or model conversion. By introducing the first time length in the process of AI / ML function and / or model conversion, the autonomous conversion of AI / ML function and / or model or the conversion of AI / ML function and / or model according to the acquired function conversion information can be performed under the condition that a preset condition is met, thereby further improving the timeliness of AI / ML function and / or model conversion, and reducing unnecessary and frequent AI / ML function and / or model conversion.

[0064] In a second aspect, embodiments of the present application provide a model conversion method, comprising:

[0065] Converting the first model.

[0066] In some embodiments, the method further comprises:

[0067] Acquiring model conversion information.

[0068] In some embodiments, the method comprises: converting the first model to a second model.

[0069] In some embodiments, the method comprises at least one of:

[0070] When the model conversion information is not acquired within the first time length, autonomously determining a second model;

[0071] When the model conversion information is acquired within the first time length, determining a second model according to the model conversion information.

[0072] In some embodiments, the method comprises at least one of:

[0073] When the model conversion information is not acquired within the first time length, autonomously determining that a target model performs model conversion on the first model;

[0074] When the model conversion information is acquired within the first time length, performing model conversion on the first model according to the model conversion information.

[0075] In some embodiments, the start point of the first time length comprises one of:

[0076] Sending monitoring information;

[0077] sending measurement information;

[0078] sending model conversion request;

[0079] sending model activation request;

[0080] sending model deactivation request;

[0081] sending model fallback request.

[0082] In some embodiments, the model conversion information comprises at least one of the following information: indication information of performing model conversion, model related information, model group related information.

[0083] In some embodiments, the method further comprises:

[0084] obtaining first information; wherein the first information comprises at least one of the following information: model related information, model group related information, and correspondence information between model and model group.

[0085] In some embodiments, the first information comprises model related information and / or model group related information that allows the terminal to autonomously perform conversion.

[0086] In some embodiments, the method further comprises:

[0087] sending second information; wherein the second information comprises at least one of the following information: model related information, model group related information, and correspondence information between model and model group.

[0088] Compared with the prior art, the model conversion method of the embodiments of the present application has the following beneficial effects:

[0089] The terminal can autonomously perform model conversion on the first model, or obtain model conversion information sent by the network side and perform model conversion on the first model based on the obtained model conversion information; wherein the first model comprises at least one first model, and the first model comprises an AI / ML model, so that the terminal can autonomously decide whether to perform AI / ML model conversion, or determine whether to perform AI / ML model conversion according to the conversion information decided by the network and indicated to the terminal, thereby improving the timeliness of AI / ML model conversion. At the same time, by introducing the first time length in the process of AI / ML model conversion, the autonomous conversion of the AI / ML model or the conversion of the AI / ML model according to the obtained model conversion information is performed under the condition of meeting the preset condition, which can further improve the timeliness of AI / ML model conversion and reduce unnecessary and frequent AI / ML model conversion.

[0090] In a third aspect, the embodiments of the present application provide a function conversion device, comprising:

[0091] The function conversion module is configured to perform function conversion on the first function.

[0092] In a fourth aspect, an embodiment of the present application provides a model conversion device, comprising:

[0093] The model conversion module is configured to perform model conversion on the first model.

[0094] In a fifth aspect, an embodiment of the present application provides a function conversion apparatus, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the function conversion method according to any one of the first aspect when executing the computer program.

[0095] In a sixth aspect, an embodiment of the present application provides a model conversion apparatus, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the model conversion method according to any one of the second aspect when executing the computer program.

[0096] In a seventh aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, wherein the computer program controls a device where the computer readable storage medium is located to execute the function conversion method according to any one of the first aspect or the model conversion method according to any one of the second aspect when the computer program runs.

[0097] In an eighth aspect, an embodiment of the present application provides a computer program product, comprising computer program / instructions, which implement the function conversion method according to any one of the first aspect or the model conversion method according to any one of the second aspect when executed by a processor.

[0098] It should be noted that the beneficial effects of the function / model conversion device, apparatus, computer readable storage medium and computer program product according to the embodiments of the present application can refer to the beneficial effects of the function / model conversion method described above, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0099] In order to more clearly illustrate the technical solutions of the present application, the drawings used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative effort based on these drawings.

[0100] FIG. 1 is a flowchart of a function conversion method according to an embodiment of the present application;

[0101] Fig. 2 is a schematic diagram of a function and / or model conversion interaction process according to an embodiment of the present application;

[0102] Fig. 3 is a flowchart of a model conversion method according to an embodiment of the present application;

[0103] Fig. 4 is a structural block diagram of a function conversion device according to an embodiment of the present application;

[0104] Fig. 5 is a structural block diagram of a model conversion device according to an embodiment of the present application;

[0105] Fig. 6 is a structural block diagram of a function conversion apparatus according to an embodiment of the present application;

[0106] Fig. 7 is a structural block diagram of a model conversion apparatus according to an embodiment of the present application. DETAILED DESCRIPTION

[0107] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0108] It should be noted that in the embodiments of the present application, the relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between the entities or operations. The terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitation, the elements defined by the statement "include" do not exclude the presence of other identical elements in the process, method, article or device including the elements. The term "a plurality of or several" refers to two or more, and the same applies to "a plurality of or several". The term "and / or" describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the three cases of A alone, A and B together, and B alone. The character " / " generally represents an "or" relationship between the front and rear associated objects.

[0109] Some terms and concepts related to the embodiments of the present application are explained below.

[0110] A terminal is a device having a wireless transceiver function in a movable or fixed position. For example, the terminal can be a cellular phone, a cordless phone, a session initiation protocol (SIP) phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), a handheld device having a wireless communication function, a smartphone, a tablet, a computer with a wireless transceiver function, a drone, a virtual reality (VR) terminal, an augmented reality (AR) terminal, and the like wearable device, a wireless terminal in industrial control, a wireless terminal deployed on a ship, an airplane, a satellite, a vehicle, a vehicle-mounted wireless terminal deployed on an unmanned vehicle, a wireless terminal in remote medical treatment, a smart grid, transportation safety, a smart city, and / or a smart home, and a terminal in a future 5th generation (5G) network or a future evolved public land mobile network (PLMN), etc., which are not limited in the embodiments of the present application. The terminal can also be referred to as user equipment (UE) at times.

[0111] The network side can be a network device (e.g., a RAN device), a network side server (e.g., an operator controllable server). The RAN (radio access network) device is a node or device for accessing a terminal to a wireless network, and the RAN device can also be referred to as a network device or a base station. For example, the RAN device can be a base station, a gNB (generation nodeB) for a 5G base station, an evolved node B (eNB), a radio network controller (RNC), a node B (NB), a base station controller (BSC), a base transceiver station (BTS), a home base station (e.g., a home evolved nodeB or a home node B, HNB), a base band unit (BBU), a transmitting and receiving point (TRP), a transmitting point (TP), a mobile switching center, etc., which are not limited in the embodiments of the present application.

[0112] Model: can also be described as AI / ML model, artificial intelligence and / or machine learning model.

[0113] Functionality: can also be described as AI / ML functionality, artificial intelligence and / or machine learning functionality.

[0114] The above function and / or model can be deployed in one of the terminal and the network side, or simultaneously deployed in the terminal and the network side. The function is applied to or corresponds to a certain scene, such as beam management (or described as beam prediction, including time domain prediction, space domain prediction), CSI (Channel State Information, Channel State Information) compression, CSI prediction, positioning (including direct positioning, auxiliary positioning) and the like. The model is applied to or corresponds to a certain configuration, such as different antenna configurations (such as the number of antenna ports), different beam numbers, and different flow numbers correspond to different models. Among them, one / many functions can include at least one / many models.

[0115] Taking the terminal deployed with the above AI / ML function and / or model as an example, due to the change of the configured parameters or the scene, the performance of the original function and / or model becomes poor, and it may no longer be applicable, and the function and / or model needs to be converted in time and quickly. Based on this, the present embodiment of the application proposes an AI / ML function and / or model conversion method, which will be described in detail below in combination with the accompanying drawings.

[0116] It should be noted that the term "conversion" in the present embodiment of the application can also be described as "switching", including at least one of activating the model and / or function, deactivating the model and / or function, and falling back (for example, falling back from the AI / ML model and / or function to the non-AI / ML model and / or function). The terms "conversion" and "switching" can be used interchangeably, and are not specifically limited in the present embodiment of the application.

[0117] Please refer to FIG. 1, which is a flow chart of a function conversion method provided by the present embodiment. The function conversion method is applied to a terminal and includes at least one of the following:

[0118] S11: performing function conversion on a first function;

[0119] S12: obtaining function conversion information.

[0120] For example, the terminal can autonomously decide to perform function conversion on the current first function, or obtain function conversion information from the network side and perform function conversion on the current first function according to the obtained function conversion information.

[0121] Further, the method comprises at least one of the following:

[0122] model conversion is performed on the first model; wherein the first function comprises at least one of the first model;

[0123] obtaining model conversion information.

[0124] Exemplarily, the terminal can autonomously decide to perform model conversion on the current first model, or obtain model conversion information from the network side, and perform function conversion on the current first model according to the obtained model conversion information.

[0125] The function conversion information comprises at least one of the following information: indication information of performing function conversion, function related information, function group related information.

[0126] Further, for the case that the first function comprises at least one first model, the function conversion information can further comprise at least one of the following information: model related information, model group related information.

[0127] The model conversion information comprises at least one of the following information: indication information of performing model conversion, model related information, model group related information.

[0128] The indication information of performing function conversion is used to instruct the terminal to perform function conversion. The indication information of performing model conversion is used to instruct the terminal to perform model conversion. The model related information can also be described as related information of a target model, and the target model is also described as a target model of conversion. The function related information can also be described as related information of a target function, and the target function is also described as a target function of conversion. It should be understood that the target function is an AI / ML function different from the first function, and the target model is an AI / ML model different from the first model.

[0129] The model related information at least comprises a model identifier or a model index indicating a model, which is used to determine the target model of conversion.

[0130] The function related information at least comprises a function identifier or a function index indicating a function, which is used to determine the target function of conversion.

[0131] The model group related information comprises at least one of the following: a model group identifier or a model group index, and an identifier or an index of a model within the model group (the model within the group can also be described as a model included in the model group), which are used to determine the target model of conversion and its corresponding model group.

[0132] The function group related information comprises at least one of the following: a function group identifier or a function group index, and an identifier or an index of a function within the function group (the function within the group can also be described as a function included in the function group), which are used to determine the target function of conversion and its corresponding function group.

[0133] For the case of function conversion based on network-side indication or decision, the terminal side can determine the corresponding target function according to the function identifier or function index in the obtained function conversion information to perform conversion of the first function, or determine the corresponding target model according to the model identifier or model index to perform conversion of the first model in the first function, or determine the corresponding target function according to the function group identifier or function group index / identifier or index of the function within the model group to perform cross-group conversion of the first function, or determine the corresponding target model according to the model group identifier or function group index / identifier or index of the function within the function group to perform cross-group conversion of the first model in the first function. The case of model conversion based on network-side indication or decision is the same, which is not repeated here.

[0134] The acquisition of the function conversion information and / or model conversion information includes the following ways:

[0135] (1) Obtained from the network-side device, such as from the base station, network management (Operation Administration and Maintenance, OAM), and the like.

[0136] (2) Obtained from the server, such as from the location management function (Location Management Function, LMF), AI / ML server, over the top (OTT, which refers to providing various application services to users through the Internet), and the like.

[0137] It should be noted that the term "function conversion information" in the embodiments of the present application can also be described as "function conversion signaling" or "function conversion command"; similarly, the term "model conversion information" can also be described as "model conversion signaling" or "model conversion command"; the terms "function conversion information", "function conversion signaling", and "function conversion command" can be used interchangeably, and the terms "model conversion information", "model conversion signaling", and "model conversion command" can be used interchangeably, which are not specifically limited in the embodiments of the present application.

[0138] In the embodiments of the present application, the terminal can autonomously perform function conversion of the first function; or obtain the function conversion information sent by the network side and perform function conversion of the first function based on the obtained function conversion information; so that the terminal can autonomously decide whether to perform AI / ML function conversion, or determine whether to perform AI / ML function conversion according to the function conversion information indicated by the network decision to the terminal, thereby improving the timeliness of AI / ML function conversion.

[0139] Specifically, the method comprises: converting the first function to the second function.

[0140] Further, the method comprises at least one of the following:

[0141] When the function conversion information is not acquired within the first time length, the second function is autonomously determined.

[0142] When the function conversion information is acquired within the first time length, the second function is determined according to the function conversion information.

[0143] Specifically, the method comprises that the first model is converted to a second model.

[0144] Further, the method comprises at least one of the following:

[0145] When the model conversion information is not acquired within the first time length, the second model is autonomously determined.

[0146] When the model conversion information is acquired within the first time length, the second model is determined according to the model conversion information.

[0147] It should be understood that the second function is an AI / ML function different from the first function, and the second model is an AI / ML model different from the first model.

[0148] Further, the method comprises at least one of the following:

[0149] When the model conversion information is not acquired within the first time length, a target model autonomously performs model conversion on the first model.

[0150] When the function conversion information is not acquired within the first time length, a target function autonomously performs function conversion on the first function.

[0151] When the model conversion information is acquired within the first time length, the first model is converted according to the model conversion information.

[0152] When the function conversion information is acquired within the first time length, the first function is converted according to the function conversion information.

[0153] In the embodiments of the present application, the first time length can be acquired in at least one of the following ways: acquired by a network side or a server, such as configuration of the first time length through related signaling; pre-defined, such as pre-defined in a protocol; determined by a terminal, and the terminal feeds back to the network side or the server.

[0154] The application scenario of the first time length includes at least one of the following: the terminal autonomously decides that the first model needs to be converted, the first model and the autonomously decided target model or second model are in different model groups, the autonomously decided target model or second model is not in the model list that allows the terminal to autonomously convert, the terminal autonomously decides that the first function needs to be converted, the first function and the autonomously decided target function or second function are in different function groups or functions, and the autonomously decided target function or second function is not in the function list that allows the terminal to autonomously convert.

[0155] It can be understood that one first time length can be applicable to all AI / ML models and / or functions, or different AI / ML models and / or functions have different lengths of the first time length, or the first time length has N length values, and different AI / ML models and / or functions are respectively associated with corresponding lengths of the first time length, or the first time length of the AI / ML model is different from the first time length of the AI / ML function, which is not specifically limited in the embodiments of the present application.

[0156] It should be noted that in the embodiments of the present application, the term "obtaining the function conversion information within the first time length" can also be described as "obtaining the function conversion information before the first time length expires" or "obtaining the function conversion information before the first time length expires". "No function conversion information is obtained within the first time length", "no function conversion information is obtained within the first time length", "no model conversion information is obtained within the first time length", "no model conversion information is obtained within the first time length", and the like are not repeated here.

[0157] For example, when the terminal does not receive the function and / or model conversion information issued by the network side within the first time length, the terminal can decide the target function and / or target model from the pre-configured function list and / or model list that allows autonomous conversion, and switch from the first function and / or first model to the target function and / or target model.

[0158] When the terminal receives the function and / or model conversion information issued by the network side within the first time length, there are the following cases:

[0159] When the terminal receives the function and / or model conversion information issued by the network side within the first time length, there are the following cases:

[0160] In the first time length, the terminal receives the indication information carrying the function and / or model conversion, the information of the target function and / or the model-related information of the target function and / or the model indicated by the network side, and the terminal switches from the first function and / or the first model to the target function and / or the target model indicated by the network side.

[0161] In the first time length, the terminal receives the indication information carrying the function and / or model conversion, the function group information and / or the model group-related information indicated by the network side, and the terminal switches from the first function and / or the first model to the target function and / or the target model in the function group and / or the model group indicated by the network side.

[0162] In the embodiment of the present application, by introducing the first time length in the function and / or model conversion process, when the terminal autonomously decides to perform AI / ML model and / or function conversion, the terminal needs to wait for the first time length. If the model and / or function conversion information is obtained within the first time length, the terminal follows the network side indication to perform the conversion; if the model and / or function conversion information is not received within the first time length, after the first time length, the terminal performs the model and / or function conversion according to the autonomous decision, as shown in FIG. 2. The embodiment of the present application introduces the first time length, which takes into account the timeliness of model and / or function conversion and the time margin of waiting for network indication, avoids the terminal waiting for network side indication indefinitely, and can also avoid missing the network side indication; on the one hand, it can avoid the problem of untimely model conversion caused by long waiting for network indication; on the other hand, it can avoid the problem of unnecessary and frequent model conversion caused by the terminal autonomous decision being different from / conflicting with the network decision, so as to clearly define the terminal behavior, reduce unnecessary and frequent model and / or function conversion, and improve the system performance.

[0163] The starting point of the first time length includes one of the following:

[0164] Send monitoring information;

[0165] Send measurement information;

[0166] Send model conversion request;

[0167] Send function conversion request;

[0168] Send model activation request;

[0169] Send model deactivation request;

[0170] Send function activation request;

[0171] Send function deactivation request;

[0172] Send model rollback request;

[0173] Send function rollback request.

[0174] In the embodiments of the present application, the start of the first time length can be that the terminal autonomously determines that model and / or function conversion is needed. The term "start of the first time length" can also be described as "start of starting the first timer".

[0175] The monitoring information can be understood as performance monitoring of AI / ML model / functionality inference (i.e., AI / ML model and / or function inference). The monitoring information includes at least one of the prediction probability of the AI / ML model and / or function (i.e., the probability of successful prediction of the AI / ML model and / or function), the throughput rate, and the location information.

[0176] The measurement information includes at least one of RSRP (Reference Signal Receiving Power), RSRQ (Reference Signal Received Quality), SINR (Signal to Interference plus Noise Ratio), RSTD (Reference Signal Time Difference), and the difference between the transmission and reception times. The RSRP, RSRQ, and SINR can be L1 (i.e., layer 1) or L3 (i.e., layer 3).

[0177] The terminal can send the monitoring information and / or the measurement information to the network side or the server, and the network side or the server can determine whether to perform model and / or function conversion. The terminal can also autonomously determine whether to perform model conversion based on the monitoring information and / or the measurement information, and send the result to the network side or the server to assist the network side or the server in subsequent configuration or scheduling.

[0178] Specifically, the method further includes:

[0179] Obtaining first information;

[0180] The first information includes at least one of the following information: model-related information, function-related information, model group-related information, function group-related information, correspondence information between the model and the model group, correspondence information between the function and the function group, and correspondence information between the model and the function.

[0181] Or the first information includes model-related information and / or model group-related information and / or function-related information and / or function group-related information that allow the terminal to make autonomous decisions (autonomous decisions include whether to make a conversion and / or a conversion target). The terminal making autonomous decisions can also be described as the terminal making autonomous conversion or the terminal autonomously making conversion, including the terminal itself determining whether to make a conversion, a conversion target, etc.

[0182] In embodiments of the application, the first information can be obtained from a network side device, such as from a base station, a network management device, etc., or from a server, such as an LMF, an AI / ML server, an OTT, etc. The term "corresponding" can be described as "mapping" or "association".

[0183] It should be noted that the corresponding information of the model and the model group includes a model identifier or a model index, a model group identifier or a model group index. It can be one-to-many or many-to-one, which is not specifically limited in embodiments of the application. For example, a certain model(s) corresponds to (or is described as belonging to) a model group, and a certain model group(s) also corresponds to (or is described as including) a model.

[0184] The corresponding information of the function and the function group includes a function identifier or a function index, a function group identifier or a function group index. It can be one-to-many or many-to-one, which is not specifically limited in embodiments of the application. For example, a certain function(s) corresponds to (or is described as belonging to) a function group, and a certain function group(s) also corresponds to (or is described as including) a function.

[0185] The corresponding information of the model and the function includes a model identifier or a model index, a function identifier or a function index. It can be one-to-many or many-to-one, which is not specifically limited in embodiments of the application. For example, a certain model(s) corresponds to (or is described as belonging to) a function, and a certain function(s) also corresponds to (or is described as including) a model.

[0186] Further, the method further includes:

[0187] Obtaining third information.

[0188] The third information includes at least one of the following information: model-related information, model group-related information, corresponding information of the model and the model group, and corresponding information of the model and the function.

[0189] Or the third information includes model-related information and / or model group-related information that allow the terminal to make autonomous conversion.

[0190] It should be understood that the role of the third information in the function and / or model conversion process can be referred to the first information, which is not repeated here.

[0191] The network side issues the first information and / or the third information, the terminal configures corresponding models and / or functions based on the first information and / or the third information, and the terminal is only allowed to perform autonomous conversion between models in a model and / or function list indicated by the first information and / or the third information. For models outside the model and / or function list indicated by the first information and / or the third information, the terminal can only perform model and / or function conversion according to an indication of the network side.

[0192] As an implementation form, the terminal is only allowed to perform autonomous conversion within the same model group. If the conversion is between models of different model groups, the terminal can only perform the conversion according to an indication of the network side. If there are model groups, the first information can include at least one model group, and each model allowed to perform autonomous conversion is associated with a corresponding model group and / or function group.

[0193] As an implementation form, the terminal is only allowed to perform autonomous conversion between different functions within the same function group. If the conversion is between functions of different function groups, the terminal can only perform the conversion according to an indication of the network side. If there are function groups, the first information can include at least one function group, and each function allowed to perform autonomous conversion is associated with a corresponding function group.

[0194] As an implementation form, the terminal can only perform autonomous conversion between different models associated with the same function. Conversion between models belonging to different functions can only be performed according to an indication of the network side.

[0195] In the embodiment of the application, the terminal is only allowed to perform autonomous conversion between models and / or functions included in the first information and / or the third information, and the terminal can only perform conversion according to an indication of the network side for models and / or functions outside the first information and / or the third information. Through the first information and / or the third information, the network side indicates which models and / or functions can be autonomously converted by the terminal, thereby explicitly indicating which models and / or which functions can be autonomously converted by the terminal, and which models and / or which functions cannot be autonomously converted by the terminal, avoiding the problem of indefinite waiting for network indication by the terminal, while allowing the terminal to autonomously perform model conversion within a certain range, and improving the degree of freedom of the terminal in implementing function and / or model conversion.

[0196] The first information and / or the third information can also be used in combination with the first time length. Specifically, when the model / function conversion does not occur in the list of models / functions indicated by the first information and / or the third information that allows autonomous conversion, or the model / function conversion occurs between different model groups / function groups, the terminal needs to wait for the first time length, and if the model and / or function conversion information is not obtained before the first time length expires, the terminal performs AI / ML model and / or function conversion according to the autonomous decision. If the model and / or function conversion information is received within the first time length, the terminal performs the relevant operation according to the indication.

[0197] It should be understood that the model and / or function list is the model-related information, function-related information, model group-related information, function group-related information, model-to-model group correspondence information, function-to-function group correspondence information, and model-to-function correspondence information indicated by the first information and / or the third information issued by the network side to the terminal.

[0198] Further, the method further comprises:

[0199] sending second information; wherein the second information comprises at least one of the following information: model-related information, function-related information, model group-related information, function group-related information, model-to-model group correspondence information, function-to-function group correspondence information, and model-to-function correspondence information.

[0200] In the embodiments of the present application, the second information comprises at least one of the model-related information, function-related information, model group-related information, function group-related information, model-to-model group correspondence information, function-to-function group correspondence information, and model-to-function correspondence information requested or desired by the terminal. Whether the network side or the server adopts the second information depends on the network implementation.

[0201] The second information can be sent to network side devices such as base stations, network management (OAM in English), etc., or servers such as LMF, AI / ML server, OTT, etc.

[0202] Exemplarily, the second information can be reported through UE auxiliary information. By reporting the second information, the terminal can help the network side to understand the situation of the terminal, and assist the network side to configure the first information and / or the third information. For example, the network side can send the first information and / or the third information to the terminal according to the second information reported by the terminal, in combination with network requirements, to instruct the terminal which models and / or which functions can be converted by the terminal autonomously, so that it is clear between which models and / or which functions can be converted by the terminal autonomously between the network side and the terminal, avoiding the problem of indefinite waiting for the network side indication by the terminal, while allowing the terminal to autonomously convert the models within a certain range, and improving the freedom of the terminal in model and / or function conversion.

[0203] Further, the method further comprises:

[0204] sending fourth information; wherein the fourth information further comprises at least one of the following information: model-related information, model group-related information, corresponding information of the model and the model group, corresponding information of the model and the function.

[0205] Similarly, the fourth information reported by the terminal can assist the network side to configure the third information, and the working process and role of the fourth information in the function and / or model conversion process can be referred to the second information, which will not be repeated here.

[0206] Please refer to FIG. 3, the embodiment of the present application further provides a model conversion method. The model conversion method is applied to a terminal, and comprises at least one of the following:

[0207] S21: performing model conversion on the first model;

[0208] S22: obtaining model conversion information.

[0209] In some embodiments, the method comprises: converting the first model to a second model.

[0210] In some embodiments, the method comprises at least one of the following:

[0211] when the model conversion information is not obtained within the first time length, autonomously determining a second model;

[0212] when the model conversion information is obtained within the first time length, determining a second model according to the model conversion information.

[0213] In some embodiments, the method comprises at least one of the following:

[0214] when the model conversion information is not obtained within the first time length, autonomously determining a target model to perform model conversion on the first model;

[0215] The model conversion information is acquired in a first time period, and the first model is converted according to the model conversion information.

[0216] In some embodiments, the start point of the first time period comprises one of the following:

[0217] sending monitoring information;

[0218] sending measurement information;

[0219] sending a model conversion request;

[0220] sending a model activation request;

[0221] sending a model deactivation request;

[0222] sending a model fallback request.

[0223] In some embodiments, the model conversion information comprises at least one of the following: indication information of model conversion, model-related information, model group-related information.

[0224] In some embodiments, the method further comprises:

[0225] acquiring first information; wherein the first information comprises at least one of the following: model-related information, model group-related information, and correspondence information between a model and a model group.

[0226] In an optional embodiment, the first information comprises model-related information and / or model group-related information that allows the terminal to autonomously perform conversion.

[0227] In an optional embodiment, the method further comprises:

[0228] sending second information; wherein the second information comprises at least one of the following: model-related information, model group-related information, and correspondence information between a model and a model group.

[0229] It should be noted that the working process of the model conversion method described in the embodiments of the present application can refer to the working process of the function conversion method described above, and the technical effects achieved by the model conversion method are the same as those of the function conversion method described above, which will not be described here again.

[0230] Referring to FIG. 4, the present application also provides a function conversion device. The function conversion device is applied to a terminal and comprises:

[0231] a function conversion module 11, configured to perform function conversion on a first function;

[0232] a function conversion information acquisition module 12, configured to acquire function conversion information.

[0233] In an alternative embodiment, the apparatus further comprises:

[0234] a model conversion module configured to perform model conversion on the first model; wherein the first function comprises at least one of the first model;

[0235] a model conversion information acquisition module configured to acquire model conversion information.

[0236] In an alternative embodiment, the first model is converted to a second model and / or the first function is converted to a second function.

[0237] In an alternative embodiment, the apparatus comprises one of the following:

[0238] a first conversion module configured to autonomously determine the second function when the function conversion information is not acquired within a first time duration;

[0239] a second conversion module configured to determine the second function according to the function conversion information when the function conversion information is acquired within the first time duration.

[0240] In an alternative embodiment, the apparatus comprises one of the following:

[0241] a third conversion module configured to autonomously determine the second model when the model conversion information is not acquired within the first time duration;

[0242] a fourth conversion module configured to determine the second model according to the model conversion information when the model conversion information is acquired within the first time duration.

[0243] In an alternative embodiment, the apparatus comprises one of the following:

[0244] a fifth conversion module configured to autonomously determine a target model to perform model conversion on the first model when the model conversion information is not acquired within the first time duration;

[0245] a sixth conversion module configured to perform model conversion on the first model according to the model conversion information when the model conversion information is acquired within the first time duration;

[0246] a seventh conversion module configured to autonomously determine a target function to perform function conversion on the first function when the function conversion information is not acquired within the first time duration;

[0247] an eighth conversion module configured to perform function conversion on the first function according to the function conversion information when the function conversion information is acquired within the first time duration.

[0248] In an alternative embodiment, the start point of the first time duration comprises one of the following:

[0249] sending monitoring information;

[0250] sending measurement information;

[0251] sending model conversion request;

[0252] sending function conversion request;

[0253] sending model activation request;

[0254] sending model deactivation request;

[0255] sending function activation request;

[0256] sending function deactivation request;

[0257] sending model fallback request;

[0258] sending function fallback request.

[0259] In an optional embodiment, the function conversion information comprises at least one of the following: indication information of performing function conversion, function related information, function group related information.

[0260] In an optional embodiment, the model conversion information comprises at least one of the following: indication information of performing model conversion, model related information, model group related information.

[0261] In an optional embodiment, the apparatus comprises:

[0262] an information obtaining module, configured to obtain first information; wherein the first information comprises at least one of the following: function related information, function group related information, correspondence information between function and function group.

[0263] In an optional embodiment, the first information comprises function related information and / or function group related information which allows the terminal to autonomously perform conversion.

[0264] In an optional embodiment, the first information further comprises at least one of the following: model related information, model group related information, correspondence information between model and model group, correspondence information between model and function.

[0265] In an optional embodiment, the first information comprises model related information and / or model group related information which allows the terminal to autonomously perform conversion.

[0266] In an optional embodiment, the apparatus comprises:

[0267] The information sending module is configured to send second information, wherein the second information comprises at least one of the following: function-related information, function group-related information, and correspondence information between a function and a function group.

[0268] In an alternative embodiment, the second information further comprises at least one of the following: model-related information, model group-related information, correspondence information between a model and a model group, and correspondence information between a model and a function.

[0269] It should be noted that the working processes of the various modules in the function conversion device according to the embodiments of the present application can refer to the working processes of the function conversion method described above, and the technical effects achieved are the same as those of the function conversion method described above, which will not be described herein again.

[0270] Referring to FIG. 5, the present application also provides a model conversion device. The model conversion device is applied to a terminal and comprises:

[0271] The model conversion module 21 is configured to perform model conversion on the first model.

[0272] The model conversion information acquisition module 22 is configured to acquire model conversion information.

[0273] In an alternative embodiment, the first model is converted to a second model.

[0274] In an alternative embodiment, the device comprises one of the following:

[0275] The first conversion module is configured to autonomously determine a second model when the model conversion information is not acquired within a first time length.

[0276] The second conversion module is configured to determine a second model according to the model conversion information when the model conversion information is acquired within a first time length.

[0277] In an alternative embodiment, the device comprises one of the following:

[0278] The third conversion module is configured to autonomously determine a target model to perform model conversion on the first model when the model conversion information is not acquired within a first time length.

[0279] The fourth conversion module is configured to perform model conversion on the first model according to the model conversion information when the model conversion information is acquired within a first time length.

[0280] In an alternative embodiment, the starting point of the first time length comprises one of the following:

[0281] Sending monitoring information;

[0282] Sending measurement information;

[0283] sending a model conversion request;

[0284] sending a model activation request;

[0285] sending a model deactivation request;

[0286] sending a model fallback request.

[0287] In an optional embodiment, the model conversion information comprises at least one of the following: indication information of performing model conversion, model-related information, model group-related information.

[0288] In an optional embodiment, the apparatus comprises:

[0289] an information obtaining module, configured to obtain first information; wherein the first information comprises at least one of the following: model-related information, model group-related information, and correspondence information between a model and a model group.

[0290] In an optional embodiment, the first information comprises model-related information and / or model group-related information that allows the terminal to autonomously perform conversion.

[0291] In an optional embodiment, the apparatus comprises:

[0292] an information sending module, configured to send second information; wherein the second information comprises at least one of the following: model-related information, model group-related information, and correspondence information between a model and a model group.

[0293] It should be noted that the working processes of the various modules in the model conversion apparatus according to the embodiments of the present application can refer to the working processes of the model conversion method described above, and the technical effects achieved are the same as those of the model conversion method described above, which will not be described here again.

[0294] Referring to FIG. 6, the embodiments of the present application further provide a function conversion device. The function conversion device comprises a first processor 31, a first memory 32, and a computer program stored in the first memory 32 and capable of running on the first processor 31. The first processor 31 implements the steps in the function conversion method, such as steps S11-S12, when executing the computer program.

[0295] For example, the computer program can be divided into one or more modules / units, which are stored in the first memory 32 and executed by the first processor 31 to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the function conversion device.

[0296] The functional conversion device can include, but is not limited to, a first processor 31 and a first memory 32. Those skilled in the art can understand that the schematic diagram is only an example of the functional conversion device, and does not constitute a limitation on the functional conversion device, and can include more or fewer components than the diagram, or combine certain components, or different components, for example, the functional conversion device can also include an input / output device, a network access device, a bus, etc.

[0297] The first processor 31 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The first processor 31 is the control center of the functional conversion device, and connects various parts of the entire functional conversion device through various interfaces and lines.

[0298] The first memory 32 can be used to store computer programs and / or modules, and the first processor 31 realizes various functions of the functional conversion device by running or executing computer programs and / or modules stored in the first memory 32, and calling data stored in the first memory 32. The first memory 32 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function (such as a sound playing function, an image playing function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phonebook, etc.), etc. In addition, the first memory 32 can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0299] The modules / units integrated in the function conversion device can be stored in a computer readable storage medium if they are realized in the form of software function units and sold or used as independent products. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can realize the steps of each method embodiment when executed by a first processor 31. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0300] Referring to FIG. 7, the model conversion device according to an embodiment of the present application is provided. The model conversion device includes a second processor 41, a second memory 42, and a computer program stored in the second memory 42 and executable on the second processor 41. The second processor 41 executes the computer program to implement the steps in the above-mentioned model conversion method, such as steps S21-S22.

[0301] It should be noted that the structure and working process of each module in the model conversion device according to the embodiment of the present application can refer to the structure and working process of each module in the function conversion device described above, which will not be repeated here.

[0302] It should be noted that the apparatus embodiments described above are only schematic and that each unit as a separate component can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed on a plurality of network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment. In addition, the connection relationship between the modules in the apparatus embodiments provided by the present application indicates that there is a communication connection between them, which can be realized as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0303] The above is the preferred embodiment of the present application, it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, can also make a plurality of improvements and refinements, these improvements and refinements are also considered to be within the scope of the present application.

Claims

1. A method for function conversion, comprising: functionally converting a first function.

2. The method of claim 1, wherein, The method further comprises: obtaining function conversion information.

3. The method of claim 1 or 2, wherein, The method comprises: model converting a first model; wherein the first function comprises at least one first model.

4. The method of claim 3, wherein, The method further comprises: obtaining model conversion information.

5. The method of claim 1 or 2, wherein, The method comprises: functionally converting the first function to a second function.

6. The method of claim 5, wherein, The method comprises at least one of: autonomously determining the second function when no function conversion information is obtained within a first time duration; determining the second function according to the function conversion information when the function conversion information is obtained within the first time duration.

7. The method of claim 2, wherein, The method comprises at least one of: autonomously determining a target function to functionally convert the first function when no function conversion information is obtained within a first time duration; functionally converting the first function according to the function conversion information when the function conversion information is obtained within the first time duration.

8. The method of claim 6 or 7, wherein, The start point of the first time duration comprises at least one of: sending monitoring information; sending measurement information; sending a function conversion request; sending a function activation request; sending a function deactivation request; sending a function rollback request.

9. The method of claims 1 to 7, wherein, The function conversion information comprises at least one of: indication information of performing function conversion, function related information, function group related information.

10. The method of claim 3 or 4, wherein, The method comprises: model converting the first model to a second model.

11. The method of claim 10, wherein, The method comprises at least one of: autonomously determining the second model when no model conversion information is obtained within a first time duration; determining the second model according to the model conversion information when the model conversion information is obtained within the first time duration.

12. The method of claim 3, wherein, The method comprises at least one of: autonomously determining a target model to model convert the first model when no model conversion information is obtained within a first time duration; model converting the first model according to the model conversion information when the model conversion information is obtained within the first time duration.

13. The method of claim 10 or 11, wherein, The start point of the first time duration comprises at least one of: sending monitoring information; sending measurement information; sending a model conversion request; sending a function conversion request; sending a model activation request; sending a model deactivation request; sending a function activation request; sending a function deactivation request; sending a model rollback request; sending a function rollback request.

14. The method of any one of claims 3 or 10-12, wherein, The model conversion information comprises at least one of: indication information of performing model conversion, model related information, model group related information.

15. The method of claim 1 or 2, wherein, The method further comprises: obtaining first information; wherein the first information comprises at least one of: function related information, function group related information, correspondence information between functions and function groups.

16. The method of claim 15, wherein, The first information comprises function related information and / or function group related information that allows the terminal to autonomously perform conversion.

17. The method of claim 15, wherein, The first information further comprises at least one of: model related information, model group related information, correspondence information between models and model groups, correspondence information between models and functions.

18. The method of claim 17, wherein, The first information comprises model related information and / or model group related information that allows the terminal to autonomously perform conversion.

19. The method of claim 1 or 2, wherein, The method further comprises: sending second information; wherein the second information comprises at least one of the following: function related information, function group related information, and correspondence between function and function group.

20. The method of claim 19, wherein, The second information further comprises at least one of the following: model related information, model group related information, correspondence between model and model group, and correspondence between model and function.

21. The method of claim 3, wherein, The method further comprises: acquiring third information; wherein the third information comprises at least one of the following: model related information, model group related information, correspondence between model and model group, and correspondence between model and function.

22. The method of claim 21, wherein, The third information comprises model related information and / or model group related information that allows the terminal to autonomously perform conversion.

23. The method of claim 3, wherein, The method further comprises: sending fourth information; wherein the fourth information further comprises at least one of the following: model related information, model group related information, correspondence between model and model group, and correspondence between model and function.

24. A model conversion method, comprising: performing model conversion on a first model.

25. The method of claim 24, wherein, The method comprises: acquiring model conversion information.

26. The method of claim 24 or 25, wherein, The method comprises conversion of the first model to a second model.

27. The method of claim 26, wherein, The method comprises at least one of the following: autonomously determining a second model when no model conversion information is acquired within a first time duration; determining a second model according to the model conversion information when the model conversion information is acquired within a first time duration.

28. The method of claim 25, wherein, The method comprises at least one of the following: autonomously determining a target model to perform model conversion on the first model when no model conversion information is acquired within a first time duration; performing model conversion on the first model according to the model conversion information when the model conversion information is acquired within a first time duration.

29. The method of claim 27 or 28, wherein, The starting point of the first time duration comprises at least one of the following: sending monitoring information; sending measurement information; sending a model conversion request; sending a model activation request; sending a model deactivation request; sending a model rollback request.

30. The method of any one of claims 24 to 28, wherein, The model conversion information comprises at least one of the following: indication information of performing model conversion, model related information, and model group related information.

31. The method of claim 24, wherein, The method further comprises: acquiring first information; wherein the first information comprises at least one of the following: model related information, model group related information, and correspondence between model and model group.

32. The method of claim 31, wherein, The first information comprises model related information and / or model group related information that allows the terminal to autonomously perform conversion.

33. The method of claim 24, wherein, The method further comprises: sending second information; wherein the second information comprises at least one of the following: model related information, model group related information, and correspondence between model and model group.

34. A function conversion apparatus, comprising: a function conversion module configured to perform function conversion on a first function.

35. A model conversion apparatus, comprising: a model conversion module configured to perform model conversion on a first model.

36. A function conversion device comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the function conversion method according to any one of claims 1 to 23 when executing the computer program.

37. A model conversion device comprising: A processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the model conversion method according to any one of claims 24 to 33 when executing the computer program.

38. A computer-readable storage medium storing a computer program, wherein, The computer readable storage medium is controlled to perform the function conversion method according to any one of claims 1 to 23, or the model conversion method according to any one of claims 22 to 30, when the computer program is running.

39. A computer program product comprising computer programs / instructions, characterized in that, The computer program / instruction implements the function conversion method according to any one of claims 1 to 23, or the model conversion method according to any one of claims 24 to 33, when executed by the processor.

Images