An information processing method, apparatus and device

Abstract

The present application provides an information processing method, device and equipment, wherein the information processing method comprises: decomposing an artificial intelligence (AI) task of a terminal into at least two different AI sub-tasks; sending primary service function chain arrangement information to a cloud computing center device according to the at least two different AI sub-tasks; receiving first information sent by the cloud computing center device; the first information comprises final service function chain chain forming decision information or confirmation information; and determining a final service function chain according to the first information. This scheme can support the implementation of the first base station and the cloud computing center device to cooperate to determine the AI service function chain arrangement based on the global perspective (second information), thereby improving the arrangement effect and solving the problem of poor arrangement effect caused by the lack of global perspective in the prior art information processing scheme for AI service function chain arrangement.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless, and in particular to an information processing method, device and equipment. BACKGROUND

[0002] With the continuous emergence of new applications, the tasks of the UE (terminal) often need to be provided by the edge network nearby. A single edge base station is limited by its computing, storage and model resources, and it is difficult to guarantee complete AI (artificial intelligence) services, so cooperation across multiple edge base stations is needed to form an AI service function chain. The training of AI models and the reasonable deployment of models, as well as the efficient and reasonable arrangement of AI service function chains, are important factors to guarantee service quality.

[0003] The current edge base station uses a local optimization heuristic algorithm to make decisions when arranging the AI service function chain. However, the AI service function chain arrangement will string together multiple edge base stations in the network topology, and needs to integrate the resources of all base stations in the network. The existing service function chain arrangement based on heuristic algorithms obtains a feasible solution to the combinatorial optimization problem through experience, lacks a global perspective, and the deviation of this feasible solution from the optimal solution cannot be predicted.

[0004] Therefore, the existing information processing scheme for AI service function chain arrangement in the prior art has the problem of poor arrangement effect due to lack of global perspective. SUMMARY

[0005] The purpose of the present application is to provide an information processing method, device and equipment to solve the problem of poor arrangement effect due to lack of global perspective in the prior art information processing scheme for AI service function chain arrangement.

[0006] To solve the above technical problems, an information processing method is provided, which is applied to a first base station and includes:

[0007] Decomposing an artificial intelligence (AI) task of a terminal into at least two different AI sub-tasks;

[0008] According to the at least two different AI sub-tasks, sending primary service function chain arrangement information to a cloud computing center device;

[0009] Receiving first information sent by the cloud computing center device; the first information includes final service function chain chaining decision information or confirmation information;

[0010] According to the first information, determining a final service function chain.

[0011] Optionally, according to the at least two different AI sub-tasks, sending primary service function chain arrangement information to a cloud computing center device includes:

[0012] Based on the second information, the service locations of each AI subtask are arranged to form a primary service function chain, resulting in primary service function chain arrangement information; wherein, the second information is the global status information corresponding to the cloud computing center equipment to which the first base station belongs; the global status information is related to the execution of AI tasks;

[0013] Send the primary service function chain orchestration information to the cloud computing center equipment.

[0014] Optionally, the second information includes at least one of the following:

[0015] Information on AI models already stored in various base stations and cloud computing center equipment;

[0016] Information on the remaining computing resources in each base station and cloud computing center equipment;

[0017] Information on the remaining storage resources in each base station and cloud computing center equipment;

[0018] Information on the physical location of each base station;

[0019] Information on the interconnection status of each base station;

[0020] Each of the base stations is connected to the cloud computing center equipment.

[0021] Optionally, the second information is inserted into the MAC layer transport block as a Media Access Control (MAC) control cell and transmitted through the transport channel;

[0022] The second information is located at the beginning of the MAC protocol data unit in the MAC layer transport block.

[0023] Optional, also includes:

[0024] Send the local status information of the first base station to the cloud computing center device;

[0025] The local state information is related to the execution of AI tasks.

[0026] Optionally, the local status information includes at least one of the following:

[0027] Information about the AI ​​model already stored in the first base station;

[0028] Information on the remaining computing resources in the first base station;

[0029] The first base station stores information about resources;

[0030] Information about the physical location of the first base station;

[0031] information of a connection status of the first base station with other base stations.

[0032] Optionally, the local state information is inserted into a MAC layer transport block as a MAC control signal element and transmitted through a transmission channel.

[0033] Optionally, the method further comprises:

[0034] performing the AI task by using the final service function chain to obtain a processing result.

[0035] sending the processing result to the terminal.

[0036] Optionally, before performing the AI task by using the final service function chain to obtain a processing result, the method further comprises:

[0037] determining a training party of an AI model corresponding to each AI subtask, and obtaining initial parameter information of the AI model by using the training party; the training party is the first base station, the cloud computing center device, or other base stations under the cloud computing center device except the first base station.

[0038] receiving final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0039] performing the AI task by using the final service function chain to obtain a processing result.

[0040] performing the AI task by using the final service function chain to obtain a processing result.

[0041] Optionally, the determining of the training party of the AI model corresponding to each AI subtask comprises:

[0042] in a case where the AI model corresponding to the AI subtask exists locally in the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station.

[0043] in a case where the AI model corresponding to the AI subtask does not exist locally in the first base station, determining whether the AI model corresponding to the AI subtask exists in the other base stations according to the second information; in a case where the AI model exists, determining that the training party of the AI model corresponding to the AI subtask is the other base stations; in a case where the AI model does not exist, determining that the training party of the AI model corresponding to the AI subtask is the cloud computing center device.

[0044] The embodiment of the application further provides an information processing method applied to a cloud computing center device, comprising:

[0045] receive primary service function chain arrangement information sent by at least one first base station;

[0046] determine whether each base station computing resource on the service function chain meets current all primary service function chain demands according to the primary service function chain arrangement information and second information, and obtain a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to execution of an AI task;

[0047] obtain first information corresponding to each first base station respectively according to the determination result; the first information includes final service function chain chaining decision information or confirmation information;

[0048] send the corresponding first information to each first base station.

[0049] Optionally, the obtaining of the first information corresponding to each first base station respectively according to the determination result includes:

[0050] in a case where the determination result indicates that each base station computing resource on the service function chain can meet current all primary service function chain demands, the first information containing the confirmation information is taken as the first information corresponding to each first base station respectively;

[0051] in a case where the determination result indicates that each base station computing resource on the service function chain cannot meet current all primary service function chain demands, the primary service function chain arrangement information is updated to determine first base stations in which primary service function chains are changed and first base stations in which primary service function chains remain unchanged; the first information containing target decision information is taken as the first information corresponding to the first base stations in which the primary service function chains are changed; the first information containing the confirmation information is taken as the first information corresponding to the first base stations in which the primary service function chains remain unchanged; the target decision information includes final service function chain chaining decision information corresponding to the first base stations in which the primary service function chains are changed.

[0052] Optionally, the method further includes:

[0053] sending the second information to each base station on the service function chain respectively;

[0054] The second information includes at least one of the following:

[0055] information of an AI model stored in each base station and the cloud computing center device;

[0056] information of remaining computing resources in each base station and the cloud computing center device;

[0057] information of remaining storage resources in each base station and the cloud computing center device;

[0058] information of physical locations of the respective base stations;

[0059] information of interconnections of the respective base stations;

[0060] The respective base stations are connected with the cloud computing center device respectively.

[0061] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

[0062] The second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0063] Optionally, the method further comprises:

[0064] receiving local state information of the respective base stations transmitted by the respective base stations respectively on the service function chain, wherein the local state information is related to execution of an AI task;

[0065] obtaining the second information according to the local state information.

[0066] Optionally, the local state information comprises at least one of the following:

[0067] information of an AI model stored by the base station;

[0068] information of remaining computing resources in the base station;

[0069] information of storage resources in the base station;

[0070] information of physical locations of the respective base stations;

[0071] information of interconnections of the respective base stations.

[0072] Optionally, the local state information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

[0073] Optionally, the method further comprises:

[0074] obtaining initial parameter information of at least one AI model;

[0075] performing a classification and aggregation operation on the initial parameter information to obtain final parameter information of the respective AI models;

[0076] transmitting the final parameter information of the AI models to base stations corresponding to the AI models.

[0077] Optionally, the obtaining of the initial parameter information of the at least one AI model comprises at least one of the following:

[0078] receiving initial parameter information of an AI model sent by at least one base station;

[0079] receiving an AI sub-task sent by the first base station, training a corresponding AI model according to the AI sub-task, and obtaining initial parameter information of the AI model.

[0080] Optionally, the training of the corresponding AI model according to the AI sub-task comprises:

[0081] training the corresponding AI model according to the AI sub-task by using a transfer learning manner.

[0082] Embodiments of the present application also provide an information processing method applied to a first base station, comprising:

[0083] decomposing an artificial intelligence (AI) task of a terminal into at least two different AI sub-tasks;

[0084] determining a training party of an AI model corresponding to each AI sub-task, and obtaining initial parameter information of the AI model by using the training party; the training party is the first base station, a cloud computing center device connected to the first base station, or other base stations except the first base station under the cloud computing center device;

[0085] receiving final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0086] Optionally, the determination of the training party of the AI model corresponding to each AI sub-task comprises:

[0087] in a case where the AI model corresponding to the AI sub-task exists locally in the first base station, determining that the training party of the AI model corresponding to the AI sub-task is the first base station;

[0088] in a case where the AI model corresponding to the AI sub-task does not exist locally in the first base station, determining whether the other base stations store the AI model corresponding to the AI sub-task according to second information; in a case where the AI model is stored, determining that the training party of the AI model corresponding to the AI sub-task is the other base stations; in a case where the AI model is not stored, determining that the training party of the AI model corresponding to the AI sub-task is the cloud computing center device.

[0089] wherein the second information is global state information corresponding to a cloud computing center device to which the first base station belongs; and the global state information is related to the execution of the AI task.

[0090] Optionally, the second information comprises at least one of:

[0091] Information of AI models stored in each base station and the cloud computing center device;

[0092] Information of remaining computing resources in each base station and the cloud computing center device;

[0093] Information of remaining storage resources in each base station and the cloud computing center device;

[0094] Information of physical locations of each base station;

[0095] Information of interconnections of each base station;

[0096] The base stations are connected to the cloud computing center device.

[0097] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

[0098] The second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0099] The embodiment of the application further provides an information processing method applied to a cloud computing center device, and the method comprises the following steps:

[0100] Obtaining initial parameter information of at least one AI model;

[0101] Performing a classification and aggregation operation on the initial parameter information to obtain final parameter information of each AI model;

[0102] Sending the final parameter information of the AI model to a base station corresponding to the AI model.

[0103] Optionally, the obtaining of the initial parameter information of at least one AI model comprises at least one of the following:

[0104] Receiving initial parameter information of an AI model sent by at least one base station;

[0105] Receiving an AI subtask sent by a first base station, training a corresponding AI model according to the AI subtask, and obtaining initial parameter information of the AI model.

[0106] Optionally, the training of the corresponding AI model according to the AI subtask comprises:

[0107] Training the corresponding AI model according to the AI subtask by using a transfer learning mode.

[0108] Optionally, the method further comprises:

[0109] Sending second information to at least one base station respectively;

[0110] Wherein, the second information is global state information corresponding to the cloud computing center device; the global state information is related to execution of an AI task.

[0111] Optionally, the second information includes at least one of the following:

[0112] Information of AI models stored in each base station and the cloud computing center device;

[0113] Information of remaining computing resources in each base station and the cloud computing center device;

[0114] Information of remaining storage resources in each base station and the cloud computing center device;

[0115] Information of physical locations of each base station;

[0116] Information of mutual connection conditions of each base station;

[0117] Wherein, each base station is connected with the cloud computing center device.

[0118] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel;

[0119] Wherein, the second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0120] The embodiment of the application further provides an information processing device applied to a first base station and comprising:

[0121] A first processing module configured to decompose an AI task of a terminal into at least two different AI subtasks;

[0122] A first sending module configured to send primary service function chain arrangement information to a cloud computing center device according to the at least two different AI subtasks;

[0123] A first receiving module configured to receive first information sent by the cloud computing center device; the first information includes final service function chain chaining decision information or confirmation information;

[0124] A first determining module configured to determine a final service function chain according to the first information.

[0125] Optionally, the sending of the primary service function chain arrangement information to the cloud computing center device according to the at least two different AI subtasks includes:

[0126] ​​According to the second information, a service position of each AI sub-task is arranged to form a primary service function chain, and primary service function chain arrangement information is obtained; wherein the second information is global state information corresponding to a cloud computing center device to which the first base station belongs; the global state information is related to execution of an AI task;

[0127] The primary service function chain arrangement information is sent to the cloud computing center device.

[0128] Optionally, the second information includes at least one of the following:

[0129] Information of AI models stored in each base station and the cloud computing center device;

[0130] Information of remaining computing resources in each base station and the cloud computing center device;

[0131] Information of remaining storage resources in each base station and the cloud computing center device;

[0132] Information of physical positions of each base station;

[0133] Information of mutual connection conditions of each base station;

[0134] The base stations are connected to the cloud computing center device respectively.

[0135] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel;

[0136] The second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0137] Optionally, the method further includes:

[0138] A second sending module is configured to send local state information of the first base station to the cloud computing center device;

[0139] The local state information is related to execution of an AI task.

[0140] Optionally, the local state information includes at least one of the following:

[0141] Information of AI models stored in the first base station;

[0142] Information of remaining computing resources in the first base station;

[0143] Information of storage resources in the first base station;

[0144] Information of a physical position of the first base station;

[0145] information of a connection status of the first base station with other base stations.

[0146] Optionally, the local state information is inserted into a MAC layer transport block as a MAC control signal element and transmitted through a transmission channel.

[0147] Optionally, the method further comprises:

[0148] The second processing module is configured to execute the AI task by using the final service function chain to obtain a processing result.

[0149] The third sending module is configured to send the processing result to the terminal.

[0150] Optionally, the method further comprises:

[0151] The third processing module is configured to, before executing the AI task by using the final service function chain to obtain a processing result, determine a training party of an AI model corresponding to each AI subtask, obtain initial parameter information of the AI model by using the training party, and the training party is the first base station, the cloud computing center device, or other base stations under the cloud computing center device except the first base station.

[0152] The second receiving module is configured to receive final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0153] The executing the AI task by using the final service function chain to obtain a processing result comprises:

[0154] Executing the AI task by using an AI model according to the final service function chain and the final parameter information to obtain a processing result.

[0155] Optionally, the determining a training party of an AI model corresponding to each AI subtask comprises:

[0156] In a case where the AI model corresponding to the AI subtask exists locally in the first base station, the training party of the AI model corresponding to the AI subtask is determined as the first base station.

[0157] In a case where the AI model corresponding to the AI subtask does not exist locally in the first base station, it is determined according to the second information whether the AI model corresponding to the AI subtask exists in the other base stations; in a case where the AI model exists, the training party of the AI model corresponding to the AI subtask is determined as the other base stations; in a case where the AI model does not exist, the training party of the AI model corresponding to the AI subtask is determined as the cloud computing center device.

[0158] The embodiment of the application further provides an information processing device applied to a cloud computing center device, comprising:

[0159] a third receiving module, configured to receive primary service function chain orchestration information sent by at least one first base station;

[0160] a second determining module, configured to determine, according to the primary service function chain orchestration information and second information, whether each base station computing resource on a service function chain meets current all primary service function chain demands, to obtain a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to execution of an AI task;

[0161] a first obtaining module, configured to obtain, according to the determination result, first information corresponding to each first base station respectively; the first information includes final service function chain chaining decision information or confirmation information;

[0162] a fourth sending module, configured to send corresponding first information to each first base station.

[0163] Optionally, the first obtaining module is configured to obtain, according to the determination result, the first information corresponding to each first base station respectively, including:

[0164] in a case where the determination result indicates that each base station computing resource on the service function chain can meet current all primary service function chain demands, the first information including confirmation information is taken as the first information corresponding to each first base station respectively;

[0165] in a case where the determination result indicates that each base station computing resource on the service function chain cannot meet current all primary service function chain demands, the primary service function chain orchestration information is updated, a first base station in which a primary service function chain is changed and a first base station in which a primary service function chain is maintained are determined, first information including target decision information is taken as the first information corresponding to the first base station in which the primary service function chain is changed, and first information including confirmation information is taken as the first information corresponding to the first base station in which the primary service function chain is maintained; the target decision information includes final service function chain chaining decision information corresponding to the first base station in which the primary service function chain is changed.

[0166] Optionally, the apparatus further includes:

[0167] a fifth sending module, configured to send the second information to each base station on the service function chain respectively;

[0168] The second information includes at least one of the following:

[0169] information of an AI model stored in each base station and the cloud computing center device;

[0170] information of remaining computing resources in each base station and the cloud computing center device;

[0171] information of remaining storage resources in each base station and the cloud computing center device;

[0172] information of physical locations of each base station;

[0173] information of interconnections of each base station;

[0174] The cloud computing center device is connected with each base station.

[0175] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

[0176] The second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0177] Optionally, the method further comprises:

[0178] a fourth receiving module configured to receive local state information of each base station on the service function chain, wherein the local state information is related to the execution of the AI task;

[0179] a third processing module configured to obtain the second information according to the local state information.

[0180] Optionally, the local state information comprises at least one of the following:

[0181] information of an AI model stored in the base station;

[0182] information of remaining computing resources in the base station;

[0183] information of storage resources in the base station;

[0184] information of a physical location of the base station;

[0185] information of an interconnection of the base station with other base stations.

[0186] Optionally, the local state information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

[0187] Optionally, the method further comprises:

[0188] a second obtaining module configured to obtain initial parameter information of at least one AI model;

[0189] a fourth processing module, configured to perform a classified aggregation operation on the initial parameter information to obtain final parameter information of each AI model;

[0190] a sixth sending module, configured to send the final parameter information of the AI model to a base station corresponding to the AI model.

[0191] Optionally, the obtaining of the initial parameter information of the at least one AI model comprises at least one of the following:

[0192] receiving initial parameter information of an AI model sent by at least one base station;

[0193] receiving an AI subtask sent by the first base station, training a corresponding AI model according to the AI subtask, and obtaining initial parameter information of the AI model.

[0194] Optionally, the training of the corresponding AI model according to the AI subtask comprises:

[0195] training the corresponding AI model according to the AI subtask by using a transfer learning manner.

[0196] Embodiments of the present application also provide an information processing device applied to a first base station, comprising:

[0197] a fifth processing module, configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks;

[0198] a sixth processing module, configured to determine a training party of an AI model corresponding to each AI subtask, and obtain initial parameter information of the AI model by using the training party; the training party is the first base station, a cloud computing center device connected to the first base station, or other base stations except the first base station under the cloud computing center device;

[0199] a fifth receiving module, configured to receive final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0200] Optionally, the determination of the training party of the AI model corresponding to each AI subtask comprises:

[0201] in a case where the AI model corresponding to the AI subtask exists locally in the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station;

[0202] If the AI ​​model corresponding to the AI ​​sub-task does not exist locally at the first base station, it is determined whether the other base stations have the AI ​​model corresponding to the AI ​​sub-task based on the second information; if they do, the other base stations are determined to be the trainers of the AI ​​model corresponding to the AI ​​sub-task; if they do not, the cloud computing center device is determined to be the trainer of the AI ​​model corresponding to the AI ​​sub-task.

[0203] The second information is the global status information corresponding to the cloud computing center equipment to which the first base station belongs; the global status information is related to the execution of AI tasks.

[0204] Optionally, the second information includes at least one of the following:

[0205] Information on AI models already stored in various base stations and cloud computing center equipment;

[0206] Information on the remaining computing resources in each base station and cloud computing center equipment;

[0207] Information on the remaining storage resources in each base station and cloud computing center equipment;

[0208] Information on the physical location of each base station;

[0209] Information on the interconnection status of each base station;

[0210] Each of the base stations is connected to the cloud computing center equipment.

[0211] Optionally, the second information is inserted into the MAC layer transport block as a Media Access Control (MAC) control cell and transmitted through the transport channel;

[0212] The second information is located at the beginning of the MAC protocol data unit in the MAC layer transport block.

[0213] This invention also provides an information processing apparatus applied to cloud computing center equipment, comprising:

[0214] The third acquisition module is used to acquire the initial parameter information of at least one AI model;

[0215] The seventh processing module is used to perform classification and aggregation operations on the initial parameter information to obtain the final parameter information of each AI model;

[0216] The seventh sending module is used to send the final parameter information of the AI ​​model to the base station corresponding to the AI ​​model.

[0217] Optionally, obtaining the initial parameter information of at least one AI model includes at least one of the following:

[0218] receive initial parameter information of an AI model sent by at least one base station;

[0219] receive an AI sub-task sent by a first base station, train a corresponding AI model according to the AI sub-task, and obtain initial parameter information of the AI model.

[0220] Optionally, the training of the corresponding AI model according to the AI sub-task comprises:

[0221] training the corresponding AI model according to the AI sub-task by using a transfer learning manner.

[0222] Optionally, the method further comprises:

[0223] an eighth sending module configured to send second information to at least one base station respectively;

[0224] The second information is global state information corresponding to the cloud computing center device, and the global state information is related to the execution of the AI task.

[0225] Optionally, the second information comprises at least one of the following:

[0226] information of AI models stored in each base station and the cloud computing center device;

[0227] information of remaining computing resources in each base station and the cloud computing center device;

[0228] information of remaining storage resources in each base station and the cloud computing center device;

[0229] information of physical positions of each base station;

[0230] information of mutual connection conditions of each base station;

[0231] The second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel.

[0232] The second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0233] The application also provides an information processing device, which is a first base station and comprises a processor and a transceiver.

[0234] The processor is configured to decompose an AI task of a terminal into at least two different AI sub-tasks.

[0235] The processor is configured to decompose an AI task of a terminal into at least two different AI sub-tasks.

[0236] According to the at least two different AI sub-tasks, the transceiver is configured to send primary service function chain arrangement information to a cloud computing center device;

[0237] The transceiver is configured to receive first information sent by the cloud computing center device; the first information includes final service function chain forming decision information or confirmation information;

[0238] According to the first information, the final service function chain is determined.

[0239] Optionally, according to the at least two different AI sub-tasks, the transceiver is configured to send primary service function chain arrangement information to a cloud computing center device, including:

[0240] According to the second information, the service positions of the AI sub-tasks are arranged to form a primary service function chain, and primary service function chain arrangement information is obtained; wherein the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; the global state information is related to the execution of the AI task;

[0241] The transceiver is configured to send the primary service function chain arrangement information to the cloud computing center device.

[0242] Optionally, the second information includes at least one of the following:

[0243] Information of AI models stored in each base station and the cloud computing center device;

[0244] Information of remaining computing resources in each base station and the cloud computing center device;

[0245] Information of remaining storage resources in each base station and the cloud computing center device;

[0246] Information of physical positions of each base station;

[0247] Information of mutual connection conditions of each base station;

[0248] The cloud computing center device is connected to each base station.

[0249] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel;

[0250] The second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0251] Optionally, the processor is further configured to:

[0252] transmit, by the transceiver, the local state information of the first base station to the cloud computing center device;

[0253] The local state information is related to execution of an AI task.

[0254] Optionally, the local state information comprises at least one of the following:

[0255] information of an AI model stored in the first base station;

[0256] information of remaining computing resources in the first base station;

[0257] information of storage resources in the first base station;

[0258] information of a physical location of the first base station;

[0259] information of a connection status of the first base station and other base stations.

[0260] Optionally, the local state information is inserted into a MAC layer transport block as a MAC control element and transmitted through a transmission channel.

[0261] Optionally, the processor is further configured to:

[0262] execute the AI task by using the final service function chain to obtain a processing result;

[0263] transmit, by the transceiver, the processing result to the terminal.

[0264] Optionally, the processor is further configured to:

[0265] determine a training party of an AI model corresponding to each AI subtask before executing the AI task by using the final service function chain to obtain a processing result, and obtain initial parameter information of the AI model by using the training party; the training party is the first base station, the cloud computing center device, or other base stations under the cloud computing center device except the first base station;

[0266] receive, by the transceiver, final parameter information obtained by the cloud computing center device according to the initial parameter information;

[0267] The executing the AI task by using the final service function chain to obtain a processing result comprises:

[0268] executing the AI task by using an AI model according to the final service function chain and the final parameter information to obtain a processing result.

[0269] Optionally, the determining a training party of an AI model corresponding to each AI subtask comprises:

[0270] In a case where the AI model corresponding to the AI subtask exists locally at the first base station, it is determined that the training party of the AI model corresponding to the AI subtask is the first base station.

[0271] In a case where the AI model corresponding to the AI subtask does not exist locally at the first base station, it is determined, according to the second information, whether the AI model corresponding to the AI subtask exists at the other base station. In a case where the AI model exists, it is determined that the training party of the AI model corresponding to the AI subtask is the other base station. In a case where the AI model does not exist, it is determined that the training party of the AI model corresponding to the AI subtask is the cloud computing center device.

[0272] The embodiment of the application further provides an information processing device, which is a cloud computing center device and comprises a processor and a transceiver.

[0273] The processor is configured to receive, by the transceiver, primary service function chain orchestration information sent by at least one first base station.

[0274] According to the primary service function chain orchestration information and second information, it is determined whether each base station computing resource on a service function chain meets current all primary service function chain demands, to obtain a determination result. The second information is global state information corresponding to the cloud computing center device. The global state information is related to execution of an AI task.

[0275] According to the determination result, first information corresponding to each first base station is obtained. The first information comprises final service function chain chaining decision information or confirmation information.

[0276] The transceiver is configured to send the corresponding first information to each first base station.

[0277] Optionally, the first information corresponding to each first base station is obtained according to the determination result, and the method comprises the following steps.

[0278] In a case where the determination result indicates that each base station computing resource on a service function chain can meet current all primary service function chain demands, first information comprising confirmation information is taken as the first information corresponding to each first base station.

[0279] In a case where the determination result indicates that the base station computing resources on the service function chain cannot satisfy all current primary service function chain demands, the primary service function chain orchestration information is updated to determine a first base station in which the primary service function chain is changed and a first base station in which the primary service function chain remains unchanged; first information containing target decision information is taken as first information corresponding to the first base station in which the primary service function chain is changed; and first information containing confirmation information is taken as first information corresponding to the first base station in which the primary service function chain remains unchanged; wherein the target decision information includes final service function chain chaining decision information corresponding to the first base station in which the primary service function chain is changed.

[0280] Optionally, the processor is further configured to:

[0281] send, by the transceiver, the second information to each base station on the service function chain respectively;

[0282] The second information includes at least one of the following:

[0283] information of AI models already stored in each base station and the cloud computing center device;

[0284] information of remaining computing resources in each base station and the cloud computing center device;

[0285] information of remaining storage resources in each base station and the cloud computing center device;

[0286] information of physical positions of each base station;

[0287] information of mutual connections of each base station;

[0288] The second information is inserted into a MAC layer transport block as a MAC control element and transmitted through a transmission channel.

[0289] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and transmitted through a transmission channel.

[0290] The second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0291] Optionally, the processor is further configured to:

[0292] receive, by the transceiver, the local state information of the base station sent by each base station on the service function chain respectively; wherein the local state information is related to the execution of the AI task;

[0293] obtain the second information according to the local state information.

[0294] Optionally, the local state information comprises at least one of the following:

[0295] information of an AI model stored in the base station;

[0296] information of remaining computing resources in the base station;

[0297] information of storage resources in the base station;

[0298] information of a physical location of the base station;

[0299] information of a connection status of the base station with other base stations.

[0300] Optionally, the local state information is inserted into a MAC layer transport block as a MAC control signal element and transmitted through a transmission channel.

[0301] Optionally, the processor is further configured to:

[0302] obtain initial parameter information of at least one AI model;

[0303] perform a classification aggregation operation on the initial parameter information to obtain final parameter information of each AI model;

[0304] transmit the final parameter information of the AI model to a base station corresponding to the AI model through the transceiver.

[0305] Optionally, the obtaining of the initial parameter information of at least one AI model comprises at least one of the following:

[0306] receiving initial parameter information of an AI model transmitted by at least one base station through the transceiver;

[0307] receiving an AI subtask transmitted by the first base station through the transceiver, training a corresponding AI model according to the AI subtask, and obtaining initial parameter information of the AI model.

[0308] Optionally, the training of the corresponding AI model according to the AI subtask comprises:

[0309] training the corresponding AI model according to the AI subtask in a manner of transfer learning.

[0310] Embodiments of the present application also provide an information processing device, which is a first base station and comprises a processor and a transceiver.

[0311] The processor is configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks.

[0312] determining a training party of an AI model corresponding to each of the AI sub-tasks, obtaining initial parameter information of the AI model by the training party; the training party is the first base station, a cloud computing center device connected with the first base station, or other base stations except the first base station under the cloud computing center device;

[0313] receiving, by the transceiver, final parameter information of the AI model obtained by the cloud computing center device according to the initial parameter information.

[0314] Optionally, the determining the training party of the AI model corresponding to each of the AI sub-tasks comprises:

[0315] in a case where the AI model corresponding to the AI sub-task exists locally in the first base station, determining that the training party of the AI model corresponding to the AI sub-task is the first base station;

[0316] in a case where the AI model corresponding to the AI sub-task does not exist locally in the first base station, determining whether the other base stations store the AI model corresponding to the AI sub-task according to second information; in a case where the AI model is stored, determining that the training party of the AI model corresponding to the AI sub-task is the other base stations; in a case where the AI model is not stored, determining that the training party of the AI model corresponding to the AI sub-task is the cloud computing center device;

[0317] wherein the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; the global state information is related to the execution of the AI task.

[0318] Optionally, the second information comprises at least one of the following:

[0319] information of AI models stored in each of the base stations and the cloud computing center device;

[0320] information of remaining computing resources in each of the base stations and the cloud computing center device;

[0321] information of remaining storage resources in each of the base stations and the cloud computing center device;

[0322] information of physical positions of each of the base stations;

[0323] information of mutual connections of each of the base stations;

[0324] wherein each of the base stations is connected with the cloud computing center device.

[0325] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel;

[0326] The second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0327] Embodiments of the present application also provide an information processing device, which is a cloud computing center device, comprising a processor and a transceiver;

[0328] The processor is configured to obtain initial parameter information of at least one AI model;

[0329] The initial parameter information is subjected to a classification and aggregation operation to obtain final parameter information of each AI model;

[0330] The transceiver is configured to send the final parameter information of the AI model to a base station corresponding to the AI model.

[0331] Optionally, the obtaining of the initial parameter information of the at least one AI model comprises at least one of the following:

[0332] The transceiver is configured to receive initial parameter information of an AI model sent by at least one base station;

[0333] The transceiver is configured to receive an AI subtask sent by a first base station, train a corresponding AI model according to the AI subtask, and obtain initial parameter information of the AI model.

[0334] Optionally, the training of the corresponding AI model according to the AI subtask comprises:

[0335] The corresponding AI model is trained according to the AI subtask by using a transfer learning manner.

[0336] Optionally, the processor is further configured to:

[0337] The transceiver is configured to send second information to at least one base station respectively;

[0338] The second information is global state information corresponding to the cloud computing center device; and the global state information is related to the execution of an AI task.

[0339] Optionally, the second information comprises at least one of the following:

[0340] Information of AI models already stored in each base station and the cloud computing center device;

[0341] Information of remaining computing resources in each base station and the cloud computing center device;

[0342] Information of remaining storage resources in each base station and the cloud computing center device;

[0343] Information of physical positions of each base station;

[0344] information of interconnection of each base station;

[0345] Each base station is connected with the cloud computing center device.

[0346] Optionally, the second information is inserted into a MAC layer transport block as a MAC control element and transmitted through a transmission channel.

[0347] The second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0348] The embodiment of the present application also provides an information processing device, including a memory, a processor and a program stored in the memory and executable on the processor; the processor implements the information processing method of the first base station side or the cloud computing center device side when executing the program.

[0349] The embodiment of the present application also provides a readable storage medium, which stores a program, and the program is executable on the processor to implement the steps of the information processing method of the first base station side or the cloud computing center device side.

[0350] The beneficial effects of the above technical solutions of the present application are as follows:

[0351] In the above scheme, the information processing method divides an AI task of a terminal into at least two different AI subtasks, sends primary service function chain arrangement information to a cloud computing center device according to the at least two different AI subtasks, receives first information sent by the cloud computing center device, the first information includes final service function chain forming decision information or confirmation information, and determines a final service function chain according to the first information, which can support the implementation of the AI service function chain arrangement based on the global perspective (second information) of the first base station and the cloud computing center device, thereby improving the arrangement effect and solving the problem of poor arrangement effect caused by the lack of global perspective in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0352] Figure 1 The information processing method flowchart of the embodiment of the present application Figure 1 ;

[0353] Figure 2 The information processing method flowchart of the embodiment of the present application Figure 2 ;

[0354] Figure 3 The information processing method flowchart of the embodiment of the present application Figure 3 ;

[0355] Figure 4 Flowchart of the information processing method of the embodiment of the present application Figure 4 ;

[0356] Figure 5 Flowchart of the model training method based on edge federated learning of the embodiment of the present application

[0357] Figure 6 Flowchart of the service process based on state feedback and orchestration collaboration mechanism of the embodiment of the present application

[0358] Figure 7 Local state representation intention of the embodiment of the present application

[0359] Figure 8 Global state representation intention of the embodiment of the present application

[0360] Figure 9 Structure diagram of the information processing device of the embodiment of the present application Figure 1 ;

[0361] Figure 10 Structure diagram of the information processing device of the embodiment of the present application Figure 2 ;

[0362] Figure 11 Structure diagram of the information processing device of the embodiment of the present application Figure 3 ;

[0363] Figure 12 Structure diagram of the information processing device of the embodiment of the present application Figure 4 ;

[0364] Figure 13 Structure diagram of the information processing device of the embodiment of the present application Figure 1 ;

[0365] Figure 14 Structure diagram of the information processing device of the embodiment of the present application Figure 2 ;

[0366] Figure 15 Structure diagram of the information processing device of the embodiment of the present application Figure 3 ;

[0367] Figure 16 Structure diagram of the information processing device of the embodiment of the present application Figure 4 . DETAILED DESCRIPTION

[0368] In order to make the technical problems, technical solutions and advantages of the present application clearer, specific embodiments will be described in detail below with reference to the accompanying drawings.

[0369] Firstly, the related content of the scheme is introduced below.

[0370] (1) The existing AI model training method is to use the powerful computing and storage resources of the cloud computing center to complete the AI model training and storage in the cloud, and the task of the UE will also be implemented in the cloud to provide on-demand services. However, there are the following problems in training all AI models in the cloud computing center: first, link congestion and data security problems caused by training data upload; second, a large amount of AI model training in the cloud will occupy a large amount of computing resources of the cloud center (i.e. cloud computing center); third, when the AI model is used, the communication between the UE and the cloud will cause the processing delay of the AI task to increase.

[0371] (2) The existing edge base stations interact through the Xn interface, and the state information content of the interaction includes the computing resources, storage resources and location information of the base station. However, the existing edge base station interaction information lacks expression of model resources, and it is difficult to provide complete information for AI service function chain arrangement. In addition, the topology structure of the edge base station network is often not in the form of full connection, and an edge base station only interacts with surrounding edge base stations, so it cannot have global knowledge when making decisions.

[0372] Based on the above, the present application provides an information processing method to solve the problem of lack of global perspective in the existing information processing scheme for AI service function chain arrangement, which is applied to a first base station, as shown in Figure 1 , which comprises:

[0373] Step 11: decomposing the artificial intelligence (AI) task of the terminal into at least two different AI sub-tasks;

[0374] Step 12: sending primary service function chain arrangement information to a cloud computing center device according to the at least two different AI sub-tasks;

[0375] Step 13: receiving first information sent by the cloud computing center device; the first information includes final service function chain chaining decision information or confirmation information;

[0376] Step 14: determining a final service function chain according to the first information.

[0377] Wherein, in the case that the first information includes the final service function chain chaining decision information, the final service function chain can be a service function chain indicated or carried by the final service function chain chaining decision information; in the case that the first information includes the confirmation information, it can be a primary service function chain, which is not limited here.

[0378] The information processing method provided by the embodiment of the application can support the implementation of the first base station and the cloud computing center device to cooperate to determine the AI service function chain arrangement based on the global perspective (second information), thereby improving the arrangement effect and solving the problem of poor arrangement effect caused by the lack of global perspective in the prior art.

[0379] The cloud computing center device sends the first information to the first base station, wherein the first information comprises final service function chain forming decision information or confirmation information; and the first base station determines the final service function chain based on the first information.

[0380] In this way, the preliminary service function chain arrangement information can be accurately obtained based on the global state information (second information) and transmitted.

[0381] In the embodiment of the application, the second information comprises at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; information of mutual connection conditions of each base station; and wherein each base station is connected to the cloud computing center device.

[0382] In this way, the second information can be clearly determined.

[0383] The second information is inserted into a MAC layer transport block as a MAC control element and transmitted through a transmission channel; and the second information is located at the start position of a MAC protocol data unit (MAC PDU) in the MAC layer transport block.

[0384] In this way, the second information can be accurately transmitted.

[0385] Further, the information processing method further comprises: sending local state information of the first base station to the cloud computing center device; and wherein the local state information is related to the execution of the AI task.

[0386] This allows the cloud computing center equipment to accurately obtain the local status information of the first base station.

[0387] The local status information includes at least one of the following: information on the AI ​​models already stored in the first base station; information on the remaining computing resources in the first base station; information on the storage resources in the first base station; information on the physical location of the first base station; and information on the connection status of the first base station with other base stations.

[0388] This allows us to clearly define the local status information of the first base station.

[0389] In this embodiment of the invention, the local state information is inserted into the MAC layer transport block as a MAC control cell and transmitted through the transport channel.

[0390] This ensures accurate transmission of the local status information of the first base station.

[0391] Furthermore, the information processing method further includes: using the final service function chain to execute the AI ​​task and obtain a processing result; and sending the processing result to the terminal.

[0392] This allows for the accurate execution of the terminal's AI tasks. Specifically, "execute the AI ​​task using the final service function chain" can include: utilizing each base station included in the final service function chain to execute the AI ​​sub-task assigned to each base station (itself); more specifically, the final service function chain can include the first base station, and correspondingly, "execute the AI ​​task using the final service function chain" can include: utilizing the first base station to execute the AI ​​sub-task assigned to the first base station, i.e., the first base station executes the assigned AI sub-task; however, it is not limited to this.

[0393] In this embodiment of the invention, before executing the AI ​​task using the final service function chain to obtain the processing result, the method further includes: determining the training party of the AI ​​model corresponding to each AI sub-task, and obtaining the initial parameter information of the AI ​​model using the training party; the training party is the first base station, the cloud computing center device, or other base stations under the cloud computing center device other than the first base station; receiving the final parameter information obtained by the cloud computing center device based on the initial parameter information; the step of executing the AI ​​task using the final service function chain to obtain the processing result includes: executing the AI ​​task using the AI ​​model based on the final service function chain and the final parameter information to obtain the processing result.

[0394] Thus, the AI task can be accurately executed, and problems such as link congestion and poor data security caused by training data uploading, occupation of a large amount of computing resources of the cloud computing center, and large AI task processing delay caused by training of all AI models in the cloud computing center can be avoided, thereby further supporting improvement of the orchestration effect and better solving the problem of poor orchestration effect caused by lack of a global perspective in the information processing scheme for AI service function chain orchestration in the prior art.

[0395] The determination of the training party of the AI model corresponding to each AI subtask includes: (1) in the case that the AI model corresponding to the AI subtask exists locally in the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station; and (2) in the case that the AI model corresponding to the AI subtask does not exist locally in the first base station, determining, according to the second information, whether the AI model corresponding to the AI subtask exists in the other base stations; in the case that the AI model exists, determining that the training party of the AI model corresponding to the AI subtask is the other base station; and in the case that the AI model does not exist, determining that the training party of the AI model corresponding to the AI subtask is the cloud computing center device.

[0396] Thus, the training party of the AI model corresponding to the AI subtask can be accurately determined.

[0397] The embodiment of the application further provides an information processing method applied to a cloud computing center device, as shown in the following formula: Figure 2 The method comprises the following steps:

[0398] Step 21: receiving primary service function chain orchestration information sent by at least one first base station;

[0399] Step 22: determining whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains according to the primary service function chain orchestration information and second information, to obtain a determination result; the second information is global state information corresponding to the cloud computing center device; and the global state information is related to execution of an AI task;

[0400] Step 23: obtaining first information corresponding to each first base station according to the determination result; the first information includes final service function chain chain forming decision information or confirmation information;

[0401] Step 24: sending corresponding first information to each first base station.

[0402] The information processing method provided by the embodiment of the application comprises the following steps: receiving primary service function chain arrangement information sent by at least one first base station; determining whether the computing resources of each base station on a service function chain meet the current demand of all primary service function chains according to the primary service function chain arrangement information and second information, and obtaining a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to the execution of an AI task; acquiring first information corresponding to each first base station according to the determination result; the first information comprises final service function chain chaining decision information or confirmation information; and sending the corresponding first information to each first base station. The AI service function chain arrangement can be determined based on the global perspective (second information) by the cooperation of the first base station and the cloud computing center device, so that the arrangement effect is improved, and the problem that the arrangement effect is poor due to the lack of global perspective in the prior art is solved.

[0403] In the embodiment, the first information corresponding to each first base station is acquired according to the determination result, and the first information comprises final service function chain chaining decision information or confirmation information.

[0404] In this way, the first information corresponding to each first base station can be accurately obtained.

[0405] Further, the information processing method further comprises: sending the second information to each base station on the service function chain respectively; wherein the second information comprises at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of the physical positions of each base station; information of the mutual connection of each base station; wherein each base station is connected to the cloud computing center device.

[0406] In this way, the second information can be accurately obtained by the base stations.

[0407] The second information is inserted into a MAC layer transport block as a MAC control information element and is transmitted through a transmission channel.

[0408] In this way, the second information can be accurately transmitted.

[0409] Further, the information processing method further includes: receiving the local state information of each base station on the service function chain respectively transmitted by the base station; wherein the local state information is related to the execution of the AI task; and obtaining the second information according to the local state information.

[0410] In this way, the second information can be accurately obtained.

[0411] The local state information includes at least one of the following: information of an AI model stored in the base station; information of remaining computing resources in the base station; information of storage resources in the base station; information of the physical location of the base station; and information of the connection status of the base station with other base stations.

[0412] In this way, the local state information can be clearly obtained.

[0413] In the embodiment of the application, the local state information is inserted into a MAC layer transport block as a MAC control information element and is transmitted through a transmission channel.

[0414] In this way, the local state information can be accurately transmitted.

[0415] Further, the information processing method further includes: obtaining initial parameter information of at least one AI model; performing a classification and aggregation operation on the initial parameter information to obtain final parameter information of each AI model; and transmitting the final parameter information of the AI model to the base station corresponding to the AI model.

[0416] In this way, the final parameter information can be accurately obtained, and the problems of link congestion and poor data security caused by training data upload, occupation of a large amount of computing resources of a cloud computing center, and large AI task processing delay when all AI model training is completed in the cloud computing center can be avoided; thereby further supporting the improvement of the orchestration effect and better solving the problem of poor orchestration effect caused by the lack of a global perspective in the prior art information processing scheme for AI service function chain orchestration.

[0417] The initial parameter information of the at least one AI model is obtained in at least one of the following ways: (1) receiving initial parameter information of an AI model sent by at least one base station; and (2) receiving an AI subtask sent by the first base station, training a corresponding AI model according to the AI subtask, and obtaining initial parameter information of the AI model.

[0418] In this way, the initial parameter information of the AI model can be obtained in multiple ways.

[0419] In the embodiment, the training of the corresponding AI model according to the AI subtask includes training the corresponding AI model according to the AI subtask by using a transfer learning method.

[0420] In this way, the training efficiency can be improved and the training cost can be reduced.

[0421] The embodiment also provides an information processing method applied to a first base station, as shown in the following. Figure 3 The information processing method includes the following steps.

[0422] Step 31: decomposing an artificial intelligence (AI) task of a terminal into at least two different AI subtasks.

[0423] Step 32: determining a training party of an AI model corresponding to each AI subtask, and obtaining initial parameter information of the AI model by using the training party; the training party is the first base station, a cloud computing center device connected to the first base station, or another base station under the cloud computing center device except the first base station.

[0424] Step 33: receiving final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0425] The information processing method provided by the embodiment decomposes an AI task of a terminal into at least two different AI subtasks, determines a training party of an AI model corresponding to each AI subtask, obtains initial parameter information of the AI model by using the training party, and receives final parameter information obtained by a cloud computing center device according to the initial parameter information. The training party is the first base station, a cloud computing center device connected to the first base station, or another base station under the cloud computing center device except the first base station. This can avoid problems such as link congestion and poor data security caused by uploading of training data, occupation of a large amount of computing resources of the cloud computing center, and large AI task processing delay when all AI models are trained in the cloud computing center, thereby further supporting improvement of orchestration effect and better solving the problem that an information processing scheme for AI service function chain orchestration in the prior art lacks a global perspective and has poor orchestration effect.

[0426] The determining of the training party of the AI model corresponding to each AI subtask comprises: (1) in the case that the AI model corresponding to the AI subtask exists locally at the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station; (2) in the case that the AI model corresponding to the AI subtask does not exist locally at the first base station, determining, according to second information, whether the AI model corresponding to the AI subtask exists at the other base stations; in the case that the AI model exists, determining that the training party of the AI model corresponding to the AI subtask is the other base stations; in the case that the AI model does not exist, determining that the training party of the AI model corresponding to the AI subtask is the cloud computing center device; wherein the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; and the global state information is related to the execution of the AI task.

[0427] In this way, the training party of the AI model corresponding to each AI subtask can be accurately determined.

[0428] In the embodiment of the application, the second information comprises at least one of the following: information of AI models that have been stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; information of mutual connection conditions of each base station; wherein each base station is connected to the cloud computing center device.

[0429] In this way, the second information can be clearly determined.

[0430] The second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel; and the second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0431] In this way, the second information can be accurately transmitted.

[0432] The embodiment of the application also provides an information processing method, which is applied to a cloud computing center device, as shown in the accompanying drawings, and comprises the following steps. Figure 4

[0433] Step 41: obtaining initial parameter information of at least one AI model;

[0434] Step 42: performing a classification and aggregation operation on the initial parameter information to obtain final parameter information of each AI model;

[0435] Step 43: sending the final parameter information of the AI model to a base station corresponding to the AI model.

[0436] ​The information processing method provided by the embodiment of the application can obtain initial parameter information of at least one AI model, perform classification and aggregation operations on the initial parameter information to obtain final parameter information of each AI model, and send the final parameter information of the AI model to a base station corresponding to the AI model, thereby avoiding problems such as link congestion and poor data security caused by training data uploading, occupation of a large amount of computing resources of a cloud computing center, and large AI task processing delay when all AI models are trained in the cloud computing center, and further improving the arrangement effect and better solving the problem of poor arrangement effect caused by the lack of a global perspective in the prior art information processing scheme for AI service function chain arrangement.

[0437] The initial parameter information of the at least one AI model includes at least one of the following: (1) receiving initial parameter information of an AI model sent by at least one base station; and (2) receiving an AI subtask sent by a first base station, training a corresponding AI model according to the AI subtask, and obtaining initial parameter information of the AI model.

[0438] In this way, the initial parameter information of the AI model can be obtained in multiple ways. The first base station can be a base station that decomposes an artificial intelligence (AI) task of a terminal into at least two different AI subtasks, but is not limited thereto.

[0439] In the embodiment of the application, the training of the corresponding AI model according to the AI subtask includes training the corresponding AI model according to the AI subtask by using a transfer learning method.

[0440] In this way, the training efficiency can be improved and the training cost can be reduced.

[0441] Further, the information processing method further includes sending second information to at least one base station, respectively, wherein the second information is global state information corresponding to the cloud computing center device, and the global state information is related to the execution of the AI task.

[0442] In this way, each base station can accurately obtain the second information.

[0443] The second information includes at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; and information of mutual connection conditions of each base station, wherein each base station is connected to the cloud computing center device.

[0444] In this way, the second information can be clearly defined.

[0445] In the embodiment of the present application, the second information is inserted into a MAC layer transport block as a MAC control signal element and is transmitted through a transmission channel; wherein the second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0446] In this way, the second information can be accurately transmitted.

[0447] The information processing method provided by the embodiment of the present application is exemplified below, and the base station is exemplified as an edge base station.

[0448] To solve the above technical problems, the embodiment of the present application provides an information processing method, which can be specifically implemented as a service method based on an edge federated learning architecture, including a model training method combining federated learning and transfer learning for service orchestration, a hierarchical state feedback and orchestration cooperation mechanism, and a design of a state table data reporting control packet, which can improve the AI model training efficiency, AI service management and orchestration flexibility in a 6G network. Specifically, the present scheme mainly includes the following contents:

[0449] (1) For the training and location deployment of AI models, the present scheme proposes a model training method combining federated learning and transfer learning for service orchestration: the edge base station (corresponding to the first base station) first decomposes the UE task into different AI subtasks, and according to the matching of the existing AI models on the edge base station and the subtasks, transmits the corresponding training data to the appropriate edge base station for model training. Based on the idea of federated learning, each edge base station uploads the local model training parameters (corresponding to the initial parameter information) to the cloud center. The cloud center classifies AI models, aggregates the same AI models, stores them in a model library (corresponding to the classification and aggregation operation on the initial parameter information to obtain the final parameter information of each AI model), and distributes the AI model parameters to the corresponding edge base station (corresponding to the step of transmitting the final parameter information of the AI model to the base station corresponding to the AI model) to complete the model parameter update. In order to reduce the training cost, and at the same time benefit from the model sharing mechanism of the model library, the cloud center can use the method of transfer learning to train new models (corresponding to the step of training the AI model corresponding to the AI subtask by using the method of transfer learning).

[0450] (2) For the service function chain arrangement problem, the scheme introduces a hierarchical state feedback and arrangement cooperation mechanism in the AI service function chain arrangement of the edge computing network: the cloud center is responsible for the collection and distribution of the network global state resources (corresponding to the second information above), and the confirmation and / or modification of the service function chain arrangement (corresponding to the determination of whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains according to the primary service function chain arrangement information and the second information, and obtaining the determination result; according to the determination result, obtaining the first information corresponding to each first base station). Among them, each edge base station reports the local state information to the cloud center, and the cloud center aggregates it into a global state table (corresponding to the global state information above) and distributes it to all edge base stations. The edge base station arranges the AI service function chain based on the global state table (corresponding to the arrangement of the service location of each AI sub-task based on the second information to form a primary service function chain, and obtaining the primary service function chain arrangement information), and uploads the arrangement decision to the cloud center for confirmation or modification (corresponding to the sending of the primary service function chain arrangement information to the cloud computing center device), and the cloud center feedbacks the confirmation and modification information to the corresponding edge base station (corresponding to the sending of the corresponding first information to each first base station).

[0451] (3) For the lack of interaction between edge base stations on computing, storage, and model resources, the scheme proposes a MAC CE measurement data reporting method: a local state table of the edge base station is inserted into the MAC control signal element (corresponding to the local state information being inserted into the MAC layer transport block as a MAC control signal element), which contains the computing resources, storage resources, model resources, and location information of the edge base station. The local state table is uploaded to the cloud center regularly, integrated into a global state table by the cloud center, and distributed to all edge base stations, so that the edge base stations can know the resource situation in the network and make reasonable AI service function chain arrangement.

[0452] The contents included in the scheme are illustrated below.

[0453] I. Model training method combining federated learning and transfer learning for service arrangement;

[0454] Arranger base station (corresponding to the first base station): for a UE (terminal), the edge base station connected with it will provide AI service function chain arrangement service for it, and this edge base station is called the arranger base station of the UE.

[0455] Global state table: records all edge base stations in the network (i.e. all edge base stations under the cloud center) and the computing resources, storage resources, model resources of the cloud center, the location information of each edge base station, the interconnection of edge base stations, etc.; can be carried in the MAC control cell and issued from the cloud center to each edge base station.

[0456] 1. The service-oriented orchestration federated learning model training method is as follows:

[0457] The UE offloads task data to the orchestrator base station; the orchestrator base station first decomposes the task into multiple AI sub-tasks, and the multiple AI sub-tasks correspond to different AI services; the orchestrator base station performs corresponding edge federated learning model training based on the global state table. Specifically:

[0458] According to the matching of the existing AI model on the edge base station and the required AI service, it is divided into the following three cases:

[0459] Case 1: there is an AI model in the orchestrator base station that matches the required AI service, then the AI sub-task data stays in the orchestrator base station for model training; corresponding to the above case where the AI model corresponding to the AI sub-task exists locally in the first base station, it is determined that the training party of the AI model corresponding to the AI sub-task is the first base station;

[0460] Case 2: there is no AI model in the orchestrator base station that matches the required AI service, but at least one edge base station that has established a connection can be queried in the global state table, and the AI model in the edge base station matches the required AI service; then the orchestrator edge base station transmits the AI sub-task to the base station (when multiple base stations are queried, the shortest one is selected) for model training; corresponding to the above case where the AI model corresponding to the AI sub-task does not exist locally in the first base station, according to the second information, it is determined whether the other base station has the AI model corresponding to the AI sub-task; in the case of existence, it is determined that the training party of the AI model corresponding to the AI sub-task is the other base station;

[0461] Case 3: there is no AI model in the orchestrator edge base station (i.e. the orchestrator base station) that matches the required AI service, and the global state table cannot find an edge base station that matches the task and has established a connection, then the orchestrator edge base station sends the sub-task to the cloud center for new model training and stores it in the global model library; corresponding to the above case where the AI model corresponding to the AI sub-task does not exist, it is determined that the training party of the AI model corresponding to the AI sub-task is the cloud computing center device.

[0462] Wherein, each edge base station can serve as an orchestrator edge base station, and the AI models trained in each edge base station are also repeated (that is, multiple base stations can train the same AI model). After local model training is completed, each edge base station uploads the model parameters (corresponding to the initial parameter information of the AI model) to the cloud center. After the cloud center classifies the AI models, it completes the aggregation of AI models of the same type and distributes the AI model parameters to the corresponding edge base station (corresponding to the classification and aggregation operation of the initial parameter information described above, obtaining the final parameter information of each AI model; sending the final parameter information of the AI model to the base station corresponding to the AI model), so that the model parameter update is completed.

[0463] Specifically, as shown in the figure, the model training method based on edge federated learning includes the following operations: Figure 5

[0464] (1) The orchestrator base station matches the AI subtask with the AI model, determines to keep the AI subtask data in the orchestrator base station for model training and sends the obtained initial model parameters to the cloud center, or sends the AI subtask data to the cloud center, uses the cloud center to train a new model (transfer learning) and stores it in the model library, or selects a connected edge base station for model training (sends the AI subtask data to the selected edge base station) and sends the obtained initial model parameters to the cloud center.

[0465] (2) The cloud center classifies and aggregates the models according to the obtained initial model parameters to obtain the final model parameters; then distributes the final model parameters to each edge base station.

[0466] 2. The transfer learning method of the cloud center is as follows:

[0467] First, classify the newly arrived training task (AI subtask): for example, data analysis, natural language processing, image and video analysis, prediction and optimization, decision making, etc.

[0468] Second, select AI model resources of the same service type (belonging to the service required by the UE) from the global model library as the original model for transfer learning. Transfer learning can be divided into the following three cases:

[0469] Case one: the source domain (that is, the data collected by the UE perception) and the target domain (that is, the data corresponding to the AI model stored in the model library) belong to the same type of data, and the model or parameters are directly shared (that is, the existing AI model is directly used);

[0470] Case two: there are similar data in the source domain and the target domain, and the allocation weight value of the original model is adjusted;

[0471] ​Case three: the source domain and the target domain do not belong to the same type of data, but there are common cross features. Through feature transformation, the features of the source domain and the target domain are mapped to the same space, so that the data distribution in the space has the same distribution of the source domain data and the target domain data, and then the traditional machine learning is carried out.

[0472] II. Hierarchical state feedback and orchestration collaboration mechanism

[0473] Under the edge federated learning architecture, local AI model training and AI service function orchestration occur at the edge base station, and service function chain modification and model aggregation occur at the cloud center, which greatly reduces the amount of data transmitted in the network, avoids UE data leakage, and improves the utilization of computing resources at the edge base station.

[0474] Specifically, as shown in Figure 6 As shown in the service flowchart based on the state feedback and orchestration collaboration mechanism, the service implementation process based on the edge federated learning architecture includes:

[0475] Step one, the cloud center issues a global state table to all edge base stations (including orchestrator base stations 1 and 2), which contains the following information: 1. existing AI models in each edge base station and the cloud center (corresponding to the information of the AI models already stored in each base station and the cloud computing center device); 2. remaining computing resources in each edge base station and the cloud center (corresponding to the information of the remaining computing resources in each base station and the cloud computing center device); 3. remaining storage resources in each edge base station and the cloud center (corresponding to the information of the remaining storage resources in each base station and the cloud computing center device); 4. physical location of each edge base station (corresponding to the information of the physical location of each base station); 5. interconnection of each edge base station (corresponding to the information of the interconnection of each base station).

[0476] Step two, UE1 sends an AI task to orchestrator base station 1 to trigger primary service function chain orchestration; orchestrator base station 1 performs primary service function chain orchestration and reports the primary service function chain to the cloud center (corresponding to the above-mentioned sending of primary service function chain orchestration information to the cloud computing center device); UE2 sends an AI task to orchestrator base station 2 to trigger primary service function chain orchestration; orchestrator base station 2 performs primary service function chain orchestration and reports the primary service function chain to the cloud center (corresponding to the above-mentioned sending of primary service function chain orchestration information to the cloud computing center device);

[0477] Specifically, the orchestrator base station 1 (or the orchestrator base station 2) can decompose the AI task of the UE 1 (or the UE 2) into a plurality of different AI sub-tasks, and then arrange the service positions of the AI sub-tasks according to the global state table to form a primary service function chain (corresponding to the above-mentioned decomposing the artificial intelligence AI task of the terminal into at least two different AI sub-tasks; arranging the service positions of each of the AI sub-tasks according to the second information to form a primary service function chain, and obtaining primary service function chain arrangement information). Among them, the primary service function chain has the characteristics of the smallest hop number and meeting all the required AI services in the AI task. The orchestrator base station 1 (or the orchestrator base station 2) uploads the primary service function chain arrangement information to the cloud center for confirmation or modification.

[0478] Step three, since each primary service function chain decision is only made at the orchestrator edge base station, lacking the global perspective of the current decision, it may lead to a certain edge service will appear on multiple primary service function chains, while the base station computing resources are insufficient to support the computing demand on all primary service chains. Therefore, the cloud center collects all primary service function chain information, based on the latest global state table, judges whether the computing resources of each edge base station on the service function chain can meet the current demand of all primary service function chains (corresponding to the above-mentioned determining whether the computing resources of each base station on the service function chain can meet the current demand of all primary service function chains according to the primary service function chain arrangement information and the second information). If it is satisfied, the cloud center confirms all primary service function chains (corresponding to the above-mentioned in the case where the determination result indicates that the computing resources of each base station on the service function chain can meet the current demand of all primary service function chains, the first information containing the confirmation information is sent as the first information corresponding to each first base station), and the UE's AI service is executed according to the service function chain made by its orchestrator base station; if it is not satisfied, the cloud center modifies the service function chain, selects other edge base station nodes to chain (corresponding to the above-mentioned in the case where the determination result indicates that the computing resources of each base station on the service function chain cannot meet the current demand of all primary service function chains, the primary service function chain arrangement information is updated to determine the first base station of the primary service function chain to be changed and the first base station of the primary service function chain to be maintained), and the final service function chain chain decision (which can be the final service function chain) is sent to the orchestrator base station of the primary service function chain to be modified (corresponding to the above-mentioned first information containing the target decision information is sent as the first information corresponding to the first base station of the primary service function chain to be changed), and the confirmation information is sent to the orchestrator base station of the primary service function chain which is not modified (i.e. does not need to be modified) (corresponding to the above-mentioned first information containing the confirmation information is sent as the first information corresponding to the first base station of the primary service function chain to be maintained). Among them, the cloud center can use game theory algorithm to make decision modification, that is, when the combined strategy of all service function chains reaches Nash equilibrium, it is the best strategy, in other words, under this strategy combination, any individual change of a service function chain node cannot make the system optimization target better.

[0479] Specifically, the orchestrator base station 1 can be taken as the orchestrator base station of the primary service function chain which is not modified, and the orchestrator base station 2 can be taken as the orchestrator base station of the primary service function chain which needs to be modified, and correspondingly, the cloud center confirms the service function chain to the orchestrator base station 1 and modifies the service function chain to the orchestrator base station 2.

[0480] Step four, the computing task of the UE (UE1 or UE2) is processed in the service function chain and the processing result is fed back to the UE (corresponding to the above-mentioned use of the final service function chain, execute the AI task, and obtain the processing result; send the processing result to the terminal). Due to the mobility of the UE, if the UE switches the connected edge base station at this moment, it is divided into two cases: one is that the edge base station connected by the UE (i.e. the base station after switching) is on the service function chain, then all the computing results on the service function chain are directly transmitted to the base station and fed back to the UE; the second is that the edge base station connected by the UE is not on the service function chain, then each base station on the service function chain broadcasts the UE position information, and transmits the computing result to the target base station through the routing forwarding mode, and finally feeds back to the UE.

[0481] Specifically, it can be assumed that the orchestrator base station 1 is the edge base station connected by UE1 and is on the service function chain, then the orchestrator base station 1 can feed back the computing result directly to the UE; it is assumed that the orchestrator base station 2 is the edge base station connected by UE2 and is not on the service function chain, then the orchestrator base station 2 can broadcast and feed back the computing result to the UE.

[0482] Step five, all edge base stations on the service function chain (including the orchestrator base station 1 and the orchestrator base station 2) feed back the local state table to the cloud center. Among them, the local state table can contain the following information: 1, the existing AI model of the local edge base station (corresponding to the information of the existing AI model of the first base station); 2, the remaining computing resources of the local edge base station (corresponding to the information of the remaining computing resources in the first base station); 3, the storage resources of the local edge base station (corresponding to the information of the storage resources in the first base station); 4, the physical location of the local edge base station (corresponding to the information of the physical location of the first base station); 4, the connection condition of the local edge base station with other base stations (corresponding to the information of the connection condition of the first base station with other base stations).

[0483] Three, state table data reporting control packet design;

[0484] Before the AI service function chain is arranged, each edge base station needs to upload the local state table to the cloud center to integrate it into a global state table. Specifically, as shown in FIG. 6, the local state table of each edge base station is uploaded to the cloud center, and the cloud center integrates the local state table of each edge base station into a global state table. Figure 7As shown, the local state table can be inserted into a MAC layer transport block as a MAC control signal element and transmitted through a transmission channel (corresponding to the above-mentioned local state information being inserted into a MAC layer transport block as a MAC control signal element and transmitted through a transmission channel). Among them, the MAC layer transport block MAC PDU can be provided with multiple MAC SDUs (service data units) and multiple H fields, and the H field can include an R field (a reserved field), an F field (an 8 or 16 bit length field), an LCID field (which can be used to carry a logical channel identifier (6 bits)), and an SDU length field (which can be used to indicate the length of the SDU (8 or 16 bits)), which can be referred to in detail Figure 7 .

[0485] Specifically, the local state table can contain the following information: 1) the existing AI model of the local edge base station (i.e. the local existing AI model); 2) the remaining computing resources of the local edge base station (i.e. the local remaining computing resources); 3) the storage resources of the local edge base station (i.e. the local remaining storage resources); 4) the physical location of the local edge base station (i.e. the local edge base station location); 5) the connection status of the local edge base station with other base stations (i.e. the local edge base station interconnection status).

[0486] Among them, it is assumed that: the local existing AI model is represented by 8 bits, for example, 0000 0000-0000 0011 can represent image recognition model, speech recognition model, content recommendation model, online decision model, etc. respectively; the local remaining computing resources are represented by 4 bits, and the local total computing resources are evenly divided into 16 levels, for example, a edge base station has 160GHz of CPU resources, 0000-1111 respectively represents that the remaining CPU resources are 0-10GHz, 10-20GHz, and so on; the local storage computing resources are also represented by 4 bits, and the local total storage resources are evenly divided into 16 levels, for example, a edge base station has 3200G of memory resources, 0000-1111 respectively represents that the remaining CPU resources are 0-200G, 200-400G, and so on; the edge base station interconnection status field can record the ID of other edge base stations connected with the edge base station, and the location of the local edge base station can adopt a unified coding rule: city code (2 characters) + network code (1 character) + manufacturer code (1 character) + room or private network code (optional: 1 character) + base station serial number (3 characters) + cell serial number (1 character), for example:

[0487] City code: BJ, representing the first city;

[0488] Network code: G, representing GSM900;

[0489] Device manufacturer code: H, representing H manufacturer;

[0490] Indoor distribution code: M, representing an indoor station;

[0491] Base station serial number: 239;

[0492] Cell serial number: 1.

[0493] Subsequently, after the cloud center receives the local state table of each edge base station, the cloud center can integrate the state table information into a global state table and broadcast the global state table to all edge base stations. In the downlink transmission, the global state table can be located at the beginning of the MAC PDU (a MAC layer transmission block) as a MAC control cell, as shown in FIG. 6. The placement of the global state table in this way is conducive to low-latency operation of the terminal. Figure 8

[0494] The global state table contains the following information: 1, existing AI models in each edge base station and the cloud center (including all existing AI models of the edge base stations); 2, remaining computing resources of each edge base station and the cloud center (including all remaining computing resources of the edge base stations); 3, remaining storage resources of each edge base station and the cloud center (including all remaining storage resources of the edge base stations); 4, physical locations of each edge base station and the cloud center (including the locations of all edge base stations); and 5, mutual connection conditions of each edge base station and the cloud center (including the mutual connection conditions of all edge base stations). Specifically, the structure of the global state table is similar to that of the local state table, and related descriptions can be referred to the related content of the local state table, which will not be described here.

[0495] It is explained above that the related content of each part can be referred to each other, and repeated parts will not be described here.

[0496] Based on the above, the scheme provided by the embodiment of the application mainly includes the following three parts:

[0497] 1. A model training method combining federated learning and transfer learning for service orchestration: the edge base station first decomposes the UE task into different AI subtasks, and according to the matching of the existing AI model on the edge base station and the subtask, the corresponding training data is transmitted to the appropriate position for model training. Each edge base station uploads the model parameters (corresponding to the initial parameter information described above) to the cloud center. The cloud center classifies the AI model, aggregates the same AI model, stores it in the model library, and distributes the AI model parameters (corresponding to the final parameter information described above) to the corresponding edge base station to complete the model parameter update. In this scheme, according to the category division of the subtask, the cloud center uses transfer learning to compare the data attributes of the source domain and the target domain, and can adjust the parameters (such as adjusting the allocation weight value described above) to more efficiently realize new model training.

[0498] ​2. Hierarchical state feedback and orchestration cooperation mechanism: Step 1, the cloud center issues a global state table to all edge base stations; Step 2, AI task decomposition, forming a primary service function chain, and the primary service function chain orchestration information is uploaded to the cloud center for confirmation or modification. Step 3, the cloud center collects all primary service function chain information, based on the latest global state table, confirms and modifies the service function chain feedback. Step 4, the UE's computing task is processed in the service function chain and the processing result is fed back to the UE. Step 5, all edge base stations feed back the local state table to the cloud center.

[0499] 3. State table data reporting control packet design: the local state table is inserted into the MAC layer transmission block as a MAC control cell and transmitted through the transmission channel. Among them, in the uplink transmission, the local state table information can include: the existing AI model (information) of the local edge base station, the remaining computing resources (information) of the local edge base station, the remaining storage resources (information) of the local edge base station, the physical location (information) of the local edge base station, and the connection status (information) of the local edge base station with other base stations; in the downlink transmission, the global state table can contain the following information: the existing AI model (information) of each edge base station and the cloud center, the remaining computing resources (information) of each edge base station and the cloud center, the remaining storage resources (information) of each edge base station and the cloud center, the physical location (information) of each edge base station, and the mutual connection status (information) of each edge base station.

[0500] In summary, the scheme provided by the embodiment of the application has the following advantages:

[0501] (1) The scheme proposes a model training method combining federated learning and transfer learning for service orchestration: using the distributed learning architecture of federated learning, the transmission of model parameters replaces the transmission of raw data (data as model input), which alleviates the problem of limited computing power of edge base stations, greatly reduces data traffic in the network, and also guarantees data security. In addition, transfer learning in the cloud center can more efficiently realize new model training.

[0502] (2) The scheme introduces a hierarchical state feedback and orchestration cooperation mechanism in the service function chain orchestration of the edge computing network: through the design and sharing of the global state table of the cloud center, each edge base station can perceive the global state information of the network, and the game theory algorithm can realize the optimal service function chain orchestration scheme.

[0503] (3) The scheme carries out the design of the state table data reporting control packet: the calculation resource information, the storage resource information, the model resource information, the location information, the interconnection information and the like of the edge base station are added in the interaction between the edge base stations, so that the global state table can be shared, and each edge base station in the network can comprehensively consider the arrangement of the AI service function chain.

[0504] The embodiment of the application further provides an information processing device applied to a first base station, as shown in the figure, comprising: Figure 9

[0505] The first processing module 91 is used for decomposing an artificial intelligence (AI) task of a terminal into at least two different AI subtasks.

[0506] The first sending module 92 is used for sending primary service function chain arrangement information to a cloud computing center device according to the at least two different AI subtasks.

[0507] The first receiving module 93 is used for receiving first information sent by the cloud computing center device; the first information comprises final service function chain chaining decision information or confirmation information.

[0508] The first determining module 94 is used for determining a final service function chain according to the first information.

[0509] The information processing device provided by the embodiment of the application can decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks, send primary service function chain arrangement information to a cloud computing center device according to the at least two different AI subtasks, receive first information sent by the cloud computing center device, wherein the first information comprises final service function chain chaining decision information or confirmation information, and determine a final service function chain according to the first information, so as to support the cooperation between a first base station and a cloud computing center device to determine AI service function chain arrangement based on a global perspective (second information), thereby improving the arrangement effect and solving the problem that the information processing scheme for AI service function chain arrangement in the prior art lacks a global perspective and has poor arrangement effect.

[0510] The sending of the primary service function chain arrangement information to the cloud computing center device according to the at least two different AI subtasks comprises the following steps: arranging the service positions of the AI subtasks according to second information to form a primary service function chain and obtain primary service function chain arrangement information, wherein the second information is global state information corresponding to the cloud computing center device to which the first base station belongs, the global state information is related to the execution of the AI task, and the primary service function chain arrangement information is sent to the cloud computing center device.

[0511] ​In the embodiment of the present application, the second information includes at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical locations of each base station; information of mutual connection conditions of each base station; wherein each base station is connected with the cloud computing center device.

[0512] The second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel; and the second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0513] Further, the information processing device further includes a second sending module configured to send local state information of the first base station to the cloud computing center device; wherein the local state information is related to execution of an AI task.

[0514] The local state information includes at least one of the following: information of AI models stored in the first base station; information of remaining computing resources in the first base station; information of storage resources in the first base station; information of a physical location of the first base station; and information of a connection condition of the first base station with other base stations.

[0515] In the embodiment of the present application, the local state information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

[0516] Further, the information processing device further includes a second processing module configured to execute the AI task by using the final service function chain to obtain a processing result; and a third sending module configured to send the processing result to the terminal.

[0517] In the embodiment of the present application, the information processing device further includes a third processing module configured to, before executing the AI task by using the final service function chain to obtain a processing result, determine a training method of an AI model corresponding to each AI subtask, obtain initial parameter information of the AI model by using the training method; the training method is the first base station, the cloud computing center device, or other base stations under the cloud computing center device except the first base station; a second receiving module configured to receive final parameter information obtained by the cloud computing center device according to the initial parameter information; and the execution of the AI task by using the final service function chain to obtain a processing result includes: executing the AI task by using the AI model according to the final service function chain and the final parameter information to obtain a processing result.

[0518] In the embodiment of the present application, the determination of the training party of the AI model corresponding to each AI sub-task includes: (1) in the case that the AI model corresponding to the AI sub-task exists locally in the first base station, determining that the training party of the AI model corresponding to the AI sub-task is the first base station; (2) in the case that the AI model corresponding to the AI sub-task does not exist locally in the first base station, determining whether the other base stations have the AI model corresponding to the AI sub-task according to the second information; in the case that the AI model exists, determining that the training party of the AI model corresponding to the AI sub-task is the other base station; in the case that the AI model does not exist, determining that the training party of the AI model corresponding to the AI sub-task is the cloud computing center device.

[0519] The implementation embodiments of the information processing method on the first base station side are applicable to the embodiments of the information processing device, and the same technical effects can be achieved.

[0520] The embodiment of the present application also provides an information processing device applied to a cloud computing center device, as shown in the figure, which comprises: Figure 10

[0521] The third receiving module 101 is configured to receive the primary service function chain orchestration information sent by at least one first base station.

[0522] The second determining module 102 is configured to determine whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains according to the primary service function chain orchestration information and second information, and obtain a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to the execution of an AI task.

[0523] The first obtaining module 103 is configured to obtain first information corresponding to each first base station according to the determination result; the first information includes final service function chain chain forming decision information or confirmation information.

[0524] The fourth sending module 104 is configured to send the corresponding first information to each first base station.

[0525] ​The information processing device provided by the embodiment of the present application receives primary service function chain arrangement information sent by at least one first base station; determines whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains according to the primary service function chain arrangement information and second information, and obtains a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to the execution of an AI task; according to the determination result, first information corresponding to each first base station is obtained; the first information includes final service function chain chaining decision information or confirmation information; the corresponding first information is sent to each first base station; the AI service function chain arrangement based on the global perspective (second information) can be supported by the first base station and the cloud computing center device to cooperate, thereby improving the arrangement effect and solving the problem that the information processing scheme for the AI service function chain arrangement in the prior art lacks a global perspective, resulting in poor arrangement effect.

[0526] According to the determination result, the first information corresponding to each first base station is obtained, including: (1) in the case where the determination result indicates that the computing resources of each base station on the service function chain can meet the current demand of all primary service function chains, the first information containing the confirmation information is taken as the first information corresponding to each first base station; (2) in the case where the determination result indicates that the computing resources of each base station on the service function chain cannot meet the current demand of all primary service function chains, the primary service function chain arrangement information is updated to determine the first base station in which the primary service function chain is changed and the first base station in which the primary service function chain remains unchanged; the first information containing the target decision information is taken as the first information corresponding to the first base station in which the primary service function chain is changed; the first information containing the confirmation information is taken as the first information corresponding to the first base station in which the primary service function chain remains unchanged; the target decision information includes the final service function chain chaining decision information corresponding to the first base station in which the primary service function chain is changed.

[0527] Further, the information processing device further includes a fifth sending module configured to send the second information to each base station on the service function chain; the second information includes at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of the physical positions of each base station; information of the mutual connection conditions of each base station; each base station is connected to the cloud computing center device.

[0528] The second information is inserted into a MAC layer transport block as a MAC control signal element and is transmitted through a transmission channel.

[0529] Further, the information processing device further includes: a fourth receiving module configured to receive local state information of the base stations on the service function chain respectively transmitted by the base stations; wherein the local state information is related to the execution of the AI task; and a third processing module configured to obtain the second information according to the local state information.

[0530] The local state information includes at least one of the following: information of an AI model stored in the base station; information of remaining computing resources in the base station; information of storage resources in the base station; information of a physical location of the base station; and information of a connection condition of the base station with other base stations.

[0531] In the embodiment of the application, the local state information is inserted into a MAC layer transport block as a MAC control signal element and is transmitted through a transmission channel.

[0532] Further, the information processing device further includes: a second obtaining module configured to obtain initial parameter information of at least one AI model; a fourth processing module configured to perform a classification and aggregation operation on the initial parameter information to obtain final parameter information of each AI model; and a sixth sending module configured to send the final parameter information of the AI model to a base station corresponding to the AI model.

[0533] The obtaining of the initial parameter information of the at least one AI model includes at least one of the following: (1) receiving initial parameter information of an AI model transmitted by at least one base station; and (2) receiving an AI subtask transmitted by the first base station, training a corresponding AI model according to the AI subtask to obtain initial parameter information of the AI model.

[0534] In the embodiment of the application, the training of the corresponding AI model according to the AI subtask includes training the corresponding AI model according to the AI subtask by using a transfer learning method.

[0535] The implementation embodiments of the information processing method on the cloud computing center device side are applicable to the embodiments of the information processing device and can achieve the same technical effects.

[0536] The embodiment of the application further provides an information processing device applied to a first base station, as shown in the figure, which includes: Figure 11

[0537] ​The fifth processing module 111 is configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks.

[0538] The sixth processing module 112 is configured to determine a training party of an AI model corresponding to each AI subtask, and obtain initial parameter information of the AI model by using the training party; the training party is the first base station, a cloud computing center device connected to the first base station, or another base station under the cloud computing center device except the first base station.

[0539] The fifth receiving module 113 is configured to receive final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0540] The information processing device provided by the embodiment of the application can decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks, determine a training party of an AI model corresponding to each AI subtask, obtain initial parameter information of the AI model by using the training party, and receive final parameter information obtained by a cloud computing center device according to the initial parameter information, thereby avoiding problems such as link congestion and poor data security caused by training data uploading, occupation of a large amount of computing resources of the cloud computing center, and large AI task processing delay when all AI models are trained in the cloud computing center, and further supporting improvement of orchestration effect and better solution to the problem of poor orchestration effect caused by lack of a global perspective in an existing information processing scheme for AI service function chain orchestration.

[0541] The determination of the training party of the AI model corresponding to each AI subtask includes: (1) in the case where the AI model corresponding to the AI subtask exists locally in the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station; and (2) in the case where the AI model corresponding to the AI subtask does not exist locally in the first base station, determining, according to second information, whether the other base station stores the AI model corresponding to the AI subtask; in the case where the AI model is stored, determining that the training party of the AI model corresponding to the AI subtask is the other base station; and in the case where the AI model is not stored, determining that the training party of the AI model corresponding to the AI subtask is the cloud computing center device; the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; and the global state information is related to execution of an AI task.

[0542] In the embodiment of the present application, the second information includes at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical locations of each base station; information of mutual connection conditions of each base station; and wherein each base station is connected to the cloud computing center device.

[0543] In the embodiment of the present application, the second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel; and wherein the second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0544] The implementation embodiments of the first base station side information processing method described above are applicable to the embodiments of the information processing device, and can achieve the same technical effects.

[0545] The embodiment of the present application further provides an information processing device, which is applied to a cloud computing center device, as shown in the accompanying drawings, and includes: Figure 12

[0546] The third acquisition module 121 is configured to acquire initial parameter information of at least one AI model.

[0547] The seventh processing module 122 is configured to perform a classification and aggregation operation on the initial parameter information to obtain final parameter information of each AI model.

[0548] The seventh sending module 123 is configured to send the final parameter information of the AI model to a base station corresponding to the AI model.

[0549] The information processing device provided by the embodiment of the present application can acquire initial parameter information of at least one AI model, perform a classification and aggregation operation on the initial parameter information to obtain final parameter information of each AI model, and send the final parameter information of the AI model to a base station corresponding to the AI model, thereby supporting the avoidance of problems such as link congestion and poor data security caused by training data upload, occupation of a large amount of computing resources of the cloud computing center, and large AI task processing delay when the cloud computing center completes the training of all AI models, and further supporting the improvement of the orchestration effect and the better solution to the problem of poor orchestration effect caused by the lack of a global perspective in the prior art information processing scheme for AI service function chain orchestration.

[0550] ​The initial parameter information of the at least one AI model is obtained, including at least one of the following: (1) receiving initial parameter information of an AI model sent by at least one base station; (2) receiving an AI subtask sent by a first base station, training a corresponding AI model according to the AI subtask, and obtaining initial parameter information of the AI model.

[0551] In the embodiment of the application, the AI model is trained according to the AI subtask by using a transfer learning method.

[0552] Further, the information processing device further comprises an eighth sending module configured to send second information to at least one base station respectively, wherein the second information is global state information corresponding to the cloud computing center device, and the global state information is related to the execution of the AI task.

[0553] The second information includes at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; information of mutual connection conditions of each base station, wherein each base station is connected to the cloud computing center device.

[0554] In the embodiment of the application, the second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel; and the second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0555] The implementation embodiments of the information processing method on the cloud computing center device side are applicable to the embodiments of the information processing device, and the same technical effects can be achieved.

[0556] The embodiment of the application further provides an information processing device, which is a first base station, as shown in Figure 13 The information processing device comprises a processor 131 and a transceiver 132.

[0557] The processor 131 is configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks.

[0558] According to the at least two different AI subtasks, the transceiver 132 is configured to send primary service function chain orchestration information to a cloud computing center device.

[0559] receive, by the transceiver 132, first information sent by the cloud computing center device; the first information comprises final service function chain forming decision information or confirmation information;

[0560] determine, according to the first information, a final service function chain.

[0561] The information processing device provided by the embodiment of the application can support the implementation of the first base station and the cloud computing center device cooperating to determine the AI service function chain arrangement based on the global perspective (second information), thereby improving the arrangement effect and solving the problem of poor arrangement effect caused by the lack of global perspective in the prior art information processing scheme for the AI service function chain arrangement.

[0562] According to the at least two different AI sub-tasks, the transceiver sends primary service function chain arrangement information to the cloud computing center device, comprising: arranging the service positions of each AI sub-task according to the second information to form a primary service function chain and obtain primary service function chain arrangement information; wherein the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; the global state information is related to the execution of the AI task; and the transceiver sends the primary service function chain arrangement information to the cloud computing center device.

[0563] In the embodiment of the application, the second information comprises at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; information of mutual connection conditions of each base station; wherein each base station is connected to the cloud computing center device.

[0564] The second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel; and the second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0565] Further, the processor is further configured to send, by the transceiver, local state information of the first base station to the cloud computing center device; wherein the local state information is related to the execution of the AI task.

[0566] The local state information includes at least one of the following: information of an AI model stored in the first base station; information of remaining computing resources in the first base station; information of storage resources in the first base station; information of a physical location of the first base station; and information of a connection status of the first base station with other base stations.

[0567] In the embodiment of the application, the local state information is inserted into a MAC layer transmission block as a MAC control signal element and is transmitted through a transmission channel.

[0568] Further, the processor is further configured to execute the AI task by using the final service function chain to obtain a processing result, and send the processing result to the terminal through the transceiver.

[0569] In the embodiment of the application, the processor is further configured to, before executing the AI task by using the final service function chain to obtain a processing result, determine a training party of an AI model corresponding to each AI subtask, obtain initial parameter information of the AI model by using the training party, the training party being the first base station, the cloud computing center device, or another base station under the cloud computing center device other than the first base station, receive final parameter information obtained by the cloud computing center device according to the initial parameter information through the transceiver, and execute the AI task by using the final service function chain and the final parameter information to obtain a processing result.

[0570] In the embodiment of the application, the determination of the training party of the AI model corresponding to each AI subtask includes: (1) in the case where the AI model corresponding to the AI subtask exists locally in the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station; and (2) in the case where the AI model corresponding to the AI subtask does not exist locally in the first base station, determining whether the AI model corresponding to the AI subtask exists in the other base station according to the second information, determining that the training party of the AI model corresponding to the AI subtask is the other base station in the case where the AI model exists, and determining that the training party of the AI model corresponding to the AI subtask is the cloud computing center device in the case where the AI model does not exist.

[0571] The implementation embodiments of the information processing method on the first base station side are applicable to the implementation embodiments of the information processing device, and the same technical effects can be achieved.

[0572] The embodiment of the application further provides an information processing device, which is a cloud computing center device, such as a cloud server. Figure 14As shown, comprising: a processor 141 and a transceiver 142;

[0573] The processor 141 is configured to receive, by the transceiver 142, primary service function chain arrangement information sent by at least one first base station;

[0574] According to the primary service function chain arrangement information and the second information, determine whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains, and obtain a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to the execution of an AI task;

[0575] According to the determination result, obtain first information corresponding to each first base station respectively; the first information includes final service function chain chain forming decision information or confirmation information;

[0576] Send the corresponding first information to each first base station through the transceiver 142.

[0577] The information processing device provided by the embodiment of the application receives the primary service function chain arrangement information sent by at least one first base station; according to the primary service function chain arrangement information and the second information, determines whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains, and obtains a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to the execution of an AI task; according to the determination result, obtain first information corresponding to each first base station respectively; the first information includes final service function chain chain forming decision information or confirmation information; send the corresponding first information to each first base station; can support the realization of the first base station and the cloud computing center device cooperation based on the global perspective (second information) to determine the AI service function chain arrangement, thereby improving the arrangement effect, and well solving the problem that the information processing scheme for the AI service function chain arrangement in the prior art lacks global perspective, resulting in poor arrangement effect.

[0578] The first information corresponding to each first base station is obtained according to the determination result, including: (1) in the case that the determination result indicates that the computing resources of each base station on the service function chain can meet the current demand of all primary service function chains, the first information containing confirmation information is taken as the first information corresponding to each first base station; (2) in the case that the determination result indicates that the computing resources of each base station on the service function chain cannot meet the current demand of all primary service function chains, the primary service function chain scheduling information is updated, the first base station in which the primary service function chain is changed and the first base station in which the primary service function chain is maintained are determined; the first information containing target decision information is taken as the first information corresponding to the first base station in which the primary service function chain is changed; the first information containing confirmation information is taken as the first information corresponding to the first base station in which the primary service function chain is maintained; wherein the target decision information includes the final service function chain forming decision information corresponding to the first base station in which the primary service function chain is changed.

[0579] Further, the processor is further configured to: send, through the transceiver, the second information to each base station on the service function chain respectively; wherein the second information includes at least one of the following: information of an AI model stored in each base station and a cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of a physical location of each base station; information of a mutual connection of each base station; wherein each base station is connected to the cloud computing center device.

[0580] The second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel; wherein the second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0581] Further, the processor is further configured to: receive, through the transceiver, the local state information of the base station sent by each base station on the service function chain respectively; wherein the local state information is related to the execution of the AI task; and obtain the second information according to the local state information.

[0582] The local state information includes at least one of the following: information of an AI model stored in the base station; information of remaining computing resources in the base station; information of storage resources in the base station; information of a physical location of the base station; information of a connection of the base station with other base stations.

[0583] In the embodiment of the present application, the local state information is inserted into a MAC layer transport block as a MAC control signal element and transmitted through a transmission channel.

[0584] Further, the processor is further configured to: obtain initial parameter information of at least one AI model; perform a classification aggregation operation on the initial parameter information to obtain final parameter information of each AI model; and send the final parameter information of the AI model to a base station corresponding to the AI model through the transceiver.

[0585] The obtaining of the initial parameter information of the at least one AI model includes at least one of the following: (1) receiving initial parameter information of an AI model sent by at least one base station through the transceiver; and (2) receiving an AI subtask sent by the first base station through the transceiver, training a corresponding AI model according to the AI subtask to obtain initial parameter information of the AI model.

[0586] In the embodiment of the present application, the training of the corresponding AI model according to the AI subtask includes training the corresponding AI model according to the AI subtask by using a transfer learning method.

[0587] The implementation embodiments of the information processing method on the cloud computing center device side are applicable to the embodiments of the information processing device and can achieve the same technical effects.

[0588] The embodiment of the present application also provides an information processing device, which is a first base station, as shown in Figure 15 The information processing device includes a processor 151 and a transceiver 152.

[0589] The processor 151 is configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks.

[0590] The processor 151 is configured to determine a training party of an AI model corresponding to each AI subtask, obtain initial parameter information of the AI model by using the training party, and determine the training party as the first base station, a cloud computing center device connected to the first base station, or another base station under the cloud computing center device other than the first base station.

[0591] The transceiver 152 is configured to receive final parameter information obtained by the cloud computing center device according to the initial parameter information.

[0592] The information processing device provided by the embodiment of the present application can avoid the problems of link congestion and poor data security caused by training data uploading, occupation of a large amount of computing resources of the cloud computing center, and large AI task processing delay, which exist when the cloud computing center completes the training of all AI models, thereby further supporting the improvement of the orchestration effect and better solving the problem of poor orchestration effect caused by the lack of a global perspective in the prior art information processing scheme for AI service function chain orchestration.

[0593] In the embodiment of the present application, the determination of the training party of the AI model corresponding to each AI subtask comprises: (1) in the case that the AI model corresponding to the AI subtask exists locally in the first base station, determining that the training party of the AI model corresponding to the AI subtask is the first base station; (2) in the case that the AI model corresponding to the AI subtask does not exist locally in the first base station, determining whether the other base station stores the AI model corresponding to the AI subtask according to second information; in the case that the AI model is stored, determining that the training party of the AI model corresponding to the AI subtask is the other base station; in the case that the AI model is not stored, determining that the training party of the AI model corresponding to the AI subtask is the cloud computing center; wherein the second information is global state information corresponding to the cloud computing center to which the first base station belongs; and the global state information is related to the execution of the AI task.

[0594] In the embodiment of the present application, the second information comprises at least one of the following: information of AI models already stored in each base station and the cloud computing center; information of remaining computing resources in each base station and the cloud computing center; information of remaining storage resources in each base station and the cloud computing center; information of physical positions of each base station; information of mutual connection conditions of each base station; wherein each base station is connected to the cloud computing center.

[0595] In the embodiment of the present application, the second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel; wherein the second information is located at the start position of a MAC protocol data unit in the MAC layer transport block.

[0596] The implementation embodiments of the first base station side information processing method are applicable to the information processing device embodiments, and the same technical effects can be achieved.

[0597] The embodiment of the present application also provides an information processing device, which is a cloud computing center device, such as Figure 16 as shown, comprising a processor 161 and a transceiver 162;

[0598] The processor 161 is configured to obtain initial parameter information of at least one AI model;

[0599] The initial parameter information is classified and aggregated to obtain final parameter information of each AI model;

[0600] The transceiver 162 is configured to send the final parameter information of the AI model to the base station corresponding to the AI model.

[0601] The information processing device provided by the embodiment of the present application obtains initial parameter information of at least one AI model, classifies and aggregates the initial parameter information to obtain final parameter information of each AI model, and sends the final parameter information of the AI model to the base station corresponding to the AI model. It can support to avoid the problems of link congestion and poor data security caused by training data upload, occupation of a large number of computing resources of the cloud computing center, large AI task processing delay and other problems caused by completing all AI model training in the cloud computing center; thereby further supporting to improve the arrangement effect, and better solving the problem of poor arrangement effect caused by lack of global perspective in the prior art information processing scheme for AI service function chain arrangement.

[0602] The initial parameter information of at least one AI model includes at least one of the following: (1) receiving the initial parameter information of the AI model sent by at least one base station through the transceiver; (2) receiving the AI subtask sent by the first base station through the transceiver, training the corresponding AI model according to the AI subtask, and obtaining the initial parameter information of the AI model.

[0603] In the embodiment of the present application, the corresponding AI model is trained according to the AI subtask, which includes training the corresponding AI model according to the AI subtask by using the transfer learning method.

[0604] Further, the processor is further configured to send second information to at least one base station through the transceiver; wherein the second information is global state information corresponding to the cloud computing center device; and the global state information is related to executing an AI task.

[0605] The second information includes at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; information of mutual connection conditions of each base station; and the like.

[0606] In the embodiment of the application, the second information is inserted into a MAC layer transport block as a MAC control element, and is transmitted through a transmission channel; and the second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

[0607] The implementation embodiments of the information processing method on the cloud computing center device side are applicable to the embodiments of the information processing device, and the same technical effects can be achieved.

[0608] The embodiment of the application further provides an information processing device, which comprises a memory, a processor, and a program stored in the memory and executable on the processor; and the processor implements the information processing method on the first base station side or the cloud computing center device side when executing the program.

[0609] The implementation embodiments of the information processing method on the first base station side or the cloud computing center device side are applicable to the embodiments of the information processing device, and the same technical effects can be achieved.

[0610] The embodiment of the application further provides a readable storage medium, which stores a program; and the program is executable on a processor to implement the steps in the information processing method on the first base station side or the cloud computing center device side.

[0611] The implementation embodiments of the information processing method on the first base station side or the cloud computing center device side are applicable to the embodiments of the readable storage medium, and the same technical effects can be achieved.

[0612] It should be noted that many functional components described in the specification are referred to as modules, so as to more particularly emphasize their implementation independence.

[0613] In the embodiments of the present application, the modules can be implemented in software, and executed by various types of processors. For example, an identified module of executable code can consist of one or more physical or logical blocks of computer instructions. For example, a module can be implemented in object-oriented programming, such as a Java® class, or in procedural programming, such as a C function. However, the executable code of an identified module need not be physically located together, but can be distributed in various places of a computer memory, and across multiple storage devices. Similarly, operational data can be identified within modules and can be

[0614] Indeed, an executable code module can be a single instruction, or many instructions, and can even be distributed over several different code segments, among different programs, and across several memory devices. Also, operational data can be identified within modules and can be

[0615] When a module is implemented in software, the module can be stored in any desired manner, including as a proprietary binary format executable by a processor, as source code, or as any other desired format. The software implementation can be stored in any desired storage medium, including a compact diskette, a floppy disk, a tape, a hard disk drive, a memory device, or a read-only memory device, among others. The software implementation can also be transmitted over a variety of different means, including a communications network, a modem, a laser, or any other desired means. The source code can be adapted by those of ordinary skill in the art to suit the particular needs of the application, while maintaining the principles of the application.

[0616] The foregoing is considered as illustrative only of the principles of the application. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the application to the exact construction and practice described. Accordingly, all such variations are intended to be included within the scope of the present application.

Claims

1. An information processing method applied to a first base station, characterized in that, The method comprises the following steps: decomposing an artificial intelligence (AI) task of a terminal into at least two different AI sub-tasks; sending primary service function chain arrangement information to a cloud computing center device according to the at least two different AI sub-tasks; receiving first information sent by the cloud computing center device; the first information comprises final service function chain chaining decision information or confirmation information; determining a final service function chain according to the first information; wherein the step of sending the primary service function chain arrangement information to the cloud computing center device according to the at least two different AI sub-tasks comprises the following steps: arranging service positions of the AI sub-tasks according to second information to form a primary service function chain and obtain primary service function chain arrangement information; wherein the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; the global state information is related to execution of the AI task; sending the primary service function chain arrangement information to the cloud computing center device.

2. The information processing method according to claim 1, characterized by, The second information comprises at least one of the following: information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical positions of each base station; information of mutual connection conditions of each base station; wherein each base station is connected to the cloud computing center device.

3. The information processing method according to claim 1, characterized by, The second information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel; wherein the second information is located at a start position of a MAC protocol data unit in the MAC layer transport block.

4. The information processing method according to claim 1, characterized by, The method further comprises the following steps: sending local state information of the first base station to the cloud computing center device; wherein the local state information is related to execution of the AI task.

5. The information processing method according to claim 4, characterized by, The local state information comprises at least one of the following: information of AI models stored in the first base station; information of remaining computing resources in the first base station; information of storage resources in the first base station; information of a physical position of the first base station; information of connection conditions of the first base station and other base stations.

6. The information processing method according to claim 4, characterized by, The local state information is inserted into a MAC layer transport block as a MAC control element and is transmitted through a transmission channel.

7. The information processing method according to claim 1, characterized by, The method further comprises the following steps: executing the AI task by using the final service function chain to obtain a processing result; sending the processing result to the terminal.

8. The information processing method according to claim 7, characterized by, Before the step of executing the AI task by using the final service function chain to obtain a processing result, the method further comprises the following steps: determining a training party of an AI model corresponding to each AI sub-task to obtain initial parameter information of the AI model by using the training party; the training party is the first base station, the cloud computing center device or other base stations under the cloud computing center device except the first base station; receiving final parameter information obtained by the cloud computing center device according to the initial parameter information; the step of executing the AI task by using the final service function chain to obtain a processing result comprises the following steps: According to the final service function chain and final parameter information, an AI model is used to execute the AI task to obtain a processing result.

9. The information processing method according to claim 8, characterized by, The method further includes: In a case where the AI sub-task corresponding AI model exists in the first base station, the training party of the AI sub-task corresponding AI model is determined as the first base station; In a case where the AI sub-task corresponding AI model does not exist in the first base station, whether the other base station stores the AI sub-task corresponding AI model is determined according to the second information; in a case where the AI sub-task corresponding AI model is stored, the training party of the AI sub-task corresponding AI model is determined as the other base station; in a case where the AI sub-task corresponding AI model is not stored, the training party of the AI sub-task corresponding AI model is determined as the cloud computing center device.

10. An information processing method applied to cloud computing center equipment, characterized in that, The method further includes: Receiving primary service function chain arrangement information sent by at least one first base station; According to the primary service function chain arrangement information and second information, determining whether the computing resources of each base station on the service function chain meet the current demand of all primary service function chains to obtain a determination result; the second information is global state information corresponding to the cloud computing center device; The global state information is related to the execution of the AI task; According to the determination result, obtaining first information corresponding to each first base station respectively; The first information includes final service function chain chain decision information or confirmation information; Sending the corresponding first information to each first base station; The information processing method further includes: Receiving local state information of each base station on the service function chain respectively sent by the base station; the local state information is related to the execution of the AI task; According to the local state information, the second information is obtained.

11. The information processing method according to claim 10, characterized by, According to the determination result, obtaining first information corresponding to each first base station respectively, includes: In a case where the determination result indicates that the computing resources of each base station on the service function chain can meet the current demand of all primary service function chains, the first information containing the confirmation information is taken as the first information corresponding to each first base station respectively; In a case where the determination result indicates that the computing resources of each base station on the service function chain cannot meet the current demand of all primary service function chains, at least one primary service function chain arrangement information is updated, the first base station of which the primary service function chain is changed and the first base station of which the primary service function chain is maintained are determined; the first information containing the target decision information is taken as the first information corresponding to the first base station of which the primary service function chain is changed; the first information containing the confirmation information is taken as the first information corresponding to the first base station of which the primary service function chain is maintained; the target decision information includes final service function chain chain decision information corresponding to the first base station of which the primary service function chain is changed.

12. The information processing method according to claim 10, characterized by, The method further includes: Sending the second information to each base station on the service function chain respectively; The second information includes at least one of the following: Information of AI models stored in each base station and the cloud computing center device; information of remaining computing resources in each base station and the cloud computing center device; information of remaining storage resources in each base station and the cloud computing center device; information of physical locations of each base station; information of mutual connection conditions of each base station; wherein each base station is connected with the cloud computing center device.

13. The information processing method according to claim 10, characterized by, The local state information includes at least one of: information of an AI model stored in the base station; information of remaining computing resources in the base station; information of storage resources in the base station; information of a physical location of the base station; information of connection conditions of the base station with other base stations.

14. The information processing method according to claim 10, characterized by, Further comprising: obtaining initial parameter information of at least one AI model; performing a classification aggregation operation on the initial parameter information to obtain final parameter information of each AI model; sending the final parameter information of the AI model to the base station corresponding to the AI model.

15. The information processing method according to claim 14, characterized by, The obtaining of the initial parameter information of at least one AI model includes at least one of: receiving initial parameter information of an AI model sent by at least one base station; receiving an AI subtask sent by the first base station, training a corresponding AI model according to the AI subtask to obtain initial parameter information of the AI model.

16. The information processing method according to claim 15, characterized by, The training of the corresponding AI model according to the AI subtask includes: training the corresponding AI model according to the AI subtask by using a transfer learning manner.

17. An information processing apparatus for a first base station, comprising: Comprising: a first processing module, configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks; a first sending module, configured to send primary service function chain arrangement information to a cloud computing center device according to the at least two different AI subtasks; a first receiving module, configured to receive first information sent by the cloud computing center device; the first information includes final service function chain chaining decision information or confirmation information; a first determining module, configured to determine a final service function chain according to the first information; wherein the sending of the primary service function chain arrangement information to the cloud computing center device according to the at least two different AI subtasks includes: arranging service locations of each AI subtask according to second information to form a primary service function chain, to obtain primary service function chain arrangement information; wherein the second information is global state information corresponding to a cloud computing center device to which the first base station belongs; the global state information is related to execution of an AI task; sending the primary service function chain arrangement information to the cloud computing center device.

18. An information processing apparatus applied to a cloud computing center device, comprising: Comprising: a third receiving module, configured to receive primary service function chain arrangement information sent by at least one first base station; a second determining module, configured to determine whether computing resources of each base station on a service function chain meet current requirements of all primary service function chains according to the primary service function chain arrangement information and second information, to obtain a determination result; the second information is global state information corresponding to the cloud computing center device; the global state information is related to execution of an AI task; a first obtaining module, configured to obtain first information corresponding to each first base station respectively according to the determination result; the first information includes final service function chain chaining decision information or confirmation information; A fourth sending module is configured to send corresponding first information to each of the first base stations. The information processing device further includes: A fourth receiving module is configured to receive local state information of each base station on the service function chain, which is respectively sent by the base station; the local state information is related to the AI task execution. A third processing module is configured to obtain the second information according to the local state information.

19. An information processing apparatus of a first base station, the information processing apparatus comprising: The information processing device includes: A processor and a transceiver; The processor is configured to decompose an artificial intelligence (AI) task of a terminal into at least two different AI subtasks. The transceiver is configured to send primary service function chain arrangement information to a cloud computing center device according to the at least two different AI subtasks. The transceiver is configured to receive first information sent by the cloud computing center device. The first information includes final service function chain chaining decision information or confirmation information. The final service function chain is determined according to the first information. The transceiver is configured to send the primary service function chain arrangement information to the cloud computing center device according to the second information, to form the primary service function chain and obtain the primary service function chain arrangement information; the second information is global state information corresponding to the cloud computing center device to which the first base station belongs; the global state information is related to the AI task execution. The transceiver is configured to send the primary service function chain arrangement information to the cloud computing center device. The information processing device includes:

20. An information processing apparatus that is a cloud computing center apparatus, comprising: A processor and a transceiver; The processor is configured to receive primary service function chain arrangement information sent by at least one first base station through the transceiver. The processor is configured to determine whether computing resources of each base station on a service function chain meet current needs of all primary service function chains according to the primary service function chain arrangement information and second information, to obtain a determination result; the second information is global state information corresponding to the cloud computing center device. The global state information is related to the AI task execution. The processor is configured to obtain first information corresponding to each of the first base stations according to the determination result. The first information includes final service function chain chaining decision information or confirmation information. The transceiver is configured to send the corresponding first information to each of the first base stations. The processor is further configured to: The transceiver is configured to receive local state information of each base station on the service function chain, which is respectively sent by the base station; the local state information is related to the AI task execution. The processor is configured to obtain the second information according to the local state information. The processor implements the information processing method in any one of claims 1 to 16 when executing the program.

21. An information processing apparatus comprising a storage, a processor, and a program stored on the storage and executable on the processor; characterized by, The program is executed by the processor to implement the steps in the information processing method in any one of claims 1 to 16.

22. A readable storage medium, having stored thereon a program, wherein the program is configured to cause a processor to perform the method according to any one of claims 1-21. ​

Images