Method and apparatus in node used for wireless communication

Abstract

The invention discloses a method and an apparatus in a node used for wireless communication. The first node receives a first configuration information block, executes M1 operations in M operations, and generates a first information block. The first configuration information block indicates the M operations, M is a positive integer greater than 1, and M1 is a positive integer not greater than M; the first information block depends on the output of the first processing, and the input of the first processing depends on the output of the M1 operations; the first processing depends on the M1 operations. According to the method, the robustness, the reliability and the accuracy of the auxiliary information are improved, and the system performance is further improved.

Description

Technical Field

[0001] The present application relates to a transmission method and device in a wireless communication system, and in particular to a scheme and device related to configuration information in a wireless communication system. Background Art

[0002] In traditional wireless communications, UE (User Equipment) reports various auxiliary information obtained by measuring downlink signals and / or channels, such as channel information, auxiliary information related to beam management (Beam Management), auxiliary information related to positioning, HARQ-ACK (Hybrid Automatic Repeat reQuest Acknowledgement) information, beam / radio link failure auxiliary information, etc. The channel information includes but is not limited to one or more of CRI (CSI-RS Resource Indicator, channel state information reference signal resource indication), RI (Rank Indicator, rank indication), PMI (Precoding Matrix Indicator, precoding indication) or CQI (Channel quality indicator). The UE can use this information to select appropriate transmission parameters or report this information. The network device selects appropriate transmission parameters for the UE according to the UE's report, such as the resident cell, MCS (Modulation and Coding Scheme, modulation and coding scheme), TPMI (Transmitted Precoding Matrix Indicator, TCI (Transmission Configuration Indication, transmission configuration indication) and other parameters. In addition, UE reporting can be used to optimize network parameters, such as better cell coverage, switching base stations based on UE location, etc.

[0003] With the adoption of new technologies, the increase in the number of antennas, the diversification of application scenarios, and the improvement of system performance requirements, traditional measurement and reporting methods will bring a lot of redundant overhead. Therefore, in NR R (release) 18, the research on AI (Artificial Intelligence) / ML (Machine Learning) technology was established to explore its impact on system performance and system design. Compared with traditional processing methods, AI / ML has the characteristics of being based on training and requiring deployment. Summary of the invention

[0004] The applicant found through research that compared with existing technologies, AI / ML functions have more special characteristics, such as higher flexibility in input and output, stronger timeliness, and the need for timely retraining / deployment, etc. Therefore, how to fully utilize the advantages of AI / ML to maximize system performance gains based on these characteristics of AI / ML is a problem that needs to be solved.

[0005] In response to the above problems, the present application discloses a solution. It should be noted that although a large number of embodiments of the present application are developed for AI / ML, the present application is also applicable to other solutions, such as traditional measurement and reporting solutions. In addition, the use of a unified solution for different scenarios (including but not limited to AI / ML-based solutions and traditional solutions) also helps to reduce hardware complexity and cost. In the absence of conflict, the embodiments in the first node of the present application and the features in the embodiments can be applied to the second node, and vice versa. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be arbitrarily combined with each other.

[0006] As an example, the interpretation of the terms in the present application refers to the definitions of the TS38 series of specification protocols of 3GPP.

[0007] As an example, the interpretation of the terms in the present application refers to the definitions of the TS28 series of specification protocols of 3GPP.

[0008] The present application discloses a method in a first node used for wireless communication, characterized by comprising:

[0009] Receive a first configuration information block, where the first configuration information block indicates M operations, where M is a positive integer greater than 1;

[0010] Execute M1 operations among the M operations, where M1 is a positive integer not greater than M;

[0011] Generate a first information block, wherein the first information block depends on an output of a first process, and an input of the first process depends on outputs of the M1 operations;

[0012] The first processing depends on the M1 operations.

[0013] As an embodiment, the problem to be solved by the present application includes how to generate the first information block when the first node is configured with M operations but only executes M1 operations therein; in the above method, the first processing for generating the first information block depends on the M1 operations, which solves this problem.

[0014] As an embodiment, the benefits of the above method include improving the robustness, reliability and accuracy of the first information block, thereby improving system performance.

[0015] As an embodiment, the benefits of the above method include more flexible use of the M operations, which improves the robustness and adaptability of the system.

[0016] As an embodiment, the characteristics of the above method include that the outputs of multiple operations among the M operations can be used together to generate the first information block, further improving the system performance; at the same time, the first node can generate the first information block using only the outputs of some operations among the M operations, improving the flexibility, robustness and adaptability of the system to different scenarios.

[0017] As an embodiment, the first node is a terminal.

[0018] According to one aspect of the present application, it is characterized by comprising:

[0019] M0 operations among the M operations are deployed, where M0 is a positive integer not greater than M1.

[0020] As an embodiment, the problem to be solved by the present application includes how to generate the first information block when only some of the multiple operations configured to be deployed are executed; in the method disclosed in the present application, the first processing for generating the first information block depends on the operations being executed, thereby solving the above problem.

[0021] As an embodiment, the benefits of the above method include adapting to the characteristics of the operation that needs to be deployed, maximizing the advantages of the AI ​​operation that needs to be deployed, and ensuring the reliability and robustness of the first information block.

[0022] According to one aspect of the present application, it is characterized in that the first configuration information block indicates M identifiers, and the M operations are respectively associated with the M identifiers.

[0023] As an embodiment, the benefits of the above method include simplified design while having good flexibility and forward compatibility.

[0024] According to one aspect of the present application, it is characterized in that any operation among the M operations except the M1 operations is invalid.

[0025] As an embodiment, the benefits of the above method include timely avoiding the negative impact of the failure operation on the first information block, ensuring the reliability of the first information block, and improving system performance.

[0026] According to one aspect of the present application, it is characterized by comprising:

[0027] The M1 operations are determined.

[0028] As an embodiment, the above method is characterized in that the first node first determines which operations need to be performed, and then determines the first processing and generates the first information block based on this.

[0029] As an embodiment, the benefits of the above method include that the generation of the first information block is appropriately adjusted according to the operation to be performed, thereby ensuring the robustness and reliability of the first information block.

[0030] As an embodiment, the benefits of the above method include adaptability to different terminals.

[0031] According to one aspect of the present application, it is characterized in that, for any operation among M2 operations among the M operations, the first node determines by itself whether to execute this operation, and the M2 is a positive integer not greater than the M.

[0032] As an embodiment, the benefits of the above method include higher flexibility and adaptability to different terminals.

[0033] As an embodiment, the benefits of the above method include reducing signaling overhead.

[0034] As an embodiment, the benefits of the above method include timely detection of failed operations and improved performance.

[0035] According to one aspect of the present application, it is characterized by comprising:

[0036] receiving a first signaling;

[0037] Among them, for any operation among M3 operations among the M operations, the first signaling indicates whether the first node performs this operation, and the M3 is a positive integer not greater than the M.

[0038] As an embodiment, the benefits of the above method include supporting joint optimization on the network side, thereby further improving system performance.

[0039] According to one aspect of the present application, it is characterized in that the first processing depends on the reliability of the M1 operations.

[0040] As an embodiment, the characteristics of the above method include that the first node can adjust the contribution of the outputs of different operations in the M1 operations to the first information block according to the reliability of the M1 operations.

[0041] As an embodiment, the benefits of the above method include further improving performance.

[0042] According to one aspect of the present application, it is characterized in that the input of the first processing includes M1 inputs, and the M1 inputs are respectively dependent on the outputs of the M1 operations. The first processing uses M1 coefficients to weight the M1 inputs respectively, and the M1 coefficients are dependent on the M1 operations.

[0043] As an embodiment, the benefits of the above method include flexible adjustment of the weights of contributions of different operations to the first information block, thereby optimizing system performance.

[0044] As an embodiment, the benefits of the above method include simple implementation and simplified operation.

[0045] According to one aspect of the present application, it is characterized by comprising:

[0046] The first information block is sent.

[0047] As an embodiment, the benefits of the above method include enhancing the robustness, reliability and accuracy of reported information.

[0048] The present application discloses a terminal, characterized in that:

[0049] The terminal includes: one or more processors and a memory;

[0050] The memory is coupled to the one or more processors, and the memory is used to store computer program codes, wherein the computer program codes include computer instructions, and the one or more processors call the computer instructions to enable the terminal to execute the method in the first node.

[0051] The present application discloses a method used in a second node of wireless communication, characterized by comprising:

[0052] Sending a first configuration information block, where the first configuration information block indicates M operations, where M is a positive integer greater than 1;

[0053] Among them, the target recipient of the first configuration information block determines the first information block, the first information block depends on the output of the first processing, the input of the first processing depends on the output of M1 operations, each of the M1 operations is one of the M operations; the target recipient of the first configuration information block performs the M1 operations, and M1 is a positive integer not greater than M; the first processing depends on the M1 operations.

[0054] As an embodiment, the second node is a base station.

[0055] According to one aspect of the present application, it is characterized by comprising:

[0056] The first information block is received.

[0057] To the M identifiers.

[0058] According to one aspect of the present application, it is characterized in that any operation among the M operations except the M1 operations is invalid.

[0059] According to one aspect of the present application, it is characterized in that the target receiver of the first configuration information block determines the M1 operations.

[0060] According to one aspect of the present application, it is characterized in that, for any operation among M2 operations among the M operations, the target receiver of the first configuration information block determines whether to perform this operation by itself, and M2 is a positive integer not greater than M.

[0061] According to one aspect of the present application, it is characterized by comprising:

[0062] Sending a first signaling;

[0063] Among them, for any operation among M3 operations among the M operations, the first signaling indicates whether the target receiver of the first configuration information block performs this operation, and the M3 is a positive integer not greater than the M.

[0064] According to one aspect of the present application, it is characterized in that the first processing depends on the reliability of the M1 operations.

[0065] According to one aspect of the present application, it is characterized in that the input of the first processing includes M1 inputs, and the M1 inputs are respectively dependent on the outputs of the M1 operations. The first processing uses M1 coefficients to weight the M1 inputs respectively, and the M1 coefficients are dependent on the M1 operations.

[0066] According to one aspect of the present application, it is characterized by comprising:

[0067] Execute a first operation;

[0068] The first information block includes first compressed CSI, and the first compressed CSI is used as an input of the first operation to generate first restored CSI.

[0069] The present application discloses a base station, characterized in that:

[0070] The base station includes: one or more processors and a memory;

[0071] The memory is coupled to the one or more processors, and the memory is used to store computer program codes, where the computer program codes include computer instructions, and the one or more processors call the computer instructions to enable the base station to execute the method in the second node.

[0072] The present application discloses a first node used for wireless communication, characterized in that it includes:

[0073] A first receiver receives a first configuration information block, wherein the first configuration information block indicates M operations, where M is a positive integer greater than 1;

[0074] A first processor executes M1 operations among the M operations, where M1 is a positive integer not greater than M;

[0075] The first processor generates a first information block, the first information block depends on an output of a first process, and an input of the first process depends on outputs of the M1 operations;

[0076] The first processing depends on the M1 operations.

[0077] The present application discloses a second node used for wireless communication, characterized in that it includes:

[0078] The second processor sends a first configuration information block, where the first configuration information block indicates M operations, where M is a positive integer greater than 1;

[0079] Among them, the target recipient of the first configuration information block determines the first information block, the first information block depends on the output of the first processing, the input of the first processing depends on the output of M1 operations, each of the M1 operations is one of the M operations; the target recipient of the first configuration information block performs the M1 operations, and M1 is a positive integer not greater than M; the first processing depends on the M1 operations.

[0080] As an embodiment, compared with the traditional solution, this application has the following advantages:

[0081] Improved robustness, reliability and accuracy of auxiliary information, such as but not limited to channel information;

[0082] Improved system performance;

[0083] Greater flexibility and adaptability;

[0084] Simplifies the implementation complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0086] Figure 1 A flowchart showing a first configuration information block, M1 operations and a first information block according to an embodiment of the present application is shown;

[0087] Figure 2 A schematic diagram of a network architecture according to an embodiment of the present application is shown;

[0088] Figure 3 A schematic diagram showing an embodiment of a wireless protocol architecture of a user plane and a control plane according to an embodiment of the present application;

[0089] Figure 4 A schematic diagram showing a first communication device and a second communication device according to an embodiment of the present application is shown;

[0090] Figure 5 shows transmission between a first node and a second node according to an embodiment of the present application;

[0091] Figure 6 A schematic diagram showing an input of a first process according to an embodiment of the present application;

[0092] Figure 7 A schematic diagram of deploying a given operation according to an embodiment of the present application is shown;

[0093] Figure 8 A schematic diagram showing M identifiers and M operations according to an embodiment of the present application is shown;

[0094] Fig. 9 A schematic diagram showing M1 operations including non-failed operations among M operations according to an embodiment of the present application is shown;

[0095] Fig.10 A schematic diagram showing a first node determining M1 operations according to an embodiment of the present application is shown;

[0096] Fig.11 A schematic diagram showing, for each of M2 operations, a first node independently determining whether to execute the operation according to an embodiment of the present application;

[0097] Fig.12 A schematic diagram showing a first signaling according to an embodiment of the present application is shown;

[0098] Fig.13 A schematic diagram showing a first process according to an embodiment of the present application that relies on M1 operations;

[0099] Fig.14 A schematic diagram showing a first process according to an embodiment of the present application that relies on M1 operations;

[0100] Fig.15 A schematic diagram showing a first process according to an embodiment of the present application that relies on M1 operations;

[0101] Fig.16 A schematic diagram showing a first process according to an embodiment of the present application that relies on M1 operations;

[0102] Fig.17 A schematic diagram showing M1 inputs and M1 coefficients according to an embodiment of the present application is shown;

[0103] Fig.18 A schematic diagram showing M1 coefficients according to an embodiment of the present application is shown;

[0104] Fig.19 A schematic diagram showing M1 coefficients according to an embodiment of the present application is shown;

[0105] Fig. 20 A schematic diagram showing M1 coefficients according to an embodiment of the present application is shown;

[0106] Fig.21 A schematic diagram showing a first compressed CSI and a first restored CSI according to an embodiment of the present application is shown;

[0107] Fig. 22 A schematic diagram of a processing system based on artificial intelligence or machine learning according to an embodiment of the present application is shown;

[0108] Fig.23 A schematic diagram based on artificial intelligence or machine learning according to an embodiment of the present application is shown;

[0109] Fig.24 A schematic diagram showing the deployment of AI functions according to an embodiment of the present application is shown;

[0110] Fig.25 A schematic diagram showing the deployment of AI functions according to an embodiment of the present application is shown;

[0111] Fig.26 A schematic diagram showing the deployment of AI functions according to an embodiment of the present application is shown;

[0112] Fig. 27 A schematic diagram of AI function deployment according to an embodiment of the present application is shown.

[0113] Fig.28A structural block diagram of a processing device used in a first node according to an embodiment of the present application is shown;

[0114] Fig.29 A structural block diagram of a processing device used in a second node according to an embodiment of the present application is shown;

[0115] Fig.30 A schematic diagram showing a first process according to an embodiment of the present application that relies on M1 operations;

[0116] Fig.31 A schematic diagram showing a first process according to an embodiment of the present application that relies on M1 operations. DETAILED DESCRIPTION

[0117] The technical solution of the present application will be further described in detail below in conjunction with the accompanying drawings. It should be noted that the embodiments and features in the embodiments of the present application can be arbitrarily combined with each other without conflict. Based on considerations such as performance, flexibility, complexity, overhead, and compatibility, those skilled in the art are motivated to flexibly combine the embodiments in different drawings without conflict, such as but not limited to the attached drawings. Figure 1 The embodiments and appendix Figure 5 -Attached Fig.31 The embodiments in Figure 5 The embodiments and appendix Figure 6 -Attached Fig.31 The embodiments in , etc.

[0118] Example 1

[0119] Embodiment 1 illustrates a flowchart of a first configuration information block, M1 operations and a first information block according to an embodiment of the present application, as shown in the attached Figure 1 As shown in the attached Figure 1 In the illustrated 100, each box represents a step. In particular, the order of the steps in the boxes does not represent a specific time sequence between the steps.

[0120] In Embodiment 1, the first node receives a first configuration information block in step 101; performs M1 operations among M operations in step 102; and generates a first information block in step 103. The first configuration information block indicates the M operations, M is a positive integer greater than 1, and M1 is a positive integer not greater than M; the first information block depends on the output of the first process, the input of the first process depends on the output of the M1 operations, and the first process depends on the M1 operations.

[0121] As an embodiment, the first configuration information block is carried by higher layer signaling.

[0122] As an embodiment, the first configuration information block is carried by RRC (Radio Resource Control) signaling.

[0123] As an embodiment, the first configuration information block is carried by an RRC IE (Information Element).

[0124] As an embodiment, the first configuration information block is carried by at least one RRC IE.

[0125] As an embodiment, the first configuration information block includes information in one or more fields in at least one RRC IE.

[0126] As an embodiment, the first configuration information block includes information in one or more fields of each RRC IE in multiple RRC IEs.

[0127] As an embodiment, the first configuration information block is an RRC IE.

[0128] As an embodiment, the first configuration information block is carried by CSI-ReportConfig IE.

[0129] As an embodiment, the first configuration information block is carried by ServingCellConfig IE.

[0130] As an embodiment, the first configuration information block is carried by CSI-MeasConfig IE.

[0131] As an embodiment, the first configuration information block is carried by ServingCellConfigCommon IE.

[0132] As an embodiment, the first configuration information block is carried by ServingCellConfigCommonSIB IE.

[0133] As an embodiment, the first configuration information block includes information in CSI-ReportConfig IE.

[0134] As an embodiment, the first configuration information block includes information in ServingCellConfig IE.

[0135] As an embodiment, the first configuration information block includes information in the CSI-MeasConfig IE.

[0136] As an embodiment, the first configuration information block includes information in ServingCellConfigCommon IE.

[0137] As an embodiment, the first configuration information block includes information in ServingCellConfigCommonSIB IE.

[0138] As an embodiment, the first configuration information block indicates the M operations in sequence.

[0139] As an embodiment, the first configuration information block includes the M operations.

[0140] As an embodiment, the first configuration information block indicates that the output of each of the M operations is used as the input of the first processing.

[0141] As an embodiment, the first configuration information block indicates that the output of each of the M operations is used to generate the input of the first process.

[0142] As an embodiment, the first configuration information block indicates that candidates for input of the first process include outputs of each of the M operations.

[0143] As an embodiment, the first configuration information block indicates that the output of each of the M operations can be used to generate the first information block.

[0144] As an embodiment, the first configuration information block indicates that the output of each of the M operations is the input for generating the first information block.

[0145] As an embodiment, the first configuration information block indicates that the output of each of the M operations provides a candidate for an input for the generation of the first information block.

[0146] As an embodiment, the first configuration information block indicates that outputs of all or part of the M operations are used to generate the first information block after the first processing.

[0147] As an embodiment, the first configuration information block instructs the first node to generate the first information block based on the output of all or part of the M operations after the first processing.

[0148] As an embodiment, each of the M operations is training-based.

[0149] As an embodiment, each of the M operations is obtained through training.

[0150] As an embodiment, the problem to be solved by the present application includes how to generate the first information block when only M1 operations among the M operations configured based on training are executed; in the method disclosed in the present application, the first processing for generating the first information block depends on the M1 operations being executed, which solves the above problem.

[0151] As an embodiment, the benefits of the above method include providing multiple input candidates for the generation of the first information block, and adjusting the generation method of the first information block according to the actual input situation, thereby maximizing the robustness, reliability and flexibility of the first information block and improving system performance.

[0152] As an embodiment, the benefits of the above method include adapting to the characteristics of training-based operations and maximizing the advantages of training-based operations.

[0153] As an embodiment, the model of each operation in the M operations is obtained through training.

[0154] As an embodiment, the training of at least one operation among the M operations is performed by the first node.

[0155] As an embodiment, the training of any one of the M operations is performed by the first node.

[0156] As an embodiment, the training of at least one operation among the M operations is performed by the sender of the first configuration information block.

[0157] As an embodiment, the training of any one of the M operations is performed by the sender of the first configuration information block.

[0158] As an embodiment, the training of at least one operation among the M operations is performed by the core network.

[0159] As an embodiment, the training of any one of the M operations is performed by the core network.

[0160] As an embodiment, the training of at least one of the M operations is performed by an AI training producer.

[0161] As an embodiment, the training of any one of the M operations is performed by an AI training producer.

[0162] As an embodiment, the training of at least one operation among the M operations is performed by an MDA function (Management Data Analytics Function).

[0163] As an embodiment, the training of any one of the M operations is performed by the MDA function.

[0164] As an embodiment, the training of at least one operation among the M operations is performed by an MDA function located at the first node.

[0165] As an embodiment, the training of any one of the M operations is performed by an MDA function located at the first node.

[0166] As an embodiment, the training of at least one operation among the M operations is performed by an MDA function located at a sender of the first configuration information block.

[0167] As an embodiment, the training of at least one operation among the M operations is performed by NWDAF (Network Data Analytics Function).

[0168] As an embodiment, the training of any one of the M operations is performed by NWDAF.

[0169] As an embodiment, the training of at least one operation among the M operations is performed by a MDAS (Management Data Analytics Service) producer.

[0170] As an embodiment, the training of any one of the M operations is performed by an MDAS producer.

[0171] As an embodiment, the training of at least one operation among the M operations is performed by a MnS (Management Service) producer.

[0172] As an embodiment, the training of any one of the M operations is performed by a MnS producer.

[0173] As an embodiment, the M operations are trained jointly.

[0174] As an embodiment, the benefits of the above method include jointly optimizing the M operations.

[0175] As an embodiment, at least two operations among the M operations are jointly trained.

[0176] As an embodiment, at least two operations among the M operations are trained separately.

[0177] As an embodiment, the benefits of the above method include improved flexibility and reduced training complexity.

[0178] As an embodiment, any two operations among the M operations are trained separately.

[0179] As an embodiment, the training of at least one operation among the M operations depends on the result of the training of at least one other operation among the M operations.

[0180] As an example, the benefits of the above method include achieving a better balance between performance, flexibility and complexity.

[0181] As an embodiment, the trainers of the M operations are the same.

[0182] As an embodiment, trainers of at least two of the M operations are different.

[0183] As an embodiment, at least one operation among the M operations includes inference.

[0184] As an embodiment, each of the M operations includes inference.

[0185] As an embodiment, the reasoning refers to AI (Artificial Intelligence) reasoning.

[0186] As an embodiment, the problem to be solved by the present application includes how to generate the first information block when only M1 operations among the M configured AI-based operations are executed; in the method disclosed in the present application, the first processing for generating the first information block depends on the M1 operations being executed, thereby solving the above problem.

[0187] As an embodiment, the benefits of the above method include ensuring the reliability and robustness of the first information block and improving system performance.

[0188] As an embodiment, the benefits of the above method include adapting to the characteristics of AI-based operations and maximizing the advantages of AI-based operations.

[0189] As an embodiment, each of the M operations is inference.

[0190] As an embodiment, each of the M operations is AI reasoning.

[0191] As an embodiment, at least one of the M operations includes AI reasoning for CSI (Channel State Information).

[0192] As an embodiment, at least one operation among the M operations is used for AI reasoning of CSI.

[0193] As an embodiment, at least one of the M operations includes AI reasoning for CSI prediction, CSI estimation or CSI compression.

[0194] As an embodiment, each of the M operations includes an AI entity.

[0195] As an embodiment, each of the M operations includes an AI entity for inference.

[0196] As an embodiment, each of the M operations includes a portion of an AI entity.

[0197] As an embodiment, each of the M operations includes a part used for reasoning in an AI entity.

[0198] As an embodiment, each of the M operations is performed by an AI entity.

[0199] As an embodiment, each of the M operations is executed by an AI entity deployed on the first node.

[0200] As an embodiment, each of the M operations is performed by an AI function.

[0201] As an embodiment, each of the M operations is executed by an AI function deployed on the first node.

[0202] As an embodiment, the AI ​​function includes an AI reasoning function.

[0203] As an embodiment, the AI ​​function includes an AI training function.

[0204] As an embodiment, the AI ​​function includes an AI management function.

[0205] As an embodiment, the AI ​​includes ML (Machine Learning).

[0206] As an embodiment, the AI ​​includes AI and ML.

[0207] As an embodiment, the AI ​​includes AI or ML.

[0208] As an embodiment, at least one operation among the M operations requires deployment.

[0209] As an embodiment, each of the M operations requires deployment.

[0210] As an embodiment, at least one operation among the M operations is obtained by loading.

[0211] As an embodiment, each of the M operations is obtained by loading.

[0212] As an embodiment, a part of the M operations needs to be deployed or loaded, and another part of the M operations does not need to be deployed or loaded.

[0213] As a sub-embodiment of the above embodiment, each operation in the another part of operations is trained at the first node.

[0214] As an embodiment, at least one operation among the M operations is loaded and obtained from a serving cell of the first node.

[0215] As an embodiment, each of the M operations is loaded and obtained from a serving cell of the first node.

[0216] As an embodiment, at least one operation among the M operations is loaded and obtained from a base station maintaining a service cell of the first node.

[0217] As an embodiment, each of the M operations is obtained from maintaining base station loading of a serving cell of the first node.

[0218] As an embodiment, at least one of the M operations is obtained by loading from the core network.

[0219] As an embodiment, each of the M operations is loaded and obtained from the core network.

[0220] As an embodiment, each of the M operations is based on artificial intelligence or machine learning.

[0221] As an embodiment, each of the M operations is based on a neural network.

[0222] As an embodiment, at least one of the M operations includes CSI compression based on artificial intelligence or machine learning.

[0223] As an embodiment, at least one of the M operations includes an encoder for CSI compression based on artificial intelligence or machine learning.

[0224] As an embodiment, at least one of the M operations includes CSI prediction or CSI estimation based on artificial intelligence or machine learning.

[0225] As an embodiment, at least one of the M operations includes beam management based on artificial intelligence or machine learning.

[0226] As an embodiment, at least one of the M operations includes positioning based on artificial intelligence or machine learning.

[0227] As an embodiment, at least one of the M operations includes positioning assisted by artificial intelligence or machine learning.

[0228] As an embodiment, the output of at least one of the M operations includes CSI.

[0229] As an embodiment, the output of each of the M operations includes CSI.

[0230] As an embodiment, the output of at least one of the M operations includes compressed CSI.

[0231] As an embodiment, the output of each of the M operations includes compressed CSI.

[0232] As an embodiment, the compressed CSI is based on artificial intelligence or machine learning.

[0233] As an embodiment, the compressed CSI is non-codebook based.

[0234] As an embodiment, the compressed CSI does not belong to the CSI defined in 3GPP Rel-18, nor does it belong to the CSI defined in versions before 3GPP Rel-18.

[0235] As an embodiment, the channel parameters recovered by the sender of the first configuration information block based on the compressed CSI are unknown to the first node.

[0236] As an embodiment, the channel parameters recovered by the receiver of the compressed CSI based on the compressed CSI are unknown to the sender of the compressed CSI.

[0237] As an embodiment, the output of at least one of the M operations includes predicted CSI or predicted beam information.

[0238] As an embodiment, the output of each of the M operations includes predicted CSI or predicted beam information.

[0239] As an embodiment, the beam information includes one or more of CRI (CSI-RS Resource Indicator), SSBRI (SS / PBCH Block Resource indicator) and RSRP (Reference Signal Received Power).

[0240] As an embodiment, an output of at least one of the M operations includes non-codebook based CSI.

[0241] As an embodiment, the output of each of the M operations includes non-codebook based CSI.

[0242] As an embodiment, the non-codebook based CSI does not belong to the CSI defined in 3GPP Rel-18, nor does it belong to the CSI defined in versions before 3GPP Rel-18.

[0243] As an embodiment, the channel parameters recovered by the sender of the first configuration information block based on the non-codebook based CSI are unknown to the first node.

[0244] As an embodiment, the channel parameters recovered by the receiver of the non-codebook based CSI based on the non-codebook based CSI are unknown to the sender of the non-codebook based CSI.

[0245] As an embodiment, the output of at least one of the M operations includes channel information.

[0246] As an embodiment, the output of each of the M operations includes channel information.

[0247] As an embodiment, the channel information includes CSI.

[0248] As an embodiment, the channel information includes one or more of PMI (Precoding Matrix Indicator), CRI, CQI (Channel Quality Indicator), RI (Rank Indicator), LI (layer indicator), SSBRI, RSRP, SINR (Signal-to-Interference-plus-Noise Ratio), capability index (CapabilityIndex) and TDCP (Time Domain Channel Properties).

[0249] As an embodiment, the channel information includes a channel impulse response.

[0250] As an embodiment, the channel information includes small-scale characteristics.

[0251] As an embodiment, the channel information includes one or more of delay spread, Doppler spread, Doppler shift, average delay and average gain.

[0252] As an embodiment, the channel information includes a channel matrix.

[0253] As an embodiment, the channel information includes an eigenvector.

[0254] As an embodiment, an input of at least one of the M operations includes measurements obtained based on CSI resources.

[0255] As an embodiment, the CSI resources include one or more of CSI-RS (Channel State Information Reference Signal) resources, SS / PBCH (Synchronisation Signal / Physical Broadcast Channel) block resources and CSI-IM (Channel State Information-Interference Measurement) resources.

[0256] As an embodiment, an input of at least one of the M operations includes channel measurement obtained based on CSI-RS resources or SS / PBCH block resources.

[0257] As an embodiment, an input of at least one of the M operations includes interference measurement obtained based on CSI-RS resources or CSI-IM resources.

[0258] As an embodiment, inputs of any two operations among the M operations respectively include measurements obtained based on different CSI resources.

[0259] As an embodiment, inputs of two operations among the M operations respectively include measurements obtained based on different CSI resources.

[0260] As an embodiment, inputs of two operations among the M operations respectively include channel measurements obtained based on different CSI resources.

[0261] As an embodiment, inputs of two operations among the M operations respectively include interference measurements obtained based on different CSI resources.

[0262] As an embodiment, an input of at least one of the M operations includes a reception quality of at least one physical channel or physical signal.

[0263] As an embodiment, the reception quality includes one or more of reception power, RSRP, SINR, BLER (BLock Error Rate, block error rate), constellation point error and LLR (log-likelihood ratio).

[0264] As an embodiment, inputs of two operations among the M operations respectively include reception qualities of different physical channels or physical signals.

[0265] As an embodiment, the input of at least one of the M operations includes positioning information.

[0266] As an embodiment, the positioning information is obtained by measuring a PRS (Positioning Reference Signal).

[0267] As an embodiment, the input of at least one of the M operations includes one or more of an ACK (acknowledgement) / NACK (negative acknowledgement) ratio, a number of NACKs, a number of beam failures, and a number of radio link failures.

[0268] As an embodiment, the M1 is equal to the M.

[0269] As an embodiment, the M1 is smaller than the M.

[0270] As an embodiment, any operation among the M1 operations is one operation among the M operations.

[0271] As an embodiment, the first node performs only the M1 operations among the M operations.

[0272] As an embodiment, the M1 is smaller than the M, and the first node executes only the M1 operations among the M operations.

[0273] As an embodiment, the M1 operations include no failed operations among the M operations.

[0274] As an embodiment, the M1 operations include all non-failed operations among the M operations.

[0275] As an embodiment, the M1 operations consist of operations that have not failed among the M operations.

[0276] As an embodiment, the input of the first processing includes the outputs of the M1 operations.

[0277] As an embodiment, the input of the first process includes the output of each of the M1 operations.

[0278] As an embodiment, the outputs of the M1 operations are used as the inputs of the first processing.

[0279] As an embodiment, the output of each of the M1 operations is used as the input of the first process.

[0280] As an embodiment, the input of the first processing includes the quantized outputs of the M1 operations.

[0281] As an embodiment, the input of the first processing includes the post-processed outputs of the M1 operations.

[0282] As an embodiment, the input of the first processing includes the post-processed output of each of the M1 operations.

[0283] As an embodiment, the outputs of the M1 operations are quantized and used as inputs of the first processing.

[0284] As an embodiment, the input of the first processing includes M1 inputs, and the M1 inputs are respectively dependent on the outputs of the M1 operations.

[0285] As an embodiment, the M1 inputs respectively include part or all of the outputs of the M1 operations.

[0286] As an embodiment, the M1 inputs respectively include part or all of the post-processed outputs of the M1 operations.

[0287] As an embodiment, the first information block includes the output of the first processing.

[0288] As an embodiment, the first information block includes all or part of the output of the first processing.

[0289] As an embodiment, the output of the first processing is used to generate the first information block.

[0290] As an embodiment, all or part of the output of the first processing is used to generate the first information block.

[0291] As an embodiment, the first information block comprises a post-processed output of the first processing.

[0292] As an embodiment, the first information block includes all or part of the post-processed output of the first processing.

[0293] As an embodiment, all or part of the output of the first processing is used to generate the first information block after post-processing.

[0294] As an embodiment, the post-processing includes one or more of quantization, truncation, matrix decomposition, domain transformation and DFT (DiscreteFourier Transform).

[0295] As an embodiment, the domain transformation includes one or more of the transformation from the angle domain to the spatial domain, the transformation from the spatial domain to the angle domain, the transformation from the time domain to the frequency domain, the transformation from the frequency domain to the time domain, the transformation from the delay domain to the frequency domain, the transformation from the frequency domain to the delay domain, the transformation from the Doppler domain to the time domain and the transformation from the time domain to the Doppler domain.

[0296] As an embodiment, the first information block includes CSI.

[0297] As an embodiment, the benefits of the above method include improving the accuracy of CSI, thereby improving the overall performance of the system.

[0298] As an embodiment, the first information block includes compressed CSI.

[0299] As an embodiment, the first information block includes compressed PMI.

[0300] As an embodiment, the first information block includes predicted beam information.

[0301] As an embodiment, the first information block includes a predicted PMI.

[0302] As an embodiment, the first information block includes channel information.

[0303] As an embodiment, the benefits of the above method include improving the accuracy of channel information, thereby improving the overall performance of the system.

[0304] As an embodiment, the first information block includes one or more of PMI, CRI, CQI, RI, LI, SSBRI, RSRP, SINR, capability index and TDCP.

[0305] As an embodiment, the first information block includes CSI, and the first configuration information block includes CSI reporting configuration.

[0306] As an embodiment, the first information block indicates the M1 operations.

[0307] As an embodiment, the first information block indicates the M1 operations from among the M operations.

[0308] As an embodiment, the first information block indicates which operation or operations among the M operations are included in the M1 operations.

[0309] As an embodiment, the first configuration information block is used to determine the relationship between the first process and the M1 operations.

[0310] As an embodiment, the first configuration information block indicates the relationship between the first processing and the M1 operations.

[0311] As an embodiment, the first processing includes calculation.

[0312] As an embodiment, the operation includes a mathematical operation.

[0313] As an embodiment, the operation includes a matrix operation.

[0314] As an embodiment, the operation includes quantization.

[0315] As an embodiment, the operation includes weighting.

[0316] As an embodiment, the operation includes weighted combining.

[0317] As an embodiment, the operation includes one or more of DFT (Discrete Fourier Transform) transformation, DFT inverse transformation and domain transformation.

[0318] As an embodiment, the domain transformation includes one or more of the transformation from the angle domain to the spatial domain, the transformation from the spatial domain to the angle domain, the transformation from the time domain to the frequency domain, the transformation from the frequency domain to the time domain, the transformation from the delay domain to the frequency domain, the transformation from the frequency domain to the delay domain, the transformation from the Doppler domain to the time domain and the transformation from the time domain to the Doppler domain.

[0319] As an embodiment, the operation includes one or more of truncation, projection, matrix decomposition, filtering, averaging and summation.

[0320] As an embodiment, the first processing includes weighting.

[0321] As an embodiment, the first processing includes weighted combining.

[0322] As an embodiment, the first processing includes subjecting the outputs of the M1 operations to mathematical operations or matrix operations to generate the first information block.

[0323] As an embodiment, the first processing includes weighting the M1 inputs separately and then merging them.

[0324] As an embodiment, the first processing includes projecting part or all of the M1 inputs onto a basis matrix respectively, and then weighting and merging the projected results.

[0325] As a sub-embodiment of the above embodiment, the first processing further includes projecting the result of the merging onto a basis matrix.

[0326] As an embodiment, the first processing includes performing DFT transformation or inverse DFT transformation on the M1 inputs respectively, and then weighting and combining the projected results.

[0327] As a sub-embodiment of the above embodiment, the first processing further includes performing DFT inverse transformation or DFT transformation on the result of the merging.

[0328] As an embodiment, the first processing includes performing domain transformation on part or all of the M1 inputs respectively, and then merging the results of the domain transformation after weighting.

[0329] As a sub-embodiment of the above embodiment, the first processing further includes performing domain transformation on the result of the merging.

[0330] As an embodiment, the first processing includes which operations depend on the M1 operations.

[0331] As an embodiment, the first processing includes which operations depend on the M1.

[0332] As an embodiment, which operations are included in the first processing depends on which operation or operations among the M operations are included in the M1 operations.

[0333] As an embodiment, the first processing includes which operations depend on the outputs of the M1 operations.

[0334] As an embodiment, the first processing includes which operations depend on the reliability of the M1 operations.

[0335] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and the first processing includes which one or more candidate operations of the plurality of candidate operations, depending on the M1 operations.

[0336] As an embodiment, the first processing includes quantization, and a criterion of the quantization depends on the M1 operations.

[0337] As an embodiment, the first processing includes quantization, and a criterion of the quantization depends on the M1.

[0338] As an embodiment, the first processing includes quantization, and a criterion of the quantization depends on which operation or operations among the M operations the M1 operations include.

[0339] As an embodiment, the first processing includes quantization, and a criterion of the quantization depends on outputs of the M1 operations.

[0340] As an embodiment, the first processing includes quantization, and a criterion of the quantization depends on the reliability of the M1 operations.

[0341] As an embodiment, the quantization criteria include scalar quantization and vector quantization.

[0342] As an embodiment, the first processing includes quantization, the first configuration information block indicates a plurality of candidate quantization criteria, and the first processing adopts which one or more of the plurality of candidate quantization criteria, depending on the M1 operations.

[0343] As an embodiment, the first processing includes quantization, and the accuracy of the quantization depends on the M1 operations.

[0344] As an embodiment, the first processing includes quantization, and the accuracy of the quantization depends on the M1.

[0345] As an embodiment, the first processing includes quantization, and the accuracy of the quantization depends on which operation or operations of the M operations are included in the M1 operations.

[0346] As an embodiment, the first processing includes quantization, and the accuracy of the quantization depends on the outputs of the M1 operations.

[0347] As an embodiment, the first processing includes quantization, and the accuracy of the quantization depends on the reliability of the M1 operations.

[0348] As an embodiment, the first processing includes quantization, the first configuration information block indicates a plurality of candidate quantization precisions, and which one or more candidate quantization precisions of the plurality of candidate quantization precisions are adopted by the first processing depends on the M1 operations.

[0349] As an embodiment, the first processing includes weighting, and the weighting coefficient depends on the M1 operations.

[0350] As an embodiment, the first processing includes weighting, and the weighting coefficient depends on the M1.

[0351] As an embodiment, the first processing includes weighting, and the weighting coefficient depends on which operation or operations among the M operations the M1 operations include.

[0352] As an embodiment, the first processing includes weighting, and the weighting coefficient depends on the outputs of the M1 operations.

[0353] As an embodiment, the first processing includes weighting, and the weighting coefficient depends on the reliability of the M1 operations.

[0354] As an embodiment, the first processing includes weighting, the first configuration information block indicates a plurality of candidate weighting coefficients, and which one or more of the plurality of candidate weighting coefficients are adopted by the first processing depends on the M1 operations.

[0355] Example 2

[0356] Embodiment 2 illustrates a schematic diagram of a network architecture according to an embodiment of the present application, as shown in the attached Figure 2 shown.

[0357] Attached Figure 2The network architecture 200 is described. The network architecture 200 is a 5G NR (New Radio) / LTE (Long-Term Evolution) / LTE-A (Long-Term Evolution Advanced) system, or the network architecture 200 is a 5G+ network architecture, or the network architecture 200 is a 6G network architecture, or the network architecture 200 is a network architecture adopted by 3GPP in the future evolution; the network architecture 200 may be called 5GS (5G System) / EPS (Evolved Packet System), or the network architecture 200 may be called 6GS (6G System); the network architecture 200 includes at least one of UE (User Equipment) 201, RAN (Radio Access Network) 202, core network 210, HSS (Home Subscriber Server) / UDM (Unified Data Management) 220 and Internet service 230. The network architecture 200 may be interconnected with other access networks, but these entities / interfaces are not shown for simplicity. As shown, the network architecture 200 provides packet switching services, but it will be readily understood by those skilled in the art that the various concepts presented throughout this application can be extended to networks or other cellular networks that provide circuit switching services. The RAN includes a node 203. The RAN may also include other nodes 204. The node 203 provides user and control plane protocol terminations toward the UE 201. The node 203 may be connected to the other nodes 204 via an Xn interface (e.g., backhaul) / X2 interface. The node 203 may also be referred to as a base station, a base transceiver station, a radio base station, a radio transceiver, a transceiver function, a basic service set (BSS), an extended service set (ESS), a TRP (transmitting receiving node) or some other suitable terminology. The core network 210 is 5GC (5G Core Network) / EPC (Evolved Packet Core), or the core network 210 is 6GC; the node 203 provides an access point to the core network 210 for the UE 201.Examples of UE201 include cellular phones, smart phones, session initiation protocol (SIP) phones, laptop computers, personal digital assistants (PDAs), satellite radios, non-terrestrial base station communications, satellite mobile communications, global positioning systems, multimedia devices, video devices, digital audio players (e.g., MP3 players), cameras, game consoles, drones, aircraft, narrowband Internet of Things devices, machine type communication devices, land vehicles, cars, wearable devices, or any other similar functional devices. Those skilled in the art may also refer to UE201 as a mobile station, a subscriber station, a mobile unit, a subscriber unit, a wireless unit, a remote unit, a mobile device, a wireless device, a wireless communication device, a remote device, a mobile subscriber station, an access terminal, a mobile terminal, a wireless terminal, a remote terminal, a handset, a user agent, a mobile client, a client, or some other suitable term. Node 203 is connected to the core network 210 via an S1 / NG interface. The core network 210 includes MME (Mobility Management Entity) / AMF (Authentication Management Field) / SMF (Session Management Function) 211, other MME / AMF / SMF214, S-GW (Service Gateway) / UPF (User Plane Function) 212 and P-GW (Packet Data Network Gateway) / UPF213. MME / AMF / SMF211 is a control node that handles signaling between UE201 and the core network 210. Generally, MME / AMF / SMF211 provides bearer and connection management. All user IP (Internet Protocol) packets are transmitted through S-GW / UPF212, and S-GW / UPF212 itself is connected to P-GW / UPF213. P-GW provides UE IP address allocation and other functions. P-GW / UPF213 is connected to Internet service 230. The Internet service 230 includes the operator's corresponding Internet protocol service, which may specifically include the Internet, intranet, IMS (IP Multimedia Subsystem) and packet switching services.

[0358] As an embodiment, the first node includes the UE201.

[0359] As an embodiment, the second node includes the node 203.

[0360] As an embodiment, the wireless link between the UE 201 and the node 203 includes a cellular network link.

[0361] As an embodiment, the sender of the first configuration information block includes the node 203.

[0362] As an embodiment, the receiver of the first configuration information block includes the UE201.

[0363] As an embodiment, the executor of the M1 operations includes the UE201.

[0364] As an embodiment, the sender of the first information block includes the UE201.

[0365] As an embodiment, the receiver of the first information block includes the node 203.

[0366] As an embodiment, the UE 201 supports operations based on AI or ML.

[0367] As an embodiment, the UE 201 supports CSI reporting based on AI or ML.

[0368] As an embodiment, the node 203 supports AI or ML based operations.

[0369] As an embodiment, the node 203 supports CSI reception based on AI or ML.

[0370] Example 3

[0371] Embodiment 3 illustrates a schematic diagram of an embodiment of a wireless protocol architecture of a user plane and a control plane according to an embodiment of the present application, as shown in the attached figure. Figure 3 shown.

[0372] Embodiment 3 shows a schematic diagram of an embodiment of a wireless protocol architecture of a user plane and a control plane according to the present application, as shown in the attached Figure 3 shown. Figure 3 is a schematic diagram illustrating an embodiment of a radio protocol architecture for a user plane 350 and a control plane 300, Figure 3The radio protocol architecture of the control plane 300 between a first communication node device (UE, gNB or RSU in V2X) and a second communication node device (gNB, UE or RSU in V2X), or between two UEs is shown with three layers: Layer 1, Layer 2 and Layer 3. Layer 1 (L1 layer) is the lowest layer and implements various PHY (physical layer) signal processing functions. The L1 layer will be referred to as PHY301 in this article. Layer 2 (L2 layer) 305 is above PHY301 and is responsible for the link between the first communication node device and the second communication node device, or between two UEs. The L2 layer 305 includes a MAC (Medium Access Control) sublayer 302, an RLC (Radio Link Control) sublayer 303, and a PDCP (Packet Data Convergence Protocol) sublayer 304, which terminate at the second communication node device. The PDCP sublayer 304 provides multiplexing between different radio bearers and logical channels. The PDCP sublayer 304 also provides security by encrypting data packets, and provides support for inter-zone mobility of the first communication node device between the second communication node device. The RLC sublayer 303 provides segmentation and reassembly of upper layer data packets, retransmission of lost data packets, and reordering of data packets to compensate for out-of-order reception due to HARQ. The MAC sublayer 302 provides multiplexing between logical and transport channels. The MAC sublayer 302 is also responsible for allocating various radio resources (e.g., resource blocks) in a cell between the first communication node devices. The MAC sublayer 302 is also responsible for HARQ operations. The RRC (Radio Resource Control) sublayer 306 in layer 3 (L3 layer) in the control plane 300 is responsible for obtaining radio resources (i.e., radio bearers) and configuring the lower layer using RRC signaling between the second communication node device and the first communication node device. The radio protocol architecture of the user plane 350 includes layer 1 (L1 layer) and layer 2 (L2 layer). The radio protocol architecture used for the first communication node device and the second communication node device in the user plane 350 is substantially the same as the corresponding layers and sublayers in the control plane 300 for the physical layer 351, the PDCP sublayer 354 in the L2 layer 355, the RLC sublayer 353 in the L2 layer 355, and the MAC sublayer 352 in the L2 layer 355, but the PDCP sublayer 354 also provides header compression for upper layer data packets to reduce radio transmission overhead.The L2 layer 355 in the user plane 350 also includes a SDAP (Service Data Adaptation Protocol) sublayer 356, which is responsible for mapping between QoS flows and data radio bearers (DRBs) to support the diversity of services. Although not shown in the figure, the first communication node device may have several upper layers above the L2 layer 355, including a network layer (e.g., an IP layer) terminated at the P-GW on the network side and an application layer terminated at the other end of the connection (e.g., a remote UE, a server, etc.).

[0373] As an example, Figure 3 The wireless protocol architecture is applicable to the first node.

[0374] As an example, Figure 3 The wireless protocol architecture is applicable to the second node.

[0375] As an embodiment, the higher layer in the present application refers to a layer above the physical layer.

[0376] As an embodiment, the first configuration information block is generated in the RRC sublayer 306.

[0377] As an embodiment, the first information block is generated in the PHY301 or the PHY351.

[0378] As an embodiment, the first signaling is generated in the RRC sublayer 306.

[0379] As an embodiment, the first signaling is generated in the MAC sublayer 302 or the MAC sublayer 352.

[0380] Example 4

[0381] Embodiment 4 illustrates a schematic diagram of a first communication device and a second communication device according to an embodiment of the present application, as shown in the attached figure. Figure 4 Attached Figure 4 is a block diagram of a first communication device 410 and a second communication device 450 communicating with each other in an access network.

[0382] The first communication device 410 includes a controller / processor 475 , a memory 476 , a receive processor 470 , a transmit processor 416 , a multi-antenna receive processor 472 , a multi-antenna transmit processor 471 , a transmitter / receiver 418 and an antenna 420 .

[0383] The second communication device 450 includes a controller / processor 459, a memory 460, a data source 467, a transmit processor 468, a receive processor 456, a multi-antenna transmit processor 457, a multi-antenna receive processor 458, a transmitter / receiver 454 and an antenna 452.

[0384] In the transmission from the first communication device 410 to the second communication device 450, at the first communication device 410, the upper layer data packets from the core network are provided to the controller / processor 475. The controller / processor 475 implements the functionality of the L2 layer. In DL (DownLink), the controller / processor 475 provides header compression, encryption, packet segmentation and reordering, multiplexing between logical and transport channels, and radio resource allocation to the second communication device 450 based on various priority metrics. The controller / processor 475 is also responsible for HARQ operations, retransmission of lost packets, and signaling to the second communication device 450. The transmit processor 416 and the multi-antenna transmit processor 471 implement various signal processing functions for the L1 layer (i.e., the physical layer). The transmit processor 416 implements coding and interleaving to facilitate forward error correction (FEC) at the second communication device 450, as well as constellation mapping based on various modulation schemes (e.g., binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), M-phase shift keying (M-PSK), M-quadrature amplitude modulation (M-QAM). The multi-antenna transmit processor 471 performs digital spatial precoding on the coded and modulated symbols, including codebook-based precoding and non-codebook-based precoding, and beamforming processing to generate one or more parallel The transmit processor 416 then maps each parallel stream to a subcarrier, multiplexes the modulated symbols with a reference signal (e.g., a pilot) in the time domain and / or frequency domain, and then uses an inverse fast Fourier transform (IFFT) to generate a physical channel carrying a time domain multi-carrier symbol stream. The multi-antenna transmit processor 471 then performs a transmit analog precoding / beamforming operation on the time domain multi-carrier symbol stream. Each transmitter 418 converts the baseband multi-carrier symbol stream provided by the multi-antenna transmit processor 471 into a radio frequency stream, and then provides it to different antennas 420.

[0385] In the transmission from the first communication device 410 to the second communication device 450, at the second communication device 450, each receiver 454 receives a signal through its corresponding antenna 452. Each receiver 454 recovers the information modulated onto the RF carrier and converts the RF stream into a baseband multi-carrier symbol stream and provides it to the receiving processor 456. The receiving processor 456 and the multi-antenna receiving processor 458 implement various signal processing functions of the L1 layer. The multi-antenna receiving processor 458 performs a receiving analog precoding / beamforming operation on the baseband multi-carrier symbol stream from the receiver 454. The receiving processor 456 uses a fast Fourier transform (FFT) to convert the baseband multi-carrier symbol stream after the receiving analog precoding / beamforming operation from the time domain to the frequency domain. In the frequency domain, the physical layer data signal and the reference signal are demultiplexed by the receiving processor 456, wherein the reference signal will be used for channel estimation, and the data signal is recovered after multi-antenna detection in the multi-antenna receiving processor 458 to any parallel stream destined for the second communication device 450. The symbols on each parallel stream are demodulated and recovered in the receiving processor 456, and soft decisions are generated. The receiving processor 456 then decodes and deinterleaves the soft decisions to recover the upper layer data and control signals transmitted by the first communication device 410 on the physical channel. The upper layer data and control signals are then provided to the controller / processor 459. The controller / processor 459 implements the functions of the L2 layer. The controller / processor 459 may be associated with a memory 460 storing program codes and data. The memory 460 may be referred to as a computer-readable medium. In DL, the controller / processor 459 provides multiplexing, packet reassembly, decryption, header decompression, and control signal processing between the transmission and logical channels to recover the upper layer data packets from the core network. The upper layer data packets are then provided to all protocol layers above the L2 layer. Various control signals may also be provided to L3 for L3 processing. The controller / processor 459 is also responsible for error detection using confirmation (ACK) and / or negative confirmation (NACK) protocols to support HARQ operations.

[0386] In the transmission from the second communication device 450 to the first communication device 410, at the second communication device 450, a data source 467 is used to provide upper layer data packets to the controller / processor 459. The data source 467 represents all protocol layers above the L2 layer. Similar to the transmission function at the first communication device 410 described in DL, the controller / processor 459 implements header compression, encryption, packet segmentation and reordering, and multiplexing between logical and transport channels based on the radio resource allocation of the first communication device 410, and implements L2 layer functions for the user plane and the control plane. The controller / processor 459 is also responsible for HARQ operations, retransmission of lost packets, and signaling to the first communication device 410. The transmit processor 468 performs modulation mapping and channel coding processing, and the multi-antenna transmit processor 457 performs digital multi-antenna spatial precoding, including codebook-based precoding and non-codebook-based precoding, and beamforming processing. Then, the transmit processor 468 modulates the generated parallel stream into a multi-carrier / single-carrier symbol stream, which is then provided to different antennas 452 via the transmitter 454 after analog precoding / beamforming operations in the multi-antenna transmit processor 457. Each transmitter 454 first converts the baseband symbol stream provided by the multi-antenna transmit processor 457 into a radio frequency symbol stream, and then provides it to the antenna 452.

[0387] In the transmission from the second communication device 450 to the first communication device 410, the function at the first communication device 410 is similar to the reception function at the second communication device 450 described in the transmission from the first communication device 410 to the second communication device 450. Each receiver 418 receives a radio frequency signal through its corresponding antenna 420, converts the received radio frequency signal into a baseband signal, and provides the baseband signal to the multi-antenna reception processor 472 and the reception processor 470. The reception processor 470 and the multi-antenna reception processor 472 jointly implement the functions of the L1 layer. The controller / processor 475 implements the L2 layer functions. The controller / processor 475 can be associated with a memory 476 that stores program codes and data. The memory 476 can be referred to as a computer-readable medium. The controller / processor 475 provides demultiplexing between transmission and logical channels, packet reassembly, decryption, header decompression, control signal processing to recover the upper layer data packets from the second communication device 450. The upper layer data packets from the controller / processor 475 can be provided to the core network. The controller / processor 475 is also responsible for error detection using an ACK and / or NACK protocol to support HARQ operations.

[0388] As an embodiment, the second communication device 450 includes: at least one processor and at least one memory, the at least one memory includes computer program code; the at least one memory and the computer program code are configured to be used with the at least one processor. The second communication device 450 device at least: receives the first configuration information block; executes the M1 operations among the M operations; generates the first information block. The first configuration information block indicates the M operations, M is a positive integer greater than 1, and M1 is a positive integer not greater than M; the first information block depends on the output of the first process, and the input of the first process depends on the output of the M1 operations; the first process depends on the M1 operations.

[0389] As an embodiment, the second communication device 450 includes: a memory storing a computer-readable instruction program, wherein the computer-readable instruction program generates actions when executed by at least one processor, and the actions include: receiving the first configuration information block; executing the M1 operations among the M operations; and generating the first information block.

[0390] As an embodiment, the first communication device 410 includes: at least one processor and at least one memory, the at least one memory includes computer program code; the at least one memory and the computer program code are configured to be used with the at least one processor. The first communication device 410 device at least: sends the first configuration information block. The first configuration information block indicates M operations, where M is a positive integer greater than 1; the target recipient of the first configuration information block determines the first information block, the first information block depends on the output of the first processing, the input of the first processing depends on the output of M1 operations, and each of the M1 operations is one of the M operations; the target recipient of the first configuration information block performs the M1 operations, where M1 is a positive integer not greater than M; the first processing depends on the M1 operations.

[0391] As an embodiment, the first communication device 410 includes: a memory storing a computer-readable instruction program, wherein the computer-readable instruction program generates an action when executed by at least one processor, and the action includes: sending the first configuration information block.

[0392] As an embodiment, the first node in the present application includes the second communication device 450.

[0393] As an embodiment, the second node in the present application includes the first communication device 410.

[0394] As an embodiment, at least one of {the antenna 452, the receiver 454, the receiving processor 456, the multi-antenna receiving processor 458, the controller / processor 459, the memory 460, and the data source 467} is used to receive the first configuration information block; and at least one of {the antenna 420, the transmitter 418, the transmitting processor 416, the multi-antenna transmitting processor 471, the controller / processor 475, and the memory 476} is used to send the first configuration information block.

[0395] As an embodiment, at least one of {the antenna 452, the receiver 454, the receiving processor 456, the multi-antenna receiving processor 458, the controller / processor 459, the memory 460, the data source 467} is used to perform the M1 operations among the M operations.

[0396] As an embodiment, at least one of {the antenna 452, the receiver 454, the receiving processor 456, the multi-antenna receiving processor 458, the controller / processor 459, the memory 460, the data source 467} is used to generate the first information block.

[0397] Example 5

[0398] Embodiment 5 illustrates a flow chart of transmission according to an embodiment of the present application; Figure 5 As shown in the attached Figure 5 In the embodiment, the second node U1 and the first node U2 are communication nodes transmitted via the air interface. Figure 5 In the process, the steps in box F51 to box F56 are respectively optional.

[0399] For the second node U1, a first operation is deployed in step S5101; a first configuration information block is sent in step S511; a first signaling is sent in step S5102; a first information block is received in step S5103; and a first operation is performed in step S5104.

[0400] For the first node U2, deploy M0 operations out of M operations in step S5201; receive the first configuration information block in step S521; receive the first signaling in step S5202; determine M1 operations out of M operations in step S5203; execute M1 operations out of M operations in step S522; generate the first information block in step S523; and send the first information block in step S5204.

[0401] In Embodiment 5, the first configuration information block indicates the M operations, where M is a positive integer greater than 1, and M1 is a positive integer not greater than M; the first information block depends on the output of the first processing, and the input of the first processing depends on the output of the M1 operations; and the first processing depends on the M1 operations.

[0402] As an embodiment, the first node U2 is the first node in this application.

[0403] As an embodiment, the second node U1 is the second node in this application.

[0404] As an embodiment, the air interface between the second node U1 and the first node U2 includes a wireless interface between a base station device and a user equipment.

[0405] As an embodiment, the air interface between the second node U1 and the first node U2 includes a wireless interface between a relay node device and a user equipment.

[0406] As an embodiment, the air interface between the second node U1 and the first node U2 includes a wireless interface between user equipments.

[0407] As an embodiment, the second node U1 is a base station maintaining a service cell of the first node U2.

[0408] As an embodiment, the first configuration information block is transmitted on PDSCH (Physical Downlink Shared Channel).

[0409] As an embodiment, the first configuration information block is transmitted on PDCCH (Physical Downlink Control Channel).

[0410] As an embodiment, the first node generates the first information block by performing the first processing.

[0411] As an example, Figure 5 The step in box F51 exists, and the first node U2 deploys M0 operations among the M operations, and M0 is a positive integer not greater than M.

[0412] As an embodiment, the deployment includes obtaining the M0 operations.

[0413] As an embodiment, the deploying includes loading the M0 operations.

[0414] As an embodiment, the deploying includes making a request to load the M0 operations.

[0415] As an embodiment, the deploying includes making a loading request for each of the M0 operations.

[0416] As an embodiment, the deployment is earlier than the reception of the first configuration information block.

[0417] As an embodiment, the deploying is later than the receiving of the first configuration information block.

[0418] As an embodiment, deployment of one of the M0 operations is earlier than the reception of the first configuration information block.

[0419] As an embodiment, deployment of one of the M0 operations is later than the reception of the first configuration information block.

[0420] As an embodiment, the M0 is smaller than the M.

[0421] As an embodiment, M0 is equal to M.

[0422] As an embodiment, the M0 is equal to the M, and the first node deploys each of the M operations.

[0423] As an embodiment, the M0 is smaller than the M, and the first node deploys only a portion of the M operations.

[0424] As a sub-embodiment of the above embodiment, another part of the M operations does not need to be deployed.

[0425] As a sub-embodiment of the above embodiment, the training of another part of the M operations is completed by the first node.

[0426] As an embodiment, the first configuration information block indicates M identifiers, and the M operations are respectively associated with the M identifiers.

[0427] As an embodiment, any operation among the M operations except the M1 operations is invalid.

[0428] As an example, Figure 5 The step in box F54 in exists, the first node U2 determines the M1 operations among the M operations.

[0429] As an embodiment, determining the M1 operations includes: determining the M1 operations independently.

[0430] As an embodiment, determining the M1 operations includes: being instructed about the M1 operations.

[0431] As an embodiment, determining the M1 operations includes: determining a part of the M1 operations by itself, and being instructed about another part of the M1 operations.

[0432] As an embodiment, for any operation among M2 operations among the M operations, the first node U2 determines by itself whether to execute this operation, and the M2 is a positive integer not greater than the M.

[0433] As an example, Figure 5 The step in box F53 exists, the first node U2 receives the first signaling; for any operation of M3 operations among the M operations, the first signaling indicates whether the first node U2 performs this operation, and M3 is a positive integer not greater than M.

[0434] As an embodiment, the first signaling is transmitted on PDSCH.

[0435] As an embodiment, the first signaling is transmitted on PDCCH (Physical Downlink Control Channel).

[0436] As an embodiment, the reception of the first signaling is later than the reception of the first configuration information block.

[0437] As an embodiment, the reception of the first signaling is earlier than the reception of the first configuration information block.

[0438] As an embodiment, the first processing depends on the reliability of the M1 operations.

[0439] As an embodiment, the input of the first processing includes M1 inputs, the M1 inputs are respectively dependent on the outputs of the M1 operations, and the first processing uses M1 coefficients to weight the M1 inputs respectively, and the M1 coefficients are dependent on the M1 operations.

[0440] As an example, Figure 5 The step in box F55 in occurs, said first node U2 sends said first information block.

[0441] As an embodiment, the first information block is transmitted on PUSCH (Physical Uplink Shared Channel).

[0442] As an embodiment, the first information block is transmitted on PUCCH (Physical Uplink Control Channel).

[0443] As an embodiment, the first information block is a report for the first configuration information block.

[0444] As an embodiment, the first information block includes CSI, the first configuration information block includes CSI reporting configuration, and the first information block includes a CSI report for the first configuration information block.

[0445] As an example, Figure 5 The step in box F55 in does not exist, and the first node uses the first information block on its own.

[0446] Generally speaking, how the first node uses the first information block is determined by the hardware device vendor. Some non-limiting implementation methods are described below:

[0447] As an embodiment, the first information block includes a PMI, and the first node determines a precoding matrix of a PUSCH according to the PMI included in the first information block.

[0448] As an embodiment, the first information block includes a CQI, and the first node determines an MCS (Modulation and Coding Scheme) of a PUSCH or a PUCCH according to the CQI included in the first information block.

[0449] As an embodiment, the first information block includes RI, and the first node determines the number of PUSCH layers according to the CQI included in the first information block.

[0450] As an embodiment, the first information block includes CRI or SSBRI, and the first node determines a spatial filter of PUSCH, PUCCH or SRS (Sounding reference signal) according to the CRI or SSBRI included in the first information block.

[0451] As an embodiment, the first information block includes CRI or SSBRI, and the first node determines the transmission power of PUSCH, PUCCH or SRS according to the CRI or SSBRI included in the first information block.

[0452] As an embodiment, the first information block includes CRI or SSBRI, and the first node determines the QCL (Quasi Co-Location) relationship or spatial reception parameters (spatial Rx parameters) of PDSCH, PDCCH or CSI-RS based on the CRI or SSBRI included in the first information block.

[0453] As an example, Figure 5 The steps in boxes F55 and F56 in the embodiment are both present, and the second node U1 performs a first operation; the first information block includes a first compressed CSI, and the first compressed CSI is used as an input of the first operation to generate a first recovered CSI.

[0454] As an example, Figure 5 The step in box F52 in is present, and the method in the second node used for wireless communication includes: deploying the first operation.

[0455] Example 6

[0456] Embodiment 6 illustrates a schematic diagram of the input of the first processing according to an embodiment of the present application; Figure 6 In Example 6, the input of the first process includes M1 inputs, and the M1 inputs are respectively dependent on the outputs of the M1 operations. Figure 6 In the example, the M1 inputs are respectively represented as input #0, ..., input #(M1-1); the M1 operations are respectively represented as operation #0, ..., operation #(M1-1).

[0457] As an embodiment, the M1 inputs are respectively outputs of the M1 operations.

[0458] As an embodiment, the M1 inputs are respectively post-processed outputs of the M1 operations.

[0459] As an embodiment, the M1 inputs respectively include part or all of the outputs of the M1 operations.

[0460] As an embodiment, the M1 inputs respectively include part or all of the post-processed outputs of the M1 operations.

[0461] As an embodiment, the outputs of the M1 operations are respectively used to generate the M1 inputs.

[0462] As an embodiment, for one of the M1 inputs, the output of the operation corresponding to the one input is used to generate the one input after post-processing.

[0463] As an embodiment, for one of the M1 inputs, the one input includes a post-processed partial or entire output of the corresponding operation.

[0464] As an embodiment, for one of the M1 inputs, all or part of the output of the operation corresponding to the one input is quantized and used to generate the one input; for another input of the M1 inputs, all or part of the output of the operation corresponding to the other input is domain transformed and used to generate the other input.

[0465] As an embodiment, for one of the M1 inputs, all or part of the output of the operation corresponding to the one input is used to generate the one input after being transformed between the angle domain and the spatial domain; for another input among the M1 inputs, all or part of the output of the operation corresponding to the other input is used to generate the other input after being transformed between the delay domain and the frequency domain.

[0466] Example 7

[0467] Embodiment 7 illustrates a schematic diagram of deploying a given operation according to an embodiment of the present application; Figure 7 As shown, the given operation is one of the M operations. In Embodiment 7, the first node makes a request to load the given operation to a first producer, and obtains the given operation from the first producer.

[0468] As an embodiment, the given operation is any operation among the M0 operations.

[0469] As an embodiment, the deployment includes obtaining the given operation.

[0470] As an embodiment, the deploying includes obtaining an AI entity.

[0471] As an embodiment, the deploying includes obtaining an AI entity that performs the given operation.

[0472] As an embodiment, the deploying includes obtaining an AI entity including an AI function for performing the given operation.

[0473] As an embodiment, the deploying includes obtaining an AI function.

[0474] As an embodiment, the deploying includes obtaining an AI function for performing the given operation.

[0475] As an embodiment, the deploying includes loading the given operation.

[0476] As an embodiment, the deploying includes making a request to load the given operation.

[0477] As an example, Figure 7 The request in is a request made by the first node to load the given operation.

[0478] As an example, Figure 7 The response in is a response to the request made by the first node to load the given operation.

[0479] As an embodiment, the first node is connected Figure 7 The response in obtains the given operation.

[0480] As an embodiment, the first node is connected Figure 7 The response in obtains an AI entity including an AI function that performs the given operation.

[0481] As an embodiment, the first node is connected Figure 7 The response in obtains the AI ​​function that performs the given operation.

[0482] As an embodiment, the first producer attaches Figure 7 The response in provides the given operation to the first node.

[0483] As an embodiment, the first producer attaches Figure 7 The response in provides, to the first node, an AI entity including an AI function for performing the given operation.

[0484] As an embodiment, the first producer attaches Figure 7 The response in provides the first node with AI functionality for performing the given operation.

[0485] As an embodiment, the deployment is performed by an AI function.

[0486] As an embodiment, the deployment is completed by an AI function deployed on the first node.

[0487] As an embodiment, the deployment is performed by an AI deployment function.

[0488] As an embodiment, the deployment is completed by an AI deployment function deployed on the first node.

[0489] As an embodiment, the deployment is performed by AI inference function.

[0490] As an embodiment, the deployment is accomplished by an AI inference function deployed on the first node.

[0491] As an embodiment, the deployment is performed by an AI entity.

[0492] As an embodiment, the deployment is completed by an AI entity deployed on the first node.

[0493] As an embodiment, the deployment is completed by an AI entity having a deployment function.

[0494] As an embodiment, the deployment is completed by an AI entity with deployment function deployed on the first node.

[0495] As an embodiment, the deployment is performed by an AI entity having an inference function.

[0496] As an embodiment, the deployment is completed by an AI entity with inference function deployed on the first node.

[0497] As an embodiment, the deploying includes obtaining the given operation from a first producer.

[0498] As an embodiment, the deploying includes making a request to the first producer to load the given operation.

[0499] As an embodiment, the deploying includes loading the given operation from a first producer.

[0500] As an embodiment, the first producer generates and provides an AI entity.

[0501] As an embodiment, the first producer generates and provides AI functionality.

[0502] As an embodiment, the first producer is the producer of the given operation.

[0503] As an embodiment, the first producer is a trained producer of the given operation.

[0504] As an embodiment, the first producer includes an AI entity producer.

[0505] As an embodiment, the first producer includes an AI function producer.

[0506] As an embodiment, the first producer includes an AI deployment producer (producer).

[0507] As an embodiment, the first producer includes an AI loading producer (producer).

[0508] As an embodiment, the first producer includes an AI training producer (producer).

[0509] As an embodiment, the first producer includes an AI inference producer.

[0510] As an embodiment, the first producer includes a producer of deployment of an AI entity.

[0511] As an embodiment, the first producer comprises a producer of loading of an AI entity.

[0512] As an embodiment, the first producer includes a producer of AI model training.

[0513] As an embodiment, the first producer includes a MnS (Management Service) producer.

[0514] As an embodiment, the first producer is a serving cell of the first node.

[0515] As an embodiment, the first producer is a base station maintaining a service cell of the first node.

[0516] As an embodiment, the first producer is a core network.

[0517] As an embodiment, the given operation is loaded and obtained from a serving cell of the first node.

[0518] As an embodiment, the given operation is obtained by loading from a base station maintaining a service cell of the first node.

[0519] As an embodiment, the given operation is loaded and obtained from the core network.

[0520] As an embodiment, the first producer is a sender of the first configuration information block.

[0521] As an embodiment, the first producer is different from the sender of the first configuration information block.

[0522] As an embodiment, the training of the given operation is performed by the first producer.

[0523] As an embodiment, producers of at least two operations among the M operations are different.

[0524] As an embodiment, at least two operations among the M operations are loaded and obtained from different producers.

[0525] As an embodiment, the producers of the M operations are the same.

[0526] As an embodiment, the M operations are loaded from the same producer.

[0527] Example 8

[0528] Embodiment 8 illustrates a schematic diagram of M identifiers and M operations according to an embodiment of the present application; Figure 8 As shown in the attached Figure 8 In the example, the M operations are respectively represented as operation #0, ..., operation #(M-1); the M identifiers are respectively represented as identifier #0, ..., identifier #(M-1).

[0529] As an embodiment, each of the M identifiers is a non-negative integer.

[0530] As an embodiment, each of the M identifiers is a character string.

[0531] As an embodiment, the M operations are respectively identified by the M identifiers.

[0532] As an embodiment, the AI ​​entity to which each operation in the M operations belongs is identified by a corresponding identifier in the M identifiers.

[0533] As an embodiment, the AI ​​function to which each operation in the M operations belongs is identified by a corresponding identifier in the M identifiers.

[0534] As an embodiment, for any operation among the M operations, the AI ​​function or AI entity that executes the operation is identified by an identifier among the M identifiers that corresponds to the operation.

[0535] As an embodiment, the benefits of the above method include simplifying the design and unifying the understanding of different AI entities or AI functions among different nodes.

[0536] As an embodiment, for any operation among the M operations, the training of this operation is identified by the identifier corresponding to this operation among the M identifiers.

[0537] As an embodiment, for any operation among the M operations, a training data set of this operation is identified by an identifier among the M identifiers corresponding to this operation.

[0538] As an embodiment, the benefits of the above method include that by identifying an AI training or an AI training data set to identify the inferences generated by this AI training or AI training data set, a consensus is established between different AI functions, further simplifying the design.

[0539] As an embodiment, the first configuration information block indicates the M operations by indicating the M identifiers.

[0540] As an embodiment, the first configuration information block indicates the M identifiers in sequence.

[0541] As an embodiment, the first configuration information block includes M configurations, and the M configurations respectively indicate the M operations.

[0542] As an embodiment, the M configurations respectively indicate the M identifiers.

[0543] As an embodiment, the M configurations respectively indicate the M operations by respectively indicating the M identifiers.

[0544] As an embodiment, the M configurations are arranged sequentially in the first configuration information block.

[0545] Example 9

[0546] Embodiment 9 illustrates a schematic diagram of M1 operations including non-failed operations among M operations according to an embodiment of the present application; Fig. 9 In Embodiment 9, any operation other than the M1 operation among the M operations is invalid.

[0547] As an embodiment, none of the M1 operations fails.

[0548] As an embodiment, the M1 operations include all non-failed operations among the M operations.

[0549] As an embodiment, the M1 operations consist of operations that have not failed among the M operations.

[0550] As an embodiment, for any operation among the M operations, the first node determines by itself whether the operation is invalid.

[0551] As an embodiment, for any operation among the M operations, the first node is indicated whether the operation fails.

[0552] As an embodiment, for any operation among a part of the M operations, the first node determines by itself whether the operation is invalid; for any operation among another part of the M operations, the first node is indicated whether the operation is invalid.

[0553] As an embodiment, for any operation among the M operations, if the operation has not failed, the first node executes the operation.

[0554] As an embodiment, for any operation among the M operations, the first node executes this operation only if this operation has not failed.

[0555] As an embodiment, for any operation among the M operations, if this operation fails, the first node does not perform this operation.

[0556] As an embodiment, for any operation among the M operations, the first node does not execute the operation only when the operation fails.

[0557] As an example, if an operation needs to be retrained, the operation is disabled.

[0558] As an embodiment, for any operation among the M operations, the first node determines by itself whether the operation needs to be retrained.

[0559] As an embodiment, for any operation among the M operations, the first node is indicated whether the operation needs to be retrained.

[0560] As an embodiment, for each operation in a part of the M operations, the first node determines by itself whether this operation needs to be retrained, and for each operation in another part of the M operations, the first node is indicated whether this operation needs to be retrained.

[0561] As an embodiment, for each operation of at least one of the M operations, the producer of this operation indicates whether this operation needs to be retrained.

[0562] As an embodiment, for any operation among the M operations, the producer of this operation indicates whether this operation needs to be retrained.

[0563] As an example, if an operation requires redeployment or reloading, the operation fails.

[0564] As an embodiment, for any operation among the M operations, the first node determines by itself whether the operation needs to be redeployed or reloaded.

[0565] As an embodiment, for any operation among the M operations, the first node is indicated whether the operation needs to be redeployed or reloaded.

[0566] As an embodiment, for each operation in a part of the M operations, the first node determines by itself whether this operation needs to be redeployed or reloaded, and for each operation in another part of the M operations, the first node is indicated whether this operation needs to be redeployed or reloaded.

[0567] As an embodiment, for each operation of at least one operation among the M operations, the producer of this operation indicates whether this operation needs to be redeployed or reloaded.

[0568] As an embodiment, for any operation among the M operations, the producer of this operation indicates whether this operation needs to be redeployed or reloaded.

[0569] As an embodiment, for each operation of at least one of the M operations, the first node compares the distribution characteristics of the input of this operation with the distribution characteristics of the data in the training data set of this operation to determine whether this operation needs to be retrained or redeployed.

[0570] As a sub-embodiment of the above embodiment, if the difference between the distribution characteristics of the input of this operation and the distribution characteristics of the data in the training data set of this operation is greater than a threshold, this operation needs to be retrained or redeployed.

[0571] As an example, if an operation fails to achieve expected performance or performance goals, the operation needs to be retrained or redeployed.

[0572] As an example, if an operation cannot achieve expected performance or performance goals, the operation fails.

[0573] As an embodiment, for any operation among the M operations, the first node determines by itself whether the operation can achieve the expected performance or performance target.

[0574] As an embodiment, for any operation among the M operations, the first node is indicated whether the operation can achieve expected performance or performance target.

[0575] As an embodiment, for each operation in a part of the M operations, the first node determines by itself whether the operation can achieve the expected performance or performance target, and for each operation in another part of the M operations, the first node is indicated whether the operation can achieve the expected performance or performance target.

[0576] As an embodiment, for each operation of at least one operation among the M operations, a producer of the operation indicates whether the operation can achieve expected performance or performance target.

[0577] As an embodiment, for any operation among the M operations, the producer of this operation indicates whether this operation can achieve the expected performance or performance target.

[0578] As an embodiment, the output of at least one of the M operations includes reliability, and for any operation of the at least one operation, the first node determines whether the operation can achieve expected performance or performance target based on the reliability.

[0579] As a sub-embodiment of the above embodiment, if the reliability included in the output of this operation is lower than a threshold, this operation cannot achieve the expected performance or performance target.

[0580] As a sub-embodiment of the above embodiment, the reliability indicates the credibility of other parts of the output of the corresponding operation.

[0581] As a sub-embodiment of the above embodiment, the reliability indicates the performance of the corresponding operation.

[0582] As a sub-embodiment of the above embodiment, the reliability indicates whether the corresponding operation reaches the performance target.

[0583] As an embodiment, for each operation of at least one of the M operations, the first node determines whether the operation can achieve the expected performance or performance target by monitoring the reception quality of the physical channel or physical signal generated based on the output of the operation.

[0584] As a sub-embodiment of the above embodiment, if the reception quality is lower than a threshold, the operation cannot achieve the expected performance or performance target.

[0585] As an embodiment, for each operation of at least one of the M operations, the first node determines whether the operation can achieve the expected performance or performance target by monitoring the difference between the input recovered based on the output of the operation and the actual input.

[0586] As a sub-embodiment of the above embodiment, if the difference is greater than a threshold, the operation cannot achieve the expected performance or performance target.

[0587] As an embodiment, if a timer corresponding to an operation expires, the operation becomes invalid.

[0588] As an embodiment, each operation of at least one operation among the M operations corresponds to a timer, and for any operation among the at least one operation, if the timer corresponding to this operation expires, this operation becomes invalid.

[0589] As an embodiment, the M operations correspond to M timers respectively. For any operation among the M1 operations, if the timer corresponding to this operation expires, this operation becomes invalid.

[0590] As an embodiment, the timer corresponding to at least one of the M operations is configurable.

[0591] As an embodiment, the timers corresponding to different operations are configured separately.

[0592] As an embodiment, the timers corresponding to at least two operations among the M operations are configured separately.

[0593] As an embodiment, at least two operations among the M operations correspond to the same timer.

[0594] As an embodiment, the timer corresponding to at least one operation among the M operations does not need to be configured.

[0595] As an embodiment, a timer is started or restarted after the training of the corresponding operation is completed.

[0596] As an embodiment, a timer is started or restarted after the deployment of the corresponding operation is completed.

[0597] As an embodiment, a timer is started or restarted after the training or retraining of the corresponding operation is completed.

[0598] As an embodiment, a timer is started or restarted after the deployment or redeployment of the corresponding operation is completed.

[0599] As an embodiment, a timer is started or restarted after the corresponding operation is performed.

[0600] As an embodiment, a timer is started or restarted after each output of a corresponding operation.

[0601] As an embodiment, when receiving a start indication for a timer, the first node starts or restarts the timer.

[0602] As an embodiment, if an operation is not updated within a time window, the operation fails.

[0603] As an embodiment, each operation of at least one operation among the M operations corresponds to a time window, and for any operation among the at least one operation, if the operation is not updated within the corresponding time window, the operation is invalid.

[0604] As an embodiment, updating of an operation includes: updating of an input of the operation.

[0605] As an embodiment, updating of an operation includes: updating of an output of the operation.

[0606] As an embodiment, updating of an operation includes: retraining of the operation.

[0607] As an embodiment, updating of an operation includes: redeploying of the operation.

[0608] As an embodiment, the time window corresponding to at least one of the M operations is configurable.

[0609] As an embodiment, the time windows corresponding to different operations are configured separately.

[0610] As an embodiment, the time windows corresponding to at least two operations among the M operations are configured separately.

[0611] As an embodiment, at least two operations among the M operations correspond to the same time window configuration.

[0612] As an embodiment, the time window corresponding to at least one operation among the M operations does not need to be configured.

[0613] As an embodiment, if an operation is indicated as failed, the operation fails.

[0614] Generally speaking, how the first node determines whether an operation is invalid is determined by the hardware equipment vendor. The following are some non-limiting implementation methods:

[0615] As an embodiment, the first node determines whether an operation fails by monitoring whether the operation reaches a performance target.

[0616] As an embodiment, the first node determines whether an operation achieves a performance target by monitoring the reception quality of a physical signal or a physical channel generated based on an output of the operation.

[0617] As an embodiment, if the reception quality of a physical signal or a physical channel generated based on the output of an operation is lower than a threshold, the operation fails to achieve the performance target.

[0618] As an embodiment, the output of an operation includes one or more precoding matrices, and the physical signal or physical channel generated based on the output of the operation refers to the physical signal or physical channel precoded by the one or more precoding matrices.

[0619] As an embodiment, the output of an operation includes one or more beams, and the physical signal or physical channel generated based on the output of the operation refers to a physical signal or physical channel using the one or more beams.

[0620] As an embodiment, the output of an operation includes one or more CQIs, and the one or more CQIs are used to determine one or more MCSs (Modulation and Coding Schemes). The physical signal or physical channel generated based on the output of the operation refers to a physical signal or physical channel using the one or more MCSs.

[0621] As an embodiment, the output of an operation includes reliability, and the first node determines whether the operation achieves a performance target through the reliability included in the output of the operation.

[0622] As an example, if an output of an operation includes a reliability below a threshold, the operation fails to achieve a performance goal.

[0623] As an embodiment, the first node determines whether the operation achieves the performance target by performing ML testing or ML simulation on a model of the operation.

[0624] As an example, if the performance of a model of an operation in an ML test or ML simulation fails to achieve a performance target, the operation fails to achieve a performance target.

[0625] As an embodiment, the first node restores the input of an operation based on the output of the operation, and then determines whether the operation achieves a performance target by comparing the restored input with the real input.

[0626] As an example, if the difference between the restored input and the real input of an operation is greater than a threshold, the operation fails to achieve the performance target.

[0627] As an embodiment, the first node determines whether an operation has achieved a performance target by comparing the distribution characteristics of the output of the operation in the reasoning phase with the distribution characteristics of the output obtained by taking the data in the training data set as input.

[0628] As an embodiment, if the difference between the distribution characteristics of the output obtained by an operation based on actual input and the distribution characteristics of the output obtained by taking data in the training data set as input is greater than a threshold, the operation cannot achieve the performance target.

[0629] As an embodiment, the first node determines whether an operation is invalid based on whether a timer corresponding to the operation has expired.

[0630] As an embodiment, the first node determines whether an operation is invalid according to whether the operation is updated within a time window.

[0631] As an embodiment, the first node determines whether an operation fails based on a signaling indication.

[0632] As an embodiment, the first node determines whether an operation is invalid by determining whether the operation needs to be retrained or redeployed.

[0633] Generally speaking, how the first node determines whether an operation requires retraining or redeployment is determined by the hardware equipment vendor. The following are some non-limiting implementation methods:

[0634] As an embodiment, the first node determines whether an operation needs to be retrained or redeployed by monitoring whether the operation reaches a performance target.

[0635] As an example, if an operation fails to achieve a performance target, the operation needs to be retrained or redeployed.

[0636] As an embodiment, the first node determines whether the operation needs to be retrained or redeployed by comparing the distribution characteristics of the real input of the operation with the distribution characteristics of the data in the training data set of the operation.

[0637] As an embodiment, if the difference between the distribution characteristics of the real input of an operation and the distribution characteristics of the data in the training data set of the operation is greater than a threshold, the operation needs to be retrained or redeployed.

[0638] Example 10

[0639] Embodiment 10 illustrates a schematic diagram of a first node determining M1 operations according to an embodiment of the present application; Fig.10 shown.

[0640] As an embodiment, the first node determining the M1 operations includes: the first node determining the M1 operations by itself.

[0641] As an embodiment, the first node determines the M1 operations including: the first node is instructed to perform the M1 operations.

[0642] As an embodiment, the first node determines the M1 operations, including: the first node determines a part of the M1 operations by itself, and the first node is instructed to determine another part of the M1 operations.

[0643] As an embodiment, determining the M1 operations includes determining that only the M1 operations among the M operations are not invalid.

[0644] As an embodiment, determining the M1 operations includes determining that each operation among the M operations except the M1 operations is invalid.

[0645] As an embodiment, for each operation of the M operations, the first node determines whether to execute the operation by determining whether the operation is invalid.

[0646] As an embodiment, for each operation in the M operations, the first node determines whether the operation belongs to the M1 operations by determining whether the operation is invalid.

[0647] As an embodiment, the first node determines on its own that any operation among the M operations except the M1 operations is invalid.

[0648] As an embodiment, the first node is instructed that any operation among the M operations except the M1 operations fails.

[0649] As an embodiment, the (M-M1) operations include all operations among the M operations except the M1 operation; for each operation among a part of the (M-M1) operations, the first node determines by itself that the operation is invalid; for each operation among another part of the (M-M1) operations, the first node is indicated that the operation is invalid.

[0650] As an embodiment, determining the M1 operations includes determining to execute only the M1 operations among the M operations.

[0651] As an embodiment, determining the M1 operations includes determining not to execute any operation other than the M1 operations among the M operations.

[0652] As an embodiment, the first node determines on its own to execute only the M1 operations among the M operations.

[0653] As an embodiment, the first node is instructed to perform only the M1 operations among the M operations.

[0654] As an embodiment, for a part of the M1 operations, the first node determines on its own to execute each operation in the part; for another part of the M1 operations, the first node is instructed to execute each operation in the another part.

[0655] As an embodiment, for each of the M operations, the first node determines on its own whether to execute the operation.

[0656] As an embodiment, for each of the M operations, the first node is instructed whether to execute this operation.

[0657] As an embodiment, for each operation in a part of the M operations, the first node determines by itself whether to execute this operation; for each operation in another part of the M operations, the first node is instructed whether to execute this operation.

[0658] Embodiment 11

[0659] Embodiment 11 illustrates a schematic diagram of a first node determining whether to execute each operation in M2 operations according to an embodiment of the present application; Fig.11 shown.

[0660] As an embodiment, M2 is equal to M.

[0661] As an embodiment, the M2 is equal to the M, and for each of the M operations, the first node determines on its own whether to execute the operation.

[0662] As an embodiment, the M2 is smaller than the M.

[0663] As an embodiment, the M2 is smaller than the M, and for any operation among the M operations except the M2 operations, the first node is instructed whether to execute this operation.

[0664] As a sub-embodiment of the above embodiment, the first node is instructed by downlink signaling whether to perform this operation.

[0665] As an embodiment, for any operation among the M operations, whether this operation is one of the M2 operations is configurable.

[0666] As an embodiment, for any operation among the M2 operations, if the first node determines that the operation is invalid, the first node does not perform the operation.

[0667] As an embodiment, for each operation of the M2 operations, the first node determines by itself whether the operation is invalid.

[0668] As an embodiment, for each of the M2 operations, the first node determines whether to execute the operation by determining whether the operation is invalid.

[0669] As an embodiment, the first node determines the M1 operations by itself.

[0670] As an embodiment, the first node determines a part of the M1 operations by itself, and the first node is instructed by the first signaling to perform another part of the M1 operations.

[0671] Generally speaking, for any operation in the M2 operations, how the first node determines whether to perform the operation is determined by the hardware device vendor. Some non-limiting implementation methods are described below:

[0672] As an embodiment, the first node determines whether to execute an operation by determining whether the operation fails.

[0673] As an embodiment, if one operation among the M operations fails, the first node does not execute the one operation.

[0674] As an embodiment, the first node determines whether to execute an operation by monitoring whether the operation reaches a performance target.

[0675] As an embodiment, if one operation among the M operations cannot achieve the performance target, the first node does not perform the one operation.

[0676] As an embodiment, the first node determines whether to execute an operation according to whether a timer corresponding to the operation expires.

[0677] As an embodiment, if the timer corresponding to one operation among the M operations expires, the first node does not perform the one operation.

[0678] As an embodiment, the first node determines whether to execute an operation according to whether the operation is updated within a time window.

[0679] As an embodiment, if one operation among the M operations is not updated within a time window, the first node does not execute the one operation.

[0680] As an embodiment, the first node determines whether to execute an operation according to whether the operation requires retraining.

[0681] As an embodiment, the first node determines whether to execute an operation according to whether the operation needs to be redeployed.

[0682] As an embodiment, if one operation among the M operations needs to be retrained or redeployed, the first node does not perform the one operation.

[0683] Example 12

[0684] Embodiment 12 illustrates a schematic diagram of the first signaling according to an embodiment of the present application; Fig.12 In Embodiment 12, for any operation among M3 operations among the M operations, the first signaling indicates whether the first node performs this operation, and M3 is a positive integer not greater than M.

[0685] As an embodiment, the first signaling includes higher layer signaling.

[0686] As an embodiment, the first signaling includes RRC (Radio Resource Control) signaling.

[0687] As an embodiment, the first signaling includes MAC CE (Medium Access Control layerControl Element).

[0688] As an embodiment, the first signaling includes DCI (Downlink Control Information).

[0689] As an embodiment, the first signaling includes RRC signaling and MAC CE.

[0690] As an embodiment, the first signaling includes higher layer signaling and DCI.

[0691] As an embodiment, the first signaling is RRC signaling.

[0692] As an embodiment, the first signaling is MAC CE signaling.

[0693] As an embodiment, the first signaling is DCI.

[0694] As an embodiment, M3 is equal to M.

[0695] As an embodiment, the M3 is smaller than the M.

[0696] As an embodiment, the M3 is equal to the M, and for each of the M operations, the first signaling indicates whether the first node performs this operation.

[0697] As an embodiment, the M3 is smaller than the M, and for each operation among the M operations except the M3 operations, the first node determines on its own whether to execute the operation.

[0698] As an embodiment, for each of the M3 operations, the first signaling indicates whether the first node performs the operation by indicating whether the operation is invalid.

[0699] As an embodiment, the first signaling indicates the M1 operations.

[0700] As an embodiment, the first signaling indicates a part of the M1 operations, and the first node determines another part of the M1 operations by itself.

[0701] As an embodiment, the first signaling explicitly indicates an operation that is executed among the M3 operations, and implicitly indicates an operation that is not executed among the M3 operations.

[0702] As an embodiment, the first signaling explicitly indicates the operations that are not executed among the M3 operations, and implicitly indicates the operations that are executed among the M3 operations.

[0703] As an embodiment, the first signaling indicates an operation that is not executed among the M3 operations.

[0704] As an embodiment, the first signaling indicates which operations are executed by indicating which operations are not executed.

[0705] As an embodiment, any operation among the M3 operations that is not indicated by the first signaling as not to be executed is executed.

[0706] As an embodiment, the first signaling indicates the operation performed among the M3 operations.

[0707] As an embodiment, the first signaling indicates which operations are not to be performed by indicating which operations are to be performed.

[0708] As an embodiment, any operation among the M3 operations that is not indicated by the first signaling to be executed is not executed.

[0709] As an embodiment, the first signaling explicitly indicates the failed operations among the M3 operations, and implicitly indicates the non-failed operations among the M3 operations.

[0710] As an embodiment, the first signaling explicitly indicates the operations that have not failed among the M3 operations, and implicitly indicates the operations that have failed among the M3 operations.

[0711] As an embodiment, the first signaling indicates a failed operation among the M3 operations.

[0712] As an embodiment, the first signaling indicates which operations are not invalid by indicating which operations are invalid.

[0713] As an embodiment, any operation among the M3 operations that is not indicated as invalid by the first signaling is not invalid.

[0714] As an embodiment, the first signaling indicates the operations that have not failed among the M3 operations.

[0715] As an embodiment, the first signaling indicates which operations are invalid by indicating which operations are not invalid.

[0716] As an embodiment, any operation among the M3 operations that is not indicated as not invalid by the first signaling is invalid.

[0717] Generally speaking, for any operation in the M3 operations, how the sender of the first signaling determines whether the operation is executed is determined by the hardware device vendor. Some non-limiting implementation methods are described below:

[0718] As an embodiment, the sender of the first signaling determines whether an operation is executed by determining whether the operation fails.

[0719] As an embodiment, if an operation fails, the operation is not performed.

[0720] As an embodiment, the sender of the first signaling determines whether an operation is executed by monitoring whether the operation reaches a performance target.

[0721] As an embodiment, if an operation cannot achieve a performance target, the operation is not performed.

[0722] As an embodiment, the sender of the first signaling generates and sends a first signal according to the output of an operation that is fed back, and determines whether the operation achieves a performance target according to the reception quality of the first signal fed back by the receiver of the first signal.

[0723] As a sub-embodiment of the above embodiment, if the reception quality is lower than a threshold, the operation cannot achieve a performance target.

[0724] As an embodiment, the sender of the first signaling receives the input and output of a fed-back operation, and the sender of the first signaling restores the input according to the output, and compares the restored input with the fed-back input to determine whether the operation is executed.

[0725] As a sub-embodiment of the above embodiment, if the error between the restored input and the fed-back input is greater than a threshold, the one operation is not performed.

[0726] As an embodiment, the executor of an operation restores the input of the operation based on the output of the operation, and feeds back the error between the restored input and the real input to the sender of the first signaling, and the sender of the first signaling determines whether the operation is executed based on the fed-back error.

[0727] As a sub-embodiment of the above embodiment, if the feedback error is greater than a threshold, the operation is not performed.

[0728] As an embodiment, the sender of the first signaling determines whether an operation is executed based on whether a timer corresponding to the operation expires.

[0729] As an embodiment, if a timer corresponding to an operation expires, the operation is not performed.

[0730] As an embodiment, the sender of the first signaling determines whether an operation is executed based on whether the operation is updated within a time window.

[0731] As an embodiment, if an operation is not updated within a time window, the operation is not executed.

[0732] As an embodiment, the sender of the first signaling determines whether the operation is executed by performing an ML test or ML simulation on a model of the operation.

[0733] As an example, if the ML test or ML simulation performance of a model of an operation fails to meet a performance target, the operation is not performed.

[0734] As an embodiment, the sender of the first signaling determines whether an operation is performed based on an instruction of the core network.

[0735] As an embodiment, if the core network indicates that an operation is not to be performed, the operation is not performed.

[0736] As an embodiment, the sender of the first signaling determines whether an operation is executed by determining whether the operation requires retraining or redeployment.

[0737] As an example, if an operation requires retraining or redeployment, the operation is not performed.

[0738] Embodiment 13

[0739] Embodiment 13 illustrates a schematic diagram of a first process according to an embodiment of the present application that relies on M1 operations; Fig.13 In Embodiment 13, the first processing depends on which operation or operations the M1 operations include.

[0740] As an embodiment, the benefits of the above method include providing different processing for the outputs of different operations to generate the first information block, which fully meets the needs of different operations and optimizes performance.

[0741] As an embodiment, the first processing depends on the M1 operations including, which operation or operations the first processing depends on the M1 operations.

[0742] As an embodiment, each operation in the M operations corresponds to an operation set, and the first processing includes the operations in the operation set corresponding to each operation in the M1 operations.

[0743] As an embodiment, the operation set corresponding to each operation of the M operations is configurable.

[0744] As an embodiment, the first configuration information block indicates the operation set corresponding to each of the M operations.

[0745] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and which one or more candidate operations of the plurality of candidate operations are included in the first processing depends on which one or more operations are included in the M1 operations.

[0746] As an embodiment, the first processing depends on whether the M1 operations include a first given operation among the M operations.

[0747] As an embodiment, the first given operation is which one of the M operations is configurable.

[0748] As an embodiment, the first given operation is configured by higher-layer parameters.

[0749] As an embodiment, the first configuration information block indicates the first given operation from among the M operations.

[0750] As an embodiment, which operations are included in the first processing depends on which operation or operations are included in the M1 operations.

[0751] As an embodiment, the first processing includes which one or more operations among the multiple candidate operations depend on whether the M1 operations include the first given operation.

[0752] As an embodiment, if the M1 operations include the first given operation, the first processing includes domain transformation.

[0753] As an embodiment, the first processing includes domain transformation only when the M1 operations include the first given operation.

[0754] As an embodiment, the first processing includes quantization, and the criterion or accuracy of the quantization depends on which operation or operations the M1 operations include.

[0755] As an embodiment, the first processing includes quantization, the first configuration information block indicates multiple candidate quantization precisions or multiple candidate quantization criteria, and which one or more of the multiple candidate quantization precisions or multiple candidate quantization criteria are adopted by the first processing depends on which one or more operations are included in the M1 operations.

[0756] As an embodiment, when the M1 operations include the first given operation, the quantization accuracy included in the first processing is higher than the quantization accuracy included in the first processing when the M1 operations do not include the first given operation.

[0757] As an embodiment, if the M1 operations include the first given operation, the first processing adopts scalar quantization; if the M1 operations do not include the first given operation, the first processing adopts vector quantization.

[0758] As an embodiment, each of the M operations corresponds to a quantization precision threshold, and the quantization precision of the first processing is not lower than the highest quantization precision threshold corresponding to the M1 operations.

[0759] As an embodiment, the quantization precision threshold corresponding to each operation of the M operations is configurable.

[0760] As an embodiment, the quantization accuracy threshold corresponding to each operation of the M operations is configured by a higher-layer parameter.

[0761] As an embodiment, the first configuration information block indicates that each operation of the M operations corresponds to the quantization accuracy threshold.

[0762] As an embodiment, the first processing comprises projecting a base matrix, wherein the base matrix depends on which operation or operations the M1 operations comprise.

[0763] As an embodiment, for one of the M1 inputs, a basis matrix used by the first processing to project the one input depends on which operation the one input corresponds to.

[0764] As an embodiment, the first processing includes projecting a basis matrix, the first configuration information block indicates a plurality of candidate basis matrices, and which one or more of the plurality of candidate basis matrices are adopted by the first processing depends on which one or more operations are included in the M1 operations.

[0765] As an embodiment, for one of the M1 inputs in the first processing, the first node determines a base matrix for the one input from the multiple candidate base matrices according to which operation the one input corresponds to.

[0766] As an embodiment, the first processing includes weighting the M1 inputs respectively, and the weighted coefficients depend on which operation or operations the M1 operations include.

[0767] As an embodiment, the weighted coefficient of the input corresponding to the operation among the M1 operations that belongs to the first operation set is greater than the weighted coefficient of the input corresponding to the operation among the M1 operations that does not belong to the first operation set, and the first operation set is configurable.

[0768] As an embodiment, the first configuration information block indicates the first operation set.

[0769] As an embodiment, the first operation set is configured by another higher-layer signaling different from the first configuration information block.

[0770] Embodiment 14

[0771] Embodiment 14 illustrates a schematic diagram of a first process according to an embodiment of the present application that relies on M1 operations; Fig.14 In Embodiment 14, the first processing depends on the outputs of the M1 operations.

[0772] As an embodiment, the benefits of the above method include adjusting the contribution of the outputs of the M1 operations to the first information block according to specific outputs, and flexibly adapting to various channel environments and application scenarios.

[0773] As an embodiment, the first processing being dependent on the M1 operations includes: the first processing being dependent on outputs of the M1 operations.

[0774] As an embodiment, the first processing depends on whether the output of the M1 operations includes a given type of CSI.

[0775] As an embodiment, the first processing includes which operations depend on the outputs of the M1 operations.

[0776] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and the first processing includes which one or more candidate operations of the plurality of candidate operations, depending on outputs of the M1 operations.

[0777] As an embodiment, if the outputs of the M1 operations include CSI of a given type, the first processing includes domain transformation.

[0778] As an embodiment, the first processing includes domain transformation only when the outputs of the M1 operations include CSI of a given type.

[0779] As an embodiment, the first processing includes quantization, and the criterion or accuracy of the quantization depends on the outputs of the M1 operations.

[0780] As an embodiment, when the outputs of the M1 operations include CSI of a given type, the precision of quantization included in the first processing is higher than the precision of quantization included in the first processing when the outputs of the M1 operations do not include CSI of a given type.

[0781] As an embodiment, if the outputs of the M1 operations include a given type of CSI, the first processing adopts scalar quantization; if the outputs of the M1 operations include a given type of CSI, the first processing adopts vector quantization.

[0782] As an embodiment, the output of any operation among the M operations is one of multiple candidate output types, the first configuration information block configures a minimum quantization accuracy threshold for each of the multiple candidate output types, and the quantization accuracy of the first processing is not lower than the minimum quantization accuracy threshold corresponding to the output of each operation among the M1 operations.

[0783] As an embodiment, the first processing includes projecting the M1 inputs onto a basis matrix respectively, and the basis matrix depends on the outputs of the M1 operations.

[0784] As an embodiment, for one of the M1 inputs, a basis matrix used by the first processing to project the one input depends on an output of an operation corresponding to the one input.

[0785] As an embodiment, the output of any operation among the M operations is one of multiple candidate output types, and the first configuration information block configures a basis matrix for each of the multiple candidate output types; for any operation among the M1 operations, the first processing uses the basis matrix corresponding to this operation to project the input corresponding to this operation.

[0786] As an embodiment, the first processing includes weighting the M1 inputs respectively, and the weighted coefficients depend on the outputs of the M1 operations.

[0787] As an embodiment, among the M1 inputs, the weighting coefficient of the input whose output of the corresponding operation includes the CSI of a given type is greater than the weighting coefficient of the input whose output of the corresponding operation does not include the CSI of the given type.

[0788] As an embodiment, the specific type of CSI includes PMI, compressed CSI, predicted CSI or predicted beam information.

[0789] As an embodiment, the specific type of CSI is configurable.

[0790] As an embodiment, the first configuration information block indicates the specific type of CSI.

[0791] As an embodiment, the output of one or more operations among the M1 operations includes reliability, and the first processing depends on the reliability included in the output of the one or more operations.

[0792] As an embodiment, the first processing includes which operations depend on the reliability of the output of the one or more operations.

[0793] As an embodiment, if the reliability included in the output of one of the one or more operations is lower than a reliability threshold, the first processing includes truncation; if the reliability of each of the one or more operations is not lower than the reliability threshold, the first processing does not include truncation.

[0794] As an embodiment, the reliability threshold is configurable.

[0795] As an embodiment, the first configuration information block indicates the reliability threshold.

[0796] As an embodiment, the first processing includes quantizing the M1 inputs respectively, and the quantization criterion or accuracy depends on the reliability included in the output of the one or more operations.

[0797] As an embodiment, for any one of the one or more inputs, the quantization criterion or quantization accuracy of the input by the first processing depends on the reliability of the output of the operation corresponding to the input.

[0798] As an embodiment, the quantization accuracy of any one of the one or more inputs increases as the reliability of the output of the operation corresponding to the input increases.

[0799] As an embodiment, the first configuration information block is used to determine the relationship between quantization accuracy and reliability.

[0800] As an embodiment, the first processing includes projecting at least part of the M1 inputs onto a basis matrix, and the basis matrix onto which the first processing projects one of the M1 inputs depends on the reliability included in the output of the one or more operations.

[0801] As an embodiment, the first processing includes weighting the M1 inputs respectively, and the weighted coefficients depend on the reliability included in the output of the one or more operations.

[0802] As an embodiment, for any operation among the one or more operations, the higher the reliability included in the output of this operation, the greater the coefficient used by the first processing to weight the input corresponding to this operation.

[0803] As an embodiment, the first configuration information block is used to determine the relationship between the weighting coefficient and the reliability.

[0804] Embodiment 15

[0805] Embodiment 15 illustrates a schematic diagram of a first process according to an embodiment of the present application that relies on M1 operations; Fig.15 In Embodiment 15, the first processing depends on the value of M1.

[0806] As an embodiment, the benefits of the above method include adjusting the contribution of each operation to the first information block according to the number of operations providing outputs, thereby providing better flexibility and robustness.

[0807] As an embodiment, the first processing being dependent on the M1 operations includes: the first processing being dependent on the value of M1.

[0808] As an embodiment, the first processing includes which operations depend on the value of M1.

[0809] As an embodiment, if the M1 is greater than a first threshold, the first processing includes domain transformation.

[0810] As an embodiment, the first processing includes domain transformation only when the M1 is greater than a first threshold.

[0811] As an embodiment, the first processing includes quantization, and a criterion of the quantization depends on the value of M1.

[0812] As an embodiment, when the M1 is greater than a first threshold, the quantization accuracy included in the first processing is higher than the quantization accuracy included in the first processing when the M1 is not greater than the first threshold.

[0813] As an embodiment, the first processing includes projecting all or part of the M1 inputs onto a basis matrix, respectively, and the basis matrix depends on the value of M1.

[0814] As an embodiment, the first processing includes weighting the M1 inputs respectively, and the weighted coefficients depend on the value of M1.

[0815] As an embodiment, for one of the M1 inputs, if the M1 is greater than a first threshold, the weighted coefficient of the input is a first candidate coefficient; if the M1 is less than the first threshold, the weighted coefficient of the input is a second candidate coefficient.

[0816] As an embodiment, the first threshold is configurable.

[0817] As an embodiment, the first configuration information block indicates the first threshold.

[0818] Example 16

[0819] Embodiment 16 illustrates a schematic diagram of a first process according to an embodiment of the present application that relies on M1 operations; Fig.16 In Embodiment 16, the first process depends on the reliability of the M1 operations.

[0820] As an embodiment, the benefits of the above method include avoiding the negative impact of low-reliability operations on the first information block and optimizing system performance.

[0821] As an embodiment, the first processing being dependent on the M1 operations includes: the first processing being dependent on the reliability of the M1 operations.

[0822] As an embodiment, the first processing includes which operations depend on the reliability of the M1 operations.

[0823] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and the first processing includes which one or more of the plurality of candidate operations, depending on the reliability of the M1 operations.

[0824] As an embodiment, if the reliability of one operation among the M1 operations is lower than a reliability threshold, the first processing includes truncation; if the reliability of each operation among the M1 operations is not lower than the reliability threshold, the first processing does not include truncation.

[0825] As an embodiment, the first processing includes quantizing the M1 inputs respectively, and the quantization criterion or accuracy depends on the reliability of the M1 operations.

[0826] As an embodiment, the first processing includes quantization, the first configuration information block indicates multiple candidate quantization precisions or multiple candidate quantization criteria, and which one or more of the multiple candidate quantization precisions or multiple candidate quantization criteria are adopted by the first processing depends on the reliability of the M1 operations.

[0827] As an embodiment, for any input among the M1 inputs, the quantization criterion or quantization accuracy of the input by the first processing depends on the reliability of the operation corresponding to the input.

[0828] As an embodiment, the quantization accuracy of any one of the M1 inputs increases as the reliability of the operation corresponding to the input increases.

[0829] As an embodiment, for the input corresponding to the operation with the highest reliability among the M1 operations, the first processing adopts the highest quantization accuracy; for the input corresponding to the operation with the second highest reliability among the M1 operations, the first processing adopts the second highest quantization accuracy; and so on.

[0830] As an embodiment, for the inputs corresponding to P1 operations with higher reliability among the M1 operations, the first processing adopts scalar quantization; for the inputs corresponding to the remaining (M1-P1) operations among the M1 operations, the first processing adopts vector quantization; the P1 is a positive integer less than the M1.

[0831] As an embodiment, the quantization accuracy of the input corresponding to the P1 operations with higher reliability among the M1 operations is higher than the quantization accuracy of the input corresponding to the remaining (M1-P1) operations among the M1 operations; the P1 is a positive integer less than the M1.

[0832] As an embodiment, the P1 is configurable.

[0833] As an embodiment, the first configuration information block indicates the P1.

[0834] As an embodiment, the first processing includes projecting part or all of the M1 inputs onto basis matrices respectively, and the basis matrix onto which the output of one operation among the M1 operations is projected by the first processing depends on the reliability of the one operation.

[0835] As an embodiment, the first processing includes weighting the M1 inputs respectively, and the weighted coefficients depend on the reliability of the M1 operations.

[0836] As an embodiment, the first configuration information block indicates multiple candidate weighting coefficients; the coefficient used by the first processing to weight any one of the M1 inputs is a candidate weighting coefficient among the multiple candidate weighting coefficients, and the first processing determines the weighting coefficient for each of the M1 inputs based on the reliability of the M1 operations.

[0837] As an embodiment, for any operation among the M1 operations, the higher the reliability of this operation, the greater the coefficient used by the first processing to weight the input corresponding to this operation.

[0838] As an embodiment, for the input corresponding to the operation with the highest reliability among the M1 operations, the first processing adopts the largest weighting coefficient; for the input corresponding to the operation with the second highest reliability among the M1 operations, the first processing adopts the second largest weighting coefficient; and so on.

[0839] As an embodiment, each of the M operations corresponds to a candidate coefficient list, and for any input among the M1 inputs, the coefficient used by the first processing to weight the input is a coefficient in the candidate coefficient list of the operation corresponding to the input, and the first node determines the weighting coefficient for the input from the corresponding candidate coefficient list based on the reliability of the operation corresponding to the input.

[0840] As an embodiment, the candidate coefficient list corresponding to each of the M operations is configurable.

[0841] As an embodiment, the candidate coefficient list corresponding to each operation in the M operations is configured by higher-level parameters.

[0842] As an embodiment, the first configuration information block indicates the candidate coefficient list corresponding to each of the M operations.

[0843] Generally speaking, how to determine the reliability of an operation is determined by the hardware equipment manufacturer. The following are some non-limiting implementation methods:

[0844] As an embodiment, the output of each operation of at least one operation among the M operations includes reliability.

[0845] As a sub-embodiment of the above embodiment, the reliability indicates the credibility of other parts of the output of the corresponding operation.

[0846] As a sub-embodiment of the above embodiment, the reliability indicates the performance of the corresponding operation.

[0847] As a sub-embodiment of the above embodiment, the reliability indicates whether the corresponding operation reaches the performance target.

[0848] As an embodiment, for each operation of at least one of the M operations, the first node determines the reliability of this operation by monitoring the reception quality of a physical channel or a physical signal generated based on the output of this operation.

[0849] As an embodiment, for each operation of at least one operation among the M1 operations, the first node restores the input based on the output of this operation, and compares the restored input with the real input to obtain the reliability of this operation.

[0850] As an embodiment, for each operation of at least one operation among the M1 operations, the first node compares the distribution characteristics of the input of this operation with the distribution characteristics of the data in the training data set of this operation to obtain the reliability of this operation.

[0851] As an embodiment, for each operation of at least one of the M1 operations, the first node compares the distribution characteristics of the output of this operation in the reasoning phase with the distribution characteristics of the output generated by this operation with the data in the training data set as input to obtain the reliability of this operation.

[0852] As an embodiment, for each operation of at least one operation among the M1 operations, the first node is indicated the reliability of this operation.

[0853] As an embodiment, for each operation of at least one operation among the M1 operations, the first signaling indicates the reliability of this operation.

[0854] Embodiment 17

[0855] Embodiment 17 illustrates a schematic diagram of M1 inputs and M1 coefficients according to an embodiment of the present application; Fig.17 In Embodiment 17, the first information block depends on the sum of the M1 inputs weighted by the M1 coefficients. Fig.17 In the example, the M1 inputs are respectively represented as input #0, ..., input #(M1-1); the M1 coefficients are respectively represented as coefficient #0, ..., coefficient #(M1-1).

[0856] As an embodiment, the first processing includes weighting the M1 inputs respectively with the M1 coefficients and then summing them.

[0857] As an embodiment, the M1 coefficients are respectively weighted coefficients of the M1 inputs.

[0858] As an embodiment, the M1 coefficients are respectively positive real numbers.

[0859] As an embodiment, the M1 coefficients are positive real numbers less than 1.

[0860] As an embodiment, the M1 coefficients are non-negative real numbers.

[0861] As an embodiment, the M1 coefficients are non-negative real numbers less than 1.

[0862] As an embodiment, the sum of the M1 coefficients is equal to 1.

[0863] As an embodiment, the M1 coefficients are dependent on the M1.

[0864] As an embodiment, the M1 coefficients depend on which operation or operations among the M operations the M1 operations include.

[0865] As an embodiment, the M1 coefficients depend on outputs of the M1 operations.

[0866] As an embodiment, the M1 coefficients depend on whether the output of the M1 operations includes a specific type of CSI.

[0867] As an embodiment, the M1 coefficients depend on the reliability of the M1 operations.

[0868] As an embodiment, the M1 coefficients are coefficients included in a plurality of candidate coefficient lists, the number of which is equal to all coefficients in a candidate coefficient list of M1, and the plurality of candidate coefficient lists are configurable.

[0869] As an embodiment, the multiple candidate coefficient lists are configured by higher layer parameters.

[0870] As an embodiment, the numbers of coefficients included in any two candidate coefficient lists among the multiple candidate coefficient lists are not equal.

[0871] As an embodiment, the numbers of coefficients included in the multiple candidate coefficient lists are 1, ..., M respectively.

[0872] As an embodiment, the numbers of coefficients included in the multiple candidate coefficient lists are 2, ..., M respectively.

[0873] As an embodiment, the first configuration information block indicates the multiple candidate coefficient lists.

[0874] As an embodiment, any coefficient in the multiple candidate coefficient lists is a non-negative real number.

[0875] As an embodiment, any coefficient in the multiple candidate coefficient lists is a non-negative real number less than 1.

[0876] As an embodiment, the sum of all coefficients in any candidate coefficient list of the multiple candidate coefficient lists is equal to 1.

[0877] As an embodiment, there is a candidate coefficient list among the multiple candidate coefficient lists that includes only one coefficient.

[0878] As an embodiment, there is a candidate coefficient list among the multiple candidate coefficient lists including multiple coefficients.

[0879] As an embodiment, if the M1 operations include a second given operation, the M1 coefficients belong to a first candidate coefficient list; if the M1 operations do not include the second given operation, the M1 coefficients belong to a second candidate coefficient list; the first candidate coefficient list and the second candidate coefficient list are respectively configurable; the second given operation is one of the M operations.

[0880] As an embodiment, the second given operation is which one of the M operations is configurable.

[0881] As an embodiment, the second given operation is configured by higher-layer parameters.

[0882] As an embodiment, the first configuration information block indicates the second given operation from among the M operations.

[0883] As an embodiment, any coefficient in the first candidate coefficient list and the second candidate coefficient list is a non-negative real number less than 1.

[0884] As an embodiment, the sum of all coefficients in the first candidate coefficient list is equal to 1, and the sum of all coefficients in the second candidate coefficient list is equal to 1.

[0885] As an embodiment, the first candidate coefficient list and the second candidate coefficient list are configured separately.

[0886] As an embodiment, the first candidate coefficient list and the second candidate coefficient list are configured by higher layer parameters respectively.

[0887] As an embodiment, the first configuration information block indicates at least one of the first candidate coefficient list and the second candidate coefficient column.

[0888] As an embodiment, the higher the reliability of the operation corresponding to any one of the M1 inputs, the larger the coefficient corresponding to the input.

[0889] As an embodiment, the M1 inputs are sorted in descending order according to the reliability of the corresponding operations, and the inputs that are ranked higher in the M1 inputs have larger corresponding coefficients.

[0890] As an embodiment, the first input is one of the M1 inputs, the first coefficient is a coefficient among the M1 coefficients corresponding to the first input, the first coefficient is one of multiple candidate weighting coefficients, and which one of the multiple candidate weighting coefficients the first coefficient is depends on the M1 operations.

[0891] As an embodiment, the multiple candidate weighting coefficients are configurable.

[0892] As an embodiment, the multiple candidate weighting coefficients are configured by higher layer parameters.

[0893] As an embodiment, the first configuration information block indicates the multiple candidate weighting coefficients.

[0894] As an embodiment, the multiple candidate weighting coefficients are configured by another higher layer signaling different from the first configuration information block.

[0895] As an embodiment, the multiple candidate weighting coefficients include a first candidate coefficient and a second candidate coefficient. If the M1 is greater than a first threshold, the first coefficient is the first candidate coefficient; if the M1 is less than the first threshold, the first coefficient is the second candidate coefficient.

[0896] As an embodiment, the first threshold is configurable.

[0897] As an embodiment, the first configuration information block indicates the first threshold.

[0898] As an embodiment, the first node determines the first coefficient from the multiple candidate weighting coefficients according to which operation or operations the M1 operations include.

[0899] As an embodiment, the multiple candidate weighting coefficients include 2 coefficients, if the M1 operations include a third given operation, the first coefficient is the smaller one of the 2 coefficients; if the M1 operations do not include the third given operation, the first coefficient is the larger one of the 2 coefficients; the third given operation is one of the M operations.

[0900] As an embodiment, the third given operation is an operation among the M operations that is different from the operation corresponding to the first input.

[0901] As an embodiment, the third given operation is which one of the M operations is configurable.

[0902] As an embodiment, the first configuration information block indicates the third given operation from among the M operations.

[0903] As an embodiment, the first node determines the first coefficient from the multiple candidate weighting coefficients according to the reliability of the operation corresponding to the first input.

[0904] As an embodiment, the multiple candidate weighting coefficients include 2 coefficients. If the reliability of the operation corresponding to the first input is higher than a reliability threshold, the first coefficient is the larger of the two coefficients; if the reliability of the operation corresponding to the first input is lower than the reliability threshold, the first coefficient is the smaller of the two coefficients.

[0905] As an embodiment, the M operations correspond to M candidate coefficient lists respectively, the multiple candidate weighted coefficients are coefficients in the M candidate coefficient lists and in the candidate coefficient lists corresponding to the operations corresponding to the first input, and the M candidate coefficient lists are configurable.

[0906] As an embodiment, there is a candidate coefficient list among the M candidate coefficient lists that includes only one coefficient.

[0907] As an embodiment, there is a candidate coefficient list among the M candidate coefficient lists including multiple coefficients.

[0908] As an embodiment, the M candidate coefficient lists are configured by higher layer parameters.

[0909] As an embodiment, the M candidate coefficient lists are configured separately.

[0910] As an embodiment, the first configuration information block indicates the M candidate coefficient lists.

[0911] As an embodiment, for an operation with the highest reliability among the M1 operations, the input weighting coefficient corresponding to this operation is the largest one among the multiple candidate weighting coefficients; for an operation with the second highest reliability among the M1 operations, the input weighting coefficient corresponding to this operation is the second largest one among the multiple candidate weighting coefficients; and so on.

[0912] As an embodiment, the number of candidate weighting coefficients included in the multiple candidate weighting coefficients is equal to M1.

[0913] As an embodiment, the multiple candidate coefficients are M1 largest coefficients among M candidate coefficients, and the M candidate coefficients are configurable.

[0914] As a sub-embodiment of the above embodiment, the first configuration information block indicates the M candidate coefficients.

[0915] As a sub-embodiment of the above embodiment, the M candidate coefficients are configured by another higher layer signaling different from the first configuration information block.

[0916] As an embodiment, the sum of the multiple candidate weighting coefficients is equal to 1.

[0917] As an embodiment, the multiple candidate weighting coefficients are multiple candidate coefficient lists, the number of coefficients included in which is equal to all coefficients in a candidate coefficient list of M1, and the multiple candidate coefficient lists are configurable.

[0918] Embodiment 18

[0919] Embodiment 18 illustrates a schematic diagram of M1 coefficients according to an embodiment of the present application; Fig.18 In Example 18, the M1 coefficients are respectively equal to the sum of the M1 reference coefficients divided by the M1 reference coefficients. Fig.18 In the figure, the M1 coefficients are respectively represented as coefficient #0, ..., coefficient #(M1-1); the M1 reference coefficients are respectively represented as reference coefficient #0, ..., reference coefficient #(M1-1).

[0920] As an embodiment, the M1 reference coefficients are respectively positive real numbers.

[0921] As an embodiment, the M1 reference coefficients are non-negative real numbers.

[0922] As an embodiment, the M1 reference coefficients are positive real numbers less than 1.

[0923] As an embodiment, the M1 reference coefficients are non-negative real numbers less than 1.

[0924] As an embodiment, the M operations correspond to M reference coefficients respectively, and the M1 reference coefficients are reference coefficients among the M reference coefficients that correspond to the M1 operations respectively.

[0925] As an embodiment, the M reference coefficients are configurable.

[0926] As an embodiment, the M reference coefficients are configured by higher layer parameters.

[0927] As an embodiment, the correspondence between the M reference coefficients and the M operations is configured by higher-level parameters.

[0928] As an embodiment, the first configuration information block indicates the M reference coefficients.

[0929] As an embodiment, the first configuration information block indicates the correspondence between the M reference coefficients and the M operations.

[0930] As an embodiment, the M reference coefficients are respectively positive real numbers.

[0931] As an embodiment, the M reference coefficients are respectively non-negative real numbers.

[0932] As an embodiment, the M reference coefficients are positive real numbers less than 1 respectively.

[0933] As an embodiment, the M reference coefficients are non-negative real numbers less than 1.

[0934] As an embodiment, the sum of the M reference coefficients is equal to 1.

[0935] As an embodiment, the first input is one of the M1 inputs, the first reference coefficient is a reference coefficient among the M1 reference coefficients corresponding to the first input, the first reference coefficient is one of multiple candidate weighting coefficients, and which one of the multiple candidate weighting coefficients the first reference coefficient is depends on the M1 operations.

[0936] As an embodiment, the first node determines the first reference coefficient from the multiple candidate weighting coefficients according to the value of M1.

[0937] As an embodiment, the multiple candidate weighting coefficients include a first candidate coefficient and a second candidate coefficient. If the M1 is greater than a first threshold, the first reference coefficient is the first candidate coefficient; if the M1 is less than the first threshold, the first reference coefficient is the second candidate coefficient.

[0938] As an embodiment, the first node determines the first reference coefficient from the multiple candidate weighting coefficients according to which operation or operations the M1 operations include.

[0939] As an embodiment, the multiple candidate weighting coefficients include 2 coefficients, if the M1 operations include a third given operation, the first reference coefficient is the smaller one of the 2 coefficients; if the M1 operations do not include the third given operation, the first reference coefficient is the larger one of the 2 coefficients; the third given operation is one of the M operations.

[0940] As an embodiment, the third given operation is an operation among the M operations that is different from the operation corresponding to the first input.

[0941] As an embodiment, the third given operation is which one of the M operations is configurable.

[0942] As an embodiment, the first configuration information block indicates the third given operation from among the M operations.

[0943] As an embodiment, the first node determines the first reference coefficient from the multiple candidate weighting coefficients according to the reliability of the operation corresponding to the first input.

[0944] As an embodiment, the multiple candidate weighting coefficients include 2 coefficients. If the reliability of the operation corresponding to the first input is higher than a reliability threshold, the first reference coefficient is the larger of the two coefficients; if the reliability of the operation corresponding to the first input is lower than the reliability threshold, the first reference coefficient is the smaller of the two coefficients.

[0945] As an embodiment, the M operations correspond to M candidate coefficient lists respectively, the multiple candidate weighted coefficients are coefficients in the M candidate coefficient lists and in the candidate coefficient lists corresponding to the operations corresponding to the first input, and the M candidate coefficient lists are configurable.

[0946] As an embodiment, the M candidate coefficient lists are configured by higher layer parameters.

[0947] As an embodiment, the M candidate coefficient lists are configured separately.

[0948] As an embodiment, the first configuration information block indicates the M candidate coefficient lists.

[0949] As an embodiment, for the input corresponding to an operation with the highest reliability among the M1 operations, the reference coefficient corresponding to this input is the largest one among the multiple candidate weighting coefficients; for the input corresponding to an operation with the second highest reliability among the M1 operations, the reference coefficient corresponding to this input is the second largest one among the multiple candidate weighting coefficients; and so on.

[0950] As an embodiment, the number of candidate weighting coefficients included in the multiple candidate weighting coefficients is equal to M1.

[0951] As an embodiment, the multiple candidate coefficients are M1 largest coefficients among M candidate coefficients, and the M candidate coefficients are configurable.

[0952] As a sub-embodiment of the above embodiment, the first configuration information block indicates the M candidate coefficients.

[0953] As a sub-embodiment of the above embodiment, the M candidate coefficients are configured by another higher layer signaling different from the first configuration information block.

[0954] As an embodiment, the sum of the multiple candidate weighting coefficients is equal to 1.

[0955] As an embodiment, the multiple candidate weighting coefficients are multiple candidate coefficient lists, the number of coefficients included in which is equal to all coefficients in a candidate coefficient list of M1, and the multiple candidate coefficient lists are configurable.

[0956] Embodiment 19

[0957] Example 19 illustrates a schematic diagram of M1 coefficients according to an embodiment of the present application; Fig.19 As shown. In Example 19, the M operations include a first candidate operation, a second candidate operation and a third candidate operation; the first candidate operation, the second candidate operation and the third candidate operation correspond to a first coefficient list, a second coefficient list and a third coefficient list respectively; the first coefficient list includes only one coefficient equal to 1 / 2, the second coefficient list includes two coefficients of 1 / 2 and 1 / 4, and the third coefficient list includes two coefficients of 1 / 2 and 1 / 4; the M1 coefficients are respectively a coefficient in the coefficient list corresponding to the corresponding operation.

[0958] As an embodiment, if the M1 operations include the first candidate operation, the second candidate operation and the third candidate operation, the M1 coefficients are respectively equal to 1 / 2, 1 / 4 and 1 / 4.

[0959] As an embodiment, if the M1 operations only include the first candidate operation and the second candidate operation, the M1 coefficients are equal to 1 / 2 and 1 / 2 respectively.

[0960] As an embodiment, if the M1 operations only include the first candidate operation and the third candidate operation, the M1 coefficients are equal to 1 / 2 and 1 / 2 respectively.

[0961] As an embodiment, if the M1 operations only include the second candidate operation and the third candidate operation, the M1 coefficients are equal to 1 / 2 and 1 / 2 respectively.

[0962] As an embodiment, the M1 coefficients depend on the value of M1.

[0963] As an embodiment, the M1 coefficients depend on which operation or operations the M1 operations include.

[0964] As an embodiment, the first processing depends on the value of M1.

[0965] As an embodiment, the first processing depends on which operation or operations the M1 operations include.

[0966] Embodiment 20

[0967] Embodiment 20 illustrates a schematic diagram of M1 coefficients according to an embodiment of the present application; Fig. 20 As shown. In embodiment 20, the M operations include a first candidate operation and a second candidate operation, and M is equal to 2; the weighting coefficients corresponding to the first candidate operation and the second candidate operation are both one of the M candidate weighting coefficients, and the M candidate weighting coefficients include 3 / 4 and 1 / 4; the M1 operation includes the first candidate operation and the second candidate operation, and the M1 input includes a first input and a second input, the first input depends on the output of the first candidate operation, and the second input depends on the output of the second candidate operation; if the reliability of the first candidate operation is higher than the reliability of the second candidate operation, the weighting coefficient of the first input is equal to 3 / 4 and the weighting coefficient of the second input is equal to 1 / 4; if the reliability of the first candidate operation is lower than the reliability of the second candidate operation, the weighting coefficient of the first input is equal to 1 / 4 and the weighting coefficient of the second input is equal to 3 / 4.

[0968] As an embodiment, the M1 coefficients depend on the outputs of the M operations.

[0969] As an embodiment, the M1 coefficients depend on the reliability of the M operations.

[0970] As an embodiment, the first processing depends on the outputs of the M operations.

[0971] As an embodiment, the first processing depends on the reliability of the M operations.

[0972] As an embodiment, the output of the first candidate operation includes reliability.

[0973] As an embodiment, the output of the second candidate operation includes reliability.

[0974] As an embodiment, the first node is indicated the reliability of at least one of the first candidate operation and the second candidate operation.

[0975] As an embodiment, the first node is indicated by downlink signaling on the reliability of at least one of the first candidate operation and the second candidate operation.

[0976] As an embodiment, the first node determines the reliability of the first candidate operation by monitoring the reception quality of a physical channel or a physical signal generated based on the output of the first candidate operation.

[0977] As an embodiment, the first node determines the reliability of the second candidate operation by monitoring the reception quality of a physical channel or a physical signal generated based on the output of the second candidate operation.

[0978] As an embodiment, the first node restores input based on the output of the first candidate operation, and compares the restored input with the real input to obtain the reliability of the first candidate operation.

[0979] As an embodiment, the first node compares the distribution characteristics of the input of the first candidate operation with the distribution characteristics of the data in the training data set of the first candidate operation to obtain the reliability of the first candidate operation.

[0980] As an embodiment, the first node compares the distribution characteristics of the output of the first candidate operation in the reasoning phase with the distribution characteristics of the output generated by the first candidate operation with the data in the training data set as input to obtain the reliability of the first candidate operation.

[0981] As an embodiment, the first node restores the input based on the output of the second candidate operation, and compares the restored input with the real input to obtain the reliability of the second candidate operation.

[0982] As an embodiment, the first node compares the distribution characteristics of the input of the second candidate operation with the distribution characteristics of the data in the training data set of the second candidate operation to obtain the reliability of the second candidate operation.

[0983] As an embodiment, the first node compares the distribution characteristics of the output of the second candidate operation in the reasoning phase with the distribution characteristics of the output generated by the second candidate operation using the data in the training data set as input to obtain the reliability of the second candidate operation.

[0984] Embodiment 21

[0985] Embodiment 21 illustrates a schematic diagram of a first compressed CSI and a first restored CSI according to an embodiment of the present application; Fig.21 In Embodiment 21, the first information block includes a first compressed CSI, and the first compressed CSI is used as an input of a first operation to generate a first restored CSI.

[0986] As an embodiment, the first compressed CSI includes compressed PMI.

[0987] As an embodiment, the first compressed CSI includes one or more of compressed CRI, compressed SSBRI, compressed RSRP and compressed SINR.

[0988] As an embodiment, the first compressed CSI is used to determine a precoding matrix.

[0989] As an embodiment, the first compressed CSI is used to determine one or more of CRI, SSBRI, RSRP and SINR.

[0990] As an embodiment, the first compressed CSI includes compressed predicted CSI.

[0991] As an embodiment, the first compressed CSI includes one or more of compressed predicted PMI, compressed predicted CQI and compressed predicted RI.

[0992] As an embodiment, the first compressed CSI includes one or more of compressed predicted CRI, compressed predicted SSBRI, compressed predicted RSRP and compressed predicted SINR.

[0993] As an embodiment, the first recovered CSI includes recovering the outputs of the M1 operations.

[0994] As an embodiment, the first recovered CSI includes recovery of the output of at least one operation among the M1 operations.

[0995] As an embodiment, the first recovery of CSI includes recovery of inputs of the M1 operations.

[0996] As an embodiment, the first recovered CSI includes recovery of an input of at least one of the M1 operations.

[0997] As an embodiment, the first recovered CSI includes decompressed CSI.

[0998] As an embodiment, the first recovered CSI includes a precoding matrix.

[0999] As an embodiment, the first recovered CSI includes PMI.

[1000] As an embodiment, the first recovered CSI includes one or more of PMI, CQI and RI.

[1001] As an embodiment, the first recovered CSI includes one or more of CRI, SSBRI, RSRP and SINR.

[1002] As an embodiment, the first recovered CSI includes one or more of PMI, CRI, CQI, RI, LI, SSBRI, RSRP, SINR, capability index and TDCP.

[1003] As an embodiment, the first recovered CSI includes one or more of predicted PMI, predicted CQI and predicted RI.

[1004] As an embodiment, the first recovered CSI includes one or more of predicted CRI, predicted SSBRI, predicted RSRP and predicted SINR.

[1005] As an embodiment, the first recovered CSI includes a channel matrix.

[1006] As an embodiment, the first recovered CSI includes a feature vector.

[1007] As an embodiment, the first operation includes an inverse operation of at least one of the M operations.

[1008] As an embodiment, the first operation includes an inverse operation of at least one of the M1 operations.

[1009] As an embodiment, the first operation is based on training.

[1010] As an embodiment, the training of the first operation is performed by a sender of the first configuration information block.

[1011] As an embodiment, the training of the first operation is performed by a core network.

[1012] As an embodiment, the first operation and one or more operations among the M operations are obtained through different training.

[1013] As an embodiment, the first operation and any one of the M operations are obtained through different training.

[1014] As an embodiment, the benefits of the above method include: saving air interface overhead, having better flexibility, being adaptable to different terminals, and having better forward compatibility.

[1015] As an embodiment, the first operation and one or more operations among the M operations are obtained through joint training.

[1016] As an embodiment, the first operation and any one of the M operations are obtained through joint training.

[1017] As an embodiment, the benefits of the above method include: optimizing performance.

[1018] As an embodiment, the training of the first operation depends on the training of one or more operations among the M operations.

[1019] As an embodiment, the training of one or more operations among the M operations depends on the training of the first operation.

[1020] As an embodiment, the benefits of the above method include: achieving a better balance between performance and complexity.

[1021] As an embodiment, the first operation requires deployment.

[1022] As an embodiment, the first operation is obtained by loading.

[1023] As an embodiment, the first operation includes inference.

[1024] As an embodiment, the first operation includes AI inference.

[1025] As an embodiment, the first operation includes AI inference for CSI.

[1026] As an embodiment, the first operation is AI inference for CSI recovery.

[1027] As an embodiment, the first operation is AI inference for CSI decompression.

[1028] As an embodiment, the first operation is performed by an AI entity deployed on the second node.

[1029] As an embodiment, the first operation is performed by an AI function deployed on the second node.

[1030] As an embodiment, the first operation is based on artificial intelligence or machine learning.

[1031] As an embodiment, the first operation is based on a neural network.

[1032] As an embodiment, the first operation includes a decoder for CSI compression based on a neural network or CNN.

[1033] Embodiment 22

[1034] Embodiment 22 illustrates a schematic diagram of a processing system based on artificial intelligence or machine learning according to an embodiment of the present application; Fig. 22 In Example 22, the third processor sends a first data set to the fourth processor and sends a second data set to the fifth processor; the fourth processor generates a target first-category parameter group based on the first data set, and the fourth processor sends the generated target first-category parameter group to the fifth processor; the fifth processor processes the second data set using the target first-category parameter group to obtain a first-category output, and the fifth processor sends the first-category output to the sixth processor. Fig. 22 In the embodiment, the first type of feedback and the second type of feedback are optional; the fourth processor includes ML training function; and the fifth processor includes ML reasoning function.

[1035] As an embodiment, the sixth processor includes ML testing functionality.

[1036] As an embodiment, the sixth processor includes performance monitoring / evaluation of the ML model.

[1037] As an embodiment, the fifth processor sends the first type of feedback to the fourth processor, and the first type of feedback is used to trigger recalculation or updating of the target first type parameter group, that is, triggering ML initial training or ML retraining.

[1038] As an embodiment, the sixth processor sends a second type of feedback to the third processor, and the second type of feedback is used to generate the first data set or the second data set, or the second type of feedback is used to trigger the sending of the first data set or the sending of the second data set.

[1039] As an embodiment, the third processor generates the first data set and the second data set based on measurement of a reference signal.

[1040] As an embodiment, the fifth processor belongs to the first node, and the sixth processor belongs to the second node.

[1041] As an embodiment, the fifth processor performs the M1 operations.

[1042] As an embodiment, the fifth processor performs one of the M1 operations.

[1043] As an embodiment, the sixth processor performs the first operation.

[1044] As an embodiment, the first type of output includes the first information block.

[1045] As an embodiment, the first type of output includes the first compressed CSI.

[1046] As an embodiment, the second data set includes measurements for a reference signal.

[1047] As an embodiment, the first data set includes training data (Training Data).

[1048] As an embodiment, the fourth processor is used to train the ML model, and the trained model is described by the target first-category parameter group.

[1049] As an embodiment, the fourth processor belongs to the first node.

[1050] The above embodiment avoids transferring the first data set to the second node.

[1051] As an embodiment, the fourth processor belongs to the second node.

[1052] The above embodiments support joint training and optimize system performance.

[1053] As an embodiment, the fourth processor belongs to the core network.

[1054] The above embodiments support joint training of the entire network, further optimizing system performance.

[1055] As an embodiment, the second data set includes inference data.

[1056] As an embodiment, the fifth processor belongs to the first node.

[1057] As an embodiment, the fifth processing machine constructs a model according to the target first-category parameter group, and then inputs the second data set into the constructed model to obtain the first-category output.

[1058] As an embodiment, the fifth processing machine generates a restored data set according to the first type of output, and an error between the restored data set and the second data set is used to generate the first type of feedback.

[1059] As an embodiment, the first type of feedback is used to reflect the performance of the trained model; when the performance of the trained model cannot meet the requirements, the fourth processor will recalculate the target first type parameter group.

[1060] As an embodiment, when the fourth processing opportunity needs to recalculate the target first-category parameter group, the corresponding operation is considered invalid.

[1061] As an embodiment, when the error is too large or has not been updated for too long, the performance of the trained model is considered to be unable to meet the requirements.

[1062] As an embodiment, when the error is too large or has not been updated for too long, the corresponding operation is considered invalid.

[1063] As an embodiment, the target first category parameter group includes: one or more of: convolution kernel size, number of convolution layers, convolution step size, pooling kernel size, pooling kernel step size, pooling function, activation function, or number of feature maps.

[1064] As an embodiment, the target first-category parameter group includes: one or more of a convolution kernel, a pooling kernel, a pooling function, an activation function, parameters of a pooling function, or parameters of an activation function.

[1065] As an embodiment, the ML includes AI.

[1066] As an embodiment, the ML includes ML and AI.

[1067] Embodiment 23

[1068] Embodiment 23 illustrates a schematic diagram based on artificial intelligence or machine learning according to an embodiment of the present application; Fig.23 Attached Fig.23 Including the second operation, the third operation, the fourth operation, the fifth operation and the sixth operation. In Example 23, the second operation and the third operation belong to the first stage, the fourth operation belongs to the second stage, the fifth operation belongs to the third stage, and the sixth operation belongs to the fourth stage. Fig.23 In the diagram, the lines with arrows indicate the sequence of the processes.

[1069] As an embodiment, the second operation includes ML training, the third operation includes ML testing, the fourth operation includes ML emulation, the fifth operation includes ML entity loading, and the sixth operation includes AI inference.

[1070] As an embodiment, the first phase includes a training phase, the second phase includes an emulation phase, the third phase includes a deployment phase, and the fourth phase includes an emulation phase.

[1071] As an embodiment, the first stage includes ML model training.

[1072] As an embodiment, the first stage includes ML model training and ML testing.

[1073] As an embodiment, the ML model training includes initial training and re-training of one or a group of ML models.

[1074] As an embodiment, the ML model training depends on training data.

[1075] As an embodiment, the ML model training includes ML entity validation.

[1076] As an embodiment, the ML entity validation is used to evaluate the performance of the ML entity.

[1077] As an embodiment, the ML entity verification relies on verification data.

[1078] As an example, if the result of ML entity verification does not meet expectations, the ML model will be retrained.

[1079] As an embodiment, the ML testing includes testing the verified ML entity to estimate the performance of the trained ML model.

[1080] As an example, if the result of the ML test meets expectations, the ML entity proceeds to the next stage; otherwise the ML model will be retrained.

[1081] As an embodiment, the ML test relies on test data.

[1082] As an embodiment, the second stage includes ML simulation, which performs inference of ML entities in a simulation environment.

[1083] As an embodiment, the ML simulation estimates the performance of ML entity reasoning in a simulation environment before using the ML entity.

[1084] As an embodiment, the second stage is optional.

[1085] As an embodiment, the third stage includes ML entity loading, and the ML entity loading is to obtain the trained ML entity to obtain the desired AI reasoning function.

[1086] As an embodiment, the third stage is optional.

[1087] As an example, when the training function and the inference function are co-located, the third stage is no longer required.

[1088] As an embodiment, the fourth stage includes AI reasoning.

[1089] As an embodiment, the ML includes AI.

[1090] As an embodiment, the AI ​​includes ML.

[1091] Embodiment 24

[1092] Embodiment 24 illustrates a schematic diagram of AI function deployment according to an embodiment of the present application; Fig.24 shown.

[1093] In Example 24, the AI ​​training function of the RAN (Radio Access Network) domain is located in the 3GPP RAN domain-specific management function, and the AI ​​inference function is located in the UE.

[1094] In embodiment 24, the RAN domain-specific management function provides AI training function management capabilities and AI inference function management capabilities.

[1095] Embodiment 25

[1096] Embodiment 25 illustrates a schematic diagram of AI function deployment according to an embodiment of the present application; Fig.25 shown.

[1097] In embodiment 25, the AI ​​training function is located in the RAN domain-specific management function, and the AI ​​inference function is located locally in the UE.

[1098] In embodiment 25, the management capability of the AI ​​training function is provided by a RAN domain-specific management function, and the management capability of the AI ​​inference is provided locally by the UE.

[1099] In the attached Fig.25 In this article, MnF refers to Management Function.

[1100] Embodiment 26

[1101] Embodiment 26 illustrates a schematic diagram of AI function deployment according to an embodiment of the present application; as shown in the attached Fig.26 shown.

[1102] In Embodiment 26, both the AI ​​training function and the AI ​​inference function are located in the UE, wherein the UE provides training and inference capabilities.

[1103] In Embodiment 26, the RAN domain-specific management function provides management capabilities for AI training functions and management capabilities for AI inference functions.

[1104] Embodiment 27

[1105] Embodiment 27 illustrates a schematic diagram of AI function deployment according to an embodiment of the present application; Fig. 27 shown.

[1106] In Embodiment 27, both the AI ​​training function and the AI ​​inference function are located in the UE.

[1107] In Embodiment 27, the management capability of the AI ​​training function and the management capability of the AI ​​inference function are both provided locally by the UE.

[1108] In the attached Fig. 27 In this article, MnF refers to Management Function.

[1109] Embodiment 28

[1110] Embodiment 28 illustrates a structural block diagram of a processing device used in a first node according to an embodiment of the present application; Fig.28 As shown in the attached Fig.28 In the embodiment, the processing device 2800 in the first node includes a first receiver 2801 and a first processor 2802.

[1111] In Embodiment 28, the first receiver 2801 receives a first configuration information block; the first processor 2802 performs M1 operations among M operations and generates a first information block.

[1112] In Example 28, the first configuration information block indicates the M operations, where M is a positive integer greater than 1, and M1 is a positive integer not greater than M; the first information block depends on the output of the first processing, and the input of the first processing depends on the output of the M1 operations; and the first processing depends on the M1 operations.

[1113] As an embodiment, each of the M operations is training-based.

[1114] As an embodiment, each of the M operations includes inference.

[1115] As an embodiment, the first node performs only the M1 operations among the M operations.

[1116] As an embodiment, the input of at least one of the M operations includes measurements obtained based on CSI resources, or the input of at least one of the M operations includes the reception quality of at least one physical channel or physical signal.

[1117] As an embodiment, the input of at least one of the M operations includes one or more of positioning information, ACK / NACK ratio, NACK number, beam failure number and wireless link failure number.

[1118] As an embodiment, the first information block includes channel information.

[1119] As an embodiment, the first information block includes CSI, and the first configuration information block includes CSI reporting configuration.

[1120] As a sub-embodiment of the above embodiment, the first information block includes a CSI report for the first configuration information block.

[1121] As an embodiment, the first processing includes computing the outputs of the M1 operations to generate the first information block; or, the first processing includes weighting the M1 inputs respectively and then merging them.

[1122] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and the first processing includes which one or more candidate operations of the plurality of candidate operations, depending on the M1 operations.

[1123] As an embodiment, the first processing includes weighting, the first configuration information block indicates a plurality of candidate weighting coefficients, and which one or more of the plurality of candidate weighting coefficients are adopted by the first processing depends on the M1 operations.

[1124] As an embodiment, the first processing includes quantization, the first configuration information block indicates multiple candidate quantization precisions or multiple candidate quantization criteria, and which one or more of the multiple candidate quantization precisions or multiple candidate quantization criteria are adopted by the first processing depends on the M1 operations.

[1125] As an embodiment, the first processor 2802 deploys M0 operations among the M operations, where M0 is a positive integer not greater than M.

[1126] As an embodiment, the first configuration information block indicates M identifiers, and the M operations are respectively associated with the M identifiers.

[1127] As an embodiment, the M operations are respectively identified by the M identifiers.

[1128] As an embodiment, any operation among the M operations except the M1 operations is invalid.

[1129] As an embodiment, if any operation among the M operations fails, the first node does not execute this operation.

[1130] As an embodiment, for any operation among the M operations, if one event in a first event set occurs, the operation fails; the first event set includes one or more of the operation requiring retraining, the operation requiring redeployment or reloading, the operation failing to achieve expected performance or performance goals, the timer corresponding to the operation expiring, the operation not being updated within a time window, and the operation being indicated as failed.

[1131] As an embodiment, the first processor 2802 determines the M1 operations.

[1132] As an embodiment, for any operation among M2 operations among the M operations, the first node determines on its own whether to execute this operation, and M2 is a positive integer not greater than M.

[1133] As an embodiment, the first receiver 2801 receives a first signaling; wherein, for any operation among M3 operations among the M operations, the first signaling indicates whether the first node performs this operation, and the M3 is a positive integer not greater than the M.

[1134] As an embodiment, the first processing depends on the reliability of the M1 operations.

[1135] As an embodiment, the input of the first processing includes M1 inputs, the M1 inputs are respectively dependent on the outputs of the M1 operations, and the first processing uses M1 coefficients to weight the M1 inputs respectively, and the M1 coefficients are dependent on the M1 operations.

[1136] As an embodiment, the first processor 2802 sends the first information block.

[1137] As an embodiment, the first node is a terminal.

[1138] As an embodiment, the first node is user equipment.

[1139] As an embodiment, the first node is a relay node device.

[1140] As an embodiment, the first receiver 2801 includes at least one of {antenna 452, receiver 454, receiving processor 456, multi-antenna receiving processor 458, controller / processor 459, memory 460, data source 467} in Embodiment 4.

[1141] As an embodiment, the first processor 2802 includes at least one of {antenna 452, transmitter 454, transmit processor 468, multi-antenna transmit processor 457, controller / processor 459, memory 460, data source 467} in Embodiment 4.

[1142] Embodiment 29

[1143] Embodiment 29 illustrates a structural block diagram of a processing device used in a second node according to an embodiment of the present application; Fig.29 As shown in the attached Fig.29 In the example, the processing device 2900 in the second node includes a second processor 2901.

[1144] In Embodiment 29, the second processor 2901 sends a first configuration information block.

[1145] In Example 29, the first configuration information block indicates M operations, where M is a positive integer greater than 1; the target recipient of the first configuration information block determines the first information block, the first information block depends on the output of a first process, the input of the first process depends on the output of M1 operations, and each of the M1 operations is one of the M operations; the target recipient of the first configuration information block performs the M1 operations, where M1 is a positive integer not greater than M; and the first process depends on the M1 operations.

[1146] As an embodiment, each of the M operations is training-based.

[1147] As an embodiment, each of the M operations includes inference.

[1148] As an embodiment, the target receiver of the first configuration information block performs only the M1 operations among the M operations.

[1149] As an embodiment, the input of at least one of the M operations includes measurements obtained based on CSI resources, or the input of at least one of the M operations includes the reception quality of at least one physical channel or physical signal.

[1150] As an embodiment, the input of at least one of the M operations includes one or more of positioning information, ACK / NACK ratio, NACK number, beam failure number and wireless link failure number.

[1151] As an embodiment, the first information block includes channel information.

[1152] As an embodiment, the first information block includes CSI, and the first configuration information block includes CSI reporting configuration.

[1153] As a sub-embodiment of the above embodiment, the first information block includes a CSI report for the first configuration information block.

[1154] As an embodiment, the first processing includes subjecting the outputs of the M1 operations to mathematical operations or matrix operations to generate the first information block; or, the first processing includes weighting the M1 inputs respectively and then merging them.

[1155] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and the first processing includes which one or more candidate operations of the plurality of candidate operations, depending on the M1 operations.

[1156] As an embodiment, the first processing includes weighting, the first configuration information block indicates a plurality of candidate weighting coefficients, and which one or more of the plurality of candidate weighting coefficients are adopted by the first processing depends on the M1 operations.

[1157] As an embodiment, the first processing includes quantization, the first configuration information block indicates multiple candidate quantization precisions or multiple candidate quantization criteria, and which one or more of the multiple candidate quantization precisions or multiple candidate quantization criteria are adopted by the first processing depends on the M1 operations.

[1158] As an embodiment, the second processor 2901 receives the first information block.

[1159] As an embodiment, the first configuration information block indicates M identifiers, and the M operations are respectively associated with the M identifiers.

[1160] As an embodiment, the M operations are respectively identified by the M identifiers.

[1161] As an embodiment, any operation among the M operations except the M1 operations is invalid.

[1162] As an embodiment, if any operation among the M operations fails, the target receiver of the first configuration information block does not perform this operation.

[1163] As an embodiment, for any operation among the M operations, if one event in a first event set occurs, the operation fails; the first event set includes one or more of the operation requiring retraining, the operation requiring redeployment or reloading, the operation failing to achieve expected performance or performance goals, the timer corresponding to the operation expiring, the operation not being updated within a time window, and the operation being indicated as failed.

[1164] As an embodiment, the target receiver of the first configuration information block determines the M1 operations.

[1165] As an embodiment, for any operation among M2 operations among the M operations, the target receiver of the first configuration information block determines whether to perform this operation by itself, and M2 is a positive integer not greater than M.

[1166] As an embodiment, the second processor 2901 sends a first signaling; wherein, for any operation of M3 operations among the M operations, the first signaling indicates whether the target receiver of the first configuration information block performs this operation, and the M3 is a positive integer not greater than the M.

[1167] As an embodiment, the first processing depends on the reliability of the M1 operations.

[1168] As an embodiment, the input of the first processing includes M1 inputs, the M1 inputs are respectively dependent on the outputs of the M1 operations, and the first processing uses M1 coefficients to weight the M1 inputs respectively, and the M1 coefficients are dependent on the M1 operations.

[1169] As an embodiment, the second processor 2901 performs a first operation; wherein the first information block includes a first compressed CSI, and the first compressed CSI is used as an input of the first operation to generate a first recovered CSI.

[1170] As an embodiment, the second processor 2901 deploys the first operation.

[1171] As an embodiment, the second node is a base station.

[1172] As an embodiment, the second node is a base station device.

[1173] As an embodiment, the second node is user equipment.

[1174] As an embodiment, the second node is a relay node device.

[1175] As an embodiment, the second processor 2901 includes at least one of {antenna 420, receiver / transmitter 418, receiving processor 470, transmitting processor 416, multi-antenna receiving processor 472, multi-antenna transmitting processor 471, controller / processor 475, memory 476} in Example 4.

[1176] Embodiment 30

[1177] Embodiment 30 illustrates a schematic diagram of a first process according to an embodiment of the present application that relies on M1 operations; Fig.30 In Embodiment 30, the M operations correspond to M operation sets respectively, and the first processing includes the operations in the operation set corresponding to each operation in the M1 operations. Fig.30 In the example, the M operations are respectively represented as operation #0, ..., operation #(M-1); the M operation sets are respectively represented as operation set #0, ..., operation set #(M-1).

[1178] As an embodiment, the operation set corresponding to each operation of the M operations is configurable.

[1179] As an embodiment, the first configuration information block indicates the operation set corresponding to each of the M operations.

[1180] As an embodiment, the first configuration information block indicates a plurality of candidate operations, and which one or more candidate operations of the plurality of candidate operations are included in the first processing depends on which one or more operations are included in the M1 operations.

[1181] Embodiment 31

[1182] Embodiment 31 illustrates a schematic diagram of a first process according to an embodiment of the present application that relies on M1 operations; Fig.31 In Example 31, for inputs corresponding to P1 operations with higher reliability among the M1 operations, the first processing adopts a first quantization precision; for inputs corresponding to the remaining (M1-P1) operations among the M1 operations, the first processing adopts a second quantization precision; the first quantization precision and the second quantization precision are configurable; and P1 is a positive integer less than M1.

[1183] As an embodiment, the first quantization accuracy is higher than the second quantization accuracy.

[1184] As an embodiment, the first quantization precision and the second quantization precision are respectively configured by higher-layer parameters.

[1185] As an embodiment, the first configuration information block indicates at least one of the first quantization precision and the second quantization precision.

[1186] As an embodiment, the first configuration information block indicates the first quantization precision and the second quantization precision.

[1187] As an embodiment, the first quantization precision and the second quantization precision are respectively configured by higher layer signaling different from the searched first configuration information block.

[1188] As an embodiment, the P1 is configurable.

[1189] As an embodiment, the first configuration information block indicates the P1.

[1190] A person of ordinary skill in the art can understand that all or part of the steps in the above method can be completed by instructing the relevant hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a hard disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module unit in the above embodiment can be implemented in the form of hardware or in the form of a software function module, and the present application is not limited to any specific form of software and hardware combination. The user equipment, terminal and UE in the present application include but are not limited to drones, communication modules on drones, remote-controlled aircraft, aircraft, small aircraft, mobile phones, tablet computers, notebooks, vehicle-mounted communication equipment, transportation tools, vehicles, RSUs, wireless sensors, Internet cards, Internet of Things terminals, RFID terminals, NB-IOT terminals, MTC (Machine Type Communication, Machine Type Communication) terminals, eMTC (enhanced MTC, enhanced MTC) terminals, data cards, Internet cards, vehicle-mounted communication equipment, low-cost mobile phones, low-cost tablet computers and other wireless communication devices. The base stations or system equipment in this application include but are not limited to macrocell base stations, microcell base stations, small cell base stations, home base stations, relay base stations, eNB, gNB, TRP (Transmitter Receiver Point), GNSS, relay satellites, satellite base stations, aerial base stations, RSU (Road Side Unit), drones, test equipment, such as wireless communication equipment such as transceivers or signaling testers that simulate some functions of base stations.

[1191] It should be understood by those skilled in the art that the present invention may be implemented in other specified forms without departing from its core or essential features. Therefore, the embodiments disclosed herein should be considered illustrative rather than restrictive in any way. The scope of the invention is determined by the appended claims rather than the preceding description, and all modifications within their equivalent meanings and regions are considered to be included therein.

Claims

1. A method used in a terminal for wireless communication, characterized in that: include: Receive a first configuration information block, where the first configuration information block indicates M operations, where M is a positive integer greater than 1; Execute M1 operations among the M operations, where M1 is a positive integer not greater than M; Generate a first information block, wherein the first information block depends on an output of a first process, and an input of the first process depends on outputs of the M1 operations; The first processing depends on the M1 operations.

2. The method according to claim 1, characterized in that include: Deploy M0 operations among the M operations, where M0 is a positive integer not greater than the M.

3. The method according to claim 1 or 2, characterized in that: The first configuration information block indicates M identifiers, and the M operations are respectively associated with the M identifiers.

4. The method according to any one of claims 1 to 3, characterized in that: Any operation among the M operations except the M1 operations is invalid.

5. The method according to any one of claims 1 to 4, characterized in that: include: The M1 operations are determined.

6. The method according to any one of claims 1 to 5, characterized in that: For any operation among M2 operations among the M operations, the first node determines on its own whether to execute this operation, and M2 is a positive integer not greater than M.

7. The method according to any one of claims 1 to 6, characterized in that: include: receiving a first signaling; For any operation among M3 operations among the M operations, the first signaling indicates whether the first node performs this operation, and M3 is a positive integer not greater than M.

8. The method according to any one of claims 1 to 7, characterized in that: The first process relies on the reliability of the M1 operations.

9. The method according to any one of claims 1 to 8, characterized in that: The input of the first process includes M1 inputs, the M1 inputs are respectively dependent on the outputs of the M1 operations, and the first process weights the M1 inputs respectively with M1 coefficients, the M1 coefficients are dependent on the M1 operations.

10. The method according to any one of claims 1 to 9, characterized in that: include: The first information block is sent.

11. A terminal, characterized in that: The terminal includes: one or more processors and a memory; The memory is coupled to the one or more processors, and the memory is used to store computer program codes, where the computer program codes include computer instructions. The one or more processors call the computer instructions to enable the terminal to execute the method according to any one of claims 1 to 10.

12. A method used in a base station for wireless communication, characterized in that: include: Sending a first configuration information block, where the first configuration information block indicates M operations, where M is a positive integer greater than 1; Among them, the target recipient of the first configuration information block determines the first information block, the first information block depends on the output of the first processing, the input of the first processing depends on the output of M1 operations, each of the M1 operations is one of the M operations; the target recipient of the first configuration information block performs the M1 operations, and M1 is a positive integer not greater than M; the first processing depends on the M1 operations.

13. The method according to claim 12, characterized in that include: The first information block is received.

14. The method according to claim 12 or 13, characterized in that The first configuration information block indicates M identifiers, and the M operations are respectively associated with the M identifiers.

15. The method according to any one of claims 12 to 14, characterized in that Any operation among the M operations except the M1 operations is invalid.

16. The method according to any one of claims 12 to 15, characterized in that The target recipient of the first configuration information block determines the M1 operations.

17. The method according to any one of claims 12 to 16, characterized in that For any operation among M2 operations among the M operations, the target receiver of the first configuration information block determines whether to execute this operation by itself, and M2 is a positive integer not greater than M.

18. The method according to any one of claims 12 to 17, characterized in that include: Sending a first signaling; Among them, for any operation among M3 operations among the M operations, the first signaling indicates whether the target receiver of the first configuration information block performs this operation, and the M3 is a positive integer not greater than the M.

19. The method according to any one of claims 12 to 18, characterized in that The first process relies on the reliability of the M1 operations.

20. The method according to any one of claims 12 to 19, characterized in that The input of the first process includes M1 inputs, the M1 inputs are respectively dependent on the outputs of the M1 operations, and the first process weights the M1 inputs respectively with M1 coefficients, the M1 coefficients are dependent on the M1 operations.

21. The method according to any one of claims 12 to 20, characterized in that include: Execute a first operation; The first information block includes first compressed CSI, and the first compressed CSI is used as an input of the first operation to generate first restored CSI.

22. A base station, characterized in that: The base station includes: one or more processors and a memory; The memory is coupled to the one or more processors, and the memory is used to store computer program codes, where the computer program codes include computer instructions, and the one or more processors call the computer instructions to enable the base station to perform the method according to any one of claims 12 to 21.

Images