Communication method and corresponding communication device

Abstract

A communication method comprises: a first device transmitting first data to a second device using a first artificial intelligence (AI) model, and transmitting second data to the second device using a second AI model; wherein the first AI model is different from the second AI model, and the first data is the same as the second data; the first device receives first information from the second device, wherein the first information is related to the first performance index and the second performance index; wherein the first performance index is a performance index when the first AI model is used for transmitting the first data; the second performance index is a performance index when the second AI model is used for transmitting the second data. According to the scheme provided by the invention, the second performance index corresponding to the second AI model can be obtained by transmitting the same data as the first AI model by using the non-activated second AI model. In this way, the accuracy of AI model switching in the communication process can be improved.

Description

Technical Field

[0001] This application relates to the field of communication technologies, and in particular, to a communication method and a corresponding communication device. Background Art

[0002] When using artificial intelligence (AI) technology, it is first necessary to train an AI model based on a training data set, and the trained AI model can be used for data inference. For the same inference function, there may be multiple AI models, and the model structure, model parameters, computational overhead, transmission overhead, etc. of each AI model may be different. During communication, for multiple AI models with the same inference function, usually one is activated for use, and the other AI models are in an inactive state as backups.

[0003] During the process of using an AI model for data transmission, it is necessary to collect the performance of the used AI model. If the performance of the AI model deteriorates, a better-performing AI model can be selected from the backup AI models for activation, and then the communication service can be switched to the newly activated AI model for processing.

[0004] Since the backup AI models are not enabled, it is impossible to accurately know whether the performance of the backup AI models is better than that of the currently used AI model. Therefore, how to determine the performance of the backup AI models has become an urgent problem to be solved. Summary of the Invention

[0005] This application provides a communication method for obtaining the performance of a backup AI model. This application also provides a corresponding communication device, system, computer-readable storage medium, computer program product, etc.

[0006] A first aspect of this application provides a communication method, including: a first device uses a first artificial intelligence (AI) model to transmit first data to a second device, and uses a second AI model to transmit second data to the second device; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data; the first device receives first information from the second device, and the first information is related to a first performance metric and a second performance metric; wherein, the first performance metric is the performance metric when the first AI model is used to transmit the first data; the second performance metric is the performance metric when the second AI model is used to transmit the second data.

[0007] In this application, the first device may be a communication device or a communication apparatus capable of supporting the communication device to implement the functions required for the communication method, such as a chip. Exemplarily, the first device is a terminal device / network device, or a chip disposed in the terminal device / network device for implementing the functions of the terminal device / network device, or other components for implementing the functions of the terminal device / network device. In the following introduction, it is described by taking the first device as a terminal device / network device as an example.

[0008] In this application, the first AI model and the second AI model may be AI models with the same functions, but the model structures, model parameters, computational overheads, transmission overheads, etc. of the two AI models may all be different or partially different. For example: both the first AI model and the second AI model are modulation and coding scheme (MCS) prediction models, but the model structures and model parameters of these two MCS prediction models are different.

[0009] In this application, the first AI model may be an activation model, and the second AI model may be a non-activation model.

[0010] In this application, the first performance metric may include one or more of the system throughput, block error ratio (BLER), signal interference noise ratio (SINR), or reference signal receive power (RSRP) when transmitting the first data using the first AI model. Of course, the first performance metric may also include other parameters that can indicate the system transmission performance.

[0011] In this application, the second performance metric may include one or more of the system throughput, BLER, SINR, or RSRP when transmitting the second data using the second AI model. Of course, the second performance metric may also include other parameters that can indicate the system transmission performance.

[0012] In this application, the first information may be the first performance metric and the second performance metric, or the result determined by the second device according to the first performance metric and the second performance metric.

[0013] In this first aspect, by transmitting the same data as the first AI model using the non-activation second AI model, the above-mentioned second performance metric corresponding to the second AI model can be obtained. In this way, the accuracy of AI model switching during the communication process can be improved.

[0014] In a possible implementation, the transmission resources for the first data and the second data are frequency division multiplex (FDM) transmission resources, time division multiplex (TDM) transmission resources, or independent transmission resources; wherein, the independent transmission resources mean that the transmission resources for the first data and the second data are different in both the time domain and the frequency domain.

[0015] In this possible implementation, the transmission resources for the first data and the second data are FDM transmission resources, that is, the first data and the second data are transmitted using different frequency domain resources at the same time. The transmission resources for the first data and the second data are TDM transmission resources, that is, the first data and the second data are transmitted using the same frequency domain resources at different times. The transmission resources for the first data and the second data are independent transmission resources, that is, the first data and the second data are transmitted using different frequency domain resources at different times. Thus, in this application, the diversity of the transmission resources for the first data and the second data is provided.

[0016] In a possible implementation, the transmission resources for the first data and the second data are indicated by first control information.

[0017] In this application, the first control information can be downlink control indicator (DCI) or sidelink control indicator (SCI).

[0018] In this application, when the transmission resources for the first data and the second data are FDM transmission resources, the first control information can be indicated in the way of start position + offset value. For example, the start position of the transmission resources for the second data in the frequency domain can be the end position of the transmission resources for the first data in the frequency domain, and the offset value can be the frequency bandwidth occupied by the transmission resources for the second data, or the number of resource elements (RE), or the number of resource blocks (RB).

[0019] In this application, when the transmission resources for the first data and the second data are TDM transmission resources, the first control information can also be indicated in the form of start position + offset value. For example, the start position of the transmission resources for the second data in the time domain can be the end position of the transmission resources for the first data in the time domain, and the offset value can be the number of time domain units occupied by the transmission resources for the second data. For example, it can be the number of time slots, the number of symbols, the number of mini (min) time slots, or the offset value is an absolute time length, for example, in milliseconds or microseconds, etc.

[0020] In this application, when the transmission resources for the first data and the second data are independent transmission resources, the first control information can directly indicate the two transmission resources and indicate the model type corresponding to each transmission resource, that is, an active model or a non - active model.

[0021] In this possible implementation, for different forms of the transmission resources for the first data and the second data, a suitable indication method is provided through a control information, which improves the flexibility of the control information indication.

[0022] In a possible implementation, the method further includes: the first device receives a first hybrid automatic repeat request (HARQ) or / and a second HARQ, where the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0023] In this possible implementation, the first HARQ can be used to indicate whether the first data is successfully transmitted, and the second HARQ can be used to indicate whether the second data is successfully transmitted. Since the first data and the second data are the same data, only the first HARQ or the second HARQ can be used to determine whether one of the data is successfully transmitted. Of course, it is also possible to jointly determine whether the first data or the second data is successfully transmitted through the first HARQ and the second HARQ. Thus, the HARQ feedback form of this application is diverse.

[0024] In a possible implementation, the first HARQ and the second HARQ are used to determine a third HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data.

[0025] In this possible implementation, the third HARQ is determined through the first HARQ and the second HARQ, and then the first data and the second data are determined whether they are successfully transmitted through the third HARQ, which can improve the accuracy of the HARQ indication.

[0026] In a possible implementation, the method further includes: a first device receives a third HARQ determined by a second device through a first HARQ and a second HARQ, where the third HARQ is the HARQ corresponding to first data and second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0027] In this possible implementation, the second device can determine the third HARQ, and the first device directly receives the third HARQ, so as to determine whether the first data and the second data are successfully transmitted.

[0028] In a possible implementation, a first performance metric is determined by the first HARQ.

[0029] In this possible implementation, the first performance metric can be directly the first HARQ, or determined indirectly by the first HARQ. When the first performance metric is directly the first HARQ, it can be that the first HARQ includes one or more of the system throughput, BLER, SINR, or RSRP; when the first performance metric is determined indirectly by the first HARQ, it can be that one or more of the system throughput, BLER, SINR, or RSRP are determined through the first HARQ. Thus, it can be seen that the HARQ in this application can not only indicate data transmission, but also indicate the performance metrics of data transmission, enhancing the indication ability of the HARQ.

[0030] In a possible implementation, a second performance metric is determined by the second HARQ.

[0031] In this possible implementation, the second performance metric can be directly the second HARQ, or determined indirectly by the second HARQ. The specific process can be understood by referring to the first performance metric.

[0032] In a possible implementation, the first performance metric is obtained within a first time period, and the second performance metric is obtained within a second time period, where the first time period and the second time period completely overlap or partially overlap.

[0033] In this possible implementation, the complete overlap of the first time period and the second time period means that the first performance metric and the second performance metric are obtained within the same time period, which can improve the reliability of the second performance metric. The partial overlap of the first time period and the second time period can be that the time used to obtain the second performance metric is shorter than the time used to obtain the first performance metric, so that the non-active model can be judged in advance.

[0034] In a possible implementation, the first information includes at least one of the following:

[0035] The first performance metric and the second performance metric;

[0036] A first indication, where the first indication is used to indicate that the performance of the first AI model is superior to that of the second AI model, or the first indication is used to indicate that the performance of the second AI model is superior to that of the first AI model;

[0037] The difference in performance between the first AI model and the second AI model; or,

[0038] A model identifier, where the model identifier is the identifier of the AI model with the optimal performance among the first AI model and the second AI model.

[0039] In this possible implementation, the second device can instruct the first device to perform different operations through different contents of the first information. Thus, it can be seen that the second device can assist the first device in calculating the processing decision of the AI model, reducing the computing pressure on the first device.

[0040] In a possible implementation, the method further includes: the first device determines operations on the first AI model and the second AI model according to the first information.

[0041] In this possible implementation, the operations on the first AI model and the second AI model may include maintaining the activation state of the first AI model, or deactivating the first AI model; activating the second AI model, stopping using the second AI model to transmit the second data, etc. Thus, it can be seen that the first device can improve the decision-making speed for processing the first AI model and the second AI model according to the first information.

[0042] In a possible implementation, the above step: using the second AI model to transmit the second data to the second device includes: when the performance of the first AI model is less than or equal to a first threshold, the first device uses the second AI model to transmit the second data.

[0043] In this possible implementation, when the performance of the first AI model is less than or equal to the first threshold, the second AI model is only activated to transmit the second data. If the performance of the first AI model is greater than the first threshold, there is no need to activate the second AI model to transmit the second data. In this way, the utilization rate of transmission resources can be improved.

[0044] In a possible implementation, the method further includes:

[0045] The first device obtains the performance P of the first AI model a and the performance P of the second AI model i ;

[0046] When the first threshold > P a > the second threshold, and P iWhen the second threshold is reached, keep the first AI model activated and keep the second AI model for data transmission;

[0047] When P a < the second threshold, and P i > the second threshold, deactivate the first AI model and activate the second AI model;

[0048] When the first threshold > P a > the second threshold, and P i > the first threshold, deactivate the first AI model and activate the second AI model;

[0049] When P a < the second threshold, and P i < the second threshold, deactivate the first AI model and fallback to the non-AI transmission mode.

[0050] In this possible implementation, through the dual-threshold setting, the acquisition of the performance metrics corresponding to the non-activated model can be advanced. When the performance of the second AI model is poor, the model can be processed quickly, which can reduce the probability of data communication interruption.

[0051] The second aspect of this application provides a communication method applied to a second device communicating with a first device. The method includes: the second device receives the first data transmitted by the first device using the first artificial intelligence (AI) model, and receives the second data transmitted by the first device using the second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data; the second device sends a first message to the first device, and the first message is related to the first performance metric and the second performance metric; wherein, the first performance metric is the performance metric when the first AI model is used to transmit the first data; the second performance metric is the performance metric when the second AI model is used to transmit the second data.

[0052] In this application, the second device can be a communication device or a communication device capable of supporting the functions required for the communication device to implement this communication method, such as a chip. Exemplarily, the second device is a terminal device / network device, or a chip provided in the terminal device / network device for implementing the functions of the terminal device / network device, or other components for implementing the functions of the terminal device / network device. In the following introduction, it is described by taking the second device as a terminal device / network device as an example.

[0053] In a possible implementation, the transmission resources of the first data and the transmission resources of the second data are frequency-division multiplexed transmission resources, time-division multiplexed transmission resources, or independent transmission resources; wherein, the independent transmission resources mean that the transmission resources of the first data and the transmission resources of the second data are different in both the time domain and the frequency domain.

[0054] In a possible implementation, the transmission resources for the first data and the transmission resources for the second data are indicated by first control information.

[0055] In a possible implementation, the method further includes: the second device sending a first HARQ and / or a second HARQ, where the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0056] In a possible implementation, the first HARQ and the second HARQ are used to determine a third HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data.

[0057] In a possible implementation, the method further includes: the second device determining a third HARQ based on the first HARQ and the second HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data; the second device sending the third HARQ.

[0058] In a possible implementation, the first performance metric is determined by the first HARQ.

[0059] In a possible implementation, the second performance metric is determined by the second HARQ.

[0060] In a possible implementation, the first performance metric is obtained within a first time period, the second performance metric is obtained within a second time period, and the first time period and the second time period fully overlap or partially overlap.

[0061] In a possible implementation, the first information includes at least one of the following:

[0062] The first performance metric and the second performance metric;

[0063] A first indication, where the first indication is used to indicate that the performance of the first AI model is better than the performance of the second AI model, or the first indication is used to indicate that the performance of the second AI model is better than the performance of the first AI model;

[0064] The difference between the performance of the first AI model and the performance of the second AI model; or,

[0065] A model identifier, where the model identifier is the identifier of the AI model with the best performance among the first AI model and the second AI model.

[0066] A third aspect of the present application provides a communication device, and the communication device includes:

[0067] A processing module, configured to transmit first data to a second device using a first artificial intelligence (AI) model and transmit second data to the second device using a second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data.

[0068] A transceiver module, configured to receive first information from the second device, where the first information is related to a first performance metric and a second performance metric; wherein, the first performance metric is the performance metric when the first AI model is used to transmit the first data; and the second performance metric is the performance metric when the second AI model is used to transmit the second data.

[0069] In a possible implementation, the transmission resources for the first data and the second data are frequency-division multiplexed transmission resources, time-division multiplexed transmission resources, or independent transmission resources; wherein, the independent transmission resources mean that the transmission resources for the first data and the second data are different in both the time domain and the frequency domain.

[0070] In a possible implementation, the transmission resources for the first data and the second data are indicated by first control information.

[0071] In a possible implementation, the transceiver module is further configured to receive a first Hybrid Automatic Repeat reQuest (HARQ) and / or a second HARQ, where the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0072] In a possible implementation, the first HARQ and the second HARQ are used to determine a third HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data.

[0073] In a possible implementation, the transceiver module is further configured to receive the third HARQ determined by the second device based on the first HARQ and the second HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0074] In a possible implementation, the first performance metric is determined by the first HARQ.

[0075] In a possible implementation, the second performance metric is determined by the second HARQ.

[0076] In a possible implementation, the first performance metric is obtained within a first time period, the second performance metric is obtained within a second time period, and the first time period and the second time period are all overlapping or partially overlapping.

[0077] In a possible implementation, the first information includes at least one of the following:

[0078] The first performance indicator and the second performance indicator;

[0079] The first indication, which is used to indicate that the performance of the first AI model is better than that of the second AI model, or the first indication is used to indicate that the performance of the second AI model is better than that of the first AI model;

[0080] The difference between the performance of the first AI model and the performance of the second AI model; or,

[0081] The model identifier, which is the identifier of the AI model with the best performance among the first AI model and the second AI model.

[0082] In a possible implementation, the processing module is further configured to determine operations on the first AI model and the second AI model according to the first information.

[0083] In a possible implementation, the processing module is specifically configured to use the second AI model to transmit the second data when the performance of the first AI model is less than or equal to the first threshold.

[0084] In a possible implementation, the processing module is further configured to: obtain the performance P of the first AI model a and the performance P of the second AI model i ;

[0085] When the first threshold > P a > the second threshold, and P i < the second threshold, keep the first AI model in the active state and keep the second AI model for data transmission;

[0086] When P a < the second threshold, and P i > the second threshold, deactivate the first AI model and activate the second AI model;

[0087] When the first threshold > P a > the second threshold, and P i > the first threshold, deactivate the first AI model and activate the second AI model;

[0088] When P a < the second threshold, and P i < the second threshold, deactivate the first AI model and fallback to the non-AI transmission mode.

[0089] A fourth aspect of this application provides a communication device, which includes:

[0090] A transceiver module, configured to receive first data transmitted by a first device using a first artificial intelligence (AI) model and second data transmitted by the first device using a second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data.

[0091] The transceiver module is further configured to send a first message to the first device, where the first message is related to a first performance metric and a second performance metric; wherein, the first performance metric is the performance metric of the first AI model when transmitting the first data; the second performance metric is the performance metric of the second AI model when transmitting the second data.

[0092] In a possible implementation, the transmission resources of the first data and the second data are frequency-division multiplexed transmission resources, time-division multiplexed transmission resources, or independent transmission resources; wherein, the independent transmission resources mean that the transmission resources of the first data and the second data are different in both the time domain and the frequency domain.

[0093] In a possible implementation, the transmission resources of the first data and the second data are indicated by a first control message.

[0094] In a possible implementation, the transceiver module is further configured to send a first Hybrid Automatic Repeat reQuest (HARQ) or / and a second HARQ, where the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0095] In a possible implementation, the first HARQ and the second HARQ are used to determine a third HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data.

[0096] In a possible implementation, a processing module is configured to determine a third HARQ based on the first HARQ and the second HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0097] The transceiver module is further configured to send the third HARQ.

[0098] In a possible implementation, the first performance metric is determined by the first HARQ.

[0099] In a possible implementation, the second performance metric is determined by the second HARQ.

[0100] In a possible implementation, the first performance metric is obtained within a first time period, the second performance metric is obtained within a second time period, and the first time period and the second time period are either fully overlapping or partially overlapping.

[0101] In one possible implementation, the first information includes at least one of the following:

[0102] The first performance metric and the second performance metric;

[0103] The first indication, where the first indication is used to indicate that the performance of the first AI model is better than that of the second AI model, or the first indication is used to indicate that the performance of the second AI model is better than that of the first AI model;

[0104] The difference between the performance of the first AI model and the performance of the second AI model; or,

[0105] The model identifier, where the model identifier is the identifier of the AI model with the best performance among the first AI model and the second AI model.

[0106] A fifth aspect of the present application provides a communication device, which includes a processor. The processor is used to call and run a computer program stored in a memory, so that the processor implements any implementation manner in the first aspect or the first aspect.

[0107] Optionally, the communication device further includes a transceiver; the processor is further used to control the transceiver to send and receive signals.

[0108] Optionally, the communication device includes a memory, and a computer program is stored in the memory.

[0109] A sixth aspect of the present application provides a communication device, which includes a processor. The processor is used to call and run a computer program stored in a memory, so that the processor implements any implementation manner in the second aspect or the second aspect.

[0110] Optionally, the communication device further includes a transceiver; the processor is further used to control the transceiver to send and receive signals.

[0111] Optionally, the communication device includes a memory, and a computer program is stored in the memory.

[0112] The communication device described in the above fifth aspect to the sixth aspect may be a device or a chip (system) in a device.

[0113] A seventh aspect of the present application provides a computer program product including instructions, characterized in that when it runs on a computer, it causes the computer to execute any implementation manner in the first aspect or the first aspect.

[0114] An eighth aspect of the present application provides a computer program product including instructions, characterized in that when it runs on a computer, it causes the computer to execute any implementation manner in the second aspect or the second aspect.

[0115] The ninth aspect of this application provides a computer-readable storage medium, including computer instructions, which when running on a computer, cause the computer to execute the implementation of the first aspect or any one of the implementation manners in the first aspect.

[0116] The tenth aspect of this application provides a computer-readable storage medium, including computer instructions, which when running on a computer, cause the computer to execute the implementation of the second aspect or any one of the implementation manners in the second aspect.

[0117] The eleventh aspect of this application provides a chip device, including a processor for connecting to a memory and calling a program stored in the memory, so that the processor executes the implementation of the first aspect or any one of the implementation manners in the first aspect.

[0118] The twelfth aspect of this application provides a chip device, including a processor for connecting to a memory and calling a program stored in the memory, so that the processor executes the implementation of the second aspect or any one of the implementation manners in the second aspect.

[0119] The thirteenth aspect of this application provides a communication system, which includes a first device and a second device. The first device can be the communication device described in the third aspect, the fifth aspect or any one of the implementation manners thereof, and the second device can be the communication device described in the fourth aspect, the sixth aspect or any one of the implementation manners thereof.

[0120] Regarding the technical effects of the second aspect and any one of the implementation manners of the second aspect, as well as the technical effects of the third to twelfth aspects, reference can be made to the technical effects of the first aspect and any one of the implementation manners of the first aspect for understanding. Description of the Drawings

[0121] Figure 1A is a schematic structural diagram of a communication system provided by an embodiment of this application;

[0122] Figure 1B is another schematic structural diagram of a communication system provided by an embodiment of this application;

[0123] Figure 2 is a schematic diagram of an embodiment of a communication method provided by an embodiment of this application;

[0124] Figure 3 is a schematic diagram of an example of FDM resource indication provided by an embodiment of this application;

[0125] Figure 4 is a schematic diagram of an example of TDM resource indication provided by an embodiment of this application;

[0126] Figure 5 is a schematic diagram of an example of independent resource indication provided by an embodiment of this application;

[0127] Figure 6A It is an exemplary schematic diagram of HARQ feedback provided by an embodiment of the present application;

[0128] Figure 6B It is another exemplary schematic diagram of HARQ feedback provided by an embodiment of the present application;

[0129] Figure 7A It is another exemplary schematic diagram of HARQ feedback provided by an embodiment of the present application;

[0130] Figure 7B It is another exemplary schematic diagram of HARQ feedback provided by an embodiment of the present application;

[0131] Figure 8A It is another exemplary schematic diagram of HARQ feedback provided by an embodiment of the present application;

[0132] Figure 8B It is another exemplary schematic diagram of HARQ feedback provided by an embodiment of the present application;

[0133] Figure 9 It is an exemplary schematic diagram of collecting and transmitting data with equal time intervals provided by an embodiment of the present application;

[0134] Figure 10 It is an exemplary schematic diagram of collecting and transmitting data with unequal time intervals provided by an embodiment of the present application;

[0135] Figure 11 It is a schematic structural diagram of a communication device provided by an embodiment of the present application;

[0136] Figure 12 It is another schematic structural diagram of a communication device provided by an embodiment of the present application;

[0137] Figure 13 It is another schematic structural diagram of a communication device provided by an embodiment of the present application;

[0138] Figure 14 It is another schematic structural diagram of a communication device provided by an embodiment of the present application;

[0139] Figure 15 It is another schematic structural diagram of a communication device provided by an embodiment of the present application. Detailed implementation manners

[0140] Next, with reference to the accompanying drawings, embodiments of the present application will be described. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Those of ordinary skill in the art can understand that with the development of technology and the emergence of new scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

[0141] In the description and claims of this application and the above-mentioned drawings, the terms "first", "second", etc. are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "comprising" and "having" 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 does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0142] An embodiment of this application provides a communication method for obtaining the performance of a standby AI model. This application also provides corresponding communication devices, systems, computer-readable storage media, computer program products, etc. Details will be described separately below.

[0143] The technical solution of the embodiment of this application can be applied to various communication systems, such as: satellite communication, fifth-generation (5G) system or new radio (NR), long term evolution (LTE) system, LTE frequency division duplex (FDD) system, LTE time division duplex (TDD), universal mobile telecommunication system (UMTS), mobile communication system after 5G network (for example, 6G mobile communication system), vehicle to everything (V2X) communication system, etc.

[0144] Figure 1A It is a schematic structural diagram of a communication system provided by an embodiment of this application.

[0145] As Figure 1A shown, the communication system applicable to this application includes a first device and a second device. The first device and the second device can be devices or chips (systems) in the devices. When the first device or the second device is a device, the first device can be a network device or a terminal device, and the second device can also be a network device or a terminal device. When the first device or the second device is a chip (system), the first device can be a chip (system) in a network device or a terminal device, and the second device can be a chip (system) in a network device or a terminal device.

[0146] From the above description, it can be seen that the solution provided by the embodiment of this application can be applied to such asFigure 1B The communication system shown

[0147] As Figure 1B shown, the communication system includes network device 101, network device 102, terminal devices 103, 104, 105, 106, and terminal device 107. Among them, network device 101 and network device 102 can communicate via a backhaul link, which can be a wired backhaul link (such as optical fiber, copper cable) or a wireless backhaul link (such as microwave).

[0148] Both network device 101 and network device 102 can provide wireless communication services for the terminal devices within their coverage areas. For example, network device 101 provides wireless communication services for terminal devices 103, 104, and 105, and network device 102 provides wireless communication services for terminal devices 105, 106, and 107. Among them, terminal device 105 is in the overlapping area of network device 101 and network device 102. Therefore, terminal device 105 can communicate with both network device 101 and network device 102. Network device 101 / network device 102 can send downlink control information (DCI) to the terminal devices within their coverage areas to control the communication between the network device and the terminal devices through DCI. The terminal devices can send uplink control information (UCI) to the network device to indicate the communication between the network device and the terminal device through UCI.

[0149] The terminal devices can communicate with each other via a sidelink (SL). For example, wireless communication can be carried out between terminal device 103 and terminal device 104, between terminal device 104 and terminal device 105, between terminal device 105 and terminal device 106, and between terminal device 106 and terminal device 107. The terminal devices can send sidelink control information (SCI) during communication to control the communication between the terminal devices through SCI.

[0150] It should be noted that the network devices and terminal devices in the above communication system are only illustrative and should not be construed as limiting the number of network devices and terminal devices.

[0151] Next, the terminal devices and network devices of the present application will be introduced.

[0152] The terminal device can be a wireless terminal device capable of receiving scheduling and indication information from a network device. The wireless terminal device can be a device that provides voice and / or data connectivity to a user, or a handheld device with wireless connection capabilities, or other processing devices connected to a wireless modem.

[0153] The terminal device, also known as user equipment (UE), mobile station (MS), mobile terminal (MT), etc., is a device that includes wireless communication capabilities (providing voice / data connectivity to a user). For example, it can be a handheld device with wireless connection capabilities, or a vehicle-mounted device, etc. Currently, some examples of terminal devices are: mobile phone, tablet computer, laptop computer, palmtop computer, mobile internet device (MID), wearable device, virtual reality (VR) device, augmented reality (AR) device, wireless terminal in industrial control, wireless terminal in vehicle-to-everything (V2X), wireless terminal in self-driving, wireless terminal in remote medical surgery, wireless terminal in smart grid, wireless terminal in transportation safety, wireless terminal in smart city, or wireless terminal in smart home, etc. For example, the wireless terminal in vehicle-to-everything (V2X) can be a vehicle-mounted device, a whole vehicle device, a vehicle-mounted module, a vehicle, etc. The wireless terminal in industrial control can be a camera, a robot, etc. The wireless terminal in smart home can be a TV, an air conditioner, a floor sweeper, a speaker, a set-top box, etc.

[0154] The network device can be a device in a wireless network. For example, the network device is a device deployed in a radio access network to provide wireless communication capabilities for terminal devices. For example, the network device can be a radio access network (RAN) node that connects a terminal device to a wireless network, and can also be called an access network device.

[0155] Network devices include, but are not limited to: evolved Node B (eNB), radio network controller (RNC), Node B (NB), base station controller (BSC), base transceiver station (BTS), home base station (e.g., home evolved Node B, or home Node B, HNB), baseband unit (BBU), access point (AP) in a wireless fidelity (WIFI) system, wireless relay node, wireless backhaul node, transmission point (TP), or transmission and reception point (TRP), etc. It can also be a network device in a 5G mobile communication system. For example, the next generation Node B (gNB) in a new radio (NR) system, transmission reception point (TRP), transmission point (TP); or one or a group (including multiple antenna panels) of antenna panels of a base station in a 5G mobile communication system; or the network device can also be a network node that constitutes a gNB or a transmission point. For example, baseband unit (BBU), or distributed unit (DU), etc.

[0156] In some deployments, a gNB may include a centralized unit (CU) and a DU. The gNB may also include an active antenna unit (AAU). The CU implements some functions of the gNB, and the DU implements some functions of the gNB. For example, the CU is responsible for processing non-real-time protocols and services, and implementing the functions of the radio resource control (RRC) and packet data convergence protocol (PDCP) layers. The DU is responsible for processing physical layer protocols and real-time services, and implementing the functions of the radio link control (RLC) layer, media access control (MAC) layer, and physical (PHY) layer. The AAU implements some physical layer processing functions, radio frequency processing, and functions related to active antennas. The information of the RRC layer will ultimately become the information of the PHY layer, or vice versa. Therefore, in this architecture, high-layer signaling (such as RRC layer signaling) can also be considered to be sent by the DU, or by the DU and the AAU. It can be understood that the network device may be a device including one or more of the CU node, DU node, and AAU node. In addition, the CU may be classified as a network device in the radio access network (RAN), or the CU may be classified as a network device in the core network (CN), and this application does not make any limitations in this regard.

[0157] To facilitate the understanding of the embodiments of this application, the following first briefly introduces the terms involved in this application.

[0158] 1. Artificial intelligence (AI) model: used to implement corresponding AI functions. The AI model may be configured based on one or more of the following parameters: structural parameters (such as at least one of the number of neural network layers, neural network width, connection relationship between layers, weights of neurons, activation functions of neurons, or biases in the activation functions), input parameters (such as the type and / or dimension of the input parameters), or output parameters (such as the type and / or dimension of the output parameters). Among them, the bias in the activation function may also be referred to as the bias of the neural network.

[0159] A model can infer an output, and the output includes one parameter or multiple parameters. The learning process, training process, or inference process of different models may be deployed in different nodes or devices, or may be deployed in the same node or device.

[0160] The neural network of the AI model can be a neural network composed of an Embedding layer and a multilayer perception (MLP), or a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a residual network, or other neural networks, etc.

[0161] 2. Activating the AI model: It refers to the AI model that is in an active state and has been officially used.

[0162] 3. Non-activated AI model: It refers to the AI model that has not been officially activated and is in a standby state, but this non-activated model can also be called to transmit data.

[0163] 4. Time-division multiplexing (TDM): TDM is a working mode in a communication system. A piece of resource can be used for different purposes at different times. For example, it is used to send signals for a period of time and to receive signals for another period of time.

[0164] 5. Frequency-division multiplexing (FDM): FDM is a working mode in a communication system. A piece of resource is divided into two parts. Each of the divided resources is used for different purposes. For example, the first half of the frequency band is used to send signals, and the second half of the frequency band is used to receive signals.

[0165] 6. Hybrid automatic repeat request (HARQ): It is to introduce a forward error correction (FEC) code into an ARQ system. This FEC can be used to correct data errors during transmission. That is, if the error is within the error correction range of the FEC, then the FEC corrects the error. If it exceeds its error correction range, then a retransmission is requested. Usually, after receiving data, the receiving end will feedback an acknowledgement character (ACK) or a negative acknowledgement character (NACK) to the sending end. Among them, ACK indicates that the data transmission is successful, and NACK indicates that the data transmission is not successful and retransmission is required.

[0166] 7. Downlink control information (DCI): It is carried by the physical downlink control channel (PDCCH) of the downlink physical control channel. The downlink control information sent by the network device to the terminal device includes uplink and downlink resource allocation, HARQ information, power control, etc.

[0167] 8. Uplink control information (DCI): Control information for communication between a terminal device and a network device.

[0168] 9. Sidelink control information (SCI): Control information for communication between terminal devices.

[0169] 10. System throughput: Refers to the amount of data blocks transmitted in the system when using an AI model to transmit data, or the amount of data blocks transmitted per unit time.

[0170] 11. Block error ratio (BLER): Refers to the ratio of the number of data blocks with errors to the total number of data blocks received by a digital circuit.

[0171] 12. Signal interference noise ratio (SINR): Refers to the ratio of the intensity of the received useful signal to the intensity of the received interference signals (noise and interference); can be abbreviated as "signal-to-interference-plus-noise ratio" or "signal-to-noise ratio".

[0172] 13. Reference signal receive power (RSRP): Can be calculated through the following relationship, RSRP = PRS * PathLoss, where the power reference signal (PRS) represents the linear average of the transmit power of each reference signal receive element (RSRE) of the corresponding cell reference signal on two time slots within the system receive bandwidth; PathLoss represents the path loss between the network device and the terminal device.

[0173] The architecture of the communication system is introduced above. Next, the communication method provided by the embodiments of the present application is introduced.

[0174] As Figure 2 shown, the communication method provided by the embodiments of the present application includes:

[0175] 201. The first device uses a first artificial intelligence (AI) model to transmit first data to the second device. Correspondingly, the second device receives the first data transmitted through the first AI model.

[0176] 202. The first device uses a second AI model to transmit second data to the second device. Correspondingly, the second device receives the second data transmitted through the second AI model.

[0177] Among them, the first AI model is different from the second AI model, and the first data is the same as the second data.

[0178] In this application, the first AI model and the second AI model can be AI models with the same function, but the model structures, model parameters, computational overheads, transmission overheads, etc. of the two AI models may all be different or partially different. For example: both the first AI model and the second AI model are modulation and coding scheme (MCS) prediction models, but the model structures and model parameters of these two MCS prediction models are different.

[0179] In this application, the first AI model can be an active model, and the second AI model can be a non-active model. It can be seen that the embodiments of this application provide a solution for using an active model and a non-active model for the same data transmission, that is, when data is transmitted using the active model, data transmission of the non-active model is accompanied at the same time, and the active model and the non-active model are used to transmit the same data. For example, when the AI model is an MCS prediction model, the MCS prediction results of the active model and the MCS prediction results of the non-active model are simultaneously used for data transmission. The active model and the non-active model are used to transmit the same data (data1), the MCS prediction result of the active model is MCS1-1, and the MCS prediction result of the non-active model is MCS2-1. Then data1 can be transmitted using MCS1-1 respectively, and data1 can be transmitted using MCS2-1. The above transmission process can continue, and the active model and the non-active model can continue to transmit data data2 in a similar manner.

[0180] 203. The second device sends the first information to the first device. Correspondingly, the first device receives the first information from the second device.

[0181] The first information is related to the first performance metric and the second performance metric; wherein, the first performance metric is the performance metric when the first AI model is used to transmit the first data; the second performance metric is the performance metric when the second AI model is used to transmit the second data.

[0182] In this application, the first performance metric may include one or more of the system throughput, transmission block error rate, signal-to-interference-plus-noise ratio, or reference signal received power when transmitting the first data using the first AI model. Of course, the first performance metric may also include other parameters that can indicate the system transmission performance.

[0183] In this application, the second performance metric may include one or more of the system throughput, transmission block error rate, signal-to-interference-plus-noise ratio, or reference signal received power when transmitting the second data using the second AI model. Of course, the second performance metric may also include other parameters that can indicate the system transmission performance.

[0184] In this application, the first information may be the first performance indicator and the second performance indicator, or the result determined by the second device according to the first performance indicator and the second performance indicator.

[0185] In the technical solution provided by the embodiment of this application, by using the inactive second AI model to transmit the same data as the first AI model, the above-mentioned second performance indicator corresponding to the second AI model can be obtained. In this way, the accuracy of AI model switching during the communication process can be improved.

[0186] Optionally, after step 203, step 204 may further be included: the first device determines operations on the first AI model and the second AI model according to the first information.

[0187] Optionally, the transmission resources of the first data and the transmission resources of the second data, that is, the transmission resources corresponding to the first AI model (activated model) and the second AI model (inactive model) may include the transmission resources of FDM, the transmission resources of TDM, and independent transmission resources. Each type of transmission resource can be indicated by a control information, but the indication methods of the control information corresponding to each type of transmission resource are different. The following are introduced separately:

[0188] 1. Transmission resources of FDM;

[0189] The transmission resources of the first data and the transmission resources of the second data are the transmission resources of FDM, that is, the first data and the second data are transmitted using different frequency domain resources at the same time. When the transmission resources of the first data and the transmission resources of the second data are the transmission resources of FDM, the first control information can be indicated in the way of starting position + offset value. For example, the starting position of the transmission resources of the second data in the frequency domain can be the ending position of the transmission resources of the first data in the frequency domain, and the offset value can be the frequency bandwidth occupied by the transmission resources of the second data, or the number of resource elements (RE), or the number of resource blocks (RB).

[0190] That is to say, transmitting data1 using the transmission resources of the first AI model and transmitting data1 using the transmission resources of the second AI model can use different frequency domain resources at the same time. The indication method of the control information in this case can be referred to Figure 3 for understanding.

[0191] Such as Figure 3As shown in the figure, control information 1 is used to indicate the transmission resources 301 of the first AI model for data1 and the transmission resources 302 of the second AI model. Among them, the transmission resources 301 of the first AI model and the transmission resources 302 of the second AI model are the same in the time domain and adjacent in the frequency domain. In control information 1, the end position of the frequency domain resources of the first AI model can be indicated as the start position of the frequency domain resources of the second AI model, and then the offset value of the frequency domain resources of the second AI model can be indicated, such as: bandwidth.

[0192] After transmitting data1, data2 can be transmitted continuously. The transmission resources 303 of the first AI model of data2 and the transmission resources 304 of the second AI model can be indicated by control information 2. Control information 2 and control information 1 are Figure 3 only different in the time domain, and the content in the frequency domain can be the same or different.

[0193] 2. Transmission resources of TDM;

[0194] The transmission resources of the first data and the transmission resources of the second data are the transmission resources of TDM, that is, the first data and the second data are transmitted using the same frequency domain resources at different times. When the transmission resources of the first data and the transmission resources of the second data are the transmission resources of TDM, the first control information can also be indicated in the way of start position + offset value. For example, the start position of the transmission resources of the second data in the time domain can be the end position of the transmission resources of the first data in the time domain, and the offset value can be the number of time domain units occupied by the transmission resources of the second data. For example: it can be the number of time slots, the number of symbols, the number of mini (min) time slots, or the offset value is the absolute time length, such as in milliseconds, microseconds, etc.

[0195] That is to say, transmitting data1 using the transmission resources of the first AI model and transmitting data1 using the transmission resources of the second AI model can be transmitted at different times on the same frequency domain resources. The indication method of the control information in this case can refer to Figure 4 for understanding.

[0196] Such as Figure 4 As shown in the figure, control information 1 is used to indicate the transmission resources 401 of the first AI model for data1 and the transmission resources 402 of the second AI model. Among them, the transmission resources 401 of the first AI model and the transmission resources 402 of the second AI model are the same in the frequency domain and adjacent in the time domain. In control information 1, the end position of the time domain resources of the first AI model can be used as the start position of the time domain resources of the second AI model, and then the offset value of the time domain resources of the second AI model can be indicated, such as: the number of time slots.

[0197] After transmitting data1, data2 can be transmitted continuously. The transmission resources 403 of the first AI model and the transmission resources 404 of the second AI model for data2 can be indicated by control information 2. Control information 2 and control information 1 only have different starting times in the time domain, and the content in the frequency domain can be the same or different. Figure 4 in the time domain.

[0198] 3. Independent transmission resources;

[0199] The transmission resources for the first data and the transmission resources for the second data are independent transmission resources, that is, the first data and the second data are transmitted using different frequency domain resources at different times. When the transmission resources for the first data and the transmission resources for the second data are independent transmission resources, the first control information can directly indicate the two transmission resources and indicate the model type corresponding to each transmission resource, that is, an active model or a non-active model.

[0200] That is to say, transmitting data1 using the transmission resources of the first AI model and transmitting data1 using the transmission resources of the second AI model can be transmitted on different frequency domain resources and different time domain resources. For the indication method of the control information in this case, reference can be made to Figure 5 for understanding.

[0201] Such as Figure 5 shown, control information 1 is used to indicate the transmission resources 501 of the first AI model and the transmission resources 502 of the second AI model for data1. Among them, the transmission resources 501 of the first AI model and the transmission resources 502 of the second AI model are different in the frequency domain and different in the time domain. In control information 1, the starting positions and offset values of the transmission resources 501 of the first AI model in the time domain and the frequency domain can be indicated respectively, and the starting positions and offset values of the transmission resources 502 of the second AI model in the time domain and the frequency domain can be indicated.

[0202] After transmitting data1, data2 can be transmitted continuously. The transmission resources 503 of the first AI model and the transmission resources 504 of the second AI model for data2 can be indicated by control information 2. Control information 2 can indicate the starting positions and offset values of the transmission resources 503 of the first AI model in the time domain and the frequency domain respectively, and indicate the starting positions and offset values of the transmission resources 504 of the second AI model in the time domain and the frequency domain.

[0203] Since both active models and non-active models are used to transmit the same data simultaneously, the embodiments of the present application provide various forms of HARQ feedback. They are introduced separately below:

[0204] 1. Feedback the first HARQ / second HARQ;

[0205] In this application, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data. The first HARQ can be used to indicate whether the first data is transmitted successfully, and the second HARQ can be used to indicate whether the second data is transmitted successfully. Since the first data and the second data are the same data, only the first HARQ or the second HARQ can be used to determine whether one of the data transmissions is successful.

[0206] When the first performance metric or the second performance metric is not related to HARQ, for example: when the first performance metric or the second performance metric is an indicator for energy detection, such as: SINR or / and RSRP, the second device does not need to feedback HARQ to the first device. In this case, the feedback is taken as only feedbacking the first HARQ for example, and reference can be made to Figure 6A for understanding.

[0207] As Figure 6A shown, for the transmission of data1 using the first AI model, the second device needs to feedback HARQ1 to the first device, and this HARQ1 is the feedback for data transmission, that is, the feedback of whether the second device successfully receives data1.

[0208] When the first performance metric or the second performance metric is related to HARQ, for example: when the first performance metric or the second performance metric is an indicator for the transmission capacity of the system, such as: throughput or / and BLER, if the second device determines the performance result of the transmission, in this case, the feedback is taken as only feedbacking the first HARQ for example, and reference can be made to Figure 6B for understanding.

[0209] As Figure 6B shown, for the transmission of data1 using the first AI model, the second device needs to feedback HARQ1 to the first device, and this HARQ1 is the feedback for data transmission, that is, the feedback of whether the second device successfully receives data1, and there is no need to feedback on the first performance metric and the second performance metric.

[0210] 2. Feedback the first HARQ and the second HARQ;

[0211] In this application, when the first performance metric or the second performance metric is related to HARQ, for example: when the first performance metric or the second performance metric is an indicator for the transmission capacity of the system, such as: throughput or / and BLER, there are the following two cases for the feedback in this situation:

[0212] 2.1. When the first device determines the performance result of the transmission;

[0213] If the first device determines the performance result, the first device needs to determine the throughput and / or BLER based on the HARQ feedback. In this case, the HARQ corresponding to the first AI model and the HARQ corresponding to the second AI model both need to be fed back to the second device.

[0214] As Figure 7A shown, the feedback for data1 transmitted for the first AI model has a data transmission function and a performance collection function. That is to say, this HARQ1 can indicate whether data1 is successfully transmitted, and the first device can also determine the first performance metric through this HARQ1. The feedback for data1 transmitted for the second AI model has a performance collection function. That is to say, the first device can also determine the second performance metric through this HARQ2.

[0215] The second device can feedback HARQ1 and HARQ2 on the same resource. In this way, the control information sent by the second device to the first device only needs to indicate one HARQ feedback resource, and the first device and the second device only need to pre - agree on the feedback order of HARQ1 corresponding to the first AI model and HARQ2 corresponding to the second AI model. Of course, HARQ1 and HARQ2 can also use independent feedback resources. In this case, the control information sent by the second device to the first device needs to indicate two HARQ feedback resources.

[0216] 2.2. The first device determines the third HARQ;

[0217] The third HARQ is the HARQ corresponding to the first data and the second data.

[0218] When the first device determines the third HARQ, the second device needs to feedback the first HARQ and the second HARQ to the first device, and then the first device determines the third HARQ based on the first HARQ and the second HARQ. That is to say, in this case, the third HARQ used to indicate data transmission is jointly determined by the transmission situation of the first AI model and the transmission situation of the second AI model. The transmission of the second AI model can be understood as a re - transmission of the transmission of the first AI model. In this way, when the transmission using the first AI model fails, the transmission of the second AI model can be jointly used for decoding, thereby improving the decoding success rate.

[0219] The determination idea of the third HARQ can be: if both the first HARQ and the second HARQ are successfully received (ACK), then the third HARQ is ACK. If the first HARQ is ACK and the second HARQ is unsuccessfully received (NACK), then the combined third HARQ is ACK; if the first HARQ is NACK and the HARQ of the second HARQ is ACK, then the third HARQ is ACK; if both the first HARQ and the second HARQ are NACK, then the first HARQ and the second HARQ are jointly decoded, and the third HARQ is determined to be ACK or NACK according to the decoding result.

[0220] This process can refer to Figure 7B for understanding. As Figure 7B shown, the feedback for data1 transmitted by the first AI model has a data transmission function and a performance collection function. That is to say, this HARQ1 can indicate whether data1 is successfully transmitted through the first AI model, and the first device can also determine the first performance index through this HARQ1. The feedback for data1 transmitted by the second AI model has a data transmission function and a performance collection function. That is to say, this HARQ2 can indicate whether data1 is successfully transmitted through the second AI model, and the first device can also determine the second performance index through this HARQ2.

[0221] After the second device feeds back HARQ1 and HARQ2 to the first device, the first device determines HARQ3 according to HARQ1 and HARQ2. The first device can determine whether data1 is successfully transmitted through HARQ3.

[0222] Regarding the feedback resources of HARQ1 and HARQ2 and the indication method for the feedback resources, you can refer to the Figure 7A corresponding introduction for details.

[0223] 3. The first device feeds back the third HARQ to the second device;

[0224] 3.1 The first performance index or the second performance index has nothing to do with HARQ;

[0225] When the second device determines the third HARQ, the second device determines the third HARQ according to the first HARQ and the second HARQ, and then the second device feeds back the third HARQ to the first device. The idea of the second device determining the third HARQ can be understood by referring to the idea of the first device determining the third HARQ mentioned above.

[0226] And when the first performance metric and / or the second performance metric is / are independent of HARQ, for example: when the first performance metric or the second performance metric is an indicator related to energy detection, such as: SINR or / and RSRP. The second device does not need to feedback the separate first HARQ and second HARQ to the first device.

[0227] This process can be referred to Figure 8A for understanding. As Figure 8A shown, the second device determines HARQ1 and HARQ2, then determines HARQ3 based on HARQ1 and HARQ2, and then feedbacks HARQ3 to the first device.

[0228] 3.2 The first performance metric or the second performance metric is related to HARQ;

[0229] The determination method of the third HARQ is the same as that described in 3.1, and can be referred to 3.1 for understanding.

[0230] In addition, when the first performance metric or the second performance metric is related to HARQ, for example: the first performance metric or the second performance metric is an indicator for representing the transmission capacity of the system, such as: throughput or / and BLER. If the first device determines the performance result, the first device needs to determine the throughput or / and BLER based on the HARQ feedback. In this case, the second device also needs to feedback the first HARQ and the second HARQ to the first device. This process can be referred to Figure 8B for understanding. As Figure 8B shown, the second device determines HARQ1 and HARQ2, then determines HARQ3 based on HARQ1 and HARQ2, and then feedbacks HARQ3 to the first device. The second device also needs to feedback HARQ1 and HARQ2 to the first device.

[0231] Regarding the above Figures 6A to 8B introduced HARQ-related schemes, the first performance metric can be directly the first HARQ, or indirectly determined by the first HARQ. When the first performance metric is directly the first HARQ, it can be that the first HARQ contains one or more of the system throughput, BLER, SINR, or RSRP; when the first performance metric is indirectly determined by the first HARQ, it can be that one or more of the system throughput, BLER, SINR, or RSRP are determined through the first HARQ. Similarly, the second performance metric can be directly the second HARQ, or indirectly determined by the second HARQ. Thus, it can be seen that the HARQ in this application can not only indicate data transmission, but also indicate the performance metrics of data transmission, enhancing the indication ability of HARQ.

[0232] In the embodiments of the present application, for the first performance indicator and the second performance indicator, the data acquisition time may be of equal duration or of unequal duration. For example, the first performance indicator is obtained within the first time period, and the second performance indicator is obtained within the second time period. When of equal duration, it means that the first time period and the second time period completely overlap, and when of unequal duration, it means that the first time period and the second time period partially overlap.

[0233] The case of equal duration can be referred to Figure 9 for understanding. For example, Figure 9 as shown, within a certain time period, M data (data1 to data M) are transmitted through the first AI model, where M is an integer greater than 1. Within this time period, M data (data1 to data M) are transmitted through the second AI model. Then, the first performance indicator can be determined by collecting the data transmitted by the system when the first AI model transmits M data, and the second performance indicator can be determined by collecting the data transmitted by the system when the second AI model transmits M data. The first performance indicator and the second performance indicator are obtained within the same time period, which can improve the reliability of the second performance indicator.

[0234] The case of unequal duration can be referred to Figure 10 for understanding. For example, Figure 10 as shown, within a certain time period, M data (data1 to data M) are transmitted through the first AI model, and within a partial time period of this time period, M data (data1 to data M) are transmitted through the second AI model. Then, the first performance indicator can be determined by collecting the data transmitted by the system when the first AI model transmits M data, and the second performance indicator can be determined by collecting the data transmitted by the system when the second AI model transmits M data. The time used to obtain the second performance indicator is shorter than the time used to obtain the first performance indicator, which can determine in advance the second AI (non-activated model), such as determining whether to continue using the second AI model to transmit the same data as the first AI model. Of course, the duration of collecting the transmission data of the second AI model can also be longer than the duration of collecting the transmission data of the first AI model, and this application does not make any limitations in this regard.

[0235] Within the duration of collecting the above transmission data, the second device can feedback the first performance indicator and the second performance indicator to the first device (such as one or more of throughput, BLER, SINR, RSRP). In this application, the first performance indicator and the second performance indicator can be in various possible forms such as the average performance, the optimal performance, and the difference between the optimal performance and the worst performance of each indicator within the duration of collecting the transmission data.

[0236] After the second device obtains the first performance indicator and the second performance indicator, it can determine the first information to be fed back to the first device, and the first information may include at least one of the following:

[0237] The first performance indicator and the second performance indicator;

[0238] The first indication, which is used to indicate that the performance of the first AI model is better than that of the second AI model, or the first indication is used to indicate that the performance of the second AI model is better than that of the first AI model;

[0239] The difference between the performance of the first AI model and the performance of the second AI model; or,

[0240] The model identifier, which is the identifier of the AI model with the best performance among the first AI model and the second AI model.

[0241] 1. The first information is the first performance indicator and the second performance indicator;

[0242] When the first information is the first performance indicator and the second performance indicator, the first device can determine the performance of the first AI model and the second AI model according to the first performance indicator and the second performance indicator. For example, when it can be determined according to the first performance indicator and the second performance indicator that the performance of the first AI model is better than that of the second AI model, the first AI model can continue to be used, and the second AI model can no longer be used to transmit the same data as the first AI model. It can also be determined according to the first performance indicator and the second performance indicator that the performance of the first AI model is worse than that of the second AI model, and the second AI model can be activated and the first AI model can be deactivated.

[0243] The first performance indicator and the second performance indicator can be various possible forms such as the average performance, the best performance, and the difference between the best performance and the worst performance of each indicator during the duration of data transmission collection.

[0244] 2. The first information is the first indication;

[0245] This first indication is used to indicate that the performance of the first AI model is better than that of the second AI model, or the first indication is used to indicate that the performance of the second AI model is better than that of the first AI model.

[0246] The second device can determine the processing of the first AI model and the second AI model according to this first indication. If the first indication is used to indicate that the performance of the first AI model is better than that of the second AI model, the first AI model can continue to be used, and the second AI model can no longer be used to transmit the same data as the first AI model. If the first indication is used to indicate that the performance of the second AI model is better than that of the first AI model, the second AI model can be activated and the first AI model can be deactivated.

[0247] 3. The first information is the difference between the performance of the first AI model and the performance of the second AI model;

[0248] When the first information is the difference between the performance of the first AI model and the performance of the second AI model, the second device can determine which of the performance of the first AI model and the performance of the second AI model is better based on this difference, and can also determine the difference in performance between the first AI model and the second AI model according to the magnitude of the difference. For example: if the difference is a positive number, it means that the performance of the first AI model is better than the performance of the second AI model; if the difference is a negative number, it means that the performance of the second AI model is better than the performance of the first AI model. Then, it can be determined how to handle the first AI model and the second AI model. The specific processing process can be understood by referring to the content where the first information is the first indication above.

[0249] 4. The first information is a model identifier;

[0250] The model identifier is the identifier of the AI model with the optimal performance among the first AI model and the second AI model.

[0251] In this case, if the first information is the identifier of the first AI model, it means that the performance of the first AI model is the best. Then, the first AI model can continue to be used, and the second AI model can be stopped from transmitting the same data as the first AI model. If the first information is the identifier of the second AI model, it means that the performance of the second AI model is the best. Then, the second AI model can be activated, and the first AI model can be deactivated.

[0252] In the embodiments of the present application, regarding the use of the second AI model to transmit the same data as the first AI model, it can be controlled by a threshold value. For example: when the performance of the first AI model is less than or equal to the first threshold, the second AI model is started to transmit the second data; if the performance of the first AI model is greater than the first threshold, the second AI model does not need to be started to transmit the second data. In this way, the utilization rate of transmission resources can be improved.

[0253] To better manage the first AI model and the second AI model, the embodiments of the present application provide a dual-threshold control mechanism. For example: through the first threshold T 1 and the second threshold T 2 to manage, where T 1 > T 2 .

[0254] For ease of understanding, hereinafter, the performance of the first AI model is represented by P a , and the performance of the second AI model is represented by P i . Then, the processing methods for the first AI model and the second AI model are described when the relationships between P a , P i and T 1 and T 2 are different.

[0255] Regarding the relationships among several parameters and the processing methods for the first AI model and the second AI model, please refer to Table 1 below for understanding.

[0256] Table 1

[0257]

[0258] As can be seen from Table 1 above, when P a > T 1 it indicates that the performance of the first AI model is very good, and there is no need to collect the performance of the second AI model. Then, the first AI model can continue to be used, and the second AI model can be stopped from transmitting the same data as the first AI model.

[0259] When T 2 < P a < T 1 & P i < T 1 it indicates that the performance of the first AI model is good, but the second AI model still needs to be used to facilitate the timely collection of the performance of the system-transmitted data and obtain the second performance indicator to facilitate the timely switching of the AI model.

[0260] When P a < T 2 & P i > T 2 it indicates that the performance of the first AI model is poor and the performance of the second AI model is good. At this time, the use of the first AI model can be stopped, the first AI model can be deactivated, and the second AI model can be activated to use the second AI model to transmit data.

[0261] When T 2 < P a < T 1 & P i > T 1 it indicates that the performance of the first AI model is good, but the performance of the second AI model is better. Then, the use of the first AI model can be stopped, the first AI model can be deactivated, and the second AI model can be activated to use the second AI model to transmit data.

[0262] When P a < T 2 & P i < T 2 it indicates that the performance of the first AI model is poor and the performance of the second AI model is also poor. At this time, the first AI model can be deactivated, the use of the second AI model can be stopped, and the system can be reverted to the non-AI transmission mode.

[0263] In the embodiments of the present application, the above performance may be the performance in a certain aspect. The units of the first threshold and the second threshold are related to the type of performance. For example, when the performance index is RSRP, the unit of the dual threshold is decibel (dB); when the performance index is throughput, the unit of the dual threshold is (kilobit per second) kbit / ms.

[0264] In the embodiments of the present application, through the dual-threshold setting, the acquisition of the performance index corresponding to the inactive model can be advanced. When the performance of the second AI model is poor, the model can be processed quickly, which can reduce the probability of data communication interruption.

[0265] It should be noted that there may be multiple second AI models introduced above. The processing for each second AI model can be understood by referring to the processing process of the second AI model introduced above.

[0266] The communication method has been introduced above. Next, the communication device provided by the embodiments of the present application will be described. Please refer to Figure 11 , Figure 11 which is a schematic structural diagram of the communication device according to the embodiments of the present application. The communication device 1100 can be used to execute Figures 2 to 10 the steps executed by the first device in the embodiments shown in

[0267] The communication device 1100 includes a transceiver module 1101 and a processing module 1102. The transceiver module 1101 can implement corresponding communication functions, and the processing module 1102 is used for data processing. The transceiver module 1101 can also be referred to as a communication interface or a communication unit.

[0268] Optionally, the communication device 1100 may further include a storage unit, which can be used to store instructions and / or data. The processing module 1102 can read the instructions and / or data in the storage unit to enable the communication device to implement the foregoing method embodiments.

[0269] The communication device 1100 can be used to execute the actions performed by the first device in the above method embodiments. The communication device 1100 can be the first device or a component configurable in the first device. The transceiver module 1101 is used to execute the receiving / sending related operations on the side of the first device in the above method embodiments, and the processing module 1102 is used to execute the processing related operations on the side of the first device in the above method embodiments.

[0270] Optionally, the transceiver module 1101 may include a sending module and a receiving module. The sending module is used to execute the sending operation in the above method embodiments. The receiving module is used to execute the receiving operation in the above method embodiments.

[0271] It should be noted that the communication device 1100 may include a sending module but not a receiving module. Alternatively, the communication device 1100 may include a receiving module but not a sending module. Specifically, it depends on whether the above-described solution executed by the communication device 1100 includes a sending action and a receiving action.

[0272] As an example, the communication device 1100 is used to execute the actions performed by the first device in the embodiment Figure 2 shown above.

[0273] The processing module 1102 is configured to transmit first data to a second device using a first artificial intelligence (AI) model and transmit second data to the second device using a second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data.

[0274] The transceiver module 1101 is configured to receive first information from the second device, where the first information is related to a first performance metric and a second performance metric; wherein, the first performance metric is the performance metric of the first AI model when transmitting the first data; the second performance metric is the performance metric of the second AI model when transmitting the second data.

[0275] Optionally, the transceiver module 1101 is further configured to receive a third Hybrid Automatic Repeat reQuest (HARQ) determined by the second device through a first HARQ and a second HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0276] Optionally, the processing module 1102 is further configured to determine operations on the first AI model and the second AI model according to the first information.

[0277] It should be understood that the specific processes of each module performing the above corresponding steps have been described in detail in the above method embodiments. For the sake of brevity, they will not be repeated here.

[0278] The processing module 1102 in the above embodiments may be implemented by at least one processor or processor-related circuits. The transceiver module 1101 may be implemented by a transceiver or transceiver-related circuits. The transceiver module 1101 may also be referred to as a communication unit or a communication interface. The storage unit may be implemented by at least one memory.

[0279] The communication device provided in the embodiments of the present application will be described below. Please refer to Figure 12 , Figure 12 which is a schematic structural diagram of the communication device in the embodiments of the present application. The communication device 1200 may be used to execute the steps performed by the second device in the embodiment Figures 2 to 10 shown above. For specific details, please refer to the relevant descriptions in the above method embodiments.

[0280] The communication device 1200 includes a transceiver module 1201. Optionally, the communication device 1200 further includes a processing module 1202. The transceiver module 1201 can implement corresponding communication functions, and the processing module 1202 is used for data processing. The transceiver module 1201 can also be referred to as a communication interface or a communication unit.

[0281] The communication device 1200 can be used to perform the actions executed by the second device in the above method embodiments. The communication device 1200 can be the second device or a component configurable in the second device. The transceiver module 1201 is used to perform the operations related to reception on the second device side in the above method embodiments.

[0282] Optionally, the transceiver module 1201 can include a sending module and a receiving module. The sending module is used to perform the sending operation in the above method embodiments. The receiving module is used to perform the receiving operation in the above method embodiments.

[0283] It should be noted that the communication device 1200 can include a sending module but not a receiving module. Or, the communication device 1200 can include a receiving module but not a sending module. Specifically, it depends on whether the above scheme executed by the communication device 1200 includes a sending action and a receiving action.

[0284] As an example, the communication device 1200 is used to perform the actions executed by the second device in the above Figure 2 illustrated embodiments.

[0285] The transceiver module 1201 is used to receive the first data transmitted by the first device using the first artificial intelligence (AI) model, and receive the second data transmitted by the first device using the second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data.

[0286] The transceiver module 1201 is further used to send the first information to the first device, and the first information is related to the first performance index and the second performance index; wherein, the first performance index is the performance index when the first AI model is used to transmit the first data; the second performance index is the performance index when the second AI model is used to transmit the second data.

[0287] Optionally, the transceiver module 1202 is further used to send the first hybrid automatic repeat request (HARQ) and / or the second HARQ, where the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

[0288] Optionally, a processing module 1202 is configured to determine a third HARQ according to a first HARQ and a second HARQ, where the third HARQ is the HARQ corresponding to first data and second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data;

[0289] A transceiver module 1202 is further configured to send the third HARQ.

[0290] It should be understood that the specific processes for each module to execute the above corresponding steps have been described in detail in the above method embodiments. For the sake of brevity, they will not be elaborated here.

[0291] The processing module 1202 in the above embodiments may be implemented by at least one processor or processor-related circuits. The transceiver module 1201 may be implemented by a transceiver or transceiver-related circuits. The transceiver module 1201 may also be referred to as a communication unit or a communication interface. The storage unit may be implemented by at least one memory.

[0292] An embodiment of the present application further provides a communication device 1300. The communication device 1300 includes a processor 1310, the processor 1310 is coupled to a memory 1320, the memory 1320 is configured to store computer programs or instructions and / or data, and the processor 1310 is configured to execute the computer programs or instructions and / or data stored in the memory 1320, so that the method in the above method embodiments is executed.

[0293] Optionally, the processor 1310 included in the communication device 1300 is one or more.

[0294] Optionally, as Figure 13 shown, the communication device 1300 may further include a memory 1320.

[0295] Optionally, the memory 1320 included in the communication device 1300 may be one or more.

[0296] Optionally, the memory 1320 may be integrated with the processor 1310 or separately provided.

[0297] Optionally, as Figure 13 shown, the communication device 1300 may further include a transceiver 1330, and the transceiver 1330 is configured to receive and / or send signals. For example, the processor 1310 is configured to control the transceiver 1330 to receive and / or send signals.

[0298] As a solution, the communication device 1300 is configured to implement the operations performed by the first device in the above method embodiments.

[0299] For example, the processor 1310 is used to implement the operations related to processing performed by the first device in the above method embodiments, and the transceiver 1330 is used to implement the operations related to transceiver performed by the first device in the above method embodiments.

[0300] As another solution, the communication device 1300 is used to implement the operations performed by the second device in the above method embodiments.

[0301] For example, the processor 1310 is used to implement the operations related to processing performed by the second device in the above method embodiments, and the transceiver 1330 is used to implement the operations related to transceiver performed by the second device in the above method embodiments.

[0302] An embodiment of the present application further provides a communication device 1400. The communication device 1400 may be the first device / second device or a chip thereof. The communication device 1400 may be used to perform the operations performed by the first device / second device in the above method embodiments.

[0303] When the communication device 1400 is the first device / second device, Figure 14 A simplified structural schematic diagram of the first device / second device is shown. As Figure 14 shown, the first device / second device includes a processor, a memory, and a transceiver. Computer program code and an AI module may be stored in the memory and / or the processor. The AI module is used to implement functions related to AI. The AI module may be implemented in a software, hardware, or a combination of software and hardware manner. For example, the AI module may include a RIC module. For example, the AI module may be a near real-time radio access network intelligent controller (RIC) or a non-real-time RIC. The transceiver includes a transmitter 1431, a receiver 1432, a radio frequency circuit (not shown in the figure), an antenna 1433, and an input / output device (not shown in the figure). The processor is mainly used to process communication protocols and communication data, control the first device / second device, execute software programs, process data of software programs, etc. The memory is mainly used to store software programs and data. The radio frequency circuit is mainly used for the conversion between baseband signals and radio frequency signals and the processing of radio frequency signals. The antenna is mainly used to transmit and receive radio frequency signals in the form of electromagnetic waves. The input / output device, such as a touch screen, a display screen, a keyboard, etc., is mainly used to receive data input by the user and output data to the user. It should be noted that some types of the first device / second device may not have an input / output device.

[0304] When data needs to be sent, after the processor performs baseband processing on the data to be sent, it outputs a baseband signal to the radio frequency circuit. The radio frequency circuit performs radio frequency processing on the baseband signal and then sends the radio frequency signal outwards in the form of electromagnetic waves through the antenna. When data is sent to the first device / second device, the radio frequency circuit receives the radio frequency signal through the antenna, converts the radio frequency signal into a baseband signal, and outputs the baseband signal to the processor. The processor converts the baseband signal into data and processes the data. For ease of explanation, Figure 14 only one memory, processor, and transceiver are shown. In an actual first device / second device product, there may be one or more processors and one or more memories. The memory may also be referred to as a storage medium or a storage device, etc. The memory may be set independently of the processor or integrated with the processor. The embodiments of the present application do not limit this.

[0305] In the embodiments of the present application, the antenna and the radio frequency circuit with transceiver functions can be regarded as the transceiver unit of the first device / second device, and the processor with processing functions can be regarded as the processing unit of the first device / second device.

[0306] As Figure 14 shown, the first device / second device includes a processor 1410, a memory 1420, and a transceiver 1430. The processor 1410 may also be referred to as a processing unit, a processing board, a processing module, a processing device, etc. The transceiver 1430 may also be referred to as a transceiver unit, a transceiver, a transceiver device, etc.

[0307] Optionally, the devices in the transceiver 1430 used to implement the receiving function can be regarded as the receiving unit, and the devices in the transceiver 1430 used to implement the sending function can be regarded as the sending unit, that is, the transceiver 1430 includes a receiver and a transmitter. The transceiver may sometimes also be referred to as a transceiver, a transceiver unit, or a transceiver circuit, etc. The receiver may sometimes also be referred to as a receiver, a receiving unit, or a receiving circuit, etc. The transmitter may sometimes also be referred to as a transmitter, a sending unit, or a transmitting circuit, etc.

[0308] For example, in one implementation, the processor 1410 is used to execute Figure 2 the processing actions on the side of the first device / second device in the shown embodiment, and the transceiver 1430 is used to execute Figure 2 the transceiver actions on the side of the first device / second device in Figure 2 For example, the transceiver 1430 is used to execute the transceiver operation of step 302 in the shown embodiment. The processor 1410 is used to execute Figure 2 the processing operation of step 303 in the shown embodiment.

[0309] It should be understood that Figure 14By way of example only and not limitation, the first device / second device including the transceiver unit and the processing unit described above may not depend on Figure 14 the structure shown.

[0310] When the communication device 1400 is a chip, the chip includes a processor and a transceiver. Among them, the transceiver may be an input / output circuit or a communication interface; the processor may be a processing unit integrated on the chip, a microprocessor, or an integrated circuit. The sending operation of the first device / second device in the above method embodiments can be understood as the output of the chip, and the receiving operation of the first device / second device in the above method embodiments can be understood as the input of the chip.

[0311] An embodiment of the present application further provides a communication device 1500, which may be the first device / second device or a chip. The communication device 1500 can be used to perform the operations performed by the first device / second device in the above method embodiments.

[0312] When the communication device 1500 is the first device / second device, for example, it is a base station. Figure 15 A simplified schematic diagram of the base station structure is shown. The base station includes a part 1510, a part 1520, and a part 1530. The part 1510 is mainly used for baseband processing and controlling the base station, etc.; the part 1510 is usually the control center of the base station and can usually be called a processor, which is used to control the base station to perform the processing operations on the first device / second device side in the above method embodiments. The part 1520 is mainly used for storing computer program codes and AI modules. The AI module is used to implement AI-related functions. The AI module can be implemented in a software, hardware, or a combination of software and hardware manner. For example, the AI module may include a RIC module. For example, the AI module may be a near-real-time RIC or a non-real-time RIC. The part 1530 is mainly used for transceiver of radio frequency signals and conversion between radio frequency signals and baseband signals; the part 1530 can usually be called a transceiver unit, a transceiver, a transceiver circuit, or a transceiver, etc. The transceiver unit of the part 1530 can also be called a transceiver or a transceiver, etc., and it includes an antenna 1533 and a radio frequency circuit (not shown in the figure), where the radio frequency circuit is mainly used for radio frequency processing. Optionally, the devices used to implement the receiving function in the part 1530 can be regarded as a receiver, and the devices used to implement the sending function can be regarded as a transmitter, that is, the part 1530 includes a receiver 1532 and a transmitter 1531. The receiver can also be called a receiving unit, a receiver, or a receiving circuit, etc., and the transmitter can be called a sending unit, a transmitter, or a transmitting circuit, etc.

[0313] The 1510 part and the 1520 part may include one or more single boards, and each single board may include one or more processors and one or more memories. The processor is used to read and execute the programs in the memory to implement baseband processing functions and control of the base station. If there are multiple single boards, they can be interconnected to enhance the processing capacity. As an alternative implementation, it is also possible that multiple single boards share one or more processors, or multiple single boards share one or more memories, or multiple single boards simultaneously share one or more processors.

[0314] For example, in one implementation, the transceiver unit of the 1530 part is used to execute Figure 2 the transceiver-related steps performed by the first device / second device in the illustrated embodiment. The processor of the 1510 part is used to execute Figure 2 the processing-related steps performed by the first device / second device in the illustrated embodiment.

[0315] It should be understood that Figure 15 merely by way of example and not limitation, the above-mentioned first device / second device including a processor, a memory, and a transceiver may not depend on Figure 15 the structure shown.

[0316] When the communication device 1500 is a chip, the chip includes a transceiver, a memory, and a processor. Among them, the transceiver may be an input / output circuit, a communication interface; the processor is a processor integrated on the chip, or a microprocessor, or an integrated circuit. The sending operation of the first device / second device in the above method embodiment can be understood as the output of the chip, and the receiving operation of the first device / second device in the above method embodiment can be understood as the input of the chip.

[0317] The embodiments of the present application also provide a computer-readable storage medium, on which computer instructions for implementing the method executed by the first device or the method executed by the second device in the above method embodiments are stored.

[0318] For example, when the computer program is executed by a computer, the computer can implement the method executed by the first device or the method executed by the second device in the above method embodiments.

[0319] The embodiments of the present application also provide a computer program product containing instructions, and when the instructions are executed by a computer, the computer implements the method executed by the first device or the method executed by the second device in the above method embodiments.

[0320] The embodiments of the present application also provide a communication system, which includes the second device and the first device in the above embodiments.

[0321] The embodiments of the present application further provide a chip device, including a processor, which is used to call the computer program or computer instructions stored in the memory, so that the processor executes the above Figures 2 to 10 method of the embodiment shown.

[0322] In a possible implementation, the input of the chip device corresponds to the receiving operation in the above Figures 2 to 10 shown embodiment, and the output of the chip device corresponds to the sending operation in the above Figures 2 to 10 shown embodiment.

[0323] Optionally, the processor is coupled to the memory through an interface.

[0324] Optionally, the chip device further includes a memory, in which computer programs or computer instructions are stored.

[0325] Wherein, the processor mentioned anywhere above can be a general-purpose central processing unit, a microprocessor, an application-specific integrated circuit (ASIC), or an integrated circuit for controlling the execution of the program of the method of the above Figures 2 to 10 shown embodiment. The memory mentioned anywhere above can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM), etc.

[0326] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the explanations and beneficial effects of the relevant content in any of the above communication devices can refer to the corresponding method embodiments provided above, and will not be elaborated here.

[0327] In the embodiments of the present application, a terminal device or a network device may include a hardware layer, an operating system layer running on the hardware layer, and an application layer running on the operating system layer. Among them, the hardware layer may include hardware such as a central processing unit (CPU), a memory management unit (MMU), and a memory (also called main memory). The operating system of the operating system layer can be any one or more computer operating systems that implement service processing through processes, for example, Linux operating system, Unix operating system, Android operating system, iOS operating system, or windows operating system, etc. The application layer may include applications such as a browser, an address book, a word processing software, and an instant messaging software.

[0328] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0329] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the couplings, direct couplings, or communication connections shown or discussed with each other can be indirect couplings or communication connections through some interfaces, devices, or units, and can be in electrical, mechanical, or other forms.

[0330] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0331] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0332] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the part that essentially contributes to the technical solution of the present application, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories, random access memories, magnetic disks, or optical discs that can store program codes.

Claims

1. A communication method, characterized in that, comprising: a first device transmits first data to a second device using a first artificial intelligence (AI) model, and transmits second data to the second device using a second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data; the first device receives first information from the second device, the first information being related to a first performance metric and a second performance metric; wherein, the first performance metric is the performance metric of the first AI model when transmitting the first data; the second performance metric is the performance metric of the second AI model when transmitting the second data.

2. The method according to claim 1, characterized in that, the transmission resources of the first data and the transmission resources of the second data are frequency-division multiplexed transmission resources, time-division multiplexed transmission resources, or independent transmission resources; wherein, the independent transmission resources mean that the transmission resources of the first data and the transmission resources of the second data are different in both the time domain and the frequency domain.

3. The method according to claim 2, characterized in that, the transmission resources of the first data and the second data are indicated by first control information.

4. The method according to any one of claims 1-3, characterized in that, the method further comprises: the first device receives a first hybrid automatic repeat request (HARQ) and / or a second HARQ, the first HARQ being the HARQ corresponding to the first data, and the second HARQ being the HARQ corresponding to the second data.

5. The method according to claim 4, characterized in that, the first HARQ and the second HARQ are used to determine a third HARQ, the third HARQ being the HARQ corresponding to the first data and the second data.

6. The method according to any one of claims 1-3, characterized in that, the method further comprises: the first device receives the third HARQ determined by the second device through the first HARQ and the second HARQ, the third HARQ being the HARQ corresponding to the first data and the second data, the first HARQ being the HARQ corresponding to the first data, and the second HARQ being the HARQ corresponding to the second data.

7. The method according to any one of claims 4-6, characterized in that, the first performance metric is determined by the first HARQ.

8. The method according to any one of claims 4-7, characterized in that, the second performance metric is determined by the second HARQ.

9. The method according to any one of claims 1-8, characterized in that, the first performance metric is obtained within a first time period, the second performance metric is obtained within a second time period, and the first time period and the second time period completely overlap or partially overlap.

10. The method according to any one of claims 1-9, characterized in that, the first information includes at least one of the following: the first performance metric and the second performance metric; A first indication, where the first indication is used to indicate that the performance of the first AI model is superior to the performance of the second AI model, or the first indication is used to indicate that the performance of the second AI model is superior to the performance of the first AI model; The difference between the performance of the first AI model and the performance of the second AI model; or, A model identifier, where the model identifier is the identifier of the AI model with the optimal performance among the first AI model and the second AI model.

11. The method according to any one of claims 1 - 10, characterized in that, the method further comprises: The first device determines operations on the first AI model and the second AI model according to the first information.

12. The method according to any one of claims 1 - 10, characterized in that, The using the second AI model to transmit second data to the second device includes: When the performance of the first AI model is less than or equal to a first threshold, the first device uses the second AI model to transmit the second data.

13. A communication method, characterized in that, applied to a second device communicating with a first device, the method comprises: The second device receives first data transmitted by the first device using a first artificial intelligence (AI) model, and receives second data transmitted by the first device using a second AI model; wherein, the first AI model is different from the second AI model, and the first data is the same as the second data; The second device sends first information to the first device, where the first information is related to a first performance metric and a second performance metric; wherein, the first performance metric is the performance metric of the first AI model when transmitting the first data; the second performance metric is the performance metric of the second AI model when transmitting the second data.

14. The method according to claim 13, characterized in that, The transmission resources of the first data and the transmission resources of the second data are frequency - division multiplexed transmission resources, time - division multiplexed transmission resources, or independent transmission resources; wherein, the independent transmission resources mean that the transmission resources of the first data and the transmission resources of the second data are different in both the time domain and the frequency domain.

15. The method according to claim 14, characterized in that, The transmission resources of the first data and the second data are indicated by first control information.

16. The method according to any one of claims 13 - 15, characterized in that, the method further comprises: The second device sends a first HARQ or / and a second HARQ, where the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data.

17. The method according to claim 16, characterized in that, The first HARQ and the second HARQ are used to determine a third HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data.

18. The method according to any one of claims 13 - 15, characterized in that, the method further comprises: The third Hybrid Automatic Repeat reQuest (HARQ) determined by the second device according to the first HARQ and the second HARQ, where the third HARQ is the HARQ corresponding to the first data and the second data, the first HARQ is the HARQ corresponding to the first data, and the second HARQ is the HARQ corresponding to the second data; The second device transmits the third HARQ.

19. The method according to any one of claims 16 - 18, wherein, The first performance metric is determined by the first HARQ.

20. The method according to any one of claims 16 - 19, wherein, The second performance metric is determined by the second HARQ.

21. The method according to any one of claims 13 - 20, wherein, The first performance metric is obtained within a first time period, the second performance metric is obtained within a second time period, and the first time period and the second time period are either fully overlapping or partially overlapping.

22. The method according to any one of claims 13 - 21, wherein, The first information includes at least one of the following: The first performance metric and the second performance metric; A first indication, where the first indication is used to indicate that the performance of the first AI model is better than the performance of the second AI model, or the first indication is used to indicate that the performance of the second AI model is better than the performance of the first AI model; The difference between the performance of the first AI model and the performance of the second AI model; or, A model identifier, where the model identifier is the identifier of the AI model with the best performance among the first AI model and the second AI model.

23. A communication device, wherein, comprises: A transceiver module and a processing module, The transceiver module is used to perform the sending step or the receiving step in the method according to any one of claims 1 - 22 above; The processing module is used to perform the steps in the method according to any one of claims 1 - 22 above other than the sending step and the receiving step.

24. A communication device, wherein, comprises at least one processor coupled to a memory; The memory is used to store programs or instructions; The at least one processor is used to execute the programs or instructions so that the device implements the method according to any one of claims 1 to 22.

25. A computer program product comprising program instructions, wherein, When the program instructions run on a computer, the computer is caused to execute the method according to any one of claims 1 to 22.

26. A computer-readable storage medium, wherein, Program instructions are stored in the computer-readable storage medium, and when the program instructions run, the method according to any one of claims 1 to 22 is caused to be executed.

27. A communication system, wherein, comprises: A first device and a second device, the first device is used to execute the method according to any one of claims 1 - 12 above, and the second device is used to execute the method according to any one of claims 13 - 22 above.

Images