Communication method, device, system, storage medium, and program product

Abstract

The present disclosure relates to a communication method, a device, a system, a storage medium, and a program product. The communication method comprises: a terminal device receives first information, the first information comprising a first identifier and a second data set, the first identifier being associated with a first model, the first model being a model deployed on both the terminal device and a network device, the first model being obtained by performing training on the basis of the first data set, and the second data set being different from the first data set; and the terminal device determines a second model on the basis of the first model and the second data set. The present disclosure can maintain a better dual-end model, and improve the efficiency and accuracy of model-related services.

Description

Communication methods, devices, systems, storage media and software products Technical Field

[0001] This disclosure relates to the field of communication technology, and in particular to communication methods, devices, systems, storage media, and program products. Background Technology

[0002] In recent years, artificial intelligence (AI) and machine learning (ML) technologies have made continuous breakthroughs in many fields. Summary of the Invention

[0003] This disclosure provides communication methods, devices, systems, storage media, and program products.

[0004] According to a first aspect of the present disclosure, a communication method is proposed, the method comprising: a terminal device receiving first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both the terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; the terminal device determining a second model based on the first model and the second dataset.

[0005] According to a second aspect of the present disclosure, a communication method is proposed, the method comprising: a network device sending first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both a terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; wherein the first model and the second dataset are used to determine the second model.

[0006] According to a third aspect of the present disclosure, a terminal device is provided, comprising: a transceiver module for receiving first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both the terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; and a processing module for determining a second model based on the first model and the second dataset.

[0007] According to a fourth aspect of the present disclosure, a network device is provided, comprising: a transceiver module for transmitting first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both a terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; wherein the first model and the second dataset are used to determine a second model.

[0008] According to a fifth aspect of the present disclosure, a terminal device is provided, comprising: one or more processors; wherein the terminal device is configured to execute the first aspect and any one of the communication methods in the first aspect.

[0009] According to a sixth aspect of the present disclosure, a network device is provided, comprising: one or more processors; wherein the network device is configured to perform the second aspect and any one of the communication methods in the second aspect.

[0010] According to a seventh aspect of the present disclosure, a communication system is provided, including a terminal device and a network device, wherein the terminal device is configured to implement the first aspect and any one of the communication methods in the first aspect, and the network device is configured to implement the second aspect and any one of the communication methods in the second aspect.

[0011] According to an eighth aspect of the present disclosure, a storage medium is provided that stores instructions which, when executed on a communication device, cause the communication device to perform a communication method as described in the first aspect and any one thereof, or the second aspect and any one thereof.

[0012] According to a ninth aspect of the present disclosure, a program product is provided, comprising: a computer program, which, when executed by a communication device, causes the communication device to perform a communication method as described in the first aspect and any one of the first aspects or the second aspect and the second aspect.

[0013] This disclosure involves receiving first information via a terminal device. The first information includes a first identifier associated with a first model. This first model is deployed on both the terminal device and network devices; it can also be called a bilateral model. The terminal device determines the first model from the models deployed on its side based on the first identifier. The second dataset included in the first information is different from the first dataset used to train the first model. The terminal device can determine the second model based on the second dataset and the first dataset. This means that a new bilateral model (i.e., the second model) can be obtained by operating on an existing bilateral model using a new dataset (i.e., the second dataset), thereby improving model performance, maintaining a better bilateral model, and improving the efficiency and accuracy of model-related services. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings required for the description of the embodiments are introduced below. The following drawings are only some embodiments of this disclosure and do not impose specific limitations on the protection scope of this disclosure.

[0015] Figure 1 is a schematic diagram of a communication system architecture according to an embodiment of the present disclosure.

[0016] Figure 2 is a schematic diagram of a communication method interaction according to an embodiment of the present disclosure.

[0017] Figure 3 is a flowchart illustrating a communication method according to an embodiment of the present disclosure.

[0018] Figure 4 is a flowchart illustrating a communication method according to an embodiment of the present disclosure.

[0019] Figure 5a is an interactive schematic diagram of a communication method according to an embodiment of the present disclosure.

[0020] Figure 5b is a schematic diagram illustrating the interaction of the model recognition method according to an embodiment of the present disclosure.

[0021] Figure 6a is a schematic diagram of the structure of the terminal device proposed in an embodiment of this disclosure.

[0022] Figure 6b is a schematic diagram of the structure of the network device proposed in an embodiment of this disclosure.

[0023] Figure 7a is a schematic diagram of the structure of a communication device proposed in an embodiment of this disclosure.

[0024] Figure 7b is a schematic diagram of the chip structure proposed in an embodiment of this disclosure. Detailed Implementation

[0025] This disclosure provides communication methods, devices, systems, storage media, and program products.

[0026] In a first aspect, embodiments of this disclosure propose a communication method, the method comprising: a terminal device receiving first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both the terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; the terminal device determining a second model based on the first model and the second dataset.

[0027] In some alternative embodiments of the first aspect, the terminal device determines a second model based on the first model and the second dataset, including: the terminal device fine-tunes the first model using the second dataset to obtain the second model.

[0028] In some alternative embodiments of the first aspect, the terminal device determines a second model based on the first model and the second dataset, including: the terminal device retrains the first model using the second dataset to obtain the second model.

[0029] In some alternative embodiments of the first aspect, the terminal device retrains the first model using the second dataset to obtain the second model, including: the terminal device retrains the first model using only the second dataset to obtain the second model.

[0030] In some alternative embodiments of the first aspect, the terminal device retrains the first model using the second dataset to obtain the second model, including: the terminal device retrains the first model using the second dataset and the first dataset to obtain the second model.

[0031] In some alternative embodiments of the first aspect, the second model is associated with the first identifier.

[0032] In some alternative embodiments of the first aspect, the method further includes: the terminal device detaching the association between the first identifier and the first model.

[0033] In some alternative embodiments of the first aspect, the first information further includes a second identifier, wherein the second model is associated with the second model.

[0034] In some alternative embodiments of the first aspect, the first information further includes first indication information, which instructs the terminal device to fine-tune the first model using the second dataset.

[0035] In some alternative embodiments of the first aspect, the first information further includes first indication information, which instructs the terminal device to retrain the first model using the second dataset.

[0036] In some alternative embodiments of the first aspect, the first information further includes second indication information, which is used to instruct the terminal device to retrain the first model using only the second dataset to obtain the second model.

[0037] In some alternative embodiments of the first aspect, the first information further includes second instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset and the first dataset to obtain the second model.

[0038] In some alternative embodiments of the first aspect, the first information further includes third indication information, the third indication information being used to indicate one of the following: fine-tuning the first model using the second dataset to obtain the second model; retraining the first model using only the second dataset to obtain the second model; or retraining the first model using both the second dataset and the first dataset to obtain the second model.

[0039] In some alternative embodiments of the first aspect, the first information further includes fourth indication information, the fourth indication information being used to indicate one of the following: fine-tuning the first model using the second dataset to obtain the second model; or retraining the first model using the second dataset and the first dataset to obtain the second model.

[0040] In some alternative embodiments of the first aspect, the first information further includes fifth indication information, the fifth indication information being used to indicate one of the following: fine-tuning the first model using the second dataset to obtain the second model, and the second model being associated with the first identifier; fine-tuning the first model using the second dataset to obtain the second model, and the second model being associated with a second identifier, the second identifier being included in the first information; retraining the first model using only the second dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0041] In some alternative embodiments of the first aspect, the method further includes: the terminal device sending a first capability and / or a second capability; wherein the first capability includes fine-tuning of the terminal device support model; and wherein the second capability includes retraining of the terminal device support model.

[0042] In some alternative embodiments of the first aspect, the first capability further includes the terminal device supporting at least one of the following: fine-tuning the first model using the second dataset to obtain a second model, wherein the second model is associated with the first identifier; fine-tuning the first model using the second dataset to obtain a second model, wherein the second model is associated with a second identifier, wherein the second identifier is included in the first information.

[0043] In some alternative embodiments of the first aspect, the second capability further includes the terminal device supporting at least one of the following: retraining the first model using only the second dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0044] In some alternative embodiments of the first aspect, the first model includes at least one of the following: an artificial intelligence (AI) model; a machine learning (ML) model.

[0045] In a second aspect, a communication method is provided, the method comprising: a network device sending first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both a terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; wherein the first model and the second dataset are used to determine a second model.

[0046] In some alternative embodiments of the second aspect, the second model is obtained by the terminal device fine-tuning the first model using the second dataset.

[0047] In some alternative embodiments of the second aspect, the second model is obtained by the terminal device retraining the first model using the second dataset.

[0048] In some alternative embodiments of the second aspect, the second model is obtained by the terminal device retraining the first model using only the second dataset.

[0049] In some alternative embodiments of the second aspect, the second model is obtained by the terminal device retraining the first model using the second dataset and the first dataset.

[0050] In some alternative embodiments of the second aspect, the second model is associated with the first identifier.

[0051] In some alternative embodiments of the second aspect, the association between the first model and the first identifier has been severed.

[0052] In some alternative embodiments of the second aspect, the first information further includes a second identifier, wherein the second model is associated with the second model.

[0053] In some alternative embodiments of the second aspect, the first information further includes first indication information, which is used to instruct the terminal device to fine-tune the first model using the second dataset.

[0054] In some alternative embodiments of the second aspect, the first information further includes first indication information, which is used to instruct the terminal device to retrain the first model using the second dataset.

[0055] In some alternative embodiments of the second aspect, the first information further includes second indication information, which is used to instruct the terminal device to retrain the first model using only the second dataset to obtain the second model.

[0056] In some alternative embodiments of the second aspect, the first information further includes second instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset and the first dataset to obtain the second model.

[0057] In some alternative embodiments of the second aspect, the first information further includes third indication information, which is used to indicate one of the following: fine-tuning the first model using the second dataset to obtain the second model; retraining the first model using only the second dataset to obtain the second model; or retraining the first model using both the second dataset and the first dataset to obtain the second model.

[0058] In some alternative embodiments of the second aspect, the first information further includes fourth indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain the second model; or retraining the first model using the second dataset and the first dataset to obtain the second model.

[0059] In some alternative embodiments of the second aspect, the first information further includes fifth indication information, the fifth indication information being used to indicate one of the following: fine-tuning the first model using the second dataset to obtain the second model, and the second model being associated with the first identifier; fine-tuning the first model using the second dataset to obtain the second model, and the second model being associated with a second identifier, the second identifier being included in the first information; retraining the first model using only the second dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0060] In some alternative embodiments of the second aspect, the method further includes: the network device receiving a first capability and / or a second capability; wherein the first capability includes fine-tuning of the terminal device support model; and wherein the second capability includes retraining of the terminal device support model.

[0061] In some alternative embodiments of the second aspect, the first capability further includes the terminal device supporting at least one of the following: fine-tuning the first model using the second dataset to obtain a second model, wherein the second model is associated with the first identifier; fine-tuning the first model using the second dataset to obtain a second model, wherein the second model is associated with a second identifier, wherein the second identifier is included in the first information.

[0062] In some alternative embodiments of the second aspect, the second capability further includes the terminal device supporting at least one of the following: retraining the first model using only the second dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second dataset and the first dataset to obtain the second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0063] In some alternative embodiments of the second aspect, the first model includes at least one of the following: an artificial intelligence (AI) model; a machine learning (ML) model.

[0064] Thirdly, a terminal device is provided, comprising: a transceiver module for receiving first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both the terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; and a processing module for determining a second model based on the first model and the second dataset.

[0065] Fourthly, a network device is provided, comprising: a transceiver module for transmitting first information, the first information including a first identifier and a second dataset, the first identifier being associated with a first model, the first model being a model deployed by both a terminal device and the network device, the first model being trained based on the first dataset, and the second dataset being different from the first dataset; wherein the first model and the second dataset are used to determine a second model.

[0066] Fifthly, a terminal device is provided, comprising: one or more processors; wherein the terminal device is configured to execute the first aspect and any one of the communication methods in the first aspect.

[0067] A sixth aspect provides a network device, comprising: one or more processors; wherein the network device is configured to perform the second aspect and any one of the communication methods in the second aspect.

[0068] A seventh aspect provides a communication system, including a terminal device and a network device, wherein the terminal device is configured to implement the first aspect and any one of the communication methods in the first aspect, and the network device is configured to implement the second aspect and any one of the communication methods in the second aspect.

[0069] Eighthly, a storage medium is provided that stores instructions, which, when executed on a communication device, cause the communication device to perform a communication method as described in the first aspect and any one thereof, or the second aspect and any one thereof.

[0070] Ninth aspect, a program product is provided, comprising: a computer program, which, when executed by a communication device, causes the communication device to perform a communication method as described in the first aspect and any one of the first aspects or the second aspect and the second aspect.

[0071] In a tenth aspect, embodiments of this disclosure provide a computer program that, when run on a computer, causes the computer to perform the methods described in an optional implementation of the first or second aspect.

[0072] Eleventhly, embodiments of this disclosure provide a chip or chip system. The chip or chip system includes processing circuitry configured to perform the methods described in the optional implementations of the first or second aspect above.

[0073] It is understood that the terminal devices, access network devices, first network elements, other network elements, core network devices, communication systems, storage media, program products, computer programs, chips, or chip systems involved in the embodiments of this disclosure are all used to execute the methods proposed in the embodiments of this disclosure. Therefore, the beneficial effects that can be achieved can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0074] This disclosure provides communication methods, devices, systems, storage media, and program products. In some embodiments, the terms "communication method" and "information processing method" can be used interchangeably, as can the terms "communication device" and "information processing device" and "communication device," and the terms "information processing system" and "communication system."

[0075] This disclosure is not exhaustive, but merely illustrative of some embodiments, and is not intended to limit the scope of protection of this disclosure. Unless otherwise specified, each step in a particular embodiment can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a particular embodiment can also be implemented as an independent embodiment, and the order of the steps in a particular embodiment can be arbitrarily interchanged. Furthermore, the optional implementation methods in a particular embodiment can be arbitrarily combined; moreover, the embodiments can be arbitrarily combined, for example, some or all steps of different embodiments can be arbitrarily combined, and a particular embodiment can be arbitrarily combined with the optional implementation methods of other embodiments.

[0076] In each of the disclosed embodiments, unless otherwise specified or in case of logical conflict, the terminology and / or descriptions of the embodiments are consistent and can be referenced by each other. The technical environments of different embodiments can be combined to form new embodiments according to their inherent logical relationships.

[0077] The terminology used in the embodiments of this disclosure is for the purpose of describing particular embodiments only and is not intended to limit the scope of this disclosure.

[0078] In this embodiment of the disclosure, unless otherwise stated, elements expressed in the singular form, such as "a," "an," "the," "the," "the," "the," "the," "the," "this," etc., can mean "one and only one," or "one or more," "at least one," etc. For example, when using articles such as "a," "an," "the," etc. in translation, the noun following the article can be understood as either a singular expression or a plural expression.

[0079] In the embodiments disclosed herein, "multiple" refers to two or more.

[0080] In some embodiments, the terms “at least one of”, “one or more”, “a plurality of”, “multiple”, etc., may be used interchangeably.

[0081] In some embodiments, the notation "at least one of A and B", "A and / or B", "A in one case, B in another", "in response to one case A, in response to another case B", etc., may include the following technical solutions depending on the situation: in some embodiments, A (execute A regardless of B); in some embodiments, B (execute B regardless of A); in some embodiments, execution is selected from A and B (A and B are selectively executed); in some embodiments, A and B (both A and B are executed). The same applies when there are more branches such as A, B, C, etc.

[0082] In some embodiments, the notation "A or B" may include the following technical solutions, depending on the situation: in some embodiments, A (execution of A regardless of B); in some embodiments, B (execution of B regardless of A); in some embodiments, execution is selected from A and B (A and B are selectively executed). The same applies when there are more branches such as A, B, C, etc.

[0083] The prefixes "first," "second," etc., used in the embodiments of this disclosure are merely for distinguishing different descriptive objects and do not impose restrictions on the position, order, priority, quantity, or content of the descriptive objects. The description of the descriptive objects is found in the claims or the context of the embodiments, and the use of prefixes should not constitute unnecessary restrictions. For example, if the descriptive object is a "field," the ordinal numbers preceding "field" in "first field" and "second field" do not restrict the position or order of the "fields." "First" and "second" do not restrict whether the "fields" they modify are in the same message, nor do they restrict the order of "first field" and "second field." Similarly, if the descriptive object is a "level," the ordinal numbers preceding "level" in "first level" and "second level" do not restrict the priority between "levels." Furthermore, the number of descriptive objects is not limited by ordinal numbers and can be one or more. For example, in "first device," the number of "devices" can be one or more. Furthermore, the objects modified by different prefixes can be the same or different. For example, if the object being described is "device", then "first device" and "second device" can be the same device or different devices, and their types can be the same or different. Similarly, if the object being described is "information", then "first information" and "second information" can be the same information or different information, and their content can be the same or different.

[0084] In some embodiments, “including A,” “containing A,” “for indicating A,” and “carrying A” can be interpreted as directly carrying A or indirectly indicating A.

[0085] In some embodiments, the terms “in response to…”, “in response to determining…”, “in the case of…”, “when…”, “if…”, “if…”, etc., can be used interchangeably.

[0086] In some embodiments, the terms “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,” and “above” can be used interchangeably, as can the terms “less than,” “less than or equal to,” “not greater than,” “less than,” “less than or equal to,” “not more than,” “lower than,” “lower than or equal to,” “not higher than,” and “below”.

[0087] In some embodiments, the apparatus and device may be interpreted as physical or virtual, and their names are not limited to the names recorded in the embodiments. In some cases, they may also be understood as "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "chip", "chip system", "entity", "body", etc.

[0088] In some embodiments, "network" can be interpreted as devices included in the network, such as access network devices, core network devices, etc.

[0089] In some embodiments, "access network device (AN device)" may also be referred to as "radio access network device (RAN device)," "base station (BS)," "radio base station," or "fixed station." In some embodiments, it may also be understood as "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," or "bandwidth part (BWP)."

[0090] In some embodiments, "terminal" or "terminal device" may be referred to as "user equipment (UE)," "user terminal," "mobile station (MS)," "mobile terminal (MT)," "subscriber station," "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," etc.

[0091] In some embodiments, the acquisition of data, information, etc., may comply with the laws and regulations of the country where the location is situated.

[0092] In some embodiments, data, information, etc., may be obtained with the user's consent.

[0093] Furthermore, each element, each row, or each column in the table of this 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.

[0094] Figure 1 is a schematic diagram of a communication system architecture according to an embodiment of the present disclosure.

[0095] As shown in Figure 1, the communication system 100 includes a terminal device 101 and a network device 102.

[0096] In some embodiments, terminal device 101 includes, but is not limited to, at least one of the following: mobile phone, wearable device, Internet of Things device, car with communication function, smart car, tablet computer, computer with wireless transceiver function, virtual reality (VR) terminal device, augmented reality (AR) terminal device, wireless terminal device in industrial control, wireless terminal device in self-driving, wireless terminal device in remote medical surgery, wireless terminal device in smart grid, wireless terminal device in transportation safety, wireless terminal device in smart city, and wireless terminal device in smart home.

[0097] In some embodiments, network device 102 may include at least one of access network device and core network device.

[0098] In some embodiments, the access network device is, for example, a node or device that connects a terminal device to a wireless network. The access network device may include, but is not limited to, at least one of the following in a 5G communication system: evolved Node B (eNB), next-generation eNB (ng-eNB), next-generation Node B (gNB), node B (NB), home node B (HNB), home evolved node B (HeNB), radio backhaul device, radio network controller (RNC), base station controller (BSC), base transceiver station (BTS), base band unit (BBU), mobile switching center, base station in a 6G communication system, open RAN, cloud RAN, base station in other communication systems, and access node in a Wi-Fi system.

[0099] In some embodiments, the technical solutions of this disclosure can be applied to the Open RAN architecture. In this case, the interfaces between or within access network devices involved in the embodiments of this disclosure can be transformed into internal interfaces of Open RAN. The processes and information interactions between these internal interfaces can be implemented by software or programs.

[0100] In some embodiments, the access network device may be composed of a central unit (CU) and a distributed unit (DU). The CU may also be called a control unit. The CU-DU structure can separate the protocol layer of the access network device. Some of the protocol layer functions are centrally controlled by the CU, while the remaining part or all of the protocol layer functions are distributed in the DU and centrally controlled by the CU. However, this is not the only possibility.

[0101] In some embodiments, a core network device may be a single device comprising one or more network elements, or it may be multiple devices or a group of devices, each comprising all or part of the aforementioned one or more network elements. Network elements may be virtual or physical. The core network may include, for example, at least one of an Evolved Packet Core (EPC), a 5G Core Network (5GCN), or a Next Generation Core (NGC).

[0102] It is understood that the communication system described in this disclosure is for the purpose of more clearly illustrating the technical solutions of this disclosure, and does not constitute a limitation on the technical solutions proposed in this disclosure. As those skilled in the art will know, with the evolution of system architecture and the emergence of new business scenarios, the technical solutions proposed in this disclosure are also applicable to similar technical problems.

[0103] The following embodiments of this disclosure can be applied to the communication system 100 shown in FIG1, or to some of the main bodies, but are not limited thereto. The main bodies shown in FIG1 are illustrative. The communication system may include all or some of the main bodies in FIG1, or may include other main bodies outside of FIG1. ​​The number and form of each main body are arbitrary. Each main body may be physical or virtual. The connection relationship between the main bodies is illustrative. The main bodies may not be connected or may be connected. The connection can be in any way, it can be a direct connection or an indirect connection, it can be a wired connection or a wireless connection.

[0104] The embodiments disclosed herein can be applied to Long Term Evolution (LTE), LTE-Advanced (LTE-A), LTE-Beyond (LTE-B), SUPER 3G, IMT-Advanced, 4th generation mobile communication system (4G), 5th generation mobile communication system (5G), 6th generation mobile communication system (6G), 5G New Radio (NR), Future Radio Access (FRA), New-Radio Access Technology (RAT), New Radio (NR), New Radio Access (NX), Future Generation Radio Access (FX), Global System for Mobile Communications (GSM), CDMA2000, Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi), IEEE 802.16 (WiMAX), and IEEE 802.20, Ultra-Wideband (UWB), Bluetooth (a registered trademark), Public Land Mobile Network (PLMN) networks, Device-to-Device (D2D) systems, Machine-to-Machine (M2M) systems, Internet of Things (IoT) systems, Vehicle-to-Everything (V2X) systems, systems utilizing other communication methods, and next-generation systems built upon them, etc. Furthermore, multiple systems can be combined (e.g., a combination of LTE or LTE-A with 5G).

[0105] The development of communication technologies has brought tremendous changes to all aspects of people's lives, such as 5G (NR, 6G, etc.). According to the vision of the International Telecommunication Union (ITU), 5G will permeate all areas of future society, building a comprehensive information ecosystem centered on the user. Specifically, 5G user experience speeds can reach 100 Mbit / s to 1 Gbit / s, supporting ultimate service experiences such as mobile virtual reality; 5G peak speeds can reach 10 Gbit / s to 20 Gbit / s, with a traffic density of 10 Mbit / s / m², supporting more than a thousandfold increase in mobile traffic; 5G connection density can reach 1 million / m², effectively supporting massive numbers of IoT devices; 5G transmission latency can be down to the millisecond level, meeting the stringent requirements of vehicle-to-everything (V2X) and industrial control; 5G can support mobile speeds of 500 km / h, providing a good user experience even in high-speed rail environments. It is conceivable that 5G, as a representative of new infrastructure, will reshape the future information society.

[0106] In recent years, artificial intelligence (AI) and machine learning (ML) technologies have achieved continuous breakthroughs in multiple fields. The ongoing development of fields such as intelligent voice and computer vision has not only brought a wide variety of applications to smart terminal devices, but has also found widespread use in education, transportation, home, healthcare, retail, security, and many other sectors, bringing convenience to people's lives while promoting industrial upgrading across various industries. AI technology is also accelerating its cross-disciplinary integration with other disciplines, combining knowledge from different fields while providing new directions and methods for the development of various disciplines.

[0107] In the 3rd generation partnership project (3GPP) Release 18, a research project on the application of artificial intelligence (AI) technology in the radio access network (RAN) working group (WG) 1 was established. This project aimed to investigate how to introduce AI technology into the RAN and explore how AI technology can assist in improving RAN transmission technology.

[0108] In research geared towards 6G, 6G systems can provide AI services across more dimensions. This mainly includes the following three aspects:

[0109] AI-enabled connectivity. This means using AI to improve communication performance, such as using AI for beam management.

[0110] Computing power services. This refers to the network side providing computing power to the terminal device side, such as assisting the terminal device in model training, model inference, and other advanced AI services. Essentially, it involves enhancing the network's transmission pipeline to improve the user experience of AI application services.

[0111] In some embodiments, AI / ML model identification (MI) was studied. AI / ML model identification is applied to two-sided model scenarios, where network devices and terminal devices each deploy an AI / ML model. These two AI / ML models form a pair, constituting a two-sided model that jointly implements a function. Due to differences in network device configuration, terminal device configuration, and service type, network devices and terminal devices typically deploy multiple pairs of two-sided models, meaning each deploys at least one AI / ML model. However, not every AI / ML model from a network device and every AI / ML model from a terminal device can form a valid two-sided model pair; certain conditions must be met for a pair to be valid. The purpose of AI / ML model identification is to identify each valid two-sided model pair from the at least one AI / ML model deployed by the network device and the at least one AI / ML model deployed by the terminal device.

[0112] A typical application example is Channel State Information (CSI) compression. The terminal device extracts features from the measured CSI using an AI / ML model, outputting a low-dimensional CSI, thus completing CSI compression. This low-dimensional CSI is then fed back to the network device. The network device receives this low-dimensional CSI and uses another AI / ML model to reconstruct the complete CSI, completing CSI decompression. The prerequisite for the network device's AI / ML model to correctly reconstruct the complete CSI is that both the network device's AI / ML model and the terminal device's AI / ML model were trained, validated, and tested using the same dataset. Therefore, one model recognition method is for the network device to send its dataset to the terminal device. The AI / ML model trained on the network device using this dataset and the AI / ML model trained on the terminal device using the same dataset form a two-sided model, thus completing model recognition.

[0113] When a network device sends a dataset to a terminal device, it also sends an identifier (ID). This identifier is used to pair the AI / ML models of the network device and the terminal device. Specifically, the network terminal device sends the identifier along with the dataset. Assuming this identifier carries a first identifier, both the network device and the terminal device will consider the AI / ML model trained on the terminal device using the dataset to be associated with this first identifier. The AI / ML model associated with this first identifier and the AI / ML model trained on the network device using the dataset form a valid two-sided model pair. Since both the network device and the terminal device maintain at least one two-sided model pair simultaneously, when the network device switches from another two-sided model to this one, it sends the first identifier to the terminal device. Upon receiving the first identifier, the terminal device switches to the AI / ML model associated with the first identifier, thus forming a valid two-sided model pair with the AI / ML model switched by the network device.

[0114] However, it is not currently supported to manipulate existing bilateral models to obtain new bilateral models.

[0115] Therefore, this disclosure provides a communication method in which a terminal device receives first information, in which a first identifier is associated with a first model. The first model is a model deployed by both the terminal device and the network device; this model can also be called a bilateral model. The terminal device determines the first model from the models deployed on the terminal device side based on the first identifier. The second dataset included in the first information is different from the first dataset used to train the first model. The terminal device can determine the second model based on the second dataset and the first dataset. That is, it can use the new dataset (i.e., the second dataset) to operate on the existing bilateral model (i.e., the first model) to obtain a new bilateral model (i.e., the second model), thereby improving model performance, maintaining a better bilateral model, and improving the efficiency and accuracy of model-related services.

[0116] Figure 2 is a schematic diagram of a communication method interaction according to an embodiment of the present disclosure. As shown in Figure 2, this embodiment of the present disclosure relates to a communication method for a communication system 100, the method including:

[0117] In step S2101, the terminal device 101 sends the first capability and / or the second capability to the network device 102.

[0118] In some embodiments, network device 102 receives a first capability and / or a second capability sent by terminal device 101.

[0119] In some embodiments, sending the first capability and / or the second capability can also be understood as reporting the first capability and / or the second capability. For example, a terminal device may passively report its capability information in response to a query signaling from a network device, and indicate in the capability information that the terminal device supports the first capability and / or the second capability. Alternatively, a terminal device may actively report its capability information, and indicate in the capability information that the terminal device supports the first capability and / or the second capability. It is understood that the terminal device capability information in this embodiment may be capability information reporting all capabilities of the terminal device, that is, reporting the first capability and / or the second capability while reporting all capabilities of the terminal device. Alternatively, the terminal device capability information may be capability information reporting only the first capability and / or the second capability, that is, the terminal device may report the first capability and / or the second capability separately, without reporting it together with other capabilities. The names of the terminal device capability information are merely exemplary; the terminal device can report the first capability and / or the second capability using any information, and this disclosure does not limit this.

[0120] Optionally, the terminal device sends a first capability to the network device. The first capability includes fine-tuning of a model supported by the terminal device. This model can be an AI model, an ML model, etc., but is not limited to these. The terminal device sends the first capability to the network device to enable the network device to configure model fine-tuning for the terminal device. For example, the terminal device deploys a first model, which is deployed on both the terminal device and the network device. The first model is trained based on a first dataset and associated with a first identifier, which is sent to the terminal device by the network device. Since the first model is deployed on both the terminal device and the network device, and the first dataset used to train the first model is sent to the terminal device by the network device, the fine-tuning of the first model can be configured by the network device. For example, by sending the first capability to the network device, the terminal device enables the network device to send a first identifier and a second dataset to the terminal device. Based on the first identifier, the terminal device can determine the first model associated with the first identifier, and based on the second dataset, it can fine-tune the first model to obtain a second model. The second dataset is different from the first dataset.

[0121] Optionally, the terminal device sends a second capability to the network device. This second capability includes the terminal device supporting model retraining. The model can be an AI model, an ML model, etc., but is not limited to these. The terminal device sends the second capability to the network device to enable the network device to configure model retraining for the terminal device. For example, the terminal device deploys a first model, which is deployed on both the terminal device and the network device. The first model is trained based on a first dataset and associated with a first identifier, which is sent to the terminal device by the network device. Since the first model is deployed on both the terminal device and the network device, and the first dataset used to train the first model is sent to the terminal device by the network device, the retraining of the first model can be configured by the network device. For example, by sending the second capability to the network device, the terminal device enables the network device to send a first identifier and a second dataset to the terminal device. The terminal device can determine the first model associated with the first identifier based on the first identifier, and can retrain the first model based on the second dataset to obtain a second model. The second dataset is different from the first dataset.

[0122] It is understood that the above optional examples can be implemented in combination; for example, the terminal device can send a first capability and a second capability to the network device.

[0123] Understandably, the second dataset used for fine-tuning and the second dataset used for retraining can be the same second dataset or different second datasets.

[0124] In some embodiments, the first capability further includes the terminal device supporting at least one of the following: fine-tuning the first model using a second dataset to obtain a second model, wherein the second model is associated with a first identifier; fine-tuning the first model using a second dataset to obtain a second model, wherein the second model is associated with a second identifier, wherein the second identifier is included in the first information.

[0125] Optionally, the first capability includes the terminal device supporting the fine-tuning of the first model using a second dataset to obtain a second model, and the second model being associated with a first identifier. For example, in this case, the network device can send the second dataset and the first identifier so that the terminal device can determine the first model based on the first identifier, fine-tune the first model based on the second dataset to obtain the second model, and associate the second model with the first identifier. The network device can also fine-tune the first model deployed on the network device based on the second dataset to obtain the second model, and associate the second model with the first identifier.

[0126] Optionally, the first capability includes the terminal device supporting the fine-tuning of the first model using a second dataset to obtain a second model, and the second model being associated with a second identifier. For example, in this case, the network device can send the second dataset, the first identifier, and the second identifier, so that the terminal device can determine the first model based on the first identifier, fine-tune the first model based on the second dataset to obtain the second model, and associate the second model with the second identifier. The network device can also fine-tune the first model deployed on the network device based on the second dataset to obtain the second model, and associate the second model with the second identifier.

[0127] It is understood that the above optional examples can be implemented in combination. For example, the first capability may include the two items mentioned above, and how the network device is configured and fine-tuned, such as configuring the second dataset and the first identifier, or configuring the second dataset, the first identifier, and the second identifier, can be determined by the network device.

[0128] It is understood that if the network device is configured with a second dataset, a first identifier, and a second identifier, the terminal device can associate the second model with the second identifier or with the first identifier. That is, even if the network device sends the second identifier, the terminal device can still associate the model with the first identifier, and this disclosure does not impose any limitations.

[0129] In some embodiments, the second capability further includes the terminal device supporting at least one of the following: retraining the first model using only the second dataset to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a first identifier; and retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0130] Optionally, the second capability includes the terminal device supporting the retraining of the first model using only the second dataset to obtain a second model, and associating the second model with the first identifier. For example, in this case, the network device can send the second dataset and the first identifier so that the terminal device can determine the first model based on the first identifier, retrain the first model based on the second dataset to obtain the second model, and associate the second model with the first identifier. Alternatively, the network device can also retrain the first model deployed on the network device based on the second dataset to obtain the second model and associate the second model with the first identifier.

[0131] Optionally, the second capability includes the terminal device supporting the retraining of the first model using a second dataset and a first dataset to obtain a second model, with the second model associated with the first identifier. For example, in this case, the network device can send the second dataset and the first identifier so that the terminal device can determine the first model based on the first identifier, retrain the first model based on the second dataset and the first dataset to obtain the second model, and associate the second model with the first identifier. Alternatively, the network device can also retrain the first model deployed on the network device based on the second dataset and the first dataset to obtain the second model, and associate the second model with the first identifier.

[0132] Optionally, the second capability includes the terminal device supporting the retraining of the first model using a second dataset and a first dataset to obtain a second model, wherein the second model is associated with a second identifier, the second identifier being included in the first information. For example, in this case, the network device can send the second dataset, the first identifier, and the second identifier, so that the terminal device can determine the first model based on the first identifier, retrain the first model based on the second dataset and the first dataset to obtain the second model, and associate the second model with the second identifier. Alternatively, the network device can also retrain the first model deployed on the network device based on the second dataset and the first dataset to obtain the second model, and associate the second model with the second identifier.

[0133] It is understood that the above optional examples can be implemented in combination. For example, the second capability may include the above three items, and how the network device is configured to retrain, for example, whether to configure the second dataset and the first identifier, or to configure the second dataset, the first identifier and the second identifier, can be determined by the network device.

[0134] It is understood that if the network device is configured with a second dataset, a first identifier, and a second identifier, the terminal device can associate the second model with the second identifier or with the first identifier. That is, even if the network device sends the second identifier, the terminal device can still associate the model with the first identifier, and this disclosure does not impose any limitations.

[0135] In some embodiments, if the terminal device supports only one of the first capability and the second capability, the network device can configure the terminal device to perform either fine-tuning or retraining. The terminal device can also determine to perform either fine-tuning or retraining based on the second dataset and the first identifier sent by the network device. For example, if the terminal device supports the first capability, the network device can configure fine-tuning for the terminal device, sending the second dataset and the first identifier to the terminal device, and the terminal device can also determine to fine-tune the first model based on the second dataset, rather than retraining or performing other operations. As another example, if the terminal device supports the second capability, the network device can configure retraining for the terminal device, sending the second dataset and the first identifier to the terminal device, and the terminal device can also determine to retrain the first model based on the second dataset, rather than performing fine-tuning or other operations.

[0136] However, if the terminal device supports both the first and second capabilities, the network device can configure fine-tuning and retraining for the terminal device. When the network device sends the second dataset and the first identifier to the terminal device, the terminal device cannot determine whether to fine-tune or retrain based on the second dataset. If the terminal device supports the first capability, but the first capability includes both of the above-mentioned items, when the network device sends the second dataset and the first identifier to the terminal device, the terminal device cannot accurately determine whether to associate the second model with the first identifier or the second identifier. If the terminal device supports the second capability, but the second capability includes at least two of the three items mentioned above, when the network device sends the second dataset and the first identifier to the terminal device, the terminal device cannot accurately determine whether to associate the second model with the first identifier or the second identifier, and whether to retrain the first model using only the second dataset or using both the second and first datasets.

[0137] Therefore, in some embodiments of this disclosure, the network device sends some indication information while sending the second dataset and the first identifier to the terminal device, in order to solve the above-mentioned problems. Specific embodiments can be found in step S2102 below.

[0138] In step S2102, network device 102 sends first information to terminal device 101.

[0139] In some embodiments, terminal device 101 receives first information sent by network device 102.

[0140] In some embodiments, the first information includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed on both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. The terminal device can determine the second model based on the first model and the second dataset.

[0141] In some embodiments, the first information may further include first instruction information, which is used to instruct the terminal device to fine-tune or retrain the first model using the second dataset.

[0142] Optionally, the first instruction information is used to instruct the terminal device to fine-tune the first model using the second dataset. The terminal device can use the second dataset to fine-tune the first model to obtain the second model.

[0143] Optionally, the first instruction information is used to instruct the terminal device to retrain the first model using the second dataset. The terminal device can use the second dataset to retrain the first model to obtain the second model.

[0144] It is understood that the second model obtained by fine-tuning the first model and the second model obtained by retraining the first model can be different second models or the same second model; this disclosure does not limit this.

[0145] Understandably, if the terminal device only supports model fine-tuning or retraining, the first instruction information can be omitted. For example, if the terminal device only supports model fine-tuning, the first instruction information is unnecessary, and the terminal device can fine-tune the first model based on the second dataset to obtain the second model. Similarly, if the terminal device only supports model retraining, the first instruction information is unnecessary, and the terminal device can retrain the first model based on the second dataset to obtain the second model.

[0146] In some embodiments, the first information may further include second instruction information, which is used to instruct the terminal device to retrain the first model using only the second dataset or using the first dataset and the second dataset.

[0147] Optionally, the second instruction information is used to instruct the terminal device to retrain the first model using only the second dataset. The terminal device can retrain the first model using only the second dataset to obtain the second model.

[0148] Optionally, the second instruction information is used to instruct the terminal device to retrain the first model using the first data set and the second dataset. The terminal device can use the first data set and the second dataset to retrain the first model and obtain the second model.

[0149] In some embodiments, the first information may include only the first indication information, only the second indication information, or both the first and second indication information.

[0150] Optionally, the first information may only include the first instruction information. For example, the terminal device may support a first capability and a second capability. For the second capability, the terminal device supports retraining the first model using only the second dataset, but does not support retraining the first model using both the first and second datasets. For the first capability, the terminal device supports fine-tuning the first model using the second dataset. That is, the terminal device supports both fine-tuning and retraining the first model, and the dataset used is always the second dataset. Therefore, the network device can instruct the terminal device to fine-tune or retrain the first model simply through the first instruction information.

[0151] Optionally, the first information may only include the second instruction information. For example, if the terminal device only supports the second capability, then the first instruction information can be omitted. Furthermore, regarding the second capability, if the terminal device supports retraining the first model using only the second dataset, or retraining the first model using both the first and second datasets, then the network device can use the second instruction information to indicate whether to retrain the first model using only the second dataset, or to support retraining the first model using both the second and first datasets.

[0152] Optionally, the first information may include first instruction information and second instruction information. For example, if the terminal device supports both the first and second capabilities, the network device can use the first instruction information to instruct the terminal device to fine-tune or retrain the first model. If the first instruction information instructs the terminal device to retrain the first model, and the terminal device supports both retraining the first model using only the second dataset and retraining the first model using both the first and second datasets, the network device can use the second instruction information to indicate whether to retrain the first model using only the second dataset or to support retraining the first model using both the second and first datasets.

[0153] In some embodiments, the first information may further include third indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain a second model; retraining the first model using only the second dataset to obtain a second model; or retraining the first model using both the second and first datasets to obtain a second model. For example, if the network device sends the second dataset and the first identifier, the terminal device may default to associating the second model with the first identifier. Alternatively, the network device may instruct the second model to be associated with the first identifier. When the terminal device determines that the second model is associated with the first identifier, the optional schemes for determining the second model based on the first model and the second dataset may include the three items in this embodiment. Therefore, the first information may include the third indication information to indicate which method is used to determine the second model when the second model is associated with the first identifier.

[0154] Optionally, the third instruction information is used to instruct the first model to be fine-tuned using the second dataset to obtain the second model. The terminal device can use the second dataset to fine-tune the first model to obtain the second model and associate the second model with the first identifier.

[0155] Optionally, the third instruction information is used to instruct that the first model be retrained using only the second dataset to obtain the second model. The terminal device can retrain the first model using only the second dataset to obtain the second model and associate the second model with the first identifier.

[0156] Optionally, the third instruction information is used to instruct the first model to be retrained using the second dataset and the first dataset to obtain the second model. The terminal device can use the second dataset and the first dataset to retrain the first model to obtain the second model and associate the second model with the first identifier.

[0157] In some embodiments, the first information may further include fourth indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain a second model; or retraining the first model using the second dataset and the first dataset to obtain a second model. For example, if the network device sends the second dataset, the first identifier, and the second identifier, the terminal device may default to associating the second model with the second identifier. Alternatively, the network device may instruct the second model to be associated with the second identifier. When the terminal device determines that the second model is associated with the second identifier, the optional schemes for determining the second model based on the first model and the second dataset may include the two items in this embodiment. Therefore, the first information may include the fourth indication information to indicate which method is used to determine the second model when the second model is associated with the second identifier.

[0158] Optionally, the fourth instruction information is used to instruct the first model to be fine-tuned using the second dataset to obtain the second model. The terminal device can use the second dataset to fine-tune the first model to obtain the second model and associate the second model with the second identifier.

[0159] Optionally, the fourth instruction information is used to instruct the first model to be retrained using the second dataset and the first dataset to obtain the second model. The terminal device can use the second dataset and the first dataset to retrain the first model to obtain the second model, and associate the second model with the second identifier.

[0160] In some embodiments, the first information further includes fifth indication information, which indicates one of the following: fine-tuning the first model using a second dataset to obtain a second model, and the second model being associated with a first identifier; fine-tuning the first model using a second dataset to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information; retraining the first model using only the second dataset to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information. For example, a network device can use third indication information to instruct a terminal device whether to fine-tune or retrain the first model, and to indicate whether to use the second dataset or both the second and first datasets, and to indicate whether the obtained second model is associated with the first identifier or the second identifier.

[0161] It is understood that the various indications included in the first information are used to instruct the terminal device to determine the scheme of the second model based on the first model and the second dataset. This disclosure only exemplifies some schemes, but is not limited to them. In some cases, the terminal device can determine a predefined scheme from the protocol. In one case, the protocol specifies some of the schemes in the above-exemplified schemes, or the terminal device supports some of the schemes, and the various indications can indicate one of the schemes in the above-exemplified schemes. For example, the third indication can be used to indicate one of (1), (2) and (3) below; or, if the protocol specifies (1) and (2), then the third indication can be used to indicate one of (1) and (2); or, if the terminal device only supports (2) and (3), then the third indication can be used to indicate one of (2) and (3). This disclosure uses the third indication as an example, but is not limited to it.

[0162] (1) Fine-tune the first model using the second dataset to obtain the second model;

[0163] (2) Retrain the first model using only the second dataset to obtain the second model;

[0164] (3) The first model is retrained using the second dataset and the first dataset to obtain the second model.

[0165] In some embodiments, the name of the first information is not limited, and it may be, for example, "configuration information".

[0166] In step S2103, the terminal device 101 determines the second model based on the first model and the second dataset.

[0167] In some embodiments, the terminal device may determine the second model based on the first model and the second dataset by referring to the manner described in the above embodiments.

[0168] Optionally, if the terminal device only supports the first capability, the terminal device can fine-tune the first model based on the second dataset.

[0169] Optionally, if the terminal device only supports the second capability, the terminal device can retrain the first model based on the second dataset.

[0170] Optionally, if the terminal device supports both the first and second capabilities, and the first information includes first instruction information, then when the first instruction information instructs the terminal device to fine-tune the first model, the terminal device can fine-tune the first model based on the second dataset. When the first instruction information instructs the terminal device to retrain the first model, the terminal device can retrain the first model based on the second dataset.

[0171] Optionally, if the terminal device only supports the second capability or the first instruction information instructs the terminal device to retrain the first model, and the second instruction information instructs the terminal device to retrain the first model using only the second dataset, then the terminal device will retrain the first model using only the second dataset.

[0172] Optionally, if the terminal device only supports the second capability or the first instruction information instructs the terminal device to retrain the first model, and the second instruction information instructs the terminal device to retrain the first model using the second dataset and the first dataset, then the terminal device retrains the first model using the second dataset and the first dataset.

[0173] Optionally, if the terminal device only supports the second capability or the first instruction information instructs the terminal device to retrain the first model, and the first information only includes the first identifier but not the second identifier, then the terminal device may retrain the first model using only the second dataset, or retrain the first model using both the first dataset and the second dataset.

[0174] Optionally, if the terminal device only supports the second capability or the first instruction information instructs the terminal device to retrain the first model, and the first information includes a first identifier and a second identifier, then the terminal device can use the second dataset and the first dataset to retrain the first model.

[0175] Understandably, if the first information includes a first identifier but not a second identifier, the second model is often associated with the first identifier. In this case, the first model can be retrained using only the second dataset. However, when the first information includes both a first and a second identifier, the second model is often associated with the second identifier. In this case, retraining the first model using only the second dataset is similar to the terminal device training the first model based on the first dataset and associating it with the first identifier. That is, it's equivalent to training a new two-sided model, rather than a two-sided model resulting from retraining the first model. Therefore, in this situation, the terminal device can use both the second and first datasets to retrain the first model and associate the second model with the second identifier.

[0176] Optionally, after the terminal device determines the second model, it can associate the second model with the first identifier.

[0177] Optionally, after the terminal device determines the second model, it can associate the second model with the second identifier.

[0178] Optionally, the terminal device may first determine that the second model is associated with the first identifier. If the second model is associated with the first identifier, the terminal device may only support one scheme to determine the second model based on the first model and the second dataset, and determine the second model.

[0179] Optionally, the terminal device may first determine that the second model is associated with the first identifier. When the second model is associated with the first identifier, the terminal device can support multiple schemes for determining the second model based on the first model and the second dataset, such as at least two of the three schemes provided in the above embodiments. The network device can indicate one of the multiple schemes through third indication information, as detailed in the above embodiments, which will not be repeated here. The terminal device can determine the second model based on the indication of the third indication information.

[0180] Optionally, the terminal device may first determine that the second model is associated with the second identifier. In the case that the second model is associated with the second identifier, the terminal device may only support one scheme for determining the second model based on the first model and the second dataset, and determine the second model.

[0181] Optionally, the terminal device may first determine that the second model is associated with the second identifier. If the second model is associated with the first identifier, the terminal device can support multiple schemes for determining the second model based on the first model and the second dataset, such as the two schemes provided in the above embodiments. The network device can indicate one of the multiple schemes through the fourth indication information; for details, please refer to the above embodiments, which will not be elaborated further in this disclosure. The terminal device can determine the second model based on the indication of the fourth indication information.

[0182] Optionally, the terminal device may determine one of several schemes for determining the second model based on the fifth indication information, such as: fine-tuning the first model using the second dataset to obtain the second model, and the second model being associated with the first identifier; fine-tuning the first model using the second dataset to obtain the second model, and the second model being associated with the second identifier, the second identifier being included in the first information; retraining the first model using only the second dataset to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second and first datasets to obtain the second model, and the second model being associated with the first identifier; retraining the first model using both the second and first datasets to obtain the second model, and the second model being associated with the second identifier, the second identifier being included in the first information.

[0183] Optionally, the terminal device may determine, from the provisions of the protocol, a scheme for determining the second model based on the first model and the second dataset. For example, the protocol may specify one of the following: fine-tuning the first model using the second dataset to obtain the second model, and the second model is associated with the first identifier; fine-tuning the first model using the second dataset to obtain the second model, and the second model is associated with the second identifier, which is included in the first information; retraining the first model using only the second dataset to obtain the second model, and the second model is associated with the first identifier; retraining the first model using both the second and first datasets to obtain the second model, and the second model is associated with the first identifier; retraining the first model using both the second and first datasets to obtain the second model, and the second model is associated with the second identifier, which is included in the first information.

[0184] In step S2104, the terminal device 101 disconnects the association between the first model and the first identifier.

[0185] In some embodiments, if the second model is associated with the first identifier, the terminal device and the network device can detach the association between the first identifier and the first model.

[0186] It is understood that step S2104 is optional. For example, if the second model is associated with the second identifier, the association between the first identifier and the first model can be left untouched; that is, the first model continues to be associated with the first identifier, and the second model is associated with the second identifier. Alternatively, even if the second model is associated with the first identifier, the association between the first identifier and the first model can be left untouched; that is, both the first model and the second model can be associated with the first identifier.

[0187] In step S2105, network device 102 disconnects the association between the first model and the first identifier.

[0188] In some embodiments, the first model is a model deployed by both the terminal device and the network device. If the second model is associated with the first identifier, the network device 102 can also detach the association between the first model and the second identifier.

[0189] It is understandable that network devices can disconnect the association between the first model and the first identifier before sending the first information. That is, this disclosure does not limit the execution order of each step, especially step S2105, which, although written after step S2104, may actually be executed before step S2104, and this disclosure does not limit it.

[0190] It is understandable that step S2105 is optional, for reasons similar to those of step S2104, and will not be repeated here.

[0191] The communication method involved in the embodiments of this disclosure may include at least one of steps S2101 to S2105. For example, steps S2102 and S2103 may be implemented as independent embodiments, but are not limited thereto.

[0192] In some embodiments, step S2101 is optional and may be omitted or replaced in different embodiments.

[0193] In some embodiments, step S2104 is optional and may be omitted or replaced in different embodiments.

[0194] In some embodiments, step S2105 is optional and may be omitted or replaced in different embodiments.

[0195] In some embodiments, other optional implementations described before or after the specification corresponding to FIG2 may be referred to.

[0196] Figure 3 is a flowchart illustrating a communication method according to an embodiment of the present disclosure. As shown in Figure 3, this embodiment of the present disclosure relates to a communication method executed by a terminal device 101, the method including:

[0197] Step S3101: Send the first capability and / or the second capability.

[0198] The optional implementation of step S3101 can be found in the optional implementation of step S2101 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0199] In some embodiments, terminal device 101 sends a first capability and / or a second capability to network device 102, but is not limited thereto, and may also send the first capability and / or the second capability to other entities.

[0200] Step S3102: Obtain the first information.

[0201] The optional implementation of step S3102 can be found in the optional implementation of step S2102 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0202] In some embodiments, terminal device 101 receives first information sent by network device 102, but is not limited thereto; it may also receive first information sent by other entities.

[0203] In some embodiments, the terminal device 101 obtains first information as defined by the protocol.

[0204] In some embodiments, the terminal device 101 obtains first information from the upper layer(s).

[0205] In some embodiments, the terminal device 101 processes the information to obtain the first information.

[0206] In some embodiments, step S3102 is omitted, and the terminal device 101 autonomously implements the function indicated by the first information, or the above function is the default or default.

[0207] Step S3103: Determine the second model based on the first model and the second dataset.

[0208] The optional implementation of step S3103 can be found in the optional implementation of step S2103 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0209] Step S3104: Disconnect the association between the first model and the first identifier.

[0210] The optional implementation of step S3104 can be found in the optional implementation of step S2104 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0211] The communication method involved in the embodiments of this disclosure may include at least one of steps S3101 to S3104. For example, steps S3102 and S3103 may be implemented as independent embodiments, but are not limited thereto.

[0212] In some embodiments, step S3101 is optional and may be omitted or replaced in different embodiments.

[0213] In some embodiments, step S3104 is optional and may be omitted or replaced in different embodiments.

[0214] In some embodiments, other optional implementations may be described before or after the specification corresponding to FIG3.

[0215] Figure 4 is a flowchart illustrating a communication method according to an embodiment of the present disclosure. As shown in Figure 4, this embodiment of the present disclosure relates to a communication method executed by a network device 102, the method comprising:

[0216] Step S4101: Acquire the first ability and / or the second ability.

[0217] The optional implementation of step S4101 can be found in the optional implementation of step S2101 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0218] In some embodiments, network device 102 receives a first capability and / or a second capability sent by terminal device 101, but is not limited thereto, and may also receive a first capability and / or a second capability sent by other entities.

[0219] In some embodiments, network device 102 acquires a first capability and / or a second capability as defined by a protocol.

[0220] In some embodiments, network device 102 obtains first capabilities and / or second capabilities from upper layer(s).

[0221] In some embodiments, network device 102 performs processing to obtain a first capability and / or a second capability.

[0222] In some embodiments, step S4101 is omitted, and the network device 102 autonomously implements the functions indicated by the first capability and / or the second capability, or the above functions are defaulted or set to default.

[0223] Step S4102: Send the first message.

[0224] The optional implementation of step S4102 can be found in the optional implementation of step S2102 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0225] In some embodiments, network device 102 sends first information to terminal device 101, but is not limited thereto; it may also send first information to other entities.

[0226] Step S4103: Disconnect the association between the first model and the first identifier.

[0227] The optional implementation of step S4103 can be found in the optional implementation of step S2105 in Figure 2, as well as other related parts in the embodiments involved in Figure 2, which will not be repeated here.

[0228] The communication method involved in the embodiments of this disclosure may include at least one of steps S4101 to S4103. For example, step S4102 may be implemented as a separate embodiment, but is not limited thereto.

[0229] In some embodiments, step S4101 is optional and may be omitted or replaced in different embodiments.

[0230] In some embodiments, step S4103 is optional and may be omitted or replaced in different embodiments.

[0231] In some embodiments, other optional implementations may be described before or after the specification corresponding to Figure 4.

[0232] Figure 5a is an interactive schematic diagram of a communication method according to an embodiment of the present disclosure. As shown in Figure 5a, the present disclosure relates to a communication method, which includes:

[0233] In step S5101, network device 102 sends first information to terminal device 101.

[0234] In step S5102, the terminal device 101 determines the second model based on the first model and the second dataset.

[0235] In some embodiments, the above methods may include the methods of the embodiments related to the communication system 100, terminal device 101, and network device 102, which will not be described again here.

[0236] This disclosure provides a model recognition method, as follows:

[0237] In some embodiments, the network device sends a first dataset and a first identifier to the terminal device, the terminal device trains a first AI / ML model based on the first dataset, and the first AI / ML model is associated with the first identifier.

[0238] Scenario 1: Fine-tuning of AI / ML models:

[0239] Method 1-1: The network device sends a second dataset and a first identifier to the terminal device. The terminal device then uses the second dataset to fine-tune the first AI / ML model associated with the first identifier to determine the second AI / ML model, which is then associated with the first identifier.

[0240] Method 1-2: The network device sends a second dataset, a first identifier, and a second identifier to the terminal device. The terminal device then uses the second dataset to fine-tune the first AI / ML model associated with the first identifier to determine the second AI / ML model, and the second AI / ML model is associated with the second identifier.

[0241] Scenario 2: AI / ML model retraining:

[0242] Method 2-1: The network device sends a second dataset and a first identifier to the terminal device. The terminal device then uses the second dataset to train a second AI / ML model, and the second AI / ML model is associated with the first identifier.

[0243] Method 2-2: The network device sends a second dataset and a first identifier to the terminal device. The terminal device then uses the second dataset and the first dataset used for the association of the first identifier to train a second AI / ML model, and the second AI / ML model is associated with the first identifier.

[0244] Method 2-3: The network device sends a second dataset, a first identifier, and a second identifier to the terminal device. The terminal device then uses the second dataset and the first dataset used for association with the first identifier to train a second AI / ML model, and the second AI / ML model is associated with the second identifier.

[0245] In some embodiments, the network device sends necessary information to the terminal device, instructing the terminal device to determine a second AI / ML model using one of the existing technologies, method 1-1, method 1-2, method 2-1, method 2-2, and method 2-3.

[0246] In some embodiments, terminal device capability reporting:

[0247] First capability: Supports fine-tuning of AI / ML models

[0248] Second capability: Supports AI / ML model retraining

[0249] Figure 5b is a schematic diagram illustrating the interaction of a model recognition method according to an embodiment of the present disclosure. As shown in Figure 5b, this embodiment of the present disclosure relates to a communication method, which includes:

[0250] In step S5201, network device 102 sends the second dataset and the first identifier to terminal device 101.

[0251] In some embodiments, the first identifier indicates a first AI / ML model, which is an existing (already deployed and trained) AI / ML model of the terminal device. The first AI / ML model is trained based on a first dataset, which is sent to the terminal device by the network device before step 1. The second dataset is used by the terminal device to train a second AI / ML model.

[0252] In some embodiments, the terminal device receives a second dataset and a first identifier sent by the network device, and trains a second AI / ML model based on the second dataset and the first identifier.

[0253] It should be understood that before step S5201, the network device sends a first dataset and a first identifier to the terminal device, the terminal device uses the first dataset to train a first AI / ML model, and the first AI / ML model is associated with the first identifier.

[0254] Optionally, the terminal device trains a second AI / ML model based on the second dataset and the first identifier, including the following possible methods:

[0255] Method 1-1: The terminal device uses the second dataset to fine-tune the first AI / ML model indicated by the first identifier, and trains the second AI / ML model.

[0256] It should be understood that if a terminal device receives an identifier sent by a network device, and that identifier is already associated with an AI / ML model that the terminal device already has (already deployed and already trained), the terminal device will consider that it needs to use the dataset received along with the identifier to fine-tune the AI / ML model associated with that identifier to generate a new AI / ML model.

[0257] Optionally, the second AI / ML model is associated with the first identifier. It should be understood that the first identifier is no longer associated with the first AI / ML model, but rather with the second AI / ML model. That is, the terminal device no longer maintains the first AI / ML model, but instead maintains the second AI / ML model. If the network device sends the first identifier instructing the terminal device to pair an AI / ML model with the network device's AI / ML model to form a valid two-sided model pair, then the terminal device will use the second AI / ML model instead of the first AI / ML model.

[0258] Method 2-1: The terminal device uses the second dataset to train the second AI / ML model.

[0259] It should be understood that if a terminal device receives an identifier sent by a network device, and that identifier is already associated with an AI / ML model that the terminal device already has (already deployed and already trained), then the terminal device will consider that it needs to retrain the AI / ML model using the dataset received along with the identifier.

[0260] Optionally, the second AI / ML model is associated with the first identifier. It should be understood that "retraining" means replacing the existing AI / ML model associated with the first identifier with an AI / ML model trained using the dataset received along with the first identifier; that is, replacing the first AI / ML model with the second AI / ML model. Therefore, the second AI / ML model also needs to be associated with the first identifier, and the terminal device no longer maintains the first AI / ML model, but instead maintains the second AI / ML model. If the network device sends the first identifier instructing the terminal device to pair an AI / ML model with the network device's AI / ML model to form a valid two-sided model pair, the terminal device will use the second AI / ML model instead of the first AI / ML model.

[0261] Method 2-2: The terminal device trains a second AI / ML model using a second dataset and a first dataset used for training a first AI / ML model for the first identifier indication.

[0262] It should be understood that if a terminal device receives a first identifier sent by a network device, and the first identifier is already associated with an AI / ML model that the terminal device already has (has been deployed and trained), then the terminal device will consider that it is necessary to retrain the AI / ML model using the dataset received along with the first identifier and the dataset used to train the AI / ML model associated with the first identifier, that is, to retrain the AI / ML model using the first dataset and the second dataset.

[0263] Optionally, the second AI / ML model is associated with the first identifier. It should be understood that this is explained in Method 2-1 regarding "retraining".

[0264] Optionally, in step S5201, the network device may also send a second identifier to the terminal device. The second identifier is used to associate with the second AI / ML model, wherein the second identifier is not associated with any existing (already deployed, already trained) AI / ML model in the terminal device.

[0265] Optionally, in method 1-1, the second AI / ML model is associated with the second identifier. It should be understood that, for ease of description, this method is referred to as method 1-2. It should be understood that the difference from method 1-1 is that in method 1-2, the first identifier is still associated with the first AI / ML model, and the terminal device maintains both the first and second AI / ML models simultaneously. If the network device sends the first identifier instructing the terminal device to pair an AI / ML model with the network device's AI / ML model to form a valid bilateral model pair, the terminal device will use the first AI / ML model; if the network device sends the second identifier instructing the terminal device to pair an AI / ML model with the network device's AI / ML model to form a valid bilateral model pair, the terminal device will use the second AI / ML model.

[0266] Optionally, in method 2-2, the second AI / ML model is associated with the second identifier. It should be understood that, for ease of description, the method will be referred to as method 2-3. It should be understood that this is the same explanation as in the previous paragraph.

[0267] Optionally, in method 2-1, the second AI / ML model is associated with the second identifier. It should be understood that this is prior art; in this case, the network device does not need to send the first identifier; that is, the terminal device does not use the first identifier or any information associated with the first identifier when determining the second AI / ML model.

[0268] It should be understood that the following is a summary of the prior art, method 1-1, method 2-1, method 2-2, method 1-2, and method 2-3:

[0269] Existing technology: If the network device sends a second identifier to the terminal device instead of a first identifier, the terminal device believes that it needs to use existing technology to determine the second AI / ML model. In other words, when the network device sends the dataset and identifier to the terminal device, if the identifier is not associated with any existing (already deployed and trained) AI / ML model in the terminal device, the terminal device determines the AI / ML model according to existing technology.

[0270] Methods 1-1, 2-1, and 2-2: If the network device sends a first identifier to the terminal device but not a second identifier, the terminal device determines that one of Methods 1-1, 2-1, and 2-2 is needed to determine the second AI / ML model. In other words, when the network device sends the dataset and identifier to the terminal device, and that identifier is associated with an existing (already deployed and trained) AI / ML model on the terminal device, the terminal device determines the AI / ML model according to one of Methods 1-1, 2-1, and 2-2.

[0271] Methods 1-2 and 2-3: If the network device sends a first identifier and a second identifier to the terminal device, the terminal device considers that one of Methods 1-2 and 2-3 is needed to determine the second AI / ML model. That is, when the network device sends the dataset to the terminal device, it also sends two identifiers. One identifier is associated with an existing (deployed and trained) AI / ML model in the terminal device, and the other identifier is not associated with any existing (deployed and trained) AI / ML model in the terminal device. The terminal device then determines the AI / ML model according to one of Methods 1-2 and 2-3.

[0272] Optionally, in step S5201, the network device may also send first information to the terminal device. This first information instructs the terminal device to use one of methods 1-1, 2-1, and 2-2 to determine the second AI / ML model. It should be understood that methods 1-1, 2-1, and 2-2 cannot be implemented simultaneously, because in all these methods, the terminal device receives the second dataset and the first identifier, which is insufficient to distinguish which method is used to determine the second AI / ML model. One solution is that only one of these three methods will be included in the standard; another solution is that when sending the second dataset and the first identifier to the terminal device, the network device also needs to send an instruction message, namely the first information, instructing the terminal device to use one of methods 1-1, 2-1, and 2-2 to determine the second AI / ML model.

[0273] Optionally, in step S5201, the network device may also send second information to the terminal device. This second information instructs the terminal device to use one of methods 1-2 and 2-3 to determine the second AI / ML model. It should be understood that methods 1-2 and 2-3 cannot be implemented simultaneously, because in both methods, the terminal device receives the second dataset, the first identifier, and the second identifier, which is insufficient to distinguish which method is used to determine the second AI / ML model. One solution is that only one of these two methods will be included in the standard; another solution is that when sending the second dataset, the first identifier, and the second identifier to the terminal device, the network device also needs to send an instruction message, namely the second information, instructing the terminal device which method of method 1-2 or 2-3 should be used to determine the second AI / ML model.

[0274] Optionally, in step S5201, the network device may also send third information to the terminal device. This third information instructs the terminal device to use one of methods 1 (methods 1-1 and 1-2) and 2 (methods 2-1, 2-2, and 2-3) to determine the second AI / ML model. For example, if the network device sends a first identifier to the terminal device but not a second identifier, the terminal device believes that one of methods 1-1, 2-1, and 2-2 needs to be used to determine the second AI / ML model. The third information instructs the terminal device to use method 1-1 to determine the second AI / ML model, or to use method 2-1 or method 2-2. It should be understood that the third information cannot be used to instruct one of methods 2-1 or 2-2. One solution is that only one of methods 2-1 and 2-2 will be incorporated into the standard; another solution is that the network device also sends a fourth message to the terminal device in step S5201, which instructs the terminal device to use one of methods 2-1 and 2-2 to determine the second AI / ML model. For example, if the network device sends a first identifier and a second identifier to the terminal device, the terminal device considers it necessary to use one of methods 1-2 and 2-3 to determine the second AI / ML model, and the third message instructs the terminal device to use one of methods 1-2 and 2-3 to determine the second AI / ML model. It should be understood that Method 1 (Method 1-1 and Method 1-2) is used for fine-tuning the AI / ML model, and Method 2 (Method 2-1, Method 2-2 and Method 2-3) is used for retraining the AI / ML model. Therefore, the third information can be understood as instructing the terminal device to fine-tune the first AI / ML model indicated by the first identifier using the second dataset, or to retrain the AI / ML model indicated by the first identifier using the second dataset (and the first dataset) (or to generate a new AI / ML model); that is, the third information is used to instruct the AI / ML model to be fine-tuned or retrained.

[0275] Optionally, in step S5201, the network device may also send a fifth message to the terminal device. The fifth message is used to instruct the terminal device to determine the second AI / ML model using one of the existing technologies, methods 1-1, 1-2, 2-1, 2-2, and 2-3. Alternatively, the fifth message is used to instruct the terminal device to determine the second AI / ML model using one of the existing technologies, methods 1-1, 1-2, 2-1, 2-2, and 2-3.

[0276] Optionally, prior to step S5201, the network device may also send a first capability and / or a second capability to the terminal device. The first capability indicates that the terminal device supports method 1 (method 1-1 and method 1-2), and the second capability indicates that the terminal device supports method 2 (method 2-1, method 2-2, and method 2-3). It should be understood that the first capability indicates whether the terminal device supports AI / ML model fine-tuning, and the second capability indicates whether the terminal device supports AI / ML model retraining.

[0277] Optionally, the first capability can also instruct the terminal device to support one or more of methods 1-1 and 1-2.

[0278] Optionally, the second capability can also instruct the terminal device to support one or more of methods 2-1, 2-2, and 2-3.

[0279] This disclosure also provides an apparatus for implementing any of the above methods. For example, an apparatus is provided that includes units or modules for implementing the steps performed by the terminal device in any of the above methods. Alternatively, another apparatus is provided that includes units or modules for implementing the steps performed by the network device (e.g., access network device, core network functional node, core network device, etc.) in any of the above methods.

[0280] It should be understood that the division of units or modules in the above device is only a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, the units or modules in the device can be implemented by a processor calling software: for example, the device includes a processor connected to a memory containing instructions. The processor calls the instructions stored in the memory to implement any of the above methods or to implement the functions of the units or modules in the above device. The processor can be, for example, a general-purpose processor, such as a Central Processing Unit (CPU) or a microprocessor, and the memory can be internal or external to the device. Alternatively, the units or modules in the device can be implemented in the form of hardware circuits. The functionality of some or all of the units or modules can be achieved through the design of these hardware circuits, which can be understood as one or more processors. For example, in one implementation, the hardware circuit is an application-specific integrated circuit (ASIC). The functionality of some or all of the units or modules is achieved through the design of the logical relationships between the components within the circuit. In another implementation, the hardware circuit can be implemented using a programmable logic device (PLD). Taking a field-programmable gate array (FPGA) as an example, it can include a large number of logic gates. The connection relationships between the logic gates are configured through configuration files, thereby achieving the functionality of some or all of the units or modules. All units or modules of the above device can be implemented entirely through processor-called software, entirely through hardware circuits, or partially through processor-called software with the remaining parts implemented through hardware circuits.

[0281] In this embodiment, the processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction read and execute capabilities, such as a Central Processing Unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), or a digital signal processor (DSP). In another implementation, the processor can implement certain functions through the logical relationships of hardware circuits. The logical relationships of the aforementioned hardware circuits are fixed or reconfigurable. For example, the processor is a hardware circuit implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and configuring the hardware circuit can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units or modules. Furthermore, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as a Neural Network Processing Unit (NPU), a Tensor Processing Unit (TPU), or a Deep Learning Processing Unit (DPU).

[0282] Figure 6a is a schematic diagram of the structure of a terminal device according to an embodiment of this disclosure. As shown in Figure 6a, the terminal device 6100 may include at least one of a transceiver module 6101 and a processing module 6102. The transceiver module 6101 is used to receive first information, which includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed by both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. The processing module 6102 is used to determine a second model based on the first model and the second dataset.

[0283] In some embodiments, the processing module 6102 determines the second model based on the first model and the second dataset in the following manner: the first model is fine-tuned using the second dataset to obtain the second model.

[0284] In some embodiments, the processing module 6102 determines the second model based on the first model and the second dataset in the following manner: the first model is retrained using the second dataset to obtain the second model.

[0285] In some embodiments, the processing module 6102 retrains the first model using the second dataset to obtain the second model in the following manner: retraining the first model using only the second dataset to obtain the second model.

[0286] In some embodiments, the processing module 6102 retrains the first model using the second dataset to obtain the second model in the following manner: the first model is retrained using the second dataset and the first dataset to obtain the second model.

[0287] In some embodiments, the second model is associated with the first identifier.

[0288] In some embodiments, the processing module 6102 is further configured to: release the association between the first identifier and the first model.

[0289] In some embodiments, the first information further includes a second identifier, and the second model is associated with a second model.

[0290] In some embodiments, the first information further includes first instruction information, which is used to instruct the terminal device to fine-tune the first model using a second dataset.

[0291] In some embodiments, the first information further includes first instruction information, which is used to instruct the terminal device to retrain the first model using a second dataset.

[0292] In some embodiments, the first information further includes second instruction information, which is used to instruct the terminal device to retrain the first model using only the second dataset to obtain the second model.

[0293] In some embodiments, the first information further includes second instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset and the first dataset to obtain the second model.

[0294] In some embodiments, the first information further includes third indication information, which is used to indicate one of the following: fine-tuning the first model using the second dataset to obtain a second model; retraining the first model using only the second dataset to obtain a second model; or retraining the first model using both the second dataset and the first dataset to obtain a second model.

[0295] In some embodiments, the first information further includes fourth indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain a second model; or retraining the first model using the second dataset and the first dataset to obtain a second model.

[0296] In some embodiments, the first information further includes fifth indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain a second model, and the second model is associated with a first identifier; fine-tuning the first model using the second dataset to obtain a second model, and the second model is associated with a second identifier, the second identifier being included in the first information; retraining the first model using only the second dataset to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0297] In some embodiments, the transceiver module 6101 is further configured to: transmit a first capability and / or a second capability; wherein the first capability includes fine-tuning of the terminal device support model; wherein the second capability includes retraining of the terminal device support model.

[0298] In some embodiments, the first capability further includes the terminal device supporting at least one of the following: fine-tuning the first model using a second dataset to obtain a second model, wherein the second model is associated with a first identifier; fine-tuning the first model using a second dataset to obtain a second model, wherein the second model is associated with a second identifier, wherein the second identifier is included in the first information.

[0299] In some embodiments, the second capability further includes the terminal device supporting at least one of the following: retraining the first model using only the second dataset to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a first identifier; and retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0300] In some embodiments, the first model includes at least one of the following: an artificial intelligence (AI) model; or a machine learning (ML) model.

[0301] Figure 6b is a schematic diagram of the network device proposed in an embodiment of this disclosure. As shown in Figure 6b, the network device 6200 may include at least one of a transceiver module 6201 and a processing module 6202. The transceiver module 6201 is used to send first information, which includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed on both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. The first model and the second dataset are used to determine the second model.

[0302] In some embodiments, the second model is obtained by the terminal device by fine-tuning the first model using a second dataset.

[0303] In some embodiments, the second model is obtained by retraining the first model using a second dataset by the terminal device.

[0304] In some embodiments, the second model is obtained by the terminal device retraining the first model using only the second dataset.

[0305] In some embodiments, the second model is obtained by the terminal device retraining the first model using the second dataset and the first dataset.

[0306] In some embodiments, the second model is associated with the first identifier.

[0307] In some embodiments, the association between the first model and the first identifier has been terminated.

[0308] In some embodiments, the first information further includes a second identifier, and the second model is associated with a second model.

[0309] In some embodiments, the first information further includes first instruction information, which is used to instruct the terminal device to fine-tune the first model using a second dataset.

[0310] In some embodiments, the first information further includes first instruction information, which is used to instruct the terminal device to retrain the first model using a second dataset.

[0311] In some embodiments, the first information further includes second instruction information, which is used to instruct the terminal device to retrain the first model using only the second dataset to obtain the second model.

[0312] In some embodiments, the first information further includes second instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset and the first dataset to obtain the second model.

[0313] In some embodiments, the first information further includes third indication information, which is used to indicate one of the following: fine-tuning the first model using the second dataset to obtain a second model; retraining the first model using only the second dataset to obtain a second model; or retraining the first model using both the second dataset and the first dataset to obtain a second model.

[0314] In some embodiments, the first information further includes fourth indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain a second model; or retraining the first model using the second dataset and the first dataset to obtain a second model.

[0315] In some embodiments, the first information further includes fifth indication information, which indicates one of the following: fine-tuning the first model using the second dataset to obtain a second model, and the second model is associated with a first identifier; fine-tuning the first model using the second dataset to obtain a second model, and the second model is associated with a second identifier, the second identifier being included in the first information; retraining the first model using only the second dataset to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0316] In some embodiments, the transceiver module 6201 is further configured to: receive a first capability and / or a second capability; wherein the first capability includes fine-tuning of the terminal device support model; wherein the second capability includes retraining of the terminal device support model.

[0317] In some embodiments, the first capability further includes the terminal device supporting at least one of the following: fine-tuning the first model using a second dataset to obtain a second model, wherein the second model is associated with a first identifier; fine-tuning the first model using a second dataset to obtain a second model, wherein the second model is associated with a second identifier, wherein the second identifier is included in the first information.

[0318] In some embodiments, the second capability further includes the terminal device supporting at least one of the following: retraining the first model using only the second dataset to obtain a second model, and the second model being associated with a first identifier; retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a first identifier; and retraining the first model using both the second and first datasets to obtain a second model, and the second model being associated with a second identifier, the second identifier being included in the first information.

[0319] In some embodiments, the first model includes at least one of the following: an artificial intelligence (AI) model; or a machine learning (ML) model.

[0320] Figure 7a is a schematic diagram of a communication device according to an embodiment of this disclosure. The communication device 7100 can be a network device, a terminal device, or a chip, chip system, or processor that supports the implementation of any of the above methods in a network device, or a chip, chip system, or processor that supports the implementation of any of the above methods in a terminal device. Optionally, the network device can be an access network device, a core network device, etc. Optionally, the terminal device can be a user equipment, etc. The communication device 7100 can be used to implement the methods described in the above method embodiments; for details, please refer to the descriptions in the above method embodiments.

[0321] As shown in Figure 7a, the communication device 7100 includes one or more processors 7101. The processor 7101 can be a general-purpose processor or a dedicated processor, such as a baseband processor or a central processing unit (CPU). The baseband processor can be used to process communication protocols and communication data, while the CPU can be used to control the communication device, execute programs, and process program data. The communication device 7100 is used to execute any of the above methods. Optionally, the communication device can be a base station, a baseband chip, a terminal device, a terminal device chip, a DU (Distributed Unit), or a CU (Computer Integrated Circuit), etc.

[0322] In some embodiments, the communication device 7100 further includes one or more memories 7102 for storing instructions. Optionally, all or part of the memories 7102 may also be located outside the communication device 7100.

[0323] In some embodiments, the communication device 7100 further includes one or more transceivers 7103. When the communication device 7100 includes one or more transceivers 7103, the transceivers 7103 perform communication steps such as sending and / or receiving in the above-described method, such as steps S2101 and S2102, but are not limited thereto. The processor 7201 performs other steps, such as steps S2103 and S2104, but is not limited thereto.

[0324] In some embodiments, a transceiver may include a receiver and / or a transmitter, which may be separate or integrated. Optionally, the terms transceiver, transceiver unit, transceiver, transceiver circuit, etc., may be used interchangeably; the terms transmitter, transmitting unit, transmitter, transmitting circuit, etc., may be used interchangeably; and the terms receiver, receiving unit, receiver, receiving circuit, etc., may be used interchangeably.

[0325] In some embodiments, the communication device 7100 may include one or more interface circuits 7104. Optionally, the interface circuit 7104 is connected to the memory 7102, and the interface circuit 7104 can be used to receive signals from the memory 7102 or other devices, and can be used to send signals to the memory 7102 or other devices. For example, the interface circuit 7104 can read instructions stored in the memory 7102 and send the instructions to the processor 7101.

[0326] The communication device 7100 described in the above embodiments may be a network device or a terminal device, but the scope of the communication device 7100 described in this disclosure is not limited thereto, and the structure of the communication device 7100 may not be limited by FIG. 7a. The communication device may be a standalone device or a part of a larger device. For example, the communication device may be: (1) a standalone integrated circuit IC, or chip, or chip system or subsystem; (2) a collection of one or more ICs, optionally, the IC collection may also include storage components for storing data and programs; (3) an ASIC, such as a modem; (4) a module that can be embedded in other devices; (5) a receiver, terminal device, smart terminal device, cellular phone, wireless device, handheld device, mobile unit, vehicle device, network device, cloud device, artificial intelligence device, etc.; (6) others, etc.

[0327] Figure 7b is a schematic diagram of the chip structure proposed in an embodiment of this disclosure. For cases where the communication device 7100 can be a chip or a chip system, please refer to the schematic diagram of the chip 7200 shown in Figure 7b, but it is not limited thereto.

[0328] Chip 7200 includes one or more processors 7201, which are used to perform any of the above methods.

[0329] In some embodiments, chip 7200 further includes one or more interface circuits 7202. Optionally, the interface circuit 7202 is connected to memory 7203, and the interface circuit 7202 can be used to receive signals from memory 7203 or other devices, and the interface circuit 7202 can be used to send signals to memory 7203 or other devices. For example, the interface circuit 7202 can read instructions stored in memory 7203 and send the instructions to processor 7201.

[0330] In some embodiments, the interface circuit 7202 performs communication steps such as sending and / or receiving in the above method, such as steps S2101 and S2102, but is not limited thereto. The processor 7201 performs other steps, such as steps S2103 and S2104, but is not limited thereto.

[0331] In some embodiments, the terms interface circuit, interface, transceiver pin, transceiver, etc., can be used interchangeably.

[0332] In some embodiments, chip 7200 further includes one or more memories 7203 for storing instructions. Optionally, all or part of the memories 7203 may be located outside of chip 7200.

[0333] This disclosure also proposes a storage medium storing instructions that, when executed on the communication device 7100, cause the communication device 7100 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 not limited thereto; it may also be a storage medium readable by other devices. Optionally, the storage medium may be a non-transitory storage medium, but not limited thereto; it may also be a temporary storage medium.

[0334] This disclosure also provides a program product that, when executed by the communication device 7100, causes the communication device 7100 to perform any of the above methods. Optionally, the program product is a computer program product.

[0335] This disclosure also proposes a computer program that, when run on a computer, causes the computer to perform any of the above methods.

Claims

1. A communication method, characterized in that, The method includes: The terminal device receives first information, which includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed by both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. The terminal device determines the second model based on the first model and the second dataset.

2. The method according to claim 1, characterized in that, The terminal device determines a second model based on the first model and the second dataset, including: The terminal device uses the second dataset to fine-tune the first model to obtain the second model.

3. The method according to claim 1, characterized in that, The terminal device determines a second model based on the first model and the second dataset, including: The terminal device uses the second dataset to retrain the first model to obtain the second model.

4. The method according to claim 3, characterized in that, The terminal device retrains the first model using the second dataset to obtain the second model, including: The terminal device retrains the first model using only the second dataset to obtain the second model.

5. The method according to claim 3, characterized in that, The terminal device retrains the first model using the second dataset to obtain the second model, including: The terminal device uses the second dataset and the first dataset to retrain the first model to obtain the second model.

6. The method according to any one of claims 1-5, characterized in that, The second model is associated with the first identifier.

7. The method according to claim 6, characterized in that, The method further includes: The terminal device disconnects the association between the first identifier and the first model.

8. The method according to claim 1, 2, or 5, characterized in that, The first information also includes a second identifier, and the second model is associated with the second model.

9. The method according to claim 2, characterized in that, The first information also includes first instruction information, which is used to instruct the terminal device to fine-tune the first model using the second dataset.

10. The method according to claim 3, characterized in that, The first information also includes first instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset.

11. The method according to claim 4, characterized in that, The first information also includes second instruction information, which is used to instruct the terminal device to retrain the first model using only the second dataset to obtain the second model.

12. The method according to claim 5, characterized in that, The first information also includes second instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset and the first dataset to obtain the second model.

13. The method according to claim 6, characterized in that, The first information also includes third indication information, which is used to indicate one of the following: The first model is fine-tuned using the second dataset to obtain the second model; The second model is obtained by retraining the first model using only the second dataset; The first model is retrained using the second dataset and the first dataset to obtain the second model.

14. The method according to claim 8, characterized in that, The first information also includes fourth indication information, which indicates one of the following: The first model is fine-tuned using the second dataset to obtain the second model; The first model is retrained using the second dataset and the first dataset to obtain the second model.

15. The method according to claim 1, characterized in that, The first information also includes a fifth indication, which indicates one of the following: The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information; The first model is retrained using only the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information.

16. The method according to any one of claims 1-15, characterized in that, The method further includes: The terminal device transmits the first capability and / or the second capability; The first capability includes fine-tuning of the terminal device support model; The second capability includes the terminal device supporting the retraining of the model.

17. The method according to claim 16, characterized in that, The first capability also includes the terminal device supporting at least one of the following: The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information.

18. The method according to claim 16, characterized in that, The second capability also includes that the terminal device supports at least one of the following: The first model is retrained using only the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information.

19. The method according to any one of claims 1-18, characterized in that, The first model includes at least one of the following: Artificial intelligence (AI) models; Machine learning (ML) models.

20. A communication method, characterized in that, The method includes: The network device sends first information, which includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed by both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. The first model and the second dataset are used to determine the second model.

21. The method according to claim 20, characterized in that, The second model is obtained by the terminal device by fine-tuning the first model using the second dataset.

22. The method according to claim 20, characterized in that, The second model is obtained by the terminal device retraining the first model using the second dataset.

23. The method according to claim 22, characterized in that, The second model is obtained by the terminal device retraining the first model using only the second dataset.

24. The method according to claim 22, characterized in that, The second model is obtained by the terminal device retraining the first model using the second dataset and the first dataset.

25. The method according to any one of claims 20-24, characterized in that, The second model is associated with the first identifier.

26. The method according to claim 25, characterized in that, The association between the first model and the first identifier has been severed.

27. The method according to claim 20, 21, or 24, characterized in that, The first information also includes a second identifier, and the second model is associated with the second model.

28. The method according to claim 21, characterized in that, The first information also includes first instruction information, which is used to instruct the terminal device to fine-tune the first model using the second dataset.

29. The method according to claim 22, characterized in that, The first information also includes first instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset.

30. The method according to claim 23, characterized in that, The first information also includes second instruction information, which is used to instruct the terminal device to retrain the first model using only the second dataset to obtain the second model.

31. The method according to claim 24, characterized in that, The first information also includes second instruction information, which is used to instruct the terminal device to retrain the first model using the second dataset and the first dataset to obtain the second model.

32. The method according to claim 25, characterized in that, The first information also includes third indication information, which is used to indicate one of the following: The first model is fine-tuned using the second dataset to obtain the second model; The second model is obtained by retraining the first model using only the second dataset; The first model is retrained using the second dataset and the first dataset to obtain the second model.

33. The method according to claim 27, characterized in that, The first information also includes fourth indication information, which indicates one of the following: The first model is fine-tuned using the second dataset to obtain the second model; The first model is retrained using the second dataset and the first dataset to obtain the second model.

34. The method according to claim 20, characterized in that, The first information also includes a fifth indication, which indicates one of the following: The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information; The first model is retrained using only the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information.

35. The method according to any one of claims 20-34, characterized in that, The method further includes: The network device receives a first capability and / or a second capability; The first capability includes fine-tuning of the terminal device support model; The second capability includes the terminal device supporting the retraining of the model.

36. The method according to claim 35, characterized in that, The first capability also includes the terminal device supporting at least one of the following: The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is fine-tuned using the second dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information.

37. The method according to claim 35, characterized in that, The second capability also includes that the terminal device supports at least one of the following: The first model is retrained using only the second dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with the first identifier; The first model is retrained using the second dataset and the first dataset to obtain the second model, and the second model is associated with a second identifier, which is included in the first information.

38. The method according to any one of claims 20-37, characterized in that, The first model includes at least one of the following: Artificial intelligence (AI) models; Machine learning (ML) models.

39. A terminal device, characterized in that, include: The transceiver module is used to receive first information, which includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed on both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. A processing module is used to determine a second model based on the first model and the second dataset.

40. A network device, characterized in that, include: The transceiver module is used to send first information, which includes a first identifier and a second dataset. The first identifier is associated with a first model, which is a model deployed on both the terminal device and the network device. The first model is trained based on the first dataset, and the second dataset is different from the first dataset. The first model and the second dataset are used to determine the second model.

41. A terminal device, characterized in that, include: One or more processors; The processor is used to execute the communication method according to any one of claims 1-19.

42. A network device, characterized in that, include: One or more processors; The processor is used to execute the communication method according to any one of claims 20-38.

43. A communication system, characterized in that, include: A terminal device and a network device, wherein the terminal device is configured to implement the communication method of any one of claims 1-19, and the network device is configured to implement the communication method of any one of claims 20-38.

44. A storage medium, characterized in that, include: The storage medium stores instructions that, when executed on a communication device, cause the communication device to perform the communication method as described in any one of claims 1-19 or 20-38.

45. A program product, characterized in that, include: A computer program, when executed by a communication device, causes the communication device to perform the communication method as described in any one of claims 1-19 or 20-38.

Images