Task processing method, first network element, terminal, access network device, and communication system

Abstract

The present disclosure relates to a task processing method, a first network element, a terminal, an access network device, and a communication system. The method comprises: a first network element acquiring an artificial intelligence (AI) task to be processed; segmenting the AI task into a plurality of AI sub-tasks; and scheduling a terminal and an access network device to jointly process the plurality of AI sub-tasks to obtain a processing result of the AI task. By means of the embodiments of the present disclosure, computing resources of each network node can be fully utilized, and flexible distribution for computation can be achieved by means of task scheduling by a core network element.

Description

Task processing method, first network element, terminal, access network equipment and communication system Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a task processing method, a first network element, a terminal, an access network device, and a communication system. Background Art

[0002] Artificial Intelligence (AI) and Machine Learning (ML) are widely used in mobile communication systems and networks. For example, in 5G communications, AI or ML uses the Network Data Analysis Function (NWDAF) as a plug-in to perform network analysis, providing AI algorithms as patching tools.

[0003] Summary of the Invention

[0004] When it comes to processing AI tasks, there are problems such as limited application scenarios, complex data transmission, and low utilization of computing resources.

[0005] The embodiments of the present disclosure provide a task processing method, a first network element, a terminal, an access network device, and a communication system.

[0006] According to a first aspect of an embodiment of the present disclosure, a task processing method is proposed, comprising: a first network element obtaining an artificial intelligence (AI) task to be processed; dividing the AI ​​task into multiple AI subtasks; and scheduling a terminal and an access network device to jointly process the multiple AI subtasks to obtain a processing result of the AI ​​task.

[0007] According to a second aspect of an embodiment of the present disclosure, a task processing method is proposed, comprising: a terminal processing a first artificial intelligence (AI) subtask based on scheduling by a first network element; wherein the first AI subtask is a subtask processed by the terminal from a plurality of subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0008] According to a third aspect of an embodiment of the present disclosure, a task processing method is proposed, comprising: an access network device processing a second artificial intelligence (AI) subtask based on scheduling by a first network element; wherein the second AI subtask is a subtask processed by the access network device from among multiple subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0009] According to the fourth aspect of the embodiment of the present disclosure, a task processing method is proposed, which includes: a first network element obtains an artificial intelligence (AI) task to be processed; divides the AI ​​task into multiple AI subtasks; the first network element dispatches a terminal and an access network device to process the multiple AI subtasks together with the first network element to obtain a processing result of the AI ​​task.

[0010] According to the fifth aspect of the embodiment of the present disclosure, a first network element is proposed, including: a transceiver module, used to obtain an artificial intelligence (AI) task to be processed; a processing module, used to divide the AI ​​task into multiple AI subtasks; the processing module is also used to schedule terminals and access network devices to jointly process the multiple AI subtasks to obtain the processing results of the AI ​​task.

[0011] According to a sixth aspect of an embodiment of the present disclosure, a terminal is proposed, comprising: a processing module, configured to process a first artificial intelligence (AI) subtask based on scheduling by a first network element; wherein the first AI subtask is a subtask processed by the terminal from among a plurality of subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0012] According to a seventh aspect of an embodiment of the present disclosure, an access network device is proposed, comprising: a processing module for processing a second artificial intelligence (AI) subtask based on the scheduling of a first network element; wherein the second AI subtask is a subtask processed by the access network device from among multiple subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0013] According to an eighth aspect of an embodiment of the present disclosure, a core network device is proposed, comprising: one or more processors; wherein the processor is used to execute the task processing method described in any one of the first aspects.

[0014] According to a ninth aspect of an embodiment of the present disclosure, a terminal is proposed, comprising: one or more processors; wherein the processor is used to execute the task processing method described in any one of the second aspects.

[0015] According to a tenth aspect of an embodiment of the present disclosure, a terminal is proposed, comprising: one or more processors; wherein the processor is used to execute the task processing method described in any one of the third aspects.

[0016] According to the eleventh aspect of an embodiment of the present disclosure, a communication system is proposed, including: a core network device, a terminal and a network device, wherein the core network device is configured to implement the task processing method described in any one of the first aspects, the terminal is configured to implement the task processing method described in any one of the second aspects, and the network device is configured to implement the task processing method described in any one of the third aspects.

[0017] According to the twelfth aspect of an embodiment of the present disclosure, a storage medium is proposed, which stores instructions. When the instructions are executed on a communication device, the communication device executes a task processing method as described in any one of the first aspect, the second aspect, or the third aspect.

[0018] Through the embodiments of the present disclosure, the computing resources of each network node can be fully utilized, and flexible computing allocation can be achieved through task scheduling of core network elements. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following drawings required for describing the embodiments are introduced. The following drawings are merely some embodiments of the present disclosure and do not impose specific limitations on the protection scope of the present disclosure.

[0020] FIG1A is a schematic diagram showing the architecture of a communication system according to an embodiment of the present disclosure.

[0021] FIG1B is a schematic diagram of a 5G communication system architecture according to an embodiment of the present disclosure.

[0022] FIG1C is a schematic diagram of a data storage architecture for storing unstructured data from any NF according to an embodiment of the present disclosure.

[0023] FIG1D is a schematic diagram of a data storage architecture according to an embodiment of the present disclosure.

[0024] FIG1E is a schematic diagram of an architecture for collecting data from any 5G cloud-native network function, shown in an embodiment of the present disclosure.

[0025] FIG2A is an interactive schematic diagram illustrating a task processing method according to an embodiment of the present disclosure.

[0026] FIG2B is a schematic diagram illustrating AI task segmentation according to an embodiment of the present disclosure.

[0027] FIG3A is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0028] FIG3B is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0029] FIG3C is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0030] FIG4A is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0031] FIG4B is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0032] FIG4C is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0033] FIG5A is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0034] FIG5B is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0035] FIG5C is a flowchart illustrating a task processing method according to an embodiment of the present disclosure.

[0036] FIG6 is an interactive diagram illustrating a task processing method according to an embodiment of the present disclosure.

[0037] FIG7 is an interactive schematic diagram illustrating a task processing method according to an embodiment of the present disclosure.

[0038] FIG8A is a schematic structural diagram of a first network element proposed in an embodiment of the present disclosure.

[0039] FIG8B is a schematic structural diagram of a terminal proposed in an embodiment of the present disclosure.

[0040] FIG8C is a schematic structural diagram of an access network device proposed in an embodiment of the present disclosure.

[0041] FIG9A is a schematic structural diagram of a communication device proposed in an embodiment of the present disclosure.

[0042] FIG9B is a schematic diagram of the structure of the chip proposed in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0043] The embodiments of the present disclosure provide a task processing method, a first network element, a terminal, an access network device, and a communication system.

[0044] In a first aspect, an embodiment of the present disclosure proposes a task processing method, which includes: a first network element obtains an artificial intelligence (AI) task to be processed; divides the AI ​​task into multiple AI subtasks; and schedules a terminal and an access network device to jointly process the multiple AI subtasks to obtain a processing result of the AI ​​task.

[0045] In the above embodiment, the first network element divides the AI ​​task into multiple subtasks and schedules the terminal and access network device to jointly process the multiple AI subtasks, thereby obtaining the processing results of the AI ​​task. Compared with a single device processing the AI ​​task and then performing transfer learning, this can greatly improve computing resource utilization and thus improve computing efficiency.

[0046] In combination with some embodiments of the first aspect, in some embodiments, the dividing the AI ​​task into multiple AI subtasks includes: sending a capability request to a second network element, and obtaining first capability information sent by the second network element in response to the capability request, the first capability information being the inherent and unchanging information of the terminal and the inherent and unchanging capability information of the access network device; obtaining second capability information sent by the access network device, the second capability information including the second capability information of the access network device and the second capability information of the terminal periodically sent to the access network device, the second capability information being the capability information of the terminal that changes in real time and the capability information of the access network device that changes in real time; based on the first capability information and the second capability information, dividing the AI ​​task into multiple AI subtasks.

[0047] In the above embodiment, by acquiring the inherent capabilities and real-time changing capabilities of the terminal and access network equipment, and then allocating AI subtasks based on the inherent capabilities and real-time changing capabilities, the AI ​​subtasks allocated to the terminal and access network equipment can be matched with their respective current computing capabilities, thereby improving computing efficiency.

[0048] In combination with some embodiments of the first aspect, in some embodiments, the AI ​​task is divided into multiple AI subtasks based on the first capability information and the second capability information, including: based on the first capability information corresponding to the terminal and the second capability information corresponding to the terminal, dividing the AI ​​task to obtain the first AI subtask to be processed by the terminal; based on the first capability information corresponding to the access network device and the second capability information corresponding to the access network device, dividing the AI ​​task to obtain the second AI subtask to be processed by the access network device; and using the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask as the third AI subtask to be processed by the first network element.

[0049] In combination with some embodiments of the first aspect, in some embodiments, the scheduling terminal and the access network device jointly process the multiple AI subtasks to obtain the processing results of the AI ​​tasks, including: obtaining the second AI subtask processing result sent by the access network device, the second AI subtask processing result is the result of the access network device processing the second AI subtask based on the first AI subtask processing result, and the first AI subtask processing result is the result of the terminal processing the first AI subtask and sending it to the access network device; processing the third AI subtask based on the second AI subtask processing result to obtain the processing result of the AI ​​task.

[0050] In combination with some embodiments of the first aspect, in some embodiments, the method further includes: periodically obtaining a status report sent by a second network element, wherein the status report includes second capability information of the access network device and the terminal; when the second capability information meets a preset condition, the AI ​​task is re-divided to obtain a plurality of re-divided AI subtasks.

[0051] In the above embodiment, by periodically obtaining loading reports, the real-time changing capabilities of the terminal and access network equipment can be periodically obtained, and then the imagined tasks can be adjusted based on their real-time changing capabilities to avoid unreasonable utilization of computing resources due to changes in capabilities.

[0052] In combination with some embodiments of the first aspect, in some embodiments, after the first network element, the terminal and the access network device jointly process the multiple AI subtasks, it also includes: obtaining the processing results of the AI ​​tasks and sending them to the second network element, and the second network element is used to store the processing results.

[0053] In combination with some embodiments of the first aspect, in some embodiments, the first capability information includes at least one of the following: the computing capability of the terminal; the resource information of the terminal; the computing capability of the access network device; and the resource information of the access network device.

[0054] In combination with some embodiments of the first aspect, in some embodiments, the second capability information includes at least one of the following: the real-time load of the terminal; the resource occupancy information of the terminal; the real-time load of the access network device; and the resource occupancy information of the access network device.

[0055] In a second aspect, an embodiment of the present disclosure proposes a task processing method, which includes: the terminal processes a first artificial intelligence (AI) subtask based on the scheduling of the first network element; wherein the first AI subtask is a subtask processed by the terminal among multiple subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0056] In the above embodiment, the first network element divides the AI ​​task into multiple subtasks and schedules the terminal and access network device to jointly process the multiple AI subtasks, thereby obtaining the processing results of the AI ​​task. Compared with a single device processing the AI ​​task and then performing transfer learning, this can greatly improve computing resource utilization and thus improve computing efficiency.

[0057] In combination with some embodiments of the second aspect, in some embodiments, the method further includes: sending first capability information to a second network element, where the first capability information is inherent and unchanging information of the terminal; periodically sending second capability information of the terminal to an access network device, where the second capability information of the terminal is capability information that changes in real time; the first capability information and the second capability information are used by the first network element to divide the AI ​​task into multiple AI subtasks.

[0058] In combination with some embodiments of the second aspect, in some embodiments, the multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; wherein, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the terminal; the second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the access network device; the third AI subtask is a subtask processed by the first network element, and is the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask.

[0059] In combination with some embodiments of the second aspect, in some embodiments, the method further includes: sending the processing result of the first AI subtask to the access network device; wherein, the processing result of the first AI subtask is used by the access network device to process the second AI subtask to obtain the processing result of the second AI subtask, and the processing result of the second AI subtask is used by the first network element to process the third AI subtask to obtain the processing result of the AI ​​task.

[0060] In combination with some embodiments of the second aspect, in some embodiments, the first capability information includes at least one of the following: the computing capability of the terminal; the resource information of the terminal; the computing capability of the access network device; and the resource information of the access network device.

[0061] In combination with some embodiments of the second aspect, in some embodiments, the second capability information includes at least one of the following: the real-time load of the terminal; the resource occupancy information of the terminal; the real-time load of the access network device; and the resource occupancy information of the access network device.

[0062] In a third aspect, an embodiment of the present disclosure proposes a task processing method, which includes: an access network device processes a second artificial intelligence (AI) subtask based on the scheduling of the first network element; wherein the second AI subtask is a subtask processed by the access network device among multiple subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0063] In the above embodiment, the first network element divides the AI ​​task into multiple subtasks and schedules the terminal and access network device to jointly process the multiple AI subtasks, thereby obtaining the processing results of the AI ​​task. Compared with a single device processing the AI ​​task and then performing transfer learning, this can greatly improve computing resource utilization and thus improve computing efficiency.

[0064] In combination with some embodiments of the third aspect, in some embodiments, the method further includes: sending first capability information to the second network element, the first capability information being inherent and unchanging information of the access network device; periodically obtaining second capability information of the terminal sent by the terminal, and sending the second capability information of the terminal and the second capability information of the access network device to the first network element; the second capability information of the access network device is capability information that changes in real time of the access network device, and the second capability information of the terminal is capability information that changes in real time of the terminal; the first capability information and the second capability information are used by the first network element to divide the AI ​​task into multiple AI subtasks.

[0065] In combination with some embodiments of the third aspect, in some embodiments, the multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; wherein, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the terminal; the second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the access network device; the third AI subtask is a subtask processed by the first network element, and is the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask.

[0066] In combination with some embodiments of the third aspect, in some embodiments, the method further includes: obtaining the processing result of the first AI subtask sent by the terminal; processing the second AI subtask based on the processing result of the first AI subtask, obtaining the processing result of the second AI subtask, and sending the processing result of the second AI subtask to the first network element; wherein the processing result of the second AI subtask is used by the first network element to process the third AI subtask to obtain the processing result of the AI ​​task.

[0067] In combination with some embodiments of the third aspect, in some embodiments, the first capability information includes at least one of the following: the computing capability of the terminal; the resource information of the terminal; the computing capability of the access network device; and the resource information of the access network device.

[0068] In combination with some embodiments of the third aspect, in some embodiments, the second capability information includes at least one of the following: the real-time load of the terminal; the resource occupancy information of the terminal; the real-time load of the access network device; and the resource occupancy information of the access network device.

[0069] In a fourth aspect, an embodiment of the present disclosure proposes a task processing method, which includes: a first network element obtains an artificial intelligence (AI) task to be processed; dividing the AI ​​task into multiple AI subtasks; and the terminal and access network equipment, based on the scheduling of the first network element, jointly process the multiple AI subtasks with the first network element to obtain the processing result of the AI ​​task.

[0070] In the fifth aspect, the embodiment of the present disclosure proposes a first network element, including: a transceiver module, used to obtain the artificial intelligence AI task to be processed; a processing module, used to divide the AI ​​task into multiple AI subtasks; the processing module is also used to schedule terminals and access network equipment to jointly process the multiple AI subtasks to obtain the processing results of the AI ​​task.

[0071] In a sixth aspect, an embodiment of the present disclosure proposes a terminal, comprising: a processing module for processing a first artificial intelligence (AI) subtask based on the scheduling of the first network element; wherein the first AI subtask is a subtask processed by the terminal from among multiple subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0072] In the seventh aspect, an embodiment of the present disclosure proposes an access network device, including: a processing module for processing a second artificial intelligence (AI) subtask based on the scheduling of the first network element; wherein the second AI subtask is a subtask processed by the access network device among multiple subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0073] In an eighth aspect, an embodiment of the present disclosure proposes a core network device, comprising: one or more processors; wherein the processor is used to execute the task processing method described in any one of the first aspects.

[0074] In a ninth aspect, an embodiment of the present disclosure proposes a terminal, which includes: one or more processors; the processor is used to execute the task processing method described in any one of the second aspects.

[0075] In the tenth aspect, an embodiment of the present disclosure proposes an access network device, comprising: one or more processors; wherein the processor is used to execute the task processing method described in any one of the third aspects.

[0076] In the eleventh aspect, an embodiment of the present disclosure proposes a communication system, comprising: a core network device, a terminal and a network device, wherein the core network device is configured to implement the task processing method described in any one of the first aspects, the terminal is configured to implement the task processing method described in any one of the second aspects, and the network device is configured to implement the task processing method described in any one of the third aspects.

[0077] In the twelfth aspect, an embodiment of the present disclosure proposes a storage medium storing instructions. When the instructions are executed on a communication device, the communication device executes a task processing method as described in any one of the first aspect, the second aspect, or the third aspect.

[0078] It is understandable that the above-mentioned terminal, access network device, first network element, second network element, core network device, communication system, storage medium, program product, computer program, chip or chip system are all used to perform the method proposed in the embodiment of the present disclosure. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects of the corresponding method and will not be repeated here.

[0079] The present disclosure provides a task processing method, a first network element, a terminal, an access network device, and a communication system. In some embodiments, the terms "task processing method" and "information processing method" and "communication method" are interchangeable; the terms "task processing device" and "information processing device" and "communication device" are interchangeable; and the terms "information processing system" and "communication system" are interchangeable.

[0080] The embodiments of the present disclosure are not exhaustive and are merely illustrative of some embodiments, and are not intended to be a specific limitation on the scope of protection of the present disclosure. In the absence of contradiction, each step in a certain embodiment can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a certain embodiment can also be implemented as an independent embodiment, and the order of the steps in a certain embodiment can be arbitrarily exchanged. In addition, the optional implementation methods in a certain embodiment can be arbitrarily combined; in addition, the embodiments can be arbitrarily combined. For example, some or all steps of different embodiments can be arbitrarily combined, and a certain embodiment can be arbitrarily combined with the optional implementation methods of other embodiments.

[0081] In each embodiment of the present disclosure, unless otherwise specified or provided for by logic, the terms and / or descriptions between the embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form a new embodiment based on their inherent logical relationships.

[0082] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure.

[0083] In the embodiments of the present disclosure, unless otherwise specified, elements expressed in the singular, such as "a", "an", "the", "above", "said", "the", "the", etc., may mean "one and only one", or "one or more", "at least one", etc. For example, when using articles such as "a", "an", "the" in English in translation, the noun following the article may be understood as a singular expression or a plural expression.

[0084] In the embodiments of the present disclosure, “plurality” refers to two or more.

[0085] In some embodiments, the terms "at least one," "one or more," "a plurality of," "multiple," etc. may be used interchangeably.

[0086] In some embodiments, descriptions such as "at least one of A and B," "A and / or B," "A in one case, B in another case," or "in response to one case A, in response to another case B" may include the following technical solutions depending on the situation: in some embodiments, A (A is executed independently of B); in some embodiments, B (B is executed independently of A); in some embodiments, execution is selected from A and B (A and B are selectively executed); and in some embodiments, A and B (both A and B are executed). The above is also applicable when there are more branches such as A, B, and C.

[0087] In some embodiments, "A or B" and other descriptions may include the following technical solutions depending on the situation: in some embodiments, A (A is executed independently of B); in some embodiments, B (B is executed independently of A); in some embodiments, execution is selected from A and B (A and B are selectively executed). The above is also applicable when there are more branches such as A, B, C, etc.

[0088] The prefixes such as "first" and "second" in the embodiments of the present disclosure are only used to distinguish different description objects and do not constitute any restriction on the position, order, priority, quantity or content of the description objects. For the statement of the description object, please refer to the description in the context of the claims or embodiments, and no unnecessary restriction should be constituted due to the use of prefixes. For example, if the description object is a "field", the ordinal number before the "field" in the "first field" and the "second field" does not limit the position or order between the "fields". "First" and "second" do not limit whether the "fields" they modify are in the same message, nor do they limit the order of the "first field" and the "second field". For another example, if the description object is a "level", the ordinal number before the "level" in the "first level" and the "second level" does not limit the priority between the "levels". For another example, the number of description objects is not limited by the ordinal number and can be one or more. Taking "first device" as an example, the number of "devices" can be one or more. In addition, the objects modified by different prefixes can be the same or different. For example, if the description object is "device", then the "first device" and the "second device" can be the same device or different devices, and their types can be the same or different; for example, if the description object is "information", then the "first information" and the "second capability information" can be the same information or different information, and their contents can be the same or different.

[0089] In some embodiments, “including A,” “comprising A,” “used to indicate A,” and “carrying A” can be interpreted as directly carrying A or indirectly indicating A.

[0090] In some embodiments, terms such as "time / frequency" and "time / frequency domain" refer to the time domain and / or the frequency domain.

[0091] In some embodiments, terms such as "in response to...", "in response to determining...", "in the case of...", "at the time of...", "when...", "if...", "if...", etc. can be used interchangeably.

[0092] In some embodiments, terms such as "greater than", "greater than or equal to", "not less than", "more than", "more than or equal to", "not less than", "higher than", "higher than or equal to", "not less than", and "above" can be replaced with each other, and terms such as "less than", "less than or equal to", "not greater than", "less than", "less than or equal to", "not more than", "lower than", "lower than or equal to", "not higher than", and "below" can be replaced with each other.

[0093] In some embodiments, devices, etc. can be interpreted as physical or virtual, and their names are not limited to the names recorded in the embodiments. Terms such as "device", "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "chip", "chip system", "entity", and "subject" can be used interchangeably.

[0094] In some embodiments, "network" can be interpreted as devices included in the network (eg, access network equipment, core network equipment, etc.).

[0095] In some embodiments, the terms "access network device (AN device)", "radio access network device (RAN device)", "base station (BS)", "radio base station" "fixed station", "node", "access point", "transmission point (TP)", "reception point (RP)", "transmission / reception point (TRP)" "panel", "antenna panel", "antenna array", "cell", "macro cell", "small cell", "femto cell", "pico cell", "sector", "cell group", "serving cell", "carrier", "component carrier", "bandwidth part (BWP)" and the like may be used interchangeably.

[0096] In some embodiments, the terms "terminal", "terminal device", "user equipment (UE)", "user terminal", "mobile station (MS)", "mobile terminal (MT)", subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, etc. can be used interchangeably.

[0097] In some embodiments, the access network device, the core network device, or the network device can be replaced by a terminal. For example, the various embodiments of the present disclosure can also be applied to a structure in which the communication between the access network device, the core network device, or the network device and the terminal is replaced by communication between multiple terminals (for example, device-to-device (D2D), vehicle-to-everything (V2X), etc.). In this case, it is also possible to set the structure in which the terminal has all or part of the functions of the access network device. In addition, terms such as "uplink" and "downlink" can also be replaced by terms corresponding to communication between terminals (for example, "side"). For example, uplink channels, downlink channels, etc. can be replaced by side channels, and uplinks, downlinks, etc. can be replaced by side links.

[0098] In some embodiments, the terminal may be replaced by an access network device, a core network device, or a network device. In this case, the access network device, the core network device, or the network device may have a structure that has all or part of the functions of the terminal.

[0099] In some embodiments, obtaining data, information, etc. may comply with the laws and regulations of the country where the data is obtained.

[0100] In some embodiments, data, information, etc. may be obtained with the user's consent.

[0101] In addition, each element, each row, or each column in the table of the embodiment of the present disclosure can be implemented as an independent embodiment, and the combination of any elements, any rows, and any columns can also be implemented as an independent embodiment.

[0102] FIG1A is a schematic diagram showing the architecture of a communication system according to an embodiment of the present disclosure.

[0103] As shown in FIG1A , a communication system 100 includes a terminal 101 , an access network device 102 , and a core network device 103 .

[0104] In some embodiments, the terminal 101 includes, for example, a mobile phone, a wearable device, an Internet of Things device, a car with communication function, a smart car, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical surgery, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, and at least one of a wireless terminal device in a smart home, but is not limited thereto.

[0105] In some embodiments, the access network device 102 is, for example, a node or device that accesses a terminal to a wireless network. The access network device may include an evolved NodeB (eNB), a next generation evolved NodeB (ng-eNB), a next generation NodeB (gNB), a node B (NB), a home node B (HNB), a home evolved nodeB (HeNB), a wireless backhaul device, a radio network controller (RNC), a base station controller (BSC), a base transceiver station (BTS), a base band unit (BBU), a mobile switching center, a base station in a 6G communication system, an open base station (Open RAN), a cloud base station (Cloud RAN), a base station in other communication systems, and at least one of an access node in a Wi-Fi system, but is not limited thereto.

[0106] In some embodiments, the technical solution of the present disclosure can be applied to the Open RAN architecture. In this case, the interfaces between or within the access network devices involved in the embodiments of the present disclosure can be transformed into internal interfaces of the Open RAN, and the processes and information interactions between these internal interfaces can be implemented through software or programs.

[0107] In some embodiments, the access network device can be composed of a centralized unit (CU) and a distributed unit (DU), where the CU can also be called a control unit. The CU-DU structure can be used to split the protocol layer of the access network device, with the functions of some protocol layers centrally controlled by the CU, and the functions of the remaining part or all of the protocol layers distributed in the DU, which is centrally controlled by the CU, but is not limited to this.

[0108] In some embodiments, the core network device 103 may be a single device including a first network element 1031, a second network element 1032, etc., or may be a plurality of devices or a device group including all or part of the first network element 1031, the second network element 1032, the third network element 1033, etc. The network element may be virtual or physical. The core network may include, for example, at least one of an evolved packet core (EPC), a 5G core network (5GCN), and a next generation core (NGC).

[0109] In some embodiments, the first network element 1031 , the second network element 1032 , and the third network element 1033 may be modules for implementing a certain function, for example.

[0110] In some embodiments, the first network element 1031 is, for example, a module for implementing a network intelligent computing function (NICF).

[0111] In some embodiments, the first network element 1031 has at least one of the following functions: a task computing function, a task scheduling function, and a task anchor function.

[0112] In some embodiments, for task computing functions, the first network element 1031 may retrieve and call data collected by the third network element 1033 and stored by the second network element 1032, and provide model training and inference decision-making during the AI ​​lifecycle, which may be specifically embodied in at least one of the following:

[0113] Optionally, the task computing function includes, for example, data and / or signaling interaction between different nodes during task execution.

[0114] Optionally, the task computing function includes, for example, collaborating with the terminal 101 and the access network device 102 to split data and / or models.

[0115] Optionally, computing functions for the task, including, for example, making graph decisions based on test data from a data plane, and providing results.

[0116] In some embodiments, for the task scheduling function, the control and scheduling phase of AI tasks can be implemented, including, for example, control information collection and scheduling resource management, for real-time adjustment of AI models and task ratios, or real-time adjustment of computing power as the network environment changes.

[0117] Optionally, information collection means that the first network element 1031 can perceive at least one of the following contents of each network node: computing power load, data processing capability, AI algorithm model and channel state information.

[0118] In some embodiments, for the anchor function, task lifecycle management can be implemented to complete task deployment, startup, deletion and monitoring based on task Quality of Service (QoS) requirements, and regulate task-related connection, computing, data and algorithm resources for coarse-grained QoS guarantee of task deployment.

[0119] In some embodiments, the second network element 1032 is, for example, a module for implementing a network data repository function (NDRF).

[0120] In some embodiments, the second network element 1032 integrates all storage-related functions, including, for example, network repository function (NRF), unified data repository (UDR), user-defined storage function (UDSF), and application data storage function (ADRF).

[0121] In some embodiments, the second network element 1032 may be used to store at least one of the following: user data (e.g., number of user registrations, service-related data, etc.), network function configuration files, network data (e.g., network data, service level agreement data, network node load data, etc.).

[0122] Exemplarily, the second network element 1032 is used to store the inherent and unchanging capability information of the terminal 101 and / or to store the inherent and unchanging capability information of the access network device 102 .

[0123] In some embodiments, the second network element 1032 can also store calculation results and provide the calculation results as historical data to users to reduce resource waste caused by redundant calculations.

[0124] In some embodiments, the third network element 1033 is, for example, a module for implementing a network data collection function (NDCF).

[0125] In some embodiments, the third network element 1033 is used to obtain real-time network information from different network functions, including, for example, real-time network traffic, network congestion, illegal access, etc.

[0126] In some embodiments, the third network element 1033 is further configured to collect data transmitted between the core network device 103 and the access network device 102 .

[0127] Optionally, after the transferred data is collected, preprocessing operations (including, for example, normalization processing, regularization processing, etc.) may be performed on the data.

[0128] In some embodiments, the third network element 1033 may be integrated with the second network element 1032 into one network element, that is, the integrated network element may be a module capable of implementing both NDCF and NDRF.

[0129] It can be understood that the communication system described in the embodiment of the present disclosure is for the purpose of more clearly illustrating the technical solution of the embodiment of the present disclosure, and does not constitute a limitation on the technical solution proposed in the embodiment of the present disclosure. Ordinary technicians in this field can know that with the evolution of the system architecture and the emergence of new business scenarios, the technical solution proposed in the embodiment of the present disclosure is also applicable to similar technical problems.

[0130] The following embodiments of the present disclosure may be applied to the communication system 100 shown in FIG1A , or a portion thereof, but are not limited thereto. The entities shown in FIG1A are illustrative only. The communication system may include all or part of the entities shown in FIG1A , or may include other entities other than those shown in FIG1A . The number and form of the entities are arbitrary, and the entities may be physical or virtual. The connection relationships between the entities are illustrative only. The entities may be connected or disconnected, and the connection may be in any manner, including direct or indirect, wired or wireless.

[0131] The embodiments of the present disclosure 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), 5G new radio (NR), future radio access (FRA), new radio access technology (RAT), new radio (NR), new radio access (NX), future generation radio access (FX), Global System for Mobile communications (GSM (registered trademark)), CDMA2000, Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi (registered trademark)), IEEE 802.16 (WiMAX (registered trademark)), IEEE 802.20, Ultra-WideBand (UWB), Bluetooth (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 utilizing other communication methods, and next-generation systems based on and extending these methods. Furthermore, multiple systems may be combined (for example, a combination of LTE or LTE-A with 5G).

[0132] Artificial Intelligence (AI) and Machine Learning (ML) are widely used in the sixth generation mobile communication system network (6 thIn 5G, AI or ML uses the network data analytic function (NWDAF) as a plug-in for network analysis, providing AI algorithms as patching tools. However, this approach has problems such as limited application scenarios, complex data transmission, and low computing resource utilization.

[0133] FIG1B is a schematic diagram of a 5G communication system architecture according to an embodiment of the present disclosure.

[0134] As shown in Figure 1B, service-based interfaces are used within the control plane. The architecture includes the following service-based interfaces and reference points. For example, N1, N2, ..., N9, etc. in Figure 1B are reference points. For example, Nnssf, Nnef, Naf, etc. are all service-based interfaces. NSSF, NEF, etc. in Figure 1B are all core network elements, and the plane formed by these core network elements can be considered the control plane. (R)AN in Figure 1B represents the access network. Reference points show how various network functions interact with each other and how network functions (NFs) in the control plane transmit data or information to other NFs through the control bus.

[0135] FIG1C is a schematic diagram of a data storage architecture for storing unstructured data from any NF, according to an embodiment of the present disclosure. As shown in FIG1C , the 5G system architecture allows any NF to store its unstructured data in an unstructured data storage function (UDSF), such as a UE context, and retrieve it from the UDSF. NFs can share a UDSF to store their respective unstructured data, or each can have its own UDSF. For example, the UDSF can be located near each NF. Nudsf in FIG1Cb represents a service interface, and N18 represents a reference point.

[0136] FIG1D is a schematic diagram of a data storage architecture according to an embodiment of the present disclosure.

[0137] As shown in Figure 1D, the 5G system architecture allows the unified data management entity (UDM), policy control function entity (PCF), and network element function (NEF) to store data in a unified data repository (UDR). The UDM and PCF store subscription data and policy data, and the NEF stores structured data and application data for exposure, including packet flow descriptions for application detection and application function (AF) request information of multiple UEs.

[0138] FIG1E is a schematic diagram of an architecture for collecting data from any 5G Cloud Native Network Functions (CNF) according to an embodiment of the present disclosure.

[0139] As shown in Figure 1E, network data analysis can be performed using NWDAF. The 5G system architecture allows NWDAF to collect data from any 5G CNF.

[0140] In some embodiments, in a 5G communication system, the following problems may occur during the use of NWDAF:

[0141] 1) Because the AI ​​function uses NWDAF for network analysis through patch integration, a large number of test reports generated during NWDAF network analysis or AI function operation may cause data security issues and excessive signaling overhead.

[0142] 2) Data related to AI computing is mainly transmitted to NWDF through the control panel in the form of requests or subscription messages, which may increase network load or cause network congestion and other problems.

[0143] 3) In 5G communication systems, the lack of collaborative computing between the core network, terminals, and access network equipment will cause computing pressure to be concentrated on a certain node, resulting in unreasonable allocation of computing resources, which in turn will cause higher network latency and insufficient data privacy protection.

[0144] Based on this, the embodiments of the present disclosure propose a task processing method to achieve coordinated processing of AI tasks among multiple nodes, thereby rationally utilizing computing resources and reducing network latency.

[0145] FIG2A is an interactive diagram of a task processing method according to an embodiment of the present disclosure. As shown in FIG2A , the present disclosure embodiment relates to a task processing method, which includes:

[0146] Step S2101: The first network element 1031 obtains an AI task to be processed.

[0147] In some embodiments, the first network element 1031 receives information from a third-party user (3 rd Party) to obtain pending AI tasks.

[0148] In some embodiments, the first network element 1031 receives an AI task sent by a third-party user.

[0149] Optionally, the first network element 1031 obtains a request sent by a third-party user, and determines an AI task to be processed based on the request sent by the third-party user.

[0150] It should be noted that AI tasks can be understood as tasks that require the use of AI models or AI functions to process, such as AI-based positioning tasks and AI-based beam prediction tasks.

[0151] In some embodiments, the first network element is a module having a Network Intelligent Computing Function (NICF).

[0152] In some embodiments, terms such as "send", "transmit", "report", "download", "transmit", "bidirectional transmission", "send and / or receive" can be used interchangeably.

[0153] In some embodiments, "obtain", "get", "get", "receive", "transmit", "bidirectional transmission", "send and / or receive" can be interchangeable, and can be interpreted as receiving from other entities, obtaining from protocols, obtaining from higher layers, obtaining by self-processing, autonomous implementation, etc.

[0154] In some embodiments, "AI model", "artificial intelligence model", "machine learning (ML) model", "ML model", "AI", "AI function", "ML function", "artificial intelligence function", and "machine learning function" can be replaced with each other.

[0155] In some embodiments, the names of information, etc. are not limited to the names described in the embodiments, and terms such as "information", "message", "signal", "signaling", "report", "configuration", "indication", "instruction", "command", "channel", "parameter", "domain", "field", "symbol", "symbol", "codeword", "codebook", "codeword", "codepoint", "bit", "data", "program", and "chip" can be used interchangeably.

[0156] In some embodiments, terms such as "certain", "preset", "preset", "setting", "indicated", "some", "any", and "first" can be interchangeable. "Specific A", "preset A", "preset A", "setting A", "indicated A", "some A", "any A", and "first A" can be interpreted as A pre-specified in a protocol, etc., or as A obtained through setting, configuration, or indication, etc., or as specific A, some A, any A, or first A, etc., but not limited to this.

[0157] In step S2102 , the first network element 1031 sends a capability request to the second network element 1032 .

[0158] In some embodiments, the second network element 1032 receives a capability request from the first network element 1031 .

[0159] In some embodiments, the second network element is a module having a network data repository function (NDRF).

[0160] Optionally, the second network element may also be a module having both NRDF and network data collection function (Network Data Collection Function, NDCF).

[0161] It is understandable that the NDCF function can be integrated into the second network element as described above, or can be deployed in an independent network element (eg, the third network element, the fourth network element, ..., the Nth network element, etc.).

[0162] It should be noted that, to facilitate understanding of the following content, the second network element in the embodiment of the present disclosure is a module having both NRDF and NDCF functions.

[0163] In some embodiments, the capability request is used to obtain capability information stored in the second network element 1032 from the second network element 1032 .

[0164] Exemplarily, information about the capabilities of the terminal 101 stored in the second network element 1032 and / or information about the capabilities of the access network device 102 stored in the second network element 1032 are obtained.

[0165] Step S2103 : The second network element 1032 sends first capability information to the first network element 1031 .

[0166] In some embodiments, the first network element 1031 receives first capability information sent from the second network element 1032 .

[0167] In some embodiments, the first information includes: inherent capability information of the terminal 101 and / or inherent capability information of the access network device 102 .

[0168] In some embodiments, the inherent capability information of the terminal 101 includes: computing capability information of the terminal 101 and / or resource information of the terminal 101 .

[0169] Exemplarily, the computing capability of the terminal 101 may be understood as data processing capability, including, for example, the amount of data that the terminal 101 can process per unit time.

[0170] Exemplarily, the resource information of the terminal 101 may be understood as network resources or hardware resources used by the terminal 101, including, for example, network resource information of the network to which the terminal 101 accesses.

[0171] In some embodiments, the inherent capability information of the access network device 102 includes: computing capability information of the access network device 102 , and / or resource information of the access network device 102 .

[0172] Exemplarily, the computing capability of the access network device 102 may be understood as data processing capability, including, for example, the amount of data that the access network device 102 can process per unit time.

[0173] Exemplarily, the resource information of the access network device 102 may be understood as network resources or hardware resources used by the access network device 102 , including, for example, network resource information of the network to which the access network device 102 accesses.

[0174] In some embodiments, the second network element 1032 sends the first capability information to the first network element 1031 based on the capability request sent by the first network element 1031 .

[0175] Optionally, the second network element 1032 sends the first capability information to the first network element 1031 in response to the capability request.

[0176] It can be understood that the first capability information sent by the first network element 1031 through the second network element 1032 can determine the inherent capabilities of the terminal 101 and / or the inherent capabilities of the access network device 102, and then can divide the AI ​​tasks based on the inherent capabilities of the two.

[0177] Step S2104 : The terminal 101 sends the second capability information of the terminal 101 to the access network device 102 .

[0178] In some embodiments, the access network device 102 receives the second capability information of the terminal 101 sent by the terminal 101 .

[0179] In some embodiments, the access network device 102 periodically receives the second capability information sent by the terminal 101 .

[0180] In some embodiments, the terminal 101 periodically sends the second capability information to the access network device 102 .

[0181] In some embodiments, the second capability information is capability information that changes in real time.

[0182] Exemplarily, the second capability information of the terminal 101 may be understood as capability information about the terminal 101 that changes in real time.

[0183] In some embodiments, the second capability information of the terminal 101 includes: real-time load information about the terminal 101 and / or resource occupancy information about the terminal 101 .

[0184] Exemplarily, the real-time load information about the terminal 101 may be understood as information indicating the real-time load status of the terminal 101 , and the resource occupancy information about the terminal 101 may be understood as information indicating the resource occupancy status of the terminal 101 .

[0185] It is understandable that the terminal 101 can provide feedback on its own changing capability information in real time by reporting its own related second capability information, so as to facilitate the first network element 1031 to perform AI task segmentation based on the information.

[0186] Step S2105 : The access network device 102 sends the second capability information to the first network element 1031 .

[0187] In some embodiments, the first network element 1031 receives the second capability information sent by the access network device 102 .

[0188] In some embodiments, the second capability information includes the second capability information of the terminal 101 and the second capability information of the access network device 102 .

[0189] It is understandable that the second capability information related content of the terminal 101 has been described in the relevant embodiments of step S2104. For detailed description, please refer to the description of the relevant embodiments and will not be repeated here.

[0190] In some embodiments, the second capability information of the access network device 102 can be understood as capability information about the access network device 102 that changes in real time.

[0191] In some embodiments, the second capability information of the access network device 102 includes: real-time load information about the access network device 102 and / or resource occupancy information about the access network device 102 .

[0192] For example, the real-time load information about the access network device 102 can be understood as information used to represent the real-time load status of the access network device 102, and the resource occupancy information about the access network device 102 can be understood as information used to represent the resource occupancy status of the access network device 102.

[0193] In some embodiments, the second capability information may also include network environment information of the terminal 101 and / or network environment information of the access network device 102.

[0194] In some embodiments, the access network device 102 periodically sends the second capability information to the first network element 1031 .

[0195] In some embodiments, the first network element 1031 periodically receives the second capability information sent by the access network device 102 .

[0196] It can be understood that the access network device 102 sends the second capability information to the first network element 1031, so that the first network element 1031 can determine the real-time capability information of the terminal 101 and the access network device 102 based on the second capability information, and then perform AI task division based on the real-time capability information of the two.

[0197] In step S2106 , the first network element 1031 divides the AI ​​task into multiple AI subtasks.

[0198] In some embodiments, the first network element 1031 divides the AI ​​task into multiple AI subtasks based on the first capability information and the second capability information.

[0199] In some embodiments, the first network element 1031 divides the AI ​​task into multiple AI subtasks that are collaboratively executed by the first network element 1031, the terminal 101, and the access network device 102 based on the first capability information and the second capability information.

[0200] Optionally, the first network element 1031 divides the AI ​​task into multiple AI subtasks based on the first capability information and the second capability information, including at least one of the following -A) to -C):

[0201] -A) based on the first capability information corresponding to the terminal 101 and the second capability information corresponding to the terminal 101, dividing the AI ​​task to obtain a first AI subtask to be processed by the terminal 101;

[0202] -B) based on the first capability information corresponding to the access network device 102 and the second capability information corresponding to the access network device 102, dividing the AI ​​task to obtain a second AI subtask to be processed by the access network device 102;

[0203] -C) The remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask are used as the third AI subtask to be processed by the first network element.

[0204] For example, FIG2B is a schematic diagram of AI task segmentation according to an embodiment of the present disclosure. As shown in FIG2B , the AI ​​task is assumed to be a 5-layer structural neural network structure, in which the forward propagation propagates from layer 1 to layer 5. The first network element 1031 can, based on -A), split the processing process of layer 1 and layer 2 of the AI ​​task and hand it over to the terminal 101 for processing. The first network element 1031, based on -B), splits the processing process of layer 3 and layer 4 of the AI ​​task and hands it over to the access network device 102 for processing. The first network element 1031 handles the remaining layer 5 after the segmentation by itself.

[0205] It is understandable that compared to the terminal 101, the access network device 102 has more powerful computing capabilities. Therefore, the first network element 1031 can split the AI ​​tasks, obtain tasks that require greater computing power, and hand them over to the access network device 102 for processing. Therefore, computing resources can be allocated more reasonably, thereby improving computing efficiency.

[0206] In some embodiments, the first network element 1031 may segment at least one of the following items of the AI ​​task: task description, task identifier, task configuration information, and Quality of Artificial Intelligence Service (QoAIS) indicator.

[0207] Optionally, the description information of the AI ​​task is used to describe the characteristics of the AI ​​task, and may include, for example, information describing the type of the AI ​​task, the AI ​​function or AI model corresponding to the AI ​​task, and the like.

[0208] Optionally, the configuration information of the AI ​​task may include, for example, a training model of the AI ​​task, data of the AI ​​task, etc.

[0209] Optionally, AI service quality indicators may include, for example, energy consumption, latency, overhead, connection quality, etc. caused by AI tasks.

[0210] In some embodiments, after the first network element 1031 completes the division of the AI ​​task (for example, while the terminal 101 or the access network device 102 executes its corresponding subtask), it may also periodically receive a status report sent by the second network element 1032.

[0211] Optionally, the status report includes second capability information of the terminal 101 and the access network device 102 in the change period.

[0212] In some embodiments, when the second capability information meets a preset condition, the first network element 1031 re-divides the AI ​​task to obtain a plurality of re-divide AI subtasks.

[0213] Exemplarily, in the second capability information, if the real-time load of the terminal 101 exceeds a preset load threshold, the first network element 1031 re-divides the AI ​​tasks to reduce the amount of tasks of the terminal 101 under the load.

[0214] Continuing with the embodiment of Figure 2B , after first network element 1031 completes AI task division, if the first AI subtask corresponding to terminal 101 is the processing task of neural network layer 1 and layer 2, terminal 101 reports real-time load information during the current cycle, and the real-time load information exceeds a preset load threshold. After obtaining this real-time load information, first network element 1031 can re-divide the AI ​​task, and the divided first AI subtask can be, for example, the processing task of neural network layer 1.

[0215] It is understandable that the redefined first AI subtask requires less computation than the previous first AI subtask, thereby making the allocation of computing resources more reasonable and improving task processing efficiency.

[0216] In some embodiments, after the execution content of the first AI subtask of terminal 101 is changed, the execution content of the second AI subtask will also change, and the execution content of the third AI subtask may or may not change.

[0217] For example, continuing with the embodiment of Figure 2B above, when the first AI subtask is redefined from the processing task of neural network layer 1 and layer 2 to the processing task of neural network layer 1, the second AI subtask is redefined by the first network element 1031 based on the second capability information of the access network device 102.

[0218] If the first network element 1031 determines, based on the second capability information of the access network device 102, that the access network device 102 can process tasks between layers 2 and 4, the second AI subtask is changed from processing tasks between layers 3 and 4 to processing tasks between layers 2 and 4. This shows that the execution content of the second AI subtask has changed, but the content of the third AI subtask has not changed.

[0219] If the first network element 1031 determines, based on the second capability information of the access network device 102, that the access network device 102 can process tasks between layers 2 and 3, it changes the second AI subtask from processing tasks between layers 3 and 4 to processing tasks between layers 2 and 3, and further changes the third AI subtask from processing tasks between layer 5 to processing tasks between layers 4 and 5. This shows that the execution content of the second AI subtask has changed, and the content of the third AI subtask has also changed.

[0220] Step S2107: The first network element 1031 configures the first AI subtask.

[0221] In some embodiments, the first network element 1031 configures a first AI subtask for the terminal 101 .

[0222] In some embodiments, the first network element 1031 configures a first AI subtask for the terminal 101 based on the configuration information.

[0223] It is understandable that the relevant description of the first AI subtask has been explained in the relevant embodiment of step S2106. For related content, please refer to the description of the corresponding embodiment and will not be repeated here.

[0224] Step S2108: The first network element 1031 configures the second AI subtask.

[0225] In some embodiments, the first network element 1031 configures a second AI subtask for the access network device 102 .

[0226] In some embodiments, the first network element 1031 configures a second AI subtask for the access network device 102 based on the configuration information.

[0227] It is understandable that the relevant description of the second AI subtask has been explained in the relevant embodiment of step S2106. For related content, please refer to the description of the corresponding embodiment and will not be repeated here.

[0228] In step S2109, the first network element 1031 configures the third AI subtask.

[0229] In some embodiments, the first network element 1031 configures a third AI subtask for itself.

[0230] It is understandable that the relevant description of the second AI subtask has been explained in the relevant embodiment of step S2106. For related content, please refer to the description of the corresponding embodiment and will not be repeated here.

[0231] In step S2110 , the terminal 101 executes the first AI subtask.

[0232] In some embodiments, the terminal 101 executes the first network element 1031 to configure the first AI subtask for the terminal 101 and obtain the first AI subtask processing result.

[0233] For example, continuing with the embodiment related to FIG2B , it is assumed that the first AI subtask configured by the first network element 1031 for the terminal 101 is the processing of neural network layers 1 and 2. The terminal 101 then performs calculations on neural network layers 1 and 2, and uses the processing results of neural network layers 1 and 2 as the processing results of the first AI subtask.

[0234] In step S2111 , the terminal 101 sends the first AI subtask processing result to the access network device 102 .

[0235] In some embodiments, the access network device 102 receives the first AI subtask processing result sent by the terminal 101.

[0236] It is understandable that the processing result of the first AI subtask can be considered as the processing result of the AI ​​task processed by terminal 101, corresponding to the processing results of neural network layer 1 and layer 2 in the relevant embodiment of Figure 2B. It is understandable that during the AI ​​task processing process, the upper layer processing result will be used as the input of the lower layer (for example, layer 3). Therefore, in order to ensure the smooth processing of the lower layer AI task (for example, the second AI subtask), the processing result of the first AI subtask is sent to the processing unit of the next subtask as input.

[0237] In step S2112, the access network device 102 executes the second AI subtask.

[0238] In some embodiments, the access network device 102 executes the first network element 1031 to configure the second AI subtask for the access network device 102 and obtain the processing result of the second AI subtask.

[0239] For example, continuing with the embodiment related to FIG2B , assume that the second AI subtask configured by the first network element 1031 for the access network device 102 is processing of neural network layers 3 and 4. The access network device 102 then performs computations on neural network layers 3 and 4, and uses the processing results of neural network layers 3 and 4 as the processing results of the second AI subtask.

[0240] In step S2113 , the access network device 102 sends the second AI subtask processing result to the first network element 1031 .

[0241] In some embodiments, the first network element 1031 receives the second AI subtask processing result sent by the access network device 102.

[0242] It is understandable that the processing result of the second AI subtask can be considered as the processing result of the AI ​​task processed by the access network device 102, corresponding to the processing results of neural network layers 3 and 4 in the relevant embodiment of Figure 2B. It is understandable that during the AI ​​task processing process, the upper layer processing result will be used as the input of the lower layer (for example, layer 5). Therefore, in order to ensure the smooth processing of the lower layer AI task (for example, the third AI subtask), the processing result of the second AI subtask is sent to the processing unit of the next subtask as input.

[0243] In step S2114, the first network element 1031 executes the third AI subtask.

[0244] In some embodiments, the first network element 1031 executes the third AI subtask and obtains a processing result.

[0245] For example, continuing with the embodiment related to FIG. 2B , assume that the third AI subtask is processing neural network layer 5. First network element 1031 then performs computational processing on neural network layer 5 and uses the processing result of neural network layer 5 as the processing result of the third AI subtask. Since the third AI subtask executed by first network element 1031 is the remaining AI task after the first and second AI subtasks, the processing result of the third AI subtask is also the final processing result of the AI ​​task.

[0246] Step S2115 , the first network element 1031 sends the processing result to the second network element 1032 .

[0247] In some embodiments, the second network element 1032 receives the processing result from the first network element 1031 .

[0248] Optionally, the second network element 1032 may also perform corresponding format conversion on the received processing result according to preset requirements.

[0249] It is understandable that, since the second network element 1032 has the NDRF function, it can store the processing results of the AI ​​task and then send the stored processing results to a third-party user or other device.

[0250] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S2101 to S2115. For example, step S2101 can be implemented as an independent embodiment, step S2102 can be implemented as an independent embodiment, step S2106 can be implemented as an independent embodiment, step S2107 can be implemented as an independent embodiment, step S2108 can be implemented as an independent embodiment, step S2109 can be implemented as an independent embodiment, step S2110 can be implemented as an independent embodiment, step S2112 can be implemented as an independent embodiment, step S2114 can be implemented as an independent embodiment, step S2102 + step S2103 can be implemented as an independent embodiment, step S2104 + step S2105 can be implemented as an independent embodiment, step S2107 + step S2108 can be implemented as an independent embodiment, and step S2111 + step S2113 can be implemented as an independent embodiment, but the present invention is not limited thereto.

[0251] In some embodiments, step S2102 and step S2103 may be executed in an interchanged order or simultaneously.

[0252] In some embodiments, step S2107, step S2108, and step S2109 may be executed in an interchanged order or simultaneously.

[0253] In some embodiments, step S2107 and step S2108 may be executed in an interchanged order or simultaneously.

[0254] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, and step S2114 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0255] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0256] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0257] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0258] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0259] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0260] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0261] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0262] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2106, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0263] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2105, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0264] In some embodiments, step S2101, step S2102, step S2103, step S2104, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, step S2112 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0265] In some embodiments, step S2101, step S2102, step S2103, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, step S2112 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0266] In some embodiments, step S2101, step S2102, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0267] In some embodiments, step S2101, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0268] In some embodiments, step S2102, step S2103, step S2104, step S2105, step S2106, step S2107, step S2108, step S2109, step S2110, step S2111, step S2112, step S2113, step S2114, and step S2115 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0269] In some embodiments, reference may be made to other optional implementations described before or after the description corresponding to FIG. 2A .

[0270] FIG3A is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG3A , the present disclosure embodiment relates to a task processing method, which includes:

[0271] Step S3101, obtain the AI ​​task to be processed.

[0272] In some embodiments, the first network element 101 receives an AI task to be processed sent by a third-party user, but is not limited thereto. The first network element 101 may also receive an AI task to be processed sent by other entities.

[0273] In some embodiments, step S3101 is omitted, and the terminal 101 autonomously implements the function of obtaining the AI ​​task to be processed, or the above function is default or acquiescent.

[0274] In some embodiments, the optional implementation of step S3101 can refer to the optional implementation of step S2101 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0275] Step S3102: Send a capability request.

[0276] In some embodiments, the optional implementation of step S3102 can refer to the optional implementation of step S2102 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0277] Step S3103: Receive first capability information.

[0278] In some embodiments, the optional implementation of step S3103 can refer to the optional implementation of step S2103 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0279] Step S3104: Receive second capability information.

[0280] In some embodiments, the optional implementation of step S3104 can refer to the optional implementation of step S2104 and step S2105 in Figure 2A, and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0281] Step S3105: Divide the AI ​​task into multiple AI subtasks.

[0282] In some embodiments, the optional implementation of step S3105 can refer to the optional implementation of step S2106 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0283] Step S3106: Configure the first AI subtask.

[0284] In some embodiments, the optional implementation of step S3106 can refer to the optional implementation of step S2107 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0285] Step S3107: Configure the second AI subtask.

[0286] In some embodiments, the optional implementation of step S3107 can refer to the optional implementation of step S2108 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0287] Step S3108: Configure the third AI subtask.

[0288] In some embodiments, the optional implementation of step S3108 can refer to the optional implementation of step S2109 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0289] Step S3109: Receive the second AI subtask processing result.

[0290] In some embodiments, the optional implementation of step S3109 can refer to the optional implementation of step S2111 and step S2113 in Figure 2A, and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0291] Step S3110: Execute the third AI subtask.

[0292] In some embodiments, the optional implementation of step S3110 can refer to the optional implementation of step S2114 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0293] Step S3111, sending processing results.

[0294] In some embodiments, the optional implementation of step S3111 can refer to the optional implementation of step S2115 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0295] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S3101 to S3111. For example, step S3101 can be implemented as an independent embodiment, step S3102 can be implemented as an independent embodiment, step S3106 can be implemented as an independent embodiment, step S3107 can be implemented as an independent embodiment, step S3108 can be implemented as an independent embodiment, step S3109 can be implemented as an independent embodiment, step S3110 can be implemented as an independent embodiment, step S3102 + step S3103 can be implemented as an independent embodiment, step S3104 + step S3105 can be implemented as an independent embodiment, and step S3107 + step S3108 can be implemented as an independent embodiment, but the present invention is not limited thereto.

[0296] In some embodiments, step S3102 and step S3103 may be executed in an interchanged order or simultaneously.

[0297] In some embodiments, step S3107, step S3108, and step S3109 may be executed in an interchanged order or simultaneously.

[0298] In some embodiments, step S3107 and step S3108 may be executed in an interchanged order or simultaneously.

[0299] In some embodiments, step S3101, step S3102, step S3103, step S3104, step S3105, step S3106, step S3107, step S3108, step S3109, and step S3110 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0300] In some embodiments, step S3101, step S3102, step S3103, step S3104, step S3105, step S3106, step S3107, step S3108, step S3109, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0301] In some embodiments, step S3101, step S3102, step S3103, step S3104, step S3105, step S3106, step S3107, step S3108, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0302] In some embodiments, step S3101, step S3102, step S3103, step S3104, step S3105, step S3106, step S3107, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0303] In some embodiments, step S3101, step S3102, step S3103, step S3104, step S3105, step S3106, step S3108, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0304] In some embodiments, step S3101, step S3102, step S3103, step S3104, step S3105, step S3107, step S3108, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0305] In some embodiments, steps S3101, S3102, S3103, S3104, S3106, S3107, S3108, S3109, S3110, and S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments. In some embodiments, step S3108 is optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0306] In some embodiments, step S3101, step S3102, step S3103, step S3105, step S3106, step S3107, step S3108, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0307] In some embodiments, step S3101, step S3102, step S3104, step S3105, step S3106, step S3107, step S3108, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0308] In some embodiments, step S3101, step S3103, step S3104, step S3105, step S3106, step S3107, step S3108, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0309] In some embodiments, step S3102, step S3103, step S3104, step S3105, step S3106, step S3107, step S3108, step S3109, step S3110, and step S3111 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0310] FIG3B is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG3B , the embodiment of the present disclosure relates to a task processing method, which includes:

[0311] Step S3201: Obtain the AI ​​task to be processed.

[0312] In some embodiments, the optional implementation of step S3201 can refer to the optional implementation of step S2101 in Figure 2A, step S3101 in Figure 3A, and other related parts in the embodiments involved in Figures 2A and 3A, which will not be repeated here.

[0313] Step S3202: Divide the AI ​​task into multiple AI subtasks.

[0314] In some embodiments, the optional implementation of step S3201 can be found in the optional implementation of steps S2101, S2102, S2103, and S2104 in Figure 2A, steps S3102, S3103, S3104, and S3105 in Figure 3A, as well as other related parts in the embodiments of Figures 2 and 3A, which will not be repeated here.

[0315] Step S3203: Receive the second AI subtask processing result.

[0316] In some embodiments, the optional implementation of step S3201 can be found in steps S2107, S2108, S2109, S2110, S2111, S2112, and S2113 of Figure 2A , the optional implementation of steps S3106, S3107, S3108, and S3109 of Figure 3A , and other related parts in the embodiments of Figures 2 and 3A , which will not be repeated here.

[0317] Step S3204: Execute the third AI subtask.

[0318] In some embodiments, the optional implementation of step S3204 can refer to the optional implementation of step S2114 in Figure 2A, step S3110 in Figure 3A, and other related parts in the embodiments involved in Figures 2A and 3A, which will not be repeated here.

[0319] Step S3205, sending the processing result.

[0320] In some embodiments, optional implementations of step S3205 may refer to the optional implementations of step S2115 in FIG. 2A , step S3111 in FIG. 3A , and other related parts in the embodiments involved in FIG. 2A and FIG. 3A , which will not be repeated here.

[0321] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S3201 to S3205. For example, step S3201 can be implemented as an independent embodiment, step S3202 can be implemented as an independent embodiment, step S32036 can be implemented as an independent embodiment, step S3204 can be implemented as an independent embodiment, step S3205 can be implemented as an independent embodiment, step S3201 + step S3202 can be implemented as an independent embodiment, step S3203 + step S3204 can be implemented as an independent embodiment, and step S3201 + step S3202 + step S3203 + step S3204 can be implemented as independent embodiments, but the present invention is not limited thereto.

[0322] In some embodiments, step S3203 and step S3204 may be executed in an interchanged order or simultaneously.

[0323] In some embodiments, step S3201, step S3202, step S3203, and step S3204 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0324] In some embodiments, step S3201, step S3202, step S3203, and step S3205 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0325] In some embodiments, step S3201, step S3202, step S3204, and step S3205 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0326] In some embodiments, step S3201, step S3203, step S3204, and step S3205 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0327] In some embodiments, step S3202, step S3203, step S3204, and step S3205 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0328] FIG3C is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG3C , the present disclosure embodiment relates to a task processing method, which includes:

[0329] Step S3301: Obtain the AI ​​task to be processed.

[0330] In some embodiments, the optional implementation of step S3301 can refer to the optional implementation of step S2101 in Figure 2A, step S3101 in Figure 3A, step S3201 in Figure 3B, and other related parts in the embodiments involved in Figures 2A, 3A, and 3B, which will not be repeated here.

[0331] Step S3302: Divide the AI ​​task into multiple AI subtasks and obtain the processing result of the AI ​​task.

[0332] In some embodiments, the optional implementation of step S3301 can refer to the optional implementation of steps S2102 to S2115 in Figure 2A, steps S3102 to S3111 in Figure 3A, and steps S3202 to S3205 in Figure 3B, as well as other related parts in the embodiments involved in Figures 2A, 3A, and 3B, which will not be repeated here.

[0333] In some embodiments, the AI ​​task is divided into multiple AI subtasks, including: sending a capability request to a second network element, and obtaining first capability information sent by the second network element in response to the capability request, the first capability information is the inherent and unchanging information of the terminal and the inherent and unchanging capability information of the access network device; obtaining second capability information sent by the access network device, the second capability information includes the second capability information of the access network device, and the second capability information of the terminal periodically sent to the access network device, the second capability information is the real-time changing capability information of the terminal and the real-time changing capability information of the access network device; based on the first capability information and the second capability information, the AI ​​task is divided into multiple AI subtasks.

[0334] In some embodiments, based on the first capability information and the second capability information, the AI ​​task is divided into multiple AI subtasks, including: based on the first capability information corresponding to the terminal and the second capability information corresponding to the terminal, the first AI subtask to be processed by the terminal is divided in the AI ​​task; based on the first capability information corresponding to the access network device and the second capability information corresponding to the access network device, the second AI subtask to be processed by the access network device is divided in the AI ​​task; and the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask are used as the third AI subtask to be processed by the first network element.

[0335] In some embodiments, the scheduling terminal and the access network device jointly process multiple AI subtasks to obtain the processing results of the AI ​​tasks, including: obtaining the second AI subtask processing result sent by the access network device, the second AI subtask processing result is the result of the access network device processing the second AI subtask based on the first AI subtask processing result, and the first AI subtask processing result is the result of the terminal processing the first AI subtask and sending it to the access network device; processing the third AI subtask based on the second AI subtask processing result to obtain the processing result of the AI ​​task.

[0336] In some embodiments, the method also includes: periodically obtaining a status report sent by a second network element, the status report including second capability information of the access network device and the terminal; when the second capability information meets a preset condition, re-dividing the AI ​​task to obtain multiple re-divided AI subtasks.

[0337] In some embodiments, after the first network element, the terminal, and the access network device jointly process multiple AI subtasks, it also includes: obtaining the processing results of the AI ​​tasks and sending them to the second network element, and the second network element is used to store the processing results.

[0338] In some embodiments, the first capability information includes at least one of the following: computing capability of the terminal; resource information of the terminal; computing capability of the access network device; resource information of the access network device.

[0339] In some embodiments, the second capability information includes at least one of the following: real-time load of the terminal; resource occupancy information of the terminal; real-time load of the access network device; resource occupancy information of the access network device.

[0340] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S3301 and S3302. For example, step S3301 may be implemented as an independent embodiment, step S3302 may be implemented as an independent embodiment, and step S3301 + step S3302 may be implemented as independent embodiments, but the present disclosure is not limited thereto.

[0341] In some embodiments, step S3301 and step S3302 may be executed in an interchanged order or simultaneously.

[0342] In some embodiments, step S3302 is optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0343] In some embodiments, step S3301 is optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0344] FIG4A is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG4A , the present disclosure embodiment relates to a task processing method, which includes:

[0345] Step S4101: Send second capability information of the terminal.

[0346] The optional implementation of step S4101 can refer to the optional implementation of step S2104 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0347] Step S4102: Obtain the configured first AI subtask.

[0348] The optional implementation of step S4102 can refer to the optional implementation of step S2107 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0349] Step S4103: Execute the first AI subtask.

[0350] The optional implementation of step S4103 can refer to the optional implementation of step S2110 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0351] Step S4104: Send the first AI subtask processing result.

[0352] In some embodiments, the optional implementation of step S4104 can refer to the optional implementation of step S2111 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0353] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S4101 to S4104. For example, step S4101 may be implemented as an independent embodiment, step S4102 may be implemented as an independent embodiment, and step S4102 + step S4103 may be implemented as independent embodiments, but the present invention is not limited thereto.

[0354] In some embodiments, step S4101 and step S4102 may be executed in an interchanged order or simultaneously.

[0355] In some embodiments, step S4101, step S4102, and step S4103 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0356] In some embodiments, step S4101, step S4102, and step S4104 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0357] In some embodiments, step S4101, step S4103, and step S4104 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0358] In some embodiments, step S4102, step S4103, and step S41043 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0359] FIG4B is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG4C , an embodiment of the present disclosure relates to a task processing method, which includes:

[0360] Step S4201: Obtain the configured first AI subtask.

[0361] The optional implementation of step S4201 can be found in step S2104 and step S2107 of Figure 2A, step S4101 and the optional implementation of step S4102 of Figure 4A, and other related parts in the embodiments involved in Figures 2A and 4A, which will not be repeated here.

[0362] Step S4202: Execute the first AI subtask and send the first AI subtask processing result.

[0363] Optional implementations of step S4202 can be found in step S2110 and step S2111 of FIG. 2A , step S4103 and the optional implementations of step S4104 of FIG. 4A , and other related parts in the embodiments involved in FIG. 2A and FIG. 4A , which will not be repeated here.

[0364] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S4101 and S4102. For example, step S4101 may be implemented as an independent embodiment, step S4102 may be implemented as an independent embodiment, and step S4101 + step S4102 may be implemented as independent embodiments, but the present invention is not limited thereto.

[0365] In some embodiments, step S4101 and step S4102 may be executed in an interchanged order or simultaneously.

[0366] In some embodiments, step S4102 is optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0367] In some embodiments, step S4101 is optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0368] FIG4C is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG4C , the embodiment of the present disclosure relates to a task processing method, which includes:

[0369] In step S4301, the terminal processes the first AI subtask to be processed based on the scheduling of the first network element.

[0370] In some embodiments, the first AI subtask is a subtask processed by the terminal among a plurality of subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0371] In some embodiments, the optional implementation of step S4301 can be found in step S2104, step S2107, step S2110, step S2111 of Figure 2A, step S4101, step S4102, step S4103, step S4104 of Figure 4A, step S4201 of Figure 4B, and the optional implementation of step S4202, as well as other related parts in the embodiments involved in Figures 2A, 4A and 4B, which will not be repeated here.

[0372] In some embodiments, the method also includes: sending first capability information to the second network element, the first capability information is information inherent to the terminal and remains unchanged; periodically sending second capability information of the terminal to the access network device, the second capability information of the terminal is capability information that changes in real time; the first capability information and the second capability information are used by the first network element to divide the AI ​​task into multiple AI subtasks.

[0373] In some embodiments, multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; wherein, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the terminal; the second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the access network device; the third AI subtask is a subtask processed by the first network element, and is the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask.

[0374] In some embodiments, the method further includes: sending the processing result of the first AI subtask to the access network device; wherein, the processing result of the first AI subtask is used by the access network device to process the second AI subtask to obtain the processing result of the second AI subtask, and the processing result of the second AI subtask is used by the first network element to process the third AI subtask to obtain the processing result of the AI ​​task.

[0375] In some embodiments, the first capability information includes at least one of the following: computing capability of the terminal; resource information of the terminal; computing capability of the access network device; resource information of the access network device.

[0376] In some embodiments, the second capability information includes at least one of the following: real-time load of the terminal; resource occupancy information of the terminal; real-time load of the access network device; resource occupancy information of the access network device.

[0377] FIG5A is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG5A , the present disclosure embodiment relates to a task processing method, which includes:

[0378] Step S5101: Receive second capability information of the terminal.

[0379] The optional implementation of step S5101 can refer to the optional implementation of step S2104 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0380] Step S5102: Send second capability information.

[0381] The optional implementation of step S5102 can refer to the optional implementation of step S2105 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0382] Step S5103: Receive the configured second AI subtask.

[0383] The optional implementation of step S5103 can refer to the optional implementation of step S2108 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0384] Step S5104: Receive the processing result of the first AI subtask.

[0385] The optional implementation of step S5104 can refer to the optional implementation of step S2111 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0386] Step S5105: Process the second AI subtask.

[0387] The optional implementation of step S5105 can refer to the optional implementation of step S2112 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0388] Step S5106: Send the second AI subtask processing result.

[0389] The optional implementation of step S5106 can refer to the optional implementation of step S2113 in Figure 2A and other related parts in the embodiment involved in Figure 2A, which will not be repeated here.

[0390] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S5101 to S5106. For example, step S5101 can be implemented as an independent embodiment, step S5102 can be implemented as an independent embodiment, step S5103 can be implemented as an independent embodiment, step S5104 can be implemented as an independent embodiment, step S5105 can be implemented as an independent embodiment, step S5106 can be implemented as an independent embodiment, step S5101 + step S5102 can be implemented as an independent embodiment, step S5103 + step S5104 can be implemented as an independent embodiment, and step S5105 + step S5106 can be implemented as an independent embodiment, but the present invention is not limited thereto.

[0391] In some embodiments, step S5102 and step S5103 may be executed in an interchanged order or simultaneously.

[0392] In some embodiments, step S5101, step S5102, step S5103, step S5104, and step S5105 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0393] In some embodiments, step S5101, step S5102, step S5103, step S5104, and step S5106 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0394] In some embodiments, step S5101, step S5102, step S5104, step S5105, and step S5106 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0395] In some embodiments, step S5101, step S5103, step S5104, step S5105, and step S5106 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0396] In some embodiments, step S5102, step S5103, step S5104, step S5105, and step S5106 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0397] FIG5B is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG5B , the embodiment of the present disclosure relates to a task processing method, which includes:

[0398] Step S5201: Receive the configured second AI subtask.

[0399] The optional implementation of step S5201 can be found in steps S2104, S2105, and S2108 of Figure 2A , the optional implementation of steps S5101, S5102, and S5103 of Figure 5A , and other related parts in the embodiments involved in Figures 2A and 5A , which will not be repeated here.

[0400] Step S5202: Process the second AI subtask.

[0401] The optional implementation of step S5202 can be found in step S2111 and step S2112 of Figure 2A, the optional implementation of step S5104 and step S5105 of Figure 5A, and other related parts in the embodiments involved in Figures 2A and 5A, which will not be repeated here.

[0402] Step S5203: Send the second AI subtask processing result.

[0403] The optional implementation of step S5203 can be found in step S2113 of FIG. 2A , the optional implementation of step S5106 of FIG. 5A , and other related parts in the embodiments involved in FIG. 2A and FIG. 5A , which will not be described in detail here.

[0404] The task processing method involved in the embodiments of the present disclosure may include at least one of steps S5201 to S5203. For example, step S5201 may be implemented as an independent embodiment, step S5202 may be implemented as an independent embodiment, and step S5202 + step S5203 may be implemented as independent embodiments, but the present invention is not limited thereto.

[0405] In some embodiments, step S5201 and step S5202 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0406] In some embodiments, step S5201 and step S5203 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0407] In some embodiments, step S5202 and step S5203 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0408] FIG5C is a flow chart of a task processing method according to an embodiment of the present disclosure. As shown in FIG5C , the embodiment of the present disclosure relates to a task processing method, which includes:

[0409] In step S5301, the access network device processes the second AI subtask to be processed based on the scheduling of the first network element.

[0410] In some embodiments, the second AI subtask is a subtask processed by the access network device among a plurality of subtasks obtained by dividing the AI ​​task to be processed by the first network element.

[0411] The optional implementation of step S5301 can be found in steps S2104, S2105, S2108, S2111, S2112, and S2113 of Figure 2A , the optional implementation of steps S5101 to S5106 of Figure 5A , the optional implementation of steps S5201 to S5203 of Figure 5B , and other related parts in the embodiments involved in Figures 2A , 5A , and 5B , which will not be repeated here.

[0412] In some embodiments, the method also includes: sending first capability information to the second network element, the first capability information being inherent and unchanging information of the access network device; periodically obtaining second capability information of the terminal sent by the terminal, and sending the second capability information of the terminal and the second capability information of the access network device to the first network element; the second capability information of the access network device is the capability information of the access network device that changes in real time, and the second capability information of the terminal is the capability information of the terminal that changes in real time; the first capability information and the second capability information are used by the first network element to divide the AI ​​task into multiple AI subtasks.

[0413] In some embodiments, multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; wherein, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the terminal; the second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the access network device; the third AI subtask is a subtask processed by the first network element, and is the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask.

[0414] In some embodiments, the method further includes: obtaining a processing result of a first AI subtask sent by the terminal; processing a second AI subtask based on the processing result of the first AI subtask, obtaining a processing result of the second AI subtask, and sending the processing result of the second AI subtask to the first network element; wherein the processing result of the second AI subtask is used by the first network element to process a third AI subtask to obtain a processing result of the AI ​​task.

[0415] In some embodiments, the first capability information includes at least one of the following: computing capability of the terminal; resource information of the terminal; computing capability of the access network device; resource information of the access network device.

[0416] In some embodiments, the second capability information includes at least one of the following: real-time load of the terminal; resource occupancy information of the terminal; real-time load of the access network device; resource occupancy information of the access network device.

[0417] FIG6 is an interactive diagram of a task processing method according to an embodiment of the present disclosure. As shown in FIG6 , the present disclosure embodiment relates to a task processing method, which includes:

[0418] Step S6101: The first network element 1031 obtains an AI task to be processed.

[0419] In some embodiments, the optional implementation of step S6101 can refer to the optional implementation of step S2101 in Figure 2A, step S3101 in Figure 3A, step S3201 in Figure 3B, and step S3301 in Figure 3C, as well as other related parts in the embodiments involved in Figures 2A, 3A, 3B and 3C, which will not be repeated here.

[0420] In step S6102 , the first network element 1031 divides the AI ​​task into multiple AI subtasks.

[0421] In some embodiments, the optional implementation of step S6102 can refer to the optional implementation of steps S2102 to S2115 of Figure 2A, steps S3102 to S3111 of Figure 3A, steps S3202 to S3205 of Figure 3B, and step S3301 of Figure 3C, as well as other related parts in the embodiments involved in Figures 2A, 3A, 3B and 3C, which will not be repeated here.

[0422] In step S6103 , the first network element 1031 schedules the terminal 101 and the access network device 102 to process multiple AI tasks together with the first network element 1031 .

[0423] In some embodiments, the optional implementation of step S6101 can refer to steps S2104, S2107, S2110, and S2111 of Figure 2A , steps S4101, S4102, S4103, and S4104 of Figure 4A , steps S4201 and S4202 of Figure 4B , the optional implementation of step S4301 of Figure 4C , and other related parts of the embodiments involved in Figures 2A , 4A , 4B , and 4C , which will not be repeated here.

[0424] The optional implementation of step S6101 can be found in steps S2104, S2105, S2108, S2111, S2112, and S2113 of Figure 2A , the optional implementation of steps S5101 to S5106 of Figure 5A , steps S5201 to S5203 of Figure 5B , the optional implementation of step S5301 of Figure 5C , and other related parts in the embodiments involved in Figures 2A , 5A , 5B , and 5C , which will not be repeated here.

[0425] In some embodiments, the above method may include the method described in the above embodiments of the communication system side, terminal side, access network device side, core network device side, first network element side, second network element side, etc., which will not be repeated here.

[0426] In some embodiments, a task processing method is also provided for AI task splitting learning based on the proposed 6G network architecture.

[0427] For example, AI computing tasks are distributed to terminals, base stations, and the network for joint training through distributed endogenous intelligence, with each network node responsible for several parts of a large AI model. The core network's intelligent computing function objectively controls the scale of the neural network executed by the UE and base station by adjusting the partition points of the AI ​​model. The network can adjust the AI ​​model partition points based on information such as UE and base station power consumption, air interface resources, and channel environment, flexibly adjust the location of intermediate layer outputs, and achieve efficient task collaborative scheduling.

[0428] It can be understood that, through the embodiments of the present disclosure, external network intelligence based on the Network Data Analytic Function (NWDAF) is converted into distributed local intelligence, effectively utilizing the computing and storage resources of network devices or terminals.

[0429] It can also be understood that the neural network layers responsible for terminals, access network devices, and core network devices are adjusted by using split learning.

[0430] It can also be understood that the computing resources of each network node can be fully utilized, and flexible computing allocation can be achieved through task scheduling of core network elements.

[0431] FIG7 is an interactive diagram of a task processing method according to an embodiment of the present disclosure. As shown in FIG7 , the present disclosure embodiment relates to a task processing method, which includes:

[0432] Step S7101, task import (Task Import).

[0433] In some embodiments, the task input is, for example, an AI task input.

[0434] Exemplarily, a third-party user imports an (AI) task into NICF to request an AI task flow, and carries a task description and performance requirements.

[0435] Step S7102, initial analysis.

[0436] In some embodiments, NICF performs preliminary analysis on the AI ​​task to determine the next strategy.

[0437] Step S7103: Information Retrieval.

[0438] In some embodiments, the information retrieval is, for example, terminal information retrieval and / or access network device information retrieval.

[0439] Exemplarily, the NICF sends a capability report (Ability_Request) to the NDRF to obtain computing capability and resource information of the terminal or access network equipment (UE / RAN), and then the NDRF retrieves the relevant information and responds to the NICF.

[0440] Step S7104: Ability Reporting.

[0441] Exemplarily, the terminal periodically reports its capability information (including real-time load and resource occupancy) to the access network device via the air interface, and then the access network device aggregates all capability information including its own capabilities and reports it to the NICF.

[0442] Step S7105, initial task deployment (Intial Task Deployment).

[0443] Exemplarily, the NICF analyzes the reported capability information and initializes the task scheduling settings, including task description, task ID, task configuration information, QoAIS indicators, and allocates the terminals, access network devices, and neural network layers that the NICF needs for training and inference.

[0444] Step S7106, acquire model or data (Acquire Model / data).

[0445] Exemplarily, the terminal, access network device, and NICF obtain NDRF and data for the AI ​​model for training respectively.

[0446] Step S7107: Model Splitting.

[0447] Exemplarily, NICF determines task deployment information and makes specific model splitting decisions, i.e., which neural network layers each network node is responsible for.

[0448] Step S7108, inference.

[0449] In some embodiments, the terminal performs inference.

[0450] Exemplarily, the terminal trains the neural network layer based on NICF's task allocation and performs inference.

[0451] Step S7109, result output (Result Output).

[0452] Exemplarily, the terminal outputs the training result to the access network device through a wireless channel.

[0453] Step S7110, inference.

[0454] In some embodiments, the access network device performs inference.

[0455] Exemplarily, the access network device takes the output of the terminal as input, and then trains the neural network layer based on the task allocation of NICF and performs inference.

[0456] Step S7111, result output (Result Output).

[0457] Exemplarily, NG-RAN outputs the inference result to NICF.

[0458] Step S7112, inference.

[0459] In some embodiments, the NICF performs inference.

[0460] For example, NICF takes the output of access network devices as input and then uses its computing power to train the remaining layers of the complex neural network model.

[0461] Step S7113: Status reporting and task scheduling.

[0462] For example, the terminal and access network equipment periodically report their status (including CSI, link bandwidth resources, channel environment, etc.) to the NDCF. The NDCF obtains real-time network load and congestion conditions, summarizes the status information, and sends them to the NICF. The NICF dynamically adjusts the allocation of neural network layers and model parameters based on the information reported by the NDCF. For example, if the computing resources or computing power of the terminal are insufficient to support it to complete a large number of computing tasks, the NICF will reduce the number of model layers responsible for the terminal and be responsible for layers with fewer neurons.

[0463] Step S7114, inference result (Inference Result).

[0464] For example, after completing multiple rounds of AI model training, NICF sends the final inference results to NDRF.

[0465] Step S7115, return result (Return Result).

[0466] Exemplarily, NDRF stores training and inference results, converts them into formats, and sends them to third-party users.

[0467] It should be noted that the access and mobility management function (AMF), NDCF, NICF, and NDRF in Figure 7 respectively represent modules with corresponding functions in the core network equipment.

[0468] In the embodiments of the present disclosure, some or all of the steps and their optional implementations may be arbitrarily combined with some or all of the steps in other embodiments, or may be arbitrarily combined with the optional implementations of other embodiments.

[0469] The embodiments of the present disclosure further provide an apparatus for implementing any of the above methods. For example, an apparatus is provided, comprising units or modules for implementing each step performed by a terminal in any of the above methods. For another example, another apparatus is provided, comprising units or modules for implementing each step performed by a network device (e.g., an access network device, a core network function node, a core network device, etc.) in any of the above methods.

[0470] It should be understood that the division of the various units or modules in the above device is merely a division of logical functions. In actual implementation, they may be fully or partially integrated into a physical entity, or they may be physically separated. In addition, the units or modules in the device may be implemented in the form of a processor calling software: for example, the device includes a processor, the processor is connected to a memory, and the memory stores instructions. The processor calls the instructions stored in the memory to implement any of the above methods or implement the functions of the various units or modules of the above device, wherein the processor is, for example, a general-purpose processor, such as a central processing unit (CPU) or a microprocessor, and the memory is a memory within the device or a memory outside the device. Alternatively, the units or modules in the device can be implemented in the form of hardware circuits, and the functions of some or all of the units or modules can be realized by designing the hardware circuits. The above-mentioned hardware circuits can be understood as one or more processors; for example, in one implementation, the above-mentioned hardware circuit is an application-specific integrated circuit (ASIC), which realizes the functions of some or all of the above units or modules by designing the logical relationship of the components in the circuit; for example, in another implementation, the above-mentioned hardware circuit can be realized by a programmable logic device (PLD). Taking a field programmable gate array (FPGA) as an example, it can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by configuring the configuration file, thereby realizing the functions of some or all of the above units or modules. All units or modules of the above devices can be realized in the form of software called by the processor, or in the form of hardware circuits, or in part by the form of software called by the processor, and the rest by hardware circuits.

[0471] In the embodiments of the present disclosure, the processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), or a digital signal processor (DSP). In another implementation, the processor can implement certain functions through the logical relationship of the hardware circuit. The logical relationship of the above-mentioned hardware circuit is fixed or reconfigurable. For example, the processor is a hardware circuit implemented by an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and implementing the hardware circuit configuration can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units or modules. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), etc.

[0472] Figure 8A is a schematic diagram of the structure of the first network element proposed in an embodiment of the present disclosure. As shown in Figure 8A, the first network element 8100 may include at least one of: a transceiver module 8101, a processing module 8102, etc. In some embodiments, the transceiver module is used to obtain an AI task to be processed. In some embodiments, the processing module is used to divide the AI ​​task into multiple AI subtasks; the scheduling terminal and access network device jointly process the multiple AI subtasks to obtain the processing results of the AI ​​task. Optionally, the transceiver module is used to perform at least one of the communication steps such as sending and / or receiving performed by the first network element 1031 in any of the above methods (for example, step S2101, step S2102, step S2103, step S2105, step S2107, step S2108, step S2113, and step S2115, but not limited thereto), which will not be repeated here. Optionally, the processing module is used to divide the AI ​​task into multiple AI subtasks, and is also used to schedule the terminal and access network device to jointly process the multiple AI subtasks to obtain the processing results of the AI ​​task. The above-mentioned processing module is also used to execute at least one of the other steps (such as step S2106, step S3109 and step S2114, but not limited to these) performed by the first network element 1031 in any of the above methods, which will not be repeated here.

[0473] In some embodiments, the processing module 8102 divides the AI ​​task into multiple AI subtasks in the following manner: sending a capability request to the second network element, and obtaining first capability information sent by the second network element in response to the capability request, the first capability information being the inherent and unchanging information of the terminal and the inherent and unchanging capability information of the access network device; obtaining second capability information sent by the access network device, the second capability information including the second capability information of the access network device and the second capability information of the terminal periodically sent to the access network device, the second capability information being the capability information that changes in real time of the terminal and the capability information that changes in real time of the access network device; based on the first capability information and the second capability information, dividing the AI ​​task into multiple AI subtasks.

[0474] In some embodiments, the processing module 8102 divides the AI ​​task into multiple AI subtasks based on the first capability information and the second capability information in the following manner, including: based on the first capability information corresponding to the terminal and the second capability information corresponding to the terminal, dividing the AI ​​task to obtain the first AI subtask to be processed by the terminal; based on the first capability information corresponding to the access network device and the second capability information corresponding to the access network device, dividing the AI ​​task to obtain the second AI subtask to be processed by the access network device; and using the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask as the third AI subtask to be processed by the first network element.

[0475] In some embodiments, the processing module 8102 schedules the terminal and the access network device in the following manner to jointly process multiple AI subtasks to obtain the processing results of the AI ​​tasks: obtain the second AI subtask processing result sent by the access network device, the second AI subtask processing result is the result of the access network device processing the second AI subtask based on the first AI subtask processing result, and the first AI subtask processing result is the result of the terminal processing the first AI subtask and sending it to the access network device; process the third AI subtask based on the second AI subtask processing result to obtain the processing result of the AI ​​task.

[0476] In some embodiments, the transceiver module 8101 is also used to periodically obtain a status report sent by the second network element, where the status report includes second capability information of the access network device and the terminal. The processing module 8102 is also used to re-divide the AI ​​task when the second capability information meets preset conditions to obtain multiple re-divide AI subtasks.

[0477] In some embodiments, the transceiver module 8101 is further used to obtain the processing result of the AI ​​task and send it to the second network element, which is used to store the processing result.

[0478] In some embodiments, the first capability information includes at least one of the following: computing capability of the terminal; resource information of the terminal; computing capability of the access network device; resource information of the access network device.

[0479] In some embodiments, the second capability information includes at least one of the following: real-time load of the terminal; resource occupancy information of the terminal; real-time load of the access network device; resource occupancy information of the access network device.

[0480] Figure 8B is a structural diagram of the terminal proposed in an embodiment of the present disclosure. As shown in Figure 8B, the terminal 8200 may include: a transceiver module 8201, and a processing module 8202. In some embodiments, the transceiver module is used to obtain the scheduling of the first network element. Optionally, the transceiver module is used to execute at least one of the communication steps such as sending and / or receiving (for example, step S2104, step S2107, step S2111, but not limited to this) performed by the terminal 101 in any of the above methods, which will not be repeated here. Optionally, the processing module is used to process the AI ​​task to be processed based on the scheduling of the first network element to obtain the processing result of the AI ​​task. At least one of the other steps (for example, step S2110, but not limited to this) performed by the terminal 101 in any of the above methods is executed, which will not be repeated here.

[0481] In some embodiments, the transceiver module 8201 is also used to send first capability information to the second network element, where the first capability information is inherent and unchanging information of the terminal; and periodically send second capability information of the terminal to the access network device, where the second capability information of the terminal is capability information that changes in real time; the first capability information and the second capability information are used by the first network element to divide the AI ​​task into multiple AI subtasks.

[0482] In some embodiments, multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; wherein, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the terminal; the second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the access network device; the third AI subtask is a subtask processed by the first network element, and is the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask.

[0483] In some embodiments, the transceiver module 8201 is also used to: send the processing result of the first AI subtask to the access network device; wherein, the processing result of the first AI subtask is used by the access network device to process the second AI subtask to obtain the processing result of the second AI subtask, and the processing result of the second AI subtask is used by the first network element to process the third AI subtask to obtain the processing result of the AI ​​task.

[0484] In some embodiments, the method also includes: when the second capability information meets a preset condition, the first network element re-divides the AI ​​task to obtain multiple re-divide AI sub-tasks, the second capability information is determined by the first network element periodically obtaining a status report sent by the second network element, and the second capability information includes: the second capability information of the access network device and the terminal.

[0485] In some embodiments, the processing result of the AI ​​task is sent from the first network element to the second network element and stored in the second network element.

[0486] In some embodiments, the first capability information includes at least one of the following: computing capability of the terminal; resource information of the terminal; computing capability of the access network device; resource information of the access network device.

[0487] In some embodiments, the second capability information includes at least one of the following: real-time load of the terminal; resource occupancy information of the terminal; real-time load of the access network device; resource occupancy information of the access network device.

[0488] Figure 8C is a schematic diagram of the structure of the access network device proposed in an embodiment of the present disclosure. As shown in Figure 8C, the access network device 8300 may include: a transceiver module 8301 and a processing module 8302. In some embodiments, the transceiver module is used to obtain the scheduling of the first network element. Optionally, the transceiver module is used to execute at least one of the communication steps such as sending and / or receiving (for example, step S2104, step S2105, step S2108, step S2111, step S2113, but not limited thereto) performed by the access network device 102 in any of the above methods, which will not be repeated here. Optionally, the processing module is used to process the AI ​​task to be processed based on the scheduling of the first network element to obtain the processing result of the AI ​​task. At least one of the other steps (for example, step S2112, but not limited thereto) performed by the access network device 102 in any of the above methods is executed, which will not be repeated here.

[0489] In some embodiments, the method also includes: sending first capability information to the second network element, the first capability information being inherent and unchanging information of the access network device; periodically obtaining second capability information of the terminal sent by the terminal, and sending the second capability information of the terminal and the second capability information of the access network device to the first network element; the second capability information of the access network device is the capability information of the access network device that changes in real time, and the second capability information of the terminal is the capability information of the terminal that changes in real time; the first capability information and the second capability information are used by the first network element to divide the AI ​​task into multiple AI subtasks.

[0490] In some embodiments, multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; wherein, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the terminal; the second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI ​​task based on the first capability information and the second capability information corresponding to the access network device; the third AI subtask is a subtask processed by the first network element, and is the remaining tasks in the AI ​​task except the first AI subtask and the second AI subtask.

[0491] In some embodiments, the transceiver module 8301 is also used to: obtain the processing result of the first AI subtask sent by the terminal; process the second AI subtask based on the processing result of the first AI subtask, obtain the processing result of the second AI subtask, and send the processing result of the second AI subtask to the first network element; wherein the processing result of the second AI subtask is used by the first network element to process the third AI subtask and obtain the processing result of the AI ​​task.

[0492] In some embodiments, the first capability information includes at least one of the following: computing capability of the terminal; resource information of the terminal; computing capability of the access network device; resource information of the access network device.

[0493] In some embodiments, the second capability information includes at least one of the following: real-time load of the terminal; resource occupancy information of the terminal; real-time load of the access network device; resource occupancy information of the access network device.

[0494] In some embodiments, the processing module can be a single module or can include multiple submodules. Optionally, the multiple submodules respectively execute all or part of the steps required to be executed by the processing module. Optionally, the processing module can be interchangeable with the processor.

[0495] Figure 9A is a schematic diagram of the structure of a communication device 9100 proposed in an embodiment of the present disclosure. Communication device 9100 can be a network device (e.g., an access network device, a core network device, etc.), a terminal (e.g., a user equipment, etc.), a chip, a chip system, or a processor that supports a network device to implement any of the above methods, or a chip, a chip system, or a processor that supports a terminal to implement any of the above methods. Communication device 9100 can be used to implement the methods described in the above method embodiments. For details, please refer to the description of the above method embodiments.

[0496] As shown in Figure 9A, the communication device 9100 includes one or more processors 9101. The processor 9101 can be a general-purpose processor or a dedicated processor, for example, a baseband processor or a central processing unit. The baseband processor can be used to process the communication protocol and communication data, and the central processing unit can be used to control the communication device (such as a base station, a baseband chip, a terminal device, a terminal device chip, a DU or a CU, etc.), execute programs, and process program data. Optionally, the communication device 9100 is used to perform any of the above methods. Optionally, one or more processors 9101 are used to call instructions to enable the communication device 9100 to perform any of the above methods.

[0497] In some embodiments, the communication device 9100 further includes one or more transceivers 9102. When the communication device 9100 includes one or more transceivers 9102, the transceiver 9102 performs at least one of the communication steps of sending and / or receiving in the above method (e.g., step S2101, step S2102, step S2103, step S2104, step S2105, step S2107, step S2108, step S2111, step S2113, and step S2115, but not limited thereto), and the processor 9101 performs at least one of the other steps (e.g., step S2106, step S2109, step S2110, step S2112, and step S2114, but not limited thereto). In an optional embodiment, the transceiver may include a receiver and / or a transmitter, and the receiver and transmitter may be separate or integrated. Optionally, terms such as transceiver, transceiver unit, transceiver, transceiver circuit, interface circuit, and interface can be replaced with each other, terms such as transmitter, transmitting unit, transmitter, and transmitting circuit can be replaced with each other, and terms such as receiver, receiving unit, receiver, and receiving circuit can be replaced with each other.

[0498] In some embodiments, the communication device 9100 further includes one or more memories 9103 for storing data. Alternatively, all or part of the memories 9103 may be located outside the communication device 9100. In alternative embodiments, the communication device 9100 may include one or more interface circuits 9104. Optionally, the interface circuits 9104 are connected to the memories 9103 and may be configured to receive data from the memories 9103 or other devices, or to send data to the memories 9103 or other devices. For example, the interface circuits 9104 may read data stored in the memories 9103 and send the data to the processor 9101.

[0499] The communication device 9100 described in the above embodiments may be a network device or a terminal, but the scope of the communication device 9100 described in the present disclosure is not limited thereto, and the structure of the communication device 9100 may not be limited by FIG. 9A. The communication device may be an independent device or may be part of a larger device. For example, the communication device may be: 1) an independent integrated circuit IC, or a chip, or a chip system or subsystem; (2) a collection of one or more ICs, optionally, the above IC collection may also include a storage component for storing data or programs; (3) an ASIC, such as a modem; (4) a module that can be embedded in other devices; (5) a receiver, a terminal device, an intelligent terminal device, a cellular phone, a wireless device, a handheld device, a mobile unit, an in-vehicle device, a network device, a cloud device, an artificial intelligence device, etc.; (6) others, etc.

[0500] 9B is a schematic diagram of the structure of a chip 9200 according to an embodiment of the present disclosure. If the communication device 9100 can be a chip or a chip system, please refer to the schematic diagram of the structure of the chip 9200 shown in FIG9B , but the present disclosure is not limited thereto.

[0501] The chip 9200 includes one or more processors 9201. The chip 9200 is configured to execute any of the above methods.

[0502] In some embodiments, chip 9200 further includes one or more interface circuits 9202. Terms such as interface circuit, interface, and transceiver pins may be used interchangeably. In some embodiments, chip 9200 further includes one or more memories 9203 for storing data. Alternatively, all or part of memory 9203 may be located external to chip 9200. Optionally, interface circuit 9202 is connected to memory 9203 and may be used to receive data from memory 9203 or other devices, or may be used to send data to memory 9203 or other devices. For example, interface circuit 9202 may read data stored in memory 9203 and send the data to processor 9201.

[0503] In some embodiments, the interface circuit 9202 performs at least one of the communication steps (e.g., steps S2101, S2102, S2103, S2104, S2105, S2107, S2108, S2111, S2113, and S2115) of the aforementioned method. For example, the interface circuit 9202 performing the communication steps (e.g., steps S2101, S2102, S2103, and S2115) of the aforementioned method means that the interface circuit 9202 performs data exchange between the processor 9201, the chip 9200, the memory 9203, or the transceiver device. In some embodiments, the processor 9201 performs at least one of the other steps (e.g., steps S2106, S2109, S2110, S2112, and S2114, but not limited thereto).

[0504] The modules and / or devices described in various embodiments, such as virtual devices, physical devices, and chips, can be arbitrarily combined or separated according to circumstances. Optionally, some or all steps can also be performed collaboratively by multiple modules and / or devices, which is not limited here.

[0505] The present disclosure also proposes a storage medium having instructions stored thereon, which, when executed on the communication device 9100, causes the communication device 9100 to execute any of the above methods. Optionally, the storage medium is an electronic storage medium. Optionally, the storage medium is a computer-readable storage medium, but is not limited thereto and may also be a storage medium readable by other devices. Optionally, the storage medium may be a non-transitory storage medium, but is not limited thereto and may also be a temporary storage medium.

[0506] The present disclosure also provides a program product, which, when executed by the communication device 9100, enables the communication device 9100 to perform any of the above methods. Optionally, the program product is a computer program product.

[0507] The present disclosure also proposes a computer program, which, when executed on a computer, causes the computer to perform any one of the above methods.

Claims

1. A task processing method, characterized in that, The method includes: A first network element obtains an artificial intelligence (AI) task to be processed; The AI task is segmented into multiple AI subtasks; A scheduling terminal and an access network device jointly process the multiple AI subtasks to obtain a processing result of the AI task.

2. The method according to claim 1, wherein The segmenting the AI task into multiple AI subtasks includes: Sending a capability request to a second network element and obtaining first capability information sent by the second network element in response to the capability request, where the first capability information is information that is inherent and unchanged for the terminal and capability information that is inherent and unchanged for the access network device; Obtaining second capability information sent by the access network device, where the second capability information includes second capability information of the access network device and second capability information of the terminal periodically sent by the terminal to the access network device, and the second capability information is capability information that changes in real time for the terminal and capability information that changes in real time for the access network device; Based on the first capability information and the second capability information, the AI task is segmented into multiple AI subtasks.

3. The method according to claim 2, wherein The segmenting the AI task into multiple AI subtasks based on the first capability information and the second capability information includes: Based on the first capability information corresponding to the terminal and the second capability information corresponding to the terminal, a first AI subtask to be processed by the terminal is segmented out of the AI task; Based on the first capability information corresponding to the access network device and the second capability information corresponding to the access network device, a second AI subtask to be processed by the access network device is segmented out of the AI task; The remaining task in the AI task other than the first AI subtask and the second AI subtask is used as a third AI subtask to be processed by the first network element.

4. The method according to claim 3, wherein The scheduling the terminal and the access network device to jointly process the multiple AI subtasks to obtain the processing result of the AI task includes: Obtaining a processing result of the second AI subtask sent by the access network device, where the processing result of the second AI subtask is the result of the access network device processing the second AI subtask based on a processing result of the first AI subtask, and the processing result of the first AI subtask is the result of the terminal processing the first AI subtask and sending it to the access network device; Based on the processing result of the second AI subtask, the third AI subtask is processed to obtain the processing result of the AI task.

5. The method according to any one of claims 1 to 4, characterized in that The method further includes: Periodically obtaining a status report sent by the second network element, where the status report includes the second capability information of the access network device and the terminal; When the second capability information meets a preset condition, the AI task is re-segmented to obtain multiple re-segmented AI subtasks.

6. The method according to claim 1, wherein After the first network element, the terminal, and the access network device jointly process the multiple AI subtasks, it further includes: Obtaining the processing result of the AI task and sending it to the second network element, and the second network element is used to store the processing result.

7. The method according to claim 2, wherein The first capability information includes at least one of the following: The computing power of the terminal; The resource information of the terminal; The computing power of the access network device; The resource information of the access network device.

8. The method according to claim 2, characterized in that, The second capability information includes at least one of the following: The real-time load of the terminal; The resource occupancy information of the terminal; The real-time load of the access network device; The resource occupancy information of the access network device.

9. A task processing method, characterized in that, The method includes: The terminal processes a first artificial intelligence (AI) sub-task based on the scheduling of a first network element; Wherein, the first AI sub-task is a sub-task processed by the terminal among the multiple sub-tasks obtained by the first network element splitting the AI task to be processed.

10. The method according to claim 9, characterized in that, The method further includes: Sending first capability information to a second network element, where the first capability information is information that is inherent and unchanged for the terminal; Periodically sending the second capability information of the terminal to the access network device, where the second capability information of the terminal is information about the capabilities of the terminal that changes in real time; The first capability information and the second capability information are used by the first network element to split the AI task into multiple AI sub-tasks.

11. The method according to claim 10, wherein The multiple AI sub-tasks include: a first AI sub-task, a second AI sub-task, and a third AI sub-task; Wherein, the first AI sub-task is a sub-task processed by the terminal and is obtained by the first network element splitting the AI task based on the first capability information and the second capability information corresponding to the terminal; The second AI sub-task is a sub-task processed by the access network device and is obtained by the first network element splitting the AI task based on the first capability information and the second capability information corresponding to the access network device; The third AI sub-task is a sub-task processed by the first network element and is the remaining task in the AI task except for the first AI sub-task and the second AI sub-task.

12. The method according to claim 11, wherein The method further includes: Sending the processing result of the first AI sub-task to the access network device; Wherein, the processing result of the first AI sub-task is used by the access network device to process the second AI sub-task to obtain the processing result of the second AI sub-task, and the processing result of the second AI sub-task is used by the first network element to process the third AI sub-task to obtain the processing result of the AI task.

13. The method according to claim 10, wherein The first capability information includes at least one of the following: The computing power of the terminal; The resource information of the terminal; The computing power of the access network device; The resource information of the access network device.

14. The method according to claim 10, wherein The second capability information includes at least one of the following: The real-time load of the terminal; The resource occupancy information of the terminal; The real-time load of the access network device; The resource occupancy information of the access network device.

15. A task processing method, characterized in that, The method includes: The access network device processes a second artificial intelligence (AI) sub-task based on the scheduling of a first network element; Wherein, the second AI sub-task is a sub-task processed by the access network device among the multiple sub-tasks obtained by the first network element splitting the AI task to be processed.

16. The method according to claim 15, characterized in that, The method further includes: Sending first capability information to a second network element, where the first capability information is information that is inherent and unchanged for the access network device; Periodically obtain the second capability information of the terminal sent by the terminal, and send the second capability information of the terminal and the second capability information of the access network device to the first network element; The second capability information of the access network device is the capability information that changes in real time of the access network device, and the second capability information of the terminal is the capability information that changes in real time of the terminal; The first capability information and the second capability information are used for the first network element to split the AI task into multiple AI subtasks.

17. The method according to claim 16, wherein The multiple AI subtasks include: a first AI subtask, a second AI subtask, and a third AI subtask; Among them, the first AI subtask is a subtask processed by the terminal, and is obtained by the first network element splitting the AI task based on the first capability information and the second capability information corresponding to the terminal; The second AI subtask is a subtask processed by the access network device, and is obtained by the first network element splitting the AI task based on the first capability information and the second capability information corresponding to the access network device; The third AI subtask is a subtask processed by the first network element, and is the remaining task in the AI task except the first AI subtask and the second AI subtask.

18. The method according to claim 17, characterized in that, The method further includes: Obtain the processing result of the first AI subtask sent by the terminal; Process the second AI subtask based on the processing result of the first AI subtask to obtain the processing result of the second AI subtask, and send the processing result of the second AI subtask to the first network element; Among them, the processing result of the second AI subtask is used for the first network element to process the third AI subtask to obtain the processing result of the AI task.

19. The method according to claim 16, wherein The first capability information includes at least one of the following: The computing power of the terminal; The resource information of the terminal; The computing power of the access network device; The resource information of the access network device.

20. The method according to claim 17, wherein The second capability information includes at least one of the following: The real-time load of the terminal; The resource occupancy information of the terminal; The real-time load of the access network device; The resource occupancy information of the access network device.

21. A task processing method, characterized in that, The method includes: The first network element obtains an artificial intelligence (AI) task to be processed; Split the AI task into multiple AI subtasks; The terminal and the access network device, based on the scheduling of the first network element, jointly process the multiple AI subtasks with the first network element to obtain the processing result of the AI task.

22. A first network element, characterized in that, Includes: A transceiver module, configured to obtain an artificial intelligence (AI) task to be processed; A processing module, configured to split the AI task into multiple AI subtasks; The processing module is further configured to schedule the terminal and the access network device to jointly process the multiple AI subtasks to obtain the processing result of the AI task.

23. A terminal, characterized in that, Includes: A processing module, configured to process the first artificial intelligence (AI) subtask based on the scheduling of the first network element; where the first AI subtask is a subtask processed by the terminal among the multiple subtasks obtained by the first network element splitting the AI task to be processed.

24. An access network device, characterized in that, Includes: A processing module, configured to process a second artificial intelligence (AI) sub-task based on the scheduling of a first network element; wherein, the second AI sub-task is a sub-task processed by the access network device among multiple sub-tasks obtained by splitting an AI task to be processed by the first network element.

25. A core network device, characterized in that, Comprising: One or more processors; Wherein, the processor is configured to execute the task processing method according to any one of claims 1 to 8.

26. A terminal, characterized in that, Comprising: One or more processors; Wherein, the processor is configured to execute the task processing method according to any one of claims 9 to 16.

27. An access network device, characterized in that, Comprising: One or more processors; Wherein, the processor is configured to execute the task processing method according to any one of claims 17 to 24.

28. A communication system, characterized in that, Comprising: A core network device, a terminal, and a network device, wherein the core network device is configured to implement the task processing method according to any one of claims 1 to 8, the terminal is configured to implement the task processing method according to any one of claims 9 to 16, and the network device is configured to implement the task processing method according to any one of claims 17 to 24.

29. A storage medium, the storage medium stores instructions, characterized in that, When the instruction runs on a communication device, the communication device is caused to execute the task processing method according to any one of claims 1 to 8, 9 to 16, or 17 to 24.

Images