Communication method and communication apparatus

Abstract

The present application provides a communication method and a communication apparatus. The method comprises: acquiring first information, the first information being used for indicating a feature extraction granularity N of a first device, the feature extraction granularity N being used for indicating a count of first resources involved when the first device performs correlation feature extraction, N being an integer greater than or equal to 1, and the correlation feature being used for processing a signal to be sent or a received signal. The present application ensures that the feature extraction granularity of an AI receiver is consistent with the feature extraction granularity of an AI transmitter, thereby improving the performance of a first device.

Description

Communication methods and communication devices

[0001] This application claims priority to Chinese Patent Application No. 202411999117.5, filed with the China National Intellectual Property Administration on December 31, 2024, entitled "Communication Method and Communication Device", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of wireless communication, and more specifically, to a communication method and a communication device. Background Technology

[0003] In wireless communication networks, the services supported are becoming increasingly diverse, leading to a wide range of demands. To meet these challenges, artificial intelligence (AI) technology is being introduced into wireless communication networks. For example, AI models are deployed in transmitters or receivers to utilize AI technology for signal generation or processing, thereby improving the performance of the transmitters or receivers. One of the key aspects of AI technology is feature extraction from the input, extracting features relevant to the output to obtain the output. Currently, how to improve receiver performance while incorporating AI technology into transmitters and receivers is a pressing issue that needs to be addressed. Summary of the Invention

[0004] This application provides a communication method and a communication device, which are intended to improve the performance of AI receivers.

[0005] Firstly, a communication method is provided. This method can be applied to a first device side; that is, the method can be executed by the first device side or by components of the first device side (such as a chip, chip system, circuit, or communication module). This application does not limit this. The following description mainly uses the first device side as an example.

[0006] It should be noted that in this application, both the network-side device and the terminal-side device can be used as the first device. In one possible implementation, the first device can be equipped with an AI transmitter or an AI receiver.

[0007] The method may include: acquiring first information, the first information being used to indicate the feature extraction granularity N of the first device, the feature extraction granularity N being used to indicate the number of first resources involved when the first device performs correlation feature extraction, N≥1 and is an integer; the correlation features being used to process the signal to be transmitted or the signal already received.

[0008] Based on the above scheme, the first device can determine the number of first resources involved in relevance feature extraction by acquiring the first information, that is, it can determine the feature extraction granularity N. This ensures that when an AI receiver is deployed on the first device side, the feature extraction granularity of the AI ​​receiver is consistent with that of the AI ​​transmitter, thereby improving the performance of the first device. Alternatively, when an AI transmitter is deployed on the first device side, it can process the signal to be transmitted based on the first information. This helps to maintain consistency with the feature extraction granularity of the AI ​​receiver; for example, if the AI ​​receiver also processes the received signal based on the first information, it will improve the performance of the AI ​​receiver.

[0009] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: processing the signal to be transmitted or the signal already received based on the first information.

[0010] In conjunction with the first aspect, in some implementations of the first aspect, processing the signal to be transmitted includes: processing the signal to be transmitted using one or more neural network models for an artificial intelligence (AI) transmitter; processing the received signal includes: processing the received signal using one or more neural network models for an AI receiver.

[0011] Based on the above scheme, the AI ​​receiver and AI transmitter can process the signal using the same feature extraction granularity N, thereby improving the performance of the AI ​​receiver and AI transmitter.

[0012] In conjunction with the first aspect, in some implementations of the first aspect, the one or more neural network models for the AI ​​transmitter are used for at least one of the following: AI modulation, AI precoding, AI piloting, and generating AI waveforms; and the one or more neural network models for the AI ​​receiver are used for at least one of the following: AI channel estimation, AI equalization, AI detection, and AI demodulation.

[0013] In conjunction with the first aspect, in some implementations of the first aspect, when the first device is a network-side device, obtaining the first information includes: receiving the first information from a terminal-side device; or when the first device is a terminal-side device, obtaining the first information includes: receiving the first information from a network-side device.

[0014] In conjunction with the first aspect, in some implementations of the first aspect, the feature extraction granularity N is any one of the following: one or more transmit resource groups (TRGs), one or more precoding resource block groups (PRGs), one or more resource block groups (RBGs), or one or more subbands.

[0015] In conjunction with the first aspect, in some implementations of the first aspect, the first information is further used to indicate a first granularity, which is the granularity of the resources used by the signal to be transmitted or the received signal; or, the method further includes: obtaining information on the first granularity, which is the granularity of the resources used by the signal to be transmitted or the received signal; the first information is used to indicate that the feature extraction granularity N is the same as the first granularity.

[0016] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: sending or receiving second information, the second information configuring the resources used by the signal to be sent or the signal already received, the configuration of the resources adopting the first granularity.

[0017] Optionally, this first-level information is either predefined by the protocol or configured by the network device.

[0018] In conjunction with the first aspect, in some implementations of the first aspect, the signal to be transmitted or the signal already received is carried on a first resource set, the first resource set including one or more subsets having the feature extraction granularity N, each of the one or more subsets having the feature extraction granularity N including N of the first resources.

[0019] In conjunction with the first aspect, in some implementations of the first aspect, the signal to be transmitted or the signal already received is carried on a first resource set, the first resource set including one or more subsets, each of the one or more subsets including one or more of the first resources, the number of the first resources included in each subset being determined by the feature extraction granularity N, and at least two of the subsets having different numbers of the first resources.

[0020] In conjunction with the first aspect, in some implementations of the first aspect, the starting position of the first resource set is the starting position of the first resource.

[0021] In conjunction with the first aspect, in some implementations of the first aspect, each subset includes a plurality of the first resources that are consecutive in one or more of the time domain, frequency domain, or spatial domain.

[0022] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: sending the first information.

[0023] In conjunction with the first aspect, in some implementations of the first aspect, obtaining the first information includes: receiving the first information.

[0024] In a second aspect, a communication apparatus is provided for performing the methods described in the first aspect and any possible implementation thereof. Specifically, the apparatus may include units and / or modules for performing the methods described in the first aspect and any possible implementation thereof, such as processing units and / or communication units.

[0025] Thirdly, a communication device is provided, the device comprising: at least one processor configured to cause the device to perform the methods described in the first aspect and any possible implementation thereof.

[0026] Optionally, the at least one processor is configured to execute computer programs or instructions to perform the methods described in the first aspect and any of its possible implementations.

[0027] Optionally, the device further includes a memory for storing the computer program or instructions.

[0028] Optionally, the at least one processor is coupled to a memory for storing the computer program or instructions. The memory may be located externally to the device.

[0029] Optionally, the device also includes a communication interface through which the processor reads instructions from memory. This can be understood as the communication interface being coupled to the processor and used to input computer programs or instructions to the processor, or to output information from the processor.

[0030] Unless otherwise specified, or if the transmission and acquisition / reception operations involved do not contradict their actual function or internal logic in the relevant description, they can be understood as output, input, or other operations, or as transmission and reception operations performed by radio frequency circuits and antennas. This application does not limit them in this regard.

[0031] Fourthly, a computer-readable storage medium is provided that stores a computer program (e.g., program code) or instructions that, when executed on a communication device, cause the communication device to perform the methods described in the first aspect and any possible implementation thereof.

[0032] Fifthly, a computer program product containing instructions is provided, which, when run on a computer, causes the computer to perform the methods described in the first aspect and any possible implementation thereof.

[0033] A sixth aspect provides a communication system including a first device, wherein the first device is configured to perform the method provided by any implementation of the first aspect. Attached Figure Description

[0034] Figure 1 is a schematic diagram of a communication system applicable to an embodiment of this application.

[0035] Figure 2 is a schematic diagram of a possible application framework in a communication system.

[0036] Figure 3 is a schematic diagram of a possible application framework in a communication system.

[0037] Figure 4 is a schematic diagram of a neuron.

[0038] Figure 5 is a schematic diagram of an AI transmitter.

[0039] Figure 6 is a schematic diagram of an AI waveform.

[0040] Figure 7 is a schematic diagram of an AI receiver.

[0041] Figure 8 is a schematic block diagram of a communication method proposed in this application.

[0042] Figure 9 is a schematic diagram of an AI transmitter deployed on the first device side according to an embodiment of this application.

[0043] Figure 10 is a schematic flowchart of another communication method proposed in this application.

[0044] Figure 11 is a schematic block diagram of a communication device provided in an embodiment of this application.

[0045] Figure 12 is a schematic block diagram of a communication device provided in an embodiment of this application.

[0046] Figure 13 is a schematic block diagram of an AI processor provided in an embodiment of this application.

[0047] Figure 14 is a schematic block diagram of a chip system provided in an embodiment of this application. Detailed Implementation

[0048] The technical solutions in this application will now be described with reference to the accompanying drawings.

[0049] Before introducing the scheme of this application, the following points should be noted.

[0050] (1) In this application, the expression " / " is used to indicate that the objects before and after are in an "or" relationship; for example, A / B can mean: A or B. The expression "and / or" is used to indicate that the objects before and after are in a relationship of either "and" or "or"; for example, A and / or B can mean the following: A exists alone, B exists alone, A and B exist simultaneously, where A and B can be single or multiple. "At least one of the following" or similar expressions are used to indicate any combination of the listed items; for example, at least one of A, B and / or C can mean the following: A exists alone, B exists alone, C exists alone, A and B exist simultaneously, B and C exist simultaneously, A and C exist simultaneously, A, B and C exist simultaneously, where A, B, and C can be single or multiple.

[0051] (2) In this application, "send" and "receive" indicate the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which may include direct transmission via the air interface or indirect transmission via the air interface by other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which may include direct reception from YY via the air interface or indirect reception from YY via the air interface by other units or modules. "Send" can also be understood as the "output" of the chip interface, and "receive" can also be understood as the "input" of the chip interface. In other words, sending and receiving can occur between devices, such as between network devices and terminal devices, or within a device, such as between components, modules, chips, software modules, or hardware modules within the device via a bus, wiring, or interface.

[0052] (3) In the various embodiments of this application, unless otherwise specified or logically conflicting, the terms and / or descriptions of different embodiments are consistent and can be referenced by each other. The technical features of different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0053] (4) In this application, "first," "second," and "#1," "#2," and "#A" are merely for descriptive convenience and are used to distinguish objects, and are not intended to limit the scope of the embodiments of this application. They are not used to describe the order or sequence of features. It should be understood that such described objects can be interchanged where appropriate so as to describe solutions other than those in the embodiments of this application.

[0054] (5) In this application, the words “exemplary,” “for example,” etc., are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as an “example” in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word “example” is intended to present the concept in a concrete manner. In the embodiments of this application, “of,” “corresponding, relevant,” and “corresponding” may sometimes be used interchangeably, and it should be noted that their intended meanings are consistent unless their distinction is emphasized.

[0055] (6) In this application, "instruction" can include direct instruction, indirect instruction, explicit instruction, implicit instruction, etc. When describing an instruction information as indicating A, it can be understood as the instruction information carrying A, carrying the identifier of A, carrying B which is associated with A, carrying the identifier of B which is associated with A, etc. In other words, if the receiving side of an instruction information can determine A based on the instruction information, it can be described as the instruction information indicating A, and the specific method of determination is not limited. When it is understood that the instruction information carries A, "instruction" can be replaced with "includes". In this case, expressions such as "send / receive instruction information, the instruction information indicates A" can be replaced with "send / receive A".

[0056] In this application, the information indicated by the instruction information is called the information to be instructed. In specific implementations, there are many ways to indicate the information to be instructed, such as, but not limited to, directly indicating the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly indicate the information to be instructed by indicating other information, where there is a relationship between the other information and the information to be instructed. It can also indicate only a part of the information to be instructed, while the other parts are known or pre-agreed upon. For example, the instruction of specific information can be achieved by using a pre-agreed (e.g., protocol-defined) arrangement of various pieces of information, thereby reducing instruction overhead to some extent. Furthermore, the information to be instructed can be sent as a whole or divided into multiple sub-information pieces, and the sending period and / or timing of these sub-information pieces can be the same or different.

[0057] The following describes the communication system to which this application applies.

[0058] The technical solutions provided in this application can be applied to various communication systems, such as 5th generation (5G) or new radio (NR) systems, long term evolution (LTE) systems, LTE frequency division duplex (FDD) systems, and LTE time division duplex (TDD) systems. The technical solutions provided in this application can also be applied to future communication networks. Furthermore, the technical solutions provided in this application can be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine-type communication (MTC), and Internet of Things (IoT) communication systems. The technical solutions provided in this application can also be applied to non-terrestrial network (NTN) systems such as inter-satellite communication and satellite communication.

[0059] As an example, a satellite communication system includes a satellite base station and terminal equipment. The satellite base station provides communication services to the terminal equipment. Satellite base stations can also communicate with each other. A satellite can act as a base station or as a terminal device. Here, "satellite" can refer to drones, hot air balloons, low-Earth orbit satellites, medium-Earth orbit satellites, high-Earth orbit satellites, etc. "Satellite" can also refer to non-terrestrial base stations or non-terrestrial equipment.

[0060] As an example, V2X communication can include: vehicle-to-vehicle (V2V) communication, vehicle-to-infrastructure (V2I) communication, vehicle-to-pedestrian (V2P) communication, and vehicle-to-network (V2N) communication.

[0061] In a communication system, a device can send signals to or receive signals from another device. These signals can include information, signaling, or data. The term "device" can also be replaced by an entity, network entity, communication device, communication module, node, communication node, etc. This application describes the system using a device as an example. For instance, a communication system can include at least one terminal device and at least one network device. The network device can send downlink signals to the terminal device, and / or the terminal device can send uplink signals to the network device. It is understood that in this application embodiment, the terminal device can be a first device, in which case the network device can be a second device; or the network device can be a first device, in which case the terminal device can be a second device.

[0062] The methods provided in this application embodiment can be executed by a terminal-side device or a network-side device. The terminal-side device can refer to the terminal device itself, a component within the terminal device (e.g., a processor, chip, or chip system), an AI entity serving the terminal device, such as a server (e.g., an over-the-top (OTT) server or a cloud server), or a logic module or software capable of implementing all or part of the terminal device's functions. The network-side device can refer to the network device itself, a component within the network device (e.g., a processor, chip, or chip system), an AI entity serving the network device, such as a RAN intelligent controller (RIC), operation administration and maintenance (OAM), or a server (e.g., an OTT server or a cloud server), or a logic module or software capable of implementing all or part of the network device's functions. Communication between the terminal-side device and the network-side device can be achieved through a communication link between the terminal device and the network device, a communication link between servers, forwarding through other communication devices outside the server, or a wired link. The following description uses terminal devices or network devices as examples. It is understood that a terminal device can be replaced by a terminal-side device, such as one or more of the aforementioned terminal-side devices, and a network device can also be replaced by a network-side device, such as one or more of the aforementioned network-side devices.

[0063] In wireless communication networks, such as mobile communication networks, the services supported by the networks are becoming increasingly diverse, thus requiring increasingly diverse demands. For example, networks need to support ultra-high speeds, ultra-low latency, and / or massive connectivity. This characteristic makes network planning, network configuration, and / or resource scheduling increasingly complex. Furthermore, as network functions become more powerful, such as supporting higher spectrum, higher-order MIMO technology, beamforming, and / or beam management, network energy efficiency has become a hot research topic. These new demands, new scenarios, and new characteristics bring unprecedented challenges to network planning, operation, and efficient operation. To meet these challenges, artificial intelligence (AI) technology can be introduced into wireless communication networks to achieve network intelligence. To support AI technology in wireless networks, AI nodes may also be introduced.

[0064] Figure 1 is a schematic diagram of a communication system applicable to the communication method of this application embodiment. As shown in Figure 1, the communication system 100 may include at least one network device, such as network device 110 shown in Figure 1. The communication system 100 may also include at least one terminal device, such as terminal device 120 and terminal device 130 shown in Figure 1. Network device 110 and terminal devices (such as terminal devices 120 and 130) can communicate via a wireless link. The communication devices in this communication system, for example, network device 110 and terminal device 120, can communicate via multi-antenna technology.

[0065] In some embodiments, the communication system 100 further includes an AI network element 140. The AI ​​network element 140 is used to perform AI-related operations, such as building training datasets or training AI models.

[0066] In one possible implementation, network device 110 can send data related to the training of the AI ​​model to AI network element 140, which then constructs a training dataset and trains the AI ​​model. For example, the data related to the training of the AI ​​model may include data reported by the terminal device. AI network element 140 can send the results of operations related to the AI ​​model to network device 110, which then forwards them to the terminal device. For example, the results of operations related to the AI ​​model may include at least one of the following: a trained AI model, model evaluation results, or test results. Exemplarily, a portion of the trained AI model may be deployed on network device 110, and another portion on the terminal device. Alternatively, the trained AI model may be deployed on network device 110. Or, the trained AI model may be deployed on the terminal device.

[0067] It should be understood that Figure 1 is only used as an example of the AI ​​network element 140 being directly connected to the network device 110. In other scenarios, the AI ​​network element 140 can also be connected to a terminal device. Alternatively, the AI ​​network element 140 can be connected to both the network device 110 and a terminal device simultaneously. Alternatively, the AI ​​network element 140 can also be connected to the network device 110 through a third-party network element. This application embodiment does not limit the connection relationship between the AI ​​network element and other network elements.

[0068] AI element 140 can also be set as a module in network devices and / or terminal devices, for example, in network device 110 or terminal device shown in Figure 1.

[0069] It should be noted that Figure 1 is a simplified schematic diagram for ease of understanding. For example, the communication system may also include other devices, such as wireless relay devices and / or wireless backhaul devices, which are not shown in Figure 1. In practical applications, the communication system may include multiple network devices or multiple terminal devices. This application embodiment does not limit the number of network devices and terminal devices included in the communication system.

[0070] In the embodiments of this application, the terminal device may also be referred to as user equipment (UE), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent, or user apparatus.

[0071] Terminal devices can be devices that provide voice / data, such as handheld devices with wireless connectivity, in-vehicle devices, etc. Currently, examples of terminals include: mobile phones, tablets, laptops, PDAs, mobile internet devices (MIDs), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self-driving vehicles, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, computing devices or other processing devices connected to a wireless modem, wearable devices, terminal devices in 5G networks, or future public land mobile communication networks. Terminal devices in a network (PLMN), etc., are not limited to this in the embodiments of this application.

[0072] By way of example and not limitation, in this embodiment, the terminal device can also be a wearable device. Wearable devices, also known as wearable smart devices, are a general term for devices that utilize wearable technology to intelligently design and develop everyday wearables, such as glasses, gloves, watches, clothing, and shoes. Wearable devices are portable devices that are worn directly on the body or integrated into the user's clothing or accessories. Wearable devices are not merely hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are feature-rich, large in size, and can achieve complete or partial functions without relying on a smartphone, such as smartwatches or smart glasses, as well as those that focus on a specific type of application function and require the use of other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0073] In this embodiment, the device for implementing the functions of the terminal device can be the terminal device itself, or it can be any device capable of supporting the terminal device in implementing those functions, such as a chip system. This device can be installed in or used in conjunction with the terminal device. In this embodiment, the chip system can be composed of chips or may include chips and other discrete components. This embodiment only uses the terminal device as an example to illustrate the device for implementing the functions of the terminal device, and does not constitute a limitation on the solution of this embodiment.

[0074] The network device in this application embodiment can be a device for communicating with a terminal device. This network device can include an access network device (i.e., an access network node) or a radio access network device, such as a base station. In this application embodiment, the radio access network device can refer to a radio access network (RAN) node (or device) that connects the terminal device to the wireless network. A base station can broadly encompass, or be replaced by, various names including: NodeB, evolved NodeB (eNB), next-generation NodeB (gNB), relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), master station, auxiliary station, motor slide retainer (MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), radio unit (RU), positioning node, etc. A base station can be a macro base station, micro base station, relay node, donor node, or similar entities, or combinations thereof. A base station can also refer to a communication module, modem, or chip installed within the aforementioned equipment or apparatus. A base station can also be a mobile switching center, equipment performing base station functions in D2D, V2X, and M2M communications, network-side equipment in future communication networks, or equipment performing base station functions in future communication networks. A base station can support networks using the same or different access technologies. Optionally, a RAN node can also be a server, wearable device, vehicle, or in-vehicle equipment. For example, the access network equipment in vehicle-to-everything (V2X) technology can be a roadside unit (RSU). The embodiments of this application do not limit the specific technologies or equipment forms used in the network equipment.

[0075] Base stations can be fixed or mobile. For example, a helicopter or drone can be configured to act as a mobile base station, and one or more cells can move depending on the location of the mobile base station. In other examples, a helicopter or drone can be configured as a device to communicate with another base station.

[0076] In some deployments, the network devices mentioned in the embodiments of this application may be devices including CU, DU, or CU and DU, or devices with control plane CU nodes (central unit-control plane (CU-CP)) and user plane CU nodes (central unit-user plane (CU-UP)) and DU nodes. For example, the network devices may include gNB-CU-CP, gNB-CU-UP, and gNB-DU.

[0077] In some deployments, multiple RAN nodes collaborate to assist terminals in achieving wireless access, with different RAN nodes each implementing some of the base station's functions. For example, RAN nodes can be CUs, DUs, CU-CPs, CU-UPs, or RUs. CUs and DUs can be configured separately or included in the same network element, such as a BBU. RUs can be included in radio frequency equipment or radio frequency units, such as RRUs, AAUs, or RRHs.

[0078] RAN nodes can support one or more types of fronthaul interfaces, with different fronthaul interfaces corresponding to DUs and RUs with different functions. If the fronthaul interface between the DU and RU is a common public radio interface (CPRI), the DU is configured to implement one or more baseband functions, and the RU is configured to implement one or more radio frequency functions. If the fronthaul interface between the DU and RU is another type of interface, relative to CPRI, some downlink and / or uplink baseband functions, such as, for downlink, precoding, digital beamforming (BF), or one or more of inverse fast Fourier transform (IFFT) / cyclic prefix addition (CP), are moved from the DU to the RU; and for uplink, one or more of digital beamforming (BF), or fast Fourier transform (FFT) / cyclic prefix removal (CP), are moved from the DU to the RU. In one possible implementation, the interface can be an enhanced common public radio interface (eCPRI). Under the eCPRI architecture, the segmentation between DU and RU differs, corresponding to different categories (Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, F.

[0079] Taking eCPRI Cat A as an example, for downlink transmission, the DU is configured to implement one or more functions before and after layer mapping (i.e., coding, rate matching, scrambling, modulation, and layer mapping), while other functions after layer mapping (e.g., resource element (RE) mapping, digital beamforming (BF), or one or more functions of inverse fast Fourier transform (IFFT) / adding cyclic prefix (CP)) are moved to the RU. For uplink transmission, the DU is configured to implement one or more functions before and after demapping (i.e., decoding, rate matching de-matching, descrambling, demodulation, inverse discrete Fourier transform (IDFT), channel equalization, and demapping), while other functions after demapping (e.g., digital BF or one or more functions of fast Fourier transform (FFT) / removing CP) are moved to the RU. It is understandable that the functional descriptions of the DU and RU corresponding to various types of eCPRI can be found in the eCPRI protocol, and will not be elaborated here.

[0080] In one possible design, the processing unit in the BBU used to implement baseband functions is called the baseband high (BBH) unit, and the processing unit in the RRU / AAU / RRH used to implement baseband functions is called the baseband low (BBL) unit.

[0081] In different systems, CU (or CU-CP and CU-UP), DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an open radio access network (ORAN / O-RAN) system, CU can also be called O-CU (open CU), DU can also be called O-DU, CU-CP can also be called O-CU-CP, CU-UP can also be called O-CU-UP, and RU can also be called O-RU. Any of the units among CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented through software modules, hardware modules, or a combination of software modules and hardware modules.

[0082] In this embodiment, the apparatus for implementing the functions of a network device can be a network device itself, or an apparatus capable of supporting the network device in implementing those functions, such as a chip system, hardware circuit, software module, or a hardware circuit plus a software module. This apparatus can be installed in or used in conjunction with the network device. In this embodiment, the example of a network device is used only to illustrate the apparatus for implementing the functions of the network device, and does not constitute a limitation on the solutions described in this embodiment.

[0083] Network devices and / or terminal devices can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; they can also be deployed on water; and they can also be deployed in the air on airplanes, balloons, and satellites. This application does not limit the scenario in which the network devices and terminal devices are located. Furthermore, terminal devices and network devices can be hardware devices, or software functions running on dedicated hardware or general-purpose hardware, such as virtualization functions instantiated on a platform (e.g., a cloud platform), or entities that include dedicated or general-purpose hardware devices and software functions. This application does not limit the specific form of the terminal devices and network devices.

[0084] In wireless communication networks, such as mobile communication networks, the services supported by the networks are becoming increasingly diverse, leading to increasingly diverse requirements. For example, networks need to support ultra-high speeds, ultra-low latency, and / or massive connectivity. This characteristic makes network planning, network configuration, and / or resource scheduling increasingly complex. Furthermore, as network functions become more powerful, such as supporting higher spectrum levels, supporting higher-order multiple-input multiple-output (MIMO) technologies, supporting beamforming, and / or supporting beam management, network energy efficiency has become a hot research topic. These new requirements, new scenarios, and new characteristics bring unprecedented challenges to network planning, operation, and efficient operation. To meet these challenges, artificial intelligence technology can be introduced into wireless communication networks to achieve network intelligence.

[0085] To support AI technology in wireless networks, AI nodes may also be introduced into the network.

[0086] Optionally, the AI ​​node can be deployed in one or more of the following locations within the communication system: access network equipment, terminal equipment, or core network equipment, etc. Alternatively, the AI ​​node can be deployed independently, for example, in a location other than any of the aforementioned devices, such as in the host or cloud server of an over-the-top (OTT) system. The AI ​​node can communicate with other devices in the communication system, which can be one or more of the following: network equipment, terminal equipment, or core network elements, etc.

[0087] It is understood that this application does not limit the number of AI nodes. For example, when there are multiple AI nodes, these nodes can be divided based on function, such as different AI nodes being responsible for different functions.

[0088] It can also be understood that AI nodes can be AI network elements or AI modules. AI nodes can be independent devices, or they can be integrated into the same device to implement different functions. Alternatively, they can be network elements in hardware devices, software functions running on dedicated hardware, or virtualization functions instantiated on a platform (e.g., a cloud platform). This application does not limit the specific form of the AI ​​nodes described above.

[0089] Figure 2 illustrates a possible application framework in a communication system. As shown in Figure 2, network elements in the communication system are connected via interfaces (e.g., NG, Xn) or air interfaces. These network element nodes, such as core network equipment, access network nodes (RAN nodes), terminal equipment, or one or more devices in operation administration and maintenance (OAM), are equipped with one or more AI modules (only one is shown in Figure 2 for clarity). An access network node can be a single RAN node or can include multiple RAN nodes, for example, including CU and DU. CU and / or DU can also be equipped with one or more AI modules. Optionally, a CU can be further divided into CU-CP and CU-UP. One or more AI models are configured in CU-CP and / or CU-UP. Exemplarily, CU and DU are connected via an F1 interface. CU and CU are connected via an Xn interface.

[0090] The AI ​​module is used to implement corresponding AI functions. Therefore, the AI ​​model can also be called an AI function or an AI function entity. AI modules deployed in different network elements can be the same or different. The AI ​​module model, configured with different parameters, can enable the AI ​​module to achieve different functions. The AI ​​module model can be configured based on one or more of the following parameters: structural parameters (e.g., at least one of the following: number of neural network layers, neural network width, inter-layer connections, neuron weights, neuron activation function, or bias in the activation function), input parameters (e.g., at least one of the following: type of input parameter, input dimension, number of input ports), or output parameters (e.g., at least one of the following: type of output parameter, output dimension, number of output ports). The bias in the activation function can also be called the bias of the neural network. The input dimension can refer to the size of an input data point; for example, when the input data is a sequence, the input dimension corresponding to that sequence can indicate the length of the sequence. The number of input ports can refer to the quantity of input data. Similarly, the output dimension can refer to the size of an output data point; for example, when the output data is a sequence, the output dimension corresponding to that sequence can indicate the length of the sequence. The number of output ports can refer to the quantity of output data.

[0091] An AI module can have one or more models. A model can infer an output, which includes one or more parameters. The learning, training, or inference processes of different models can be deployed on different nodes or devices, or they can be deployed on the same node or device.

[0092] The network device can be a network device equipped with one or more AI modules. The network device can be one or more devices in the core network, access network node (RAN node), or OAM as shown in Figure 2. For example, the AI ​​module can be the RIC shown in Figure 3, such as a near real-time RIC or a non-real-time RIC. For example, the near real-time RIC is set in the RAN node (e.g., in CU, DU), while the non-real-time RIC is set in the OAM, cloud server, core network device, or other network device. The RIC can obtain data (e.g., a subset of data) from multiple terminal devices from the RAN node (e.g., CU, CU-CP, CU-UP, DU, and / or RU), reassemble it into a training dataset #2, and train based on the training dataset #2. Exemplarily, the near real-time RIC and the non-real-time RIC can also be set up separately as a network element; the network device can be a near real-time RIC or a non-real-time RIC.

[0093] Figure 3 illustrates a possible application framework in a communication system. As shown in Figure 3, the communication system includes a RAN intelligent controller (RIC). For example, the RIC can be the AI ​​module shown in Figure 2, used to implement AI-related functions. RICs include near-real-time RICs (near-RT RICs) and non-real-time RICs (non-RT RICs). Non-real-time RICs primarily process non-real-time information, such as data that is not sensitive to latency, with latency in the order of seconds. Real-time RICs primarily process near-real-time information, such as data that is relatively sensitive to latency, with latency in the order of tens of milliseconds.

[0094] Near real-time RICs are used for model training and inference. For example, they are used to train AI models and then use those models for inference. Near real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, and / or RUs) and / or terminals. This information can be used as training data or input data for inference. Optionally, near real-time RICs can deliver inference results to RAN nodes and / or terminals. Optionally, inference results can be exchanged between CUs and DUs, and / or between DUs and RUs. For example, a near real-time RIC delivers an inference result to a DU, which then sends it to an RU.

[0095] Non-real-time RICs are also used for model training and inference. For example, they can be used to train AI models and then use those models for inference. Non-real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, and / or RUs) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to RAN nodes and / or terminals. Optionally, inference results can be exchanged between CUs and DUs, and / or between DUs and RUs; for example, a non-real-time RIC delivers inference results to a DU, which then forwards them to an RU.

[0096] Near real-time RICs and non-real-time RICs can also be configured as separate network elements. Optionally, near real-time RICs and non-real-time RICs can also be part of other devices. For example, near real-time RICs can be set in RAN nodes (e.g., CU, DU), while non-real-time RICs can be set in OAM, cloud servers, core network devices, or other network devices.

[0097] To facilitate understanding of the solutions in the embodiments of this application, the terms that may be involved in the embodiments of this application are explained below.

[0098] 1. AI Model

[0099] An AI model is an algorithm or computer program that enables AI functionality. It represents the mapping relationship between the model's input and output. An AI model can be understood as a function model that maps an input of a certain dimension to an output of a certain dimension; its parameters are obtained through machine learning training. For example, f(x) = ax 2 +b is a quadratic function model, which can be viewed as an AI model. a and b correspond to the parameters of this AI model, and can be obtained through machine learning training. An AI model can also be called a model, an AI function, or a feature. One AI function can correspond to one or more AI models.

[0100] AI models can be neural networks, linear regression models, decision tree models, support vector machines (SVM), Bayesian networks, Q-learning models, or other machine learning (ML) models.

[0101] 2. Neural Network (NN)

[0102] Neural networks are a specific implementation of AI or machine learning. According to the general approximation theorem, neural networks can theoretically approximate any continuous function, thus enabling them to learn arbitrary mappings.

[0103] Neural networks can be composed of neural units, which can refer to units represented by x. s A neural network is a computational unit that takes an intercept of 1 as input. It is formed by connecting many of these individual neural units together; that is, the output of one neural unit can be the input of another. The input of each neural unit can be connected to the local receptive field of the previous layer to extract features from that local receptive field, which can be a region composed of several neural units. A neural unit is also called a neuron. Each neuron performs a weighted summation operation on its own input values ​​and generates an output through a nonlinear function. The weights of the weighted summation operation of neurons in a neural network and the nonlinear function are called the parameters of the neural network. The parameters of all neurons in a neural network constitute the parameters of that neural network.

[0104] Figure 4 is a schematic diagram of a neuron. As shown in Figure 4, assume the neuron's input is x = [x0, x1, ..., x...]. n ], where the weights corresponding to each input element in x are w = [w0, w1, ..., w n The bias of the weighted summation is b. The nonlinear function is... The output is y. For example, if the nonlinear function is y = f(z) = max(0, z), then the output of this neuron is... For example, if the nonlinear function of a neuron is y = f(z) = z, then the output of that neuron is... b can take various possible values, such as decimals, integers (0, positive integers, or negative integers), or complex numbers. The nonlinear functions of different neurons in a neural network can be the same or different.

[0105] Neural networks typically consist of multiple layers, each containing one or more neurons. Increasing the depth and / or width of a neural network can enhance its expressive power, providing more robust information extraction and abstract modeling capabilities for complex systems. The depth of a neural network can refer to the number of layers it comprises, while the number of neurons in each layer can be called the width of that layer. In one implementation, the neural network includes an input layer and an output layer. The input layer processes the received input through neurons and then passes the result to the output layer, which obtains the output of the neural network. In another implementation, the neural network includes an input layer, hidden layers, and an output layer. The input layer processes the received input through neurons and then passes the result to the hidden layer, which in turn passes the result to the output layer or an adjacent hidden layer. Finally, the output layer obtains the output of the neural network. A neural network can include one or more sequentially connected hidden layers, without limitation. During the training process of the neural network, a loss function can be defined. The loss function describes the difference between the output value of the neural network and the ideal target value; this application does not limit the specific form of the loss function. The training process of a neural network involves adjusting the network parameters, such as the number of layers, width, weights of neurons, and / or parameters in the activation functions of neurons, to make the loss function value less than a threshold value or meet the target requirements.

[0106] Taking neural networks as an example, the AI ​​model involved in this application can be a self-attention network. The AI ​​model involved in this application can also be other neural networks, such as deep neural networks (DNNs). Depending on the network construction method, DNNs can include feedforward neural networks (FNNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs), etc. FNNs, CNNs, and RNNs are all constructed based on neurons.

[0107] The following section introduces self-attention networks.

[0108] Self-attention networks, also known as self-attention mechanisms, will be used as an example for ease of description. A self-attention mechanism is a neural network model based on attention mechanisms and is widely used in Transformer models. In a self-attention mechanism, each element is represented by a vector. To calculate the relationship between each element and other elements, three matrices are introduced: a query matrix, a key matrix, and a value matrix. These three matrices are typically denoted as W. Q W K W V W Q W K W V Features can be extracted from each element of the input sequence through linear transformation. First, the input sequence X is compared with matrix W. Q W K W V Multiplying them yields three new vectors: Q, K, and V. In other words, Q = XW. Q K = XW K V = XW V Furthermore, the features corresponding to the input sequence X are obtained according to the following formula:

[0109] Among them, K T This represents the transpose of vector K, d k This represents the number of columns in vector K.

[0110] 3. AI transmitter and AI receiver

[0111] In a communication system, the transmitter is responsible for processing the raw signals to be transmitted. These signals include voice, data, images, etc., and need to be converted into electrical signals suitable for transmission over a wireless channel. For example, the transmitter may perform at least one of the following operations on the raw signal: modulation, layer mapping, and antenna port mapping, to generate an electrical signal suitable for transmission over the wireless channel. The receiver is responsible for processing the received signals to obtain the raw baseband signals. These baseband signals can then be converted into a form suitable for user use (e.g., voice, images, etc.). For example, the receiver may perform at least one of the following operations on the received signal: channel estimation, equalization / multiple-input multiple-output (MIMO) detection, demodulation, etc., to output a bitstream.

[0112] Figure 5 is a schematic diagram of an AI transmitter. As shown in Figure 5, the AI ​​transmitter includes one or more AI models. There is no explicit mathematical expression between the input and output of these AI models; instead, the input is processed by the AI ​​model to produce the output. One of the core working principles of an AI transmitter is feature extraction, which involves extracting features related to the output from the input to obtain the output. For example, if the input is a bitstream, the output after passing through the AI ​​transmitter is a transmitted symbol. One method of feature extraction is to use the self-attention mechanism mentioned above. For instance, this self-attention mechanism can be included in the AI ​​module of the AI ​​transmitter. During the training process of the AI ​​transmitter, the model parameters involved in the self-attention mechanism can be obtained; that is, the self-attention mechanism and the functions implemented by the AI ​​transmitter are trained together. The functions implemented by the AI ​​transmitter can include one or more of the following: AI modulation, AI precoding, AI pilot or AI reference signal, and AI waveform.

[0113] It should be noted that the AI ​​model in the AI ​​transmitter can be used for one or more of the following: AI modulation, AI precoding, AI pilot or AI reference signal, and AI waveform; or, the AI ​​transmitter is an AI model with other transmitter functions; or, the AI ​​transmitter is an AI model with multiple of the above functions. The AI ​​model in the AI ​​transmitter may also include one of DNN, CNN, FNN, and RNN. This application does not limit the AI ​​model in the AI ​​transmitter.

[0114] In one implementation, as shown in Figure 6, when the AI ​​transmitter includes waveform functionality, the output of the AI ​​transmitter can be a frequency domain transmission signal or a time domain transmission signal. When the output of the AI ​​transmitter is a frequency domain transmission signal, it can be transformed by IFFT to obtain a time domain transmission signal.

[0115] For example, when the AI ​​transmitter includes AI modulation functionality, the training data involved in training the AI ​​transmitter includes tensor B corresponding to the bits and tensor H corresponding to the channel information, as well as label X corresponding to each input. label Tag X label If the target of the transmitted signal output by the AI ​​transmitter is to be represented, then feature extraction is performed on the training data to obtain the features related to the training data. Then, the mean squared error (MSE) loss function is used for backpropagation to update the parameters of the AI ​​transmitter and train the AI ​​transmitter. Then, the extracted modulation-related information is input into the AI ​​transmitter to obtain the AI ​​modulation result corresponding to the input.

[0116] Figure 7 is a schematic diagram of an AI receiver. As shown in Figure 7, the AI ​​receiver includes one or more AI models. There is no explicit mathematical expression between the input and output of these AI models; instead, the input is processed by the AI ​​model to produce the output. One of the core working principles of an AI receiver is feature extraction, which involves extracting features related to the output from the input to obtain the output. For example, if the input is a received signal, the output after passing through the AI ​​receiver is a bitstream. One method of feature extraction is to use the self-attention mechanism mentioned above. For instance, this self-attention mechanism can be included in the AI ​​module of the AI ​​receiver. During the training process of the AI ​​receiver, the model parameters involved in the self-attention mechanism can be obtained; that is, the self-attention mechanism and the functions implemented by the AI ​​receiver are trained together. The functions implemented by the AI ​​receiver can include one or more of AI channel estimation, AI equalization, AI detection, and AI demodulation.

[0117] It should be noted that the AI ​​model of the AI ​​receiver can be one or more of AI channel estimation, AI equalization, AI detection, and AI demodulation. Alternatively, the AI ​​receiver can be an AI model with other transmitter functions; or, the AI ​​receiver can be an AI model with multiple of the above functions. The AI ​​model in the AI ​​receiver can also include any of DNN, CNN, FNN, and RNN. This application does not limit the AI ​​model in the AI ​​receiver.

[0118] For example, when an AI receiver has AI demodulation capabilities, it performs AI demodulation on the input. For instance, if the input is an AI-modulated signal output from an AI transmitter, the AI ​​receiver will demodulate the AI-modulated signal. In this case, the loss function of the AI ​​receiver is the MSE loss function. The loss function is calculated and backpropagated to update the parameters of the AI ​​receiver, thus training the AI ​​receiver. The extracted demodulation-related information is then input into the AI ​​receiver to obtain the AI ​​demodulation result corresponding to the input.

[0119] For example, the functions implemented by the AI ​​transmitter and the AI ​​receiver can be corresponding. For instance, when the AI ​​transmitter is used for AI modulation, the AI ​​receiver is used for AI demodulation. Or, for example, when the AI ​​transmitter is used for AI precoding, the AI ​​receiver is used for AI channel estimation. Both the AI ​​transmitter and the AI ​​receiver perform feature extraction on the input information to further obtain the output result.

[0120] In the current protocol, the base station informs the UE of the precoding granularity through the precoding resource block group (PRG) parameter. The UE assumes that the precoding is the same on all resources in the PRG. In this way, the UE can perform channel estimation with a granularity less than or equal to that of the PRG to avoid the problem of channel discontinuity caused by different precoding.

[0121] However, the PRG parameter indicates the precoding granularity, which may not be suitable for the feature extraction granularity between the AI ​​transmitter and the AI ​​receiver. Furthermore, a mismatch between the number of resources involved in processing each input and the number of resources involved in processing each input by the AI ​​receiver can degrade receiver performance. For example, if the feature extraction granularity of the AI ​​receiver is smaller than that of the AI ​​transmitter, the AI ​​receiver cannot utilize complete feature information. Conversely, if the feature extraction granularity of the AI ​​receiver is larger than that of the AI ​​transmitter, the AI ​​receiver may extract incorrect feature information, thus worsening its performance.

[0122] In view of this, this application provides a communication method and a communication device, wherein the first device obtains the feature extraction granularity to achieve the alignment of the feature extraction granularity, thereby ensuring the performance of the first device.

[0123] The methods provided by the embodiments of this application will be described in detail below with reference to the accompanying drawings. The embodiments provided by this application can be applied to the scenarios shown in the above figures and are not limited thereto. It should be understood that the embodiments of this application can be applied to scenarios involving communication between a sending end and a receiving end. Specifically, the technical solutions of this application are applicable to uplink transmission, downlink transmission, or sidelink transmission scenarios, etc.

[0124] Figure 8 is a schematic diagram of a communication method 800 provided in an embodiment of this application. For ease of description, a first device is used as an example for illustrative purposes. The first device can be replaced by components of the first device (e.g., a chip, chip system, circuit, or communication module). The method 800 shown in Figure 8 may include the following steps.

[0125] S810. Obtain first information, which is used to indicate the feature extraction granularity N of the first device. The feature extraction granularity N is used to indicate the number of first resources involved when the first device performs correlation feature extraction. N≥1 and is an integer. The correlation features are used to process the signal to be transmitted or the signal already received.

[0126] The first device can be equipped with either an AI transmitter or an AI receiver. For example, if the first device is a network-side device, it will have an AI transmitter deployed on it; or if the first device is a terminal-side device, it will have an AI transmitter deployed on it.

[0127] For example, the feature extraction granularity N can be understood as the number of first resources involved in the first device's relevance feature extraction, where the first resource can be a resource element (RE). For instance, when the first device deploys an AI transmitter, it can map the input information to the first device onto several REs. These REs then correspond to the input information of the first device. When the first device extracts relevance features from this input information, the number of REs involved in a single relevance feature extraction operation, such as the number of REs mapped to this input information, is the feature extraction granularity N of the first device.

[0128] It should be noted that the first device can extract relevant features from the input information in multiple steps. For example, if the input information is mapped to 48 REs, then the 48 REs can be divided into groups of 12. The first device can then extract relevant features three times, with each extraction involving 12 REs. Therefore, the feature extraction granularity N is 12.

[0129] For example, correlation features are used to process either the signal to be transmitted or the signal already received. This can be understood as extracting correlation features from the signal to be transmitted and obtaining the output signal of the first device based on these features. For instance, the signal to be transmitted is processed using one or more neural network models for an AI transmitter. During AI modulation by the AI ​​transmitter, modulation-related features are extracted from the signal to be transmitted, and the output signal, which is the modulated signal, is obtained based on the obtained modulation-related features. Alternatively, correlation features are extracted from the received signal to obtain the original signal. For instance, the received signal is processed using one or more neural network models for an AI transmitter. During AI demodulation by an AI receiver, demodulation-related features are extracted from the received signal, and the output signal, which is the original signal corresponding to the input signal, is obtained based on the obtained demodulation-related features.

[0130] For example, the first device in this application embodiment can be a network-side device or a terminal-side device. The network-side device can be a network device or a device served by a network device; the terminal-side device can be a terminal device or a device served by a terminal device, and this application does not limit this. Optionally, when the first device is a network device or a terminal device, method 800 further includes step S820.

[0131] S820. Based on the first information, process the signal to be sent or the signal already received.

[0132] Specifically, based on the feature extraction granularity N indicated by the first information, correlation features are extracted from the signal to be transmitted, or from the received signal. Then, subsequent processing is performed based on the extracted correlation features to obtain the output of the first device.

[0133] For example, when an AI transmitter is deployed on the first device side, the AI ​​transmitter deployed in the first device can be used for AI modulation, AI precoding, AI pilot (reference signal), or generating AI waveforms. As another example, when an AI receiver is deployed on the first device side, the AI ​​receiver deployed in the first device can be used for AI channel estimation, AI equalization, AI detection, or AI demodulation.

[0134] In other words, the first device can be used to perform AI modulation on the signal to be transmitted, or AI precoding, or AI pilot (reference signal), or to generate AI waveforms. Alternatively, the first device can be used to perform AI modulation on the received signal, or AI channel estimation, or equalization, or AI detection, or AI demodulation.

[0135] To more clearly illustrate the process by which the first device processes the signal to be transmitted or the signal already received based on the first information, the following explanation will be based on the example of an AI transmitter deployed on the first device side, and the AI ​​model in the AI ​​transmitter being a self-attention network.

[0136] Figure 9 is a schematic diagram of an AI transmitter deployed on the first device side according to an embodiment of this application. The explanation will be based on an example where the AI ​​transmitter's inputs are bit information and channel information, and its output is frequency domain transmitted symbols.

[0137] As shown in Figure 9, the AI ​​model of this AI transmitter is a self-attention network. The number of REs involved in the extraction of relevance features by the self-attention network is N. That is, the input of the self-attention network is the bit information and channel information corresponding to N REs, and the output of the self-attention network is the transmission symbol corresponding to the N REs. The transmission symbol corresponding to each RE is related to the bits and channel information on the N REs.

[0138] Specifically, the acquired bit information and channel information are converted into tensors of at least two dimensions, where at least two dimensions represent orthogonal frequency division multiplexing (OFDM) symbols and subcarriers, respectively. For example, the tensor corresponding to a bit is B, with dimensions K×L×M, where K represents the number of subcarriers, L represents the number of OFDM symbols, and M is the modulation order, i.e., each RE corresponds to M bits. Considering the channel information in the MIMO system, the dimension of the channel information tensor H is transformed to K×L×N. t ×N r N t N r These represent the number of transmit antennas and receive antennas corresponding to the channel or equivalent channel, respectively. After arranging and merging the acquired bit information and channel information according to the RE (Recursive Array) method, a three-dimensional tensor is obtained. That is, at least one dimension of the aforementioned tensor represents M+N. t ×N r After arranging the data in the RE (Recursive Extraction) manner, the input information corresponding to each RE is M+N. t ×N r The vector.

[0139] Furthermore, the bit information corresponding to these N REs is used as input to a self-attention network, and the signals on these N REs are output through the self-attention network. These N REs represent the number of REs involved in the self-attention network's extraction of relevance features, and N is less than or equal to K×L. The input information of these N REs is X, with a dimension of N×(M+N). t ×N r X can be viewed as having N dimensions and M+N t ×N r The vector.

[0140] For example, the self-attention network performs embedding processing on the input vector corresponding to each RE, obtaining a new vector, denoted as vector #1 (also known as the token vector). Vector #1 has a dimension of N1, where N1 is a hyperparameter of the self-attention network. The embedding processing on the input information X yields a tensor composed of N token vectors. The dimension is N*N1. Here, when the number of REs for joint feature extraction is N, the number of vectors #1 is also N. For ease of description, the following description uses vector #1 as a token vector.

[0141] For example, by performing three linear transformations on each of the N token vectors, we obtain three vectors corresponding to each token vector. The three tensors formed by the three vectors corresponding to the N tokens are Q, K, and V, respectively. That is,

[0142] Furthermore, attention operations are performed on Q, K, and V corresponding to each token vector, as shown in the following formula:

[0143] In the above formula, Q and K are both linear transformations of the same token vector. Therefore, the matrix multiplication of the transpose of Q and K can be understood as calculating the correlation matrix for N token vectors. That is, the feature extraction of the self-attention network is based on the extraction of the correlation features of N token vectors.

[0144] The output of the self-attention network is denoted as tensor X1, where X1 has dimensions N*N1. An AI transmitter can include one or more self-attention networks. For example, for a second layer of self-attention network, X1 is subjected to three linear transformations again to obtain new Q, K, V tensors. Attention operations are then performed on these new Q, K, V tensors to obtain the output X2 of the second layer of self-attention network. The AI ​​transmitter can also include other neural network layers, such as any of the following: multi-layer perceptron (MLP), DNN, CNN, FNN, and RNN. These other neural network layers can be placed before, after, or between two self-attention networks. For example, after the first layer of self-attention network, an MLP is added, i.e., X1' = W1*X1 + b1, where W1 and b1 are the parameters of the MLP. The second layer of self-attention network performs linear transformations on X1' to obtain new Q, K, V tensors. Attention operations are then performed on these new Q, K, V tensors to obtain the output X2 of the second layer of self-attention network.

[0145] Taking the AI ​​transmitter as a single-layer self-attention network as an example, as shown in Figure 9, N REs are transformed into N token vectors, i.e., the matrix in Figure 9. The matrix Let be an N*1 dimensional matrix, where the matrix is... Each row in the matrix represents one of N token vectors, or in other words, the matrix... The matrix has N rows of elements that correspond one-to-one with N token vectors. respectively with W q W K and W v Multiplying these yields Q, K, and V. Further, attention operations are performed on Q, K, and V according to the above formula to obtain X1. A linear transformation is then applied to X to obtain X... out X out The dimension is N*2, which represents the output of the transmitted signals on the N REs. The transmitted signal on each RE is a complex number, and the 2 in the dimension represents the real part and the imaginary part of the complex number.

[0146] Therefore, each of the N output REs corresponds to the information on the N input REs. The AI ​​transmitter can be trained independently or jointly with the AI ​​receiver.

[0147] Let X be the output of the aforementioned AI transmitter. out For example, the training data required for independent training of an AI transmitter includes the AI ​​transmitter's input {B, H} and the label X corresponding to each input. label Tag X label This represents the target of the transmitted signal output by the AI ​​transmitter, which can be obtained, for example, through a traditional non-AI transmitter or by human design. Backpropagation is performed by calculating a loss function to update the parameters of the AI ​​transmitter, making its output as close as possible to the label, thus completing the training of the AI ​​transmitter. For example, the loss function could be the mean squared error (MSE)(X... out ,X label ).

[0148] The joint training of the AI ​​transmitter and AI receiver can be understood as passing the transmitted signal output by the AI ​​transmitter through channel H and adding noise n to obtain the received signal Y = H*X. out +n, the received signal Y is input into the AI ​​receiver, which outputs N log-likelihood ratios (LLRs) on REs. Using the bit B input from the AI ​​transmitter as a label, the loss function is calculated for backpropagation to update the parameters of the AI ​​transmitter, or jointly update the parameters of the AI ​​transmitter and AI receiver, so that the output LLR is as close as possible to the input bit B, thus completing the training of the AI ​​transmitter. For example, the loss function can be binary cross entropy (BCE), i.e., BCE(sigmod(LLR), B), where During this process, the self-attention mechanisms included in both the AI ​​transmitter and the AI ​​receiver are also trained.

[0149] In this embodiment, the first device obtains first information to determine the number of first resources involved in relevance feature extraction, that is, to determine the feature extraction granularity N. This ensures that when an AI receiver is deployed on the first device side, the feature extraction granularity of the AI ​​receiver is consistent with that of the AI ​​transmitter, thereby improving the performance of the first device.

[0150] To facilitate understanding of the solutions in the embodiments of this application, the solutions in the embodiments of this application will be described in detail below with reference to FIG10. FIG10 is described using a first device and a second device as examples of the execution entities. It can be understood that in the embodiments of this application, the first device can be a network-side device, in which case the second device can be a terminal-side device; or, the first device can be a terminal-side device, in which case the second device can be a network-side device.

[0151] The first device can be replaced by components of the first device (e.g., a chip, chip system, circuit, or communication module), and the second device can be replaced by components of the second device (e.g., a chip, chip system, circuit, or communication module). Furthermore, the steps described below as being performed by a single execution entity can also be divided into steps performed by multiple execution entities, which can be logically and / or physically separated. The method 1000 shown in Figure 10 may include the following steps.

[0152] It should be noted that, in the embodiments of this application, method 1000 can be applied to uplink transmission, side-transmission scenarios, or downlink transmission scenarios. For ease of understanding, method 1000 is described from the perspectives of case 1 (i.e., uplink transmission) and case 2 (i.e., downlink transmission). The side-transmission scenario can also be referred to the description of case 1 or case 2, and will not be repeated here. It is understood that the solution in the embodiments of this application is also applicable to other scenarios involving corresponding transmitters and receivers.

[0153] Case 1: In the uplink transmission case, an AI receiver is deployed on the first device side and an AI transmitter is deployed on the second device side, specifically including steps S1010 to S1040.

[0154] S1010, The first device acquires the first information.

[0155] The first information is used to indicate the feature extraction granularity N of the first device. The feature extraction granularity N is used to indicate the number of first resources involved when the first device performs correlation feature extraction. N≥1 and is an integer. The correlation features are used to process the signal to be transmitted or the signal already received.

[0156] The following describes how the first device acquires the first information, using methods one through three.

[0157] Method 1: The first piece of information is pre-configured.

[0158] At this point, the first device acquiring the first information can be understood as the first device acquiring the first information through a pre-configured method. Alternatively, the first device configures its own feature extraction granularity N so that it can process the received signal using this configured granularity. For example, when extracting correlation features from the received signal, N feature extraction units (REs) are used. Then, correlation features are extracted from the received signal using these N REs, thereby obtaining the original signal corresponding to the received signal.

[0159] Optionally, method 1000 further includes: the first device sending first information to the second device. Correspondingly, the second device receives the first information from the first device. This enables the second device, which deploys the AI ​​transmitter, to acquire feature extraction granularity N.

[0160] Method 2: The first piece of information is pre-ordered.

[0161] At this point, the first device acquiring the first information can be understood as acquiring the first information through a predefined method. Alternatively, the feature extraction granularity N is predefined so that the first device can process the received signal using the predefined feature extraction granularity N. For example, when performing correlation feature extraction on the received signal, N feature extraction units (REs) are used. Then, correlation features are extracted from the received signal on these N REs, thereby obtaining the original signal corresponding to the received signal.

[0162] Optionally, method 1000 further includes: the first device sending first information to the second device. Correspondingly, the second device receives the first information from the first device. This enables the second device, which deploys the AI ​​transmitter, to acquire feature extraction granularity N.

[0163] Method 3: The first device receives first information from the second device. In this case, the first device acquires the first information, including receiving the first information from the second device.

[0164] For example, when the first device is a network-side device, the first device acquires the first information, including receiving the first information from the terminal-side device. Taking the network-side device as a base station and the terminal-side device as a UE as an example, when the second device sends the first information to the first device, it can be understood that the UE reports the first information to the base station. At this time, the UE's reporting of the first information can be periodic, semi-persistent, or non-periodic.

[0165] In one implementation, the start time of the effective duration of the first information can be the same as the time when the UE sends the first information to the base station.

[0166] In one implementation, assuming the UE sends the first information to the base station at time t1, the start time of the effective duration of the first information is time t1+a, the end time of the effective duration of the first information is time t1+b, or the end time of the effective duration of the first information is time t1+a+b, or the end time of the effective duration of the first information is time t2, where t2 is the time when the UE will report the first information to the base station again, or the end time of the effective duration of the first information is time t+a, where b≥a, and a and b are predefined or configured by the base station.

[0167] In one implementation, assuming the UE sends the first information to the base station at time t, the start time of the effective duration of the first information is time t+a, and the end time of the effective duration of the first information is time t+b, or the end time of the effective duration of the first information is time t+a+b.

[0168] For example, when the first device is a terminal-side device, the first device acquires the first information, including: receiving the first information from the network-side device.

[0169] Specifically, before the first device receives the first information from the second device, the second device obtains the first information through a pre-configured or pre-defined method, and then the second device sends the first information to the first device.

[0170] S1020, The second device acquires the first information.

[0171] The following describes how the first device acquires the first information, using methods one through three.

[0172] Method 1: The first piece of information is pre-configured.

[0173] At this point, the second device acquiring the first information can be understood as acquiring the first information through a pre-configured method. Alternatively, the second device can configure its own feature extraction granularity N, enabling it to process the received signal using this configured granularity. For example, when extracting correlation features from the received signal, N feature extraction units (REs) are used. Then, correlation features are extracted from these N REs to obtain the original signal corresponding to the received signal.

[0174] Optionally, the second device sends first information to the first device. Correspondingly, the first device receives the first information from the second device.

[0175] Method 2: The first piece of information is pre-ordered.

[0176] At this point, the second device acquiring the first information can be understood as acquiring the first information through a predefined method. Alternatively, the feature extraction granularity N is predefined so that the second device can process the received signal using the predefined feature extraction granularity N. For example, when performing correlation feature extraction on the received signal, N REs are used. Then, correlation features are extracted from the received signal on these N REs, thereby obtaining the original signal corresponding to the received signal.

[0177] Optionally, the second device sends first information to the first device. Correspondingly, the first device receives the first information from the second device.

[0178] Method 3: The second device receives the first information from the first device.

[0179] At this time, the second device acquires the first information, including receiving the first information from the first device.

[0180] For example, when the second device is a network-side device, the second device obtains the first information by receiving the first information from the network-side device.

[0181] For example, when the second device is a terminal-side device, the second device obtains the first information by receiving the first information from the terminal-side device.

[0182] Specifically, before the second device receives the first information from the first device, the first device obtains the first information through a pre-configured or pre-defined method, and then sends the first information to the second device.

[0183] It should be noted that when the first information is pre-configured or pre-defined, the value of the feature extraction granularity N can also be pre-defined or pre-configured. For example, the value of the feature extraction granularity N can be configured through radio resource control (RRC) signaling, medium access control (MAC) signaling, or physical layer signaling. Specifically, the base station can configure at least one value of N, or at least one value of N can be pre-defined through the protocol. Further, the base station indicates the value of the feature extraction granularity N to the UE through downlink control information (DCI). That is, the DCI is the first information.

[0184] It should be noted that the execution order of steps S1010 and S1020 is not limited in the embodiments of this application. For example, step S1010 may occur before step S1020; or step S1010 may occur after step S1020.

[0185] For example, the feature extraction granularity N is any of the following: one or more emission resource groups (TRGs), one or more precoding resource block groups (PRGs), one or more resource block groups (RBGs), or one or more subbands.

[0186] In one implementation, the feature extraction granularity N is one or more transmitter resource groups (TRGs). This can be understood as follows: if a TRG includes N first resources, then the feature extraction granularity N is the same as the number of first resources included in that TRG. Alternatively, if each of multiple TRGs includes the same or different number of first resources, then the feature extraction granularity is the sum of the number of first resources included in these multiple TRGs.

[0187] For example, the first resource may include one or more time-domain resources; and / or, the first resource may include one or more frequency-domain resources; and / or, the first resource may include one or more spatial-domain resources. Wherein, the time-domain resource may be an orthogonal frequency division multiplexing (OFDM) symbol, time slot, or subframe. The frequency-domain resource may be a subcarrier or a resource block (RB). The spatial-domain resource may be a transport stream or a layer.

[0188] As an example, a signal to be transmitted or a signal already received is carried on a first resource set, which includes one or more subsets having the feature extraction granularity N, each of the one or more subsets having the feature extraction granularity N including N of the first resources.

[0189] The first resource set includes one or more subsets with the feature extraction granularity N. This can be understood as follows: when the first resource set includes one subset, it can be considered a TRG (True Resource Group), and in this case, the subset includes N first resources. When the first resource set includes multiple subsets, it can be considered as multiple TRGs, and each of these subsets includes N first resources. In other words, the first resource set includes multiple subsets with the same number of first resources; or, in other words, each subset in the first resource set includes the same number of first resources.

[0190] As another example, the signal to be transmitted or the signal already received is carried on a first resource set, which includes one or more subsets. Each subset includes one or more of the first resources, and the number of the first resources included in each subset is determined by the feature extraction granularity N. At least two subsets among the plurality of subsets have different numbers of the first resources included. For example, the first resource set includes two subsets, subset #1 and subset #2. In this case, subset #1 can be regarded as TRG#1, and subset #2 can be regarded as TRG#2. Then, the number of the first resources included in TRG#1 can be obtained by N× mod N, and TRG#1 can be regarded as the first subset in the first resource set. The number of the first resources included in TRG#2 can be obtained by (X+Y) mod N, and TRG#2 can be regarded as the last subset in the first resource set. Here, X and Y can be predefined or configured by the base station. For example, if the first resource set includes two or more subsets, then, except for the first and last subsets, the number of first resources included in the other subsets is N. The starting position of the first subset is the same as the starting position of the first resource set, and the starting position of the other subsets is the next first resource after the ending position of the previous subset.

[0191] In this context, the starting position of the first resource set is the starting position of the first resource. This can be understood as the position of the first resource in the first resource set being the starting position of the first resource.

[0192] For example, each subset includes multiple first resources that are continuous in one or more of the time domain, frequency domain, or spatial domain. This can be understood as follows: when each subset includes multiple first resources, these multiple first resources can be continuous time-domain resources or frequency-domain resources.

[0193] In one implementation, the feature extraction granularity N is one or more precoding resource block groups (PRGs). When the feature extraction granularity N is a single PRG, it can be understood that a PRG includes N resources; in this case, the feature extraction granularity N is the same as the number of resources included in that PRG. Alternatively, when the feature extraction granularity N is multiple PRGs, each PRG may include the same or different number of resources, and the feature extraction granularity is the sum of the number of resources included in these multiple PRGs. Here, PRG indicates the granularity of the precoding resources.

[0194] In one implementation, the feature extraction granularity N is one or more RBGs. When the feature extraction granularity N is a single RBG, it can be understood that the number of resources included in that RBG is N; therefore, the feature extraction granularity N is the same as the number of resources included in that RBG. Alternatively, when the feature extraction granularity N is multiple RBGs, the number of resources included in each of the multiple RBGs can be the same or different; in this case, the feature extraction granularity is the sum of the number of resources included in these multiple RBGs. Here, RBG is used to indicate the granularity of the allocated frequency domain resources.

[0195] In one implementation, the feature extraction granularity N is one or more subbands. When the feature extraction granularity N is a single subband, it can be understood that the number of resources included in a subband is N; therefore, the feature extraction granularity N is the same as the number of resources included in that subband. Alternatively, when the feature extraction granularity is multiple subbands, and each of the multiple subbands includes a different number of resources, then the feature extraction granularity is the sum of the number of resources included in these multiple subbands. The subband is used to indicate the granularity of feedback channel state information (CSI). The granularity of feedback CSI can be understood as the granularity of the resources occupied by CSI in the time or frequency domain.

[0196] Optionally, the first information is further used to indicate a first granularity, which is the granularity of the resources used by the signal to be transmitted or the signal already received. For example, the first granularity is the number of resources included in the PRG, or the first granularity is the number of resources included in the RBG, or the first granularity is the number of resources included in the subband.

[0197] Optionally, the first information may also indicate the relationship between the first granularity and the feature extraction granularity N, such as the number of first granularities included in the feature extraction granularity N, that is, the multiple of the first granularity.

[0198] Optionally, the relationship between the first granularity and the feature extraction granularity N can be preset.

[0199] For example, when the feature extraction granularity N is one or more PRGs, or when the feature extraction granularity N is one or more RBGs, or when the feature extraction granularity N is one or more subbands, method 1000 may further include the following steps.

[0200] S1001, The first device acquires information of the first granularity.

[0201] Wherein, the first granularity is the granularity of the resources used by the signal to be transmitted or the signal already received; the first information is used to indicate that the feature extraction granularity is the same as the first granularity. For example, the information of the first granularity can be predefined or pre-configured. Pre-configuration can also be understood as the base station configuring itself.

[0202] S1002, The first device sends the second information to the second device.

[0203] Correspondingly, the second device receives the second information from the first device.

[0204] The second information configures the resources used by the signal to be sent or the signal already received, and the configuration adopts the first granularity.

[0205] For example, this configuration using the first granularity can be understood as the first device configuring the first granularity itself and then indicing the first granularity to the second device. The first granularity can be the same as the precoding granularity indicated by the PRG. Alternatively, the first granularity can be the same as the frequency domain resource allocation granularity indicated by the RBG. Yet another example is that the first granularity can be the same as the feedback CSI granularity indicated by the subband.

[0206] For example, the second information can also be predefined.

[0207] For a detailed description of the feature extraction granularity N, the first resource, and the relevance features, please refer to step S810 above, which will not be repeated here.

[0208] S1030. The second device processes the signal to be transmitted based on the first information.

[0209] For example, processing the signal to be transmitted includes: processing the signal to be transmitted using one or more neural network models for an AI transmitter.

[0210] In this process, the signal to be transmitted is processed using one or more neural network models used in the AI ​​transmitter. This can be understood as the AI ​​transmitter deployed on the second device side using one or more neural network models to process the signal to be transmitted.

[0211] For example, one or more neural network models for an AI transmitter are used for at least one of the following: AI modulation, AI precoding, AI piloting, and generating AI waveforms.

[0212] In one implementation, the signal to be transmitted is processed using a neural network model of an AI transmitter. For example, this neural network model is used for AI modulation. After extracting relevant features from the signal to be transmitted at a feature extraction granularity N, the neural network model is used to modulate the extracted relevant features to obtain a modulated signal, such as a first signal.

[0213] In one implementation, the signal to be transmitted is processed using multiple neural network models of an AI transmitter. For example, these multiple neural network models include neural network model #1 and neural network model #2. Neural network model #1 is used for AI modulation, and neural network model #2 is used for AI precoding. Both neural network models #1 and #2 involve N first resources when extracting correlation features from the signal to be transmitted. Neural network model #1 can extract correlation features from the signal to be transmitted according to feature extraction granularity N, and then modulate the extracted correlation features to obtain a modulated signal, such as the first signal. Neural network model #2 can extract correlation features from the signal to be transmitted according to feature extraction granularity N, and then precode the extracted correlation features to obtain a precoded signal, such as the first signal. Specifically, neural network model #1 can first modulate the signal to be transmitted, and then input the modulated signal into neural network model #2, using neural network model #2 to precode the modulated signal to obtain the first signal. Alternatively, neural network model #2 precodes the signal to be transmitted to obtain a precoded signal, and then inputs the precoded signal into neural network model #1. Neural network model #1 modulates the precoded signal to obtain the first signal.

[0214] Furthermore, method 1000 may also include step S1031.

[0215] S1031, The second device sends a first signal to the first device. Correspondingly, the first device receives the first signal from the second device.

[0216] S1040. The first device processes the received signal based on the first information.

[0217] For example, processing the received signal includes: processing the received signal using one or more neural network models for an AI receiver.

[0218] The process of processing the received signals using one or more neural network models for the AI ​​receiver can be understood as the AI ​​receiver deployed on the first device side processing the received signals using one or more neural network models.

[0219] For example, one or more neural network models for an AI receiver are used for at least one of the following: AI channel estimation, AI equalization, AI detection, and AI demodulation.

[0220] In one implementation, the received signal is processed using a neural network model of an AI receiver. Here, the received signal can be the aforementioned first signal. For example, this neural network model is used for AI demodulation. This neural network can extract relevant features from the received signal at a feature extraction granularity N, and then demodulate the extracted relevant features to obtain the demodulated signal, which is the signal to be transmitted corresponding to the first signal.

[0221] In one implementation, the received signal is processed using multiple neural network models of the AI ​​transmitter. For example, these multiple neural network models include neural network model #1 and neural network model #2. Neural network model #1 is used for AI demodulation, and neural network model #2 is used for AI channel estimation. Both neural network models #1 and #2 involve N first resources when extracting correlation features from the signal to be transmitted. Neural network model #1 can then extract correlation features from the received signal at a feature extraction granularity of N, and then demodulate the extracted correlation features to obtain the demodulated signal, i.e., the signal to be transmitted corresponding to the first signal. Neural network model #2 can also extract correlation features from the received signal at a feature extraction granularity of N, and then perform channel estimation on the extracted correlation features to obtain the channel-estimated signal. Specifically, neural network model #1 can first demodulate the received signal, and then input the demodulated signal into neural network model #2, which then performs channel estimation on the demodulated signal. Alternatively, neural network model #2 performs channel estimation on the received signal, and then inputs the channel-estimated signal into neural network model #1. Neural network model #1 demodulates the channel-estimated signal to obtain the signal to be transmitted corresponding to the first signal.

[0222] Case 2: In the downlink transmission case, an AI transmitter is deployed on the first device side and an AI receiver is deployed on the second device side, specifically including steps S1050 to S1080.

[0223] S1050, The first device acquires the first information.

[0224] The following describes how the first device acquires the first information, using methods one and two as examples.

[0225] Method 1: The first piece of information is pre-configured.

[0226] At this point, the first device acquiring the first information can be understood as the first device acquiring the first information through a pre-configured method. Alternatively, the first device configures its own feature extraction granularity N so that it can process the received signal using this configured granularity. For example, when extracting correlation features from the received signal, N feature extraction units (REs) are used. Then, correlation features are extracted from the received signal using these N REs, thereby obtaining the original signal corresponding to the received signal.

[0227] Method 2: The first piece of information is pre-ordered.

[0228] At this point, the first device acquiring the first information can be understood as acquiring the first information through a predefined method. Alternatively, the feature extraction granularity N is predefined so that the first device can process the received signal using the predefined feature extraction granularity N. For example, when performing correlation feature extraction on the received signal, N feature extraction units (REs) are used. Then, correlation features are extracted from the received signal on these N REs, thereby obtaining the original signal corresponding to the received signal.

[0229] Optionally, method 1000 further includes: the first device sending first information to the second device. Correspondingly, the second device receives the first information from the first device. This enables the second device, which deploys the AI ​​transmitter, to acquire feature extraction granularity N.

[0230] S1060, The second device acquires the first information.

[0231] The following describes how the first device acquires the first information, using methods one through three.

[0232] Method 1: The first piece of information is pre-configured.

[0233] At this point, the second device acquiring the first information can be understood as acquiring the first information through a pre-configured method. Alternatively, the second device can configure its own feature extraction granularity N, enabling it to process the received signal using this configured granularity. For example, when extracting correlation features from the received signal, N feature extraction units (REs) are used. Then, correlation features are extracted from these N REs to obtain the original signal corresponding to the received signal.

[0234] Optionally, the second device sends first information to the first device. Correspondingly, the first device receives the first information from the second device.

[0235] Method 2: The first piece of information is pre-ordered.

[0236] At this point, the second device acquiring the first information can be understood as acquiring the first information through a predefined method. Alternatively, the feature extraction granularity N is predefined so that the second device can process the received signal using the predefined feature extraction granularity N. For example, when performing correlation feature extraction on the received signal, N REs are used. Then, correlation features are extracted from the received signal on these N REs, thereby obtaining the original signal corresponding to the received signal.

[0237] Optionally, the second device sends first information to the first device. Correspondingly, the first device receives the first information from the second device.

[0238] Method 3: The second device receives the first information from the first device.

[0239] At this time, the second device acquires the first information, including receiving the first information from the first device.

[0240] For example, when the second device is a network-side device, the second device obtains the first information by receiving the first information from the network-side device.

[0241] For example, when the second device is a terminal-side device, the second device obtains the first information by receiving the first information from the terminal-side device.

[0242] Specifically, before the second device receives the first information from the first device, the first device obtains the first information through a pre-configured or pre-defined method, and then sends the first information to the second device.

[0243] For a detailed description of steps S1050 and S1060, please refer to steps S1010 and S1020 above, which will not be repeated here.

[0244] It should be noted that when the feature extraction granularity N is one or more PRGs, or when the feature extraction granularity N is one or more RBGs, or when the feature extraction granularity N is one or more subbands, method 1000 further includes the following steps.

[0245] S1050a, The first device acquires information about the first particle size.

[0246] S1050b, The first device sends the second information to the second device.

[0247] Correspondingly, the second device receives the second information from the first device.

[0248] For details regarding steps S1050a and S1050b, please refer to steps S1001 and S1002 above, which will not be repeated here.

[0249] S1070. The first device processes the signal to be transmitted based on the first information.

[0250] The detailed description of how the first device processes the signal to be sent based on the first information can be found in step S1030 above. It is only necessary to replace the second device as the executing entity with the first device and the first signal with the second signal.

[0251] Furthermore, method 1000 also includes step S1071.

[0252] S1071, The first device sends a second signal to the second device. Correspondingly, the second device receives the second signal from the first device.

[0253] For a detailed description of step S1071, please refer to step S1030 above. Simply replace the first signal with the second signal.

[0254] S1080. The second device processes the received signal based on the first information.

[0255] The detailed description of how the second device processes the received signal based on the first information can be found in step S1040 above. It is only necessary to replace the execution subject with the second device and the first signal with the second signal.

[0256] In this embodiment, the first device obtains first information to determine the number of first resources involved in relevance feature extraction, i.e., to determine the feature extraction granularity N. This ensures that when an AI transmitter is deployed on the first device side, the AI ​​receiver deployed on the second device side also obtains the feature extraction granularity N, or vice versa. This maintains consistency in the feature extraction granularity between the AI ​​transmitter and the AI ​​receiver, thereby improving the performance of the AI ​​receiver.

[0257] Figure 11 is a schematic diagram of a communication device 1200 provided in an embodiment of this application. The communication device 1200 includes a transceiver unit 1210 and a processing unit 1220. The transceiver unit 1210 can be used to implement corresponding communication functions. The transceiver unit 1210 can also be referred to as a communication interface or a communication unit. The processing unit 1220 can be used to perform processing, such as determining information bits.

[0258] Optionally, the device 1200 may further include a storage unit, which can be used to store instructions and / or data, and the processing unit 1220 can read the instructions and / or data in the storage unit to enable the device to implement the aforementioned method embodiments.

[0259] In a first possible design, the device 1200 can be the first device in the foregoing embodiments, which can implement the steps or processes corresponding to those performed by the first device in the above method embodiments. Specifically, the transceiver unit 1210 can be used to perform transceiver-related operations (such as sending and / or receiving data or messages) of the first device in the above method embodiments, and the processing unit 1220 can be used to perform processing-related operations of the first device in the above method embodiments, or operations other than transceiver (such as operations other than sending and / or receiving data or messages).

[0260] In one possible implementation, the processing unit 1220 is used to acquire first information, which is used to indicate the feature extraction granularity N of the first device. The feature extraction granularity N is used to indicate the number of first resources involved when the first device performs correlation feature extraction, where N≥1 and is an integer. The correlation features are used to process the signal to be transmitted or the signal already received.

[0261] Optionally, when the first device is a network-side device, the transceiver unit 1210 is used to receive the first information from the terminal-side device; or when the first device is a terminal-side device, the transceiver unit 1210 is used to receive the first information from the network-side device.

[0262] It should be understood that the specific process of each unit performing the above-mentioned corresponding steps has been described in detail in the above method embodiments, and will not be repeated here for the sake of brevity.

[0263] It should also be understood that the device 1200 here is embodied in the form of a functional unit. The term "unit" here can refer to an application-specific integrated circuit (ASIC), electronic circuitry, a processor (e.g., a shared processor, a proprietary processor, or a group processor, etc.) and memory for executing one or more software or firmware programs, integrated logic circuitry, and / or other suitable components supporting the described functions. In an alternative example, those skilled in the art will understand that the device 1200 can be specifically the communication device in the above embodiments, and can be used to execute the various processes and / or steps corresponding to the communication device in the above method embodiments; to avoid repetition, these will not be described again here.

[0264] The apparatus 1200 of each of the above-described schemes has the function of implementing the corresponding steps performed by the communication device (such as the first device) in the above-described methods. The function can be implemented in hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions; for example, the transceiver unit can be replaced by a transceiver (e.g., the transmitting unit in the transceiver unit can be replaced by a transmitter, and the receiving unit in the transceiver unit can be replaced by a receiver), and other units, such as processing units, can be replaced by processors, each performing the transceiver operations and related processing operations in the respective method embodiments.

[0265] In addition, the transceiver unit 1210 may also be a transceiver circuit (for example, it may include a receiving circuit and a transmitting circuit), and the processing unit may be a processing circuit.

[0266] It should be noted that the device in Figure 11 can be the communication device (such as the first device) in the foregoing embodiments, or it can be a chip or a chip system, such as a system on a chip (SoC). The transceiver unit can be an input / output circuit or a communication interface; the processing unit is a processor, microprocessor, or integrated circuit integrated on the chip. No limitations are imposed here.

[0267] Figure 12 is a schematic diagram of another communication device 1300 provided in an embodiment of this application. The device 1300 includes a processor 1310, which is coupled to a memory 1320. The memory 1320 is used to store computer programs or instructions and / or data. The processor 1310 is used to execute the computer programs or instructions stored in the memory 1320, or to read the data stored in the memory 1320, so as to execute the methods in the above method embodiments.

[0268] Optionally, there may be one or more processors 1310.

[0269] Optionally, the memory 1320 may be one or more.

[0270] Alternatively, the memory 1320 can be integrated with the processor 1310, or it can be set separately.

[0271] Optionally, as shown in FIG12, the device 1300 further includes a transceiver 1330 for receiving and / or transmitting signals. For example, a processor 1310 is used to control the transceiver 1330 to receive and / or transmit signals.

[0272] As an example, processor 1310 may have the functions of processing unit 1220 shown in FIG11, memory 1320 may have the functions of storage unit, and transceiver 1330 may have the functions of transceiver unit 1210 shown in FIG11.

[0273] As one option, the device 1300 is used to implement the operations performed by the communication device (such as the first device) in the various method embodiments described above.

[0274] For example, processor 1310 is used to execute computer programs or instructions stored in memory 1320 to implement the relevant operations of the communication device in the various method embodiments described above.

[0275] It should be understood that the processor mentioned in the embodiments of this application can be a central processing unit (CPU), a processor used for AI, or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0276] For example, a processor for AI may include one or more of the following: graphics processing unit (GPU), neural processing unit (NPU), tensor processing unit (TPU), and data processing unit (DPU).

[0277] For example, one possible implementation of a processor for AI could be the AI ​​processor 1400 shown in Figure 13.

[0278] As shown in Figure 13, the AI ​​processor 1400 may include one or more of the following: an AI core, a digital vision pre-processing (DVPP) module, a task scheduler (TS), an L3 cache, an AI CPU, a control CPU, an L2 cache, a universal serial bus (USB) interface, a network interface card (NIC), a peripheral component interconnect express (PCIe) interface (PCIe is a high-speed serial computer expansion bus standard), a double data rate (DDR) / high bandwidth memory (HBM) interface, a generation purpose input / output (GPIO) / inter-integrated circuit (I2C) bus, etc. It is understood that the specific meanings of these terms are well known to those skilled in the art and will not be elaborated upon here.

[0279] It should also be understood that the memory mentioned in the embodiments of this application can be volatile memory and / or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM). For example, RAM can be used as an external cache. By way of example and not limitation, RAM includes the following forms: static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0280] It should be noted that when the processor is a general-purpose processor, DSP, ASIC, FPGA, or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, the memory (storage module) can be integrated into the processor.

[0281] It should also be noted that the memory described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0282] Figure 14 is a schematic diagram of a chip system 1500 provided in an embodiment of this application. The chip system 1500 (or may also be called a processing system) includes logic circuitry 1510 and an input / output interface 1520.

[0283] The logic circuit 1510 can be a processing circuit in the chip system 1500. The logic circuit 1510 can be coupled to a memory unit, calling instructions from the memory unit, enabling the chip system 1500 to implement the methods and functions of the embodiments of this application. The input / output interface 1520 can be an input / output circuit in the chip system 1500, outputting processed information from the chip system 1500, or inputting data or signaling information to be processed into the chip system 1500 for processing.

[0284] As one option, the chip system 1500 is used to implement the operations performed by the communication device (such as the first device) in the various method embodiments described above.

[0285] For example, logic circuit 1510 is used to implement processing-related operations performed by a communication device (such as the first device) in the above method embodiments; input / output interface 1520 is used to implement sending and / or receiving-related operations performed by a communication device (such as the first device) in the above method embodiments.

[0286] This application also provides a computer-readable storage medium storing a computer program or instructions for implementing the methods executed by a communication device (such as a first device) in the above-described method embodiments. For example, when the computer program or instructions are run on the communication device, the communication device (such as the first device) causes the communication device (such as method 1000) to execute the above-described methods.

[0287] This application also provides a computer program product comprising instructions that, when executed by a computer, implement the methods performed by a communication device (such as a first device) in the above-described method embodiments. For example, when the computer program or instructions are run on the communication device, the communication device (such as the first device) performs the above-described methods (such as method 1000).

[0288] This application also provides a communication system, which includes a first device and a second device in the embodiments described above. For example, the system includes the first device and the second device in the embodiment of FIG10.

[0289] The explanations and beneficial effects of the relevant contents in any of the devices provided above can be found in the corresponding method embodiments provided above, and will not be repeated here.

[0290] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of apparatus or units may be electrical, mechanical, or other forms.

[0291] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. For example, the computer can be a personal computer, a server, or a network device, etc. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks, SSDs). For example, the aforementioned available media include, but are not limited to, various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0292] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A communication method, characterized in that, The method, which applies to a first device or a chip in a first device, includes: Obtain first information, which is used to indicate the feature extraction granularity N of the first device. The feature extraction granularity N is used to indicate the number of first resources involved when the first device extracts relevant features. N ≥ 1 and is an integer. The relevant features are used to process the signal to be sent or the signal already received.

2. The method according to claim 1, characterized in that, The method further includes: Based on the first information, the signal to be sent or the signal already received is processed.

3. The method according to claim 1 or 2, characterized in that, Processing the signal to be transmitted includes: The signal to be transmitted is processed using one or more neural network models for artificial intelligence (AI) transmitters; The processing of the received signal includes: The received signal is processed using one or more neural network models for an AI receiver.

4. The method according to claim 3, characterized in that, The one or more neural network models used for the AI ​​transmitter are used in at least one of the following: AI modulation, AI precoding, AI piloting, and AI waveform generation; The one or more neural network models used in the AI ​​receiver are for at least one of the following: AI channel estimation, AI equalization, AI detection, and AI demodulation.

5. The method according to any one of claims 1-4, characterized in that, The feature extraction granularity N is one of the following: One or more transmit resource groups (TRGs), one or more precoding resource block groups (PRGs), one or more resource block groups (RBGs), and one or more subbands.

6. The method according to any one of claims 1-5, characterized in that, The first information is also used to indicate a first granularity, which is the granularity of the resources used by the signal to be transmitted or the signal already received; or, The method further includes: Obtain information at a first granularity, where the first granularity is the granularity of the resources used by the signal to be sent or the signal already received; The first information is used to indicate that the feature extraction granularity N is the same as the first granularity.

7. The method according to claim 6, characterized in that, The method further includes: Sending or receiving second information, the second information configuring the resources used by the signal to be sent or the signal already received, the configuration of the resources adopting the first granularity.

8. The method according to any one of claims 1-7, characterized in that, The signal to be transmitted or the signal already received is carried on a first resource set, which includes one or more subsets having the feature extraction granularity N, and each subset of the one or more subsets having the feature extraction granularity N includes N of the first resources.

9. The method according to any one of claims 1-7, characterized in that, The signal to be transmitted or the signal already received is carried on a first resource set, the first resource set including one or more subsets, each subset including one or more of the first resources, the number of the first resources included in each subset being determined by the feature extraction granularity N, and at least two subsets among the plurality of subsets having different numbers of the first resources.

10. The method according to claim 8 or 9, characterized in that, The starting position of the first resource set is the starting position of the first resource.

11. The method according to any one of claims 8-10, characterized in that, Each subset includes multiple first resources that are consecutive in one or more of the time domain, frequency domain, or spatial domain.

12. The method according to any one of claims 1-11, characterized in that, N is predefined in the protocol.

13. The method according to any one of claims 1-12, characterized in that, Also includes: Send the first message.

14. The method according to any one of claims 1-13, characterized in that, The step of obtaining the first information includes: receiving the first information.

15. A communication device, characterized in that, Includes modules or units for performing the method according to any one of claims 1 to 14.

16. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program or instructions that, when executed on a communication device, cause the communication device to perform the method as described in any one of claims 1 to 14.

17. A computer program product, characterized in that, The computer program product includes a computer program or instructions that, when executed on a communication device, cause the communication device to perform the method as described in any one of claims 1 to 14.

Images