Communication method and related apparatus

Abstract

Provided in the embodiments of the present application are a communication method and a related apparatus. The method comprises: receiving first indication information, wherein the first indication information is used for indicating a packeting specification of data of a first application, and remote computing support is provided to the data of the first application by a first communication apparatus on which a large model is deployed; packaging data to be sent of the first application according to the packeting specification, so as to obtain a first data packet; sending the first data packet to the first communication apparatus; locally caching data, other than the first data packet, in the data to be sent of the first application; and receiving a computing result for the data of the first application. By using the method, the waste of communication resources can be reduced.

Description

A communication method and related apparatus

[0001] This application claims priority to Chinese Patent Application No. 202510032719.1, filed on January 8, 2025, entitled "A Communication Method and Related Device", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communication technology, and in particular to a communication method and related apparatus. Background Technology

[0003] With the development of Artificial Intelligence (AI) technology, AI models are gradually evolving from traditional small-scale neural network models (such as Multilayer Perceptron (MLP), Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN)) to large-scale neural network models based on Transformers. Simultaneously, terminal applications based on large models are constantly emerging, such as multimodal real-time dialogue application services like ChatGPT-4o. In such applications, the large service model of ChatGPT-4o is typically deployed on cloud servers. These cloud servers, where the large service model is deployed, can be called compute nodes (FeINs), also known as the compute schedulers for the large model inference service. These compute nodes (FeINs) can employ the Virtual Large Language Model (vLLM) framework or other frameworks. The vLLM framework manages requests within the large model inference engine, maximizing computational efficiency. The computing node (FeIN) provides computing resources for terminal applications. As shown in Figure 1, the terminal (such as user equipment (UE)) sends service requests (including application-related data) directly or indirectly to the computing node (FeIN). The computing node (FeIN) performs inference calculations uniformly and then returns the inference calculation results to the terminal.

[0004] In practical applications, the computing power of the compute node (FeIN) has an upper limit, and the processing latency of service requests increases with the number of concurrent service requests. Therefore, the compute node (FeIN) may not be able to process service requests from the terminal immediately. If not processed in time, it may trigger the retransmission of the service request, which will cause a single service request to consume many times more communication resources. Summary of the Invention

[0005] This application discloses a communication method and related apparatus that can save communication resources.

[0006] In a first aspect, embodiments of this application provide a communication method, the method comprising:

[0007] Receive first indication information, wherein the first indication information is used to indicate the data packet specifications of the first application, and the data of the first application is remotely computed by a first communication device that has deployed a large model;

[0008] The data of the first application to be sent is packaged according to the packetization specification to obtain a first data packet; optionally, the data size of the first data packet is less than or equal to the packetization specification.

[0009] Send the first data packet to the first communication device;

[0010] The remaining data in the data of the first application to be sent, excluding the first data packet, is cached locally;

[0011] Receive the calculation results for the data of the first application.

[0012] In the above method, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, thus preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0013] In conjunction with the first aspect, in one possible implementation of the first aspect, the first indication information includes the package specification.

[0014] In conjunction with the first aspect, in one possible implementation of the first aspect, the first indication information includes a first parameter, the first parameter representing available computing resources for computing data of the first application, the first parameter being used to calculate the package specification, and the method further includes: determining the package specification based on the first parameter, which is equivalent to the first indication information indirectly indicating the package specification.

[0015] In another possible implementation of the first aspect, in conjunction with the first aspect or any of the above possible implementations of the first aspect, the method further includes: sending a first service request, wherein the first service request is used to query available computing resources for the first application.

[0016] In another possible implementation of the first aspect, in conjunction with the first aspect or any of the above possible implementations of the first aspect, the first service request includes a first identifier, and the first identifier has a direct or indirect correspondence with the first application.

[0017] In another possible implementation of the first aspect, in conjunction with the first aspect or any of the above possible implementations of the first aspect, the first parameter includes the remaining number of tokens that can be accommodated and / or the remaining number of floating-point operations per second (FLOPS).

[0018] In conjunction with the first aspect or any of the above possible implementations of the first aspect, in another possible implementation of the first aspect, the method is applied to a second communication device, and receiving the first indication information includes:

[0019] Receive first instruction information from a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

[0020] In conjunction with the first aspect or any of the above possible implementations of the first aspect, in yet another possible implementation of the first aspect, the first indication information is further used to indicate a first time interval, and the method further includes:

[0021] A second data packet is sent after the first time interval, the second data packet being a data packet obtained from the remaining data packets.

[0022] Secondly, embodiments of this application provide a communication method, the method comprising:

[0023] Receive second instruction information, wherein the second instruction information includes a first parameter; the first parameter represents available computing resources for computing on data of the first application, the data of the first application being remotely computed by a first communication device deployed with a large model;

[0024] The data package specifications of the first application are determined based on the second instruction information;

[0025] Send a first indication message to a second communication device, wherein the first indication message is used to indicate the packet size of the data of the first application, and the packet size is used to limit the data size of the data packets sent to the first application.

[0026] In the above method, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, thus preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0027] In conjunction with the second aspect, one possible implementation of the second aspect also includes:

[0028] Receive a first data packet from a second communication device, wherein the first data packet is a data packet obtained by assembling the data of the first application according to the packet specification;

[0029] The first data packet is sent to the first communication device.

[0030] In a further possible implementation of the second aspect, in conjunction with the second aspect or any of the above possible implementations of the second aspect, determining the data packet specifications of the first application based on the second indication information includes:

[0031] The data package specification of the first application corresponding to the first parameter in the second indication information is determined according to the stored calculation rules or correspondence.

[0032] In a further possible implementation of the second aspect, in conjunction with the second aspect or any of the above possible implementations of the second aspect, determining the data packet specifications of the first application based on the second indication information includes:

[0033] If the channel status of other communication devices is better than that of the second communication device, then the computing resources corresponding to the first parameter in the second indication information will be preferentially allocated to the other communication devices, wherein the other communication devices and the second communication device are of the same type.

[0034] Based on the remaining resources after the computing resources corresponding to the first parameter are allocated, the packet specification of the data for the first application of the second communication device is determined.

[0035] In conjunction with the second aspect or any of the above-described possible implementations of the second aspect, another possible implementation of the second aspect further includes:

[0036] Receive a first service request from the second communication device, wherein the first service request is used to query available computing resources for the first application;

[0037] Send the first service request to the first communication device.

[0038] In conjunction with the second aspect or any of the above possible implementations of the second aspect, in yet another possible implementation of the second aspect, receiving the second instruction information includes:

[0039] Receive second instruction information from the first communication device.

[0040] In conjunction with the second aspect or any of the above possible implementations of the second aspect, in yet another possible implementation of the second aspect, the method is applied to a third communication device; the receiving of the second indication information includes:

[0041] Receive a second instruction message from a fourth communication device, wherein the fourth communication device is a management function node of the third communication device.

[0042] In another possible implementation of the second aspect, in conjunction with the second aspect or any of the above possible implementations of the second aspect, the first service request includes a first identifier, which has a direct or indirect correspondence with the first application.

[0043] In another possible implementation of the second aspect, in conjunction with the second aspect or any of the above possible implementations of the second aspect, the first parameter includes the remaining number of tokens that can be accommodated and / or the remaining number of floating-point operations per second (FLOPS).

[0044] In conjunction with the second aspect or any of the above possible implementations of the second aspect, in yet another possible implementation of the second aspect, the second indication information is further used to indicate the average computational latency within the second time period, and the step of determining the data packet specifications of the first application based on the second indication information includes:

[0045] The communication latency requirement is determined based on the total processing latency requirement of the data from the first application and the average calculation latency.

[0046] The specifications of the first packet group are determined based on the communication latency requirements;

[0047] The specifications of the second group of packages are determined based on the second instruction information;

[0048] Among them, the smaller of the first package specification and the second package specification is used as the package specification of the data of the first application.

[0049] In conjunction with the second aspect or any of the above possible implementations of the second aspect, in yet another possible implementation of the second aspect, the method further includes:

[0050] Send a third indication message to the first communication device, wherein the third indication message includes the second time period, and the third indication message is used to request the average calculation delay within the second time period.

[0051] In conjunction with the second aspect or any of the above-described possible implementations of the second aspect, another possible implementation of the second aspect further includes:

[0052] Receive a first feedback message from the first communication device, wherein the first feedback message indicates the effect of the first communication device in calculating the data of the first application;

[0053] If the effect of calculating the data of the first application does not meet the preset target, the second time period is updated, and the updated second time period is used for the next calculation of the average calculation latency.

[0054] Thirdly, embodiments of this application provide a communication method applied to a first communication device, the method comprising:

[0055] A first parameter is determined based on the maximum computing resources used to compute data from the first application and the computing resources already used to compute data from the first application, wherein the first parameter represents the available computing resources used to compute data from the first application.

[0056] Send a second instruction message, wherein the second instruction message includes the first parameter or the packet specification of the data of the first application, the first parameter is used to calculate the packet specification, and the packet specification is used to limit the data size of the data packet of the first application; the first communication device is used to provide remote calculation support for the data of the first application to the second communication device through the deployed large model.

[0057] In conjunction with the third aspect, in another possible implementation of the third aspect, the first parameter includes the remaining number of tokens that can be held and / or the remaining number of floating-point operations per second (FLOPS).

[0058] In combination with the third aspect or any of the above possible implementations of the third aspect, in yet another possible implementation of the third aspect, the first parameter specifically includes the average remaining number of tokens that can be accommodated within the reference duration and / or the remaining number of floating-point operations per second (FLOPS).

[0059] In a further possible implementation of the third aspect, in conjunction with the third aspect or any of the above possible implementations, the method further includes:

[0060] A first service request is received from the third communication device, wherein the first service request is used to query available computing resources for the first application.

[0061] In a further possible implementation of the third aspect, in conjunction with the third aspect or any of the above possible implementations, the method further includes:

[0062] A second service request is received from a fourth communication device, wherein the second service request is used to query the available computing resources for the first application within a reference time period.

[0063] In conjunction with the third aspect or any of the above possible implementations of the third aspect, in yet another possible implementation of the third aspect, the sending of the second instruction information includes:

[0064] Send a second instruction message to the fourth communication device.

[0065] In conjunction with the third aspect or any of the above possible implementations of the third aspect, in yet another possible implementation of the third aspect, the sending of the second instruction information includes:

[0066] Send a second instruction message to a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

[0067] Fourthly, embodiments of this application provide a communication method, the method comprising:

[0068] Send a second service request to the first communication device, wherein the second service request is used to request the availability of computing resources for the first application within a reference time period;

[0069] The system receives second indication information from the first communication device, wherein the second indication information includes the first parameter or the packet specification of the data of the first application, the first parameter represents the available computing resources for computing the data of the first application, the first parameter is used to calculate the packet specification, and the packet specification is used to limit the data size of the data packet sent to the first application; the first communication device is used to provide remote computing support for the data of the first application to the second communication device through a deployed large model.

[0070] The second instruction information is sent to a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

[0071] In conjunction with the fourth aspect, in one possible implementation of the fourth aspect, the first parameter specifically includes the average remaining number of tokens that can be accommodated within the reference duration and / or the remaining number of floating-point operations per second (FLOPS).

[0072] Fifthly, embodiments of this application provide a communication device (such as a second communication device), which can be a terminal or a device or functional module within a terminal, wherein:

[0073] The communication device includes a module for performing the method described in the first aspect or any possible implementation thereof;

[0074] Alternatively, the communication device includes a processor for performing the method described in the first aspect or any possible implementation thereof.

[0075] Sixthly, embodiments of this application provide a communication device (such as a third communication device), which can be a network device (such as a gNB) or a device or functional module within a network device, wherein:

[0076] The communication device includes a module for performing the method described in the second aspect or any possible implementation thereof;

[0077] Alternatively, the communication device includes a processor for performing the method described in the second aspect or any possible implementation thereof.

[0078] In a seventh aspect, embodiments of this application provide a communication device (such as a first communication device), which can be a computing node (such as FeIN) or a device or functional module within a computing node, wherein:

[0079] The communication device includes a module for performing the method described in the third aspect or any possible implementation of the third aspect;

[0080] Alternatively, the communication device includes a processor for performing the method described in the third aspect or any possible implementation thereof.

[0081] Eighthly, embodiments of this application provide a communication device (such as a fourth communication device), which can be an OAM or a device or functional module in an OAM, wherein:

[0082] The communication device includes a module for performing the method described in the fourth aspect or any possible implementation of the fourth aspect;

[0083] Alternatively, the communication device includes a processor for performing the method described in the fourth aspect or any possible implementation thereof.

[0084] Ninthly, embodiments of this application provide a communication device, characterized in that it includes a logic circuit and an interface, the logic circuit and the interface being coupled; the interface is used for inputting and / or outputting information, wherein:

[0085] The logic circuit is used to perform the method described in the first aspect or any possible implementation thereof, or...

[0086] The logic circuit is used to execute the method described in the second aspect or any possible implementation thereof, or...

[0087] The logic circuit is used to execute the method described in the third aspect or any possible implementation thereof, or...

[0088] The logic circuit is used to perform the method described in the fourth aspect or any possible implementation of the fourth aspect.

[0089] Tenthly, embodiments of this application provide a computer-readable storage medium for storing a computer program, wherein:

[0090] When the computer program is executed, it is capable of implementing the first aspect or any possible implementation of the first aspect, or...

[0091] When the computer program is executed, it is capable of implementing the second aspect or any possible implementation of the second aspect, or...

[0092] When the computer program is executed, it is capable of implementing the third aspect or any possible implementation of the third aspect, or...

[0093] When the computer program is executed, it is capable of implementing the fourth aspect or any possible implementation of the fourth aspect.

[0094] Eleventhly, embodiments of this application provide a communication system, which includes a first communication device, a second communication device, a third communication device, and a fourth communication device, wherein:

[0095] The second communication device is used to perform the method described in the first aspect or any possible implementation of the first aspect; the third communication device is used to perform the method described in the second aspect or any possible implementation of the second aspect; the first communication device is used to perform the method described in the third aspect or any possible implementation of the third aspect; and the fourth communication device is used to perform the method described in the fourth aspect or any possible implementation of the fourth aspect. Attached Figure Description

[0096] The accompanying drawings used in the embodiments of this application are described below.

[0097] Figure 1 is a schematic diagram of the architecture of a communication system provided in an embodiment of this application;

[0098] Figure 2 is a schematic diagram of the architecture of another communication system provided in an embodiment of this application;

[0099] Figure 3 is a flowchart illustrating a communication method provided in an embodiment of this application;

[0100] Figure 4 is a flowchart illustrating another communication method provided in an embodiment of this application;

[0101] Figure 5 is a flowchart illustrating another communication method provided in an embodiment of this application;

[0102] Figure 6 is a flowchart illustrating another communication method provided in an embodiment of this application;

[0103] Figure 7 is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0104] Figure 8 is a schematic diagram of the structure of a communication device provided in an embodiment of this application;

[0105] Figure 9 is a schematic diagram of the structure of a communication device provided in an embodiment of this application. Detailed Implementation

[0106] The embodiments of this application are described below with reference to the accompanying drawings.

[0107] Please refer to Figure 2, which is a schematic diagram of a communication system provided in an embodiment of this application. The communication system includes a first communication device 201, a second communication device 202, and a third communication device 203. Optionally, it also includes a fourth communication device 204, wherein:

[0108] The first communication device 201 deploys devices, nodes, or clusters of nodes with large-scale models. Large-scale models refer to machine learning models with massive parameters and complex computational structures, typically built from deep neural networks, possessing billions or even hundreds of billions of parameters. The purpose of large-scale models is to improve their expressive power and predictive performance, enabling them to handle more complex tasks and data. Large-scale models have wide applications in various fields, including natural language processing, computer vision, speech recognition, and recommendation systems. By training on massive amounts of data, large-scale models learn complex patterns and features, exhibiting stronger generalization capabilities and making accurate predictions on unseen data. This type of large-scale model is also called a foundation model. Ultra-large models are a subset of large-scale models, with a much larger number of parameters than ordinary large-scale models. For example, large-scale models can be various artificial neural network model structures such as large language models, vLLM, MLP, CNN, RNN, Transformer, Graph Neural Network (GNN), autoencoder (AE), encoder, and decoder. In this embodiment, the first communication device 201, which deploys a large model, can perform functions such as model inference or cloud / mobile edge computing (MEC) offloading. Therefore, it can provide application data computing support for multiple devices. If one of the multiple devices (such as the second communication device) has a need to compute the corresponding application data, it will send the corresponding application data (directly or indirectly) to the first communication device. After the first communication device performs the computation on the corresponding application data, it will feed back the computation result to the corresponding device. Since the computation task is concentrated in the first communication device, the computation pressure of the multiple devices can be reduced.

[0109] Optionally, the first communication device 201 may be a node or a device within a node, such as a computing node (FeIN), a task execution function (TEF), or a computation execution function (CEF).

[0110] The second communication device 202 is a device or a component in a device that has an application installed. The second communication device generates application data when running the application. The application data needs to be calculated to complete subsequent related operations. In this embodiment, the calculation task of the application data is completed remotely with the assistance of the first communication device. After the first communication device completes the calculation, it feeds back the calculation result to the second communication device.

[0111] Optionally, the second communication device can be user equipment (UE), wireless terminal equipment, mobile terminal equipment, device-to-device (D2D) terminal equipment, vehicle-to-everything (V2X) terminal equipment, machine-to-machine / machine-type communications (M2M / MTC) terminal equipment, Internet of Things (IoT) terminal equipment, light UE, reduced capability UE (REDCAP UE), subscriber unit, subscriber station, mobile station, remote station, access point (AP), remote terminal, access terminal, user terminal, user agent, or user device, etc. For example, this can include mobile phones (or "cellular" phones), smartphones, computers with mobile terminal devices, portable, pocket-sized, handheld, computer-embedded mobile devices, laptop computers, wireless data cards, tablet computers, wireless modems, etc. Examples include personal communication service (PCS) phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, and personal digital assistants (PDAs). It also includes limited devices, such as devices with low power consumption, limited storage capacity, or limited computing power. Examples include information sensing devices such as barcode scanners, radio frequency identification (RFID), sensors, global positioning systems (GPS), and laser scanners.

[0112] By way of example and not limitation, in this embodiment, the second communication device can also be a wearable device. Wearable devices, also known as wearable smart devices or smart wearable devices, are a general term for devices that utilize wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, 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 a specific type of application function and require the use of other devices such as smartphones, such as various smart bracelets, smart helmets, and smart jewelry for vital sign monitoring.

[0113] The second communication device described above, if located in a vehicle (e.g., placed inside or installed inside the vehicle), can be considered a vehicle-mounted second communication device, also known as an on-board unit (OBU). In the embodiments of this application, the second communication device may further include a relay. Alternatively, it can be understood that anything capable of data communication with a base station can be considered a second communication device.

[0114] In this embodiment of the application, the second communication device may be the corresponding device or device in the device mentioned above, such as a chip system. The chip system may be composed of chips or may include chips and other discrete devices.

[0115] The third communication device 203 serves as a communication node between the first communication device 201 and the second communication device 202, and / or a communication node between other communication devices and the second communication device. For example, the third communication device 203 may be an access network node, a Task Control Function (TCF), or a Computation Control Function (CCF) in the network accessed by the second communication device 202. The TCF or CCF is responsible for functions such as selecting computing nodes, partitioning computing tasks, discovering computing power registration, and scheduling computing. The access network node may include access network (AN) equipment, such as a base station (e.g., an access point), which can refer to a device in the access network that communicates with wireless terminal devices via one or more cells over the air interface. Alternatively, for example, in a vehicle-to-everything (V2X) technology, the access network node may be a roadside unit (RSU). The base station can be used to convert received air frames and IP packets to each other, acting as a router between the terminal device and the rest of the access network, which may include an IP network. RSUs can be fixed infrastructure entities that support V2X applications and can exchange messages with other entities that support V2X applications. Access network nodes can also coordinate the management of air interface attributes. For example, access network nodes can include base transceiver stations (BTS) in Global System for Mobile Communication (GSM) or Code Division Multiple Access (CDMA) networks, NBs (NodeBs) in Wideband Code Division Multiple Access (WCDMA), evolved base stations (NodeBs, eNBs, or e-NodeBs) in Long Term Evolution-Advanced (LTE) or Long Term Evolution-Advanced (LTE-A) systems, or next-generation node Bs (gNBs) in 5G NR systems (also known as NR systems), or radio controllers, centralized units (CUs), and distributed units (DUs) in cloud radio access networks (Cloud RAN) systems.In addition, the access network node can also be other devices that have deployed network access functions, such as wearable devices, vehicle-mounted devices, and transmission and reception points (TRPs). This application embodiment is not limited to these.

[0116] In this embodiment of the application, the device for implementing the function of the access network node can be the access network node itself, or it can be a device that enables the access network node to implement the function, such as a chip system, which can be installed in the access network node.

[0117] The fourth communication device 204 is a network element or function that manages the third communication device. For example, the fourth communication device can be a network operation administration and maintenance (OAM) function, including a network management system (NMS) and an element management system (EMS). The NMS undertakes the operation, management and maintenance functions of the network and can be called a cross-domain management system. It can directly manage the EMS. The EMS can manage one or more network elements of a certain category and can also be called a domain management system or a single-domain management system. In the RAN or core network (CN) domain, the EMS of each domain can manage network elements within the domain. For example, the EMS of the RAN domain can manage network elements such as gNB, and the EMS of the CN domain can manage network elements such as the Network Data Analytics Function (NWDAF).

[0118] The network elements or functions mentioned above can be deployed in pairs, independently, or jointly with other existing network elements or functions not mentioned above. For example, CCF can be deployed jointly with gNB or with NWDAF. Furthermore, TxF / CxF can be wholly or partially located within the RAN or CN domain (where different x values ​​correspond to different network elements), and should be managed by the EMS of the corresponding domain. The third communication device (such as gNB) interacts with the second communication device (such as UE) during operation.

[0119] Optionally, in the Open Radio Access Network (ORAN) network domain, the Service Management and Orchestration (SMO) function plays a similar role to the NMS in the network architecture. It is responsible for the operation, management, and maintenance of various network services and orchestration functions. The network elements it directly manages can be heterogeneous, including gNB, NWDAF, TxF / CxF, etc. Data / model information in the network can be exposed to and interacted with the over-the-top (OTT) server corresponding to the second communication device (such as the UE) through the NMS or SMO. The data / model information of the second communication device (such as the UE) can then interact with the OTT server.

[0120] When performing application data computation based on the above architecture, taking the first communication device (computing node FeIN), the second communication device as the terminal, and the third communication device as gNB as an example, the specific execution process can be as follows: As shown in Figure 3, the service request sent by the terminal is forwarded to the computing node (FeIN) via the base station (gNodeB, gNB) / Computation Control Function (CCF). The computing node (FeIN) determines whether it is overloaded based on its local computing power (affecting the service latency guarantee of computing service requests in the queue). If it is overloaded, admission control is performed, and the remaining service requests will be directly rejected or postponed. For example, if a terminal request (typically containing multimodal data such as text, images, or video) arrives at the compute node (FeIN) via the current communication network and is forwarded by the gNB / CCF (forwarding consumes communication resources such as time-frequency resources in a wireless network, typically quantized into time-frequency resource blocks, i.e., resource blocks (RBs)), the compute node (FeIN) may become overloaded and unable to process the request due to excessive computational load requirements (e.g., in large model inference, text and images can be quantized into tokens, with a fixed amount of computational resources required per token, and the number of tokens indicating the computational load requirement). In this case, compute node (FeIN) admission control is usually triggered, meaning requests that trigger compute node (FeIN) overload are discarded, wasting the communication resources spent on transmitting the service request to the compute node (FeIN). To complete the computation of the service request, the terminal (or the corresponding application) will trigger a retransmission of the service request. If the compute node (FeIN) is idle when the retransmitted service request arrives, the service request can be computed; otherwise, the service request must be retransmitted again.

[0121] It can be seen that admission of compute nodes (FeIN) can cause a single service request to consume many times more communication resources.

[0122] To avoid wasting communication resources, the following section proposes an improved communication method based on Figures 4 to 6.

[0123] Please refer to Figure 4, which is a flowchart illustrating a communication method provided in an embodiment of this application. This method can be implemented based on the architecture shown in Figure 2, or on other architectures. The method includes, but is not limited to, the following steps:

[0124] Step S401: The second communication device sends a first service request.

[0125] Optionally, the second communication device (such as the UE) can send the first service request to the third communication device (such as the gNB). After receiving the first service request, the third communication device forwards it to the first communication device (such as the FeIN). The message formats of the first service request received by the third communication device and the first service request sent out can be the same or different, but their substantive functions are the same. They are both used to query available computing resources for the first application. For example, the first service request carries a first identifier, which has a direct or indirect correspondence with the first application. The first identifier indicates that available computing resources for the first application are to be queried. For example, the first identifier can be one or more of the following: application identifier, device identifier, quality of service (QoS) identifier, 5G QoS Identifier (5QI).

[0126] Optionally, the identifiers related to the first application carried in the first service request received and the first service request sent by the third communication device may be the same or different. For example, the identifier related to the first application in the first service request received by the third communication device may be the identifier of the first application, while the identifier related to the first application in the first service request sent by the third communication device may be a device identifier. However, there is an association between the device identifier and the identifier of the first application, so it can also indirectly identify the first application. Similarly, when the third communication device relays messages between the first communication device and the second communication device, it may also involve converting the identifiers related to the first application in the corresponding messages. For example, it may convert the direct identifier of the first application into an indirect identifier before forwarding, or convert the indirect identifier of the first application into a direct identifier before forwarding. Specific messages will not be described again later.

[0127] In one alternative scheme, after receiving the first service request, the third communication device knows that it needs to send the first service request to the first communication device. In another alternative scheme, after receiving the first service request, the third communication device needs to query which node provides the computing service for the first application based on the first application corresponding to the first service request. It can be understood that the third communication device can determine the node that needs to provide the computing service for the first application based on the first identifier in the first service request. If it is found that the first communication device provides the computing service for the first application, the third communication device will send the first service request to the first communication device.

[0128] Accordingly, the first communication device receives the first service request. The first communication device responds to the first service request and determines the first parameters.

[0129] Step S402: The first communication device determines the first parameter based on the maximum computing resources used to calculate the data of the first application and the computing resources already used to calculate the data of the first application.

[0130] It is understood that the first communication device is configured with the maximum computing resources for computing data for the first application. Since the first communication device is centrally responsible for computing tasks for the data of the first application, it may also be providing computing services for the data of the first application to other devices (such as other UEs), thus occupying a portion of computing resources. This portion of computing resources is called the used computing resources for computing data for the first application. Therefore, based on the maximum computing resources corresponding to the first application and the used computing resources, the available computing resources for computing data for the first application can be obtained. The size of this portion of resources can be represented by the first parameter.

[0131] It is understandable that once the application (business) is determined, for example, if it is determined to be for the first application, then the model or algorithm for calculating available computing resources is determined. For example, the available computing resources (i.e., the first parameter) can be obtained by directly subtracting the used computing resources from the maximum computing resources; or the maximum computing resources can be subtracted from the used computing resources to obtain an intermediate parameter, and then the intermediate parameter can be calculated accordingly (such as multiplying by a preset weight) to obtain the available computing resources (i.e., the first parameter); of course, other parameters can also be introduced. In general, the calculation of available computing resources (i.e., the first parameter) requires the maximum computing resources and the used computing resources.

[0132] Optionally, the first parameter includes the remaining number of tokens that can be accommodated, the remaining floating point operations per second (FLOPS), the upper limit of FLOPS and the currently used FLOPS (used to calculate the remaining FLOPS), the FLOPS allocated to the first application and the currently used FLOPS of the first application (used to calculate the remaining FLOPS), the peak computing power of the first application and the current computing power of the first application (e.g., including but not limited to, the Time to First Token of a large language model, the time to generate each token, etc., the remaining number of tokens that can be accommodated can be calculated based on the peak and current values ​​of this indicator), and so on. Of course, it can also be other parameters that can directly or indirectly reflect the size of computing resources.

[0133] In this embodiment of the application, in addition to providing centralized computing services for the first application, the first communication device can also provide centralized computing services for other applications. Maximum computing resources can also be configured for other applications. Other applications also have corresponding models or algorithms for calculating available computing resources (such as the first parameter). The execution process and principle of providing services to other applications are the same as those of the first application, and will not be described in detail hereafter.

[0134] In this embodiment, the first parameter is the first parameter within the estimated reference time period. Several calculation methods are illustrated below:

[0135] For example, based on the maximum computing resources and used computing resources corresponding to the first application recorded over a period of time, the changing trend of available computing resources for computing the data of the first application is analyzed, and then the first parameter within the reference time period is determined based on the changing trend.

[0136] For example, based on the available computing resources for calculating the data of the first application during a historical reference time period (such as 8:00-12:00 the previous day), the available computing resources for calculating the data of the first application during the current reference time period (such as 8:00-12:00 today) are determined, thereby obtaining the first parameter within the reference data segment.

[0137] For example, the first parameter of the reference time period can be predicted by using information such as the reference time period, the maximum computing resources corresponding to the first application, and the computing resources already used as model inputs.

[0138] For example, multiple computation time points are taken in a recent period, and the available computing resources for computing the data of the first application corresponding to each computation time point are calculated (the calculation method has been introduced earlier). Then, the average value of the available computing resources calculated for multiple time points is taken to obtain the first parameter within the reference time period.

[0139] Optionally, the reference time period may be carried in the first service request and thus indicated to the first communication device, or it may be pre-configured to the first communication device so that the first communication device can calculate the first parameter corresponding to the reference time period.

[0140] It is understandable that since the first parameter is for use within a reference time period, the second communication device does not need to request the first parameter again within the reference time period, nor does the network need to actively send the first parameter to the second communication device. This avoids frequent queries, frequent instructions, and frequent interactions. Although a small amount of accuracy is sacrificed (because it is not a real-time first parameter, but only a near-real-time first parameter), the process is greatly simplified and signaling overhead is saved.

[0141] Step S403: The first communication device sends the second instruction information to the third communication device.

[0142] The second indication information includes the first parameter. Optionally, the second indication information also includes the identifier of the first application, which can directly identify the first application or indirectly identify the first application.

[0143] Optionally, this second indication information may also be referred to as computational load indication information or other names.

[0144] Correspondingly, the third communication device receives the second instruction information.

[0145] It is understood that the third communication device can determine, based on the relevant information in the second instruction information (such as the identifier of the first application), that the available computing resources represented by the first parameter in the second instruction information are available computing resources for computing the data of the first application.

[0146] Step S404: The third communication device determines the data packet specifications of the first application based on the second instruction information.

[0147] Specifically, the packet specification is used to limit the size of the data packet sent by the first application. Therefore, this specification can also be referred to as size, dimension, etc., and the unit of measurement can be bits or other units for measuring data size. Optionally, the data packet of the first application consists of a packet header determined according to the transmission protocol used (such as TCP, UDP, IP, PDCP) and a payload carrying the data of the first application, or the data packet of the first application is the payload carrying the data of the first application.

[0148] There are many ways to calculate package specifications. To make it easier to understand, the following example is provided:

[0149] Case 1, determining the data packet specifications of the first application based on the second indication information includes:

[0150] The data package specification of the first application corresponding to the first parameter in the second instruction information is determined according to the stored calculation rules or correspondence.

[0151] For example, the third communication device can pre-store a mapping table (e.g., a correspondence) for the first application. This mapping table includes multiple parameter records and multiple packet specification records, and the correspondence between the multiple parameter records and the multiple packet specification records is marked. Therefore, the packet specification corresponding to the first parameter can be found from this mapping table. It should be noted that in one approach, different applications can share a single mapping table; in another approach, different applications correspond to their own separate mapping tables.

[0152] For example, the third communication device can pre-store a calculation rule (or algorithm model) for the first application. The first parameter can be used as the input parameter of the calculation rule, and the output parameters of the calculation rule include the packet specification. Optionally, the calculation rule can be configured based on experience or obtained by training a model based on historical prior data; there is no specific limitation. It should be noted that in one approach, different applications can share a single calculation rule; in another approach, different applications correspond to their own calculation rules. For example, the first parameter mentioned above is the number of tokens T available for calculating the data of the first application, and the data size of a single token in the first application is S (for example, for Chinese character generation, a single token corresponds to an average of 2 Chinese characters, which, according to UTF-16 encoding, occupies 4 bytes, i.e., the corresponding data size is 4 bytes). Then, the packet specification can be equal to S multiplied by T. Of course, in addition to the number of tokens T and the data size of a single token S, other parameters or algorithms may be used to calculate the packet specification; this is not limited here.

[0153] It should be noted that when determining the package specifications based on the first parameter, regardless of the method used for determination, the first parameter and the package specifications should have a positive correlation (optionally, this positive correlation should be maintained at least within a certain period of the first parameter; beyond that period, the package specifications may be set to a constant value). For example, this positive correlation can be understood as follows: if the determined package specifications are larger, one reason is that the first parameter used to calculate the package specifications is larger.

[0154] It should be noted that if the first parameter is 0, that is, the computing resources configured in the first communication device to calculate the data of the first application are already fully loaded, i.e., the computing resources have been used up, then the first parameter can be 0, and the calculated packet specification can also be 0.

[0155] Case 2: Besides the second communication device having data from the first application that requires the first communication device's assistance in calculation, other communication devices (such as other UEs) also have data from the first application that requires the first communication device's assistance in calculation (these other communication devices are of the same type as the second communication device, for example, both are UEs). Therefore, the third communication device can allocate the resources available on the first communication device for calculating the data from the first application, taking into account the situations of the second and other communication devices. The allocation criteria can be set as needed (e.g., allocation based on channel status, allocation based on priority, etc.), and then determine the packetization specifications of the second and other communication devices based on the allocation. For example, determining the packetization specifications of the data from the first application based on the second indication information includes:

[0156] If the channel status of other communication devices is better than that of the second communication device, the computing resources corresponding to the first parameter in the second indication information are preferentially allocated to the other communication devices, and the packet specification is configured for the other communication devices according to the computing resources allocated to them.

[0157] Then, based on the remaining resources after the computing resources corresponding to the first parameter are allocated to other communication devices, the packetization specification for the first application data of the second communication device is determined. It should be noted that the method of determining the packetization specification based on the remaining resources can refer to the principle of determining the packetization specification based on the first parameter in Case 1 above; therefore, the details will not be repeated. Optionally, if all the computing resources corresponding to the first parameter are allocated to other communication devices with better channel conditions, then the remaining computing resources available for allocation to the second communication device are 0. Therefore, the determined packetization specification is 0, meaning that other communication devices are currently allowed to send the first application data according to a certain specification, but the second communication device is currently not allowed to send the first application data.

[0158] In this embodiment of the application, optionally, when determining the packet specification, in addition to considering the first parameter, the packet header of the data packet can also be considered. For example, if the amount of data that the remaining computing resources of the first communication device can calculate is determined to be data A according to the first parameter, and the amount of data corresponding to the packet header of the data packet is data B, then the data size limited by the packet specification can be data A plus data B.

[0159] Step S405: The third communication device sends the first instruction information to the second communication device.

[0160] The first indication information is used to indicate the data packaging specifications of the first application. Optionally, the first indication information may include an identifier of the first application, indicating that the indicated packaging specifications are for the first application.

[0161] Accordingly, the second communication device receives the first instruction information. Optionally, the first instruction information may also be referred to as communication instruction information.

[0162] Step S406: The second communication device assembles the data of the first application to be sent into packets according to the packet assembly specifications to obtain the first data packet.

[0163] Specifically, the packet specification is used to limit the size of the data packet sent to the first application. Therefore, the data to be sent is packetized according to the packet specification, and the resulting data packet is called the first data packet.

[0164] In one alternative, the packet specification can limit the total data size of the entire data packet; therefore, the data size in the first data packet must be less than or equal to the data size limited by the packet specification.

[0165] In another alternative scheme, the packet specification can limit the size of the service data in the data packet. Therefore, the amount of service data in the first data packet must be less than or equal to the amount of data limited by the packet specification.

[0166] Optionally, if the package specification in the first instruction information is 0, it means that package assembly is not required, and step S406 is not executed; if the package specification in the first instruction information is not 0, then step S406 can be executed.

[0167] Step S407: The second communication device sends the first data packet to the first communication device.

[0168] For example, a second communication device sends a first data packet. After receiving the first data packet, a third communication device sends it back to the first communication device. The third communication device can either transmit the first data packet directly or process it before transmission. This processing does not change the core business data within the first data packet, i.e., it does not change the data related to the first application. For instance, the first data packet might be an IP packet. Both the first and second communication devices can recognize this IP packet. However, for transmission purposes, the third communication device, upon receiving the IP packet, needs to encapsulate it using a specific protocol. Then, it sends the encapsulated data packet to the first communication device. Upon receiving the encapsulated data packet, the first communication device decapsulates it using the same protocol to obtain the original IP packet. Since the core business data in the original IP packet and the encapsulated data packet are the same, both the data packet received and sent by the third communication device are called the first data packet. However, upon closer examination, the two are not entirely identical.

[0169] Accordingly, the first communication device receives the first data packet.

[0170] Step S408: The first communication device performs calculations on the first data packet using available computing resources for calculating data from the first application.

[0171] Step S409: The second communication device locally caches the remaining data in the data of the first application to be sent, excluding the first data packet.

[0172] It is understandable that, given the limitations of the packet size specifications, it is impossible to transmit all the data of the first application in one packet. Therefore, the remaining data outside the first packet is cached for transmission in the next packet.

[0173] In other words, in this embodiment, the buffer only carries a portion of the data from the first application according to the packet assembly specification, namely the first data packet (which can be an IP data packet or a data packet of other protocols). Optionally, if the first data packet does not include a header, then the buffer can be used to store the first data packet plus the header; the remaining part of the data from the first application is cached locally (e.g., stored in other types of storage units besides the buffer). It is understood that the buffer size affects the BSR, thereby affecting air interface scheduling.

[0174] Step S410: The second communication device sends the second data packet.

[0175] Specifically, the second data packet is a data packet obtained by repackaging the remaining data. The second communication device can send the second data packet after the first communication device has released the resources for calculating the data of the first application.

[0176] Optionally, the second communication device sends a second data packet, and after receiving the second data packet, the third communication device sends the second data packet to the first communication device. The third communication device may transmit the second data packet transparently, or it may process the second data packet accordingly before transmitting it. The processing here does not change the core business data in the second data packet, that is, it does not change the data related to the first application.

[0177] Optionally, the first communication device may proactively indicate to the second communication device (directly or indirectly) that computing resources have been released via a notification message.

[0178] Optionally, the second communication device may send (directly or indirectly) a request to the first communication device to request the first communication device to notify the second communication device after releasing computing resources.

[0179] Optionally, the second communication device can re-obtain the packet assembly specification in the same way as the previous process of obtaining the packet assembly specification, and then reassemble the remaining data according to the re-obtained packet assembly specification to obtain the second data packet, and then send the second data packet.

[0180] In one alternative approach, after sending the first data packet, the second communication device sends the second data packet after a first time interval. The method for determining this first time interval is not limited here; alternatively, there may be no explicit time interval, but rather a corresponding trigger condition is set, and the second data packet can be sent when the trigger condition is met.

[0181] Accordingly, the first communication device receives the second data packet.

[0182] Step S411: The first communication device performs calculations on the second data packet using available computing resources for calculating the data of the first application.

[0183] It is understood that both steps S411 and S408 mention the available computing resources for calculating the data of the first application. Since the execution times of steps S411 and S408 are different, the operating state of the first communication device is also different. Therefore, the size of the "available computing resources for calculating the data of the first application" mentioned in steps S411 and S408 may be different.

[0184] Step S412: The first communication device sends the calculation results of the data for the first application.

[0185] In this embodiment, the first communication device first receives a first data packet and performs calculations on it, then receives a second data packet and performs calculations on it. These two (or possibly more) calculations yield a single overall result, which is the result of calculating the data from the first application received from the second communication device. That is, although the data from the first application to be sent is sent separately, the calculation result is still a single, overall result. This overall calculation result is then sent.

[0186] Optionally, the first communication device sends the calculation result, and after receiving the calculation result, the third communication device sends the calculation result to the second communication device. The third communication device may transmit the calculation result transparently, or it may process the calculation result accordingly before transmitting it. The processing here does not change the core business data in the calculation result; for example, it may only encapsulate the transmission protocol to make it easier for the other end to receive.

[0187] Step S413: The second communication device receives the calculation results of the data for the first application.

[0188] It is understandable that after receiving the calculation result of the data from the first application, the second communication device can continue to run the first application based on the calculation result. During the running of the first application, new data may be generated again. The newly generated data of the first application can also be sent to the first communication device for centralized calculation in accordance with the aforementioned process, and then the calculation result is fed back to the second communication device.

[0189] In the embodiments of this application, the steps listed above are only illustrative examples. Some steps may be omitted or replaced with other steps, or new steps may be added, which can also constitute new feasible solutions. In addition, the order of description of the steps above does not represent the order of execution of the steps. The order of execution of the steps may be the same as the order of description or different from the order of description (but the logic must be coherent and reasonable).

[0190] In the embodiments of this application, the technical solutions described in the above steps can be modified to form new feasible technical solutions.

[0191] For example, the second instruction information includes the first parameter. After receiving the second instruction information from the first communication device, the third communication device does not calculate the packet specification based on the second instruction information. Instead, it sends the second instruction information to the second communication device, which then calculates the packet specification based on the first parameter in the second instruction information. That is, the operation related to generating the packet specification in step S404 above is completed by the second communication device. How the second communication device uses the packet specification after obtaining it has been described previously and will not be repeated here.

[0192] For example, after the first communication device generates the first parameter, it continues to generate the packet specification based on the first parameter. The operation of generating the packet specification in step S404 above is completed by the first communication device. After generating the packet specification, the first communication device sends a second instruction message to the third communication device. In this case, the second instruction message does not include the first parameter, but includes the packet specification. After receiving the second instruction message, the third communication device sends the second instruction message to the second communication device. Therefore, the second communication device can obtain the packet specification based on the second instruction message. How the second communication device uses the packet specification after obtaining it has been introduced previously and will not be repeated here.

[0193] For example, in the entire execution process, there is no intermediate node between the first communication device and the second communication device, that is, there is no third communication device. Therefore, the first communication device and the second communication device can communicate directly. For example, the first communication device can directly send the second instruction information to the second communication device. The second instruction information may include the packet specification or include the first parameter used to calculate the packet specification.

[0194] In the method described in Figure 4, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0195] Please refer to Figure 5, which is a flowchart illustrating a communication method provided in an embodiment of this application. This method can be implemented based on the architecture shown in Figure 2, or on other architectures. The method includes, but is not limited to, the following steps:

[0196] Step S501: The fourth communication device sends a second service request to the first communication device.

[0197] Optionally, the fourth communication device is used to query available computing resources for other devices. That is, the actual user of the computing resources is another device, but the fourth communication device can actively trigger the query process for computing resources and then send the query results directly or indirectly to that other device (e.g., the second communication device). The second service request is used to request a query for available computing resources for the first application within a reference duration. Therefore, the second service request may include the reference duration (or time period) and the first identifier. The first identifier has a direct or indirect correspondence with the first application. The first identifier indicates that available computing resources for the first application are to be queried. For example, the first identifier may be one or more of the following: application identifier, device identifier, Quality of Service (QoS) identifier, 5G QoS Identifier (5QI).

[0198] Optionally, the first service request may also include a terminal identifier of the second communication device, indicating that the result of the query through the first service request is for use by the second communication device, or that the second communication device is querying available computing resources for the first application.

[0199] Accordingly, the first communication device receives the second service request. The first communication device responds to the second service request and determines the first parameter within the reference time period.

[0200] Step S502: The first communication device determines the first parameter based on the maximum computing resources used to calculate the data of the first application and the computing resources already used to calculate the data of the first application.

[0201] It is understood that the first communication device is configured with the maximum computing resources for computing data for the first application. Since the first communication device is centrally responsible for computing tasks for the data of the first application, it may also be providing computing tasks for the data of the first application to other devices, thus occupying a portion of computing resources. This portion of computing resources is called the used computing resources for computing data for the first application. Therefore, based on the maximum computing resources corresponding to the first application and the used computing resources, the available computing resources for computing data for the first application can be obtained. The size of this portion of resources can be represented by the first parameter.

[0202] It is understandable that once the application (business) is determined, for example, if it is determined to be for the first application, then the model or algorithm for calculating available computing resources is determined. For example, the available computing resources (i.e., the first parameter) can be obtained by directly subtracting the used computing resources from the maximum computing resources; or the maximum computing resources can be subtracted from the used computing resources to obtain an intermediate parameter, and then the intermediate parameter can be calculated accordingly (such as multiplying by a preset weight) to obtain the available computing resources (i.e., the first parameter); of course, other parameters can also be introduced. In general, the calculation of available computing resources (i.e., the first parameter) requires the maximum computing resources and the used computing resources.

[0203] Optionally, the first parameter may include the remaining number of tokens that can be accommodated and / or the remaining number of floating point operations per second (FLOPS). Of course, it may also be other parameters that can directly or indirectly reflect the size of computing resources.

[0204] In this embodiment of the application, in addition to providing centralized computing services for the first application, the first communication device can also provide centralized computing services for other applications. Maximum computing resources can also be configured for other applications. Other applications also have corresponding models or algorithms for calculating available computing resources (such as the first parameter). The execution process and principle of providing services to other applications are the same as those of the first application, and will not be described in detail hereafter.

[0205] In this embodiment, the first parameter is the first parameter within the estimated reference time period. Several calculation methods are illustrated below:

[0206] For example, based on the maximum computing resources and used computing resources corresponding to the first application recorded over a period of time, the changing trend of available computing resources for computing the data of the first application is analyzed, and then the first parameter within the reference time period is determined based on the changing trend.

[0207] For example, based on the available computing resources for calculating the data of the first application during a historical reference time period (such as 8:00-12:00 the previous day), the available computing resources for calculating the data of the first application during the current reference time period (such as 8:00-12:00 today) are determined, thereby obtaining the first parameter within the reference data segment.

[0208] For example, the first parameter of the reference time period can be predicted by using information such as the reference time period, the maximum computing resources corresponding to the first application, and the computing resources already used as model inputs.

[0209] For example, multiple computation time points are taken in a recent period, and the available computing resources for computing the data of the first application corresponding to each computation time point are calculated (the calculation method has been introduced earlier). Then, the average value of the available computing resources calculated for multiple time points is taken to obtain the first parameter within the reference time period.

[0210] Optionally, the reference time period may be carried in the second service request and thus indicated to the first communication device, or it may be pre-configured to the first communication device so that the first communication device can calculate the first parameter corresponding to the reference time period.

[0211] It is understandable that since the first parameter is for use within a reference time period, the second communication device does not need to request the first parameter again within the reference time period, nor does the network need to actively send the first parameter to the second communication device. This saves frequent queries, frequent instructions, and frequent interactions. Although a small amount of accuracy is sacrificed (because it is not a real-time first parameter, but only a near-real-time first parameter), the process is greatly simplified and signaling overhead is saved.

[0212] Step S503: The first communication device sends the second instruction information to the fourth communication device.

[0213] The second indication information includes a reference time period and a first parameter. Optionally, the second indication information also includes an identifier of the first application, which can directly identify the first application or indirectly identify the first application.

[0214] Optionally, this second indication information may also be referred to as computational load indication information or other names.

[0215] Correspondingly, the fourth communication device receives the second instruction information.

[0216] It is understood that the fourth communication device can determine, based on the relevant information in the second instruction information (such as the identifier of the first application), that the available computing resources represented by the first parameter in the second instruction information are available computing resources for computing the data of the first application, and can know that they are available computing resources for the predicted reference time period.

[0217] Step S504: The fourth communication device sends the second instruction information to the third communication device.

[0218] Correspondingly, the third communication device receives the second instruction information.

[0219] It is understood that the third communication device can determine, based on the relevant information in the second indication information (such as the identifier of the first application), that the available computing resources represented by the first parameter in the second indication information are available computing resources for computing the data of the first application, and can know that the first parameter specifically indicates the available computing resources for the predicted reference time period.

[0220] Step S505: The third communication device determines the packet specifications of the data of the first application based on the second instruction information.

[0221] Specifically, the packet specification is used to limit the data size of the data packets sent to the first application, and optionally, it is used to limit the data size of the data packets sent to the first application within a reference time period. There are many ways to calculate the packet specification; for ease of understanding, an example is given below:

[0222] Case 1, determining the data packet specifications of the first application based on the second indication information includes:

[0223] The data package specification of the first application corresponding to the first parameter in the second instruction information is determined according to the stored calculation rules or correspondence.

[0224] For example, the third communication device can pre-store a mapping table (e.g., a correspondence) for the first application. This mapping table includes multiple parameter records and multiple packet specification records, and the correspondence between the multiple parameter records and the multiple packet specification records is marked. Therefore, the packet specification corresponding to the first parameter can be found from this mapping table. It should be noted that in one approach, different applications can share a single mapping table; in another approach, different applications correspond to their own separate mapping tables.

[0225] For example, the third communication device can pre-store a calculation rule (or algorithm model) for the first application. The first parameter can be used as the input parameter of the calculation rule, and the output parameters of the calculation rule include the packet specification. Optionally, the calculation rule can be configured based on experience or obtained by training a model based on historical prior data; there is no specific limitation. It should be noted that in one approach, different applications can share a single calculation rule; in another approach, different applications correspond to their own calculation rules. For example, the first parameter mentioned above is the number of tokens T available for calculating the data of the first application, and the data size of a single token in the first application is S (for example, for Chinese character generation, a single token corresponds to an average of 2 Chinese characters, which, according to UTF-16 encoding, occupies 4 bytes, i.e., the corresponding data size is 4 bytes). Then, the packet specification can be equal to S multiplied by T. Of course, in addition to the number of tokens T and the data size of a single token S, other parameters or algorithms may be used to calculate the packet specification; this is not limited here.

[0226] It should be noted that when determining the package specifications based on the first parameter, regardless of the method used for determination, the first parameter and the package specifications should have a positive correlation (optionally, this positive correlation should be maintained at least within a certain period of the first parameter; beyond that period, the package specifications may be set to a constant value). For example, this positive correlation can be understood as follows: if the determined package specifications are larger, one reason is that the first parameter used to calculate the package specifications is larger.

[0227] It should be noted that if the first parameter is 0, that is, the computing resources configured in the first communication device to calculate the data of the first application are already fully loaded, i.e., the computing resources have been used up, then the first parameter can be 0, and the calculated packet specification can also be 0.

[0228] Case 2: Besides the second communication device having data from the first application that requires the first communication device's assistance in calculation, other communication devices (such as other UEs) also have data from the first application that requires the first communication device's assistance in calculation (these other communication devices are of the same type as the second communication device, for example, both are UEs). Therefore, the third communication device can allocate the resources available on the first communication device for calculating the data from the first application, taking into account the situations of the second and other communication devices. The allocation criteria can be set as needed (e.g., allocation based on channel status, allocation based on priority, etc.), and then determine the packetization specifications of the second and other communication devices based on the allocation. For example, determining the packetization specifications of the data from the first application based on the second indication information includes:

[0229] If the channel status of other communication devices is better than that of the second communication device, the computing resources corresponding to the first parameter in the second indication information are preferentially allocated to the other communication devices, and the packet specification is configured for the other communication devices according to the computing resources allocated to them.

[0230] Then, based on the remaining resources after the computing resources corresponding to the first parameter are allocated to other communication devices, the packetization specification for the first application data of the second communication device is determined. It should be noted that the method of determining the packetization specification based on the remaining resources can refer to the principle of determining the packetization specification based on the first parameter in Case 1 above; therefore, the details will not be repeated. Optionally, if all the computing resources corresponding to the first parameter are allocated to other communication devices with better channel conditions, then the remaining computing resources available for allocation to the second communication device are 0. Therefore, the determined packetization specification is 0, meaning that other communication devices are currently allowed to send the first application data according to a certain specification, but the second communication device is currently not allowed to send the first application data.

[0231] In this embodiment of the application, optionally, when determining the packet specification, in addition to considering the first parameter, the packet header of the data packet can also be considered. For example, if the amount of data that the remaining computing resources of the first communication device can calculate is determined to be data A according to the first parameter, and the amount of data corresponding to the packet header of the data packet is data B, then the data size limited by the packet specification can be data A plus data B.

[0232] Step S506: The third communication device sends the first instruction information to the second communication device.

[0233] The first indication information is used to indicate the data packaging specifications of the first application. Optionally, the first indication information may include an identifier of the first application, indicating that the indicated packaging specifications are for the first application.

[0234] Accordingly, the second communication device receives the first instruction information. Optionally, the first instruction information may also be referred to as communication instruction information.

[0235] Step S507: The second communication device assembles the data of the first application to be sent into packets according to the packet assembly specifications to obtain the first data packet.

[0236] Specifically, the packetization specification is used to limit the size of the data packets sent to the first application. Therefore, the data to be sent is packetized according to this packetization specification, and the resulting data packet is called the first data packet. For example, packets are packetized according to this packetization specification within a reference time period. Optionally, the aforementioned first indication information may carry the reference time period.

[0237] In one alternative, the packet specification can limit the total data size of the entire data packet; therefore, the data size in the first data packet must be less than or equal to the data size limited by the packet specification.

[0238] In another alternative scheme, the packet specification can limit the size of the service data in the data packet. Therefore, the amount of service data in the first data packet must be less than or equal to the amount of data limited by the packet specification.

[0239] Optionally, if the package specification in the first instruction information is 0, it means that package assembly is not required, and step S507 is not executed; if the package specification in the first instruction information is not 0, then step S507 can be executed.

[0240] Step S508: The second communication device sends the first data packet to the first communication device.

[0241] For example, a second communication device sends a first data packet. After receiving the first data packet, a third communication device sends it back to the first communication device. The third communication device can either transmit the first data packet directly or process it before transmission. This processing does not change the core business data within the first data packet, i.e., it does not change the data related to the first application. For instance, the first data packet might be an IP packet. Both the first and second communication devices can recognize this IP packet. However, for transmission purposes, the third communication device, upon receiving the IP packet, needs to encapsulate it using a specific protocol. Then, it sends the encapsulated data packet to the first communication device. Upon receiving the encapsulated data packet, the first communication device decapsulates it using the same protocol to obtain the original IP packet. Since the core business data in the original IP packet and the encapsulated data packet are the same, both the data packet received and sent by the third communication device are called the first data packet. However, upon closer examination, the two are not entirely identical.

[0242] Accordingly, the first communication device receives the first data packet.

[0243] Step S509: The first communication device performs calculations on the first data packet using available computing resources for calculating data from the first application.

[0244] Step S510: The second communication device locally caches the remaining data in the data of the first application to be sent, excluding the first data packet.

[0245] It is understandable that, given the limitations of the packet size specifications, it is impossible to transmit all the data of the first application in one packet. Therefore, the remaining data outside the first packet is cached for transmission in the next packet.

[0246] In other words, in this embodiment, the buffer only carries a portion of the data from the first application according to the packet assembly specification, namely the first data packet (which can be an IP data packet or a data packet of other protocols). Optionally, if the first data packet does not include a header, then the buffer can be used to store the first data packet plus the header; the remaining part of the data from the first application is cached locally (e.g., stored in other types of storage units besides the buffer). It is understood that the buffer size affects the BSR, thereby affecting air interface scheduling.

[0247] Step S511: The second communication device sends the second data packet.

[0248] Specifically, the second data packet is a data packet obtained by repackaging the remaining data. The second communication device can send the second data packet after the first communication device has released the resources for calculating the data of the first application.

[0249] Optionally, the second communication device sends a second data packet, and after receiving the second data packet, the third communication device sends the second data packet to the first communication device. The third communication device may transmit the second data packet transparently, or it may process the second data packet accordingly before transmitting it. The processing here does not change the core business data in the second data packet, that is, it does not change the data related to the first application.

[0250] Optionally, the first communication device may proactively indicate to the second communication device (directly or indirectly) that computing resources have been released via a notification message.

[0251] Optionally, the second communication device may send (directly or indirectly) a request to the first communication device to request the first communication device to notify the second communication device after releasing computing resources.

[0252] Optionally, the second communication device can re-obtain the packet assembly specification in the same way as the previous process of obtaining the packet assembly specification, and then reassemble the remaining data according to the re-obtained packet assembly specification to obtain the second data packet, and then send the second data packet.

[0253] In one alternative approach, the second communication device sends the first data packet, followed by a first time interval, before sending the second data packet. For example, the first time interval is obtained as follows:

[0254] Accordingly, the first communication device receives the second data packet.

[0255] Step S512: The first communication device performs calculations on the second data packet using the available computing resources for calculating the data of the first application.

[0256] It is understood that both steps S512 and S509 mention the available computing resources for calculating the data of the first application. Since the execution times of steps S512 and S509 are different, the operating state of the first communication device is also different. Therefore, the size of the "available computing resources for calculating the data of the first application" mentioned in steps S512 and S509 may be different.

[0257] Step S513: The first communication device sends the calculation results of the data for the first application.

[0258] In this embodiment, the first communication device first receives a first data packet and performs calculations on it, then receives a second data packet and performs calculations on it. These two (or possibly more) calculations yield a single overall result, which is the result of calculating the data from the first application received from the second communication device. That is, although the data from the first application to be sent is sent separately, the calculation result is still a single, overall result. This overall calculation result is then sent.

[0259] Optionally, the first communication device sends the calculation result, and after receiving the calculation result, the third communication device sends the calculation result to the second communication device. The third communication device may transmit the calculation result transparently, or it may process the calculation result accordingly before transmitting it. The processing here does not change the core business data in the calculation result; for example, it may only encapsulate the transmission protocol to make it easier for the other end to receive.

[0260] Step S514: The second communication device receives the calculation results of the data for the first application.

[0261] It is understandable that after receiving the calculation result of the data from the first application, the second communication device can continue to run the first application based on the calculation result. During the running of the first application, new data may be generated again. The newly generated data of the first application can also be sent to the first communication device for centralized calculation in accordance with the aforementioned process, and then the calculation result is fed back to the second communication device.

[0262] In the embodiments of this application, the steps listed above are only illustrative examples. Some steps may be omitted or replaced with other steps, or new steps may be added, which can also constitute new feasible solutions. In addition, the order of description of the steps above does not represent the order of execution of the steps. The order of execution of the steps may be the same as the order of description or different from the order of description (but the logic must be coherent and reasonable).

[0263] In the embodiments of this application, the technical solutions described in the above steps can be modified to form new feasible technical solutions.

[0264] For example, the second instruction information includes the first parameter. After receiving the second instruction information from the first communication device, the third communication device does not calculate the packet specification based on the second instruction information. Instead, it sends the second instruction information to the second communication device, which then calculates the packet specification based on the first parameter in the second instruction information. That is, the operation related to generating the packet specification in step S505 above is completed by the second communication device. How the second communication device uses the packet specification after obtaining it has been described previously and will not be repeated here.

[0265] For example, the second instruction information includes the first parameter. After receiving the second instruction information from the first communication device, the fourth communication device calculates the packet specification based on the second instruction information. That is, the operation related to generating the packet specification in step S505 above is completed by the fourth communication device. After obtaining the packet specification, the fourth communication device sends it to the third communication device, which then sends it to the second communication device. How the second communication device uses the packet specification has been described previously and will not be repeated here.

[0266] For example, after the first communication device generates the first parameter, it continues to generate the packet specification based on the first parameter. The operation of generating the packet specification in step S505 above is completed by the first communication device. After generating the packet specification, the first communication device sends a second instruction message to the third communication device. In this case, the second instruction message does not include the first parameter, but includes the packet specification. After receiving the second instruction message, the third communication device sends the second instruction message to the second communication device. Therefore, the second communication device can obtain the packet specification based on the second instruction message. How the second communication device uses the packet specification after obtaining it has been introduced previously and will not be repeated here.

[0267] For example, in the entire execution process, there is no intermediate node between the first communication device and the second communication device, that is, there is no third communication device. Therefore, the first communication device and the second communication device can communicate directly. For example, the first communication device can directly send the second instruction information to the second communication device. The second instruction information may include the packet specification or include the first parameter used to calculate the packet specification.

[0268] In the method described in Figure 5, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0269] Please refer to Figure 6, which is a flowchart illustrating a communication method provided in an embodiment of this application. This method can be implemented based on the architecture shown in Figure 2, or on other architectures. The method includes, but is not limited to, the following steps:

[0270] Step S601: The second communication device sends a first service request.

[0271] Optionally, the second communication device (such as the UE) can send the first service request to the third communication device (such as the gNB). After receiving the first service request, the third communication device forwards it to the first communication device (such as the FeIN). The message formats of the first service request received by the third communication device and the first service request sent out can be the same or different, but their substantive functions are the same. They are both used to query available computing resources for the first application. For example, the first service request carries a first identifier, which has a direct or indirect correspondence with the first application. The first identifier indicates that available computing resources for the first application are to be queried. For example, the first identifier can be one or more of the following: application identifier, device identifier, quality of service (QoS) identifier, 5G QoS Identifier (5QI).

[0272] Optionally, the identifiers related to the first application carried in the first service request received and the first service request sent by the third communication device may be the same or different. For example, the identifier related to the first application in the first service request received by the third communication device may be the identifier of the first application, while the identifier related to the first application in the first service request sent by the third communication device may be a device identifier. However, there is an association between the device identifier and the identifier of the first application, so it can also indirectly identify the first application. Similarly, when the third communication device relays messages between the first communication device and the second communication device, it may also involve converting the identifiers related to the first application in the corresponding messages. For example, it may convert the direct identifier of the first application into an indirect identifier before forwarding, or convert the indirect identifier of the first application into a direct identifier before forwarding. Specific messages will not be described again later.

[0273] In one alternative scheme, after receiving the first service request, the third communication device knows that it needs to send the first service request to the first communication device. In another alternative scheme, after receiving the first service request, the third communication device needs to query which node provides the computing service for the first application based on the first application corresponding to the first service request. It can be understood that the third communication device can determine the node that needs to provide the computing service for the first application based on the first identifier in the first service request. If it is found that the first communication device provides the computing service for the first application, the third communication device will send the first service request to the first communication device.

[0274] Accordingly, the first communication device receives the first service request. The first communication device responds to the first service request and determines the first parameters.

[0275] Step S602: The third communication device determines the transmission bandwidth based on the network slice information.

[0276] In this embodiment of the application, slice information can be allocated to the second communication device and the first application according to corresponding rules, for example:

[0277] In one case, the second communication device can determine the corresponding network slice information once it has been identified.

[0278] In another case, once the first application is determined, the corresponding network slice information can be determined.

[0279] In another case, the corresponding network slice information can be determined based on the second communication device and the first application.

[0280] Once the network slice information is determined, the transmission bandwidth can be further determined when the second communication device transmits data for the first application.

[0281] Step S603: The first communication device determines the first parameter based on the maximum computing resources used to calculate the data of the first application and the computing resources already used to calculate the data of the first application.

[0282] It is understood that the first communication device is configured with the maximum computing resources for computing data for the first application. Since the first communication device is centrally responsible for computing tasks for the data of the first application, it may also be providing computing tasks for the data of the first application to other devices, thus occupying a portion of computing resources. This portion of computing resources is called the used computing resources for computing data for the first application. Therefore, based on the maximum computing resources corresponding to the first application and the used computing resources, the available computing resources for computing data for the first application can be obtained. The size of this portion of resources can be represented by the first parameter.

[0283] It is understandable that once the application (business) is determined, for example, if it is determined to be for the first application, then the model or algorithm for calculating available computing resources is determined. For example, the available computing resources (i.e., the first parameter) can be obtained by directly subtracting the used computing resources from the maximum computing resources; or the maximum computing resources can be subtracted from the used computing resources to obtain an intermediate parameter, and then the intermediate parameter can be calculated accordingly (such as multiplying by a preset weight) to obtain the available computing resources (i.e., the first parameter); of course, other parameters can also be introduced. In general, the calculation of available computing resources (i.e., the first parameter) requires the maximum computing resources and the used computing resources.

[0284] Optionally, the first parameter may include the remaining number of tokens that can be accommodated and / or the remaining number of floating point operations per second (FLOPS). Of course, it may also be other parameters that can directly or indirectly reflect the size of computing resources.

[0285] In this embodiment of the application, in addition to providing centralized computing services for the first application, the first communication device can also provide centralized computing services for other applications. Maximum computing resources can also be configured for other applications. Other applications also have corresponding models or algorithms for calculating available computing resources (such as the first parameter). The execution process and principle of providing services to other applications are the same as those of the first application, and will not be described in detail hereafter.

[0286] In this embodiment, the first parameter is the first parameter within the estimated reference time period. Several calculation methods are illustrated below:

[0287] For example, based on the maximum computing resources and used computing resources corresponding to the first application recorded over a period of time, the changing trend of available computing resources for computing the data of the first application is analyzed, and then the first parameter within the reference time period is determined based on the changing trend.

[0288] For example, based on the available computing resources for calculating the data of the first application during a historical reference time period (such as 8:00-12:00 the previous day), the available computing resources for calculating the data of the first application during the current reference time period (such as 8:00-12:00 today) are determined, thereby obtaining the first parameter within the reference data segment.

[0289] For example, the first parameter of the reference time period can be predicted by using information such as the reference time period, the maximum computing resources corresponding to the first application, and the computing resources already used as model inputs.

[0290] For example, multiple computation time points are taken in a recent period, and the available computing resources for computing the data of the first application corresponding to each computation time point are calculated (the calculation method has been introduced earlier). Then, the average value of the available computing resources calculated for multiple time points is taken to obtain the first parameter within the reference time period.

[0291] Optionally, the reference time period may be carried in the first service request and thus indicated to the first communication device, or it may be pre-configured to the first communication device so that the first communication device can calculate the first parameter corresponding to the reference time period.

[0292] It is understandable that since the first parameter is for use within a reference time period, the second communication device does not need to request the first parameter again within the reference time period, nor does the network need to actively send the first parameter to the second communication device. This saves frequent queries, frequent instructions, and frequent interactions. Although a small amount of accuracy is sacrificed (because it is not a real-time first parameter, but only a near-real-time first parameter), the process is greatly simplified and signaling overhead is saved.

[0293] Step S604: The first communication device sends a second instruction message to the third communication device.

[0294] The second indication information includes the first parameter. Optionally, the second indication information also includes the identifier of the first application, which can directly identify the first application or indirectly identify the first application.

[0295] Optionally, this second indication information may also be referred to as computational load indication information or other names.

[0296] Correspondingly, the third communication device receives the second instruction information.

[0297] It is understood that the third communication device can determine, based on the relevant information in the second instruction information (such as the identifier of the first application), that the available computing resources represented by the first parameter in the second instruction information are available computing resources for computing the data of the first application.

[0298] Step S605: The third communication device determines the second packet specification of the data of the first application based on the second instruction information.

[0299] Specifically, the second packet specification is used to limit the size of the data packet sent to the first application. There are many ways to calculate the second packet specification; for ease of understanding, an example is given below:

[0300] Case 1, determining the second package specification of the data of the first application based on the second indication information includes:

[0301] The second package specification of the data of the first application corresponding to the first parameter in the second instruction information is determined according to the stored calculation rules or correspondence.

[0302] For example, the third communication device can pre-store a mapping table (e.g., a correspondence) for the first application. This mapping table includes multiple parameter records and multiple packet specification records, and the correspondence between the multiple parameter records and the multiple packet specification records is marked. Therefore, the second packet specification corresponding to the first parameter can be found from this mapping table. It should be noted that in one approach, different applications can share a single mapping table; in another approach, different applications correspond to their own separate mapping tables.

[0303] For example, the third communication device can pre-store a calculation rule (or algorithm model) for the first application. The first parameter can be used as the input parameter of the calculation rule, and the output parameter of the calculation rule includes the second packet specification. Optionally, the calculation rule can be configured based on experience or obtained by training a model based on historical prior data; there is no specific limitation. It should be noted that in one approach, different applications can share a single calculation rule; in another approach, different applications correspond to their own calculation rules. For example, the first parameter mentioned above is the number of tokens T available for calculating the data of the first application, and the data size of a single token in the first application is S (for example, for Chinese character generation, a single token corresponds to an average of 2 Chinese characters, which, according to UTF-16 encoding, occupies 4 bytes, i.e., the corresponding data size is 4 bytes). Then, the second packet specification can be equal to S multiplied by T. Of course, in addition to the number of tokens T and the data size of a single token S, other parameters or algorithms may be used to calculate the second packet specification; there is no limitation here.

[0304] It should be noted that when determining the specifications of the second package based on the first parameter, regardless of the method used for determination, the first parameter and the specifications of the second package should have a positive correlation (optionally, this positive correlation should be maintained at least within a certain period of the first parameter; beyond that period, the specifications of the second package may be set to a constant value). For example, this positive correlation can be understood as follows: if the determined specifications of the second package are larger, one reason is that the first parameter used to calculate the specifications of the second package is larger.

[0305] It should be noted that if the first parameter is 0, that is, the computing resources configured in the first communication device to calculate the data of the first application are already fully loaded, i.e., the computing resources have been used up, then the first parameter can be 0, and the calculated second packet specification can also be 0.

[0306] Case 2: Besides the second communication device having data from the first application that requires the first communication device's assistance in calculation, other communication devices (such as other UEs) also have data from the first application that requires the first communication device's assistance in calculation (these other communication devices are of the same type as the second communication device, for example, both are UEs). Therefore, the third communication device can allocate the resources available on the first communication device for calculating the data from the first application, taking into account the situations of the second and other communication devices. The allocation criteria can be set as needed (e.g., allocation based on channel status, allocation based on priority, etc.), and then determine the packet specifications of the second and other communication devices based on the allocation. For example, determining the second packet specifications of the data from the first application based on the second indication information includes:

[0307] If the channel status of other communication devices is better than that of the second communication device, the computing resources corresponding to the first parameter in the second indication information are preferentially allocated to the other communication devices, and the packet specification is configured for the other communication devices according to the computing resources allocated to them.

[0308] Then, based on the remaining resources after the computing resources corresponding to the first parameter are allocated to other communication devices, the second packet specification for the first application data of the second communication device is determined. It should be noted that the method of determining the second packet specification based on the remaining resources can refer to the principle of determining the second packet specification based on the first parameter in Case 1 above; therefore, the details will not be repeated. Optionally, if all the computing resources corresponding to the first parameter are allocated to other communication devices with better channel conditions, then the remaining computing resources available for allocation to the second communication device are 0. Therefore, the determined second packet specification is 0, meaning that other communication devices are currently allowed to send the first application data according to a certain specification, but the second communication device is currently not allowed to send the first application data.

[0309] In this embodiment of the application, optionally, when determining the second group of packet specifications, in addition to considering the first parameter, the packet header of the data packet can also be considered. For example, if the amount of data that the remaining computing resources of the first communication device can calculate is determined to be data A according to the first parameter, and the amount of data corresponding to the packet header of the data packet is data B, then the data size limited by the second group of packet specifications can be data A plus data B.

[0310] Step S606: The third communication device determines the communication latency requirement based on the total processing latency requirement and calculation latency of the data of the first application.

[0311] It is understandable that the total processing latency requirement for the data of the first application is a known quantity, usually determined by the Service Level Agreement (SLA). Therefore, once the data object to be processed is determined, the total processing latency requirement corresponding to that data is also determined. Optionally, the total processing latency requirement for the data of the first application can be determined by the third communication device according to pre-configured rules or information (such as the SLA). For example, it can be pre-configured by a third-party application or network management function. Different applications may have different total latency requirements. If a node needs to know the total processing latency requirement for the data of a certain application, it can query the corresponding total latency requirement through the identifier of that application. Therefore, the identifier of the first application can be carried in the information sent to the third communication device so that the third communication device can query the total latency requirement of the first application based on the identifier of the first application. Of course, the total processing latency requirement for the data of the first application can also be carried in the information sent by the second communication device to the third communication device (e.g., determined by the second communication device according to the SLA), or it can be carried in the information sent by other devices to the second communication device.

[0312] Optionally, the computational latency reflects the computational speed of the first communication device and therefore originates from the first communication device. For example, the computational latency can be carried in the second indication information sent by the first communication device to the third communication device. In this embodiment, the computational latency can be an average value within a second time period or one or more instantaneous values. Optionally, when there are multiple instantaneous values, the computational latency for future computations can be predicted using regression or a trained model, based on regression or model training methods. Therefore, regardless of whether the second indication information carries an average value or an instantaneous value, the computational latency for the next computation can ultimately be obtained (which can be understood as the minimum value required for computation, i.e., the shortest time required to complete the computation). Optionally, the third communication device can first send third indication information to the first communication device, which includes the second time period and is used to request the average computational latency within the second time period. Of course, the third indication information can also be combined with other information and sent to the first communication device.

[0313] Furthermore, determining the communication latency requirement based on the total processing latency requirement and computation latency of the data in the first application may include subtracting the computation latency from the total processing latency requirement of the data in the first application to obtain the communication latency requirement; optionally, other parameters and / or algorithms may be introduced for further calculation based on the subtraction to obtain a communication latency that better meets the needs of the scenario, such as introducing weights or introducing compensation amounts.

[0314] Step S607: The third communication device determines the first packet specification based on the communication latency requirements and transmission bandwidth.

[0315] For example, multiplying the communication latency requirement by the transmission bandwidth yields the total amount of data transmitted within that latency period. The size of this total data amount is used as the first packet specification. It should be noted that, in determining the first packet specification, in addition to multiplying the communication latency requirement by the transmission bandwidth, other parameters and / or algorithms can be introduced to obtain a first packet specification that better meets the needs of the scenario.

[0316] Step S608: The third communication device sends the first instruction information to the second communication device.

[0317] The first indication information is used to indicate the packet size of the data of the first application. The packet size is either the first packet size or the second packet size obtained earlier. For example, it is the smaller of the first packet size and the second packet size. It can be understood that the first packet size is calculated from the computing power dimension of the first communication device, and the second packet size is calculated from the latency requirement dimension. In order to ensure that the data packets sent by the second communication device can be calculated normally and meet the latency requirement, the smaller packet size is selected from the first packet size and the second packet size to limit the size of the data packets sent by the second communication device.

[0318] Optionally, the first instruction information may include an identifier of the first application, indicating that the package specification of the instruction is for the first application.

[0319] Accordingly, the second communication device receives the first instruction information. Optionally, the first instruction information may also be referred to as communication instruction information.

[0320] Step S609: The second communication device assembles the data of the first application to be sent into packets according to the packet assembly specifications to obtain the first data packet.

[0321] Specifically, the packet specification is used to limit the size of the data packet sent to the first application. Therefore, the data to be sent is packetized according to the packet specification, and the resulting data packet is called the first data packet.

[0322] In one alternative, the packet specification can limit the total data size of the entire data packet; therefore, the data size in the first data packet must be less than or equal to the data size limited by the packet specification.

[0323] In another alternative scheme, the packet specification can limit the size of the service data in the data packet. Therefore, the amount of service data in the first data packet must be less than or equal to the amount of data limited by the packet specification.

[0324] Optionally, if the package specification in the first instruction information is 0, it means that package assembly is not required, and step S609 is not executed; if the package specification in the first instruction information is not 0, then step S609 can be executed.

[0325] Step S610: The second communication device sends the first data packet to the first communication device.

[0326] For example, a second communication device sends a first data packet. After receiving the first data packet, a third communication device sends it back to the first communication device. The third communication device can either transmit the first data packet directly or process it before transmission. This processing does not change the core business data within the first data packet, i.e., it does not change the data related to the first application. For instance, the first data packet might be an IP packet. Both the first and second communication devices can recognize this IP packet. However, for transmission purposes, the third communication device, upon receiving the IP packet, needs to encapsulate it using a specific protocol. Then, it sends the encapsulated data packet to the first communication device. Upon receiving the encapsulated data packet, the first communication device decapsulates it using the same protocol to obtain the original IP packet. Since the core business data in the original IP packet and the encapsulated data packet are the same, both the data packet received and sent by the third communication device are called the first data packet. However, upon closer examination, the two are not entirely identical.

[0327] Accordingly, the first communication device receives the first data packet.

[0328] Step S611: The first communication device performs calculations on the first data packet using available computing resources for calculating data from the first application.

[0329] Step S612: The second communication device locally caches the remaining data in the data of the first application to be sent, excluding the first data packet.

[0330] It is understandable that, given the limitations of the packet size specifications, it is impossible to transmit all the data of the first application in one packet. Therefore, the remaining data outside the first packet is cached for transmission in the next packet.

[0331] In other words, in this embodiment, the buffer only carries a portion of the data from the first application according to the packet assembly specification, namely the first data packet (which can be an IP data packet or a data packet of other protocols). Optionally, if the first data packet does not include a header, then the buffer can be used to store the first data packet plus the header; the remaining part of the data from the first application is cached locally (e.g., stored in other types of storage units besides the buffer). It is understood that the buffer size affects the BSR, thereby affecting air interface scheduling.

[0332] Step S613: The second communication device sends the second data packet.

[0333] Specifically, the second data packet is a data packet obtained by repackaging the remaining data. The second communication device can send the second data packet after the first communication device has released the resources for calculating the data of the first application.

[0334] Optionally, the second communication device sends a second data packet, and after receiving the second data packet, the third communication device sends the second data packet to the first communication device. The third communication device may transmit the second data packet transparently, or it may process the second data packet accordingly before transmitting it. The processing here does not change the core business data in the second data packet, that is, it does not change the data related to the first application.

[0335] Optionally, the first communication device may proactively indicate to the second communication device (directly or indirectly) that computing resources have been released via a notification message.

[0336] Optionally, the second communication device may send (directly or indirectly) a request to the first communication device to request the first communication device to notify the second communication device after releasing computing resources.

[0337] Optionally, the second communication device can re-obtain the packet assembly specification in the same way as the previous process of obtaining the packet assembly specification, and then reassemble the remaining data according to the re-obtained packet assembly specification to obtain the second data packet, and then send the second data packet.

[0338] In one alternative approach, the second communication device sends the first data packet, followed by a first time interval, before sending the second data packet. For example, the first time interval is obtained as follows:

[0339] Accordingly, the first communication device receives the second data packet.

[0340] Step S614: The first communication device performs calculations on the second data packet using the available computing resources for calculating the data of the first application.

[0341] It is understood that both steps S614 and S611 mention the available computing resources for calculating the data of the first application. Since the execution times of steps S614 and S611 are different, the operating state of the first communication device is also different. Therefore, the size of the "available computing resources for calculating the data of the first application" mentioned in steps S614 and S611 may be different.

[0342] Step S615: The first communication device sends the calculation results of the data for the first application.

[0343] In this embodiment, the first communication device first receives a first data packet and performs calculations on it, then receives a second data packet and performs calculations on it. These two (or possibly more) calculations yield a single overall result, which is the result of calculating the data from the first application received from the second communication device. That is, although the data from the first application to be sent is sent separately, the calculation result is still a single, overall result. This overall calculation result is then sent.

[0344] Optionally, the first communication device sends the calculation result, and after receiving the calculation result, the third communication device sends the calculation result to the second communication device. The third communication device may transmit the calculation result transparently, or it may process the calculation result accordingly before transmitting it. The processing here does not change the core business data in the calculation result; for example, it may only encapsulate the transmission protocol to make it easier for the other end to receive.

[0345] Step S616: The second communication device receives the calculation results of the data for the first application.

[0346] It is understandable that after receiving the calculation result of the data from the first application, the second communication device can continue to run the first application based on the calculation result. During the running of the first application, new data may be generated again. The newly generated data of the first application can also be sent to the first communication device for centralized calculation in accordance with the aforementioned process, and then the calculation result is fed back to the second communication device.

[0347] In the embodiments of this application, the steps listed above are only illustrative examples. Some steps may be omitted or replaced with other steps, or new steps may be added, which can also constitute new feasible solutions. In addition, the order of description of the steps above does not represent the order of execution of the steps. The order of execution of the steps may be the same as the order of description or different from the order of description (but the logic must be coherent and reasonable).

[0348] Step S617: The first communication device sends the first feedback information.

[0349] The first feedback message indicates the effectiveness of the first communication device in calculating the data of the first application. For example, the first communication device previously generated (or predicted) the average calculation latency of the data of the first application in a second time period. Subsequently, the first packet specification is obtained based on the average calculation latency, and the second communication device sends data packets (such as the first data packet) according to the final packet specification (the first packet specification or the second packet specification). Then, for this series of processes, the effectiveness of the series of processes can be evaluated by corresponding indicators or rules (such as the latency requirements of SLA), for example, whether the preset target is achieved or not.

[0350] Correspondingly, the third communication device receives the first feedback information.

[0351] Step S618: The third communication device updates the second time period based on the first feedback information.

[0352] The second time period used in calculating the average computation latency may affect the data transmission and computation performance of the first application. Therefore, if the computation performance of the first application's data does not meet the preset target, the third communication device updates the second time period. The updated second time period is used for the next average computation latency calculation; that is, the average computation latency calculated subsequently is the average computation latency within the updated second time period. If the computation performance of the first application's data meets the preset target, the second time period does not need to be updated.

[0353] In this embodiment of the application, optionally, the communication time of the second data packet can be paralleled with the calculation time of the first data packet, thereby ensuring end-to-end service latency by ensuring the communication time and calculation time of the first packet.

[0354] In the embodiments of this application, the technical solutions described in the above steps can be modified to form new feasible technical solutions.

[0355] For example, the second instruction information includes a first parameter. After receiving the second instruction information from the first communication device, the third communication device does not calculate the second packet specification based on the second instruction information. Instead, it sends the second instruction information to the second communication device, which then calculates the second packet specification based on the first parameter in the second instruction information. That is, the operation related to generating the second packet specification in step S605 above is completed by the second communication device. How the second communication device uses the second packet specification after obtaining it has been described previously and will not be repeated here.

[0356] For example, after the first communication device generates the first parameter, it continues to generate the second packet specification based on the first parameter. The operation of generating the second packet specification in step S605 is completed by the first communication device. After generating the second packet specification, the first communication device sends a second indication message to the third communication device. In this case, the second indication message does not include the first parameter, but includes the second packet specification. After receiving the second indication message, the third communication device sends the second indication message to the second communication device. Therefore, the second communication device can obtain the second packet specification based on the second indication message. How the second communication device uses the second packet specification after obtaining it has been described previously and will not be repeated here.

[0357] For example, in the entire execution process, there is no intermediate node between the first communication device and the second communication device, that is, there is no third communication device. Therefore, the first communication device and the second communication device can communicate directly. For example, the first communication device can directly send the second instruction information to the second communication device. The second instruction information may include the second packet specification or include the first parameter used to calculate the second packet specification.

[0358] In addition, the idea of ​​introducing a first package specification and comparing it with a second package specification, selecting the smaller package specification as the final package specification, can be combined with the embodiment shown in Figure 5 to form a new optional solution.

[0359] In the method described in Figure 6, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0360] The following describes the communication device provided in the embodiments of this application.

[0361] This application divides the communication device into functional modules according to the above method embodiments. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing module. The integrated modules can be implemented in hardware or as software functional modules. It should be noted that the module division in this application is illustrative and only represents one logical functional division; other division methods may be used in actual implementation. The communication device of the embodiments of this application will be described in detail below with reference to Figures 7 to 9.

[0362] Figure 7 is a schematic diagram of a communication device provided in an embodiment of this application. As shown in Figure 7, the communication device includes a processing module 1801 and a transceiver module 1802. The transceiver module 1802 can implement corresponding communication functions, and the processing module 1801 is used for data processing. The transceiver module 1802 can also be referred to as an interface, a communication interface, or a communication module, etc.

[0363] In some embodiments of this application, the communication device can be used to perform the actions performed by the second communication device in the above method embodiments, such as the second communication device being the terminal itself or a chip or functional module configurable in the terminal. In still other embodiments of this application, the communication device can be used to perform the actions performed by the third communication device in the above method embodiments, such as the third communication device being the network device itself or a chip or functional module configurable in the network device. In still other embodiments of this application, the communication device can be used to perform the actions performed by the first communication device in the above method embodiments, such as the first communication device being the computing node itself or a chip or functional module configurable in the computing node. In still other embodiments of this application, the communication device can be used to perform the actions performed by the fourth communication device in the above method embodiments, such as the fourth communication device being the OAM itself or a chip or functional module configurable in the OAM. Specifically, the transceiver module 1802 is used to perform the transceiver-related operations in the above method embodiments, and the processing module 1801 is used to perform the processing-related operations in the above method embodiments. The processing module 1801 can perform the corresponding operations by calling a computer program or by performing the corresponding operations through corresponding hardware circuits. The transceiver module 1802 can perform transceiver operations independently, or it can perform corresponding transceiver operations under the control of the processing module 1801.

[0364] For example, the communication device shown in Figure 7 can be a second communication device (such as a terminal or a device in a terminal, such as a chip). The processing module 1801 and the transceiver module 1802 in this communication device can respectively perform the following operations:

[0365] The transceiver module 1802 is used to receive first indication information, wherein the first indication information is used to indicate the packet specifications of the data of the first application, and the data of the first application is remotely computed by a first communication device that has deployed a large model;

[0366] The processing module 1801 is used to assemble the data of the first application to be sent into packets according to the packet assembly specification to obtain a first data packet; optionally, the data size of the first data packet is less than or equal to the packet assembly specification.

[0367] The transceiver module 1802 is used to send the first data packet to the first communication device;

[0368] The processing module 1801 is used to locally cache the remaining data in the data of the first application to be sent, excluding the first data packet;

[0369] The transceiver module 1802 is used to receive the calculation results of the data for the first application.

[0370] In the above method, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, thus preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0371] In one possible implementation, the first indication information includes the package specification.

[0372] In one possible implementation, the first indication information includes a first parameter, which represents the available computing resources for computing on the data of the first application. The first parameter is used to calculate the package specification. The processing module 1801 is also used to determine the package specification based on the first parameter, which is equivalent to the first indication information indirectly indicating the package specification.

[0373] In another possible implementation, the method further includes sending a first service request, wherein the first service request is used to query available computing resources for the first application.

[0374] In another possible implementation, the first service request includes a first identifier, which has a direct or indirect correspondence with the first application.

[0375] In another possible implementation, the first parameter includes the remaining number of tokens that can be held and / or the remaining number of floating-point operations per second (FLOPS).

[0376] In yet another possible implementation, the method is applied to a second communication device, wherein, in receiving the first indication information, the transceiver module 1802 is specifically used for:

[0377] Receive first instruction information from a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

[0378] In another possible implementation, the first indication information is further used to indicate a first time interval, and the transceiver module 1802 is further used to:

[0379] A second data packet is sent after the first time interval, the second data packet being a data packet obtained from the remaining data packets.

[0380] Reusing Figure 7, in some other embodiments of this application, exemplaryly, the communication device shown in Figure 7 can be a third communication device (such as a network device or a device within a network device), and the processing module 1801 and transceiver module 1802 in this communication device can respectively perform the following operations:

[0381] The transceiver module 1802 is used to receive second indication information, wherein the second indication information includes a first parameter; the first parameter represents the available computing resources for computing on the data of the first application, the data of the first application being remotely computing supported by a first communication device that has deployed a large model;

[0382] Processing module 1801 is used to determine the data packaging specifications of the first application based on the second indication information;

[0383] The transceiver module 1802 is used to send first indication information to the second communication device, wherein the first indication information is used to indicate the packet specification of the data of the first application, and the packet specification is used to limit the data size of the data packet sent by the first application.

[0384] In the above method, network communication and computing functions work together to guide communication transmission based on the available computing resources in the first communication device. This ensures that the data of the first application transmitted by the second communication device is within the computing power of the first communication device, thus preventing the data of the first application sent by the second communication device from exceeding the computing power of the first communication device and triggering the access control of the first communication device. Since the access control is not triggered, the data sent by the second communication device to the first communication device will not be discarded, thus avoiding the waste of communication resources.

[0385] In one possible implementation:

[0386] The transceiver module 1802 is also configured to receive a first data packet from a second communication device, wherein the first data packet is a data packet obtained by assembling the data of the first application according to the packet assembly specification;

[0387] The transceiver module 1802 is also used to send the first data packet to the first communication device.

[0388] In another possible implementation, regarding the determination of the data packet specifications of the first application based on the second indication information, the processing module 1801 is specifically used for:

[0389] The data package specification of the first application corresponding to the first parameter in the second indication information is determined according to the stored calculation rules or correspondence.

[0390] In another possible implementation, regarding the determination of the data packet specifications of the first application based on the second indication information, the processing module 1801 is specifically used for:

[0391] If the channel status of other communication devices is better than that of the second communication device, then the computing resources corresponding to the first parameter in the second indication information will be preferentially allocated to the other communication devices, wherein the other communication devices and the second communication device are of the same type.

[0392] Based on the remaining resources after the computing resources corresponding to the first parameter are allocated, the packet specification of the data for the first application of the second communication device is determined.

[0393] In yet another possible implementation:

[0394] The transceiver module 1802 is also configured to receive a first service request from the second communication device, wherein the first service request is used to query available computing resources for the first application;

[0395] The transceiver module 1802 is also used to send the first service request to the first communication device.

[0396] In yet another possible implementation, regarding the receiving of the second indication information, the transceiver module 1802 is specifically used for:

[0397] Receive second instruction information from the first communication device.

[0398] In another possible implementation, the method is applied to a third communication device; the transceiver module 1802, in the invention for receiving the second indication information, is specifically used for:

[0399] Receive a second instruction message from a fourth communication device, wherein the fourth communication device is a management function node of the third communication device.

[0400] In another possible implementation, the first service request includes a first identifier, which has a direct or indirect correspondence with the first application.

[0401] In another possible implementation, the first parameter includes the remaining number of tokens that can be held and / or the remaining number of floating-point operations per second (FLOPS).

[0402] In another possible implementation, the second indication information is further used to indicate the average computational latency within the second time period. The invention for determining the data packet specifications of the first application based on the second indication information, specifically, the processing module 1801 is used for:

[0403] The communication latency requirement is determined based on the total processing latency requirement of the data from the first application and the average calculation latency.

[0404] The specifications of the first packet group are determined based on the communication latency requirements;

[0405] The specifications of the second group of packages are determined based on the second instruction information;

[0406] Among them, the smaller of the first package specification and the second package specification is used as the package specification of the data of the first application.

[0407] In yet another possible implementation:

[0408] The transceiver module 1802 is further configured to send a third indication information to the first communication device, wherein the third indication information includes the second time period, and the third indication information is used to request the acquisition of the average calculation delay within the second time period.

[0409] In conjunction with the second aspect or any of the above-described possible implementations of the second aspect, another possible implementation of the second aspect further includes:

[0410] Receive a first feedback message from the first communication device, wherein the first feedback message indicates the effect of the first communication device in calculating the data of the first application;

[0411] If the effect of calculating the data of the first application does not meet the preset target, the second time period is updated, and the updated second time period is used for the next calculation of the average calculation latency.

[0412] Reusing Figure 7, in some other embodiments of this application, exemplaryly, the communication device shown in Figure 7 can be a first communication device (such as a computing node or a device in a computing node), and the processing module 1801 and transceiver module 1802 in this communication device can respectively perform the following operations:

[0413] The processing module 1801 is used to determine a first parameter based on the maximum computing resources used to compute data of the first application and the computing resources already used to compute data of the first application, wherein the first parameter represents the available computing resources used to compute data of the first application.

[0414] The transceiver module 1802 is used to send second indication information, wherein the second indication information includes the first parameter or the packet specification of the data of the first application, the first parameter is used to calculate the packet specification, and the packet specification is used to limit the data size of the data packet of the first application; the first communication device is used to provide remote calculation support for the data of the first application to the second communication device through the deployed large model.

[0415] In one possible implementation, the first parameter includes the remaining number of tokens that can be held and / or the remaining number of floating-point operations per second (FLOPS).

[0416] In another possible implementation, the first parameter specifically includes the average remaining number of tokens that can be accommodated within the reference duration and / or the remaining number of floating-point operations per second (FLOPS).

[0417] In another possible implementation of the third aspect, combining the third aspect or any of the above possible implementations:

[0418] The transceiver module 1802 is also configured to receive a first service request from the third communication device, wherein the first service request is used to query available computing resources for the first application.

[0419] In yet another possible implementation:

[0420] The transceiver module 1802 is also configured to receive a second service request from a fourth communication device, wherein the second service request is used to query the available computing resources for the first application within a reference time period.

[0421] In yet another possible implementation, the transceiver module is further configured to: send the second instruction information.

[0422] Send a second instruction message to the fourth communication device.

[0423] In yet another possible implementation, regarding the transmission of the second indication information, the transceiver module is further configured to:

[0424] Send a second instruction message to a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

[0425] Reusing Figure 7, in some other embodiments of this application, exemplaryly, the communication device shown in Figure 7 can be a fourth communication device (such as OAM or a device in OAM), and the processing module 1801 and transceiver module 1802 in this communication device can respectively perform the following operations:

[0426] The transceiver module 1802 is used to send a second service request to the first communication device, wherein the second service request is used to request the available computing resources for the first application within a reference time period;

[0427] The transceiver module 1802 is further configured to receive second indication information from the first communication device, wherein the second indication information includes the first parameter or the packet specification of the data of the first application, the first parameter representing the available computing resources for computing the data of the first application, the first parameter being used to calculate the packet specification, and the packet specification being used to limit the data size of the data packet to be sent to the first application; the first communication device is configured to provide remote computing support for the data of the first application to the second communication device through a deployed large model;

[0428] The transceiver module 1802 is also used to send the second indication information to a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

[0429] In one possible implementation, the first parameter specifically includes the average remaining number of tokens that can be accommodated within the reference duration and / or the remaining number of floating-point operations per second (FLOPS).

[0430] The specific descriptions of the transceiver module and processing module shown in the above embodiments are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the above method embodiments, which will not be described in detail here.

[0431] The communication device according to the embodiments of this application has been described above. The possible product forms of the communication device are described below. Any product possessing the functions of the communication device described in FIG7 above falls within the protection scope of the embodiments of this application.

[0432] The following description is merely an example and does not limit the product form of the communication device in the embodiments of this application to this.

[0433] In one possible implementation, in the communication device shown in FIG7, the processing module 1801 can be one or more processors, and the transceiver module 1802 can be a transceiver, or the transceiver module 1802 can also be a transmitting module and a receiving module. The transmitting module can be a transmitter, and the receiving module can be a receiver. The transmitting module and the receiving module are integrated into one device, such as a transceiver. In the embodiments of this application, the processor and the transceiver can be coupled, etc., and the connection method of the processor and the transceiver is not limited in the embodiments of this application. In the process of executing the above method, the process of sending information in the above method can be the process of the processor outputting the above information. When outputting the above information, the processor outputs the above information to the transceiver so that the transceiver can transmit it. After the above information is output by the processor, it may need to undergo other processing before reaching the transceiver. Similarly, the process of receiving information in the above method can be the process of the processor receiving the input above information. When the processor receives the input information, the transceiver receives the above information and inputs it into the processor. Furthermore, after the transceiver receives the above information, the above information may need to undergo other processing before being input into the processor.

[0434] As shown in Figure 8, the communication device 190 includes one or more processors 1920 and transceivers 1910. Exemplarily, the transceiver 1910 is used to perform the functions or steps implemented by the transceiver module 1802 shown in Figure 7, and the processor 1920 is used to perform the functions or steps implemented by the processing module 1801 shown in Figure 7. Detailed descriptions of the processor 1920 and transceiver 1910 can be found in Figure 7 or the method embodiments shown above, and will not be elaborated further here.

[0435] The descriptions of the relevant steps and information in the above embodiments can be found in the descriptions of the method embodiments above, and will not be detailed here.

[0436] In various implementations of the communication device shown in Figure 8, the transceiver may include a receiver for performing a receiving function (or operation) and a transmitter for performing a transmitting function (or operation). The transceiver is also used to communicate with other devices / appliances via a transmission medium.

[0437] Optionally, the communication device 190 may further include one or more memories 1930 for storing program instructions and / or data. The memory 1930 is coupled to the processor 1920. The coupling in this embodiment is an indirect coupling or communication connection between devices, units, or modules, and can be electrical, mechanical, or other forms, used for information exchange between devices, units, or modules. The processor 1920 may operate in conjunction with the memory 1930. The processor 1920 may execute program instructions stored in the memory 1930. Optionally, at least one of the above-mentioned memories may be included in the processor.

[0438] This embodiment does not limit the specific connection medium between the transceiver 1910, processor 1920, and memory 1930. In Figure 8, the memory 1930, processor 1920, and transceiver 1910 are connected via a bus 1940, indicated by a thick line. The connection methods between other components are merely illustrative and not intended to be limiting. The bus can be categorized as an address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used in Figure 8, but this does not imply that there is only one bus or one type of bus.

[0439] In the embodiments of this application, the processor may be 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, discrete gate or transistor logic devices, discrete hardware components, etc., and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as being executed by a hardware processor, or being executed by a combination of hardware and software modules within the processor.

[0440] In this application embodiment, the memory may include, but is not limited to, non-volatile memory such as hard disk drive (HDD) or solid-state drive (SSD), random access memory (RAM), erasable programmable read-only memory (EPROM), read-only memory (ROM), or compact disc read-only memory (CD-ROM), etc. Memory is any storage medium capable of carrying or storing program code having instruction or data structure forms, and capable of being read and / or written by a computer (such as the communication device shown in this application), but is not limited to this. The memory in this application embodiment may also be a circuit or any other device capable of implementing storage functions, used to store program instructions and / or data.

[0441] Processor 1920 is primarily used for processing communication protocols and data, controlling the entire communication device, executing software programs, and processing software program data. Memory 1930 is primarily used for storing software programs and data. Transceiver 1910 may include control circuitry and an antenna. The control circuitry is primarily used for converting baseband signals to radio frequency signals and processing radio frequency signals. The antenna is primarily used for transmitting and receiving radio frequency signals in the form of electromagnetic waves. Input / output devices, such as touchscreens, displays, and keyboards, are primarily used for receiving user input data and outputting data to the user.

[0442] When the communication device is powered on, the processor 1920 can read the software program in the memory 1930, interpret and execute the instructions of the software program, and process the data of the software program. When data needs to be transmitted wirelessly, the processor 1920 performs baseband processing on the data to be transmitted and outputs the baseband signal to the radio frequency (RF) circuit. The RF circuit processes the baseband signal and transmits the RF signal outward in the form of electromagnetic waves through the antenna. When data is sent to the communication device, the RF circuit receives the RF signal through the antenna, converts the RF signal into a baseband signal, and outputs the baseband signal to the processor 1920. The processor 1920 converts the baseband signal back into data and processes the data.

[0443] In another implementation, the radio frequency circuitry and antenna can be set up independently of the processor performing baseband processing. For example, in a distributed scenario, the radio frequency circuitry and antenna can be arranged remotely, independent of the communication device.

[0444] The communication device shown in this application embodiment may also have more components than those in Figure 8, and this application embodiment does not limit this. The methods executed by the processor and transceiver shown above are only examples, and the specific steps executed by the processor and transceiver can be referred to the methods described above.

[0445] In another possible implementation, in the communication device shown in Figure 7, the processing module 1801 can be one or more logic circuits, and the transceiver module 1802 can be an input / output interface, or a communication interface, or an interface circuit, or an interface, etc. Alternatively, the transceiver module 1802 can also be a transmitting module and a receiving module. The transmitting module can be an output interface, and the receiving module can be an input interface. The transmitting module and the receiving module are integrated into one module, such as an input / output interface. As shown in Figure 9, the communication device shown in Figure 9 includes a logic circuit 2001 and an interface 2002. That is, the processing module 1801 can be implemented using the logic circuit 2001, and the transceiver module 1802 can be implemented using the interface 2002. The logic circuit 2001 can be a chip, a processing circuit, an integrated circuit, or a system-on-a-chip (SoC) chip, etc., and the interface 2002 can be a communication interface, an input / output interface, pins, etc. For example, Figure 9 illustrates the communication device as a chip, which includes the logic circuit 2001 and the interface 2002.

[0446] In this embodiment, the logic circuit and the interface can also be coupled to each other. The specific connection method of the logic circuit and the interface is not limited in this embodiment. For example, the logic circuit 2001 can be used to execute the functions or steps implemented by the processing module 1801 shown in FIG. 7, and the interface 2002 can be used to execute the functions or steps implemented by the transceiver module 1802 shown in FIG. 7. For a detailed description of the logic circuit 2001 and the interface 2002, please refer to FIG. 7 or the method embodiment shown above, which will not be detailed here.

[0447] The above description of the communication device is only an example. For a detailed description of the communication device shown in Figure 9, please refer to the above method embodiment or Figure 7 or Figure 8. It will not be described in detail here.

[0448] The communication device shown in the embodiments of this application can implement the method provided in the embodiments of this application in hardware form, or it can implement the method provided in the embodiments of this application in software form, etc., and the embodiments of this application do not limit it in this way.

[0449] The descriptions of relevant steps and information in the above embodiments can be found in the method embodiments described above, and will not be detailed here. For the specific implementation methods of the embodiments shown in Figure 9, please also refer to the above embodiments, which will not be detailed here.

[0450] This application also provides a communication system, which includes one or more of a first communication device, a second communication device, a third communication device, and a fourth communication device. These communication devices can interact with each other to perform all or part of the steps in any of the foregoing method embodiments.

[0451] In addition, this application also provides a computer program for implementing the operations and / or processes performed by various communication devices in the method provided in this application.

[0452] This application also provides a computer-readable storage medium storing computer code that, when executed on a computer, causes the computer to perform the operations and / or processes performed by various communication devices in the methods provided in this application.

[0453] This application also provides a computer program product comprising computer code or a computer program that, when run on a computer, causes the operations and / or processes performed by various entities in the method provided in this application to be executed.

[0454] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, or it may be an electrical, mechanical, or other form of connection.

[0455] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the technical effects of the solutions provided in the embodiments of this application.

[0456] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0457] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned readable storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0458] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A communication method, characterized in that, include: Receive first indication information, wherein the first indication information is used to indicate the data packet specifications of the first application, and the data of the first application is remotely computed by a first communication device that has deployed a large model; The data of the first application to be sent is packaged according to the packetization specification to obtain the first data packet; Send the first data packet to the first communication device; The remaining data in the data of the first application to be sent, excluding the first data packet, is cached locally; Receive the calculation results for the data of the first application.

2. The method according to claim 1, characterized in that, The first indication information includes the package specifications.

3. The method according to claim 1, characterized in that, The first indication information includes a first parameter, which represents the available computing resources for calculating the data of the first application. The first parameter is used to calculate the package specification. The method further includes: The package specifications are determined based on the first parameter.

4. The method according to any one of claims 1-3, characterized in that, Also includes: Send a first service request, wherein the first service request is used to query available computing resources for the first application.

5. The method according to any one of claims 1-4, characterized in that, The first service request includes a first identifier, which has a direct or indirect correspondence with the first application.

6. The method according to any one of claims 1-5, characterized in that, The first parameter includes the remaining number of tokens that can be held and / or the remaining number of floating-point operations per second (FLOPS).

7. The method according to any one of claims 1-6, characterized in that, The method is applied to a second communication device, wherein receiving the first indication information includes: Receive first instruction information from a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

8. The method according to any one of claims 1-7, characterized in that, The first indication information is also used to indicate a first time interval, and the method further includes: A second data packet is sent after the first time interval, the second data packet being a data packet obtained from the remaining data packets.

9. A communication method, characterized in that, include: Receive second instruction information, wherein the second instruction information includes a first parameter; the first parameter represents available computing resources for computing on data of the first application, the data of the first application being remotely computed by a first communication device deployed with a large model; The data package specifications of the first application are determined based on the second instruction information; Send a first indication message to a second communication device, wherein the first indication message is used to indicate the packet size of the data of the first application, and the packet size is used to limit the data size of the data packets sent to the first application.

10. The method according to claim 9, characterized in that, Also includes: Receive a first data packet from a second communication device, wherein the first data packet is a data packet obtained by assembling the data of the first application according to the packet specification; The first data packet is sent to the first communication device.

11. The method according to claim 9 or 10, characterized in that, Determining the data packet specifications of the first application based on the second indication information includes: The data package specification of the first application corresponding to the first parameter in the second indication information is determined according to the stored calculation rules or correspondence.

12. The method according to claim 9 or 10, characterized in that, Determining the data packet specifications of the first application based on the second indication information includes: If the channel status of other communication devices is better than that of the second communication device, then the computing resources corresponding to the first parameter in the second indication information will be preferentially allocated to the other communication devices, wherein the other communication devices and the second communication device are of the same type. Based on the remaining resources after the computing resources corresponding to the first parameter are allocated, the packet specification of the data for the first application of the second communication device is determined.

13. The method according to any one of claims 9-12, characterized in that, Also includes: Receive a first service request from the second communication device, wherein the first service request is used to query available computing resources for the first application; Send the first service request to the first communication device.

14. The method according to claim 13, characterized in that, The receipt of the second indication information includes: Receive second instruction information from the first communication device.

15. The method according to any one of claims 9-12, characterized in that, Applied to a third communication device; the receiving of the second indication information includes: Receive a second instruction message from a fourth communication device, wherein the fourth communication device is a management function node of the third communication device.

16. The method according to any one of claims 9-15, characterized in that, The first service request includes a first identifier, which has a direct or indirect correspondence with the first application.

17. The method according to any one of claims 9-16, characterized in that, The first parameter includes the remaining number of tokens that can be held and / or the remaining number of floating-point operations per second (FLOPS).

18. The method according to any one of claims 9-17, characterized in that, The second indication information is also used to indicate the average computation latency within the second time period. Determining the data packet specifications of the first application based on the second indication information includes: The communication latency requirement is determined based on the total processing latency requirement of the data from the first application and the average calculation latency. The specifications of the first packet group are determined based on the communication latency requirements; The specifications of the second group of packages are determined based on the second instruction information; Among them, the smaller of the first package specification and the second package specification is used as the package specification of the data of the first application.

19. The method according to claim 18, characterized in that, Also includes: Send a third indication message to the first communication device, wherein the third indication message includes the second time period, and the third indication message is used to request the average calculation delay within the second time period.

20. The method according to claim 18 or 19, characterized in that, Also includes: Receive a first feedback message from the first communication device, wherein the first feedback message indicates the effect of the first communication device in calculating the data of the first application; If the effect of calculating the data of the first application does not meet the preset target, the second time period is updated, and the updated second time period is used for the next calculation of the average calculation latency.

21. A communication method, characterized in that, Applied to a first communication device, the method includes: A first parameter is determined based on the maximum computing resources used to compute data from the first application and the computing resources already used to compute data from the first application, wherein the first parameter represents the available computing resources used to compute data from the first application. Send a second instruction message, wherein the second instruction message includes the first parameter or the packet specification of the data of the first application, the first parameter is used to calculate the packet specification, and the packet specification is used to limit the data size of the data packet of the first application; the first communication device is used to provide remote calculation support for the data of the first application to the second communication device through the deployed large model.

22. The method according to claim 21, characterized in that, Also includes: A first service request is received from the third communication device, wherein the first service request is used to query available computing resources for the first application.

23. The method according to claim 21 or 22, characterized in that, Also includes: A second service request is received from a fourth communication device, wherein the second service request is used to query the available computing resources for the first application within a reference time period.

24. The method according to any one of claims 21-23, characterized in that, The sending of the second instruction information includes: Send a second instruction message to the fourth communication device.

25. The method according to any one of claims 21-24, characterized in that, The sending of the second instruction information includes: Send a second instruction message to a third communication device, wherein the third communication device is a communication node between the first communication device and the second communication device.

26. A communication device, characterized in that, The communication device includes a module for performing the method as described in any one of claims 1-25; or, the communication device includes a processor for performing the method as described in any one of claims 1-25.

27. A communication device, characterized in that, The communication device includes a module for performing the method as described in any one of claims 1-25; or, the communication device includes a processor for performing the method as described in any one of claims 1-25.

28. A communication device, characterized in that, Includes logic circuits and interfaces, wherein the logic circuits and interfaces are coupled; The interface is used for inputting and / or outputting information, and the logic circuit is used for performing the method as described in any one of claims 1-25.

29. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program, which, when executed, performs the method as described in any one of claims 1-25.

30. A communication system, characterized in that, The device includes a first communication device, a second communication device, and a third communication device, wherein the second communication device is configured to perform the method as described in any one of claims 1-8, the third communication device is configured to perform the method as described in any one of claims 9-20, and the first communication device is configured to perform the method as described in any one of claims 21-25.

Images