Model management method based on real-time intelligence, and related device

Abstract

The present disclosure relates to the technical field of communications, and provides a model management method based on real-time intelligence, and a related device. The method comprises: receiving a service scene request sent by a network element, wherein the service scene request is used for requesting an artificial intelligence model that the network element needs to use in a service scene; forwarding the service scene request to a near-real-time wireless intelligent controller, so that the near-real-time wireless intelligent controller searches for the artificial intelligence model and sends the artificial intelligence model to a real-time wireless intelligent controller; and receiving the artificial intelligence model sent by the near-real-time wireless intelligent controller, and sending the artificial intelligence model to the network element. By means of the cooperation with a near-real-time wireless intelligent controller, an artificial intelligence model can be fed back to a network element having a service scene request with low time delay, and real-time service is provided.

Description

Model management method and related equipment based on real-time intelligence

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This disclosure claims priority to Chinese patent application number 202311214421.X, filed on September 19, 2023, entitled “Model management method, device, system, equipment and medium based on real-time intelligence”, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present disclosure relates to the field of communication technology, and in particular to a model management method based on real-time intelligence and related equipment. Background Art

[0004] With the implementation of 5G research and the development of 6G research, 6G research revolves around the intelligent connection of all things. Based on this, the inherent intelligence of 6G has received more and more attention. In the research of inherent intelligence, intelligence can be divided into non-real-time intelligence, near real-time intelligence and real-time intelligence through the division of latency requirements.

[0005] In related technologies, for real-time intelligence, how to manage artificial intelligence (AI) models and meet the real-time requirements of low latency is an urgent problem to be solved.

[0006] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute prior art known to ordinary technicians in the field.

[0007] Summary of the Invention

[0008] The present disclosure provides a model management method and related equipment based on real-time intelligence, which at least to a certain extent overcomes the management of artificial intelligence models based on real-time intelligence in related technologies to solve the low-latency network's own needs and the real-time needs of business scenarios.

[0009] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by practice of the present disclosure.

[0010] In a first aspect, embodiments of the present disclosure provide a real-time intelligence-based model management method, applied to a real-time wireless intelligent controller; the method comprises:

[0011] Receive a business scenario request sent by a network element; the business scenario request is used to request the network element to use an artificial intelligence model in the business scenario;

[0012] Forwarding the business scenario request to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller;

[0013] Receive the artificial intelligence model sent by the near real-time wireless intelligent controller, and send the artificial intelligence model to the network element.

[0014] In a possible embodiment, forwarding the service scenario request to the near real-time wireless intelligent controller includes:

[0015] Detecting whether the artificial intelligence model exists in the real-time wireless intelligent controller;

[0016] If the real-time wireless intelligent controller does not have the artificial intelligence model, determining whether the computing resources of the real-time wireless intelligent controller are sufficient;

[0017] If the computing resources are insufficient, the service scenario request is forwarded to a near real-time wireless intelligent controller.

[0018] In a possible embodiment, the business scenario request includes a requested artificial intelligence model, and / or input parameters, output parameters, and model type that need to be input into the artificial intelligence model.

[0019] In a possible embodiment, after receiving the service scenario request sent by the network element, the method further includes:

[0020] Detecting whether the artificial intelligence model exists in the real-time wireless intelligent controller;

[0021] If it is determined that the artificial intelligence model is available, requesting screening of the artificial intelligence model according to the business scenario;

[0022] Sending the artificial intelligence model to the network element.

[0023] In a possible embodiment, after receiving the service scenario request sent by the network element, the method further includes:

[0024] Detecting whether the artificial intelligence model exists in the real-time wireless intelligent controller;

[0025] If it is determined that the artificial intelligence model is not available, determining whether the computing resources of the real-time wireless intelligent controller are sufficient;

[0026] If the computing resources are sufficient, the artificial intelligence model is trained and evaluated in real time until a trained artificial intelligence model is obtained;

[0027] The trained artificial intelligence model is sent to the network element.

[0028] In a possible embodiment, after receiving the artificial intelligence model sent by the near real-time wireless intelligent controller, the method further includes:

[0029] In the real-time model library, the artificial intelligence model is cached.

[0030] In a possible embodiment, the method further includes:

[0031] Receive network data in various real-time business scenarios;

[0032] sending the network data and the data processing instruction to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller classifies the network data according to the data processing instruction; and

[0033] The training process of multiple intelligent models is completed based on the network data.

[0034] In a second aspect, embodiments of the present disclosure provide a model management method based on real-time intelligence, applied to a near real-time wireless intelligent controller, the method comprising:

[0035] Receive a business scenario request sent by a real-time wireless intelligent controller; the business scenario request is used to request an artificial intelligence model that a network element needs to use in the business scenario;

[0036] The artificial intelligence model is searched and sent to the real-time wireless intelligent controller.

[0037] In a possible embodiment, searching for the artificial intelligence model and sending it to the real-time wireless intelligent controller includes:

[0038] Requesting testing of the artificial intelligence model according to the business scenario;

[0039] The artificial intelligence model is searched in a near real-time model library and sent to the real-time wireless intelligent controller.

[0040] In a possible embodiment, the method further includes:

[0041] Receiving network data and data processing instructions sent by the real-time wireless intelligent controller;

[0042] According to the data processing instructions, the network data is classified, and a training process of multiple intelligent models is completed based on the network data.

[0043] In a third aspect, embodiments of the present disclosure provide a real-time intelligent model management device, which is applied to a real-time wireless intelligent controller; comprising:

[0044] A first receiving unit is configured to receive a business scenario request sent by a network element; the business scenario request is used to request an artificial intelligence model that the network element needs to use in the business scenario;

[0045] a sending unit configured to forward the service scenario request to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller searches for the artificial intelligence model and sends the artificial intelligence model to the real-time wireless intelligent controller;

[0046] The first transmission unit is configured to receive the artificial intelligence model sent by the near real-time wireless intelligent controller and send the artificial intelligence model to the network element.

[0047] In a fourth aspect, embodiments of the present disclosure provide a model management device based on real-time intelligence, applied to a near-real-time wireless intelligent controller; comprising:

[0048] A second receiving unit is configured to receive a service scenario request sent by the real-time wireless intelligent controller; the service scenario request is used to request an artificial intelligence model that the network element needs to use in the service scenario;

[0049] The second transmission unit is configured to search for the artificial intelligence model and send it to the real-time wireless intelligent controller.

[0050] In a fifth aspect, embodiments of the present disclosure provide a radio access network, comprising: a programmable intelligent layer and a programmable radio access network RAN ​​plane;

[0051] The programmable RAN plane includes a central unit-control plane and network elements; the programmable intelligent layer includes a central unit-control plane and a near real-time wireless intelligent controller; the central unit-control plane includes a real-time wireless intelligent controller; the real-time wireless intelligent controller and the near real-time wireless intelligent controller are connected via a controller interaction interface;

[0052] The real-time wireless intelligent controller is configured to execute the method described in any one of the first aspects; the near real-time wireless intelligent controller is configured to execute the method described in any one of the second aspects.

[0053] In a sixth aspect, an embodiment of the present disclosure provides an electronic device, comprising: a processor; and a memory configured to store executable instructions of the processor; wherein the processor is configured to execute the methods described in the first and second aspects above by executing the executable instructions.

[0054] In a seventh aspect, an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, which implements the methods described in the first and second aspects above when the computer program is executed by a processor.

[0055] In an eighth aspect, according to another aspect of the present disclosure, a computer program product or computer program is further provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform any of the methods described above.

[0056] The embodiments of the present disclosure provide a real-time intelligence-based model management method and related equipment, specifically relating to a real-time intelligence-based model management method, apparatus, wireless access network, equipment, medium, and program product. The method includes: receiving a business scenario request sent by a network element, the business scenario request is used to request an artificial intelligence model that the network element needs to use in the business scenario; forwarding the business scenario request to a near-real-time wireless intelligent controller so that the near-real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller; receiving the artificial intelligence model sent by the near-real-time wireless intelligent controller and sending the artificial intelligence model to the network element. By collaborating with the near-real-time wireless intelligent controller, the artificial intelligence model can be fed back to the network element with a low-latency business scenario request, providing real-time business services, saving the storage resources of the real-time wireless intelligent controller, increasing the utilization of communication resources, and improving communication quality.

[0057] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification, are used to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0059] FIG1 shows a schematic structural diagram of a wireless access network according to an embodiment of the present disclosure;

[0060] FIG2 shows a schematic structural diagram of another wireless access network according to an embodiment of the present disclosure;

[0061] FIG3 is a flow chart showing a model management method based on real-time intelligence in an embodiment of the present disclosure;

[0062] FIG4 is a flow chart showing another model management method based on real-time intelligence in an embodiment of the present disclosure;

[0063] FIG5 is a flow chart showing another model management method based on real-time intelligence in an embodiment of the present disclosure;

[0064] FIG6 shows an interactive schematic diagram of a model management method based on real-time intelligence in an embodiment of the present disclosure;

[0065] FIG7 shows a schematic structural diagram of a model management device based on real-time intelligence in an embodiment of the present disclosure;

[0066] FIG8 shows a schematic structural diagram of another model management device based on real-time intelligence in an embodiment of the present disclosure;

[0067] FIG9 shows a schematic structural diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0068] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be embodied in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0069] In addition, the accompanying drawings are merely schematic illustrations of the present disclosure and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0070] The evolution of 6G networks will see the addition of intelligent elements, placing new demands on the openness and programmability of wireless access networks. The intelligent interconnection of everything proposed by 6G is placing increasing importance on research into endogenous intelligence. Intelligence can be categorized as non-real-time, near-real-time, and real-time. Due to latency requirements, real-time intelligence primarily addresses services requiring latency under 100ms, such as beamforming and physical layer coding on the RAN itself. Near-real-time intelligence can handle services with latency between 100ms and 1s. Service scenarios for near-real-time intelligence can include Quality of Service (QoS) guarantees and Radio Access Network (RAN) slicing.

[0071] However, for low-latency services, such as those with latency requirements within 100ms, there is currently no good way to handle them in the real-time intelligent 6G network. Based on this, the embodiments of the present disclosure provide a model management method based on real-time intelligence, which is aimed at deploying a real-time wireless intelligent controller in the wireless access network to realize the processing of low-latency service scenario requests.

[0072] The real-time intelligence-based model management method disclosed in the present invention can be applied to electronic devices and also to wireless access networks.

[0073] FIG1 shows a schematic structural diagram of a wireless access network that can be applied to an embodiment of the present disclosure.

[0074] As shown in FIG. 1 , a radio access network 100 may include a programmable intelligent layer 110 and a programmable radio access network plane 120 .

[0075] The programmable intelligent layer 110 includes a central unit-control plane 130 (CU-CP) and a near real-time wireless intelligent controller 140 .

[0076] The programmable RAN plane 120 includes a central unit-control plane 130 and network elements 150 .

[0077] The network element 150 may include a central unit-user plane 151 (Central Unit-User Plane, CU-UP), an open distributed unit 152 (Open Distributed Unit, O-DU) and an open radio unit 153 (Open Radio Unit, O-RU).

[0078] The central unit-control plane 130 includes a real-time wireless intelligent controller 160 .

[0079] The real-time wireless intelligent controller 160 is connected to the near-real-time wireless intelligent controller 140 via a controller interaction interface.

[0080] In a possible embodiment, the real-time wireless intelligent controller 160 can be configured to receive a business scenario request sent by a network element; the business scenario request is used to request the artificial intelligence model that the network element needs to use in the business scenario; forward the business scenario request to the near-real-time wireless intelligent controller so that the near-real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller; receive the artificial intelligence model sent by the near-real-time wireless intelligent controller, and send the artificial intelligence model to the network element.

[0081] The near real-time wireless intelligent controller 140 can be configured to receive a business scenario request sent by a real-time wireless intelligent controller; the business scenario request is used to request the artificial intelligence model that the network element needs to use in the business scenario; search for the artificial intelligence model and send it to the real-time wireless intelligent controller.

[0082] The service scenario request may include a beamforming request and / or a physical layer coding request, etc.

[0083] When the open distributed unit 152 sends a business scenario request to the real-time wireless intelligent controller 160, the way in which it receives the returned artificial intelligence may include the following possible communication methods.

[0084] In a possible embodiment, the real-time wireless intelligent controller 160 may send the artificial intelligence model through a conventional communication method.

[0085] In another possible embodiment, the real-time wireless intelligent controller 160 and the open distributed unit 152 can be connected via a model interaction interface. The artificial intelligence model can be directly delivered via the model interaction interface.

[0086] Furthermore, when the open radio unit 153 sends a service scenario request to the real-time wireless intelligent controller 160, the manner in which the real-time wireless intelligent controller 160 receives the returned artificial intelligence may include the following possible communication modes.

[0087] There is no direct interface between the open radio unit 153 and the real-time wireless intelligent controller 160. It is connected to the open distributed unit 152 through the front-end transmission (Fronthaul) interface, and then forwards the request to the central unit-control plane 130 using the F (Fronthaul) 1-c (Control) interface to interact with the real-time wireless intelligent controller 160.

[0088] In a possible embodiment, the returned artificial intelligence model can be received through the above-mentioned communication method.

[0089] In another possible embodiment, the artificial intelligence model can be sent to the open distributed unit 152 through the model interaction interface, and forwarded to the open radio unit 153 by the open distributed unit 152.

[0090] Furthermore, the central unit-user plane 151 is connected to the open distributed unit 152 via an F1-u (User) interface. The central unit-control plane 130 is connected to the central unit-user plane 151 via an E1 interface.

[0091] The central unit-user plane 151, the central unit-control plane 130 and the near real-time wireless intelligent controller 140 are also in communication connection, and communicate through a conventional communication interface connection.

[0092] In a possible implementation, FIG2 shows a schematic structural diagram of another wireless access network; as shown in FIG2 , each module includes specific functional definitions.

[0093] In a possible embodiment, in the programmable intelligent layer 110 , the central unit-control plane 130 includes a protocol function 131 , an AI scenario management unit 132 , and a real-time wireless intelligent controller 160 .

[0094] The near real-time wireless intelligent controller 140 may specifically include an interface management unit 141 , an AI management unit 142 , a near real-time intelligent network function 143 , a network function management 144 , a near real-time model library 145 and a database 146 .

[0095] The real-time wireless intelligent controller 160 may include an interface management unit 161 , a wireless network status collection unit 162 , an AI / machine learning (ML) flow management unit 163 , a real-time scenario AI model agent unit 164 , a real-time model library 165 and a local database 166 .

[0096] In a possible embodiment, in the programmable RAN plane 120 , the central unit-user plane 151 may specifically include an AI scenario management unit 1511 and a protocol function 1512 .

[0097] The open distributed unit 152 may specifically include an AI scenario management unit 1521 and a protocol function 1522 .

[0098] The open radio unit 153 may specifically include a radio frequency (RF) unit 1531 .

[0099] In a possible embodiment, the real-time model library 165 in the real-time wireless intelligent controller 160 can be configured to store models that have been trained in the near real-time intelligent controller in the corresponding real-time scenario.

[0100] The local database 166 in the real-time wireless intelligent controller 160 can collect and receive network data in various real-time business scenarios, upload it to the near-real-time wireless intelligent controller 140, and send data processing instructions to the near-real-time wireless intelligent controller 140 so that the near-real-time wireless intelligent controller 140 can classify the network data according to the data processing instructions; and complete the training process of various intelligent models based on the network data.

[0101] The AI / ML flow management unit 163 in the real-time wireless intelligent controller 160 can be mainly configured to complete the processing, splitting and decision-making of AI model requests in real-time scenarios, decide which AI model to select based on different business scenario requests, and return the AI ​​model to the network element, deploy it on a specific network element side, complete the AI ​​reasoning function, and solve the business needs in the network element.

[0102] Furthermore, the AI ​​management unit in the near real-time wireless intelligent controller 140 can serve as a near real-time control unit in the near real-time wireless intelligent controller 140, configured to complete the management of relevant AI models in the near real-time wireless intelligent controller, including training, reasoning, and pushing of AI models.

[0103] The near real-time model library 145 in the near real-time wireless intelligent controller 140 stores AI models required by the near real-time and real-time wireless intelligent controllers within a certain coverage area.

[0104] The protocol function refers to the protocol used in the communication process, which defines the rules and methods for exchanging information between communication devices.

[0105] The architecture of the wireless access network provided in the embodiments of the present disclosure is a possible architecture for the future 6G wireless access network. Based on the real-time intelligence required for the intelligent connection of all things in 6G, the real-time wireless intelligent controller 160 is deployed in a new architecture within the future 6GRAN. The real-time wireless intelligent controller 160 can also be flexibly deployed in the corresponding wireless access network architecture based on the development of future technologies. Based on the design and development of endogenous intelligence in 6G, the real-time wireless intelligent controller 160 is added to the wireless access network to realize the processing of low-latency business scenario requests.

[0106] This exemplary implementation is described in detail below with reference to the accompanying drawings and examples.

[0107] First, an embodiment of the present disclosure provides a model management method based on real-time intelligence. This method can be executed by any electronic device with computing and processing capabilities, or by a real-time wireless intelligent controller in a wireless access network. The following process takes execution by a real-time wireless intelligent controller in a wireless access network as an example.

[0108] FIG3 shows a flow chart of a model management method based on real-time intelligence in an embodiment of the present disclosure. As shown in FIG3 , the model management method based on real-time intelligence provided in an embodiment of the present disclosure includes the following steps:

[0109] S302: Receive a business scenario request sent by a network element, where the business scenario request is used to request an artificial intelligence model that the network element needs to use in the business scenario.

[0110] In a possible embodiment, the service scenario request sent by the network element in the present disclosure may be a request corresponding to scenarios such as related beamforming and physical layer encoding and decoding in an open distributed unit.

[0111] Artificial intelligence models can include multiple models, for example, models that can be used to process requests such as beamforming and physical layer encoding and decoding, or models for other low-latency business scenario requests, such as those below 100ms.

[0112] In one possible embodiment, the business scenario request includes a requested artificial intelligence model, and / or input parameters, output parameters, and model type that need to be input into the artificial intelligence model.

[0113] For example, the network element can analyze the business that needs to be processed, determine the artificial intelligence model that needs to be used, and then send it to the real-time wireless intelligent controller through a business scenario request to directly request the artificial intelligence model.

[0114] For example, the network element can upload all the received parameters and network data to the real-time wireless intelligent controller. The AI / ML flow management unit in the real-time wireless intelligent controller makes decisions based on the parameters and model types to determine the artificial intelligence model that needs to be sent to the network element.

[0115] S304: Forward the business scenario request to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller.

[0116] In a possible embodiment, the service scenario request may be directly forwarded to the near real-time wireless intelligent controller for processing by the near real-time wireless intelligent controller.

[0117] Due to storage space limitations, the real-time wireless intelligent controller cannot store a large number of AI models to provide services for real-time services. Therefore, the real-time wireless intelligent controller can be deployed in the programmable intelligent layer of the wireless access network. Through new interfaces, such as the controller interaction interface, the real-time wireless intelligent controller and the near-real-time wireless intelligent controller are connected. The real-time wireless intelligent controller mainly deploys communication resources and performs fast forwarding and other processing for requests with low latency requirements.

[0118] In another possible embodiment, AI models corresponding to some commonly used business scenario requests can be deployed in the real-time wireless intelligent controller and stored in the real-time model library. This approach can handle most business scenario requests and improve the processing efficiency of business scenario requests. The AI / ML flow management unit determines whether the AI ​​model for the business scenario request exists in the real-time model library.

[0119] For example, if the artificial intelligence model requested by the business scenario is not available in the real-time model library, the business scenario request is forwarded to the near real-time wireless intelligent controller.

[0120] For example, if it is detected whether there is an artificial intelligence model in the real-time wireless intelligent controller, if it is determined that there is an artificial intelligence model, the artificial intelligence model is filtered in the real-time model library according to the business scenario request, and the artificial intelligence model is sent to the network element.

[0121] Through the above method, some AI models are deployed in the real-time model library of the real-time wireless intelligent controller, and the AI ​​models required for some business scenario requests corresponding to low latency requirements are deployed in the near-real-time wireless intelligent controller. The specific deployment method can be to sort the number of AI models sent to the network element within a preset time period, and according to the remaining storage resources, select the AI ​​models with the highest number of AI models sent to the network element. The specific number of AI models selected can be determined according to the storage resources. The more storage resources, the more AI models can be selected.

[0122] In another possible embodiment, the AI / ML stream management unit determines that if the artificial intelligence model requested by the business scenario is not in the real-time model library, the business scenario request can be directly forwarded to the near real-time wireless intelligent controller, or it can be further determined whether the computing power resources of the real-time wireless intelligent controller are sufficient.

[0123] For example, if computing power resources are sufficient, the artificial intelligence model is trained and evaluated in real time until a trained artificial intelligence model is obtained, and the trained artificial intelligence model is sent to the network element.

[0124] For example, if computing resources are insufficient, the service scenario request is forwarded to the near real-time wireless intelligent controller.

[0125] S306: Receive the artificial intelligence model sent by the near real-time wireless intelligent controller, and send the artificial intelligence model to the network element.

[0126] In one possible embodiment, after receiving an AI model from a near-real-time wireless intelligent controller, the AI ​​model is cached in a real-time model library before being forwarded to the network element. This allows the near-real-time wireless intelligent controller to collaborate with the next service request, further improving service processing efficiency.

[0127] The real-time model library informs the AI / ML stream management unit that the corresponding AI model has been added, and requests the AI / ML stream management unit to complete corresponding operations such as model recording and registration. At the same time, the AI / ML stream management unit dispatches the required model to the corresponding network element according to a series of processes to realize AI model inference and other functions.

[0128] In another possible embodiment, after receiving the AI ​​model sent by the near-real-time wireless intelligent controller, it can be directly forwarded to the network element. Due to insufficient storage resources, the AI ​​model stored in the real-time wireless intelligent controller is already commonly used. The AI ​​model corresponding to the current request is not commonly used, so it can be selected to not occupy storage resources, saving storage resources.

[0129] Through the methods provided by the embodiments of the present disclosure and the near-real-time wireless network intelligent controller, it is easy to realize the linkage of AI model requirements in real-time and near-real-time scenarios, improve support for AI demand scenarios, and provide real-time business services.

[0130] In a possible implementation, FIG4 shows another model management method based on real-time intelligence, as shown in FIG4 , including the following steps:

[0131] S402: Receive a business scenario request sent by a network element, where the business scenario request is used to request an artificial intelligence model that the network element needs to use in the business scenario.

[0132] S404: Determine whether an artificial intelligence model exists in the real-time wireless intelligent controller; if so, execute S406; if not, execute S408.

[0133] S406: Return the artificial intelligence model to the network element.

[0134] S408: Determine whether the computing power resources of the real-time wireless intelligent controller are sufficient; if sufficient, execute S410; if insufficient, execute S414.

[0135] S410: Train and evaluate the artificial intelligence model in real time until a trained artificial intelligence model is obtained.

[0136] S412: Send the trained artificial intelligence model to the network element.

[0137] S414: Forward the service scenario request to the near real-time wireless intelligent controller.

[0138] S416: Receive the artificial intelligence model sent by the near real-time wireless intelligent controller, and send the artificial intelligence model to the network element.

[0139] In one possible implementation, FIG5 shows a flow chart of a real-time intelligent model management method provided by an embodiment of the present disclosure, which is applied to a near real-time wireless intelligent controller, including the following steps:

[0140] S502: Receive a business scenario request sent by a real-time wireless intelligent controller, where the business scenario request is used to request an artificial intelligence model that a network element needs to use in a business scenario.

[0141] S504: Searching for an artificial intelligence model and sending it to a real-time wireless intelligent controller.

[0142] In one possible embodiment, an artificial intelligence model is detected according to a business scenario request, the artificial intelligence model is searched in a near real-time model library, and sent to a real-time wireless intelligent controller.

[0143] For example, after receiving a business scenario request, the AI ​​management unit deployed in the near real-time wireless intelligent controller can match the AI ​​model required by the business scenario request and obtain matching information of the AI ​​model.

[0144] The AI ​​management unit in the near-real-time wireless intelligent controller can send AI model matching information to the near-real-time model library, which then filters the library to find the AI ​​model corresponding to the business scenario request. The near-real-time model library matches the corresponding AI model and transmits the AI ​​model for the real-time scenario to the real-time wireless intelligent controller via the corresponding communication interface, namely the controller interaction interface.

[0145] In one possible embodiment, if no AI model is found that matches the business scenario request, relevant network data can be retrieved from the database for training. Alternatively, a message indicating that no AI model was found is returned to the real-time wireless intelligent controller, requesting the controller to provide relevant network data for real-time training and evaluation of the AI ​​model before it is issued.

[0146] In another possible embodiment, since the storage resources and computing power resources in the real-time wireless intelligent controller are limited by other conditions, the training and evaluation processes of some commonly used AI models can be completed by the near-real-time wireless intelligent controller.

[0147] Exemplarily, the near real-time wireless intelligent controller can receive network data and data processing instructions sent by the real-time wireless intelligent controller, classify the network data according to the data processing instructions, and complete the training process of multiple intelligent models based on the network data.

[0148] In one possible embodiment, the AI ​​model deployed in the real-time wireless intelligent controller is also trained by the near-real-time wireless intelligent controller. After the training and evaluation are completed, the AI ​​model deployed in the real-time wireless intelligent controller is updated.

[0149] Furthermore, FIG6 shows an interactive flow chart of a model management method based on real-time intelligence, as shown in FIG6 , including the following steps:

[0150] S602: The network element sends a service scenario request to the real-time wireless intelligent controller.

[0151] S604: The AI / ML flow management unit in the real-time wireless intelligent controller receives a service scenario request sent by the network element. The service scenario request is used to request the network element to use an artificial intelligence model in the service scenario.

[0152] Among them, the business scenario request includes the requested artificial intelligence model, and / or the input parameters, output parameters and model type that need to be input into the artificial intelligence model.

[0153] S606: The AI / ML flow management unit detects that the real-time wireless intelligent controller does not have an artificial intelligence model.

[0154] S608: The AI / ML flow management unit determines that the computing resources of the real-time wireless intelligent controller are insufficient.

[0155] S610: The AI / ML flow management unit forwards the service scenario request to the near real-time wireless intelligent controller.

[0156] S612: The AI ​​management unit matches the AI ​​model required by the business scenario request and obtains matching information of the AI ​​model.

[0157] S614: The AI ​​management unit filters the AI ​​model corresponding to the business scenario request in the near real-time model library according to the matching information of the AI ​​model.

[0158] S616: The near real-time model library returns the AI ​​model to the real-time wireless intelligent controller.

[0159] S618: The real-time wireless intelligent controller caches the received AI model in the real-time model library.

[0160] S620: The real-time model library notifies the AI / ML stream management unit that the AI ​​model has been cached.

[0161] S622: The AI / ML flow management unit sends the AI ​​model to the network element.

[0162] Based on the same inventive concept, the present disclosure also provides a real-time intelligent model management device, as shown in the following embodiment. Since the principle of solving the problem in this device embodiment is similar to that in the above method embodiment, the implementation of this device embodiment can refer to the implementation of the above method embodiment, and the repeated parts will not be repeated.

[0163] Figure 7 shows a structural diagram of a real-time intelligence-based model management device in an embodiment of the present disclosure. As shown in Figure 7, the real-time intelligence-based model management device 70 is applied to a real-time wireless intelligent controller, including: a first receiving unit 701, configured to receive a business scenario request sent by a network element; the business scenario request is used to request the network element to use an artificial intelligence model in the business scenario; a sending unit 702, configured to forward the business scenario request to the near-real-time wireless intelligent controller, so that the near-real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller; a first transmission unit 703, configured to receive the artificial intelligence model sent by the near-real-time wireless intelligent controller, and send the artificial intelligence model to the network element.

[0164] In the embodiments of the present disclosure, the examples and application scenarios implemented by the above units and corresponding steps are the same, but are not limited to the contents disclosed in the above method embodiments. It should be noted that the above units as part of the device can be executed in a computer system such as a set of computer executable instructions.

[0165] Figure 8 shows a structural diagram of a real-time intelligence-based model management device in an embodiment of the present disclosure. As shown in Figure 8, the real-time intelligence-based model management device 80 is applied to a near-real-time wireless intelligent controller, including: a second receiving unit 801, configured to receive a business scenario request sent by the real-time wireless intelligent controller; the business scenario request is used to request the network element to use an artificial intelligence model in the business scenario; a second transmission unit 802, configured to search for the artificial intelligence model and send it to the real-time wireless intelligent controller.

[0166] Those skilled in the art will appreciate that various aspects of the present disclosure may be implemented as systems, methods, or program products. Therefore, various aspects of the present disclosure may be implemented in the following forms: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which may be collectively referred to herein as "circuits," "modules," or "systems."

[0167] The electronic device 900 according to this embodiment of the present disclosure is described below with reference to Figure 9. The electronic device 900 shown in Figure 9 is merely an example and should not limit the functions and scope of use of the embodiments of the present disclosure.

[0168] As shown in Figure 9, electronic device 900 is implemented as a general-purpose computing device. Components of electronic device 900 may include, but are not limited to, the aforementioned at least one processing unit 910, the aforementioned at least one storage unit 920, and a bus 930 connecting various system components (including storage unit 920 and processing unit 910).

[0169] The storage unit stores program code, which can be executed by the processing unit 910, so that the processing unit 910 performs the steps of various exemplary embodiments of the present disclosure described in the "Exemplary Methods" section above. For example, the processing unit 910 can perform the steps of any of the above method embodiments.

[0170] The storage unit 920 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 9201 and / or a cache memory unit 9202 , and may further include a read-only memory unit (ROM) 9203 .

[0171] The storage unit 920 may also include a program / utility 9204 having a set (at least one) of program modules 9205, such program modules 9205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0172] Bus 930 may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0173] The electronic device 900 can also communicate with one or more external devices 940 (e.g., a keyboard, a pointing device, a Bluetooth device, etc.), one or more devices that enable a user to interact with the electronic device 900, and / or any device that enables the electronic device 900 to communicate with one or more other computing devices (e.g., a router, a modem, etc.). Such communication can occur via an input / output (I / O) interface 950. Furthermore, the electronic device 900 can communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet) via a network adapter 960. As shown, the network adapter 960 communicates with other modules of the electronic device 900 via a bus 930. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device 900, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0174] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0175] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods of the above embodiments.

[0176] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided. The computer-readable storage medium may be a readable signal medium or a readable storage medium. A program product capable of implementing the above-mentioned method of the present disclosure is stored thereon. In some possible implementations, various aspects of the present disclosure may also be implemented in the form of a program product, which includes program code. When the program product is run on a terminal device, the program code is used to cause the terminal device to execute the steps according to various exemplary embodiments of the present disclosure described in the "Exemplary Methods" section above of this specification.

[0177] More specific examples of computer-readable storage media in the present disclosure may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fibers, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0178] In the present disclosure, a computer-readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, which carries readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium that can transmit, propagate, or transfer a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0179] Alternatively, the program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0180] In a specific implementation, the program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and the like, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a standalone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0181] It should be noted that although several modules or units of the device for action execution are mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be concretized in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0182] Furthermore, although the steps of the method of the present disclosure are described in a particular order in the accompanying drawings, this does not require or imply that the steps must be performed in this particular order, or that all steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0183] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0184] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the appended claims.

Claims

1. A model management method based on real-time intelligence, wherein: Applied to a real-time wireless intelligent controller; the method comprises: Receive a business scenario request sent by a network element; the business scenario request is used to request an artificial intelligence model that the network element needs to use in the business scenario; Forwarding the service scenario request to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller; Receive the artificial intelligence model sent by the near real-time wireless intelligent controller, and send the artificial intelligence model to the network element.

2. The method according to claim 1, wherein: The forwarding the service scenario request to the near real-time wireless intelligent controller includes: Detecting whether the artificial intelligence model exists in the real-time wireless intelligent controller; If the real-time wireless intelligent controller does not have the artificial intelligence model, determining whether the computing resources of the real-time wireless intelligent controller are sufficient; If the computing resources are insufficient, the business scenario request is forwarded to the near real-time wireless intelligent controller.

3. The method according to claim 1, wherein: The business scenario request includes the requested artificial intelligence model, and / or the input parameters, output parameters and model type that need to be input into the artificial intelligence model.

4. The method according to claim 1, wherein: After receiving the service scenario request sent by the network element, the method further includes: Detecting whether the artificial intelligence model exists in the real-time wireless intelligent controller; If it is determined that the artificial intelligence model is available, requesting to filter the artificial intelligence model according to the business scenario; The artificial intelligence model is sent to the network element.

5. The method according to claim 1, wherein: After receiving the service scenario request sent by the network element, the method further includes: Detecting whether the artificial intelligence model exists in the real-time wireless intelligent controller; If it is determined that the artificial intelligence model is not available, determining whether the computing resources of the real-time wireless intelligent controller are sufficient; If the computing resources are sufficient, the artificial intelligence model is trained and evaluated in real time until a trained artificial intelligence model is obtained; The trained artificial intelligence model is sent to the network element.

6. The method according to claim 1, wherein: After receiving the artificial intelligence model sent by the near real-time wireless intelligent controller, the method further includes: In the real-time model library, the artificial intelligence model is cached.

7. The method according to claim 1, wherein: The method further comprises: Receive network data in various real-time business scenarios; sending the network data and the data processing instruction to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller classifies the network data according to the data processing instruction; and The training process of various intelligent models is completed according to the network data.

8. A model management method based on real-time intelligence, wherein: Applied to a near real-time wireless intelligent controller, the method comprises: Receiving a service scenario request sent by a real-time wireless intelligent controller; the service scenario request is used to request an artificial intelligence model that a network element needs to use in a service scenario; The artificial intelligence model is searched and sent to the real-time wireless intelligent controller.

9. The method according to claim 8, wherein: The step of searching for the artificial intelligence model and sending it to the real-time wireless intelligent controller includes: Requesting detection of the artificial intelligence model according to the business scenario; The artificial intelligence model is searched in the near real-time model library and sent to the real-time wireless intelligent controller.

10. The method according to claim 8, wherein: The method further comprises: Receiving network data and data processing instructions sent by the real-time wireless intelligent controller; According to the data processing instructions, the network data is classified, and the training process of multiple intelligent models is completed according to the network data.

11. A model management device based on real-time intelligence, wherein: Applied to real-time wireless intelligent controllers; including: A first receiving unit is configured to receive a business scenario request sent by a network element; the business scenario request is used to request an artificial intelligence model that the network element needs to use in the business scenario; A sending unit, configured to forward the service scenario request to the near real-time wireless intelligent controller, so that the near real-time wireless intelligent controller searches for the artificial intelligence model and sends it to the real-time wireless intelligent controller; The first transmission unit is configured to receive the artificial intelligence model sent by the near real-time wireless intelligent controller and send the artificial intelligence model to the network element.

12. A model management device based on real-time intelligence, wherein: Applied to near real-time wireless intelligent controllers; including: A second receiving unit is configured to receive a service scenario request sent by the real-time wireless intelligent controller; the service scenario request is used to request an artificial intelligence model that a network element needs to use in a service scenario; The second transmission unit is configured to search for the artificial intelligence model and send it to the real-time wireless intelligent controller.

13. A wireless access network, wherein: include: Programmable intelligent layer and programmable radio access network RAN ​​plane; The programmable RAN plane includes a central unit-control plane and network elements; The programmable intelligent layer includes a central unit-control plane and a near real-time wireless intelligent controller; the central unit-control plane includes a real-time wireless intelligent controller; the real-time wireless intelligent controller is connected to the near real-time wireless intelligent controller via a controller interaction interface; The real-time wireless intelligent controller is configured to execute the method described in any one of claims 1 to 7; the near real-time wireless intelligent controller is configured to execute the method described in any one of claims 8 to 10.

14. An electronic device, wherein: include: processor; as well as a memory configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 and 8 to 10 by executing the executable instructions.

15. A computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 and 8 to 10 is implemented.

16. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 7 and 8 to 10 is implemented.

Images