Wireless communication method and related device

Abstract

The embodiment of the invention provides a wireless communication method and a related device, and the method comprises the steps: building a merging model parameter through employing the model parameters of a CSI prediction model locally trained by a plurality of UE, and transmitting the merging model parameter to the UE which needs to update the local CSI prediction model, thereby enabling the UE to directly update the local CSI prediction model through employing the received merging model parameter, and improving the user experience. According to the embodiment of the invention, local training is not needed, and the generalization of the CSI prediction model is improved and the universality of the model among different UEs is improved by the combination model obtained according to the local CSI prediction models of the multiple UEs. Moreover, the model parameters of the local models of the multiple UEs with similar CSI data distribution are combined to obtain the combined model parameters, so that the reasonability of the models participating in combination is ensured.

Description

Technical Field

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

[0002] In the fifth-generation (5G) mobile communication system, a network device needs to obtain channel state information (CSI) between a terminal device and the network device, and further perform resource scheduling for uplink / downlink data transmission according to the CSI.

[0003] The CSI prediction scheme based on artificial intelligence (AI) / machine learning (ML) is: inputting a plurality of historical CSI information and / or current CSI information into a CSI prediction model to output a prediction result of the CSI at a future moment.

[0004] However, locally training and updating the CSI prediction model based on the terminal device may result in poor generalization of the CSI prediction model, and moreover, the update speed is slow and a large amount of data is required. Summary of the Invention

[0005] In view of this, this application provides a wireless communication method and related devices to solve at least some of the above problems, and the disclosed technical solutions are as follows:

[0006] In a first aspect, this application provides a wireless communication method applied to a first terminal. The method includes: sending a model update request message for requesting to update the model parameters of the CSI prediction model of the first terminal, where the model update request message includes a first CSI data distribution feature sequence for characterizing the distribution features of the CSI data collected by the first terminal; receiving a model update indication message sent by a network device for indicating the first terminal to update the model parameters of the CSI prediction model, where the model update indication message includes fusion model parameters corresponding to the CSI prediction model, and the fusion model parameters are obtained by the network device according to the model parameters of the CSI prediction models in at least one second terminal; and responding to the model update information, and updating the model parameters of the CSI prediction model based on the fusion model parameters carried in the model update indication message.

[0007] In this way, the model parameters of the CSI prediction models locally trained by multiple UEs are used to construct combined model parameters, and the combined model parameters are sent to the UEs that need to update the local CSI prediction models. In this way, the UEs can directly update the local CSI prediction models by using the received combined model parameters without local training. Moreover, the combined model obtained based on the local CSI prediction models of multiple UEs improves the generalization of the CSI prediction model and the universality of the model among different UEs. Among them, the model parameters of the local models of multiple UEs with similar CSI data distributions are combined to obtain the combined model parameters, thereby ensuring the rationality of the models participating in the combination.

[0008] In a possible implementation manner of the first aspect, the process of obtaining the first CSI data distribution feature sequence includes: performing Hash encoding on the obtained CSI data to obtain the first CSI data distribution feature sequence. In this way, encoding the CSI data of the UE can greatly reduce the data volume, thereby reducing the communication overhead. Moreover, using the distance between the encoded feature sequences to determine the similarity of data distributions simplifies the judgment process and also provides a basis for reasonably selecting models for model combination.

[0009] In a possible implementation manner of the first aspect, performing Hash encoding on the obtained CSI data to obtain the first CSI data distribution feature sequence includes: generating K random complex vectors, where K represents the number of carrier frequencies; extracting L the most recently obtained CSI tensors from the CSI data set obtained from the first terminal, and respectively converting them into CSI feature complex vectors, where the dimensions of the random complex vectors and the CSI feature complex vectors are the same; respectively performing inner product operations on each CSI feature complex vector and K random complex vectors to obtain K × L inner product operation results, and converting each inner product operation result into binary bits; arranging the K × L binary bits in sequence to obtain the first CSI data distribution feature sequence.

[0010] In a possible implementation manner of the first aspect, respectively performing inner product operations on each CSI feature complex vector and K random complex vectors to obtain K × L inner product operation results, and converting each inner product operation result into binary bits includes: converting each inner product operation result into binary bits according to the following formula:

[0011]

[0012] where Denote a random complex vector The k-th complex vector in Denote the UE i The CSI eigen complex vector obtained by converting the CSI tensor at time t Denote And The binary bit obtained after performing the inner product calculation Denote And The real part of the inner product operation result of is greater than 0 Denote the most recent L acquisition times, and S denotes the total number of CSI tensors included in the CSI dataset

[0013] In a possible implementation of the first aspect, arrange K × L Binary bits in order to obtain the first CSI data distribution feature sequence, including: arranging according to the following formula

[0014]

[0015] Where Denote the CSI data distribution feature sequence corresponding to the CSI dataset of the UE i

[0016] ​Second aspect, the present application also provides a wireless communication method, which is applied to a network device. The method includes: receiving a model update request message sent by a first terminal, where the model update request message is used to request an update of the model parameters of the CSI prediction model of the first terminal, and the model update request message includes a first CSI data distribution feature sequence, and the first CSI data distribution feature sequence is used to characterize the distribution features of the CSI data collected by the first terminal; sending a similarity discrimination indication message to a second terminal, where the similarity discrimination indication message is used to instruct the second terminal to determine the similarity between its own CSI data distribution features and the first CSI data distribution feature sequence of the first terminal, and the similarity discrimination indication message includes a model merging process ID and the first CSI data distribution feature sequence; receiving a similarity discrimination response message sent by the second terminal, where the similarity discrimination response message includes the device ID of the second terminal, the model merging process ID, and the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal; determining a candidate terminal list for participating in model merging according to the similarities in each similarity discrimination response message, where the candidate terminal list includes the device IDs corresponding to the candidate terminals; sending a model upload indication message, where the model upload indication message is used to instruct the second terminal to report the CSI prediction model parameters, and the model upload indication message includes the model merging process ID and the candidate terminal list; receiving a model upload response message, where the model upload response message includes the model parameters of the CSI prediction model in the second terminal; performing model parameter fusion on the model parameters sent by each second terminal to obtain fused model parameters; sending a model update indication message to the first terminal, where the model update indication message is used to instruct the first terminal to update the model parameters of the CSI prediction model, and the model update indication message includes the model merging process ID and the fused model parameters.

[0017] In a possible implementation manner of the second aspect, the method further includes: within a preset waiting time threshold after sending the similarity discrimination indication message, if the similarity discrimination response message is not received, sending a model training indication message to the first terminal, where the model training indication message is used to instruct the first terminal to perform local training update on the CSI prediction model, and the model training indication message includes the model merging process ID.

[0018] In a possible implementation manner of the second aspect, determining a candidate terminal list for participating in model merging according to the similarities in each similarity discrimination response message includes: selecting a preset number of second terminals in descending order of similarity to obtain the candidate terminal list. In this way, the model parameters of the CSI prediction models of other terminals whose CSI data distribution is similar to that of the terminal initiating the model update request can be selected for update through the CSI data distribution similarity, improving the rationality and usability of model merging.

[0019] In a possible implementation of the second aspect, a candidate terminal list is obtained by selecting a preset number of second terminals in descending order of similarity, including: selecting a preset number of second terminals in ascending order of the distance between the CSI data distribution feature sequence of each second terminal and the first CSI data distribution feature sequence to obtain a candidate terminal list.

[0020] In a possible implementation of the second aspect, a candidate terminal list is obtained by selecting a preset number of second terminals in ascending order of the distance between the CSI data distribution feature sequence of each second terminal and the first CSI data distribution feature sequence, including: determining a first number of second terminals that send similarity discrimination response messages; if the first number is less than a preset number threshold, determining that all second terminals that send similarity discrimination response messages form a candidate terminal list; if the first number is greater than the preset number threshold, selecting a preset number threshold from the first number in ascending order of the distance between the CSI data distribution feature sequence of each second terminal and the first CSI data distribution feature sequence to form a candidate terminal list. In this way, the number of terminals participating in model merging can be reasonably restricted, avoiding too long time to obtain the fused model parameters due to too many terminals participating in model merging, and improving the speed of the network device responding to the model update request.

[0021] In a possible implementation of the second aspect, model parameter fusion is performed on the model parameters sent by each second terminal to obtain fused model parameters, including: obtaining an average model parameter corresponding to the model parameters sent by each second terminal by using the arithmetic mean method, and obtaining the average variance value between each received model parameter and the average model parameter; normalizing the negative exponent of each average variance value to obtain a weight coefficient corresponding to each received model parameter; performing weighted summation on each received model parameter to obtain fused model parameters.

[0022] In a third aspect, the present application further provides a wireless communication method, which is applied to a second terminal. The method includes: receiving a similarity discrimination indication message sent by a network device, where the similarity discrimination indication message is used to instruct the second terminal to determine the similarity between its own CSI data distribution feature sequence and the first CSI data distribution feature sequence of a first terminal, and the similarity discrimination indication message includes a model merging process ID and the first CSI data distribution feature sequence, and the first CSI data distribution feature sequence is used to characterize the distribution feature of the CSI data collected by the first terminal; responding to the similarity discrimination indication message, and obtaining the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal; when the similarity meets a preset condition, sending a similarity discrimination response message to the network device, where the similarity discrimination response message includes the device ID of the second terminal, the model merging process ID, and the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal; receiving a model upload indication message sent by the network device, where the model upload indication message is used to instruct the second terminal to report CSI prediction model parameters, and the model upload indication message includes the model merging process ID and a candidate terminal list, and the candidate terminal list is obtained by the network device according to the similarities between the CSI data distribution feature sequences of each second terminal and the first terminal; after determining that the second terminal is included in the candidate terminal list, sending the model parameters of the local CSI prediction model to the network device, and the model parameters are used to enable the network device to obtain the fusion model parameters corresponding to the CSI prediction model.

[0023] In a possible implementation manner of the third aspect, responding to the similarity discrimination indication message and obtaining the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal includes: responding to the similarity discrimination indication message, and obtaining the model merging process ID and the first CSI data distribution feature sequence; obtaining the retention time of the CSI prediction model of the second terminal, and if the retention time is greater than a preset retention time threshold, obtaining the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal. In this way, it can be ensured that the CSI prediction model of the second terminal obtains a sufficient number of model monitoring data, and the model will be used to participate in the model update of other terminals only after it is stable, ensuring the stability of the models participating in the merging.

[0024] In a possible implementation manner of the third aspect, obtaining the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal includes: calculating the Hamming distance between the CSI data distribution feature sequence of the second terminal and the first CSI data distribution feature sequence of the first terminal, and the Hamming distance is negatively correlated with the similarity.

[0025] In a possible implementation manner of the third aspect, the similarity meeting the preset condition includes: the Hamming distance is less than a preset distance threshold.

[0026] Fourthly, the present application further provides a communication device, including a processing module and a transceiver module. The communication device is configured to execute the method according to any one of the first to third aspects.

[0027] Fifthly, the present application further provides a communication device, including: a memory for storing computer instructions; a processor for executing the computer program or computer instructions stored in the memory, so that the communication device executes the method according to any one of the first to third aspects.

[0028] Sixthly, the present application further provides a computer storage medium for storing a computer program, which is used to implement the method according to any one of the first to third aspects when the computer program is executed.

[0029] Seventhly, the present application further provides a computer program product, the computer program of which, when run, causes the method according to any one of the first to third aspects to be executed. Description of the Drawings

[0030] Figure 1 It is a schematic structural diagram of a communication system;

[0031] Figure 2 It is a flowchart of a wireless communication method provided by an embodiment of the present application;

[0032] Figure 3 It is a flowchart of another wireless communication method provided by an embodiment of the present application;

[0033] Figure 4 It is a flowchart of yet another wireless communication method provided by an embodiment of the present application;

[0034] Figure 5 It is a schematic structural diagram of a communication device provided by an embodiment of the present application;

[0035] Figure 6 It is a schematic structural diagram of another communication device provided by an embodiment of the present application. Detailed Embodiments

[0036] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. The terms used in the following embodiments are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and appended claims of the present application, the singular forms "a", "an", "the", "above", "said", "this" are also intended to include the expression "one or more", unless the context clearly indicates otherwise. It should also be understood that in the embodiments of the present application, "one or more" means one, two or more than two; "and / or" describes the association relationship of associated objects and means that three relationships may exist; for example, A and / or B may mean: A exists alone, A and B exist simultaneously, or B exists alone, where A and B may be singular or plural. The character " / " generally means an "or" relationship between the associated objects before and after.

[0037] Reference to "one embodiment" or "some embodiments" or the like described in this specification means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the present application. Thus, the phrases "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in another way. The terms "comprising", "including", "having" and their variants all mean "including but not limited to", unless otherwise specifically emphasized in another way.

[0038] The "multiple" involved in the embodiments of the present application means two or more. It should be noted that in the description of the embodiments of the present application, the terms "first", "second", etc. are only used for the purpose of distinguishing descriptions and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying an order.

[0039] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: 3GPP communication systems, fourth-generation (4G) mobile communication technologies, such as long term evolution (LTE) systems, 5G mobile communication systems, 5G new radio (NR) communication systems, vehicle-to-everything (NRV2X) systems, and can also be applied to systems with hybrid networking of LTE and 5G, or non-terrestrial network (NTN) systems, device-to-device (D2D) communication systems, machine-to-machine (M2M) communication systems, Internet of Things (IoT), and other next-generation communication systems, such as communication systems evolved after 5G, such as sixth-generation (6G) communication systems, etc., and can also be non-3GPP communication systems, which are not limited in the present application. The following takes a 5G communication system (such as an NR communication system) as an example for illustration.

[0040] To facilitate the understanding of the embodiments of the present application, Figure 1 the application scenario of the present application is described by taking the Figure 1 shown communication system architecture as an example. As Figure 1 shown, the communication system includes a network device 101 and a terminal device 102. It can be understood that

[0041] only one possible communication system architecture to which the embodiments of the present application can be applied is shown. In other possible scenarios, the communication system architecture may also include other devices.

[0042] The terminal device can be a mobile terminal device, such as a mobile phone (or also called a "cellular" phone, mobile phone), computer, and data card. For example, it can be a portable, pocket-sized, handheld, computer-integrated, or vehicle-mounted mobile device that exchanges voice and / or data with a radio access network. For example, devices such as personal communication service (PCS) phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), tablets (Pads), computers with wireless transceiver functions, etc. The wireless terminal device can also be called a system, subscriber unit, subscriber station, mobile station, mobile station (MS), remote station, access point (AP), remote terminal device, access terminal device, user terminal device, user agent, subscriber station (SS), customer premises equipment (CPE), terminal, user equipment (UE), mobile terminal (MT), etc. The terminal device can also be a wearable device and the next-generation communication system. For example, the terminal device in a 5G communication system or the terminal device in a future evolved communication system, etc.

[0043] The network device 101 can be a device in a wireless network. For example, the network device can be a RAN node (or device) that connects a terminal device to the wireless network, and can also be referred to as a base station. Currently, some RAN devices can include: the new generation base station (generation Node B, gNodeB) in a 5G communication system, transmission reception point (TRP), evolved Node B (eNB), radio network controller (RNC), Node B (NB), base station controller (BSC), base transceiver station (BTS), home base station (e.g., home evolved Node B, or home Node B, HNB), base band unit (BBU), or wireless fidelity (Wi-Fi) AP, etc. Additionally, in a network structure, the network device can include a centralized unit (CU) node, or a distributed unit (DU) node, or a RAN device including a CU node and a DU node.

[0044] In some deployments, the gNB may include a CU and a DU. The gNB may also include a radio unit (RU). The CU implements some functions of the gNB, and the DU implements some functions of the gNB. For example, the CU implements the functions of the radio resource control (RRC) and packet data convergence protocol (PDCP) layers, and the DU implements the functions of the radio link control (RLC), media access control (MAC), and physical (PHY) layers. Since the information of the RRC layer will ultimately become the information of the PHY layer, or is transformed from the information of the PHY layer, thus, in this architecture, high-layer signaling, such as RRC layer signaling or PHCP layer signaling, can also be considered to be sent by the DU, or sent by the DU + RU. It can be understood that the network device can be a CU node, or a DU node, or a device including a CU node and a DU node. In addition, the CU can be classified as a network device in the radio access network (RAN), or the CU can be classified as a network device in the core network (CN), which is not limited here. Among them, in the ORAN system, the CU can also be called an O-CU, the DU can also be called an open (O)-DU, the CUCP can also be called an O-CU-CP, the CU-UP can also be called an O-CUP-UP, and the RU can also be called an O-RU.

[0045] Among them, the network device can send configuration information to the terminal device (for example, carried in a scheduling message and / or an indication message), and the terminal device further performs network configuration according to the configuration information, so that the network configuration between the network device and the terminal device is aligned; or, through the network configuration preset in the network device and the network configuration preset in the terminal device, the network configuration between the network device and the terminal device is aligned. Specifically, "aligned" means that when the network device and the terminal device exchange messages, the two have the same understanding of the carrier frequency for sending and receiving the exchange message, the determination of the exchange message type, the meaning of the field information carried in the exchange message, or other configurations of the exchange message.

[0046] In addition, in other possible cases, the network device can be other devices that provide wireless communication functions for the terminal device. The embodiments of the present application do not limit the specific technologies and specific device forms adopted by the network device. For the convenience of description, the embodiments of the present application do not limit.

[0047] The network device may further include a core network device. For example, the core network device may include an access and mobility management function (AMF), a user plane function (UPF), a session management function (SMF), etc.

[0048] In the embodiments of the present application, the device for implementing the functions of the network device may be the network device or a device capable of supporting the network device to implement the functions, such as a chip system, and this device may be installed in the network device. Below, taking the device for implementing the functions of the network device as the network device as an example, the technical solutions provided in the embodiments of the present application are described.

[0049] Please refer to Figure 2 , which shows a flowchart of a wireless communication method provided in the embodiments of the present application. This method may be applied to Figure 1 the communication system shown in Figure 2 As shown, the method may include the following steps:

[0050] S101, the first terminal obtains a first CSI data distribution feature sequence according to the local CSI data set.

[0051] The first terminal is a terminal device that needs to update the local CSI prediction model. The local CSI data set includes historical CSI information and / or current CSI information obtained by the first terminal.

[0052] The CSI information may include CSI measurement results obtained by performing CSI measurement (or referred to as RS resource channel measurement, channel measurement, or pilot channel measurement, etc.) according to a channel state information reference signal (CSI-RS). Among them, the CSI measurement results may include at least one of channel response information, a channel-quality indicator (CQI), a rank index (RI) of downlink data transmission, and a precoding matrix index (PMI).

[0053] Performing Hash coding on the CSI data set obtained by the first terminal to obtain a CSI data distribution feature binary coding sequence, that is, the first CSI data distribution feature sequence. In this way, only the encoded feature sequence needs to be transmitted to the network device without transmitting the data set entity, thereby reducing the air interface communication overhead. Moreover, it provides a basis for simplifying the discrimination of the similarity of the CSI data set distribution between different UEs.

[0054] S102, The first terminal sends a model update request message to the network device.

[0055] The corresponding network device receives the model update request message. The model update request message is used to request to obtain available CSI prediction model parameters and may carry the first CSI data distribution feature sequence.

[0056] S103, The network device sends a similarity discrimination indication message to the second terminal.

[0057] The corresponding second terminal receives the similarity discrimination indication message. The similarity discrimination indication message is used to instruct the terminal to judge the similarity between the local CSI data distribution feature and the CSI data distribution feature of the first terminal.

[0058] Exemplarily, the similarity discrimination indication message may carry the model merging process ID and the first CSI data distribution feature sequence. The model merging process ID refers to the unique identifier uniformly assigned by the network device for the model merging process, and different model merging processes correspond to different IDs. For example, the unique identifier assigned by the network device for the merging process of the CSI prediction model initiated by UE i is ID1; when UEj initiates a CSI prediction model merging request, the unique identifier assigned by the network device for this model merging process is ID2. The first CSI data distribution feature sequence is used to enable other terminals to perform the similarity between their local CSI data distribution and the CSI data distribution of the first terminal.

[0059] In an exemplary embodiment, the network device may send the similarity discrimination indication message in a broadcast manner. The second terminal includes one or more terminals that can receive the similarity discrimination indication message. For example, in a scenario where only one terminal receives the similarity discrimination indication message, the second terminal includes one terminal; in a scenario where multiple terminals receive the similarity discrimination indication message, the second terminal includes multiple terminals.

[0060] S104, The second terminal discriminates the CSI data distribution similarity according to the CSI data distribution feature sequence corresponding to the local CSI data set and the first CSI data distribution feature sequence.

[0061] After receiving the similarity discrimination indication message, the second terminal parses the message to obtain the model merging process ID and the first CSI data distribution feature sequence. Further, the second terminal performs Hash encoding on the local CSI data set to obtain the corresponding CSI data distribution feature sequence, or the second CSI data distribution feature sequence for short. Finally, the similarity between the second CSI data distribution feature sequence and the first CSI data distribution feature sequence is obtained. For example, the hamming distance between the two feature sequences can be obtained. If the hamming distance is greater than the preset threshold, it is determined that the two CSI data distribution feature sequences are not similar. If the hamming distance is less than or equal to the preset threshold, it is determined that the two CSI data distribution feature sequences are similar.

[0062] S105. The second terminal sends a similarity discrimination response message to the network device.

[0063] The corresponding network device receives the similarity discrimination response message. The similarity discrimination response message is used to respond to the similarity discrimination indication message sent by the network device.

[0064] After determining that its own CSI data distribution feature is similar to that of the first terminal, the second terminal sends a similarity discrimination response message to the network device. The similarity discrimination response message may carry the device ID of the second terminal, the model merging process ID, and the distance between the second CSI data distribution feature sequence and the first CSI data distribution feature sequence. If it is determined that the CSI data distribution features of the second terminal and the first terminal are not similar, no similarity discrimination response message is sent to the network device.

[0065] By calculating the distance of the CSI data distribution features between UEs to discriminate the similarity of the CSI data distribution, the discrimination process is simplified, and at the same time, the accuracy of selecting the model parameters for model merging is improved.

[0066] S106. The network device determines a candidate terminal list for model merging according to the similarity discrimination response messages sent by each second terminal.

[0067] All second terminals that receive the similarity discrimination indication message sent by the network device will judge the similarity between their own CSI data distribution features and the CSI data distribution features of the first terminal. If it is determined that the CSI data distribution features of the two terminals are similar, a similarity discrimination response message carrying the distance between the two CSI data distribution feature sequences is sent to the network device. In other words, the network device may receive multiple similarity discrimination response messages sent by the second terminals.

[0068] In an exemplary embodiment, the network device may wait to receive a similarity discrimination response message after sending a similarity discrimination indication message. Until the waiting time reaches the time threshold, it detects whether a similarity discrimination response message is received. In this embodiment, the network device receives similarity discrimination response messages sent by at least one second terminal.

[0069] Further, the network device selects a preset number of UEs in ascending order according to the distances between the CSI data distribution characteristics carried by each similarity discrimination response message, and determines them as a candidate terminal list participating in model merging.

[0070] In an exemplary embodiment, the preset number may be , where N is the number of actually responding UEs, represents a preset maximum number threshold of responding UEs. If N > , then select responding UEs to form a candidate terminal list; if N ≤ , then select N responding UEs to form a candidate terminal list.

[0071] In another exemplary embodiment, a ratio between the number of terminals included in the candidate terminal list and the actually received similarity discrimination response messages may be set, such as rounding up 80%. The preset number is determined according to this ratio and the number of actually responding UEs.

[0072] The candidate terminal list includes the terminal IDs of all terminals participating in the CSI prediction model merging. The terminal ID is the unique identifier of the terminal device and may be the International Mobile Equipment Identity (IMEI), the International Mobile Subscriber Identity (IMSI), or the Temporary Mobile Subscriber Identity (TMSI), etc.

[0073] S107. The network device sends a model upload indication message.

[0074] Correspondingly, the second terminal receives the model upload indication message, and this message is used to instruct the terminal to report the model parameters of the specified model to the network device.

[0075] Exemplarily, the network device may send the model upload indication message in a broadcast manner. Among them, the model upload indication message may carry the model merging process ID and the candidate terminal list.

[0076] S108. The second terminal sends a model upload response message to the network device.

[0077] The corresponding network device receives a model upload response message, which is a response message to the model upload indication message.

[0078] Each second terminal that receives the model upload indication message parses the model upload indication message to obtain a model merging process ID and a candidate terminal list. Then, the second terminal checks whether the model merging process ID carried in the model upload indication message is the same as the model merging process ID carried in the similarity discrimination response message sent by itself. If they are the same, it continues to check whether the candidate terminal list contains its own device ID (i.e., the ID of the second terminal). If it contains, it determines that it needs to respond to the model upload indication message and further sends a model upload response message containing the model merging process ID and the local CSI prediction model parameters to the network device; if it does not contain, it determines that it does not need to respond to the model upload indication message, that is, there is no need to send a model upload response message to the network device.

[0079] S109. The network device merges the model parameter sets sent by each second terminal to obtain a fused model parameter set.

[0080] Exemplarily, in a possible scenario, the network device receives model parameter sets sent by multiple second terminals. In this scenario, the network device can use a low-complexity training-free model merging algorithm to merge the model parameters of the CSI prediction models sent by each second terminal to obtain a fused model parameter set.

[0081] In another possible scenario, the network device only receives a model parameter set sent by one second terminal. In this scenario, the received model parameter set can be directly used as the fused model parameter set.

[0082] S110. The network device sends a model update indication message to the first terminal.

[0083] The corresponding first terminal receives the model update indication message. The model update indication message is used to instruct the terminal to update the model parameters of the local specified model. Among them, the model update indication message may carry a model merging process ID and a fused model parameter set.

[0084] S111. The first terminal updates the parameters of the local CSI prediction model according to the fused model parameter set in the model update indication message.

[0085] After receiving the model update indication message, the first terminal parses the indication message to obtain a model merging process ID and a fused model parameter set, and further determines whether the model merging process ID in the indication message is the same as the model merging process ID in the sent model merging request. If they are the same, it updates the model parameters of the local CSI prediction model based on the fused model parameter set carried in the indication message to obtain an updated CSI prediction model.

[0086] S112. The first terminal uses the updated model to perform CSI prediction and sends the CSI prediction result to the network device.

[0087] The first terminal uses the updated CSI prediction model to perform CSI prediction to obtain a CSI prediction result and sends it to the network device, so that the network device performs resource scheduling according to the CSI prediction result sent by the terminal.

[0088] The wireless communication method provided in this embodiment constructs combined model parameters by using the model parameters of the CSI prediction models locally trained by multiple UEs, and sends the combined model parameters to the UEs that need to update the local CSI prediction model. In this way, the UEs can directly use the received combined model parameters to update the local CSI prediction models without local training. Moreover, the combined model obtained from the local CSI prediction models of multiple UEs improves the generalization of the CSI prediction model and the universality of the model among different UEs. Specifically, in the process of constructing the combined model by using the local models of multiple UEs, the model parameters of the local models of multiple UEs with similar CSI data distributions are combined to obtain the combined model parameters, so as to ensure the rationality of the models participating in the combination. Further, when discriminating the similarity of CSI data distributions among UEs, the CSI data set of the UE is encoded to obtain a CSI data distribution feature sequence, the distance between the CSI data distribution feature sequences corresponding to two UEs is calculated, and further the similarity of the CSI data distributions of the two UEs is discriminated according to the distance. Encoding the CSI data of the UE can greatly reduce the data volume, thereby reducing the communication overhead. Moreover, using the distance between the encoded feature sequences to discriminate the similarity of data distributions simplifies the judgment process and provides a basis for reasonably selecting models for model combination.

[0089] Please refer to Figure 3 , which shows the flowchart of another wireless communication method provided in the embodiment of the present application. This method is applicable to Figure 1 the communication system shown in Figure 3 as shown in

[0090] S201. The first terminal obtains a first CSI data distribution feature sequence according to the local CSI data set.

[0091] S202. The first terminal sends a model update request message to the network device.

[0092] S203. The network device sends a similarity discrimination indication message to the second terminal.

[0093] For the implementation processes of S201 to S203 in this embodiment, please refer to Figure 2The relevant content of S101 to S103 will not be elaborated here.

[0094] S204, the network device detects that the waiting time reaches the time threshold Twait and does not receive the similarity discrimination response message, and sends a model training instruction message to the first terminal.

[0095] After sending the similarity discrimination instruction message, the network device waits to receive the similarity discrimination response message. If the waiting time reaches the time threshold and no similarity discrimination response message sent by any terminal is received, it is considered that there are no model parameters that can be used for model merging.

[0096] In an exemplary embodiment, the network device can maintain a timer. When the network device sends the similarity discrimination instruction message, the timer is started for timing. When the timing reaches the time threshold, the timer stops working. If the network device does not receive any similarity discrimination response message until the timer stops timing, it is determined that there are no model parameters that can be merged currently, and the first terminal is notified to update the model through model training.

[0097] S205, the first terminal responds to the model training instruction message and performs local CSI prediction model training.

[0098] After receiving the model training instruction message, the first terminal performs model training on the local CSI prediction model to update the model.

[0099] S206, the first terminal performs CSI prediction according to the updated model and sends the CSI prediction result to the network device.

[0100] The wireless communication method provided in this embodiment, after detecting that there is no model that can participate in model merging, notifies the terminal to update the model through local training.

[0101] Please refer to Figure 4 , which shows a flowchart of another wireless communication method provided in an embodiment of the present application. This embodiment is further elaborated in detail on the basis of the embodiment shown in Figure 2 As shown in Figure 4 , the method may include the following steps:

[0102] S301, the first terminal generates a first CSI data distribution feature Hash coding sequence according to the local CSI data set.

[0103] The first terminal is a UE that needs to update the CSI prediction model. Exemplarily, when the first terminal detects that an event triggering model update occurs, it performs the actions described in S301. Among them, the event triggering model update may include an event where the performance monitoring metric of the CSI prediction model meets a preset condition. For example, when the squared generalized cosine similarity (SGCS cosine similarity) monitored by the CSI prediction model is lower than the corresponding preset threshold, or when the normalized mean square error (NMSE) monitored by the model is higher than the corresponding preset threshold, it is determined that the event triggering model update occurs.

[0104] In an exemplary embodiment, the process of obtaining the CSI data distribution feature Hash sequence is as follows:

[0105] A1. Randomly generate K complex vectors using a standard normal distribution or a Cauchy distribution. The complex vector is a K×N r ×N t -dimensional complex vector, which can be denoted as , where K represents the number of carrier frequencies, N r represents the number of antenna ports of the UE, and N t represents the number of antenna ports of the network device. These random complex vectors are shared among all UEs.

[0106] A2. The first terminal (UE i ) sequentially extracts the most recent L channel state tensors from the local CSI data set , and converts each channel state tensor into a K×N r ×N t -dimensional channel state feature complex vector. The conversion formula is as follows:

[0107] (1)

[0108] In formula 1, represents the complex vector obtained by converting the channel state tensor i of the UE at time t, and the function vec() is used to rearrange the matrix into a vector; in r ×N t represents the set of K×N r ×N t -dimensional channel state tensors; S is the total number of channel tensors included in the data set (i.e., the total number of times of collecting CSI data),

[0109]

[0110] represents the collection time range of the most recent L collection times.

[0109] A3, perform an inner product operation on each complex vector of channel state features obtained by converting A2 with the random complex vectors obtained by A1, and convert them into binary bits. Among them, the inner product calculation result can be converted into binary bits according to the following formula:

[0110] (2)

[0111] Among them, in formula 2 represents and the complex vector of features corresponding to the binary bits obtained after performing the inner product calculation, represents and the real part of the inner product of is greater than 0 (indicating a positive correlation relationship between the two vectors), that is, if and the real part of the inner product of is greater than 0, then = 1, otherwise = 0.

[0112] A4, arrange the L binary bits obtained by A3 in the order shown by the following formula to obtain the CSI data distribution feature Hash coding sequence of the UE i .

[0113] (3)

[0114] Among them, replaces the local CSI data set of the UE i and is used as the basis for comparing the similarity of CSI data distribution between the UE and other UEs. In this way, the transmission of the entire CSI data set i can be avoided, so the communication overhead is reduced. Moreover, the comparison algorithm for the similarity of CSI data distribution between UEs can be simplified. In addition, the length of

[0115] can be flexibly adjusted by changing K and L to achieve the accuracy required to describe the similarity of channel data distribution between UEs. S302, the first terminal sends a model update request message to the network device.

[0116] The model update request message carries the CSI data distribution feature Hash coding sequence of the first terminal (UE

[0117] i ).

[0118] S303, the network device broadcasts a similarity discrimination indication message.

[0119] ​The network device receives the model update request message sent by the first terminal and parses to obtain the model merge process ID and the CSI data distribution feature Hash coding sequence of the UE carried in the request message, then generates a similarity discrimination indication message for the CSI data distribution and broadcasts the message, where the indication message carries the model merge process ID and the CSI data distribution feature Hash coding sequence of the UE. i After that, it generates a similarity discrimination indication message for the CSI data distribution and broadcasts the message, where the indication message carries the model merge process ID and the CSI data distribution feature Hash coding sequence of the UE. i The CSI data distribution feature Hash coding sequence of the UE.

[0120] S304. The second terminal that receives the similarity discrimination indication message determines whether the retention time of the current CSI prediction model is greater than or equal to the preset retention time threshold. If so, execute S305; otherwise, end the current process.

[0121] Setting the preset retention time threshold can ensure that after the model is put into use, it has gone through a sufficient number of CSI measurement cycles (for example, one cycle is 20 ms), so as to obtain a sufficient number of model monitoring data, and ensure that the model is stable and available before using the model to participate in the model update of other UEs, that is, it ensures the stability of the models participating in the merge. For example, the preset retention time threshold can be set to more than 100 ms, and this application does not limit this.

[0122] After the second terminal receives the similarity discrimination indication message and parses the message to determine that the message is used to indicate the terminal to perform CSI data distribution feature similarity discrimination, it obtains the retention time of the local CSI prediction model and compares the size relationship between the retention time and the preset retention time threshold. The retention time of the exemplary CSI prediction model refers to the time difference between the current moment and the latest update time of the CSI prediction model.

[0123] In addition, the network device sends the similarity discrimination indication message in a broadcast manner, so the number of second terminals that receive the message may be multiple, and the actions of this step are applicable to all terminals that receive the message.

[0124] S305. The second terminal determines whether the Hamming distance between its own CSI data distribution feature Hash coding sequence and the CSI data distribution feature Hash coding sequence of the first terminal is less than the preset distance threshold. If so, execute S306; otherwise, end the current process.

[0125] After the second terminal determines that the retention time of the local CSI prediction model is greater than or equal to the preset retention time threshold, it calculates the Hamming distance between its own CSI data distribution feature Hash coding sequence and the CSI data distribution feature Hash coding sequence of the first terminal, and determines whether the Hamming distance is less than the preset distance threshold. If so, it indicates that the CSI data distribution features of the two terminals are similar, and continue to execute S306; if not, it indicates that the CSI data distribution features of the two terminals are not similar.

[0126] In an exemplary embodiment, the Hamming distance can be calculated according to the following formulas (4) to (5):

[0127] (4)

[0128] Formula 4 performs modulo-two addition (i.e., logical exclusive OR operation) on each bit in and to obtain a new binary complex vector ; represents the CSI data distribution feature Hash coding sequence corresponding to UE i , represents the CSI data distribution feature Hash coding sequence corresponding to UE j .

[0129] (5)

[0130] Formula 5 means taking the number of "1"s in as the Hamming distance i between the CSI data distribution feature Hash coding sequences of UE j and UE , represents the b-th bit in. The meanings of K and L can be referred to the relevant explanations of Formula 1.

[0131] Furthermore, the data distribution similarity between UE i and UE j can be determined according to the following formula:

[0132] (6)

[0133] wherein, represents a preset distance threshold, represents a Boolean value of the CSI data distribution similarity between UE i and UE j . Among them, if the Hamming distance between the CSI data distribution feature Hash coding sequences of UE i and UE j is less than the distance threshold, the corresponding Boolean value is true, and the corresponding value is "1"; if the Hamming distance is greater than or equal to the distance threshold, the corresponding Boolean value is false, and the corresponding value is "0".

[0134] S306. The second terminal sends a similarity discrimination response message to the network device.

[0135] After the second terminal determines that the CSI data distribution characteristics of itself and the first terminal are similar, it sends a similarity discrimination response message to the network device. The response message carries the device ID of the second terminal, the model merging process ID, and the above Hamming distance.

[0136] S307. The network device determines whether the waiting time is greater than or equal to a preset waiting time threshold; if so, it executes S308, otherwise it waits for a certain duration and then continues to execute S307.

[0137] In an exemplary embodiment, the network device can maintain a timer. When the network device sends a similarity discrimination indication message, it starts the timer for timing. When the timing reaches the time threshold, the timer stops timing. The network device detects that the waiting time has not reached the waiting time threshold, waits for a certain duration, and then continues to detect whether the waiting time reaches the waiting time threshold; if it detects that the waiting time reaches the waiting time threshold, it continues to execute S308.

[0138] For example, the waiting time threshold can be any value within the range greater than 100 ms and less than 500 ms. This application does not impose special restrictions on the specific value of the waiting time threshold.

[0139] S308. The network device determines whether it has received a similarity discrimination response message; if it has received the similarity discrimination response message, it executes S309, otherwise it executes S316.

[0140] After the network device detects that the timer stops timing, it detects whether it has received a similarity discrimination response message sent by the terminal. If it has received the similarity discrimination response message, it determines that there is a terminal whose CSI data distribution characteristics are similar to those of the first terminal, and continues to execute S309; if it has not received any similarity discrimination response message sent by the terminal, it determines that there is no terminal whose CSI data distribution characteristics are similar to those of the first terminal, and continues to execute S316.

[0141] S309. The network device selects a preset number of second terminals in ascending order of the Hamming distance to obtain a candidate terminal list.

[0142] After the network device receives similarity discrimination response messages sent by multiple terminals, it obtains the model merging process ID and the Hamming distance carried in the response messages. If the model merging process ID carried in the response message is the same as the model merging process ID in the similarity discrimination indication message, it sorts the Hamming distances in ascending order and selects a preset number of second terminals to form a candidate terminal list, which includes the device IDs of the selected second terminals.

[0143] In an exemplary embodiment, the preset number can be , where N is the number of UEs actually responding, represents a preset maximum number of responding UEs threshold. If N > , then select responding UEs to form a candidate terminal list; if N ≤ , then select N responding UEs to form a candidate terminal list.

[0144] S310, the network device broadcasts a model upload indication message.

[0145] After determining the candidate terminal list, the network device broadcasts a model upload indication message, which is used to indicate the model parameters of the model that the terminal needs to upload. Among them, this message can carry the model merging process ID and the candidate terminal list.

[0146] S311, the second terminal determines whether it needs to respond to the model upload indication message; if so, execute S311, otherwise end the current process.

[0147] After receiving the model upload indication message broadcast by the network device, the second terminal parses the indication message to obtain the model merging process ID and the candidate terminal list. Then, the second terminal determines whether the model merging process ID in the model upload indication message is the same as the model merging process ID carried in the similarity discrimination response message sent by itself. If they are different, the second terminal does not need to respond to the model upload indication message; if they are the same, the second terminal continues to determine whether the candidate terminal list contains its own device ID. If it contains, it is determined that it needs to respond to the model upload indication message, otherwise it does not need to respond to the model upload indication message.

[0148] S312, the second terminal sends the model parameters of the local CSI prediction model to the network device.

[0149] After determining that it needs to respond to the received model upload indication message, the second terminal sends the model merging process ID and the CSI prediction model parameters .

[0150] S313, the network device performs a low-complexity training-free model merging algorithm on the received CSI prediction model parameters to obtain the merged model parameters.

[0151] In an exemplary embodiment, the main idea of the low-complexity training-free model merging algorithm includes:

[0152] B1, using the arithmetic mean method to obtain the average model parameters of each responding UE . Among them, the responding UE refers to the terminal that responds to the model upload indication message sent by the network device.

[0153] Exemplarily, the formula for obtaining the average model parameters is as follows:

[0154] (7)

[0155] Wherein, represents the number of UEs included in the candidate terminal list, and the meaning of

[0156] B2, calculate the average variance value between the model parameters of each responding UE and the average model parameters by using the following formula ( );

[0157] (8)

[0158] Wherein, represents the average variance value between the model parameters of the i-th UE and the average model parameters; represents the number of parameters in the model parameter set. It is assumed that the models of each UE are isomorphic, that is, the number of model parameters of each UE is the same; represents the model parameter set in the -th parameter, represents the -th parameter in the average model parameter set.

[0159] B3, normalize the negative exponent of the average variance value of the model parameters of each responding UE by using the following formula to obtain the weight coefficient ( ):

[0160] (9)

[0161] Wherein, represents the negative exponent of the average variance value of the model parameters of the i-th UE, represents the number of UEs in the candidate terminal list.

[0162] B4, weight the model parameters by using the weight coefficient corresponding to each candidate terminal by using the following formula to obtain the fused model parameters :

[0163] (10)

[0164] Wherein, represents the model parameter set of the i-th UE, represents the weight coefficient of the model parameters of the i-th UE, represents the number of UEs in the candidate terminal list.

[0165] S314, the network device sends a model update indication message to the first terminal.

[0166] The network device obtains the fusion model parameters of multiple candidate UEs After that, a model update indication message is sent to the terminal that initiated the model merging request (i.e., the first terminal in this embodiment), which is used to instruct the UE to update the UE's local model parameters. The message may carry the model merging process ID and the fusion model parameters. .

[0167] S315: The first terminal updates model parameters of a local CSI prediction model based on the fusion model parameters.

[0168] After receiving the model update indication message, the first terminal parses the message to obtain the model merging process ID and fusion model parameters , update the model parameters in the local CSI prediction model to the fusion model parameters , and obtain the updated model.

[0169] S316, the network device sends a model training instruction message to the first terminal.

[0170] If the network device determines in S308 that no similarity determination response message is received from any terminal, a model training indication message is sent to the terminal that initiated the model update request. The message is used to instruct the terminal to perform model training locally to update model parameters. The message may carry a model merging process ID.

[0171] S317: The first terminal performs local CSI prediction model training and updates model parameters.

[0172] After receiving the model training indication message, the first terminal performs model training on the local CSI prediction model to update the model parameters.

[0173] S318: The first terminal uses the updated CSI prediction model to perform CSI prediction and sends the CSI prediction result to the network device.

[0174] The first terminal updates the CSI prediction model using the fusion model parameters, or performs model training and updates the model locally. The updated model is used to perform CSI prediction, and the CSI prediction result is sent to the network device.

[0175] In an exemplary embodiment, for the model update request message, similarity determination response message, and model upload response message sent by the above-mentioned terminal, the above-mentioned message can be carried by a field in the uplink signaling of the RRC layer, for example, a newly defined field (or reused existing field) in the CSI reporting configuration signaling RRC CSI-ReportConfig of the RRC layer carries the above-mentioned message. The network device identifies the specific message content by parsing the corresponding field in the RRC CSI-ReportConfig signaling.

[0176] In an exemplary embodiment, for the similarity discrimination indication message, model upload indication message, model update indication message, and model training indication message sent by the above-mentioned network device, the above messages can be carried by newly defined fields (or reused existing fields) in the RRC layer downlink signaling. For example, the newly defined fields in the CSI resource configuration signaling RRC CSI-ResourceConfig of the RRC layer carry the above messages. The terminal identifies the specific message content by parsing the corresponding fields in the RRC CSI-ResourceConfig.

[0177] Corresponding to the above-mentioned wireless communication method embodiments, the present application also provides communication device embodiments.

[0178] Figure 5 FIG. is a schematic structural diagram of a communication device provided by an embodiment of the present application. The communication device can be a terminal device, or a device (such as a chip) in the terminal device, or a device that can be used in combination with the terminal device; or the communication device can be a network device, or a device (such as a chip) in the network device, or a device that can be used in combination with the network device.

[0179] As Figure 5 shown, the communication device may include a transceiver module 110 and a processing module 120. Specifically, the processing module 120 is used to process data, which can be the data received by the transceiver module 110, and the processed data can also be sent by the transceiver module 110.

[0180] The processing module 120 is used to execute the processing procedures of the terminal device or the network device in the above-mentioned communication method embodiments.

[0181] In one implementation, the communication device can correspond to the UE in the above method embodiments and can be used to execute each step and / or process performed by the UE in the above method embodiments.

[0182] In another implementation, the communication device can correspond to the network device in the above method embodiments and can be used to execute each step and / or process performed by the network device in the above method embodiments.

[0183] For other possible implementations of the communication device, reference can be made to the relevant descriptions of the functions of the foregoing terminal device or network device, which will not be elaborated here.

[0184] Figure 6 FIG. is another schematic block diagram of the communication device provided by an embodiment of the present application.

[0185] The communication device may be a UE, a network device, a chip, a chip system, or a processor that implements the above method. The communication device can be used to implement the method described in the above method embodiments. For specific details, please refer to the descriptions in the above method embodiments.

[0186] As Figure 6 shown, the communication device may include one or more processors 201. The processor 201 may also be referred to as a processing unit or a processing module and can implement certain control functions. The processor 201 may be a general-purpose processor or a dedicated processor, etc. For example, it may be a baseband processor or a central processing unit. The baseband processor can be used to process communication protocols and communication data, and the central processing unit can be used to control the communication device (such as a base station, a baseband chip, a user, a user chip), execute software programs, and process the data of software programs.

[0187] In an alternative design, the processor 201 may also store instructions and / or data, and the instructions and / or data can be run by the processor 201, so that the communication device executes the method described in the above method embodiments.

[0188] In another alternative design, the communication device may include a communication interface 202 for implementing receiving and sending functions. For example, the communication interface 202 may be a transceiver circuit, an interface, an interface circuit, or a transceiver, etc. The transceiver circuit, interface, interface circuit, or transceiver for implementing receiving and sending functions may be separate or integrated together. The above transceiver circuit, interface, interface circuit, or transceiver can be used for reading and writing code / data, or the above transceiver circuit, interface, interface circuit, or transceiver can be used for signal transmission or transfer.

[0189] Optionally, the communication device may include one or more memories 203, on which instructions may be stored. These instructions can be run on the processor 201, so that the communication device executes the method described in the above method embodiments. Optionally, data may also be stored in the memory 203. Optionally, instructions and / or data may also be stored in the processor 201. The processor 201 and the memory 203 may be provided separately or integrated together.

[0190] It should be understood that in a possible design, each step in the method embodiments provided in this application can be completed by the integrated logic circuit of the hardware in the processor or the instructions in software form. The steps of the method disclosed in combination with the embodiments of this application can be directly embodied as being executed and completed by the hardware processor, or executed and completed by the combination of the hardware and software modules in the processor. The software module can be located in mature storage media in the art such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers, etc. This storage media is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method. To avoid repetition, it will not be described in detail here.

[0191] In one implementation, the communication device can correspond to the UE in the above method embodiments and can be used to execute each step and / or process executed by the UE in the above method embodiments. The processor 201 can be used to execute the instructions stored in the memory 203, and when the processor 201 executes the instructions stored in the memory, the processor 201 is used to execute each step and / or process of the above method embodiment corresponding to the terminal.

[0192] In another implementation, the communication device can correspond to the network device in the above method embodiments and can be used to execute each step and / or process executed by the network device in the above method embodiments. The processor 201 can be used to execute the instructions stored in the memory 203, and when the processor 201 executes the instructions stored in the memory, the processor 201 is used to execute each step and / or process of the above method embodiment corresponding to the network device.

[0193] It can be understood that the above processor can be one or more chips. For example, the processor can be a field programmable gate array (FPGA), can be an application specific integrated circuit (ASIC), can also be a system on chip (SoC), can also be a central processor unit (CPU), can also be a network processor (NP), can also be a digital signal processing circuit (DSP), can also be a micro controller unit (MCU), can also be a programmable logic device (PLD) or other integrated chips.

[0194] It can be understood that the memory in the embodiments of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and directrambus RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0195] The embodiments of the present application further provide a computer-readable storage medium, in which instructions are stored. When the instructions are run on one or more computing devices, the one or more computing devices are caused to execute the wireless communication method described in the foregoing embodiments.

[0196] The computer-readable storage medium can be a non-transitory computer-readable storage medium. For example, the non-transitory computer-readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0197] An embodiment of this application also provides a computer program product. When the computer program product is executed by one or more computing devices, the one or more computing devices execute any of the foregoing wireless communication methods. The computer program product may be a software installation package. In the case where any of the foregoing wireless communication methods needs to be used, the computer program product can be downloaded and executed on a computer.

[0198] An embodiment of this application also provides a processor, including: an input circuit, an output circuit, and a processing circuit. Among them: The processing circuit is used to receive a signal through the input circuit and transmit the signal through the output circuit, so that the processor executes the wireless communication method described in the foregoing embodiment.

[0199] In a specific implementation process, the processor may be one or more chips, the input circuit may be input pins, the output circuit may be output pins, and the processing circuit may be transistors, gate circuits, flip-flops, and various logic circuits, etc. The input signal received by the input circuit may be received and input by, for example, but not limited to, a receiver. The signal output by the output circuit may be output to, for example, but not limited to, a transmitter and transmitted by the transmitter. Moreover, the input circuit and the output circuit may be the same circuit, and this circuit is used as the input circuit and the output circuit at different times. The embodiments of this application do not limit the specific implementation manners of the processor and various circuits.

[0200] An embodiment of this application also provides a chip system, which includes one or more processors, and is used to call and run instructions stored in a memory from the memory, so that the wireless communication method described in the foregoing embodiment is executed. The chip system may be composed of chips, or may include chips and other discrete devices. Among them, the chip system may include an input circuit or interface for sending information or data, and an output circuit or interface for receiving information or data.

[0201] In the embodiments of this application, the terms and English abbreviations are all exemplary examples given for convenience of description, and should not constitute any limitation to this application. This application does not exclude the possibility of defining other terms that can achieve the same or similar functions in existing or future protocols.

[0202] In the foregoing embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this application are generated in whole or in part.

[0203] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be in an electrical, mechanical, or other form.

[0204] It should be understood that in various embodiments of the present application, the size of the serial numbers of each process does not mean the sequence of execution. The execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0205] In summary, the above description is only a preferred embodiment of the technical solution of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A wireless communication method, characterized in that, Applied to a first terminal, the method includes: Sending a model update request message for requesting to update the CSI prediction model parameters of the first terminal. The model update request message includes a first CSI data distribution feature sequence for characterizing the distribution features of the CSI data collected by the first terminal; Receiving a model update indication message sent by a network device for indicating the first terminal to update the model parameters of the CSI prediction model. The model update indication message includes the fusion model parameters corresponding to the CSI prediction model, and the fusion model parameters are obtained by the network device according to the model parameters of the CSI prediction models in at least one second terminal; Responding to the model update indication message and updating the model parameters of the CSI prediction model based on the fusion model parameters carried in the model update indication message.

2. The method according to claim 1, characterized in that The process of obtaining the first CSI data distribution feature sequence includes: Performing Hash encoding on the obtained CSI data to obtain the first CSI data distribution feature sequence.

3. The method according to claim 2, wherein The performing Hash encoding on the obtained CSI data to obtain the first CSI data distribution feature sequence includes: Generating K random complex vectors, where K represents the number of carrier frequencies; Extracting L recently obtained CSI tensors from the CSI data set obtained by the first terminal and respectively converting them into CSI feature complex vectors, and the random complex vectors have the same dimension as the CSI feature complex vectors; Performing inner product operations on each CSI feature complex vector and the K random complex vectors respectively to obtain K×L inner product operation results, and converting each inner product operation result into a binary bit; Arranging the K×L binary bits in order to obtain the first CSI data distribution feature sequence.

4. The method according to claim 3, characterized in that The performing inner product operations on each CSI feature complex vector and the K random complex vectors respectively to obtain K×L inner product operation results, and converting each inner product operation result into a binary bit includes: Converting each inner product operation result into a binary bit according to the following formula: Among them, represents the k-th complex vector in the random complex vector, represents the CSI eigen complex vector obtained by converting the CSI tensor of the UE i at time t; represents and the binary bit obtained after performing the inner product calculation, represents and the real part of the inner product operation result of is greater than 0; represents the most recent L acquisition times, and S represents the total number of CSI tensors included in the CSI dataset.

5. The method according to claim 3, wherein The arranging the K×L binary bits in order to obtain the first CSI data distribution feature sequence includes: Performing arrangement according to the following formula: Among them, represents the CSI data distribution feature sequence corresponding to the CSI data set of the UE i .

6. A wireless communication method, characterized in that, Applied to a network device, the method includes: Receiving a model update request message sent by a first terminal for requesting to update the CSI prediction model parameters of the first terminal. The model update request message includes a first CSI data distribution feature sequence for characterizing the distribution features of the CSI data collected by the first terminal; Sending a similarity discrimination indication message to a second terminal for indicating the second terminal to determine the similarity between its own CSI data distribution features and the first CSI data distribution feature sequence of the first terminal. The similarity discrimination indication message includes a model merging process ID and the first CSI data distribution feature sequence; Receive the similarity discrimination response message sent by the second terminal, where the similarity discrimination response message includes the device ID of the second terminal, the model merging process ID, and the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal; Determine a candidate terminal list for participating in model merging according to the similarities in each of the similarity discrimination response messages, where the candidate terminal list includes the device IDs corresponding to the candidate terminals; Send a model upload indication message, which is used to instruct the second terminal to report the CSI prediction model parameters, and the model upload indication message includes the model merging process ID and the candidate terminal list; Receive a model upload response message, where the model upload response message includes the model parameters of the CSI prediction model in the second terminal; Perform model parameter fusion on the model parameters sent by each second terminal to obtain fused model parameters; Send a model update indication message to the first terminal, where the model update indication message is used to instruct the first terminal to update the model parameters of the CSI prediction model, and the model update indication message includes the model merging process ID and the fused model parameters.

7. The method according to claim 6, wherein The method further includes: Within a preset waiting time threshold after sending the similarity discrimination indication message, if the similarity discrimination response message is not received, send a model training indication message to the first terminal, where the model training indication message is used to instruct the first terminal to perform local training and update on the CSI prediction model, and the model training indication message includes the model merging process ID.

8. The method according to claim 6 or 7, characterized in that, The determining a candidate terminal list for participating in model merging according to the similarities in each of the similarity discrimination response messages includes: Select a preset number of second terminals in descending order of similarity to obtain the candidate terminal list.

9. The method according to claim 8, characterized in that The selecting a preset number of second terminals in descending order of similarity to obtain the candidate terminal list includes: Select a preset number of second terminals in ascending order of the distance between the CSI data distribution feature sequence of each second terminal and the first CSI data distribution feature sequence to obtain the candidate terminal list.

10. The method according to claim 9, characterized in that, The selecting a preset number of second terminals in ascending order of the distance between the CSI data distribution feature sequence of each second terminal and the first CSI data distribution feature sequence to obtain the candidate terminal list includes: Determine the first number of second terminals that send the similarity discrimination response message; If the first number is less than a preset number threshold, determine that all second terminals that send the similarity discrimination response message form the candidate terminal list; If the first number is greater than the preset number threshold, select the preset number threshold of second terminals from the first number in ascending order of the distance between the CSI data distribution feature sequence of each second terminal and the first CSI data distribution feature sequence to form the candidate terminal list.

11. The method according to claim 6, wherein The performing model parameter fusion on the model parameters sent by each second terminal to obtain fused model parameters includes: Obtain the average model parameters corresponding to the model parameters sent by each second terminal using the arithmetic mean method, and obtain the average variance value between each received model parameter and the average model parameters; Normalize the negative exponent of each average variance value to obtain the weight coefficient corresponding to each received model parameter; Perform weighted summation on each received model parameter to obtain the fused model parameters.

12. A wireless communication method, characterized in that, Applied to the second terminal, the method includes: Receive a similarity discrimination indication message sent by the network device. The similarity discrimination indication message is used to instruct the second terminal to determine the similarity between its own CSI data distribution feature sequence and the first CSI data distribution feature sequence of the first terminal. The similarity discrimination indication message includes a model merging process ID and the first CSI data distribution feature sequence, and the first CSI data distribution feature sequence is used to characterize the distribution feature of the CSI data collected by the first terminal; In response to the similarity discrimination indication message, obtain the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal; When the similarity meets the preset conditions, send a similarity discrimination response message to the network device. The similarity discrimination response message includes the device ID of the second terminal, the model merging process ID, and the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal; Receive a model upload indication message sent by the network device. The model upload indication message is used to instruct the second terminal to report CSI prediction model parameters. The model upload indication message includes the model merging process ID and a candidate terminal list, and the candidate terminal list is obtained by the network device according to the similarity between the CSI data distribution feature sequences of each second terminal and the first terminal; After determining that the second terminal is included in the candidate terminal list, send the model parameters of the local CSI prediction model to the network device. The model parameters are used to enable the network device to obtain the fused model parameters corresponding to the CSI prediction model.

13. The method according to claim 12, characterized in that, The step of, in response to the similarity discrimination indication message, obtaining the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal includes: In response to the similarity discrimination indication message, obtain the model merging process ID and the first CSI data distribution feature sequence; Obtain the retention time of the prediction model of the second terminal. If the retention time is greater than a preset retention time threshold, obtain the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal.

14. The method according to claim 12, characterized in that, The step of obtaining the similarity between the CSI data distribution feature sequences of the second terminal and the first terminal includes: Calculate the Hamming distance between the CSI data distribution feature sequence of the second terminal and the first CSI data distribution feature sequence of the first terminal. The Hamming distance is negatively correlated with the similarity.

15. The method according to claim 14, wherein The similarity meeting the preset conditions includes: the Hamming distance is less than a preset distance threshold.

16. A communication device, characterized in that, It includes a processing module and a transceiver module, and the communication device is used to execute the method described in any one of claims 1 to 15.

17. A communication device, characterized in that, It includes: a memory for storing computer instructions; a processor for executing the computer program or computer instructions stored in the memory, so that the communication device executes the method described in any one of claims 1 to 15.

Images