Communication method and apparatus

Abstract

The present disclosure provides a communication method and apparatus, which are applied to the technical field of communications. The method comprises: acquiring a plurality of pieces of data, the plurality of pieces of data comprising first data, second data, and network feedback data, the first data being data input into a first model, the second data being data output by the first model based on the first data, and the network feedback data being network quality-related data obtained by evaluating the second data at a network node; and training the first model according to the first data, the second data, and the network feedback data. The described solution improves communication system performance by means of training a first model.

Description

Communication method and apparatus

[0001] The present application claims priority to the Chinese patent application No. 202411605509.9, filed on November 11, 2024, with the State Intellectual Property Office of China, and the Chinese patent application No. 202411605509.9 has the invention name of “A communication method and apparatus”, the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the field of communication technology, in particular to a communication method and apparatus. BACKGROUND

[0003] With the continuous development of wireless communication technology, future communication systems can be integrated with artificial intelligence (AI) technology, for example, the integration of future mobile communication systems and AI models. There are a large number of agents based on AI models in future communication networks, and these agents can achieve future network services through data interaction. Since the AI model is a pre-trained model, it needs to be optimized according to the actual service effect of its inference decision when applied in the network, so as to adjust the generation strategy of the model. However, when the AI model is used in the network, it is difficult to predict the pros and cons of the strategy generated by the AI model. The strategy can only be judged by the actual running results of the service after execution, and in the process of network operation, it is impossible to intervene in real time to judge the pros and cons of the decision of the AI model. These situations may result in poor performance of the communication system. SUMMARY

[0004] Embodiments of the present application provide a communication method and apparatus, which can improve the performance of the communication system by training the first model.

[0005] In a first aspect, embodiments of the present application provide a communication method, which can be executed by a first network element, or a module (such as a processor, a chip, or a chip system, etc.) applied to the first network element, or a logic node, a logic module, or software that can realize all or part of the functions of the first network element. The method comprises:

[0006] Obtaining a plurality of data, the plurality of data comprising first data, second data, and network feedback data, the first data being data input to a first model, the second data being data output by the first model after the first data is input to the first model, and the network feedback data being network quality related data obtained by evaluating the second data by a network node.

[0007] Training the first model according to the first data, the second data, and the network feedback data.

[0008] By obtaining the plurality of data, the first network element can train the first model according to the first data, the second data, and the network feedback data in the plurality of data, which is beneficial to optimize the decision of the agent by using the network feedback, thereby improving the performance of the communication system.

[0009] In a possible design, type information of each data in the plurality of data is obtained, the type information including input data, output data, or feedback data; and the plurality of data is classified according to the type information. By obtaining the type information, the first network element classifies the plurality of data to obtain the first data, the second data, and the network feedback data, which is beneficial to standardize the data format of the network feedback in the subsequent process, thereby optimizing the decision of the agent by using the network feedback and improving the performance of the communication system.

[0010] In another possible design, identification information corresponding to the first data, identification information corresponding to the second data, and identification information corresponding to the network feedback data are obtained; and the first data, the second data, and the network feedback data are processed according to the identification information. By obtaining the identification information, the first network element can process the first data, the second data, and the network feedback data according to the identification information, thereby standardizing the data format of the network feedback, which is beneficial to optimize the decision of the agent by using the network feedback in the subsequent process and improve the performance of the communication system.

[0011] In another possible design, the identification information includes a task identifier; and the first data, the second data, and the network feedback data corresponding to a same task are determined according to the task identifier. By processing the first data, the second data, and the network feedback data according to the task identifier, the data format of the network feedback can be standardized, which is beneficial to optimize the decision of the agent by using the network feedback in the subsequent process and improve the performance of the communication system.

[0012] In another possible design, the identification information includes a task identifier and a time sequence; the first data, the second data, and the network feedback data corresponding to a same task are determined according to the task identifier; and the first data, the second data, and the network feedback data corresponding to the same task are numbered according to the time sequence. By aligning the first data, the second data, and the network feedback data according to the task identifier and the time sequence, the order information of the collaborative decision of the plurality of agents can be carried when the data format of the network feedback is standardized, which is beneficial to improve the effectiveness of optimizing the decision of the agent by using the network feedback in the subsequent process.

[0013] In another possible design, the first data corresponding to the same task includes first input data and second input data, the second data corresponding to the same task includes first output data and second output data, the first input data corresponds to the first output data, the second input data corresponds to the second output data, a payload of the first output data is the same as a payload of the second input data, and a time sequence of the first output data is before a time sequence of the second input data; and a number corresponding to the first output data is added by a first value according to the order of the time sequence of the first output data and the time sequence of the second input data, to obtain a number corresponding to the second input data. By numbering the first data, the second data, and the network feedback data corresponding to the same task, sequence information of collaborative decision-making of multiple agents can be carried when implementing data format standardization of network feedback, which is beneficial to improving effectiveness of subsequent optimization of decision-making of the agent by using the network feedback.

[0014] In another possible design, the first data corresponding to the same task includes first input data, second input data, and third input data, the second data corresponding to the same task includes first output data, second output data, and third output data, the first input data corresponds to the first output data, the second input data corresponds to the second output data, the third input data corresponds to the third output data, a payload of the first output data is the same as a payload of the second input data and a payload of the third input data respectively, the time sequence of the first output data is before a time sequence of the second input data, and the time sequence of the second input data is the same as a time sequence of the third input data; the number corresponding to the first output data is added by a first value according to the order of the time sequence of the first output data and the time sequence of the second input data, to obtain a number corresponding to the second input data; and the number corresponding to the third input data is determined to be equal to the number corresponding to the second input data according to the order of the time sequence of the second input data and the time sequence of the third input data. By numbering the first data, the second data, and the network feedback data corresponding to the same task, sequence information of collaborative decision-making of multiple agents can be carried when implementing data format standardization of network feedback, which is beneficial to improving effectiveness of subsequent optimization of decision-making of the agent by using the network feedback.

[0015] In another possible design, the first data corresponding to the same task includes first input data, second input data, and third input data, the second data corresponding to the same task includes first output data and second output data and third output data, the first input data corresponds to the first output data, the second input data corresponds to the second output data, the third input data corresponds to the third output data, a payload of the third input data includes a payload of the first output data and a payload of the second output data, and a time sequence of the third input data is after time sequences of the first output data and the second output data; a number corresponding to the third input data is obtained by adding a first value to a number corresponding to the first output data according to an order of the time sequences of the first output data, the second output data, and the third input data, and the time sequence of the first output data is after the time sequence of the second output data; or a number corresponding to the third input data is obtained by adding a first value to a number corresponding to the second output data according to an order of the time sequences of the first output data, the second output data, and the third input data, and the time sequence of the second output data is after the time sequence of the first output data. By numbering the first data, the second data, and the network feedback data corresponding to the same task, the order information of the collaborative decision of the plurality of agents can be carried when the data format of the network feedback is standardized, and the effectiveness of the optimization of the decision of the agent by using the network feedback can be improved.

[0016] In another possible design, the first data includes at least one of the following: user demand information, retrieval augmentation generation (RAG) information, or prompt information. This facilitates the standardization of the data format of the network feedback, and thus the optimization of the decision of the agent by using the network feedback and the improvement of the performance of the communication system.

[0017] In another possible design, the second data includes at least one of the following: a task decomposition result, tool invocation information, or configuration information of the network node. This facilitates the standardization of the data format of the network feedback, and thus the optimization of the decision of the agent by using the network feedback and the improvement of the performance of the communication system.

[0018] In another possible design, the network feedback data includes at least one of the following: resource utilization, network throughput, task completion accuracy, or average task waiting delay. This facilitates the standardization of the data format of the network feedback, and thus the optimization of the decision of the agent by using the network feedback and the improvement of the performance of the communication system.

[0019] Secondly, embodiments of this application provide a communication device configured to implement the methods and functions described in the first aspect. The communication device is implemented in hardware / software. It includes modules corresponding to the aforementioned functions. The communication device can be a first network element, a chip, chip system, or processor supporting the first network element in implementing the aforementioned methods, or a logical node, logical module, or software capable of implementing all or part of the functions of the first network element.

[0020] Thirdly, embodiments of this application provide a communication device including one or more processors. The one or more processors enable the communication device to implement the methods in any possible design or implementation of the first aspect described above.

[0021] Fourthly, embodiments of this application provide a communication system including at least one first network element, the first network element being used to perform the steps in the first aspect described above.

[0022] Fifthly, embodiments of this application provide a computer-readable storage medium for storing a computer program that, when executed, causes the method described in any one of the first aspects to be implemented.

[0023] In a sixth aspect, embodiments of this application provide a computer program product containing instructions that, when executed, cause the method described in any one of the first aspects to be implemented.

[0024] In a seventh aspect, embodiments of this application provide a chip including a processor and a communication interface for communicating with external or internal devices, the processor enabling the chip to implement the methods described in the above aspects.

[0025] In one possible design, the chip may further include a memory storing computer programs or instructions, which the processor executes, either from the stored computer programs or instructions or derived from other programs or instructions. When the computer program or instructions are executed, the processor causes the chip to implement the methods described above.

[0026] In another possible design, the chip can be integrated into a communication system or communication device. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of this application or the background art, the accompanying drawings used in the embodiments of this application or the background art will be described below.

[0028] Figure 1 is a schematic diagram of a possible application framework in a communication system;

[0029] Figure 2 is a schematic diagram of another possible application framework in a communication system;

[0030] Figure 3a is a schematic diagram of a communication system applicable to the communication method of this application embodiment;

[0031] Figure 3b is a schematic diagram of another communication system applicable to the communication method of the present application embodiment;

[0032] Figure 4 is a schematic diagram of an AI application framework;

[0033] Figure 5 is a flowchart illustrating a communication method provided in an embodiment of this application;

[0034] Figure 6 is a schematic diagram of a numbering method provided in an embodiment of this application;

[0035] Figure 7 is a schematic diagram of another numbering method provided in an embodiment of this application;

[0036] Figure 8 is a schematic diagram of another numbering method provided in an embodiment of this application;

[0037] Figure 9 is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0038] Figure 10 is a schematic diagram of the structure of a communication device provided in an embodiment of this application. Detailed Implementation

[0039] The following explanations of some of the terms used in this application are provided to facilitate understanding by those skilled in the art.

[0040] (1) Large model: refers to a deep learning model with a very large number of parameters, usually used to process large-scale datasets and capable of learning complex features of the data. In this application, the large model belongs to the category of AI models, which can be simply referred to as a model. The solution in this application is not limited to large models, but is also applicable to models with fewer or more parameters.

[0041] (2) Agent: An entity that can make autonomous behavioral decisions, take actions to interact with the environment and continuously improve. In this application, it specifically refers to an agent implemented based on an AI model.

[0042] The embodiments of this application are described below with reference to the accompanying drawings.

[0043] It should be understood that in the description of this application, "at least one" means one or more, and "multiple" means two or more. In addition, the words "first," "second," etc., unless otherwise stated, are used only for the purpose of distinguishing descriptions and should not be construed as indicating or implying relative importance or order.

[0044] It should be understood that in the description of this application, the indication includes direct indication (also known as explicit indication) and implicit indication. Direct indication information A refers to information A being included; implicit indication information A refers to information A being indicated through the correspondence between information A and information B, and the direct indication information B. The correspondence between information A and information B can be predefined, pre-stored, pre-burned, or pre-configured.

[0045] It should be understood that, in the description of this application, information C is used to determine information D, including both situations where information D is determined solely based on information C and situations where it is determined based on information C and other information. Furthermore, information C can also be used to determine information D indirectly, for example, where information D is determined based on information E, and information E is determined based on information C.

[0046] Furthermore, in this application, "network element A sends message A to network element B" can be understood as network element B being the destination of message A or an intermediate network element in the transmission path between the destination and network element B, which may include sending the message directly or indirectly to network element B. Similarly, "network element B receives message A from network element A" can be understood as network element A being the source of message A or an intermediate network element in the transmission path between the source and network element A, which may include receiving the message directly or indirectly from network element A. The message may undergo necessary processing between the source and destination, such as format changes, but the destination can understand a valid message from the source. Similar expressions in this application can be interpreted in a similar way and will not be elaborated further here.

[0047] The technical solutions provided in this application can be applied to various communication systems, such as 5G or new radio (NR) systems, long term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, LTE time division duplex (TDD) systems, wireless local area network (WLAN) systems, satellite communication systems, future communication systems, or integrated systems of multiple systems. The technical solutions provided in this application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), and Internet of Things (IoT) communication systems, or other communication systems.

[0048] Optionally, the technical solution provided in this application can also be applied to low-frequency scenarios below 6 GHz, as well as high-frequency scenarios above 6 GHz, terahertz, optical communication, etc., and this application does not limit it in this regard.

[0049] In a communication system, a network element can send signals to or receive signals from another network element. These signals can include information, signaling, or data. The term "network element" can also be replaced by an entity, network entity, device, communication equipment, communication module, node, communication node, etc. This application uses a network element as an example for description. For instance, a communication system can include at least one terminal device and at least one network device. The network device can send downlink signals to the terminal device, and / or the terminal device can send uplink signals to the network device.

[0050] In the embodiments of this application, the terminal device may also be referred to as user equipment (UE), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent, or user apparatus.

[0051] Terminal equipment can be a device that provides voice / data communication, such as a handheld device or vehicle-mounted device with wireless connectivity. Currently, examples of terminals include: mobile phones, tablets, laptops, PDAs, mobile internet devices (MIDs), wearable devices, VR devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, computing devices or other processing devices connected to a wireless modem, wearable devices, terminal devices in 5G networks, or terminal devices in future evolved public land mobile networks (PLMNs), etc., and the embodiments of this application are not limited to these.

[0052] By way of example and not limitation, in this embodiment, the terminal device can also be a wearable device. Wearable devices, also known as wearable smart devices, are a general term for devices that utilize wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, jewelry, and shoes. Wearable devices are portable devices that are worn directly on the body or integrated into the user's clothing or accessories. Wearable devices are not merely hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are feature-rich, large in size, and can achieve complete or partial functions without relying on a smartphone, such as smartwatches or smart glasses, as well as those that focus on only one type of application function and require the use of other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0053] In this embodiment, the device used to implement the functions of the terminal device can be the terminal device itself, or a device capable of supporting the terminal device in implementing those functions, such as a chip, chip system, or processor. It can also be a logic node, logic module, or software capable of implementing all or part of the terminal device's functions. This device can be installed in the terminal device or used in conjunction with the terminal device. In this embodiment, the chip system can be composed of chips or include chips and other discrete devices. This embodiment only uses the terminal device as an example to illustrate the device used to implement the functions of the terminal device, and does not constitute a limitation on the solution of this embodiment.

[0054] The network side in this application embodiment may include network devices, wherein the network devices include devices for communicating with terminal devices, and the network devices include access network devices or radio access network devices, such as base stations; or, OAM devices or CN devices for operation, management and maintenance. In this application embodiment, the access network device may refer to a radio access network (RAN) node (or device) that connects the terminal device to the wireless network. A base station can broadly encompass, or be replaced by, various names including: NodeB, evolved NodeB (eNB), next-generation NodeB (gNB) for 5G systems, relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), master station, auxiliary station, motor slide retainer (MSR) node, home base station, network controller, access node, radio node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), radio unit (RU), positioning node, etc. A base station can be a macro base station, micro base station, relay node, donor node, or similar entities, or combinations thereof. A base station can also refer to a communication module, modem, or chip installed within the aforementioned equipment or apparatus. A base station can also be a mobile switching center, a device that performs base station functions in D2D, V2X, and M2M communications, or a device that performs base station functions in future communication systems. A base station can support networks using the same or different access technologies. Optionally, a RAN node can also be a server, wearable device, vehicle, or in-vehicle equipment. For example, the access network equipment in V2X technology can be a roadside unit (RSU). The embodiments of this application do not limit the specific technologies or equipment forms used in the network equipment.

[0055] Base stations can be fixed or mobile. For example, a helicopter or drone can be configured to act as a mobile base station, and one or more cells can move depending on the location of the mobile base station. In other examples, a helicopter or drone can be configured as a device to communicate with another base station.

[0056] In one possible scenario, the radio access network (RAN) device can also be a module or unit that performs some of the functions of a base station; for example, it could be a CU (Control Unit) or a DU (User Unit). In another possible scenario, multiple RAN devices collaborate to assist a terminal in achieving radio access, with each RAN device performing some of the base station's functions. For example, the RAN device could be a CU, DU, CU-control plane (CP), CU-user plane (UP), or RU (Runner Unit). The CU and DU can be configured separately or included in the same network element, such as in a BBU (Base Station Unit). The RU can be included in radio frequency (RF) equipment or RF units, such as in an RRU (Radio Router Unit), AAU (Automatic Access Unit), or RRH (Radio Reception Headquarters).

[0057] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an ORAN system, CU can also be called an open central unit (O-CU), DU can also be called an open distributed unit (O-DU), CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. For ease of description, this application uses DU and RU as examples. Any of the DU and RU units in this application can be implemented through software modules, hardware modules, or a combination of software and hardware modules. The embodiments of this application do not limit the specific technology or specific device form used in the wireless access network equipment. For ease of description, the following description uses a base station as an example of a wireless access network equipment.

[0058] RAN nodes can support one or more types of fronthaul interfaces, with different fronthaul interfaces corresponding to DUs and RUs with different functions. If the fronthaul interface between the DU and RU is a common public radio interface (CPRI), the DU is configured to implement one or more baseband functions, and the RU is configured to implement one or more radio frequency functions. If the fronthaul interface between the DU and RU is another type of interface, relative to CPRI, it moves some downlink and / or uplink baseband functions—for example, for downlink, precoding, digital beamforming (BF), or one or more of inverse fast Fourier transform (IFFT) / adding a cyclic prefix (CP), from the DU to the RU; and for uplink, digital BF, or one or more of fast Fourier transform (FFT) / removing CP, from the DU to the RU. In one possible implementation, this interface can be an enhanced common public radio interface (eCPRI). Under the eCPRI architecture, the partitioning methods between DU and RU are different, corresponding to different types (category, Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, F.

[0059] Taking eCPRI Cat A as an example, for downlink transmission, layer mapping is used as the dividing line. The DU is configured to implement one or more functions preceding layer mapping (i.e., encoding, rate matching, scrambling, modulation, and layer mapping), while other functions following layer mapping (e.g., resource element (RE) mapping, digital BF, or IFFT / CP removal) are implemented in the RU. For uplink transmission, de-RE mapping is used as the dividing line. The DU is configured to implement one or more functions preceding de-mapping (i.e., decoding, rate matching de-mapping, descrambling, demodulation, inverse discrete Fourier transform (IDFT), channel equalization, and de-RE mapping), while other functions following de-mapping (e.g., digital BF or FFT / CP removal) are implemented in the RU. It is understood that descriptions of the functions of the DU and RU corresponding to various types of eCPRI can be found in the eCPRI protocol and will not be elaborated here.

[0060] In this application embodiment, the apparatus for implementing the functions of a network device can be a network device itself; it can also be an apparatus capable of supporting the network device in implementing that function, such as a chip system, hardware circuit, software module, or hardware circuit plus software module; or it can be a logical node, logical module, or software capable of implementing all or part of the functions of a network device. This apparatus can be installed in the network device or used in conjunction with the network device. In this application embodiment, only the network device is used as an example to illustrate the apparatus for implementing the functions of the network device, and this does not constitute a limitation on the solutions of this application embodiment.

[0061] Network devices and / or terminal devices can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; they can also be deployed on water; and they can also be deployed in the air on airplanes, balloons, and satellites. This application does not limit the scenario in which the network devices and terminal devices are located. Furthermore, terminal devices and network devices can be hardware devices, or software functions running on dedicated hardware, or software functions running on general-purpose hardware, such as virtualization functions instantiated on a platform (e.g., a cloud platform), or entities that include dedicated or general-purpose hardware devices and software functions. This application does not limit the specific form of the terminal devices and network devices.

[0062] In wireless communication networks, such as mobile communication networks, the services supported by the networks are becoming increasingly diverse, leading to increasingly diverse requirements. For example, networks need to support ultra-high speeds, ultra-low latency, and / or massive connectivity. This characteristic makes network planning, network configuration, and / or resource scheduling increasingly complex. Furthermore, as network functions become more powerful, such as supporting higher spectrum levels, supporting higher-order multiple-input multiple-output (MIMO) technologies, supporting beamforming, and / or supporting beam management, network energy efficiency has become a hot research topic. These new requirements, new scenarios, and new characteristics bring unprecedented challenges to network planning, operation, and efficient operation. To meet these challenges, AI technology can be introduced into wireless communication networks to achieve network intelligence.

[0063] To support AI technology in wireless networks, AI nodes may also be introduced into the network.

[0064] Optionally, the AI ​​node can be deployed in one or more of the following locations within the communication system: access network equipment, terminal equipment, or core network equipment, etc. Alternatively, the AI ​​node can be deployed independently, for example, in a location other than any of the aforementioned devices, such as in the host or cloud server of an over-the-top (OTT) system. The AI ​​node can communicate with other devices in the communication system, which can be, for example, one or more of the following: network equipment, terminal equipment, or core network elements, etc.

[0065] It is understood that this application does not limit the number of AI nodes. For example, when there are multiple AI nodes, these nodes can be divided based on function, such as different AI nodes being responsible for different functions.

[0066] It can also be understood that AI nodes can be independent devices, or they can be integrated into the same device to achieve different functions. Alternatively, they can be network elements in hardware devices, software functions running on dedicated hardware, or virtualization functions instantiated on a platform (e.g., a cloud platform). This application does not limit the specific form of the aforementioned AI nodes.

[0067] AI nodes can be AI network elements or AI modules.

[0068] Figure 1 illustrates a possible application framework in a communication system. Network elements in the communication system are connected via interfaces (e.g., NG, Xn) or air interfaces. These network element nodes, such as core network equipment, access network nodes (RAN nodes), terminals, or one or more devices in the OAM, are equipped with one or more AI modules (only one is shown in Figure 1 for clarity). The access network node can be a single RAN node or can include multiple RAN nodes, for example, including CU and DU. The CU and / or DU can also be equipped with one or more AI modules. Optionally, the CU can be further divided into CU-CP and CU-UP. One or more AI models are configured in the CU-CP and / or CU-UP.

[0069] The AI ​​module is used to implement corresponding AI functions. AI modules deployed in different network elements can be the same or different. Depending on the parameter configuration, the AI ​​module can implement different functions. The AI ​​module model can be configured based on one or more of the following parameters: structural parameters (e.g., at least one of the following: number of neural network layers, neural network width, inter-layer connections, neuron weights, neuron activation function, or bias in the activation function), input parameters (e.g., type and / or dimension of input parameters), or output parameters (e.g., type and / or dimension of output parameters). The bias in the activation function can also be referred to as the neural network bias.

[0070] An AI module can have one or more models. A model can infer an output, which includes one or more parameters. The learning, training, or inference processes of different models can be deployed on different nodes or devices, or they can be deployed on the same node or device.

[0071] Figure 2 illustrates another possible application framework in a communication system. The communication system includes a RAN intelligent controller (RIC). For example, the RIC could be the AI ​​module shown in Figure 1, used to implement AI-related functions. The RIC includes near-real-time (near-RT) RICs and non-real-time (non-RT) RICs. Non-real-time RICs primarily process non-real-time information, such as data that is not sensitive to latency, with latency in the order of seconds. Real-time RICs primarily process near-real-time information, such as data that is relatively sensitive to latency, with latency in the order of tens of milliseconds.

[0072] The near real-time RIC is used for model training and inference. For example, it can be used to train an AI model and then use that AI model for inference. The near real-time RIC can obtain network-side and / or terminal-side information from RAN nodes (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or terminals. This information can be used as training data or inference data. Optionally, the near real-time RIC can deliver inference results to RAN nodes and / or terminals. Optionally, inference results can be exchanged between CU and DU, and / or between DU and RU. For example, the near real-time RIC delivers the inference result to the DU, and the DU sends it to the RU.

[0073] The non-real-time RIC is also used for model training and inference. For example, it can be used to train an AI model and then use that model for inference. The non-real-time RIC can obtain network-side and / or terminal-side information from RAN nodes (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to the RAN nodes and / or terminals. Optionally, inference results can be exchanged between CU and DU, and / or between DU and RU. For example, the non-real-time RIC delivers the inference results to the DU, which then forwards them to the RU.

[0074] The near real-time RIC and non-real-time RIC can also be set up as separate network elements. Optionally, the near real-time RIC and non-real-time RIC can also be part of other devices. For example, the near real-time RIC can be set in the RAN node (e.g., in CU, DU), while the non-real-time RIC can be set in OAM, cloud server, core network equipment, or other network equipment.

[0075] As shown in Figure 3a, which is a schematic diagram of a communication system applicable to the communication method of this application embodiment, the communication system may include at least one network device, such as network device 301 shown in Figure 3a; the communication system may also include at least one terminal device, such as terminal device 302 and terminal device 303 shown in Figure 3a. Network device 301 and terminal devices (such as terminal devices 302 and 303) can communicate via a wireless link. The communication devices in this communication system, for example, network device 301 and terminal device 302, can communicate via multi-antenna technology.

[0076] As shown in Figure 3b, Figure 3b is a schematic diagram of another communication system applicable to the communication method of this application embodiment. Compared with the communication system shown in Figure 3a, the communication system shown in Figure 3b further includes an AI network element 304. The AI ​​network element 304 is used to perform AI-related operations, such as constructing a training dataset or training an AI model.

[0077] In one possible implementation, network device 301 can send data related to the training of the AI ​​model to AI network element 304, which then constructs a training dataset and trains the AI ​​model. For example, the data related to the training of the AI ​​model may include data reported by the terminal device. AI network element 304 can send the results of operations related to the AI ​​model to network device 301, which then forwards them to the terminal device. For example, the results of operations related to the AI ​​model may include at least one of the following: a trained AI model, model evaluation results, or test results.

[0078] It should be understood that Figure 3b is only used as an example of the AI ​​network element 304 being directly connected to the network device 301. In other scenarios, the AI ​​network element 304 can also be connected to the terminal device. Alternatively, the AI ​​network element 304 can be connected to both the network device 301 and the terminal device simultaneously. Alternatively, the AI ​​network element 304 can also be connected to the network device 301 through a third-party network element. This application embodiment does not limit the connection relationship between the AI ​​network element and other network elements.

[0079] AI element 304 can also be set as a module in network devices and / or terminal devices, for example, in network device 301 or terminal device shown in Figure 3a.

[0080] It should be noted that Figures 3a and 3b are simplified schematic diagrams for ease of understanding. For example, the communication system may also include other devices, such as wireless relay devices and / or wireless backhaul devices, which are not shown in Figures 3a and 3b. In practical applications, the communication system may include multiple network devices or multiple terminal devices. This application does not limit the number of network devices and terminal devices included in the communication system.

[0081] In existing technologies, generative pre-trained transformers (GPTs) can be optimized using reinforcement learning from human feedback (RLHF). Specifically, human evaluation of the GPT's output can be used to score the output, a reward model can be trained, and then the reward model can be used to provide feedback to the GPT's output. Reinforcement learning can then be used to adjust the GPT's generation strategy.

[0082] With the continuous development of wireless communication technology and AI technology, in order to enable more people to enjoy the convenience brought by intelligent services and further promote the realization of intelligent and inclusive future communication, future communication systems can be integrated with AI technology. For example, it can be the integration of future mobile communication systems with AI models. There are a large number of intelligent agents based on AI models in future communication networks. These intelligent agents can realize future network services through data interaction. However, when AI models are used in the network, they have the following main drawbacks: (1) The quality of the strategy can only be judged from the actual operation results of the service after execution. (2) During network operation, the quality of the AI ​​model's decision cannot be judged in real time. These drawbacks may lead to poor performance of the communication system.

[0083] To address the aforementioned shortcomings, network feedback reinforcement learning (NFRL) techniques can be used to optimize the model. Figure 4 illustrates a schematic diagram of an AI application framework. Specifically, an agent can send input data to an AI model (e.g., a large model). After receiving the input data, the AI ​​model processes it to obtain output data and then sends the output data back to the agent. Upon receiving the output data, the agent can send it to an action module, which executes the policy corresponding to the output data. After the policy is actually executed, network feedback is obtained through the network environment. This network feedback is used to optimize the AI ​​model using NFRL techniques. However, the specific data to be included in the network feedback and how to standardize the data format are currently unknown.

[0084] The technical solution provided in this application will be described in detail below with reference to more accompanying drawings.

[0085] As shown in Figure 5, Figure 5 is a flowchart illustrating a communication method provided in an embodiment of this application. This communication method includes, but is not limited to, the following steps:

[0086] S501: Acquire multiple data, including first data, second data, and network feedback data. The first data is the data input to the first model, the second data is the data output by the first model after the first data is processed, and the network feedback data is the network quality-related data obtained by evaluating the second data through network nodes.

[0087] The first model can be a pre-trained model; the first data can include at least one of the following: user demand information, retrieval-augmented generation (RAG) information, or prompt information; the second data can include at least one of the following: task decomposition results, tool call information, or network node configuration information; the network feedback data can include at least one of the following: resource utilization, network throughput, task completion accuracy, or average task latency; and the network node can include at least one of the following: terminal equipment, network equipment, or core network elements.

[0088] For example, in a smart city application scenario, if a user requests to sense the traffic flow at a certain intersection, the first data may include user demand information. The user demand information is input into the first model to obtain the second data. This second data may include the task decomposition results and / or tool call information corresponding to sensing the traffic flow at that intersection. The terminal device, network device, and / or core network elements are deployed and executed according to the task decomposition results and / or tool call information, which can yield the task completion accuracy. In other words, the network feedback data may include the task completion accuracy.

[0089] For example, in the application scenarios of FDD or TDD systems, network device A needs to send a data block to terminal device A. The first data may include a data transmission request. The data transmission request is input into the first model to obtain the second data. The second data may include the configuration information used by network device A and terminal device A to transmit the data block. Network device A and terminal device A configure and transmit the data block based on the configuration information. According to the actual transmission situation, the resource utilization rate corresponding to this data transmission can be determined. That is, the network feedback data may include the resource utilization rate.

[0090] Specifically, in order to train the first model using network feedback, the first network element needs to collect first data, second data, and network feedback data from different modules and different time periods.

[0091] The first network element can be a data collection module (e.g., memory) in the network. The data collection module can be deployed in the core network or in other devices or network elements. This application does not limit this.

[0092] Optionally, the first and second data may come from the intelligent agent, and the network feedback data may come from terminal devices, network devices, or network elements of the core network.

[0093] Optionally, the first network element can also obtain the type information of each data in multiple data sets.

[0094] The type information includes input data, output data, or reward data.

[0095] For example, the acquired data can be shown in Table 1. Specifically, the data includes data 1, data 2, data 3, etc. Each data includes type information and payload; where the type information for data 1 is input; the type information for data 2 is output; the type information for data 3 is reward; and so on.

[0096] Table 1

[0097] Optionally, the first network element can also classify multiple data points according to type information to obtain first data, second data, and network feedback data. Specifically, data with type information of "input" is identified as first data, data with type information of "output" is identified as second data, and data with type information of "reward" is identified as network feedback data.

[0098] For example, for the multiple data in Table 1 above, since the type information of data 1 is input, data 1 is identified as the first data; since the type information of data 2 is output, data 2 is identified as the second data; since the type information of data 3 is reward, data 3 is identified as network feedback data.

[0099] S502: Train the first model based on the first data, the second data, and the network feedback data.

[0100] Specifically, after the first network element obtains the first data, the second data, and the network feedback data, in order to train the first model, it also needs to obtain the identification information corresponding to the first data, the identification information corresponding to the second data, and the identification information corresponding to the network feedback data. Then, based on the identification information, the first data, the second data, and the network feedback data are processed to obtain the data for training the first model.

[0101] The identification information may include a task identifier (task ID) and / or a time sequence (TS). The task identifier may be assigned when the network receives a service request, and all subsequent operations related to that service request will carry this task identifier. The time sequence is used to distinguish the order in which multiple agents make collaborative decisions.

[0102] For example, in smart city applications, when a user requests to sense traffic flow at a certain intersection, the network assigns a task ID of 213 to identify it. Subsequent task decomposition planning, tool invocation, and business deployment execution will all use "213" for identification.

[0103] For example, the first data corresponding to task A includes input data 1 and input data 2. The time series corresponding to input data 1 is the first moment within a certain time period, and the time series corresponding to input data 2 is the second moment within the same time period. If the first moment is before the second moment, it means that the time series corresponding to input data 1 is before the time series corresponding to input data 2.

[0104] For example, the acquired data can also be shown in Table 2. Specifically, the data includes data A, data B, data C, etc. Each data includes a task identifier, type information, and payload; among them, the task identifier corresponding to data A is "213" and the type information is input; the task identifier corresponding to data B is "213" and the type information is output; the task identifier corresponding to data C is "714" and the type information is input; the task identifier corresponding to data D is "213" and the type information is reward; the task identifier corresponding to data E is "714" and the type information is output; the task identifier corresponding to data F is "714" and the type information is reward; and so on, which will not be repeated here.

[0105] Table 2

[0106] In one possible implementation, the identification information includes a task identifier, and the first network element can determine the first data, second data, and network feedback data corresponding to the same task based on the task identifier.

[0107] Specifically, when an agent in a network processes multiple tasks, there will be different inputs, outputs, and network node feedbacks corresponding to different tasks. In this regard, the first network element can integrate the acquired multiple data according to the task identifier, and identify the data with the same task identifier as the first data, second data, and network feedback data corresponding to the same task. Then, the first data, second data, and network feedback data corresponding to the same task are used as the data for training the first model.

[0108] For example, for multiple data points in Table 2 above, based on the task identifier, the first data, second data, and network feedback data corresponding to the same task in Table 2 are determined, as shown in Table 3. Specifically, for task identifier "213", the first data (i.e., type information is input) includes data A, the second data (i.e., type information is output) includes data B, and the network feedback data (i.e., type information is reward) includes data D; for task identifier "714", the first data (i.e., type information is input) includes data C, the second data (i.e., type information is input) includes data E, and the network feedback data (i.e., type information is reward) includes data F; others are similar and will not be elaborated here.

[0109] Table 3

[0110] It should be noted that if a task identifier is missing at least one of the corresponding first data, second data, or network feedback data, it means that the collection of data for that task identifier has failed. Therefore, all data corresponding to that task identifier will not be used in the training of the first model.

[0111] In another possible implementation, the identification information includes task identifier and time series. The first network element can first determine the first data, second data and network feedback data corresponding to the same task based on the task identifier, and then number the first data, second data and network feedback data corresponding to the same task according to the time series.

[0112] First, there may be more than one agent in the network. When multiple agents work together to complete a task, there will be multiple inputs and outputs, and there will be a temporal sequence relationship between them. In this regard, the first network element can integrate the multiple data according to the task identifier, and identify the data with the same task identifier as the first data, second data and network feedback data corresponding to the same task.

[0113] In this application, the first data and the second data corresponding to the same task have a one-to-one mapping relationship. Furthermore, for multiple first data corresponding to the same task, the payloads of different first data can be the same or different, and this application does not limit this.

[0114] For example, the first data corresponding to task A includes input data 1 and input data 2, and the second data corresponding to task A includes output data 1 and output data 2. If output data 1 is the data output by the first model after input data 1, and output data 2 is the data output by the first model after input data 2, then it means that input data 1 corresponds to output data 1, and input data 2 corresponds to output data 2.

[0115] Optionally, the network feedback data corresponding to a certain task can be obtained by the network node evaluating any one of the multiple second data points corresponding to that task. Specifically, the agent can randomly select one second data point from multiple second data points as the target decision, and the network node obtains the network feedback data corresponding to that task by evaluating the specific execution of the target decision.

[0116] For example, the first data corresponding to task A includes input data 1 and input data 2, and the second data corresponding to task A includes output data 1 and output data 2. Input data 1 corresponds to output data 1, and input data 2 corresponds to output data 2. If the agent selects output data 2 as the target decision, then by executing the decision corresponding to output data 2, the business requirements corresponding to input data 2 can be realized. The network node obtains the network feedback data corresponding to task A by evaluating the specific execution.

[0117] Secondly, based on the time series, the first data, second data, and network feedback data corresponding to the same task are numbered to determine the order of collaborative decision-making among multiple agents. The specific process mainly includes the following three scenarios:

[0118] Specifically, the first and second data that have a mapping relationship use the same number; for network feedback data, their corresponding number can be set to "0xFF".

[0119] Case 1: The numbers increase continuously. Specifically, the output data of the previous agent may be the input data of the next agent. In this case, the number corresponding to the input data of the next agent is equal to the number corresponding to the output data of the previous agent plus the first value.

[0120] In one possible design, the first data corresponding to the same task includes first input data and second input data, and the second data corresponding to the same task includes first output data and second output data. The first input data corresponds to the first output data, and the second input data corresponds to the second output data. The payload of the first output data is the same as the payload of the second input data, and the time series of the first output data precedes the time series of the second input data. The first network element can add a first value to the number corresponding to the first output data according to the chronological order of the time series of the first output data and the time series of the second input data to obtain the number corresponding to the second input data.

[0121] As shown in Figure 6, Figure 6 is a schematic diagram of a numbering method provided in an embodiment of this application. Multiple intelligent agents collaborate to complete a task. These agents include, but are not limited to, agent 1, agent 2, and agent 3. The output data of agent 1 is the input data of agent 2, and the output data of agent 2 is the input data of agent 3. The output data of agent 1 is numbered TS1, with a first value of 1. Based on the time sequence of the output data of agent 1 and the time sequence of the input data of agent 2, the input data of agent 2 is numbered TS1+1=TS2. Similarly, based on the time sequence of the output data of agent 2 and the time sequence of the input data of agent 3, the input data of agent 3 is numbered TS2+1=TS3.

[0122] Scenario 2: Same ID. Specifically, the output data of the previous agent may be sent to two agents in parallel. In this case, the input data of the two agents can use the same ID.

[0123] In one possible design, the first data corresponding to the same task includes first input data, second input data, and third input data; the second data corresponding to the same task includes first output data, second output data, and third output data; the first input data corresponds to the first output data; the second input data corresponds to the second output data; and the third input data corresponds to the third output data. The load of the first output data is the same as the load of the second input data and the load of the third input data, respectively. The time series of the first output data precedes the time series of the second input data, and the time series of the second input data and the time series of the third input data are the same. The first network element can add a first value to the number corresponding to the first output data according to the chronological order of the time series of the first output data and the time series of the second input data to obtain the number corresponding to the second input data; and determine that the number corresponding to the third input data is equal to the number corresponding to the second input data according to the chronological order of the time series of the second input data and the time series of the third input data.

[0124] As shown in Figure 7, Figure 7 is a schematic diagram of another numbering method provided in an embodiment of this application. Multiple agents collaborate to complete a task. These agents include, but are not limited to, Agent 1, Agent 2, and Agent 3. The output data of Agent 1 is the input data of Agent 2 and Agent 3, respectively. The output data of Agent 1 is numbered TS1, with a first value of 1. Based on the chronological order of the output data of Agent 1 and the input data of Agent 2, the input data of Agent 2 is numbered TS1+1=TS2. Since the time sequence of the input data of Agent 2 is the same as that of Agent 3, the input data of Agent 3 is numbered TS2.

[0125] Case 3: The numbers increase discontinuously. Specifically, the numbers corresponding to the input and output data of different agents increase continuously on the concurrent line. If the output data of two agents is the input data of a certain agent, then the number corresponding to the input data of that certain agent is equal to the maximum number corresponding to the output data of the two agents plus a first value.

[0126] In one possible design, the first data corresponding to the same task includes first input data, second input data, and third input data; the second data corresponding to the same task includes first output data, second output data, and third output data; the first input data corresponds to the first output data; the second input data corresponds to the second output data; the third input data corresponds to the third output data; the payload of the third input data includes the payload of the first output data and the payload of the second output data; the time series of the third input data follows the time series of the first and second output data. The first network element can, according to the chronological order of the time series of the first, second, and third output data, add a first value to the number corresponding to the first output data to obtain the number corresponding to the third input data, where the time series of the first output data follows the time series of the second output data; or, the first network element can also, according to the chronological order of the time series of the first, second, and third input data, add a first value to the number corresponding to the second output data to obtain the number corresponding to the third input data, where the time series of the second output data follows the time series of the first output data.

[0127] As shown in Figure 8, Figure 8 is a schematic diagram of another numbering method provided in an embodiment of this application. Multiple intelligent agents collaborate to complete a task. These multiple intelligent agents include, but are not limited to, Intelligent Agent 1, Intelligent Agent 2, Intelligent Agent 3, and Intelligent Agent 4. The output data of Intelligent Agent 1 is the input data of Intelligent Agent 2, and the output data of Intelligent Agent 2 and Intelligent Agent 3 are the input data of Intelligent Agent 4. The output data of Intelligent Agent 1 is numbered TS1, the output data of Intelligent Agent 3 is numbered TS1, and the first value is 1. Based on the chronological order of the output data of Intelligent Agent 1 and the input data of Intelligent Agent 2, the input data of Intelligent Agent 2 is numbered TS1+1=TS2. Since the output data of Intelligent Agent 2 follows the output data of Intelligent Agent 3, the input data of Intelligent Agent 4 is numbered TS2+1=TS3.

[0128] Finally, the first network element can sort the first data, second data, and network feedback data corresponding to the same task according to the number, and use the sorted first data, second data, and network feedback data corresponding to the same task as the data for training the first model.

[0129] For example, the sorted first data, second data, and network feedback data can be shown in Table 4. Specifically, the same task can include multiple first data and multiple second data. Each first / second data includes a task identifier, type information, number, and payload. Among them, the task identifier corresponding to data 1 is "213", the type information is input, and the number is "1"; the task identifier corresponding to data 2 is "213", the type information is output, and the number is "1"; the task identifier corresponding to data 3 is "213", the type information is input, and the number is "2"; the task identifier corresponding to data 4 is "213", the type information is output, and the number is "2"; the task identifier corresponding to data 5 is "213", the type information is reward, and the number is "0xFF"; others are similar and will not be described in detail here.

[0130] Table 4

[0131] Optionally, if the output data of the previous agent is the input data of the next agent, then some input data can be omitted.

[0132] For example, for multiple data in Table 4 above, if the load of data 2 is the same as the load of data 3, it means that data 2 is the output data of the previous agent and data 3 is the input data of the next agent. Data 3 can be omitted, as shown in Table 5.

[0133] Table 5

[0134] Alternatively, time series can be replaced by time indexes, which are timestamps (e.g., dates and times) corresponding to each piece of data, giving the data a clear temporal order.

[0135] Optionally, the identification information may also include agent ID, which is used to filter out the data of a specific agent for individual fine-tuning.

[0136] In this embodiment, by acquiring multiple data sets, the first network element can train a first model based on the first data, second data, and network feedback data. By acquiring identification information, the first network element can integrate the collected data sets based on the task identifier in the identification information, connecting the first data, second data, and network feedback data belonging to the same task. The first network element can also number the first data, second data, and network feedback data corresponding to the same task in the multiple data sets based on the time series in the identification information. This allows the data to carry the order information of collaborative decision-making among multiple agents when standardizing the data format of network feedback, which helps to improve the effectiveness of optimizing agent decisions using network feedback and improve the performance of the communication system.

[0137] The methods of the embodiments of this application have been described in detail above. The following is a description of the apparatus provided in the embodiments of this application.

[0138] As shown in Figure 9, which is a schematic diagram of a communication device according to an embodiment of this application, the communication device includes an acquisition module 901 and a processing module 902. The detailed descriptions of each module are as follows.

[0139] The acquisition module 901 is used to acquire multiple data, including first data, second data, and network feedback data. The first data is the data input to the first model, the second data is the data output by the first model after the first data is processed, and the network feedback data is the network quality-related data obtained by evaluating the second data through network nodes.

[0140] The processing module 902 is used to train the first model based on the first data, the second data, and the network feedback data.

[0141] Optionally, the acquisition module 901 is also used to acquire type information for each data in the multiple data, including input data, output data, or feedback data; the processing module 902 is also used to classify the multiple data according to the type information.

[0142] Optionally, the acquisition module 901 is further configured to acquire the identification information corresponding to the first data, the identification information corresponding to the second data, and the identification information corresponding to the network feedback data; the processing module 902 is further configured to process the first data, the second data, and the network feedback data according to the identification information.

[0143] Optionally, the identification information includes a task identifier; the processing module 902 is also used to determine the first data, the second data, and the network feedback data corresponding to the same task based on the task identifier.

[0144] Optionally, the identification information includes a task identifier and a time series; the processing module 902 is further configured to determine the first data, second data, and network feedback data corresponding to the same task based on the task identifier; the processing module 902 is further configured to number the first data, second data, and network feedback data corresponding to the same task based on the time series.

[0145] Optionally, the first data corresponding to the same task includes first input data and second input data, and the second data corresponding to the same task includes first output data and second output data. The first input data corresponds to the first output data, and the second input data corresponds to the second output data. The payload of the first output data is the same as the payload of the second input data, and the time series of the first output data precedes the time series of the second input data. The processing module 902 is further configured to add a first value to the number corresponding to the first output data according to the chronological order of the time series of the first output data and the time series of the second input data to obtain the number corresponding to the second input data.

[0146] Optionally, the first data corresponding to the same task includes first input data, second input data, and third input data; the second data corresponding to the same task includes first output data, second output data, and third output data; the first input data corresponds to the first output data; the second input data corresponds to the second output data; and the third input data corresponds to the third output data. The payload of the first output data is the same as the payload of the second input data and the payload of the third input data, respectively. The time series of the first output data precedes the time series of the second input data, and the time series of the second input data is the same as the time series of the third input data. The processing module 902 is further configured to add a first value to the number corresponding to the first output data according to the chronological order of the time series of the first output data and the time series of the second input data to obtain the number corresponding to the second input data. The processing module 902 is further configured to determine that the number corresponding to the third input data is equal to the number corresponding to the second input data according to the chronological order of the time series of the second input data and the time series of the third input data.

[0147] Optionally, the first data corresponding to the same task includes first input data, second input data, and third input data; the second data corresponding to the same task includes first output data, second output data, and third output data; the first input data corresponds to the first output data; the second input data corresponds to the second output data; the third input data corresponds to the third output data; the payload of the third input data includes the payload of the first output data and the payload of the second output data; the time series of the third input data is after the time series of the first output data and the time series of the second output data; the processing module 902 is further configured to add a first value to the number corresponding to the first output data according to the chronological order of the time series of the first output data, the time series of the second output data, and the time series of the third input data to obtain the number corresponding to the third input data; the time series of the first output data is after the time series of the second output data; or the processing module 902 is further configured to add a first value to the number corresponding to the second output data according to the chronological order of the time series of the first output data, the time series of the second output data, and the time series of the third input data to obtain the number corresponding to the third input data; the time series of the second output data is after the time series of the first output data.

[0148] Optionally, the first data includes at least one of the following: user demand information, search-enhanced RAG information, or prompt information.

[0149] Optionally, the second data includes at least one of the following: task decomposition results, tool call information, or network node configuration information.

[0150] Optionally, network feedback data may include at least one of the following: resource utilization, network throughput, task completion accuracy, or average task latency.

[0151] It should be noted that the implementation of each module can also refer to the corresponding descriptions of the method embodiments shown in Figures 5-8, and execute the methods and functions performed by the second network element in the above embodiments.

[0152] Figure 10 is a schematic diagram of a communication device provided in an embodiment of this application. The communication device can be a chip or processing system in a communication system or communication device, and can implement any of the methods and functions in any of the foregoing embodiments.

[0153] As shown in Figure 10, the communication device includes a processor 1001, which is configured to perform the actions described in the above method embodiments. Optionally, the communication device also includes a transceiver 1002. Optionally, the communication device also includes a memory 1003, which is configured to store a computer program. The processor 1001 retrieves and runs the computer program from the memory 1003 to control the transceiver 1002 to transmit and receive signals. Optionally, the communication device may also include an antenna, which is configured to transmit uplink data or uplink control signaling output by the transceiver 1002 via a wireless signal.

[0154] The processor 1001, transceiver 1002 and memory 1003 can communicate with each other through internal connection channels to transmit control and / or data signals.

[0155] The processor 1001 and the memory 1003 can be combined into a single processing device. The processor 1001 is configured to execute the program code stored in the memory 1003 to achieve the above-mentioned functions. In specific implementations, the memory 1003 can be integrated into the processor 1001 or independent of the processor 1001.

[0156] The transceiver 1002 described above can also be referred to as a transceiver unit or transceiver module. The transceiver 1002 may include a receiver (or receiver circuit) and a transmitter (or transmitter circuit). The receiver is configured to receive signals, and the transmitter is configured to transmit signals.

[0157] It should be understood that the communication device shown in Figure 10 can implement the various processes in the method embodiments shown in Figures 5-8. The operation and / or function of each module in the communication device are respectively for implementing the corresponding processes in the above method embodiments. For details, please refer to the description in the above method embodiments; to avoid repetition, detailed descriptions are appropriately omitted here.

[0158] The processor 1001 can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor 1001 can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, etc. The communication bus 1004 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is used in Figure 10, but this does not indicate that there is only one bus or one type of bus. The communication bus 1004 is configured to enable communication between these components. In this embodiment, the transceiver 1002 is configured to communicate with other node devices for signaling or data. The memory 1003 may include volatile memory, such as nonvolatile random access memory (NVRAM), phase change RAM (PRAM), magnetoresistive RAM (MRAM), etc., and may also include nonvolatile memory, such as at least one disk storage device, electrically erasable programmable read-only memory (EEPROM), flash memory devices, such as NOR flash memory or NAND flash memory, semiconductor devices, such as solid-state disks (SSDs), etc. The memory 1003 may also be at least one storage device located remotely from the aforementioned processor 1001. The memory 1003 may also store a set of computer program code or configuration information. The processor 1001 may also execute the program stored in the memory 1003. The processor 1001 can cooperate with the memory 1003 and the transceiver 1002 to perform any of the methods and functions involved in the above-described embodiments.

[0159] This application also provides a chip system including a processor for supporting a communication system or communication device to implement the functions involved in any of the above embodiments, such as generating first data, second data and network feedback data corresponding to the same task or processing the data features involved in the above methods.

[0160] This application also provides a computer program product, which includes a computer program that, when run on a computer, causes the computer to perform the method of any one of the embodiments shown in FIG5-FIG8.

[0161] This application also provides a computer-readable medium storing a computer program that, when run on a computer, causes the computer to perform the method of any one of the embodiments shown in FIG5-FIG8.

[0162] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the communication device, the unit or module within the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0163] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0164] It should be understood that the "and / or" appearing in the embodiments of this application is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0165] It should be understood that in the embodiments of this application, "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean that B is determined solely based on A; B can also be determined based on A and / or other information.

[0166] It should be understood that the symbol " / " appearing in the embodiments of this application can indicate that the preceding and following objects are in an "or" relationship. Additionally, the symbol " / " can also represent a division sign, i.e., performing a division operation. For example, A / B can mean A divided by B.

[0167] It should be understood that some or all of the steps in the embodiments of this application may be performed. These steps or operations are merely examples. In the embodiments of this application, other operations or variations of various operations may also be performed. Furthermore, the steps may be performed in different orders as presented in the embodiments of this application, and it is not necessary to perform all the operations in the embodiments of this application.

[0168] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. Any modifications, equivalent substitutions, or improvements made within the principles of this application should be included within the scope of protection of this application.

Claims

1. A communication method, characterized in that, include: Multiple data are acquired, including first data, second data, and network feedback data. The first data is the data input to the first model, the second data is the data output by the first model after the first data is processed, and the network feedback data is the network quality-related data obtained by evaluating the second data through network nodes. The first model is trained based on the first data, the second data, and the network feedback data.

2. The method as described in claim 1, characterized in that, The method further includes: Obtain the type information of each data in the plurality of data, wherein the type information includes input data, output data, or feedback data; The multiple data are classified according to the type information.

3. The method as described in claim 1 or 2, characterized in that, The step of training the first model based on the first data, the second data, and the network feedback data includes: Obtain the identification information corresponding to the first data, the identification information corresponding to the second data, and the identification information corresponding to the network feedback data; The first data, the second data, and the network feedback data are processed according to the identification information.

4. The method as described in claim 3, characterized in that, The identification information includes a task identifier; the processing of the first data, the second data, and the network feedback data based on the identification information includes: Based on the task identifier, determine the first data, second data, and network feedback data corresponding to the same task.

5. The method as described in claim 3, characterized in that, The identification information includes a task identifier and a time series; the processing of the first data, the second data, and the network feedback data based on the identification information includes: Based on the task identifier, determine the first data, second data, and network feedback data corresponding to the same task; Based on the time series, the first data, second data, and network feedback data corresponding to the same task are numbered.

6. The method as described in claim 5, characterized in that, The first data corresponding to the same task includes first input data and second input data, and the second data corresponding to the same task includes first output data and second output data. The first input data corresponds to the first output data, and the second input data corresponds to the second output data. The payload of the first output data is the same as the payload of the second input data, and the time series of the first output data is before the time series of the second input data. The step of numbering the first data, second data, and network feedback data corresponding to the same task according to the time series includes: Based on the chronological order of the time series of the first output data and the time series of the second input data, the number corresponding to the first output data is added to the first value to obtain the number corresponding to the second input data.

7. The method as described in claim 5, characterized in that, The first data corresponding to the same task includes first input data, second input data, and third input data. The second data corresponding to the same task includes first output data, second output data, and third output data. The first input data corresponds to the first output data, the second input data corresponds to the second output data, and the third input data corresponds to the third output data. The payload of the first output data is the same as the payload of the second input data and the payload of the third input data, respectively. The time series of the first output data is earlier than the time series of the second input data, and the time series of the second input data and the time series of the third input data are the same. The step of numbering the first data, second data, and network feedback data corresponding to the same task according to the time series includes: According to the chronological order of the time series of the first output data and the time series of the second input data, the number corresponding to the first output data is added to the first value to obtain the number corresponding to the second input data; Based on the chronological order of the time series of the second input data and the third input data, the number corresponding to the third input data is determined to be equal to the number corresponding to the second input data.

8. The method as described in claim 5, characterized in that, The first data corresponding to the same task includes first input data, second input data, and third input data. The second data corresponding to the same task includes first output data, second output data, and third output data. The first input data corresponds to the first output data, the second input data corresponds to the second output data, and the third input data corresponds to the third output data. The payload of the third input data includes the payload of the first output data and the payload of the second output data. The time series of the third input data is after the time series of the first output data and the time series of the second output data. The step of numbering the first data, second data, and network feedback data corresponding to the same task according to the time series includes: According to the chronological order of the time series of the first output data, the time series of the second output data, and the time series of the third input data, the number corresponding to the first output data is added to the first value to obtain the number corresponding to the third input data, and the time series of the first output data is after the time series of the second output data. or According to the chronological order of the time series of the first output data, the second output data, and the third input data, the number corresponding to the second output data is added to the first value to obtain the number corresponding to the third input data. The time series of the second output data is after the time series of the first output data.

9. The method according to any one of claims 1-8, characterized in that, The first data includes at least one of the following: user demand information, search-enhanced RAG information, or prompt information.

10. The method according to any one of claims 1-9, characterized in that, The second data includes at least one of the following: task decomposition results, tool call information, or configuration information of the network node.

11. The method according to any one of claims 1-10, characterized in that, The network feedback data includes at least one of the following: resource utilization, network throughput, task completion accuracy, or average task latency.

12. A communication device, characterized in that, include: The acquisition module is used to acquire multiple data, including first data, second data, and network feedback data. The first data is the data input to the first model, the second data is the data output by the first model after the first data is processed, and the network feedback data is network quality-related data obtained by evaluating the second data through network nodes. The processing module is used to train the first model based on the first data, the second data, and the network feedback data.

13. The apparatus as claimed in claim 12, characterized in that, The acquisition module is further configured to acquire type information for each of the plurality of data, wherein the type information includes input data, output data, or feedback data; The processing module is further configured to classify the plurality of data according to the type information.

14. The apparatus as claimed in claim 12 or 13, characterized in that, The acquisition module is further configured to acquire the identification information corresponding to the first data, the identification information corresponding to the second data, and the identification information corresponding to the network feedback data; The processing module is further configured to process the first data, the second data, and the network feedback data according to the identification information.

15. The apparatus as claimed in claim 14, characterized in that, The identification information includes a task identifier; The processing module is further configured to determine the first data, the second data, and the network feedback data corresponding to the same task based on the task identifier.

16. The apparatus as claimed in claim 14, characterized in that, The identification information includes task identifier and time sequence; The processing module is further configured to determine the first data, the second data, and the network feedback data corresponding to the same task based on the task identifier; The processing module is further configured to number the first data, second data, and network feedback data corresponding to the same task according to the time series.

17. The apparatus as claimed in claim 16, characterized in that, The first data corresponding to the same task includes first input data and second input data, and the second data corresponding to the same task includes first output data and second output data. The first input data corresponds to the first output data, and the second input data corresponds to the second output data. The payload of the first output data is the same as the payload of the second input data, and the time series of the first output data is before the time series of the second input data. The processing module is further configured to add a first value to the number corresponding to the first output data according to the chronological order of the time sequence of the first output data and the time sequence of the second input data, so as to obtain the number corresponding to the second input data.

18. The apparatus as claimed in claim 16, characterized in that, The first data corresponding to the same task includes first input data, second input data, and third input data. The second data corresponding to the same task includes first output data, second output data, and third output data. The first input data corresponds to the first output data, the second input data corresponds to the second output data, and the third input data corresponds to the third output data. The payload of the first output data is the same as the payload of the second input data and the payload of the third input data, respectively. The time series of the first output data is earlier than the time series of the second input data, and the time series of the second input data and the time series of the third input data are the same. The processing module is further configured to add a first value to the number corresponding to the first output data according to the chronological order of the time sequence of the first output data and the time sequence of the second input data, so as to obtain the number corresponding to the second input data. The processing module is further configured to determine, according to the chronological order of the time series of the second input data and the time series of the third input data, that the number corresponding to the third input data is equal to the number corresponding to the second input data.

19. The apparatus as claimed in claim 16, characterized in that, The first data corresponding to the same task includes first input data, second input data, and third input data. The second data corresponding to the same task includes first output data, second output data, and third output data. The first input data corresponds to the first output data, the second input data corresponds to the second output data, and the third input data corresponds to the third output data. The payload of the third input data includes the payload of the first output data and the payload of the second output data. The time series of the third input data is after the time series of the first output data and the time series of the second output data. The processing module is further configured to add a first value to the number corresponding to the first output data according to the chronological order of the time sequence of the first output data, the time sequence of the second output data, and the time sequence of the third input data, to obtain the number corresponding to the third input data, wherein the time sequence of the first output data is after the time sequence of the second output data. or The processing module is further configured to add a first value to the number corresponding to the second output data according to the chronological order of the time series of the first output data, the time series of the second output data, and the time series of the third input data, to obtain the number corresponding to the third input data, wherein the time series of the second output data is after the time series of the first output data.

20. The apparatus according to any one of claims 12-19, characterized in that, The first data includes at least one of the following: user demand information, search-enhanced RAG information, or prompt information.

21. The apparatus according to any one of claims 12-20, characterized in that, The second data includes at least one of the following: task decomposition results, tool call information, or configuration information of the network node.

22. The apparatus according to any one of claims 12-21, characterized in that, The network feedback data includes at least one of the following: resource utilization, network throughput, task completion accuracy, or average task latency.

23. A communication device, characterized in that, Includes a processor that causes the communication device to perform the method of any one of claims 1-11.

24. A communication system, characterized in that, It includes a first network element, which is used to perform the method of any one of claims 1-11.

25. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a computer program that, when executed by a processor, causes the method as described in any one of claims 1-11 to be implemented.

26. A chip, characterized in that, The chip includes a processor and a communication interface for communicating with external or internal devices, and the processor enables the chip to implement the method as described in any one of claims 1-11.

27. A computer program product, characterized in that, The computer program product includes a computer program that, when run on a computer, causes the computer to perform the method according to any one of claims 1-11.

Images