Communication method and apparatus

Abstract

A communication method and apparatus, which relate to the technical field of communications, and can implement collaborative training of a first model and a second model, and improve the generalization capability of the first model. The method comprises: a terminal device sending a local data set to a first device, receiving a first result from the first device, sending a second result to the first device, and receiving a third result from the first device; and sending an updated local second model to a second device on the basis of the third result, wherein the first result is a result obtained after forward propagation of the local data set in a first parameter set of the first model; the second result is a result obtained after forward propagation of the first result in an initialized local second model; the third result is a result obtained after back propagation of a fourth result in a third parameter set of the first model; the fourth result is a result obtained after forward propagation of the second result and a fifth result in the third parameter set of the first model; and the fifth result is a result obtained after forward propagation of the first result in a second parameter set of the first model.

Description

Communication method and apparatus

[0001] The present application claims priority from the Chinese patent application No. 202410687773.5 filed on May 29, 2024, and entitled "Communication method and apparatus", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

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

[0003] In order to meet the needs of rapid development of the network, a large pre-trained model (LPM) will become one of the key technologies for network development. By training part of the parameters of the LPM, the tasks of the terminal device can be more efficiently met. Among them, part of the parameters of the LPM can be trained through a low-rank adaptation of large language models (LoRA) model.

[0004] The fine-tuning operation for downstream users only involves adjusting and saving a small part of trainable parameters locally, and these fine-tuning improvements will not be fed back to the global version of the LPM, and will not affect the main structure and parameters of the LPM on the server, resulting in that the generalization ability of the LPM itself is not enhanced. SUMMARY

[0005] The present application provides a communication method and apparatus, which can perform collaborative training on a first model (such as an LPM model) and a second model (such as a LoRA model), and enhance the generalization ability of the first model.

[0006] In a first aspect, a communication method is provided, which can be performed by a terminal device. The terminal device can refer to the terminal device itself, a component (e.g., a processor, a chip, or a chip system) in the terminal device, or a logic module or software capable of implementing all or part of the functions of the terminal device. The method includes: sending a local data set to a first device; receiving a first result from the first device; sending a second result to the first device according to the first result; receiving a third result from the first device; and sending an updated local second model to a second device. The first result is a result of forward propagation of the local data set in a first parameter set of a first model; the second result is a result of forward propagation of the first result in an initialized local second model; the third result is a result of back propagation of a fourth result in a third parameter set of the first model; the fourth result is a result of forward propagation of the second result and a fifth result in the third parameter set of the first model; the fifth result is a result of forward propagation of the first result in a second parameter set of the first model; and the updated local second model is obtained by updating the local second model according to the third result.

[0007] Based on the first aspect, when training the model, the first model can be collaboratively trained based on the first parameter set, the second parameter set, the third parameter set, and the second model, thereby improving the generalization ability of the first model and providing favorable guidance for subsequent training of the local second model. Meanwhile, the terminal device can update the local second model and send it to the second device, thereby achieving knowledge sharing of the terminal device.

[0008] In a possible design, model configuration information is received from the second device, and the local second model is initialized according to the model configuration information. The model configuration information includes recommended parameters and matrix dimensions.

[0009] In a possible design, the capability information of the terminal device is sent to the second device, and the model configuration information can be determined according to the capability information of the terminal device.

[0010] Based on the possible design, the capability information of the terminal device sent by the terminal device to the second device and the model configuration information sent by the second device to the terminal device take into account the computing capability of the terminal device, so that the local second model of each terminal device is more suitable for the actual processing capability of the terminal device, thereby achieving better training effect.

[0011] In a possible design, the updated local second model is obtained by updating the local second model according to the third result, including: determining an update gradient value of the local second model according to the third result; and updating the local second model according to the update gradient value of the local second model, to obtain the updated local second model.

[0012] In a possible design, the local dataset includes knowledge of the terminal device related to a user task. The knowledge of different terminal devices related to the user task can be homogeneous or heterogeneous.

[0013] Based on this possible design, the structure of the knowledge of the terminal device related to the user task is not limited, and the use of the data structure of the terminal device is more flexible.

[0014] In a second aspect, the present application provides a communication method, which can be executed by a first device. In the case where no special description is given, the first device in the present application can refer to the first device itself, a component (for example, a processor, a chip, or a chip system) in the first device, or a logic module or software capable of realizing all or part of the functions of the first device. The method includes: receiving a local dataset from a plurality of terminal devices; sending a first result corresponding to each terminal device to the plurality of terminal devices; receiving a plurality of second results from the plurality of terminal devices; sending a third result corresponding to each terminal device to the plurality of terminal devices; and updating a second parameter set of a first model according to a global second model. The first result corresponding to the terminal device is a result of forward propagation of the local dataset of the terminal device in the first parameter set of the first model; the second result of the terminal device is a result of forward propagation of the first result corresponding to the terminal device in the initialized local second model of the terminal device; the third result is a result of backward propagation of a fourth result in a third parameter set of the first model; the fourth result is a result of forward propagation of the second result and a fifth result in the third parameter set of the first model; the fifth result is a result of forward propagation of the first result in the second parameter set of the first model; the global second model is received from a second device; and the global second model is obtained by aggregating the updated local second models of the plurality of terminal devices.

[0015] Based on the second aspect, when the model is trained, the first parameter set, the second parameter set, the third parameter set of the first model, and the second model can be used for collaborative training, so as to improve the generalization ability of the first model and provide favorable guidance for subsequent training of the local second model. Meanwhile, the terminal device can also update the local second model and send it to the second device, so as to realize knowledge sharing of the terminal device.

[0016] In a possible design, the local dataset includes knowledge of the terminal device related to a user task. The knowledge of different terminal devices related to the user task can be homogeneous or heterogeneous.

[0017] Based on this possible design, the structure of the knowledge of the terminal device related to the user task is not limited, and the use of the data structure of the terminal device is more flexible.

[0018] In a third aspect, the present application provides a communication apparatus, which can be applied to the terminal device in the first aspect to implement the functions performed by the terminal device. The communication apparatus can be the terminal device, a chip or chip system or system on chip, etc. of the terminal device. The communication apparatus can perform the functions performed by the terminal device through hardware or corresponding software. The hardware or software includes one or more modules corresponding to the functions. For example, the transceiver module and the processing module. The transceiver module can perform the following transceiving operations independently or in cooperation with the processing module. Similarly, the processing module can perform the following processing operations independently or in cooperation with the transceiver module.

[0019] For example, the transceiver module can be configured to send the local data set to the first device, receive the first result from the first device, send the second result to the first device according to the first result, receive the third result from the first device, and send the updated local second model to the second device. The first result is the result of the forward propagation of the local data set in the first parameter set of the first model. The second result is the result of the forward propagation of the first result in the initialized local second model. The third result is the result of the backward propagation of the fourth result in the third parameter set of the first model. The fourth result is the result of the forward propagation of the second result and the fifth result in the third parameter set of the first model. The fifth result is the result of the forward propagation of the first result in the second parameter set of the first model. The updated local second model is obtained by updating the local second model according to the third result.

[0020] Optionally, the transceiver module and the processing module of the communication apparatus in the third aspect can also perform the corresponding functions in any possible design of the first aspect. For details, refer to the detailed description in the method examples. The beneficial effects achieved can also be seen from the foregoing related content.

[0021] In a fourth aspect, the present application provides a communication apparatus, which can be applied to the first device in the second aspect to implement the functions performed by the first device. The communication apparatus can be the network device, a chip or chip system or system on chip, etc. of the network device. The communication apparatus can perform the functions performed by the network device through hardware or corresponding software. The hardware or software includes one or more modules corresponding to the functions. For example, the transceiver module and the processing module. The transceiver module can perform the following transceiving operations independently or in cooperation with the processing module. Similarly, the processing module can perform the following processing operations independently or in cooperation with the transceiver module.

[0022] The transceiver module can be configured to receive the local data set from the plurality of terminal devices, send the first result corresponding to each terminal device to the plurality of terminal devices, receive the second result from the plurality of terminal devices, send the third result corresponding to each terminal device to the plurality of terminal devices, and the processing module can be configured to update the second parameter set of the first model according to the global second model. The first result corresponding to the terminal device is a result of forward propagation of the local data set of the terminal device in the first parameter set of the first model. The second result of the terminal device is a result of forward propagation of the first result of the terminal device in the initialized local second model of the terminal device. The third result is a result of back propagation of the fourth result in the third parameter set of the first model. The fourth result is a result of forward propagation of the second result and the fifth result in the third parameter set of the first model. The fifth result is a result of forward propagation of the first result in the second parameter set of the first model. The global second model is received from the second device. The global second model is obtained by aggregating the updated local second models of the plurality of terminal devices.

[0023] Optionally, the transceiver module and the processing module of the communication device in the fourth aspect can also perform the corresponding functions in any possible design of the second aspect, and the specific functions can be referred to the detailed description in the method examples. The beneficial effects achieved can also be referred to the foregoing related content.

[0024] In the fifth aspect, the present application provides a communication device, which comprises one or more processors; and the one or more processors are configured to execute computer programs or instructions, and when the one or more processors execute the computer programs or instructions, the positioning method in any one of the first aspect to the second aspect is executed.

[0025] In a possible design, the communication device further comprises one or more memories coupled to the one or more processors, and the one or more memories are configured to store the computer programs or instructions. In a possible implementation, the memory is located outside the communication device. In another possible implementation, the memory is located inside the communication device. In the present application, the processor and the memory can also be integrated into one device, i.e., the processor and the memory can also be integrated together. In a possible implementation, the communication device further comprises a transceiver configured to receive information and / or send information.

[0026] In a possible design, the communication device further comprises one or more communication interfaces coupled to the one or more processors, and the one or more communication interfaces are configured to communicate with other modules outside the communication device.

[0027] In a sixth aspect, the present application provides a communication apparatus (physical apparatus), comprising an interface circuit and a logic circuit; the interface circuit is configured to input and / or output information; the logic circuit is configured to perform the communication method according to any one of the first aspect to the second aspect, process and / or generate information according to the information.

[0028] In a seventh aspect, the present application provides a computer readable storage medium, which stores computer instructions or programs, when the computer instructions or programs are run on a computer, the communication method according to any one of the first aspect to the second aspect is performed.

[0029] In an eighth aspect, the present application provides a computer program product comprising computer instructions, when the computer instructions are run on a computer, the communication method according to any one of the first aspect to the second aspect is performed.

[0030] In a ninth aspect, the present application provides a computer program, when the computer program is run on a computer, the communication method according to any one of the first aspect to the second aspect is performed.

[0031] In a tenth aspect, the present application provides a chip, comprising: one or more processors, the processor is coupled with one or more memories, the memory is configured to store programs or instructions, when the programs or instructions are executed by the processor, the communication method according to any one of the first aspect to the second aspect is performed.

[0032] The technical effects brought by any one of the third aspect to the fourth aspect can be referred to the technical effects brought by any one of the first aspect to the second aspect, which will not be repeated here.

[0033] In an eleventh aspect, the present application provides a communication system, which can comprise a communication apparatus for performing the communication method according to the first aspect or any possible design of the first aspect, a communication apparatus for performing the communication method according to the second aspect or any possible design of the second aspect. BRIEF DESCRIPTION OF DRAWINGS

[0034] FIG. 1 is a schematic diagram of a communication system according to an embodiment of the present application;

[0035] FIG. 2 is a schematic diagram of a communication apparatus according to an embodiment of the present application;

[0036] FIG. 3 is a flowchart of a communication method according to an embodiment of the present application;

[0037] FIG. 4 is a schematic diagram of a communication apparatus according to an embodiment of the present application;

[0038] FIG. 5 is a schematic diagram of a communication apparatus according to an embodiment of the present application. DETAILED DESCRIPTION

[0039] Before the embodiments of the present application are described, technical terms involved in the embodiments of the present application are described.

[0040] To meet the needs of the rapid development of the network, large-pretrained models (LPMs) will become one of the key technologies for network development. By pre-training LPM models on a large amount of data sets, the purpose of enabling the model to capture and store general knowledge and patterns within the data set is achieved. With the rapid growth of the demand for LPM services by terminal devices, the network load caused by adjusting / training all parameters of the LPM is too large, thus deriving a fine-tuning technology that replaces adjusting / training all parameters by fine-tuning / training part of the parameters, such as training part of the parameters of the LPM through a low-rank adaptation of large language models (LoRA) model. The fine-tuning technology of fine-tuning / training part of the parameters can reduce the network load, enabling the LPM to generalize the tasks of terminal devices more efficiently, and providing an important driving force for the diversification of emerging LPM services.

[0041] Although the fine-tuning technology provides these attractive and valuable advantages, it still faces many challenges in a wireless network distributed environment, especially when fine-tuning LPMs in a distributed manner.

[0042] When fine-tuning / training the local LoRA model on the terminal device, the model training can utilize the knowledge directly related to the user task on the terminal device.

[0043] However, the training process of the local LoRA model on the terminal device has the following disadvantages:

[0044] The knowledge directly related to the user task on each terminal device is retained on the terminal device, limiting the sharing of different knowledge. In an ideal centralized learning environment, the knowledge directly related to the user task collected from all terminal devices can be shared and used. However, in a wireless network distributed environment, when fine-tuning / training the local LoRA model on each terminal device, the fine-tuned / trained parameters are retained on the terminal device, limiting the sharing of knowledge between different terminal devices, i.e., other terminal devices cannot utilize the fine-tuned / trained parameters used by the terminal device to optimize the performance of the local LoRA model on other terminal devices.

[0045] In addition, the fine-tuned / trained parameters of the local LoRA model on each terminal device are not fed back to the global version of the LPM, and do not affect the main structure and parameters of the LPM. The generalization ability of the LPM itself is not enhanced.

[0046] Furthermore, the computing resources of different types of terminal devices are different, resulting in that the parameters of the fine-tuning / training of the local LoRA models of different types of terminal devices are inconsistent, and there is a problem of difficulty in aligning the fine-tuned models for attempting to maintain or update a global fine-tuned model, so that a global fine-tuned model cannot be obtained.

[0047] In order to improve the above problems, a distributed collaborative learning framework can be used for model training.

[0048] Among them, the distributed collaborative learning framework shows unique potential in knowledge sharing, which allows the training of local LoRA models on terminal devices, where each local LoRA model reflects the data characteristics of the knowledge directly related to the user task on the terminal device. By aggregating different local LoRA models to build a global LoRA model, the knowledge directly related to the user task on each terminal device is effectively integrated and utilized while maintaining the privacy of the knowledge directly related to the user task on each terminal device. Based on the distributed collaborative learning framework for model training allows LPM to learn general knowledge while also capturing knowledge directly related to the user task on different terminal devices.

[0049] For example, a federated learning method can be used for model training. Based on the federated learning method, the LPM and the local model on the terminal device are trained, and each terminal device can use a homogeneous LoRA model to learn the knowledge directly related to the user task on each terminal device. The network side device can aggregate the local LoRA model of each terminal device to obtain a global LoRA model to realize the sharing of the knowledge directly related to the user task on different terminal devices; the global knowledge carried by the aggregated global LoRA model can guide the local LoRA model and enhance the generalization ability of the LPM. This method trains the local model on each terminal device to obtain the updated local model of each terminal device, and then aggregates the updated local models of different terminal devices to build a global LoRA model.

[0050] However, in the above model training method based on federated learning, there is a great limitation in the heterogeneity of the individualized knowledge of the terminal devices, especially when the categories of the individualized knowledge of each terminal device are different or the difference between the individualized knowledge is large. The global knowledge carried by the aggregated global LoRA model may not bring beneficial guidance to the training of the local LoRA model. At the same time, the computing and communication resources of each terminal device are different, resulting in that the LoRA models supported by these resource configurations for computing are also different.

[0051] To solve the technical problem, the embodiment of the present application provides a communication method, in which a terminal device sends a local data set to a first device; receives a first result from the first device; according to the first result, the terminal device sends a second result to the first device; the terminal device receives a third result from the first device; and the terminal device sends an updated local second model to a second device. Wherein, the first result is a result of forward propagation of the local data set in a first parameter set of a first model; the second result is a result of forward propagation of the first result in an initialized local second model; the third result is a result of backward propagation of a fourth result in a third parameter set of the first model; the fourth result is a result of forward propagation of the second result and a fifth result in the third parameter set of the first model; the fifth result is a result of forward propagation of the first result in a second parameter set of the first model; and the updated local second model is obtained by updating the local second model according to the third result.

[0052] In the embodiment of the present application, when the model training is performed, the first model can be collaboratively trained based on the first parameter set, the second parameter set, the third parameter set of the first model, and the second model, so as to improve the generalization ability of the first model and provide favorable guidance for the subsequent training of the local second model. At the same time, the terminal device can also update the local second model and send it to the second device, so as to realize the knowledge sharing of the terminal device.

[0053] The implementation of the embodiment of the present application will be described in detail below in combination with the drawings of the specification.

[0054] The communication method provided by the embodiments of the present application can be used in any communication system, which can be a third generation partnership project (3GPP) communication system, for example, a long term evolution (LTE) system, and can also be a fifth generation (5G) mobile communication system, a system of mixed networking of LTE and 5G, an NR system, an NR vehicle to everything (V2X) system, a device-to-device (D2D) communication system, a machine to machine (M2M) communication system, an internet of things (IoT), a narrow band-internet of things (NB-IoT) system, a global system for mobile communications (GSM) system, an enhanced data rate for GSM evolution (EDGE) system, a wideband code division multiple access (WCDMA) system, a code division multiple access (CDMA) 2000 system, a time division-synchronization code division multiple access (TD-SCDMA) system, an enhanced mobile broadband (eMBB) system, an ultra-reliable and low-latency communication (URLLC) system, an enhanced machine-type communication (eMTC) system, and various types of next-generation communication systems such as a sixth generation (6G) mobile communication system, and can also be a non-terrestrial network (NTN) system (such as a satellite communication system), a non-3GPP communication system, and the like, without limitation.

[0055] In addition, the embodiments of the present application are applicable to homogeneous network and heterogeneous network scenarios, and are not limited to transmission points. The embodiments of the present application are applicable to multi-point cooperative transmission between macro base stations and macro base stations, micro base stations and micro base stations, and macro base stations and micro base stations, and are applicable to frequency division duplex (FDD) and time division duplex (TDD) systems. The embodiments of the present application can be applicable to low frequency scenarios (sub 6G) and high frequency scenarios (6G and above), terahertz, optical communication, etc., without limitation.

[0056] The communication system provided by the embodiments of the present application is described below with reference to FIG. 1.

[0057] FIG. 1 is a schematic diagram of a communication system provided by an embodiment of the present application. As shown in FIG. 1, the communication system can include at least one terminal device, at least one network device, and at least one core network device.

[0058] The terminal device in FIG. 1 can be a device with wireless transceiving function or a chip or chip system that can be provided in the device, and can allow a user to access a network, which is a device for providing voice and / or data connectivity to a user. The terminal device can also be referred to as a user equipment (UE), a subscriber unit, a terminal, or a mobile station (MS) or a mobile terminal (MT), etc.

[0059] Exemplarily, the terminal device in FIG. 1 can be a mobile phone, a tablet computer, or a computer with wireless transceiver function. The terminal device can also be a user station, a mobile station, a remote station, a remote terminal device, a mobile terminal device, a user terminal device, a wireless communication device, a user agent, a user equipment, a cellular phone, a cordless phone, a session initiation protocol (SIP) phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), a handheld device with wireless communication function, a computing device, a processing device connected to a wireless modem, an in-vehicle device, a wearable device, a terminal device in Internet of Things, a household appliance, a virtual reality (VR) terminal, an augmented reality (AR) terminal, a wireless terminal in industrial control, a wireless terminal in unmanned driving, a wireless terminal in remote medical treatment, a wireless terminal in smart power grid, a wireless terminal in smart city, a wireless terminal in smart home, a vehicle with vehicle-to-vehicle (V2V) communication capability, a smart connected vehicle, a drone with unmanned aerial vehicle to unmanned aerial vehicle (UAV to UAV, U2U) communication capability, a terminal device in future network, or a terminal device in future evolved public land mobile network (PLMN), etc. It can be understood that the terminal device and the mobile user can be completely independent. All information related to the user can be stored in a subscriber identity module (SIM) card, which can be used on the terminal device. The terminal device can send and / or receive signals through the air interface to complete interaction with the network device.

[0060] The network device in FIG. 1 can be any device deployed in an access network and capable of wireless communication with the terminal device, can also be a chip or chip system that can be provided in the above device, can also be a logic node or a logic module or a software-implemented function, and is mainly responsible for functions such as wireless physical control function on the air interface side, resource scheduling, wireless resource management, quality of service management, data compression and encryption, wireless access control, and mobility management. Specifically, the network device can be a device supporting wired access or a device supporting wireless access.

[0061] Exemplary network devices can be composed of one or more access network (AN) / radio access network (RAN) nodes. AN / RAN nodes can be various types of base stations such as satellite base stations, continue evolution NodeBs (gNBs), transmission reception points (TRPs), evolved NodeBs (eNBs), radio network controllers (RNCs), NodeBs (NBs), base station controllers (BSCs), base transceiver stations (BTSs), home base stations (e.g., home eNBs or home NBs, HNB), macro base stations, micro base stations, pico base stations, femto base stations, relay stations, balloon stations, drone stations, wireless backhaul nodes, base band units (BBUs), or wireless fidelity (Wi-Fi) access points (APs), etc. It can be understood that network devices can be ground-based devices or non-ground-based devices (e.g., satellites, drones, high-altitude communication devices, etc.). In addition, in communication systems using different wireless access technologies, the names of network devices with base station functions can be different, which is not limited in the present application.

[0062] In yet another example, network devices can include a BBU and a remote radio unit (RRU). The BBU and the RRU can be placed in different locations, for example, the RRU is pulled away and placed in a high traffic area, and the BBU is placed in a central machine room. The BBU and the RRU can also be placed in the same machine room. The BBU and the RRU can also be different components under one rack.

[0063] In still another example, network devices can also be devices including a centralized unit (CU) node, or including a distributed unit (DU) node, or including a CU node and a DU node. For example, network devices can be divided into a CU and a DU from a logical function perspective, and the functions of part of the protocol layers are placed in the CU for centralized control, and the rest or all of the protocol layers are distributed in the DU and controlled by the CU. The CU and the DU can be separately arranged or can be included in the same network element, such as a BBU. Furthermore, the centralized unit CU can also be divided into a control plane (CU-CP) and a user plane (CU-UP).

[0064] In another example, the network device can also be a device including a radio unit (RU), or including a CU, a DU and a RU. The RU can be included in a radio frequency device or a radio frequency unit, for example, included in a RRU, an active antenna unit (AAU) or a remote radio head (RRH).

[0065] It can be understood that the CU (or CU-CP and CU-UP), DU or RU can also have different names in different systems, but those skilled in the art can understand their meanings. For example, in an open radio access network (O-RAN) system, the CU can also be referred to as O-CU (open CU), the DU can also be referred to as O-DU, the CU-CP can also be referred to as O-CU-CP, the CU-UP can also be referred to as O-CU-UP, and the RU can also be referred to as O-RU. For the convenience of description, the CU, CU-CP, CU-UP, DU and RU are taken as examples for description in the present application. Any one of the CU (or CU-CP, CU-UP), DU and RU in the present application can be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0066] In the above, the core network device in FIG. 1 can be a general term for various functional entities for managing users, data transmission, completing a bearer for a service, providing an interface to an external network as a bearer network, and network device configuration.

[0067] For example, the core network device can include a mobility management network element, a session management network element, a user plane network element, and the like, without limitation. The mobility management network element can be an access and mobility management function (AMF). The session management network element can be a session management function (SMF). The user plane network element can be a user plane function (UPF).

[0068] Based on the above description of the terminal device, the network device and the core network device, optionally, the communication method provided by the embodiments of the present application can be implemented by the terminal device or the network device, or by components of the terminal device or the network device.

[0069] In a specific implementation, each of the terminal device, the network device, and the core network device shown in FIG. 1 can adopt the component structure shown in FIG. 2, or include the components shown in FIG. 2. FIG. 2 is a component diagram of a communication apparatus 200 according to an embodiment of the present application. The communication apparatus 200 can be a terminal device or a chip or system on chip in the terminal device, or a network device or a chip or system on chip in the network device, or a core network device or a chip or system on chip in the core network device. As shown in FIG. 2, the communication apparatus 200 includes a processor 201, a transceiver 202, and a communication line 203.

[0070] Further, the communication apparatus 200 can further include a memory 204. The processor 201, the memory 204, and the transceiver 202 can be connected through the communication line 203.

[0071] The processor 201 can be a central processing unit (CPU), a general processor network processor (NP), a digital signal processing (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. The processor 201 can also be other devices with processing functions, such as a circuit, a device, or a software module, without limitation.

[0072] The transceiver 202 is configured to communicate with other devices or other communication networks. The other communication networks can be an Ethernet, a radio access network (RAN), a wireless local area network (WLAN), and the like. The transceiver 202 can be a module, a circuit, a transceiver, or any device capable of communication.

[0073] The communication line 203 is configured to transmit information between components included in the communication apparatus 200.

[0074] The memory 204 is configured to store instructions. The instructions can be a computer program.

[0075] The memory 204 can be a read-only memory (ROM) or other type of static storage device that can store static information and / or instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and / or instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magnetic disk storage or other magnetic storage devices, etc., without limitation.

[0076] It should be noted that the memory 204 can exist independently of the processor 201 or can be integrated with the processor 201. The memory 204 can be used to store instructions or program codes or some data, etc. The memory 204 can be located within the communication apparatus 200 or outside the communication apparatus 200, without limitation. The processor 201 is configured to execute the instructions stored in the memory 204 to implement the communication method provided by the embodiments described below.

[0077] In an example, the processor 201 can include one or more CPUs, such as CPU0 and CPU1 in FIG. 2.

[0078] As an optional implementation, the communication apparatus 200 includes multiple processors, for example, in addition to the processor 201 in FIG. 2, the communication apparatus 200 can further include a processor 207.

[0079] As an optional implementation, the communication apparatus 200 further includes an output device 205 and an input device 206. For example, the input device 206 is a keyboard, a mouse, a microphone, a joystick, etc., and the output device 205 is a display screen, a speaker, etc.

[0080] It should be noted that the communication apparatus 200 can be a desktop computer, a laptop computer, a network server, a mobile phone, a tablet computer, a wireless terminal, an embedded device, a chip system, or a device having a similar structure to that in FIG. 2. In addition, the constituent structure shown in FIG. 2 does not constitute a limitation on the communication apparatus, and the communication apparatus can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0081] In the embodiments of the present application, the chip system can be composed of a chip or can include a chip and other discrete devices.

[0082] In addition, the actions, terms, etc. involved between the embodiments of the present application can be mutually referred to without limitation. The message names or parameter names in the messages exchanged between the various devices in the embodiments of the present application are only examples, and other names can also be used in specific implementations without limitation.

[0083] The communication method provided by the embodiments of the present application will be described below with reference to FIG. 1 and FIG. 3.

[0084] The terminal device can be any terminal device in the communication system shown in FIG. 1, the first device can be any network device or core network device in the communication system shown in FIG. 1, the second device can be any network device or core network device in the communication system shown in FIG. 1, and the first device and the second device can be the same device or different devices without limitation.

[0085] FIG. 3 is a flowchart of a communication method provided by an embodiment of the present application. As shown in FIG. 3, the method can include the following steps.

[0086] In step 301, one or more terminal devices send a local data set of each terminal device to a first device, and correspondingly, the first device receives the local data set from each terminal device.

[0087] The local data set of the terminal device includes knowledge of the terminal device related to a user task. The knowledge of different terminal devices related to the user task is homogeneous or heterogeneous.

[0088] For example, the local data set of the terminal device can include one or more of the following: general knowledge of the terminal device, or personalized knowledge of the terminal device.

[0089] The general knowledge can be knowledge shared by or common to all terminal devices, and the personalized knowledge can be knowledge unique to each terminal device.

[0090] For example, the knowledge related to the user task can include specific feature recognition, language understanding, etc.

[0091] In step 302, the first device sends a first result corresponding to the terminal device to one or more terminal devices, and correspondingly, the one or more terminal devices receive the first result from the first device.

[0092] The first result corresponding to the terminal device is a result of forward propagation of the local data set of the terminal device in a first parameter set of a first model.

[0093] For each terminal device, the first device can forward propagate the local data set of the terminal device in the first parameter set of the first model to obtain a first result corresponding to each terminal device, and send the first result to each terminal device respectively.

[0094] The training parameters in the first model include the first parameter set, the second parameter set, and the third parameter set.

[0095] Optionally, in the forward propagation process of the first model, the forward propagation can be sequentially performed through the first parameter set, the second parameter set, and the third parameter set.

[0096] Optionally, the second parameter set includes training parameters in the first model that are trained / adjusted by the second model.

[0097] The first model is deployed in the first device, and can be used for network self-use or third-party service.

[0098] For example, the first model can be an LPM model. By setting the LPM model in the first device, the network can add the LPM function, and can provide continuous fine-tuning / training parameter service for the LPM model deployed on the network.

[0099] For example, the second model can be a LoRA model. By fine-tuning / training the second parameter set of the LPM using the LoRA model, the network load can be reduced, and the fine-tuning technology of adjusting / training part of the parameters is replaced to realize the model training performance close to adjusting / training all parameters.

[0100] Optionally, when the first device sends the first result to the terminal device, the first result can be taken as a payload and sent to the terminal device through a data transmission channel of a user plane.

[0101] In step 303, one or more terminal devices send a second result of the terminal device to the first device; correspondingly, the first device receives the second result from the one or more terminal devices.

[0102] The second result of the terminal device is a result of forward propagation of the first result corresponding to the terminal device in the initialized local second model of the terminal device.

[0103] For each terminal device, the terminal device can forward propagate the first result corresponding to the terminal device in the initialized local second model of the terminal device to obtain the second result of each terminal device.

[0104] Optionally, the local second model is a second model stored locally in the terminal device.

[0105] For example, the local second model can be a local LoRA model.

[0106] Optionally, the one or more terminal devices can also initialize the local second model before using the local second model, and the method of initialization can refer to the following description:

[0107] The one or more terminal devices initialize the local second model according to the initialization configuration.

[0108] The initialization configuration includes one or more of the following: computing resources or model configuration information.

[0109] The computing resources are used to indicate the computing capability of the terminal device itself, and the model configuration information is sent by the second device to the terminal device. The model configuration information includes recommended parameters and matrix dimensions.

[0110] The recommended parameters are used to configure the parameters that need to be adjusted / trained by the local second model, and the matrix dimensions are used to indicate that the matrix dimensions of the two models are the same when the local second model is collaboratively trained with the first model.

[0111] For example, the recommended parameters are the rank of low-rank decomposition, and the rank is the recommended configuration of the local second model by the second device.

[0112] The terminal device reports the computing capability to the second device according to the computing resources of the terminal device, and the second device issues corresponding training parameters to the terminal device according to the computing capability.

[0113] For example, when the terminal device is a high-performance server or a professional computing device, the computing capability of the terminal device is high, and the second device issues a large number and types of local second model training parameters to the terminal device; when the terminal device is a smartphone or a small edge device, the computing capability of the terminal device is low, and the second device issues a small number and types of local second model training parameters to the terminal device compared to the former.

[0114] For example, during the initialization configuration process, k is 0, and the initialized local second model of the terminal device i is:

[0115] Optionally, when the one or more terminal devices send the second results of the terminal devices to the first device, the second results can be used as payloads and sent to the first device through a data transmission channel of a user plane.

[0116] Step 304, the first device sends the third results corresponding to the terminal devices to the one or more terminal devices; correspondingly, the one or more terminal devices receive the third results from the first device.

[0117] The third result corresponding to the terminal device is a result of the fourth result after being back propagated in the third parameter set of the first model; the fourth result is a result of the second result and the fifth result after being forward propagated in the third parameter set of the first model; and the fifth result is a result of the first result after being forward propagated in the second parameter set of the first model.

[0118] For each terminal device, the first device can forward propagate the first result corresponding to the terminal device in the second parameter set of the first model to obtain the fifth result corresponding to each terminal device; the first device can forward propagate the fifth result corresponding to the terminal device and the second result in the third parameter set of the first model to obtain the fourth result corresponding to each terminal; and the first device can back propagate the fourth result corresponding to the terminal device in the third parameter set of the first model to obtain the third result corresponding to each terminal device.

[0119] For example, the third result corresponding to the terminal device is a loss value (Loss), and the value of the Loss of the terminal device i is represented as L i .

[0120] Optionally, when the first device sends the third result corresponding to the terminal device to one or more terminal devices, the third result can be taken as a payload and sent to the terminal device through a data transmission channel of a user plane.

[0121] In step 305, one or more terminal devices send the updated local second model of each terminal device to the second device; correspondingly, the second device receives the updated local second model from one or more terminal devices.

[0122] The terminal device can update the local second model based on the third result to obtain the updated local second model.

[0123] For example, the terminal device can update the local second model according to the following formula:

[0124] wherein, represents the second model; θ L represents the training parameter of the second model; i represents the terminal device (for example, i = 1 represents the terminal device 1); t is a global round, which represents the number of times of updating the first model according to the local second model; k is a local round, which represents the number of times of training the local second model in cooperation with the first model, and the maximum value of k is set in the local second model.

[0125] The maximum value of k can be predefined by a communication protocol, preconfigured by the first device, or preconfigured by the second device, without limitation.

[0126] Based on the description of the parameters in the above formula, the formula for updating the local second model can be understood as: superimposing the updated gradient value adjusted by the learning rate on the second model of the t and k rounds to obtain the second model of the k+1 round, that is, the updated second model. Wherein, the updated gradient value is represented as η represents the learning rate, which is used to determine the update step length of updating the local second model.

[0127] The terminal device and the first device perform one or more collaborative training on the local second model and the first model of the terminal device according to steps 301-305. When the number of training reaches the maximum value of k, the terminal device sends the updated local second model when k is the maximum value to the second device.

[0128] Optionally, when the one or more terminal devices send the updated local second model to the second device, the updated local second model can be used as a payload and sent to the second device through a data transmission channel of the user plane.

[0129] Step 306, the second device sends the global second model to the first device; correspondingly, the first device receives the global second model sent by the second device.

[0130] Optionally, the second device can aggregate the aligned local second model based on the aggregator.

[0131] Optionally, the second device aligns the updated local second model to obtain the aligned local second model, and aggregates the aligned local second model to obtain the global second model.

[0132] Illustratively, the second device can align the updated local second model of each terminal device according to the following formula:

[0133] Wherein, A and B are low-rank matrices.

[0134] Optionally, the A and B matrices can be determined during the training process using the second model.

[0135] It can be understood that the training method of the second model is to train the second model in the form of A and B two low-rank matrices. Therefore, the second device receives the updated local second model of the terminal device in the form of A and B two low-rank matrices. Then, the second device needs to align A and B two low-rank matrices to obtain the aligned local second model.

[0136] In the above formula, indicates the aligned second model parameter; the description of i, t, and k is referred to the above step 305, which will not be repeated here.

[0137] Exemplarily, the global second model can be obtained by aggregating the aligned local second models of the U terminal devices according to the following formula:

[0138] D i represents the number of samples of the local data set of the terminal device i; represents the aligned second model parameter; θ L The descriptions of i and t are the same as those in step 305, which will not be repeated here.

[0139] The second device aggregates the aligned local second models of the U terminal devices to realize the sharing of the personalized knowledge of the U terminal devices.

[0140] Optionally, when the second device sends the global second model to the first device, the global second model can be taken as a payload and sent to the first device through a data transmission channel of a user plane.

[0141] In step 307, the first device updates the second parameter set of the first model according to the global second model.

[0142] The first device updates the first model by loading the global second model into the second parameter set of the first model. The first device loads the global second model into the second parameter set of the first model according to the following formula:

[0143] wherein θ g represents the second model used for training the second parameter set of the first model; represents loading the global second model into the second model used for training the second parameter set of the first model; t is a global round, which represents the number of times of updating the first model according to the local second model; μ is a coefficient for determining the loading step of the global second model.

[0144] If the inference accuracy of the updated first model does not meet the requirement, the terminal device and the first device continue to perform collaborative training, and the second device sends the local second model configuration information of the next round of training to the terminal device; otherwise, the collaborative training of the terminal device and the first device is ended.

[0145] wherein whether the inference accuracy of the first model meets the accuracy requirement is determined according to the test result of the updated first model tested on a test machine. The updated first model is tested (for example, tested by a test machine), and the test result is obtained. Whether the accuracy requirement is met is determined according to the test result.

[0146] Based on the method shown in Figure 3, the first model on the first device and the local second model on the terminal device are collaboratively trained, wherein, during the collaborative training, the first parameter set, the second parameter set, and the third parameter set of the first model are used, thereby providing favorable guidance for the training of the local second model by the first model. While achieving personalized knowledge sharing of the terminal device, the generalization capability of the first model is enhanced.

[0147] It can be understood that, in the embodiments of the present application, the execution subject can execute part or all of the steps in the embodiments of the present application, and these steps or operations are only examples, and the embodiments of the present application can also execute other operations or variations of various operations. In addition, each step can be executed in a different order as presented in the embodiments of the present application, and it is possible that not all operations in the embodiments of the present application are executed.

[0148] The above mainly introduces the scheme provided by the embodiments of the present application from the perspective of interaction between devices. It can be understood that, in order to achieve the above functions, each device comprises a hardware structure and / or a software module for executing each function. Those skilled in the art should easily realize that, in combination with the algorithm steps of each example described in the embodiments disclosed herein, the present application can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in hardware or computer software driven hardware depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0149] The embodiments of the present application can divide the functional modules of each device according to the above method examples, for example, each functional module can be divided according to each function, or two or more functions can be integrated in one processing module. The above integrated module can be realized in the form of hardware or software functional module. It should be noted that the division of modules in the embodiments of the present application is illustrative, and is only a logical functional division. Actual implementation can have another division manner.

[0150] In the case of dividing each functional module according to each function, Figure 4 shows a communication apparatus 400, which can execute the actions performed by the terminal device, or the first device, or the second device in the method shown in Figure 3. All related contents of each step involved in the above method embodiments can be referred to the function description of the corresponding functional module, and the technical effects that can be obtained are referred to the above method embodiments, which will not be described here again.

[0151] The communication apparatus 400 can include a transceiver module 401 and a processing module 402. For example, the communication apparatus 400 can be a communication device, or a chip or other combination device or component having the above-mentioned transmitting device function applied in the communication device. When the communication apparatus 400 is a communication device, the transceiver module 401 can be a transceiver, which can include an antenna and a radio frequency circuit, etc. The processing module 402 can be a processor (or processing circuit), for example, a baseband processor, which can include one or more CPUs. When the communication apparatus 400 is a component having the above-mentioned transmitting device function, the transceiver module 401 can be a radio frequency unit. The processing module 402 can be a processor (or processing circuit), for example, a baseband processor. When the communication apparatus 400 is a chip system, the transceiver module 401 can be an input / output interface of a chip (for example, a baseband chip). The processing module 402 can be a processor (or processing circuit) of the chip system, which can include one or more central processing units. It should be understood that the transceiver module 401 in the embodiments of the present application can be implemented by a transceiver or a transceiver-related circuit component. The processing module 402 can be implemented by a processor or a processor-related circuit component (or processing circuit).

[0152] For example, the transceiver module 401 can be configured to perform all the transceiving operations performed by the transmitting device in the embodiment shown in FIG. 3, and / or other processes for supporting the technologies described herein. The processing module 402 can be configured to perform all the operations performed by the transmitting device in the embodiment shown in FIG. 3, except the transceiving operations, and / or other processes for supporting the technologies described herein.

[0153] As another implementation manner, the transceiver module 401 in FIG. 4 can be replaced by a transceiver which can integrate the functions of the transceiver module 401. The processing module 402 can be replaced by a processor which can integrate the functions of the processing module 402. Further, the transmitting device 400 shown in FIG. 4 can further include a memory.

[0154] Alternatively, when the processing module 402 is replaced by a processor and the transceiver module 401 is replaced by a transceiver, the transmitting device 400 related to the embodiments of the present application can also be the communication apparatus 500 shown in FIG. 5.

[0155] The processor can be a logic circuit 501, and the transceiver can be an interface circuit 502. Further, the communication apparatus 500 shown in FIG. 5 can further include a memory 503.

[0156] The embodiments of the present application also provide a computer program product, which can realize the functions of any of the above-mentioned method embodiments when executed by a computer.

[0157] The embodiment of the present application further provides a computer program which can realize the functions of any of the above method embodiments when executed by a computer.

[0158] The embodiment of the present application further provides a computer readable storage medium. All or part of the processes of the above method embodiments can be instructed by a computer program to relevant hardware to complete, the program can be stored in the above computer readable storage medium, and the program can include the processes of the above method embodiments when executed. The computer readable storage medium can be an internal storage unit of the terminal (including a data sending terminal and / or a data receiving terminal) of any of the above embodiments, for example, a hard disk or a memory of the terminal. The computer readable storage medium can also be an external storage device of the terminal, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card and the like. Further, the computer readable storage medium can include both the internal storage unit and the external storage device of the terminal. The computer readable storage medium is used to store the above computer program and other programs and data required by the terminal. The computer readable storage medium can also be used to temporarily store data that has been output or will be output.

[0159] It should be noted that the terms "first" and "second" and the like in the specification of the present application, claims and drawings are used to distinguish different objects, and are not used to describe a specific order. "First", "second" are only for description purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more features. In the description of the present embodiment, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0160] In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed or can optionally include other steps or units inherent to the process, method, product or device.

[0161] It should be understood that in the present application, "at least one" means one or more. "Multiple" means two or more. "At least two" means two or three and more. "And / or" is used to describe the relationship between the associated objects, which means that there can be three relationships. For example, "A and / or B" can mean that there are three cases: only A, only B, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c, can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple. "When" and "if" both mean that under certain objective circumstances, the corresponding processing will be done, not limited to time, and does not require a judgment action when implemented, nor does it mean that there are other limitations.

[0162] In the embodiments of the present application, the words such as "exemplary" or "for example" are used to mean an example, illustration, or description. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the words "exemplary" or "for example" are intended to present the relevant concept in a specific manner for understanding.

[0163] In the present application, "sending information to (a terminal device)" can be understood as that the destination of the information is the terminal device. It can include direct or indirect sending of information to the terminal device. "Receiving information from (a terminal device)" can be understood as that the source of the information is the terminal device, and it can include direct or indirect receiving of information from the terminal device. The information can be processed as necessary between the source and the destination of the information transmission, such as format change, etc., but the destination can understand the valid information from the source.

[0164] Through the description of the above embodiments, those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of functional modules is taken as an example, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0165] In several embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented by other manners. For example, the apparatus embodiments described above are merely illustrative, for example, the division of the modules or units is merely a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another apparatus, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units or components shown or discussed can be indirect coupling or communication connection through some interfaces, apparatuses or units, and can be electrical, mechanical or other forms.

[0166] The units described as separate components can or can not be physically separated, and the components shown as units can be one physical unit or a plurality of physical units, that is, can be located in one place or can be distributed to a plurality of different places. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0167] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0168] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a readable storage medium. Based on such understanding, the technical scheme of the embodiment of the present application can be embodied in the form of a software product in essence or all or part of the technical scheme. The software product is stored in a storage medium and includes a plurality of instructions for causing an apparatus (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a ROM, a RAM, a magnetic disk or an optical disk and various program code storage media.

Claims

1. A communication method characterized by comprising: The method comprises: sending a local data set to a first device; receiving a first result from the first device; the first result is a result of forward propagation of the local data set in a first parameter set of a first model; sending a second result to the first device according to the first result; the second result is a result of forward propagation of the first result in an initialized local second model; receiving a third result from the first device; the third result is a result of back propagation of a fourth result in a third parameter set of the first model; the fourth result is a result of forward propagation of the second result and a fifth result in the third parameter set of the first model; the fifth result is a result of forward propagation of the first result in a second parameter set of the first model; sending an updated local second model to a second device; the updated local second model is obtained by updating the local second model according to the third result.

2. The method of claim 1, wherein, The method further comprises: receiving model configuration information from a second device; the model configuration information comprises recommended parameters and matrix dimensions; initializing the local second model according to the model configuration information.

3. The method of claim 2, wherein, The method further comprises: sending capability information of a terminal device to the second device; the model configuration information can be determined according to the capability information of the terminal device.

4. The method of claim 1, wherein, The updated local second model obtained by updating the local second model according to the third result comprises: determining an update gradient value of the local second model according to the third result; updating the local second model according to the update gradient value of the local second model to obtain the updated local second model.

5. The method of claim 1, wherein: the local data set comprises knowledge of the terminal device related to a user task, and the knowledge of different terminal devices related to the user task can be homogeneous or heterogeneous.

6. A communication method characterized by comprising: The method comprises: receiving local data sets from a plurality of terminal devices; sending a first result corresponding to each terminal device to the plurality of terminal devices; the first result corresponding to the terminal device is a result of forward propagation of the local data set of the terminal device in a first parameter set of a first model; receiving a plurality of second results from the plurality of terminal devices; the second result of the terminal device is a result of forward propagation of the first result corresponding to the terminal device in an initialized local second model of the terminal device; sending a third result corresponding to each terminal device to the plurality of terminal devices; the third result is a result of back propagation of a fourth result in a third parameter set of the first model; the fourth result is a result of forward propagation of the second result and a fifth result in the third parameter set of the first model; the fifth result is a result of forward propagation of the first result in a second parameter set of the first model; receiving a global second model from a second device; the global second model is obtained by aggregating updated local second models of the plurality of terminal devices; updating the second parameter set of the first model according to the global second model.

7. The method of claim 6, wherein, the local data set is knowledge of the terminal device related to a user task, and the knowledge of different terminal devices related to a user task can be homogeneous or heterogeneous.

8. A communication device, characterized by The communication device comprises a processor; the processor is used to run a computer program or instruction, so that the communication method as claimed in any one of claims 1-5 is executed, or the communication method as claimed in claim 6 or 7 is executed.

9. The communication apparatus according to claim 8, wherein Further comprising a memory for storing the computer program or instruction.

10. The communication apparatus according to claim 8 or 9, wherein, Further comprising a communication interface for communicating with modules outside the communication device.

11. A communications device, characterized by The communication device comprises an interface circuit and a logic circuit; the interface circuit is used to input and / or output information; the logic circuit is used to execute the communication method as claimed in any one of claims 1-5, or execute the communication method as claimed in claim 6 or 7, process and / or generate the information according to the information.

12. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions or programs, when the computer instructions or programs are run on a computer, so that the communication method as claimed in any one of claims 1-5 is executed, or the communication method as claimed in claim 6 or 7 is executed.

13. A computer program product, characterised in that, The computer program product comprises computer instructions; when part or all of the computer instructions are run on a computer, so that the communication method as claimed in any one of claims 1-5 is executed, or the communication method as claimed in claim 6 or 7 is executed.

14. A communications device, characterized by It comprises a module or unit for executing the communication method as claimed in any one of claims 1-5, or a module or unit for executing the communication method as claimed in claim 6 or 7.

Images