Data set collection method and device, model training method and device and communication equipment

Abstract

The invention discloses a data set collection method and device, a model training method and device and communication equipment, and belongs to the technical field of communication, and the data set collection method comprises the steps that first equipment obtains communication measurement related data and perception measurement related data; and the first device determines a target data set for model training based on the communication measurement related data and the perception measurement related data.

Description

Technical Field

[0001] This application belongs to the field of communication technologies, and particularly relates to a method for collecting a data set, a method for training a model, a device, and a communication device. Background Art

[0002] With the development of Artificial Intelligence (AI) technology, AI technology is expected to be applied to communication technologies. For example, in related technologies, it has been proposed to use an AI neural network model to implement communication functions such as beam prediction, Channel State Information (CSI) prediction, or positioning services. An AI neural network model needs to collect relevant data (i.e., training samples) for model training to be obtained. In related technologies, the data used for model training basically comes from the data obtained through communication measurements. However, the data obtained through communication measurements is relatively limited, which results in limited inference accuracy of the trained AI model. Summary of the Invention

[0003] Embodiments of this application provide a method for collecting a data set, a method for training a model, a method for storing a model, a device, and a communication device, which can solve the problem in related technologies that the inference accuracy of the trained AI model is limited due to relatively limited training samples.

[0004] In a first aspect, a method for collecting a data set is provided, which is executed by a first device. The method includes:

[0005] The first device obtains communication measurement-related data and perception measurement-related data;

[0006] The first device determines a target data set for model training based on the communication measurement-related data and the perception measurement-related data.

[0007] In a second aspect, a method for collecting a data set is provided, which is executed by a second device. The method includes:

[0008] The second device receives a first request from the first device, where the first request is used to request perception measurement-related data required for model training;

[0009] The second device performs a first operation, where the first operation includes any one of the following:

[0010] The second device sends a target signal based on the first request;

[0011] The second device determines configuration information of the target signal based on the first request;

[0012] The second device sends the configuration information of the target signal based on the first request;

[0013] The second device sends data related to sensing measurements based on the first request;

[0014] Wherein, the target signal is a signal used for sensing.

[0015] In a third aspect, a method for collecting a data set is provided, which is executed by a third device. The method includes:

[0016] The third device sends first information to the first device, and the first information includes at least one of communication measurement related data and sensing measurement related data;

[0017] The first information includes at least one of the following:

[0018] Communication measurement quantities and sensing measurement quantities, where the communication measurement quantities and the sensing measurement quantities are determined to belong to the same training sample;

[0019] Communication measurement quantities and a third indication, where the communication measurement quantities and the sensing measurement quantities indicated by the third indication are determined to belong to the same training sample;

[0020] Communication measurement quantities, where the communication measurement quantities and the sensing measurement quantities of the most recent training sample are determined to belong to the same training sample;

[0021] Or,

[0022] The first information includes at least one of the following:

[0023] A data set identifier, and the data set corresponding to the data set identifier includes sensing measurement related data;

[0024] A configuration identifier of the sensing measurement quantity;

[0025] An identifier of the sensing measurement quantity;

[0026] An identifier of the sensing measurement link;

[0027] An identifier of the sensing service;

[0028] An identifier of the sensing service type;

[0029] Sensing measurement quantities;

[0030] A timestamp of the sensing measurement quantity;

[0031] Information about the sending device of the sensing measurement quantity;

[0032] Coordinate information of the sensing measurement quantity;

[0033] Information for indicating a performance metric of the sensing measurement quantity;

[0034] Information for indicating the source of the sensed measurement quantity;

[0035] Information for indicating the type of the sensed measurement quantity;

[0036] Information for indicating the use of the sensed measurement quantity;

[0037] Information for indicating the sensing mode of the sensed measurement quantity;

[0038] Communication measurement quantity;

[0039] Communication measurement resource identifier;

[0040] Timestamp of the communication measurement quantity;

[0041] Information for indicating the first time window;

[0042] Information for indicating the first valid time;

[0043] Configuration identifier of the target signal;

[0044] Configuration information of the target signal;

[0045] Information of the sending device of the target signal;

[0046] Information of the receiving device of the target signal;

[0047] Wherein, the target signal is a signal used for sensing.

[0048] In a fourth aspect, a model training method is provided, which is executed by a fourth device, and the method includes:

[0049] The fourth device receives second information from a first device, and the second information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensing measurement-related data;

[0050] The fourth device uses the target data set to train a target AI unit.

[0051] In a fifth aspect, a data set collection method is provided, which is executed by a fifth device, and the method includes:

[0052] The fifth device receives third information from a first device, or receives fourth information from a fourth device, and the third information or the fourth information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensing measurement-related data;

[0053] The fifth device stores the target data set and the identifier of the target data set.

[0054] In a sixth aspect, a model storage method is provided, which is executed by a sixth device. The method includes:

[0055] The sixth device receives fifth information from a fourth device. The fifth information includes a target AI unit, an identifier of the target AI unit, and an identifier of a target data set. The data set corresponding to the identifier of the target data set includes communication measurement-related data and perception measurement-related data;

[0056] The sixth device stores the fifth information.

[0057] In a seventh aspect, a data set collection device is provided, which is applied to a first device. The device includes:

[0058] A first processing module, configured to obtain communication measurement-related data and perception measurement-related data;

[0059] A second processing module, configured to determine a target data set for model training based on the communication measurement-related data and the perception measurement-related data.

[0060] In an eighth aspect, a data set collection device is provided, which is applied to a second device. The device includes:

[0061] A receiving module, configured to receive a first request from the first device. The first request is used to request perception measurement-related data required for model training;

[0062] A processing module, configured to perform a first operation. The first operation includes any one of the following:

[0063] Based on the first request, send a target signal;

[0064] Based on the first request, determine configuration information of the target signal;

[0065] Based on the first request, send the configuration information of the target signal;

[0066] Based on the first request, send perception measurement-related data;

[0067] Wherein, the target signal is a signal used for perception.

[0068] In a ninth aspect, a data set collection device is provided, which is applied to a third device. The device includes:

[0069] A sending module, configured to send first information to the first device. The first information includes at least one of communication measurement-related data and perception measurement-related data;

[0070] The first information includes at least one of the following:

[0071] Communication measurement quantities and sensing measurement quantities, where the communication measurement quantities and the sensing measurement quantities are determined to belong to the same training sample;

[0072] Communication measurement quantities and a third indication, where the communication measurement quantities and the sensing measurement quantities indicated by the third indication are determined to belong to the same training sample;

[0073] Communication measurement quantities, where the communication measurement quantities and the sensing measurement quantities of the most recent training sample are determined to belong to the same training sample;

[0074] Or,

[0075] The first information includes at least one of the following:

[0076] A dataset identifier, where the dataset corresponding to the dataset identifier includes sensing measurement-related data;

[0077] A configuration identifier of sensing measurement quantities;

[0078] An identifier of sensing measurement quantities;

[0079] An identifier of a sensing measurement link;

[0080] An identifier of a sensing service;

[0081] An identifier of a sensing service type;

[0082] Sensing measurement quantities;

[0083] A timestamp of sensing measurement quantities;

[0084] Information about the sending device of sensing measurement quantities;

[0085] Coordinate information of sensing measurement quantities;

[0086] Information for indicating performance metrics of sensing measurement quantities;

[0087] Information for indicating the source of sensing measurement quantities;

[0088] Information for indicating the type of sensing measurement quantities;

[0089] Information for indicating the use of sensing measurement quantities;

[0090] Information for indicating the sensing mode of sensing measurement quantities;

[0091] Communication measurement quantities;

[0092] A communication measurement resource identifier;

[0093] A timestamp of communication measurement quantities;

[0094] Information for indicating a first time window;

[0095] Information for indicating a first valid time;

[0096] Configuration identifier of the target signal;

[0097] Configuration information of the target signal;

[0098] Information of the sending device of the target signal;

[0099] Information of the receiving device of the target signal;

[0100] Wherein, the target signal is a signal used for sensing.

[0101] In a tenth aspect, a model training device is provided, which is applied to a fourth device. The device includes:

[0102] A receiving module, configured to receive second information from a first device, where the second information includes a target data set and an identifier of the target data set; the target data set includes communication measurement related data and sensing measurement related data;

[0103] A processing module, configured to use the target data set to train a target AI unit.

[0104] In an eleventh aspect, a data set collection device is provided, which is applied to a fifth device. The device includes:

[0105] A receiving module, configured to receive third information from a first device, or receive fourth information from a fourth device, where the third information or the fourth information includes a target data set and an identifier of the target data set; the target data set includes communication measurement related data and sensing measurement related data;

[0106] A processing module, configured to store the target data set and the identifier of the target data set;

[0107] Wherein, the first device is a device for collecting the target data set, the fourth device is a device for model training, and the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set.

[0108] In a twelfth aspect, a model storage device is provided, which is applied to a sixth device. The device includes:

[0109] A receiving module, configured to receive fifth information from a fourth device, where the fifth information includes a target AI unit, an identifier of the target AI unit, and an identifier of a target data set, and the data set corresponding to the identifier of the target data set includes communication measurement related data and sensing measurement related data;

[0110] The first processing module is used to store the fifth information.

[0111] In a thirteenth aspect, a communication device is provided. The communication device includes a processor and a memory. The memory stores a program or instruction that can run on the processor. When the program or instruction is executed by the processor, it implements the steps of the method described in the first aspect, or implements the steps of the method described in the second aspect, or implements the steps of the method described in the third aspect, or implements the steps of the method described in the fourth aspect, or implements the steps of the method described in the fifth aspect, or implements the steps of the method described in the sixth aspect.

[0112] In a fourteenth aspect, a communication device is provided, including a processor and a communication interface. The processor is configured to: The first device obtains communication measurement-related data and perception measurement-related data; The first device determines a target data set for model training based on the communication measurement-related data and the perception measurement-related data.

[0113] In a fifteenth aspect, a communication device is provided, including a processor and a communication interface. The communication interface is configured to: receive a first request from a first device, where the first request is used to request perception measurement-related data required for model training; The processor is configured to: perform a first operation, where the first operation includes any one of the following:

[0114] Based on the first request, send a target signal;

[0115] Based on the first request, determine configuration information of the target signal;

[0116] Based on the first request, send the configuration information of the target signal;

[0117] Based on the first request, send perception measurement-related data;

[0118] Wherein, the target signal is a signal used for perception.

[0119] In a sixteenth aspect, a communication device is provided, including a processor and a communication interface. The communication interface is configured to: send first information to a first device, where the first information includes at least one of communication measurement-related data and perception measurement-related data;

[0120] The first information includes at least one of the following:

[0121] Communication measurement quantities and perception measurement quantities, where the communication measurement quantities and the perception measurement quantities are determined to belong to the same training sample;

[0122] Communication measurement quantity and a third indication, where the communication measurement quantity and the perception measurement quantity indicated by the third indication are determined to belong to the same training sample;

[0123] Communication measurement quantity, where the communication measurement quantity and the perception measurement quantity of the most recent training sample are determined to belong to the same training sample;

[0124] Or,

[0125] The first information includes at least one of the following:

[0126] Dataset identifier, where the dataset corresponding to the dataset identifier includes perception measurement-related data;

[0127] Configuration identifier of the perception measurement quantity;

[0128] Identifier of the perception measurement quantity;

[0129] Identifier of the perception measurement link;

[0130] Identifier of the perception service;

[0131] Identifier of the perception service type;

[0132] Perception measurement quantity;

[0133] Timestamp of the perception measurement quantity;

[0134] Information of the sending device of the perception measurement quantity;

[0135] Coordinate information of the perception measurement quantity;

[0136] Information for indicating the performance metric of the perception measurement quantity;

[0137] Information for indicating the source of the perception measurement quantity;

[0138] Information for indicating the type of the perception measurement quantity;

[0139] Information for indicating the use of the perception measurement quantity;

[0140] Information for indicating the perception mode of the perception measurement quantity;

[0141] Communication measurement quantity;

[0142] Communication measurement resource identifier;

[0143] Timestamp of the communication measurement quantity;

[0144] Information for indicating the first time window;

[0145] Information for indicating the first valid time;

[0146] Configuration identifier of the target signal;

[0147] Configuration information of the target signal

[0148] Information of the sending device of the target signal

[0149] Information of the receiving device of the target signal

[0150] Wherein, the target signal is a signal used for sensing.

[0151] In a seventeenth aspect, a communication device is provided, including a processor and a communication interface. Wherein, the communication interface is configured to: receive second information from a first device, the second information including a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensing measurement-related data; the processor is configured to: use the target data set to train a target AI unit.

[0152] In an eighteenth aspect, a communication device is provided, including a processor and a communication interface. Wherein, the communication interface is configured to: receive third information from a first device, or receive fourth information from a fourth device, the third information or the fourth information including a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensing measurement-related data; the processor is configured to: store the target data set and the identifier of the target data set; wherein, the first device is a device that collects the target data set, the fourth device is a device for model training, the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set.

[0153] In a nineteenth aspect, a communication device is provided, including a processor and a communication interface. Wherein, the communication interface is configured to: receive fifth information from a fourth device, the fifth information including a target AI unit, an identifier of the target AI unit, and an identifier of a target data set, the data set corresponding to the identifier of the target data set including communication measurement-related data and sensing measurement-related data; the processor is configured to: store the fifth information.

[0154] In a twentieth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, it implements the steps of the method as described in the first aspect, or implements the steps of the method as described in the second aspect, or implements the steps of the method as described in the third aspect, or implements the steps of the method as described in the fourth aspect, or implements the steps of the method as described in the fifth aspect, or implements the steps of the method as described in the sixth aspect.

[0155] In a twenty - first aspect, a wireless communication system is provided, including: a first device and a second device. The first device can be used to execute the steps of the method described in the first aspect, and the second device can be used to execute the steps of the method described in the second aspect.

[0156] In a twenty - second aspect, a chip is provided. The chip includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the method described in the first aspect, or to implement the method described in the second aspect, or to implement the steps of the method described in the third aspect, or to implement the steps of the method described in the fourth aspect, or to implement the steps of the method described in the fifth aspect, or to implement the steps of the method described in the sixth aspect.

[0157] In a twenty - third aspect, a computer program / program product is provided. The computer program / program product is stored in a storage medium. The program / program product is executed by at least one processor to implement the steps of the method described in the first aspect, or to implement the method described in the second aspect, or to implement the steps of the method described in the third aspect, or to implement the steps of the method described in the fourth aspect, or to implement the steps of the method described in the fifth aspect, or to implement the steps of the method described in the sixth aspect.

[0158] In an embodiment of the present application, the first device acquires communication measurement - related data and perception measurement - related data; the first device determines a target data set for model training based on the communication measurement - related data and the perception measurement - related data. In this way, the input information for model training can be expanded, the model training effect can be improved, and thus the trained AI model can have a high inference accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0159] Figure 1 is a block diagram of a wireless communication system provided by an embodiment of the present application;

[0160] Figure 2 is a schematic diagram of different perception modes of communication - perception integration;

[0161] Figure 3 is a flowchart of a data set collection method provided by an embodiment of the present application;

[0162] Figure 4 is a flowchart of a data set collection method provided by an embodiment of the present application;

[0163] Figure 5 is a flowchart of a data set collection method provided by an embodiment of the present application;

[0164] Figure 6It is a flowchart of a model training method provided by an embodiment of the present application;

[0165] Figure 7 It is a flowchart of a dataset collection method provided by an embodiment of the present application;

[0166] Figure 8 It is a flowchart of a model storage method provided by an embodiment of the present application;

[0167] Figure 9 It is a flowchart of Embodiment 1 provided by an embodiment of the present application;

[0168] Figures 10 to 12 It is an example diagram of possible situations of using a perceptual measurement quantity in a sample;

[0169] Figure 13 It is a flowchart of Embodiment 2 provided by an embodiment of the present application;

[0170] Figure 14 It is a flowchart of Embodiment 3 provided by an embodiment of the present application;

[0171] Figure 15 It is a flowchart of Embodiment 4 provided by an embodiment of the present application;

[0172] Figure 16 It is a flowchart of Embodiment 5 provided by an embodiment of the present application;

[0173] Figure 17 It is a flowchart of Embodiment 6 provided by an embodiment of the present application;

[0174] Figure 18 It is a flowchart of Embodiment 7 provided by an embodiment of the present application;

[0175] Figures 19 to 20 It is an example diagram corresponding to Embodiment 7;

[0176] Figure 21 It is a flowchart of Embodiment 8 provided by an embodiment of the present application;

[0177] Figure 22 It is a flowchart of Embodiment 9 provided by an embodiment of the present application;

[0178] Figure 23 It is a structural diagram of a dataset collection device provided by an embodiment of the present application;

[0179] Figure 24 It is a structural diagram of a dataset collection device provided by an embodiment of the present application;

[0180] Figure 25 It is a structural diagram of a dataset collection device provided by an embodiment of the present application;

[0181] Figure 26 It is a structural diagram of a model training device provided by an embodiment of the present application;

[0182] Figure 27 It is a structural diagram of a data set collection device provided by an embodiment of the present application;

[0183] Figure 28 It is a structural diagram of a model storage device provided by an embodiment of the present application;

[0184] Figure 29 It is a structural diagram of a communication device provided by an embodiment of the present application;

[0185] Figure 30 It is a schematic diagram of the hardware structure of a terminal provided by an embodiment of the present application;

[0186] Figure 31 It is a schematic diagram of the hardware structure of a network-side device provided by an embodiment of the present application;

[0187] Figure 32 It is a schematic diagram of the hardware structure of another network-side device provided by an embodiment of the present application. Specific embodiments

[0188] Next, the technical solutions in the embodiments of the present application will be clearly described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present application.

[0189] The terms "first", "second", etc. in the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such terms can be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are usually of the same category, and the number of objects is not limited. For example, the first object can be one or multiple. In addition, "or" in the present application means at least one of the connected objects. For example, "A or B" covers three scenarios, namely, Scenario 1: including A and not including B; Scenario 2: including B and not including A; Scenario 3: including both A and B. The character " / " generally indicates an "or" relationship between the associated objects before and after.

[0190] The term "indication" in this application can be either a direct indication (or an explicit indication) or an indirect indication (or an implicit indication, a hidden indication). Among them, a direct indication can be understood as that the sender clearly tells the receiver specific information, operations to be performed, request results, etc. in the sent indication; an indirect indication can be understood as that the receiver determines the corresponding information according to the indication sent by the sender, or makes a judgment and determines the operations to be performed or request results, etc. according to the judgment result.

[0191] It should be noted that the technology described in the embodiments of this application is not limited to the Long Term Evolution (LTE) / LTE-Advanced (LTE-A) system, and can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA), or other systems. The terms "system" and "network" in the embodiments of this application are often used interchangeably, and the described technology can be used in the above-mentioned systems and radio technologies, as well as in other systems and radio technologies. The following description describes the New Radio (NR) system for example purposes, and uses NR terms in most of the following descriptions, but these technologies can also be applied to systems other than the NR system, such as the 6th Generation (6G) communication system. th Generation, 6G) communication system.

[0192] Figure 1The block diagram of a wireless communication system to which the embodiments of the present application can be applied is shown. The wireless communication system includes a terminal 11 and a network-side device 12. Among them, the terminal 11 can be a mobile phone, a tablet personal computer, a laptop computer, a notebook computer, a personal digital assistant (PDA), a handheld computer, a netbook, an ultra-mobile personal computer (UMPC), a mobile internet device (MID), an augmented reality (AR), a virtual reality (VR) device, a robot, a wearable device, a flight vehicle, a vehicle user equipment (VUE), a shipborne device, a pedestrian user equipment (PUE), a smart home (home devices with wireless communication functions, such as refrigerators, TVs, washing machines or furniture, etc.), a game console, a personal computer (PC), a teller machine or a self-service machine, etc. Wearable devices include: smart watches, smart bracelets, smart earphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart ankle chains, etc.), smart wristbands, smart clothing, etc. Among them, the vehicle user equipment can also be called a vehicle terminal, a vehicle controller, a vehicle module, a vehicle component, a vehicle chip or a vehicle unit, etc. In addition to the above terminal devices, it can also be a chip inside the terminal, such as a modem chip or a system on chip (SoC). It should be noted that the specific type of the terminal 11 is not limited in the embodiments of the present application. The network-side device 12 can include an access network device or a core network device. Among them, the access network device can also be called a radio access network (RAN) device, a radio access network function or a radio access network unit. The access network device can include a base station, a wireless local area network (WLAN) access point (AP) or a wireless fidelity (WiFi) node, etc.Among them, the base station can be referred to as Node B (NB), Evolved Node B (eNB), the next generation Node B (gNB), New Radio Node B (NR Node B), access point, Relay Base Station (RBS), Serving Base Station (SBS), Base Transceiver Station (BTS), radio base station, radio transceiver, Basic Service Set (BSS), Extended Service Set (ESS), home Node B (HNB), home evolved Node B, Transmission Reception Point (TRP), or some other suitable term in the art. As long as the same technical effect is achieved, the base station is not limited to specific technical terms. It should be noted that in the embodiments of this application, the base station in the NR system is mainly taken as an example for introduction, and the specific type of the base station is not limited.

[0193] The core network device may include, but is not limited to, at least one of the following: core network node, core network function, Mobility Management Entity (MME), Access and Mobility Management Function (AMF), Session Management Function (SMF), User Plane Function (UPF), Policy Control Function (PCF), Policy and Charging Rules Function (PCRF), Edge Application Server Discovery Function (EASDF), Unified Data Management (UDM), Unified Data Repository (UDR), Home Subscriber Server (HSS), Centralized network configuration (CNC), Location Management Function (LMF), Network Repository Function (NRF), Network Exposure Function (NEF), Local NEF (L-NEF), Binding Support Function (BSF), Application Function (AF), etc. It should be noted that in the embodiments of this application, the core network devices in the NR system are mainly used as examples for introduction, and the specific types of core network devices are not limited.

[0194] The related technologies will be briefly introduced below.

[0195] Related Technology 1: AI Technology and Communication Technology

[0196] 1) AI-based Beam Prediction

[0197] In millimeter-wave wireless communication, both communication transceiver ends, such as base stations and user equipment (UE, also known as terminals), are configured with multiple analog beams. For the same UE, when measuring using different transmit analog beams and receive analog beams, the measured channel quality varies. How to quickly and accurately find the transceiver beam group with the highest channel quality from all possible combinations of transmit and receive analog beams is the key to affecting transmission quality. After introducing the AI neural network model (abbreviated as AI unit / AI model), the terminal can effectively predict the receive analog beam or transmit analog beam with the highest channel quality based on historical channel quality information and report it to the network side.

[0198] In 5G, beam prediction mainly uses beam quality information obtained based on CSI measurement, beam identification information, etc. as model inputs. In the AI-based beam prediction use cases discussed in 5G, during model training, a training sample includes model inputs and ground truth / labels. Among them, the model inputs include beam quality, beam identification / CSI resource identification, or measurement timestamp obtained based on CSI measurement; the ground truth / labels include the strongest beam identification or the beam intensity of each beam (measured actually at a future moment).

[0199] 2) AI-based CSI prediction

[0200] In the AI-based CSI prediction use cases discussed in 5G, during model training, a training sample includes model inputs and ground truth / labels. Among them, the model inputs include the historical channel matrix, precoding matrix indicator (PMI), channel eigenvector / eigenvalue, or measurement timestamp obtained based on CSI measurement; the ground truth / labels include the channel matrix, PMI, or channel eigenvector / eigenvalue (measured actually at a future moment).

[0201] 3) AI-based positioning

[0202] In the AI-based positioning use cases discussed in 5G, during model training, a training sample includes model inputs and ground truth / labels. Among them, the model inputs include measurement results of the time-domain channel and positioning reference signals; the ground truth / labels include the distance relative to the base station or the time of arrival (TOA) (measured actually).

[0203] Related technology 2: Communication and sensing integration

[0204] In addition to communication capabilities, mobile communication systems such as B5G systems or 6G systems will also possess sensing capabilities. The sensing capabilities, that is, one or more devices with sensing capabilities, can sense information such as the orientation, distance, or speed of a target object through the transmission and reception of wireless signals, or detect, track, identify, or image a target object, event, or environment, etc. With the deployment of small base stations with high-frequency band and large bandwidth capabilities such as millimeter waves and terahertz in the 6G network, the sensing resolution will be significantly improved compared to centimeter waves, enabling the 6G network to provide more refined sensing services. A typical sensing function and application scenario are shown in Table 1.

[0205] Table 1

[0206]

[0207] Communication-sensing integration (abbreviated as communication-sensing integration) means that in the same system, through spectrum sharing and hardware sharing, the integrated design of communication and sensing functions is realized. While the system is transmitting information, it can sense information such as orientation, distance, and speed, detect, track, and identify target devices or events. The communication system and the sensing system complement each other, achieving an improvement in overall performance and bringing a better service experience.

[0208] The integration of communication and radar belongs to a typical application of communication-sensing integration (communication-sensing fusion). In the past, radar systems and communication systems were strictly separated due to different research objects and focuses of attention, and the two systems were studied independently in most scenarios. In fact, both radar and communication systems are typical ways of information transmission, acquisition, processing, and exchange, and there are many similarities in terms of working principles, system architectures, and frequency bands. The design of the integration of communication and radar has great feasibility, which is mainly reflected in the following aspects: First, both the communication system and the sensing system are based on the electromagnetic wave theory, and use the transmission and reception of electromagnetic waves to complete the acquisition and transmission of information; Second, both the communication system and the sensing system have structures such as antennas, transmitters, receivers, and signal processors, and there is a large overlap in hardware resources; With the development of technology, there is also an increasing overlap in their working frequency bands; In addition, there are similarities in key technologies such as signal modulation, reception detection, and waveform design. The integration of communication and radar systems can bring many advantages, such as cost savings, size reduction, power consumption reduction, spectrum efficiency improvement, mutual interference reduction, etc., thus improving the overall performance of the system.

[0209] According to the different sensing signal sending nodes and receiving nodes, it can be divided into 6 basic sensing modes. As Figure 2 shown, specifically including:

[0210] (1) Base station echo sensing. In this sensing mode, base station A sends a sensing signal and performs sensing measurement by receiving the echo of the sensing signal.

[0211] (2) Air interface sensing between base stations. In this sensing mode, base station B receives the sensing signal sent by base station A and performs sensing measurements.

[0212] (3) Uplink air interface sensing. In this sensing mode, base station A receives the sensing signal sent by terminal A and performs sensing measurements.

[0213] (4) Downlink air interface sensing. In this sensing mode, terminal B receives the sensing signal sent by base station B and performs sensing measurements.

[0214] (5) Terminal echo sensing. In this sensing mode, terminal A sends a sensing signal and performs sensing measurements by receiving the echo of the sensing signal.

[0215] (6) Sidelink sensing between terminals. In this sensing mode, terminal B receives the sensing signal sent by terminal A and performs sensing measurements.

[0216] It should be noted that Figure 2 each sensing mode in [the above] takes a sensing signal sending node and a sensing signal receiving node as examples. In an actual system, one or more different sensing modes can be selected according to different sensing use cases and sensing requirements, and there can be one or more sending nodes and receiving nodes for each sensing mode. Figure 2 The sensing targets in [the above] take people and vehicles as examples, and it is assumed that neither people nor vehicles carry or install signal transceiver devices. The sensing targets in the actual scenario will be more diverse.

[0217] The following introduces three ways to obtain sensing results:

[0218] Obtain sensing results with A sending and B receiving:

[0219] Node A sends a sensing reference signal (or called a sensing measurement signal), node B receives the sensing reference signal, and node B obtains the sensing measurement quantity / sensing result. Among them, node A can be base station 1, node B can be the target UE, or a UE near the target UE, or base station 2.

[0220] Obtain sensing results with self - sending and self - receiving:

[0221] Node A sends a sensing reference signal, node A receives the sensing reference signal, and node A obtains the sensing measurement quantity / sensing result. Among them, node A can be the target UE, or the serving base station of the target UE, or other base stations.

[0222] Obtain sensing results through sensors:

[0223] The network node obtains the sensed measurement quantity / sensed result through the sensor-like sensing devices deployed by itself or in the environment.

[0224] It should be noted that, without special indication, the sensed measurement quantity, sensed measurement result, etc. can all be understood as the sensed result.

[0225] Sensor-like sensing is a sensing method that performs sensing services through means other than the communication-sensing integrated system. Typical devices include lidar, millimeter-wave radar, vision sensors (such as monocular vision sensors, binocular vision sensors, infrared sensors, etc.), inertial measurement units (IMUs), and various other sensors (such as rain gauges, thermometers, hygrometers, etc.).

[0226] In 6G, the network can obtain the sensed result through sensors or sensed measurement signals. If the sensed result can be used as an auxiliary input to the model, the inference accuracy of the AI model (see Explanation 1 later) can be further improved. However, in Related Technology 1, in the physical layer AI use cases discussed in 5G currently, the collection of model training data does not involve the acquisition of sensing information and how to utilize sensing information. Related Technology 2 is a common method for obtaining the sensed measurement quantity / sensed result, and does not involve how to use it as the training data of the AI model to improve the inference accuracy of the AI model. Currently, there is no mature solution for using the sensed result and communication measurement result as the input data of the model for model training. In view of this, the embodiments of this application propose a solution for collecting model training data assisted by sensing, enabling network nodes, such as terminals, base stations, etc., to obtain sensing information as auxiliary information for the AI model for model training, which can improve the training effect of the AI model and thus improve the inference accuracy of the AI model.

[0227] The embodiments of this application involve the interaction of multiple communication devices, and each communication device can be represented by various network nodes. The functions corresponding to each communication device and the network nodes representing each communication device can be seen in Table 2.

[0228] Table 2

[0229]

[0230] In the embodiments of this application, the target signal refers to a dedicated signal used for sensing, such as a CSI reference signal (CSI-RS), a CSI-RS for tracking (TRS), a sounding reference signal (SRS), or a synchronization signal, etc. The target signal can be called a sensing reference signal or a sensed measurement signal.

[0231] The following will combine the accompanying drawings to provide a detailed description of the dataset collection method and the model training method provided by the embodiments of the present application.

[0232] Figure 3 The flowchart of a dataset collection method provided by an embodiment of the present application is shown. As Figure 3 shown, the dataset collection method includes the following steps:

[0233] Step 301: The first device obtains communication measurement-related data and sensing measurement-related data;

[0234] Step 302: The first device determines a target dataset for model training based on the communication measurement-related data and the sensing measurement-related data.

[0235] Communication measurement-related data: Data related to communication measurement. For example, the communication measurement-related data includes one of the following: CSI measurement quantity (or CSI measurement result), CSI measurement resource identifier, reference signal received power (RSRP) of a beam, reference signal received quality (RSRQ) of a beam, signal-to-noise and interference ratio (SINR) of a beam, RSRP of a cell channel, RSRQ of a cell channel, SINR of a cell channel, received signal strength indication (RSSI) of a cell channel, cell channel impulse response, precoding matrix indicator (PMI), rank indicator (RI), channel quality indicator (CQI), beam identifier, subband identifier.

[0236] Sensing measurement-related data: Data related to sensing measurement, such as data including sensing measurement configuration, sensing results (such as sensing measurement quantities), sensing performance, sensing characteristic description information, etc.

[0237] The target dataset can be composed of multiple training samples. It can be that some training samples include communication measurement-related data and sensing measurement-related data, and some training samples only include communication measurement-related data; or all training samples include communication measurement-related data and sensing measurement-related data. The embodiments of the present application do not limit this.

[0238] In the embodiments of the present application, the first device acquires communication measurement-related data and perception measurement-related data; the first device determines a target data set for model training based on the communication measurement-related data and the perception measurement-related data. In this way, the input information for model training can be expanded, the effect of model training can be improved, and thus the trained AI model can have a high inference accuracy.

[0239] In some embodiments, the method further includes:

[0240] The first device uses the target data set to train the target AI unit.

[0241] In the embodiments of the present application, in addition to being a data set collection device, the first device can also be a model training device, that is, the first device and the fourth device are combined.

[0242] In this implementation manner, the first device can use the target data set collected by itself to train the target AI unit.

[0243] In some embodiments, the method further includes:

[0244] The first device generates an identifier of the target data set.

[0245] The identifier of the target data set is associated with the perception characteristics of the target data set; or,

[0246] The identifier of the target data set is associated with the perception characteristics and communication characteristics (also known as wireless communication characteristics) of the target data set.

[0247] In this implementation manner, when the first device generates an identifier of the target data set, it can be understood that the first device generates the identifier of the target data set based on the perception characteristics of the target data set, that is, the first device associates the identifier of the target data set with the perception characteristics of the target data set. Or, when the first device generates an identifier of the target data set, it can be understood that the first device generates the identifier of the target data set based on the perception characteristics and communication characteristics of the target data set, that is, the first device associates the identifier of the target data set with the perception characteristics and communication characteristics of the target data set. In this way, the perception characteristics (and communication characteristics) of the target data set can be known through the identifier of the target data set, and the perception characteristics (and communication characteristics) of the target data set can provide guidance for subsequent model training processes and model inference processes. For example, maintaining the consistency of data (distribution) characteristics during model training and model inference is beneficial to improving the effect of model training and the accuracy of model inference.

[0248] The sensing characteristics can be expressed as sensing additional conditions / auxiliary conditions. Correspondingly, the communication characteristics can be expressed as communication additional conditions / auxiliary conditions.

[0249] Exemplarily, the communication characteristics include at least one of the following: beam direction, beam width, antenna downtilt, cell radius, cell channel characteristics (urban macrocells (Uma), urban microcells (UMi), rural macrocells (RMa)), line of sight (LOS) / non-line of sight (NLOS), indoor / outdoor, number of antenna ports (horizontal / vertical).

[0250] Optionally, the sensing measurement related data includes at least one of the following information:

[0251] Sensing measurement quantity;

[0252] Indicator of the sensing measurement quantity;

[0253] Timestamp of the sensing measurement quantity;

[0254] Information of the sending device of the sensing measurement quantity;

[0255] Information of the receiving device of the sensing measurement quantity;

[0256] Coordinate information of the sensing measurement quantity;

[0257] Information for indicating the performance index of the sensing measurement quantity;

[0258] Information for indicating the source of the sensing measurement quantity;

[0259] Information for indicating the type of the sensing measurement quantity;

[0260] Information for indicating the sensing mode of the sensing measurement quantity;

[0261] Configuration information of the target signal;

[0262] Information of the sending device of the target signal;

[0263] Information of the receiving device of the target signal;

[0264] Wherein, the target signal is a signal used for sensing.

[0265] In the embodiments of the present application, for the type of the sensing measurement quantity, reference can be made to Explanation 2 later, and for the configuration information of the target signal, reference can be made to Explanation 3 later.

[0266] The timestamp of the sensed measurement quantity may refer to the measurement timestamp of the sensed measurement quantity or the reception timestamp of the sensed measurement quantity, and its meaning can be flexibly determined according to specific circumstances.

[0267] In some embodiments, the method further includes:

[0268] The first device sends a first request to the second device, and the first request is used to request the sensed measurement-related data required for model training.

[0269] For example, when the first device is a terminal, it can send a first request to its serving base station. The first request can reflect the first device's need for the sensed information required for model training data, which is beneficial for the first device to collect model training data that meets the requirements, thereby facilitating the improvement of the model training effect.

[0270] Optionally, the first request includes at least one of the following information:

[0271] Information indicating that the model is in the training phase;

[0272] A dataset identifier, and the dataset corresponding to the dataset identifier includes sensed measurement-related data;

[0273] A configuration identifier of the sensed measurement quantity;

[0274] An identifier of the sensed measurement quantity;

[0275] Configuration information of the sensed measurement quantity required for model training;

[0276] Information indicating the number threshold of the sensed measurement quantity required for model training;

[0277] Information indicating the source of the sensed measurement quantity required for model training;

[0278] Information indicating the type of the sensed measurement quantity required for model training;

[0279] Information indicating the sensing mode of the sensed measurement quantity required for model training;

[0280] Information indicating the sensing requirement;

[0281] Configuration information of the target signal;

[0282] Information about the sending device of the target signal;

[0283] Information about the receiving device of the target signal;

[0284] Wherein, the target signal is a signal used for sensing.

[0285] Each of the above identifiers can be a predefined identifier.

[0286] The above instructions can be either implicit instructions or explicit instructions, and the embodiments of the present application do not limit this.

[0287] In some embodiments, the first device collects a target data set for model training, including:

[0288] The first device generates training samples according to the obtained communication measurement-related data and perception measurement-related data, and the target data set includes the training samples.

[0289] A training sample may include model input data and ground truth / label data, and the model input data may include communication measurement-related data or perception measurement-related data.

[0290] The first device obtains communication measurement-related data through communication measurements (such as CSI measurement, Radio Resource Management (RRM) measurement).

[0291] There are many ways for the first device to obtain perception measurement-related data. It can be achieved by the first device performing perception measurements or obtained from other devices. For specific details, refer to the embodiments provided later.

[0292] The first device generates training samples according to the obtained communication measurement-related data and perception measurement-related data. It can be understood that the first device synchronizes samples of the communication measurement-related data and perception measurement-related data to determine whether to combine the communication measurement-related data and perception measurement-related data as a training sample. For example, the first device synchronizes samples based on the communication measurement-related data to determine the perception measurement-related data that can belong to the same training sample as the communication measurement-related data.

[0293] In the embodiments of the present application, through sample synchronization, each generated training sample can include communication measurement-related data and perception measurement-related data as much as possible, so that each training sample has multi-dimensional data such as communication and perception as much as possible. This is beneficial to expanding the input information of the AI model and thus beneficial to improving the training effect of the AI model.

[0294] The following separately describes optional implementation manners of sample synchronization.

[0295] As an optional implementation manner, the perception measurement-related data includes a first perception measurement quantity;

[0296] The first device determines a target data set for model training based on the communication measurement-related data and the perception measurement-related data, including at least one of the following:

[0297] The first device generates a first training sample based on the target timestamp, the timestamp of the first sensed measurement quantity, and the effective duration of the first sensed measurement quantity, and determines the target data set based on the first training sample;

[0298] The first device generates a first training sample based on the first indication and the target timestamp, and determines the target data set based on the first training sample;

[0299] The first device generates a first training sample based on whether a second indication is received, and determines the target data set based on the first training sample;

[0300] Wherein, the target timestamp is the measurement timestamp of the first communication measurement quantity, and the first communication measurement quantity is the last communication measurement quantity corresponding to the true value or label in the communication measurement related data;

[0301] The first indication is used to indicate the failure time or the failure time difference of the first sensed measurement quantity;

[0302] The second indication is used to indicate the failure of the first sensed measurement quantity.

[0303] This embodiment lies in that the first device determines whether to generate a training sample from the sensed result and the communication measurement result based on the invalid time of the sensed result. In this way, it can be ensured that the sensed measurement related data in the training sample is timely data, thereby improving the quality of the model training sample and further enhancing the training accuracy of the AI model.

[0304] Optionally, the first device generates a first training sample based on the target timestamp, the timestamp of the first sensed measurement quantity, and the effective duration of the first sensed measurement quantity, including:

[0305] When the time corresponding to the target timestamp is earlier than the first time, the first device generates a first training sample including the communication measurement related data and the first sensed measurement quantity; the first time is the time corresponding to the timestamp of the first sensed measurement quantity plus the effective duration of the first sensed measurement quantity;

[0306] The first device generates a first training sample based on the first indication and the target timestamp, including:

[0307] When the time corresponding to the target timestamp is earlier than the second time, the first device generates a first training sample including the communication measurement related data and the first sensed measurement quantity; the second time is the failure time of the first sensed measurement quantity determined based on the first indication;

[0308] The first device generates a first training sample based on whether a second indication is received, including:

[0309] In a case where the first device does not receive the second indication, the first device generates a first training sample including the communication measurement related data and the first perception measurement quantity.

[0310] It should be noted that in this embodiment, the first training sample generated by the first device and including the first perception measurement quantity does not limit that there is only a measurement quantity in the first training sample, and may also include other perception measurement related data.

[0311] Specifically, reference can be made to the embodiments provided later.

[0312] Optionally, the method for determining the effective duration of the first perception measurement quantity includes at least one of the following:

[0313] In a case where a first effective duration is pre-configured or defined, the effective duration of the first perception measurement quantity is the first effective duration;

[0314] In a case where a second effective duration is pre-configured or defined and the first target signal carries a time adjustment value, the effective duration of the first perception measurement quantity is determined according to the second effective duration and the time adjustment value;

[0315] In a case where the first target signal carries a third effective duration, the effective duration of the first perception measurement quantity is the third effective duration;

[0316] Wherein, the first target signal is the target signal corresponding to the first perception measurement quantity, and the target signal is a signal used for perception.

[0317] Here, the first effective duration can be understood as a general absolute effective duration, which can be used as the effective duration of any perception measurement quantity; the second effective duration can be understood as a general reference effective duration, which can be used as the reference effective duration of any perception measurement quantity; the third effective duration can be understood as the effective duration unique to the first perception measurement quantity, which is only used as the effective duration of the first perception measurement quantity.

[0318] Specifically, reference can be made to the embodiments provided later.

[0319] It should be noted that the first perception measurement quantity is any perception measurement quantity, not a specific perception measurement quantity.

[0320] As another optional embodiment, the first device determines a target data set for model training based on the communication measurement related data and the perception measurement related data, including:

[0321] The first device receives first information from a third device, where the first information includes at least one of communication measurement-related data and sensing measurement-related data;

[0322] Based on the first information, the first device generates training samples and determines the target data set based on the training samples; alternatively, the first device generates training samples based on the first information and stored historical data, and determines the target data set based on the training samples;

[0323] Wherein, the historical data includes at least one of communication measurement-related data and sensing measurement-related data.

[0324] This embodiment lies in that the first device generates training samples based on the first information (or the first information and historical data) sent by the third device.

[0325] When the first information only includes communication measurement-related data, the first device can generate training samples according to the first information and historical sensing measurement-related data. Correspondingly, when the first information only includes sensing measurement-related data, the first device can generate training samples according to the first information and historical communication measurement-related data. When the first information includes both communication measurement-related data and sensing measurement-related data, if the communication measurement-related data and the sensing measurement-related data therein are data with sample synchronization by the third device, the first device can directly use the communication measurement-related data and the sensing measurement-related data therein as a training sample; if the communication measurement-related data and the sensing measurement-related data therein are data without sample synchronization by the third device, the first device can perform sample synchronization on the communication measurement-related data and the sensing measurement-related data therein to obtain a training sample.

[0326] Optionally, the first information includes at least one of the following:

[0327] A data set identifier, where the data set corresponding to the data set identifier includes sensing measurement-related data;

[0328] A configuration identifier of a sensing measurement quantity;

[0329] An identifier of a sensing measurement quantity;

[0330] An identifier of a sensing measurement link;

[0331] An identifier of a sensing service;

[0332] An identifier of a sensing service type;

[0333] A sensing measurement quantity;

[0334] A timestamp of a sensing measurement quantity;

[0335] Information of the sending device of the sensed measurement quantity;

[0336] Coordinate information of the sensed measurement quantity;

[0337] Information for indicating the performance index of the sensed measurement quantity;

[0338] Information for indicating the source of the sensed measurement quantity;

[0339] Information for indicating the type of the sensed measurement quantity;

[0340] Information for indicating the use of the sensed measurement quantity;

[0341] Information for indicating the sensing mode of the sensed measurement quantity;

[0342] Communication measurement quantity;

[0343] Communication measurement resource identifier;

[0344] Timestamp of the communication measurement quantity;

[0345] Information for indicating the first time window;

[0346] Information for indicating the first valid time;

[0347] Configuration information of the target signal;

[0348] Configuration identifier of the target signal;

[0349] Information of the sending device of the target signal;

[0350] Information of the receiving device of the target signal;

[0351] Wherein, the target signal is a signal used for sensing.

[0352] This embodiment may include the following four alternative cases:

[0353] Case 1: The first information includes a second sensed measurement quantity, a timestamp of the second sensed measurement quantity, a second communication measurement quantity, and a timestamp of the second communication measurement quantity;

[0354] Based on the first information, the first device generates a training sample, including at least one of the following:

[0355] When the time difference between the timestamp of the second sensed measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to the first time window indicated by the first information, the first device generates a second training sample including the second communication measurement quantity and the second sensed measurement quantity;

[0356] In the case where the time difference between the timestamp of the second sensed measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to a predefined second time window, the first device generates a second training sample including the second communication measurement quantity and the second sensed measurement quantity.

[0357] It should be noted that in this case, the second training sample generated by the first device, which includes the second communication measurement quantity and the second sensed measurement quantity, does not limit that the second training sample only contains measurement quantities, but may also include other communication measurement-related data and other sensed measurement-related data. The subsequent training samples can be understood in this way. To avoid repetition, this will not be elaborated further.

[0358] In this case, the first information includes communication measurement-related data and sensed measurement-related data, and the third device does not perform sample synchronization. The first device performs sample synchronization based on the first information to obtain a training sample.

[0359] For details, reference can be made to the embodiments provided later.

[0360] Case 2: The first information includes a third sensed measurement quantity and the timestamp of the third sensed measurement quantity;

[0361] The first device generates a training sample based on the first information and the stored historical data, including:

[0362] The first device generates a third training sample including a third communication measurement quantity and the third sensed measurement quantity;

[0363] Wherein, the third communication measurement quantity includes at least one of the following:

[0364] The time difference between the timestamp of the historical data and the timestamp of the third sensed measurement quantity is less than or equal to the communication measurement quantity of the first time window indicated by the first information;

[0365] The time difference between the timestamp of the historical data and the timestamp of the third sensed measurement quantity is less than or equal to the communication measurement quantity of a predefined second time window.

[0366] It should be noted that in this case, the first information includes a third sensed measurement quantity and the timestamp of the third sensed measurement quantity, which can be understood as follows: The first information can only include sensed measurement-related data (such as the first information only includes the third sensed measurement quantity and the timestamp of the third sensed measurement quantity), and the first information includes both sensed measurement-related data and communication measurement-related data (such as the first information, in addition to including the third sensed measurement quantity and the timestamp of the third sensed measurement quantity, also includes one or more communication measurement quantities (and the timestamps of the communication measurement quantities)).

[0367] In this case, the third device does not perform sample synchronization, and the first device performs sample synchronization based on the first information and historical data to obtain training samples.

[0368] For details, refer to the embodiments provided later.

[0369] Case 3: The first information includes a fourth communication measurement quantity and a time stamp of the fourth communication measurement quantity;

[0370] The first device generates training samples based on the first information and the stored historical data, including:

[0371] The first device generates a fourth training sample including the fourth communication measurement quantity and a fourth perception measurement quantity;

[0372] Wherein, the fourth communication measurement quantity includes at least one of the following:

[0373] The time difference between the time stamp of the historical data and the time stamp of the fourth communication measurement quantity is less than or equal to the perception measurement quantity of the first time window indicated by the first information;

[0374] The time stamp of the historical data is less than or equal to the perception measurement quantity of the first effective time indicated by the first information;

[0375] The time stamp of the historical data is less than or equal to the perception measurement quantity of a predefined second effective time.

[0376] It should be noted that in this case, the first information includes a fourth communication measurement quantity and a time stamp of the fourth communication measurement quantity, which can be understood as follows: The first information may only include communication measurement-related data (such as the first information only includes the fourth communication measurement quantity and the time stamp of the fourth communication measurement quantity), or the first information includes both perception measurement-related data and communication measurement-related data (such as the first information includes, in addition to the fourth communication measurement quantity and the time stamp of the fourth communication measurement quantity, one or more perception measurement quantities (and time stamps of the perception measurement quantities)).

[0377] In this case, the third device does not perform sample synchronization, and the first device performs sample synchronization based on the first information and historical data to obtain training samples.

[0378] For details, refer to the embodiments provided later.

[0379] Case 4: The first device generates training samples based on the first information, including at least one of the following:

[0380] When the first information includes a fifth communication measurement and a fifth sensing measurement, the first device generates a fifth training sample including the fifth communication measurement and the fifth sensing measurement;

[0381] When the first information includes a fifth communication measurement and a third indication, the first device generates a fifth training sample including the fifth communication measurement and a fifth sensing measurement, where the fifth sensing measurement is the sensing measurement corresponding to the third indication;

[0382] When the first information includes a fifth communication measurement, the first device generates a fifth training sample including the fifth communication measurement and a fifth sensing measurement, where the fifth sensing measurement is the sensing measurement of the most recent training sample.

[0383] Optionally, the third indication includes at least one of the following:

[0384] The identifier of the training sample;

[0385] The identifier of the resource used for the communication measurement result;

[0386] The reporting identifier of the communication measurement result;

[0387] The identifier of the resource used for the sensing measurement;

[0388] The reporting identifier of the sensing measurement;

[0389] The identifier of the sensing measurement.

[0390] In this case, the first information includes at least communication measurement-related data, and the third device has performed sample synchronization. The first device obtains the synchronized data based on the first information (or the first information and historical data) to obtain the training sample.

[0391] For details, please refer to the embodiments provided later.

[0392] It should be noted that the sensing measurement (or communication measurement) in each of the above cases is an arbitrary sensing measurement (or communication measurement), not a specific sensing measurement (or communication measurement).

[0393] The first device can also send the collected target data set to other devices, such as the fourth device, the fifth device, or the sixth device. When the first device sends the target data set to other devices, it can also send the identifier of the target data set. When the first device sends the target data set and the identifier of the target data set to other devices, it can also carry target information for indicating the sensing characteristics of the target data set, so that other devices can obtain the sensing characteristics of the target data set through the target information.

[0394] It should be noted that the perceptual characteristics of the target dataset can provide guidance for the subsequent training process and inference process of the target AI model. For example, maintaining the consistency of the data (distribution) characteristics during model training and model inference is beneficial to improving the training effect and inference accuracy of the target AI model.

[0395] In some embodiments, the method further includes:

[0396] The first device sends second information to the fourth device. The second information includes the target dataset and the identifier of the target dataset. The fourth device is a device for model training. The identifier of the target dataset is associated with the perceptual characteristics of the target dataset, or the identifier of the target dataset is associated with the perceptual characteristics and communication characteristics of the target dataset.

[0397] Optionally, the second information further includes target information;

[0398] The target information is used to indicate the perceptual characteristics of the target dataset, or the target information is used to indicate the perceptual characteristics and communication characteristics of the target dataset.

[0399] In some embodiments, the method further includes:

[0400] The first device sends third information to the fifth device. The third information includes the target dataset and the identifier of the target dataset. The fifth device is a device for storing the model training dataset.

[0401] Optionally, the third information further includes target information;

[0402] The target information is used to indicate the perceptual characteristics of the target dataset, or the target information is used to indicate the perceptual characteristics and communication characteristics of the target dataset.

[0403] Optionally, the target information is used to indicate at least one of the following:

[0404] Whether the samples in the target dataset use perceptual measurement quantities;

[0405] The number of perceptual measurement quantities used by the samples in the target dataset;

[0406] The proportion of perceptual measurement quantities used by the samples in the target dataset;

[0407] Whether the number of perceptual measurement quantities used by the samples in the target dataset meets the minimum quantity threshold;

[0408] Whether the proportion of perceptual measurement quantities used by the samples in the target dataset meets the minimum proportion threshold;

[0409] The proportion of the perception measurement quantities used in the samples of the target dataset that meet timeliness;

[0410] The quantity of the perception measurement quantities used in the samples of the target dataset that meet timeliness;

[0411] A first threshold, where the first threshold is a ratio threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in a sample;

[0412] A second threshold, where the second threshold is a ratio threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in a sample.

[0413] The above items reflect the perception characteristics of the dataset from the perspective of the usage of perception measurement quantities.

[0414] Optionally, the target information is used to indicate at least one of the following:

[0415] A dataset identifier, where the dataset corresponding to the dataset identifier includes perception measurement-related data;

[0416] A configuration identifier of the perception measurement quantity;

[0417] An identifier of the perception measurement quantity;

[0418] An identifier of the perception measurement link;

[0419] An identifier of the perception service;

[0420] An identifier of the perception service type;

[0421] The parameter items of the perception measurement quantities used in the samples of the target dataset, where the parameter items include at least one of Doppler of the diameter, time delay, power, and angle;

[0422] The sending device of the perception measurement quantities used in the samples of the target dataset;

[0423] The receiving device of the perception measurement quantities used in the samples of the target dataset;

[0424] The coordinates of the perception measurement quantities used in the samples of the target dataset;

[0425] The source of the perception measurement quantities used in the samples of the target dataset;

[0426] The perception mode of the perception measurement quantities used in the samples of the target dataset;

[0427] The processing level of the perception measurement quantities used in the samples of the target dataset;

[0428] The sensing service of the sensing measurement quantity used by the samples in the target dataset;

[0429] The use of the sensing measurement quantity used by the samples in the target dataset;

[0430] The performance index of the sensing measurement quantity used by the samples in the target dataset;

[0431] The sensing measurement quantity used by the samples in the target dataset is a fixed sensing measurement quantity or a combination of sensing measurement quantities;

[0432] The sensing measurement quantity used by the samples in the target dataset is a variable sensing measurement quantity or a combination of sensing measurement quantities;

[0433] The number of sensing measurement quantities used by the samples in the target dataset is variable;

[0434] The configuration information of the second target signal;

[0435] Wherein, the second target signal is the target signal corresponding to the sensing measurement quantity used by the samples in the target dataset, and the target signal is the signal used for sensing.

[0436] The above items reflect the sensing characteristics of the dataset from the perspective of the sensing configuration situation.

[0437] The above is the first device-side embodiment of the present application. The second device-side embodiment, the third device-side embodiment, the fourth device-side embodiment, the fifth device-side embodiment, and the sixth device-side embodiment of the present application will be described below respectively.

[0438] Figure 4 The flowchart of a dataset collection method provided by the embodiment of the present application is shown. As Figure 4 shown, the dataset collection method includes the following steps:

[0439] Step 401: The second device receives a first request from the first device, and the first request is used to request the sensing measurement-related data required for model training;

[0440] Step 402: The second device performs a first operation, wherein the first operation includes any one of the following:

[0441] The second device sends a target signal based on the first request;

[0442] The second device determines the configuration information of the target signal based on the first request;

[0443] The second device sends the configuration information of the target signal based on the first request;

[0444] The second device sends perception measurement-related data based on the first request;

[0445] Wherein, the target signal is a signal used for perception.

[0446] In the embodiments of the present application, in addition to representing a data request receiving device, the second device can also represent a target signal sending device, that is, the second device and the seventh device can be co-located in the same network node.

[0447] Specifically, when the first operation includes sending a target signal, the second device and the seventh device are co-located; when the first operation includes sending configuration information of the target signal, the device receiving the configuration information of the target signal can be used as the seventh device.

[0448] When the first operation includes sending a target signal, the recipient of the target signal is not unique. For example, the second device can send and receive the target signal by itself, the second device can send the target signal to the first device, the second device can also send the target signal to the neighboring base station, the second device can also send the target signal to the UE near the target UE, and the second device can also send the target signal to the eighth device.

[0449] Correspondingly, when the first operation includes sending configuration information of the target signal, the recipient of the configuration information of the target signal is also not unique.

[0450] In some embodiments, when the first operation includes sending perception measurement-related data, the second device represents both the second device and the third device (i.e., the perception result sending device), that is, the second device and the third device are co-located in the same network node. The second device sends perception measurement-related data based on the first request, which can be that the second device directly sends perception measurement-related data to the first device based on the first request.

[0451] Optionally, the first request includes at least one of the following information:

[0452] Information indicating that the model is in the training stage;

[0453] A dataset identifier, and the dataset corresponding to the dataset identifier includes perception measurement-related data;

[0454] A configuration identifier of the perception measurement quantity;

[0455] An identifier of the perception measurement quantity;

[0456] Configuration information of the perception measurement quantity required for model training;

[0457] Information indicating a number threshold of the perception measurement quantity required for model training;

[0458] Information indicating the source of the perception measurement quantities required for model training;

[0459] Information indicating the types of perception measurement quantities required for model training;

[0460] Information indicating the perception mode of the perception measurement quantities required for model training;

[0461] Information indicating the perception requirements;

[0462] Configuration information of the target signal;

[0463] Information of the sending device of the target signal;

[0464] Information of the receiving device of the target signal;

[0465] Wherein, the target signal is a signal used for perception.

[0466] In addition, the second device can also represent the model training device, that is, the second device and the fourth device are co-located at the same network node. That is to say, the second device can receive the target data set and the identifier of the target data set from the first device, and use the target data set for AI model training.

[0467] For the relevant descriptions of the embodiments of the present application, reference can be made to Figure 3 the relevant descriptions of the method embodiments, and the same technical effects can be achieved. To avoid repetition, no further details are provided here.

[0468] Figure 5 The flowchart of a data set collection method provided by the embodiments of the present application is shown. As Figure 5 shown, the data set collection method includes the following steps:

[0469] Step 501: The third device sends first information to the first device, and the first information includes at least one of communication measurement related data and perception measurement related data.

[0470] In the embodiments of the present application, the third device can perform sample synchronization or not.

[0471] For the case where the third device does not perform sample synchronization, the first information may include at least one of the following:

[0472] Data set identifier, and the data set corresponding to the data set identifier includes perception measurement related data;

[0473] Configuration identifier of the perception measurement quantity;

[0474] Identifier of the perception measurement quantity;

[0475] Identifier of the perception measurement link;

[0476] Identification of the sensing service

[0477] Identification of the sensing service type

[0478] Sensing measurement quantity

[0479] Timestamp of the sensing measurement quantity

[0480] Information of the sending device of the sensing measurement quantity

[0481] Coordinate information of the sensing measurement quantity

[0482] Information for indicating the performance index of the sensing measurement quantity

[0483] Information for indicating the source of the sensing measurement quantity

[0484] Information for indicating the type of the sensing measurement quantity

[0485] Information for indicating the sensing mode of the sensing measurement quantity

[0486] Communication measurement quantity

[0487] Communication measurement resource identification

[0488] Timestamp of the communication measurement quantity

[0489] Information for indicating the first time window

[0490] Information for indicating the first valid time

[0491] Configuration identification of the target signal

[0492] Configuration information of the target signal

[0493] Information of the sending device of the target signal

[0494] Information of the receiving device of the target signal

[0495] Wherein, the target signal is a signal used for sensing.

[0496] For the case where the third device performs sample synchronization, the first information may include at least one of the following:

[0497] The fifth communication measurement quantity and the fifth sensing measurement quantity, and the fifth communication measurement quantity and the fifth sensing measurement quantity are determined to belong to the same training sample;

[0498] The fifth communication measurement quantity and the third indication, and the fifth communication measurement quantity and the sensing measurement quantity indicated by the third indication are determined to belong to the same training sample;

[0499] The fifth communication measurement quantity, and the fifth communication measurement quantity and the perception measurement quantity of the most recent training sample are determined to belong to the same training sample.

[0500] Optionally, the third indication includes at least one of the following:

[0501] The identifier of the training sample;

[0502] The identifier of the resource used for the communication measurement result;

[0503] The reporting identifier of the communication measurement result;

[0504] The identifier of the resource used for the perception measurement quantity;

[0505] The reporting identifier of the perception measurement quantity;

[0506] The identifier of the perception measurement quantity.

[0507] It should be noted that regardless of whether the third device performs sample synchronization, the third device can be referred to as a perception result sending device. For the case where the third device performs sample synchronization, the third device may not send the communication measurement result and the perception result simultaneously in the first information. However, since it can implicitly indicate the perception result it has sent through the third indication, the third device has actually sent the perception result before, that is, the third device is a perception result sending device.

[0508] In some embodiments, the third device can cooperate with the second device, that is, the third device can receive a first request from the first device and send the first information to the first device based on the first request.

[0509] For the relevant descriptions of the embodiments of the present application, reference can be made to Figure 3 the relevant descriptions of the method embodiments, and the same technical effects can be achieved. To avoid repetition, details are not described herein.

[0510] Figure 6 The flowchart of a model training method provided by the embodiments of the present application is shown. As Figure 6 shown, the model training method includes the following steps:

[0511] Step 601: The fourth device receives a second information from the first device, and the second information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and perception measurement-related data;

[0512] Step 602: The fourth device uses the target data set to train a target AI unit.

[0513] Optionally, the perception measurement-related data includes at least one of the following information:

[0514] Perceived measurement quantity;

[0515] Indicator of the perceived measurement quantity;

[0516] Timestamp of the perceived measurement quantity;

[0517] Information on the sending device of the perceived measurement quantity;

[0518] Information on the receiving device of the perceived measurement quantity;

[0519] Coordinate information of the perceived measurement quantity;

[0520] Information for indicating the performance metrics of the perceived measurement quantity;

[0521] Information for indicating the source of the perceived measurement quantity;

[0522] Information for indicating the type of the perceived measurement quantity;

[0523] Information for indicating the perception mode of the perceived measurement quantity;

[0524] Configuration information of the target signal;

[0525] Information on the sending device of the target signal;

[0526] Information on the receiving device of the target signal;

[0527] Wherein, the target signal is a signal used for perception.

[0528] Optionally, the second information further includes target information;

[0529] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[0530] Optionally, the method further includes:

[0531] The fourth device sends fourth information to the fifth device, the fourth information includes the target data set and the identifier of the target data set, and the fifth device is a device storing the model training data set.

[0532] Optionally, the fourth information further includes target information;

[0533] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[0534] Optionally, the method further includes:

[0535] The fourth device sends fifth information to the sixth device, where the fifth information includes the target AI unit, the identifier of the target AI unit, and the identifier of the target data set, and the sixth device is the device storing the target AI unit.

[0536] Optionally, the fifth information further includes target information;

[0537] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[0538] Optionally, the target information is used to indicate at least one of the following:

[0539] Whether a perception measurement quantity is used for the samples in the target data set;

[0540] The quantity of perception measurement quantities used for the samples in the target data set;

[0541] The proportion of perception measurement quantities used for the samples in the target data set;

[0542] Whether the quantity of perception measurement quantities used for the samples in the target data set meets a minimum quantity threshold;

[0543] Whether the proportion of perception measurement quantities used for the samples in the target data set meets a minimum proportion threshold;

[0544] The proportion of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for the samples in the target data set;

[0545] The quantity of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for the samples in the target data set;

[0546] A first threshold, where the first threshold is a proportion threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in a sample;

[0547] A second threshold, where the second threshold is a proportion threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in a sample.

[0548] Optionally, the target information is used to indicate at least one of the following:

[0549] A data set identifier, where the data set corresponding to the data set identifier includes perception measurement-related data;

[0550] A configuration identifier of the perception measurement quantity;

[0551] An identifier of the perception measurement quantity;

[0552] An identifier of the perception measurement link;

[0553] Identification of the sensing service

[0554] Identification of the sensing service type

[0555] Parameter items of the sensing measurement quantity used by the samples in the target dataset, where the parameter items include at least one of Doppler of the diameter, time delay, power, and angle

[0556] Transmitting device of the sensing measurement quantity used by the samples in the target dataset

[0557] Receiving device of the sensing measurement quantity used by the samples in the target dataset

[0558] Coordinates of the sensing measurement quantity used by the samples in the target dataset

[0559] Source of the sensing measurement quantity used by the samples in the target dataset

[0560] Sensing mode of the sensing measurement quantity used by the samples in the target dataset

[0561] Processing level of the sensing measurement quantity used by the samples in the target dataset

[0562] Sensing service of the sensing measurement quantity used by the samples in the target dataset

[0563] Purpose of the sensing measurement quantity used by the samples in the target dataset

[0564] Performance index of the sensing measurement quantity used by the samples in the target dataset

[0565] The sensing measurement quantity used by the samples in the target dataset is a fixed sensing measurement quantity or a combination of sensing measurement quantities

[0566] The sensing measurement quantity used by the samples in the target dataset is a variable sensing measurement quantity or a combination of sensing measurement quantities

[0567] The number of sensing measurement quantities used by the samples in the target dataset is variable

[0568] Configuration information of the second target signal

[0569] Wherein, the second target signal is the target signal corresponding to the sensing measurement quantity used by the samples in the target dataset, and the target signal is a signal used for sensing

[0570] For the relevant descriptions of the embodiments of this application, reference can be made to Figure 3 The relevant descriptions of the method embodiments, and the same technical effects can be achieved. To avoid repetition, no further elaboration will be provided

[0571] Figure 7 The flowchart of a method for collecting a dataset provided by an embodiment of the present application is shown. As Figure 7 shown, the method for collecting a dataset includes the following steps:

[0572] Step 701: The fifth device receives the third information from the first device, or receives the fourth information from the fourth device, where the third information or the fourth information includes the target dataset and the identifier of the target dataset; the target dataset includes communication measurement-related data and perception measurement-related data;

[0573] Step 702: The fifth device stores the target dataset and the identifier of the target dataset.

[0574] Wherein, the first device is the device that collects the target dataset, the fourth device is the device for model training, the identifier of the target dataset is associated with the perception characteristics of the target dataset, or the identifier of the target dataset is associated with the perception characteristics and communication characteristics of the target dataset.

[0575] Optionally, the perception measurement-related data includes at least one of the following information:

[0576] Perception measurement quantity;

[0577] Indicator of the perception measurement quantity;

[0578] Timestamp of the perception measurement quantity;

[0579] Information of the sending device of the perception measurement quantity;

[0580] Information of the receiving device of the perception measurement quantity;

[0581] Coordinate information of the perception measurement quantity;

[0582] Information for indicating the performance index of the perception measurement quantity;

[0583] Information for indicating the source of the perception measurement quantity;

[0584] Information for indicating the type of the perception measurement quantity;

[0585] Information for indicating the perception mode of the perception measurement quantity;

[0586] Configuration information of the target signal;

[0587] Information of the sending device of the target signal;

[0588] Information of the receiving device of the target signal;

[0589] Wherein, the target signal is the signal used for perception.

[0590] Optionally, the third information or the fourth information further includes target information;

[0591] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[0592] Optionally, the target information is used to indicate at least one of the following:

[0593] Whether the samples in the target data set use perception measurement quantities;

[0594] The quantity of perception measurement quantities used by the samples in the target data set;

[0595] The proportion of perception measurement quantities used by the samples in the target data set;

[0596] Whether the quantity of perception measurement quantities used by the samples in the target data set meets the minimum quantity threshold;

[0597] Whether the proportion of perception measurement quantities used by the samples in the target data set meets the minimum proportion threshold;

[0598] The proportion of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used by the samples in the target data set;

[0599] The quantity of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used by the samples in the target data set;

[0600] A first threshold, which is a proportion threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in one sample;

[0601] A second threshold, which is a proportion threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in one sample.

[0602] Optionally, the target information is used to indicate at least one of the following:

[0603] A data set identifier, and the data set corresponding to the data set identifier includes perception measurement-related data;

[0604] A configuration identifier of the perception measurement quantity;

[0605] An identifier of the perception measurement quantity;

[0606] An identifier of the perception measurement link;

[0607] An identifier of the perception service;

[0608] An identifier of the perception service type;

[0609] The parameter items of the perception measurement used for the samples in the target dataset, where the parameter items include at least one of Doppler of the diameter, time delay, power, and angle;

[0610] The transmitting device of the perception measurement used for the samples in the target dataset;

[0611] The receiving device of the perception measurement used for the samples in the target dataset;

[0612] The coordinates of the perception measurement used for the samples in the target dataset;

[0613] The source of the perception measurement used for the samples in the target dataset;

[0614] The perception mode of the perception measurement used for the samples in the target dataset;

[0615] The processing level of the perception measurement used for the samples in the target dataset;

[0616] The perception service of the perception measurement used for the samples in the target dataset;

[0617] The use of the perception measurement used for the samples in the target dataset;

[0618] The performance index of the perception measurement used for the samples in the target dataset;

[0619] The perception measurement used for the samples in the target dataset is a fixed perception measurement or a combination of perception measurements;

[0620] The perception measurement used for the samples in the target dataset is a variable perception measurement or a combination of perception measurements;

[0621] The number of perception measurements used for the samples in the target dataset is variable;

[0622] The configuration information of the second target signal;

[0623] Wherein, the second target signal is the target signal corresponding to the perception measurement used for the samples in the target dataset, and the target signal is the signal used for perception.

[0624] For the relevant descriptions of the embodiments of the present application, reference can be made to Figure 3 the relevant descriptions of the method embodiments, and the same technical effects can be achieved. To avoid repetition, no further elaboration is provided here.

[0625] Figure 8 shows the flowchart of a model storage method provided by an embodiment of the present application. As Figure 8 shown, the model storage method includes the following steps:

[0626] Step 801: The sixth device receives fifth information from the fourth device. The fifth information includes a target AI unit, an identifier of the target AI unit, and an identifier of a target data set. The data set corresponding to the identifier of the target data set includes communication measurement-related data and perception measurement-related data;

[0627] Step 802: The sixth device stores the fifth information.

[0628] Optionally, the perception measurement-related data includes at least one of the following information:

[0629] Perception measurement quantity;

[0630] Indicator of the perception measurement quantity;

[0631] Timestamp of the perception measurement quantity;

[0632] Information of the sending device of the perception measurement quantity;

[0633] Information of the receiving device of the perception measurement quantity;

[0634] Coordinate information of the perception measurement quantity;

[0635] Information indicating the performance index of the perception measurement quantity;

[0636] Information indicating the source of the perception measurement quantity;

[0637] Information indicating the type of the perception measurement quantity;

[0638] Information indicating the perception mode of the perception measurement quantity;

[0639] Configuration information of the target signal;

[0640] Information of the sending device of the target signal;

[0641] Information of the receiving device of the target signal;

[0642] Wherein, the target signal is a signal used for perception.

[0643] Optionally, the method further includes:

[0644] The sixth device associates the identifier of the target AI unit with the identifier of the target data set.

[0645] Optionally, the fifth information further includes target information;

[0646] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[0647] Optionally, the target information is used to indicate at least one of the following:

[0648] whether a perception measurement quantity is used for a sample in the target dataset;

[0649] the quantity of perception measurement quantities used for a sample in the target dataset;

[0650] the proportion of perception measurement quantities used for a sample in the target dataset;

[0651] whether the quantity of perception measurement quantities used for a sample in the target dataset meets a minimum quantity threshold;

[0652] whether the proportion of perception measurement quantities used for a sample in the target dataset meets a minimum proportion threshold;

[0653] the proportion of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for a sample in the target dataset;

[0654] the quantity of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for a sample in the target dataset;

[0655] a first threshold, which is a proportion threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in a sample;

[0656] a second threshold, which is a proportion threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in a sample.

[0657] Optionally, the target information is used to indicate at least one of the following:

[0658] a dataset identifier, and the dataset corresponding to the dataset identifier includes perception measurement-related data;

[0659] a configuration identifier of a perception measurement quantity;

[0660] an identifier of a perception measurement quantity;

[0661] an identifier of a perception measurement link;

[0662] an identifier of a perception service;

[0663] an identifier of a perception service type;

[0664] parameter items of the perception measurement quantities used for a sample in the target dataset, and the parameter items include at least one of Doppler of a diameter, time delay, power, and angle;

[0665] the sending device of the perception measurement quantities used for a sample in the target dataset;

[0666] A receiving device for the perception measurement quantities used by the samples in the target dataset;

[0667] The coordinates of the perception measurement quantities used by the samples in the target dataset;

[0668] The source of the perception measurement quantities used by the samples in the target dataset;

[0669] The perception mode of the perception measurement quantities used by the samples in the target dataset;

[0670] The processing level of the perception measurement quantities used by the samples in the target dataset;

[0671] The perception service of the perception measurement quantities used by the samples in the target dataset;

[0672] The use of the perception measurement quantities used by the samples in the target dataset;

[0673] The performance indicators of the perception measurement quantities used by the samples in the target dataset;

[0674] The perception measurement quantity used by the samples in the target dataset is a fixed perception measurement quantity or a combination of perception measurement quantities;

[0675] The perception measurement quantity used by the samples in the target dataset is a variable perception measurement quantity or a combination of perception measurement quantities;

[0676] The number of perception measurement quantities used by the samples in the target dataset is variable;

[0677] The configuration information of the second target signal;

[0678] Wherein, the second target signal is the target signal corresponding to the perception measurement quantity used by the samples in the target dataset, and the target signal is the signal used for perception.

[0679] For the relevant descriptions of the embodiments of the present application, reference can be made to Figure 3 the relevant descriptions of the method embodiments, and the same technical effects can be achieved. To avoid repetition, no further elaboration will be provided here.

[0680] Specific embodiments are provided below to exemplarily illustrate the interaction process involved in the embodiments of the present application.

[0681] Embodiment 1: The serving base station (referred to as the base station for short) sends a perception reference signal, the UE receives the perception reference signal, and the UE collects training data

[0682] In this embodiment, the UE represents the first device and the eighth device (i.e., the first device and the eighth device are integrated in the UE), and the base station represents the second device and the seventh device (i.e., the second device and the seventh device are integrated in the base station).

[0683] As shown in Figure 9 the following steps are included:

[0684] 1. The base station sends a sensing reference signal to the UE;

[0685] 2. The UE receives the sensing reference signal sent by the base station, obtains sensing measurement quantities, synchronizes the sensing measurement quantities and the CSI measurement results, and generates a data set for model training.

[0686] Before step 1, the UE sends a first request to the base station for requesting the sensing measurement quantities required for model training.

[0687] The first request can reflect the UE's requirements, which can be implicitly reflected or explicitly reflected. The first request may include at least one of the following:

[0688] AI lifecycle management process indication, for example: indicating that the model is in the training stage;

[0689] Data set identifier, which can implicitly indicate, for example: the configuration of a certain sensing reference signal or sensing measurement quantity used in the previous training process, the reporting configuration of the sensing measurement quantity used in the previous training process; (implicit indication)

[0690] Identifier of a predefined sensing measurement quantity, which can implicitly indicate what kind of sensing measurement quantity or combination of sensing measurement quantities is required; (implicit indication)

[0691] Configuration identifier of a predefined sensing measurement quantity, which can implicitly indicate the processing level, configuration density, etc. of the required sensing measurement quantity; (implicit indication)

[0692] Information indicating the source of the required sensing measurement quantity, for example, measurement from a sensing reference signal (i.e., integrated communication and sensing), or from sensor sensing, or which sensing method from measurement from a sensing reference signal (i.e., integrated communication and sensing) (i.e., the 6 basic sensing methods in related technology 2), or which type of sensor from sensor sensing;

[0693] Information indicating the sending device of the sensing reference signal (which can also be understood as the source of the sensing measurement quantity), for example, serving base station, neighboring base station, nearby UE;

[0694] Information indicating the receiving device of the sensing reference signal (which can also be understood as the source of the sensing measurement quantity), for example, serving base station, neighboring base station, nearby UE, target UE;

[0695] Information indicating the type of the required sensing measurement quantity (see explanation 2 later);

[0696] Configuration information of the sensing reference signal (see Explanation 3 below).

[0697] Information indicating the total number of required sensing measurement quantities.

[0698] Information indicating the minimum number of required sensing measurement quantities.

[0699] Information indicating the sensing requirement (see Explanation 4 below).

[0700] In this embodiment, for the combination of sensing measurement quantities corresponding to the identification of the sensing measurement quantities, refer to Table 3.

[0701] Table 3

[0702] Identification of the sensed measurement quantity Corresponding sensed measurement quantity used 1 Doppler, delay, power / amplitude / energy, angle, etc. of the path 2 Doppler, delay, power / amplitude / energy of the path 3 Doppler, delay of the path 4 Doppler, angle, Doppler resolution, angle resolution of the path

[0703] In this embodiment, when the serving base station of the UE only represents the second device and does not represent the seventh device, that is, the transmitting device of the sensing reference signal is not the serving base station of the UE, how does the serving base station of the UE coordinate the configuration of the sensing reference signal with the seventh device? The following provides three coordination methods:

[0704] Method 1: The seventh device determines the configuration of the sensing reference signal by itself and then tells the serving base station of the UE. In this method, the serving base station of the UE receives the configuration of the sensing reference signal sent by the seventh device.

[0705] Method 2: The serving base station of the UE determines the corresponding configuration of the sensing reference signal according to the first request and then notifies (or forwards) it to the seventh device. In this method, the serving base station of the UE receives the feedback from the seventh device on the configuration of the sensing reference signal. For example: The seventh device agrees to the configuration of the sensing reference signal; or, the seventh device does not agree to the configuration of the sensing reference signal and negotiates the configuration of the sensing reference signal with the serving base station of the UE.

[0706] Method 3: The seventh device first determines the configuration of the sensing reference signal and then sends it to the serving base station of the UE for negotiation. In this method, the serving base station of the UE receives the configuration of the sensing reference signal sent by the seventh device and, based on the first request, feeds back the configuration difference or the desired configuration to the seventh device.

[0707] In step 2, sample synchronization can be understood as generating training samples. A data set can include multiple training samples, and a training sample can include model input data and true value / label data. Therefore, data set collection can be understood as the collection of training samples. The collection of training samples can be non - continuous and can last for a long time.

[0708] In step 2, the UE performs sample synchronization on the sensing measurement quantities and the CSI measurement results, including the following two methods:

[0709] Method 1:

[0710] The UE determines whether to combine the perception measurement quantity and the CSI measurement quantity as a training sample based on the timestamp of the perception reference signal measurement, the effective duration of the perception measurement quantity, the CSI measurement quantity (such as L1-RSRP), and the timestamp of the CSI measurement.

[0711] Specifically, if the timestamp of the last CSI measurement corresponding to the true value / label is earlier than the timestamp of the perception measurement + the effective duration, the UE combines the perception measurement quantity and the CSI measurement result (including the model input and the true value) as a training sample.

[0712] Method 2:

[0713] The UE determines whether to combine the perception measurement quantity and the CSI measurement quantity as a training sample based on the perception reference signal measurement quantity, the perception control message, the CSI measurement quantity, and the timestamp of the CSI measurement.

[0714] Specifically, it includes at least one of the following:

[0715] If the UE obtains the perception measurement quantity and does not receive a message indicating the invalidation of the perception result from the base station, the UE combines the perception measurement quantity and the CSI measurement quantity as a training sample;

[0716] If the timestamp of the last CSI measurement corresponding to the true value / label is earlier than the invalidation time indicated by the first indication, the UE combines the perception measurement quantity and the CSI measurement quantity as a training sample. Among them, the first indication is used to indicate the invalidation of the perception result or includes the time difference of the invalidation of the perception result. The time difference can also be predefined.

[0717] In the above Method 1, the UE needs to obtain the timeliness (i.e., the effective duration) of the perception measurement quantity to synchronize the perception measurement quantity and the CSI measurement result for sampling. The methods for the UE to obtain the timeliness include the following three:

[0718] Method 1: Initially configure or predefine an effective duration of the perception measurement quantity (such as 1 s, which can be implicitly indicated or explicitly indicated), and no additional effective duration is carried when sending the perception reference signal each time;

[0719] Method 2: Initially configure or predefine an effective duration of the perception measurement quantity, and carry an adjustment value (such as -100 ms, which can be implicitly indicated or explicitly indicated) when sending the perception reference signal each time;

[0720] Method 3: Carry the absolute effective duration (such as 900 ms, which can be implicitly indicated or explicitly indicated) when sending the perception reference signal each time.

[0721] After step 2, the UE can generate an identifier for the data set based on the sensing characteristics of the data set. That is to say, the UE associates the identifier of the data set with the sensing characteristics of the data set. Alternatively, the UE can generate an identifier for the data set based on the sensing characteristics and communication characteristics of the data set. That is to say, the UE associates the identifier of the data set with the sensing characteristics and communication characteristics of the data set.

[0722] The UE can also send the collected data set to other devices, such as the fourth device, the fifth device, or the sixth device. When the UE sends the data set to other devices, it can also send the identifier of the data set.

[0723] Specifically, the UE can send the data set and the identifier of the data set to the fourth device, and the fourth device performs model training to generate a target AI unit. The fourth device can associate the identifier of the generated target AI unit with the identifier of the data set. The fourth device can also send the target AI unit, the identifier of the target AI unit, and the identifier of the data set to the sixth device.

[0724] The UE can send the data set and the identifier of the data set to the fifth device, and the fifth device stores the data set and the identifier of the data set.

[0725] When the UE sends the data set and the identifier of the data set to other devices, it can also carry target information for indicating the sensing characteristics of the data set, so that other devices can obtain the sensing characteristics of the data set through the target information.

[0726] The target information can indicate, for example, at least one of the following:

[0727] Whether the samples in the data set use sensing measurement quantities;

[0728] The number of sensing measurement quantities used by the samples in the data set;

[0729] The proportion of the sensing measurement quantities used by the samples in the data set;

[0730] Whether the number of sensing measurement quantities used by the samples in the data set meets the minimum quantity threshold;

[0731] Whether the proportion of the sensing measurement quantities used by the samples in the data set meets the minimum proportion threshold;

[0732] The proportion of the sensing measurement quantities used by the samples in the data set that meet the timeliness;

[0733] The number of sensing measurement quantities used by the samples in the data set that meet the timeliness;

[0734] A first threshold, where the first threshold is the proportion threshold between the number of valid sensing measurement quantities in a sample and the total number of sensing measurement quantities;

[0735] Second threshold, where the second threshold is a ratio threshold between the number of actually used perception measurement quantities in a sample and the total number of the maximum supportable perception measurement quantities.

[0736] The above items reflect the perception characteristics of the data set from the perspective of the usage of perception measurement quantities.

[0737] Examples are provided below to illustrate the possible situations of using perception measurement quantities in a sample.

[0738] Example 1: As Figure 10 shown, if the first timestamp corresponding to step 5 is earlier than the perception result expiration timestamp corresponding to step 6, and the timestamp for obtaining the perception measurement quantity in step 2 is earlier than the second timestamp corresponding to step 4, then the (sample, perception measurement quantity) is marked as valid. Among them, the first timestamp is the timestamp of the last CSI measurement corresponding to the true value / label, and the second timestamp is the earliest timestamp of the CSI reference signal measurement corresponding to the model input.

[0739] Suppose the sample includes N perception measurement quantities, and M perception measurement quantities are valid.

[0740] If M / N >= the first threshold, then the sample is marked as effectively using the perception result. The first threshold can default to 1 or be configured as a number less than 1.

[0741] Example 2: As Figure 11 shown, if the timestamp for obtaining the perception measurement quantity in step 2 is later than the second timestamp corresponding to step 4, then the (sample, perception measurement quantity) is marked as unused. Among them, the second timestamp is the earliest timestamp of the CSI reference signal measurement corresponding to the model input.

[0742] Suppose the sample includes N supportable perception measurement quantities, and L perception measurement quantities are used.

[0743] If L > 0, then the sample is marked as having used perception measurement quantities.

[0744] If L / N >= the second threshold, then the sample is marked as having used the minimum number of perception measurement quantities. The second threshold can default to 0 or be configured as a number greater than 0 or less than 1.

[0745] Example 3: As Figure 12 shown, if the perception result expiration timestamp corresponding to step 6 is earlier than the first timestamp corresponding to step 5, then the (sample, perception measurement quantity) is marked as invalid (or unused). Among them, the first timestamp is the timestamp of the last CSI measurement corresponding to the true value / label.

[0746] The target information can also indicate at least one of the following:

[0747] Parameter items of the perception measurement used for samples in the dataset, such as Doppler, time delay, power, or angle of the path, etc.;

[0748] Transmitting device of the perception measurement used for samples in the dataset;

[0749] Receiving device of the perception measurement used for samples in the dataset;

[0750] Coordinates of the perception measurement used for samples in the dataset;

[0751] Source of the perception measurement used for samples in the dataset;

[0752] Perception mode of the perception measurement used for samples in the dataset;

[0753] Processing level of the perception measurement used for samples in the dataset;

[0754] Perception service of the perception measurement used for samples in the dataset;

[0755] Purpose of the perception measurement used for samples in the dataset;

[0756] Performance indicators of the perception measurement used for samples in the dataset;

[0757] The perception measurement used for samples in the dataset is a fixed perception measurement or a combination of perception measurements;

[0758] The perception measurement used for samples in the dataset is a variable perception measurement or a combination of perception measurements;

[0759] The number of perception measurements used for samples in the dataset is variable;

[0760] Configuration information of the perception reference signal corresponding to the perception measurement used for samples in the dataset.

[0761] The above items reflect the perception characteristics of the dataset from the perspective of the perception configuration.

[0762] It should be noted that the perception characteristics of the dataset can provide guidance for subsequent model training processes and model inference processes. For example, maintaining the consistency of data (distribution) characteristics during model training and model inference, which is beneficial to improving the training effect of the AI model and the inference accuracy of the model.

[0763] Embodiment 2: The serving base station (referred to as the base station for short) transmits a perception reference signal, the UE receives the perception reference signal and feeds it back to the base station, and the base station collects training data

[0764] In this embodiment, the base station represents the first device and the seventh device (i.e., the first device and the seventh device are co-located in the base station), and the UE represents the third device and the eighth device (i.e., the third device and the eighth device are co-located in the UE).

[0765] As Figure 13 shown, it includes the following steps:

[0766] 1. The base station sends a sensing reference signal to the UE;

[0767] 2. The UE receives the sensing reference signal sent by the base station, performs sensing measurements, obtains sensing measurement quantities or sensing results, and the UE can also perform CSI measurements to obtain CSI measurement results;

[0768] 3. The UE reports at least one of the following information to the base station:

[0769] Sensing measurement quantities;

[0770] CSI measurement quantities;

[0771] Timestamps of sensing measurement quantities;

[0772] Timestamps of CSI measurement quantities;

[0773] Associated sensing measurement quantity indications;

[0774] Timestamps of associated sensing measurement quantities.

[0775] Figure 13 In, step 3 includes 3a and 3b, where 3a is the reporting of sensing measurement quantities and 3b is the reporting of CSI measurement quantities.

[0776] 4. The base station generates a data set for model training based on the information reported by the UE (or the information reported by the UE and the stored historical data).

[0777] After step 4, the base station generates an identifier for the data set according to the sensing characteristics of the data set.

[0778] In this embodiment, the information reported by the UE to the base station can be information after sample synchronization (or sample alignment), that is, the sample synchronization is performed by the UE; the information reported by the UE to the base station can also be information without sample synchronization, that is, the sample synchronization is performed by the base station.

[0779] In the case where the sample synchronization is performed by the UE, the information reported by the UE can include at least one of the following:

[0780] Sensing measurement quantities and CSI measurement quantities (including model inputs and true values, the same below) that have been time-aligned and can be used for one training sample;

[0781] After time alignment, the CSI measurement of a training sample + a third indication, where the third indication is used to indicate the sensing measurement that has been sent to the base station; wherein, the third indication may include a sample indication (such as an identifier) that has been sent to the base station, or a CSI measurement indication (such as a resource identifier / CSI reporting identifier) that has been sent to the base station, or a sensing measurement indication (such as an identifier) that has been sent to the base station;

[0782] The CSI measurement of a training sample, without a third indication. In this case, the sensing measurement of the most recent training sample is used by default.

[0783] When sample synchronization is performed by the base station, the information reported by the UE may include at least one of the following:

[0784] Sensing measurement and CSI measurement. When the time difference between the received times of a certain sensing measurement and a certain CSI measurement reported by the UE is less than a certain time window, the base station may consider that the sensing measurement and the CSI measurement belong to the same training sample; or, when the difference between the timestamps of a certain sensing measurement reported by the UE and a certain CSI measurement that the base station has received (or has stored) is within a certain time window, the base station may consider that the sensing measurement and the CSI measurement belong to the same training sample; or, when the difference between the timestamps of a certain CSI measurement reported by the UE and a certain sensing measurement that the base station has received (or has stored) is within a certain time window, the base station may consider that the sensing measurement and the CSI measurement belong to the same training sample;

[0785] CSI measurement and the timestamp of the CSI measurement. When the difference between the timestamp of a certain CSI measurement and the timestamp of a received (or stored) sensing measurement is within a certain time window, or when the received sensing measurement is within the valid time, the base station may consider that the CSI measurement and the received sensing measurement belong to the same training sample;

[0786] Sensing measurement and the timestamp of the sensing measurement. When the difference between the timestamp of a certain sensing measurement and the timestamp of a received (or stored) CSI measurement is within a certain time window, the base station may consider that the sensing measurement and the received CSI measurement belong to the same training sample.

[0787] The above time window may be a predefined time window or a time window reported by the UE.

[0788] Embodiment 3: The serving base station (referred to as the base station for short) sends the sensing measurement of the sensor to the UE, and the UE collects training data

[0789] In this embodiment, the UE represents the first device, and the base station represents the second device and the third device (i.e., the second device and the third device are co-located in the base station).

[0790] As Figure 14 shown, it includes the following steps:

[0791] 1. The base station sends sensor-based sensing measurement quantities to the UE;

[0792] 2. The UE receives the sensing measurement quantities of the sensor sent by the base station, and synchronizes the sensing measurement quantities and the CSI measurement results to generate a data set for model training.

[0793] The sensing measurement quantities of the sensor may also include some auxiliary information (contained in the first information), such as:

[0794] Data set identifier, which can implicitly indicate, for example, the configuration of a certain sensing reference signal or sensing measurement quantity used in the previous training process, and the reporting configuration of the sensing measurement quantity used in the previous training process; (implicit indication)

[0795] Pre-defined identifier of the sensing measurement quantity, which can implicitly indicate what kind of sensing measurement quantity or combination of sensing measurement quantities; (implicit indication)

[0796] Pre-defined configuration identifier of the sensing measurement quantity, which can implicitly indicate the processing level, configuration density, etc. of the sensing measurement quantity; (implicit indication)

[0797] Information indicating the source of the sensing measurement quantity, for example, measurement from the sensing reference signal (i.e., integrated communication and sensing), or from sensor sensing, or which sensing method from the measurement of the sensing reference signal (i.e., integrated communication and sensing), or which type of sensor from sensor sensing;

[0798] Information indicating the sending device of the sensing reference signal (which can also be understood as the source of the sensing measurement quantity), for example, serving base station, neighboring base station, nearby UE;

[0799] Information indicating the receiving device of the sensing reference signal (which can also be understood as the source of the sensing measurement quantity), for example, serving base station, neighboring base station, nearby UE, target UE;

[0800] Information indicating the type of the sensing measurement quantity;

[0801] Configuration information of the sensing reference signal;

[0802] Information indicating the total number of the sensing measurement quantities;

[0803] Information indicating the minimum number of the sensing measurement quantities.

[0804] Before step 1, the UE sends a first request to the base station to request the sensing measurement quantities required for model training. (Same as Embodiment 1)

[0805] In step 2, the method for the UE to synchronize the sensing measurement quantities and the CSI measurement results is similar to that in Embodiment 1, including the following two ways:

[0806] Way 1:

[0807] Based on the measurement timestamp or reception timestamp of the sensing measurement quantities of the sensor, the effective duration of the sensing measurement quantities, the CSI measurement quantities (such as Layer 1 RSRP (L1-RSRP)), and the timestamp of the CSI measurement, the UE determines whether to combine the sensing measurement quantities and the CSI measurement quantities as a training sample.

[0808] Specifically, it includes one of the following:

[0809] If the timestamp of the last CSI measurement corresponding to the ground truth / label is earlier than the measurement timestamp of the sensing measurement quantities + the effective duration, the UE combines the sensing measurement quantities and the CSI measurement results (including the model input and the ground truth) as a training sample.

[0810] If the timestamp of the last CSI measurement corresponding to the ground truth / label is earlier than the reception timestamp of the sensing measurement quantities + the effective duration, the UE combines the sensing measurement quantities and the CSI measurement results (including the model input and the ground truth) as a training sample.

[0811] Way 2:

[0812] Based on the sensing measurement quantities of the sensor, the sensing control message, the CSI measurement quantities, and the timestamp of the CSI measurement, the UE determines whether to combine the sensing measurement quantities and the CSI measurement quantities as a training sample.

[0813] Specifically, it includes at least one of the following:

[0814] If the UE obtains the sensing measurement quantities and does not receive a message indicating the invalidation of the sensing result from the base station, the UE combines the sensing measurement quantities and the CSI measurement quantities as a training sample.

[0815] If the timestamp of the last CSI measurement corresponding to the ground truth / label is earlier than the invalidation time indicated by the first indication, the UE combines the sensing measurement quantities and the CSI measurement quantities as a training sample. Herein, the first indication is used to indicate the invalidation of the sensing result or includes the time difference of the invalidation of the sensing result. The time difference can also be predefined.

[0816] In the above Method 1, the UE needs to obtain the timeliness (i.e., the effective duration) of the sensing measurement quantity to synchronize the samples of the sensing measurement quantity and the CSI measurement result. (Same as Figure 9 the illustrated Embodiment 1)

[0817] The rest is the same as Figure 9 the illustrated Embodiment 1. To avoid repetition, it will not be elaborated here.

[0818] Embodiment 4: The UE sends the sensing measurement quantity of the sensor to the serving base station (abbreviated as the base station), and the base station collects training data

[0819] In this embodiment, the base station represents the first device, and the UE represents the third device.

[0820] As Figure 15 shown, it includes the following steps:

[0821] 1. The base station sends the reporting configuration of the sensing measurement quantity of the sensor to the UE;

[0822] 2. The UE reports to the base station at least one of the following information:

[0823] The sensing measurement quantity;

[0824] The CSI measurement quantity;

[0825] The timestamp of the sensing measurement quantity;

[0826] The timestamp of the CSI measurement quantity;

[0827] The associated sensing measurement quantity indication;

[0828] The timestamp of the associated sensing measurement quantity.

[0829] Figure 15 In, step 2 includes 2a and 2b, where 2a is the reporting of the sensing measurement quantity of the sensor, and 2b is the reporting of the CSI measurement quantity.

[0830] 3. The base station generates a data set for model training based on the information reported by the UE (or the information reported by the UE and the stored historical data).

[0831] After step 3, the base station generates an identifier for the data set according to the sensing characteristics of the data set.

[0832] In this embodiment, the information reported by the UE to the base station can be the information after sample synchronization (or sample alignment), that is, the sample synchronization is performed by the UE; the information reported by the UE to the base station can also be the information without sample synchronization, that is, the sample synchronization is performed by the base station. (Same as Figure 13 the illustrated Embodiment 2)

[0833] Embodiment 5: The first base station (serving base station) sends a sensing reference signal, the second base station (neighboring cell base station) receives the sensing reference signal, the first base station sends the sensing measurement quantity to the UE, and the UE collects training data

[0834] In this embodiment, the UE represents the first device, the second base station represents the third device and the eighth device (i.e., the third device and the eighth device are co-located in the second base station), and the first base station represents the third device and the seventh device. The second device may be the first base station or a core network function (such as AMF, LMF).

[0835] As Figure 16 shown, it includes the following steps:

[0836] 1. The first base station sends a sensing reference signal to the second base station;

[0837] 2. The second base station obtains a sensing measurement quantity or a sensing result based on the sensing reference signal;

[0838] 3. The second base station sends at least one of the following information to the first base station:

[0839] Sensing measurement quantity;

[0840] Timestamp of the sensing measurement quantity;

[0841] Valid duration of the sensing measurement quantity;

[0842] Resource indication (such as identification) of the sensing measurement quantity;

[0843] Other relevant information of the sensing measurement quantity.

[0844] 4. The first base station sends at least one of the following information to the UE:

[0845] Sensing measurement quantity;

[0846] Timestamp of the sensing measurement quantity;

[0847] Valid duration of the sensing measurement quantity;

[0848] Indication of the sensing reference signal receiving device (in this embodiment, it is to indicate the second base station, and this indication can be an implicit indication or an explicit indication);

[0849] Resource indication (such as identification) of the sensing measurement quantity;

[0850] Other relevant information of the sensing measurement quantity.

[0851] 5. The UE synchronizes the sensing measurement quantity and the CSI measurement result based on the information sent by the first base station, and generates a data set for model training.

[0852] In this embodiment, other relevant information of the sensed measurement quantity (included in the first information or the target information) includes, for example, at least one of the following:

[0853] Dataset identifier, which can implicitly indicate, for example: the configuration of a certain sensed reference signal or sensed measurement quantity used in the previous training process, the reporting configuration of the sensed measurement quantity used in the previous training process; (implicit indication)

[0854] Pre-defined identifier of the sensed measurement quantity, which can implicitly indicate what kind of sensed measurement quantity or what combination of sensed measurement quantities; (implicit indication)

[0855] Pre-defined configuration identifier of the sensed measurement quantity, which can implicitly indicate the processing level, configuration density, etc. of the sensed measurement quantity; (implicit indication)

[0856] Sensed measurement link identification information (used to distinguish from which sensed link the sensed measurement result comes, which device sends and which device receives);

[0857] Sensing mode information (monostatic sensing mode or bistatic sensing mode, or at least one of the 6 sensing modes);

[0858] Sensed signal configuration identification information (used to distinguish the measurement of which sensed signal the sensed measurement result comes from);

[0859] Sensing service information (such as sensing service ID);

[0860] Sensing service type information (such as sensing service type ID);

[0861] Data subscription ID;

[0862] Measurement result usage information (such as communication, sensing, communication and sensing, for AI inference, training, etc.);

[0863] Device information of the sensed measurement (such as UE ID, device location, device orientation, device movement information);

[0864] Coordinate information of the measurement result (the value of the measurement quantity), for example, whether it is the result based on the local coordinate system or the global coordinate system, the description information of the local coordinate system, such as the rotation angles relative to the global coordinate system: α (bearing angle), β (down tilt angle), and γ (tilt angle);

[0865] Performance index information corresponding to the measurement result, such as resolution (which can be time delay resolution, ranging resolution, angle measurement resolution, Doppler resolution, speed measurement resolution, imaging resolution, etc., that is, the granularity of the value of the reported measurement quantity), SINR, sensing SINR (the ratio of the power of the path associated with the sensing target to the power of noise and interference), etc.

[0866] The performance metrics corresponding to the sensing results (e.g., sensing SINR) can be used as model inputs together with the results of sensing measurements. For example, in addition to the Doppler value of the path on the Doppler spectrum being used as an input for training / inference, the ratio of the power of this path to the power of noise and interference is also used as an input to represent the accuracy or reliability of this Doppler value.

[0867] Before step 1, the UE sends a first request to the first base station to request the sensing measurement quantities required for model training.

[0868] (Same as Figure 9 Embodiment 1 shown)

[0869] After the first base station receives the first request and before step 1, the first base station sends a sensing reference signal configuration to the second base station.

[0870] In step 5, the method for the UE to synchronize the sensing measurement quantities and the CSI measurement results is similar to that in Embodiment 1 and includes the following two methods:

[0871] Method 1:

[0872] The UE determines whether to combine the sensing measurement quantities and the CSI measurement quantities as a training sample based on the measurement timestamp or reception timestamp of the sensing measurement quantities, the effective duration of the sensing measurement quantities, the CSI measurement quantity (e.g., L1-RSRP), and the timestamp of the CSI measurement.

[0873] Specifically, it includes one of the following:

[0874] If the timestamp of the last CSI measurement corresponding to the ground truth / label is earlier than the measurement timestamp of the sensing measurement quantities + the effective duration, the UE combines the sensing measurement quantities and the CSI measurement results (including the model input and the ground truth) as a training sample;

[0875] If the timestamp of the last CSI measurement corresponding to the ground truth / label is earlier than the reception timestamp of the sensing measurement quantities + the effective duration, the UE combines the sensing measurement quantities and the CSI measurement results (including the model input and the ground truth) as a training sample.

[0876] Method 2:

[0877] The UE determines whether to combine the sensing measurement quantities and the CSI measurement quantities as a training sample based on the sensing measurement quantities, the sensing control message, the CSI measurement quantity, and the timestamp of the CSI measurement.

[0878] Specifically, it includes at least one of the following:

[0879] If the UE obtains the sensing measurement quantity and does not receive a message indicating the invalidation of the sensing result from the base station, the UE combines the sensing measurement quantity with the CSI measurement quantity as a training sample;

[0880] If the timestamp of the last CSI measurement corresponding to the true value / label is earlier than the invalidation time indicated by the first indication, the UE combines the sensing measurement quantity with the CSI measurement quantity as a training sample. Among them, the first indication is used to indicate the invalidation of the sensing result, or includes the time difference of the invalidation of the sensing result. The time difference can also be predefined.

[0881] In the above method 1, the UE needs to obtain the timeliness (i.e., the effective duration) of the sensing measurement quantity to synchronize the samples of the sensing measurement quantity and the CSI measurement result. (Same as Figure 9 the shown Embodiment 1)

[0882] The rest is the same as Figure 9 the shown Embodiment 1. To avoid repetition, it will not be elaborated here.

[0883] Embodiment 6: The first base station (serving base station) sends a sensing reference signal, the second base station (neighboring base station) receives the sensing reference signal, the second base station sends the sensing measurement quantity to the UE, and the UE performs training data collection

[0884] In this embodiment, the UE represents the first device, the second base station represents the third device and the eighth device (i.e., the third device and the eighth device are co-located in the second base station), and the first base station represents the seventh device.

[0885] In this embodiment, the second device can be the first base station or a core network function (such as AMF, LMF).

[0886] In this embodiment, the UE has also established a connection with the second base station. Therefore, the second base station can directly send the sensing measurement quantity to the UE.

[0887] It should be noted that how does the second base station know to send the sensing measurement quantity to the UE? Specifically, when the first base station sends the sensing reference signal to the second base station, or when configuring the sensing reference signal, it is necessary to indicate the receiving device of the sensing measurement quantity, that is, the UE.

[0888] As Figure 17 shown, it includes the following steps:

[0889] 1. The first base station sends a sensing reference signal to the second base station;

[0890] 2. The second base station obtains the sensing measurement quantity or the sensing result based on the sensing reference signal;

[0891] 3. The second base station sends at least one of the following information to the UE:

[0892] Perceived measurement quantity;

[0893] Timestamp of the perceived measurement quantity;

[0894] Valid duration of the perceived measurement quantity;

[0895] Other relevant information of the perceived measurement quantity;

[0896] Indicator of the perceived reference signal receiving device (in this embodiment, it is an indicator of the second base station, and this indicator can be an implicit indicator or an explicit indicator).

[0897] 4. The UE synchronizes the perceived measurement quantity and the CSI measurement result based on the information sent by the second base station to generate a dataset for model training.

[0898] Before step 1, the UE sends a first request to the first base station to request the perceived measurement quantity required for model training.

[0899] (Same as Embodiment 1)

[0900] After the first base station receives the first request and before step 1, the first base station sends a perceived reference signal configuration to the second base station.

[0901] The rest is the same as Figure 16 Embodiment 5 shown. To avoid repetition, it will not be elaborated here.

[0902] Embodiment 7: The serving base station (abbreviated as the base station) sends a perceived reference signal, the second UE receives the perceived reference signal, the base station sends the perceived measurement quantity to the target UE (or the first UE), and the target UE collects training data

[0903] In this embodiment, the target UE represents the first device, the base station represents the third device and the seventh device (i.e., the third device and the seventh device are co-located in the base station), and the second UE represents the third device and the eighth device.

[0904] In this embodiment, the second device can be the first base station or a core network function (such as AMF, LMF).

[0905] In this embodiment, the second UE is a UE near the target UE.

[0906] As Figure 18 shown, it includes the following steps:

[0907] 1. The base station sends a perceived reference signal to the second UE;

[0908] 2. The second UE obtains a perceived measurement quantity or a perceived result based on the perceived reference signal;

[0909] 3. The second UE sends at least one of the following information to the base station:

[0910] Perceived measurement quantity;

[0911] Timestamp of the perceived measurement quantity;

[0912] Valid duration of the perceived measurement quantity;

[0913] Resource indication (such as identification) of the perceived measurement quantity;

[0914] Other relevant information of the perceived measurement quantity.

[0915] 4. The base station sends at least one of the following information to the target UE:

[0916] Perceived measurement quantity;

[0917] Timestamp of the perceived measurement quantity;

[0918] Valid duration of the perceived measurement quantity;

[0919] Resource indication (such as identification) of the perceived measurement quantity;

[0920] Other relevant information of the perceived measurement quantity;

[0921] Indication of the perceived reference signal receiving device (in this embodiment, it is to indicate the second UE, and this indication can be an implicit indication or an explicit indication).

[0922] 5. The target UE synchronizes the perceived measurement quantity and the CSI measurement result based on the information sent by the base station to generate a data set for model training.

[0923] It should be noted that the serving base station can directly receive information from the second UE or be forwarded through other devices. For example, if the second UE is a neighboring cell UE, the second UE first sends the information to the neighboring cell base station, and then the neighboring cell base station forwards it to the serving base station.

[0924] Before step 1, the target UE sends a first request to the base station to request the perceived measurement quantity required for model training.

[0925] (Same as Figure 9 the embodiment 1 shown)

[0926] After the first base station receives the first request and before step 1, the first base station sends a perceived reference signal configuration to the second UE.

[0927] In this embodiment, when the serving base station of the target UE only represents the second device and does not represent the seventh device, that is, the sending device of the perceived reference signal is not the serving base station of the target UE, how does the serving base station of the target UE coordinate the configuration of the perceived reference signal with the seventh device? (Same as embodiment 1)

[0928] Specifically, refer to Figure 19 and Figure 20 . Figure 19 In this case, the second UE belongs to the first base station (i.e., the serving base station), the target UE represents the first device, the first base station represents the third device, the second UE represents the third device and the eighth device, and the second base station represents the seventh device. Figure 20 In this case, the second UE belongs to the second base station, the target UE represents the first device, the first base station represents the third device, the second UE represents the third device and the eighth device, and the second base station represents the third device and the seventh device.

[0929] The rest is the same as the Figure 16 embodiment 5 shown. To avoid repetition, it will not be elaborated here.

[0930] Embodiment 8: The serving base station (also known as the first base station) sends a sensing reference signal, the second UE receives the sensing reference signal, the second UE sends the sensing measurement quantity to the target UE (or the first UE), and the target UE performs training data collection

[0931] In this embodiment, the target UE represents the first device, the serving base station represents the seventh device, and the second UE represents the third device and the eighth device.

[0932] In this embodiment, the second device can be the first base station or a core network function (such as AMF, LMF).

[0933] In this embodiment, the second UE is a UE near the target UE, the target UE has established a sidelink with the second UE, and the second UE can directly send the sensing measurement quantity to the target UE.

[0934] The main process of this embodiment is similar to that of Embodiment 6, except that the device receiving the sensing reference signal is different.

[0935] In this embodiment, the serving base station sends a sensing reference signal to the second UE. In Embodiment 6, the serving base station sends a sensing reference signal to the second base station.

[0936] As Figure 21 shown, it includes the following steps:

[0937] 1. The base station sends a sensing reference signal to the second UE;

[0938] 2. The second UE obtains a sensing measurement quantity or a sensing result based on the sensing reference signal;

[0939] 3. The second UE sends at least one of the following information to the target UE:

[0940] The sensing measurement quantity;

[0941] The timestamp of the sensing measurement quantity;

[0942] The effective duration of the sensed measurement quantity;

[0943] Other relevant information of the sensed measurement quantity;

[0944] The indication of the device receiving the sensing reference signal (in this embodiment, it is to indicate the second UE, and this indication can be an implicit indication or an explicit indication).

[0945] 4. Based on the information sent by the second UE, the target UE synchronizes the sensed measurement quantity and the CSI measurement result to generate a data set for model training.

[0946] The rest is the same as Figure 17 Embodiment 6 shown, and for the sake of avoiding repetition, it will not be elaborated here.

[0947] Embodiment 9: The base station sends a sensing reference signal, the base station receives the sensing reference signal, the base station sends the sensed measurement quantity to the UE, and the UE collects training data

[0948] In this embodiment, the UE represents the first device, and the base station represents the third device, the seventh device, and the eighth device.

[0949] In this embodiment, the second device can be a base station or a core network function (such as AMF, LMF).

[0950] In this embodiment, the base station performing the self-transmission and self-reception operation may be the serving base station or the neighboring cell base station. In addition, the other nearby UEs may also perform the self-transmission and self-reception operation, and this embodiment will not make specific descriptions about this.

[0951] As Figure 22 shown, it includes the following steps:

[0952] 1. The base station sends a sensing reference signal and receives the sensing reference signal to obtain the sensed measurement quantity or the sensing result;

[0953] 2. The base station sends at least one of the following information to the UE:

[0954] The sensed measurement quantity;

[0955] The timestamp of the sensed measurement quantity;

[0956] The effective duration of the sensed measurement quantity;

[0957] Other relevant information of the sensed measurement quantity;

[0958] The indication of the device receiving the sensing reference signal (in this embodiment, it is to indicate the base station, and this indication can be an implicit indication or an explicit indication).

[0959] 3. Based on the information sent by the base station, the UE synchronizes the perception measurement quantities and the CSI measurement results to generate a data set for model training.

[0960] Before step 1, the UE sends a first request to the base station to request the perception measurement quantities required for model training. (Same as

[0961] Embodiment 1)

[0962] The rest is the same as Figure 16 Embodiment 5 shown, and for the sake of avoiding repetition, it will not be elaborated here.

[0963] The above are the specific embodiments provided by the embodiments of this application.

[0964] Regarding the relevant terms of this application, the explanations are as follows:

[0965] Explanation 1: AI model

[0966] The AI model can also be referred to as an AI unit, a machine learning (ML) model, an ML unit, an AI structure, an AI function, an AI feature, a neural network, a neural network function, a neural network function, etc. Or, the AI unit / AI model can also refer to a processing unit that can implement specific algorithms, formulas, processing flows, capabilities, etc. related to AI. Or, the AI unit / AI model can be a processing method, algorithm, function, module, or unit for a specific data set. Or, the AI unit / AI model can be a processing method, algorithm, function, module, or unit running on AI / ML-related hardware such as GPUs, NPUs, TPUs, ASICs, etc. This application does not make specific limitations on this.

[0967] Optionally, the specific data set includes the input or output of the AI unit / AI model.

[0968] Optionally, the identifier of the AI unit / AI model can be an AI model identifier, an AI structure identifier, an AI algorithm identifier, etc. Or, the identifier of the specific data set associated with the AI unit / AI model. Or, the identifier of the specific scenario, environment, channel characteristics, device, etc. related to AI / ML. Or, the identifier of the function, feature, ability, or module, etc. related to AI / ML. This application does not make specific limitations on this.

[0969] Explanation 2: Types of perception measurement quantities

[0970] For the perception measurement quantities related to integrated communication and sensing, the following types are included:

[0971] (1) The first-level measurement quantity (received signal / original channel information) includes: the complex result of the received signal / channel response, amplitude / phase, in-phase (I) / quadrature (Q) channels and their operation results (operations include addition, subtraction, multiplication, division, matrix addition, subtraction, multiplication, matrix transpose, trigonometric relation operations, square root operation, power operation, etc., as well as the threshold detection results and maximum / minimum value extraction results of the above operation results; operations also include Fast Fourier Transform (FFT) / Inverse Fast Fourier Transform (IFFT), Discrete Fourier Transform (DFT) / Inverse Discrete Fourier Transform (IDFT), 2D-FFT, 3D-FFT, matched filtering, autocorrelation operation, wavelet transform, digital filtering, etc., as well as the threshold detection results and maximum / minimum value extraction results of the above operation results);

[0972] Here, H is not limited to the channels between the serving base station and the UE, but also includes the sensing channels obtained by other base stations through self-transmission and self-reception. Multiple scenarios are included. The H sensed by the base station through self-transmission and self-reception is different from the H measured by the UE itself. For example, whether the moving target in the environment is caused by the movement of obstacles or the movement of the UE itself. This can achieve interference cancellation and is beneficial to 3D environment reconstruction.

[0973] It should be noted that the above H represents the received signal / original channel information.

[0974] The effective time of the sensing measurement quantity can be long or short. If it is relatively static sensing information (for example, unchanged for several years), it can be directly sent; if it is obtained through real-time measurement, it is necessary to configure which devices send the sensing signal and which devices receive the sensing signal. This can improve the prediction accuracy.

[0975] (2) The second-level measurement quantity (basic measurement quantity) includes: time delay, Doppler, angle, intensity, and their multi-dimensional combined representations; for example, it can be a time delay value, Doppler value, or time delay power spectrum, Doppler power spectrum, velocity power spectrum, angle power spectrum, time delay-Doppler spectrum, time delay-angle spectrum, Doppler-angle spectrum, time delay-Doppler-angle spectrum, etc.

[0976] (3) The third-level measurement quantity: sensing result / sensing intermediate result

[0977] Basic attributes / status include: distance, velocity, orientation, spatial position, acceleration;

[0978] Advanced attributes / status, including: whether the target exists, trajectory, action, expression, vital signs, quantity, imaging result, weather, air quality, shape, material, composition;

[0979] Environmental reconstruction result, trajectory.

[0980] For the measurement quantities sensed by sensors, the following types are included:

[0981] (1) Measurement quantities related to lidar, including at least one of the following:

[0982] Lidar point cloud data, and each point in the lidar point cloud data includes: X / Y / Z position information, and additional information;

[0983] The angle and distance of the target obtained from the lidar point cloud data;

[0984] The visual features of the target identified from the lidar point cloud data, such as: people, vehicles, etc.;

[0985] The number of targets identified from the lidar point cloud data.

[0986] Among them, the additional information in the lidar point cloud data includes at least one of the following:

[0987] Intensity: The echo intensity of the laser pulse that generates the lidar point;

[0988] Number of echoes: The number of echoes is the total number of echoes of a given pulse;

[0989] Point classification: Each post-processed lidar point can have a classification that defines the type of object reflecting the lidar pulse, and the lidar points can be divided into many categories, such as: ground, bare surface, top of the tree canopy, and water area, etc.;

[0990] RGB: The RGB band can be used as an attribute of the lidar data, and this attribute usually comes from the image collected during the lidar measurement;

[0991] GPS time: The GPS timestamp when the laser point is emitted from the aircraft;

[0992] Scanning angle;

[0993] Scanning direction: The traveling direction of the laser scanning mirror, with the value 1 representing the positive scanning direction and the value 0 representing the negative scanning direction.

[0994] (2) Measurement quantities related to vision, including at least one of the following:

[0995] Visual image;

[0996] The luminosity of the image pixels;

[0997] RGB values of image pixels;

[0998] Visual features of the objects recognized from the image, such as: people, vehicles, etc.;

[0999] Angles and distances of the objects recognized from the image (especially for binocular vision);

[1000] Number of the objects recognized from the image.

[1001] (3) Measurement quantities related to radar, including at least one of the following:

[1002] Radar point cloud, and each point in the point cloud includes at least one of: distance / speed / azimuth angle / elevation angle, or at least one of X / Y / Z / speed;

[1003] Distances, speeds, and angles of the recognized objects;

[1004] Radar imaging;

[1005] Number of the objects.

[1006] (4) Measurement quantities related to the inertial measurement unit, including at least one of the following:

[1007] Acceleration: at least one of the three directions of X / Y / Z;

[1008] Speed: at least one of the three directions of X / Y / Z;

[1009] Angular velocity: at least one of the three axes of X / Y / Z.

[1010] (5) Measurement quantities of a position sensor such as GNSS, including the position information of the device and the relative distance / angle information between the device and a specific object.

[1011] (6) Other measurement quantities, including at least one of the following: whether the object exists, trajectory, action, expression, vital signs, quantity, imaging result, weather, air quality, shape, material, composition, etc.

[1012] Explanation 3: Signal configuration information (which can be called the configuration information of the target signal, the configuration information of the sensing reference signal, etc.)

[1013] Signal configuration information, including at least one of the following:

[1014] Signal resource identifier (ID), used to distinguish different signal resource configurations;

[1015] Signal usage indicates that the signal is for communication (such as channel measurement, channel estimation, synchronization, carrying data information, etc.), for sensing, or for both communication and sensing. Specifically, it can also be the signal for which sensing service or which type of sensing service. The definitions of the sensing service and sensing service type can refer to Explanation 3.

[1016] Waveform, for example, the waveform is Orthogonal Frequency Division Multiplexing (OFDM), Single-carrier Frequency-Division Multiple Access (SC-FDMA), Orthogonal Time Frequency Space (OTFS), Chirp waveform, Frequency Modulated Continuous Wave (FMCW), pulse signal, etc.;

[1017] Subcarrier spacing, for example, the subcarrier spacing of the OFDM system is 30 KHz.

[1018] Guard interval, the guard interval is the time interval between the moment when the signal ends transmission and the moment when the latest echo signal of the signal is received; this parameter is proportional to the maximum sensing distance; for example, it can be calculated by c / (2R max ), where R max is the maximum sensing distance (belonging to sensing requirement information). For example, for a self-transmitting and self-receiving sensing signal, R max represents the maximum distance from the sensing signal transceiver point to the signal emission point; in some cases, the cyclic prefix (CP) of the OFDM signal can serve as the minimum guard interval; c is the speed of light.

[1019] Starting frequency domain position, that is, the starting frequency point, which can also be the starting RE, RB index;

[1020] Starting time domain position, that is, the starting time point, which can also be the starting symbol index, time slot index, frame index;

[1021] Ending frequency domain position, that is, the ending frequency point, which can be represented by the ending RE, RB index;

[1022] Ending time domain position, that is, the ending time point, which can be represented by the ending RE, RB index;

[1023] Frequency domain resource length, that is, the frequency domain bandwidth. The frequency domain bandwidth is inversely proportional to the range resolution. The frequency domain bandwidth B of each of the first signals satisfies B≥c / (2ΔR), where c is the speed of light and ΔR is the range resolution;

[1024] The time-domain resource length, also known as the burst duration, is inversely proportional to the Doppler resolution;

[1025] The frequency-domain resource interval represents the interval between adjacent signal frequency-domain resource units, which can be expressed by the number of REs or RBs, or by the density value (Density). For example, Density = 1 means that there is one RE in each RB for carrying signals. The frequency-domain resource interval is inversely proportional to the maximum unambiguous distance / delay. Among them, for an OFDM system, when subcarriers are continuously mapped, the frequency-domain interval is equal to the subcarrier interval;

[1026] The time-domain resource interval is the time interval between two adjacent signal resource units, and the time-domain resource interval is associated with the maximum unambiguous Doppler frequency shift or the maximum unambiguous speed;

[1027] Time-domain resource characteristics, periodic transmission, semi-persistent transmission, aperiodic transmission;

[1028] Signal power, for example, taking a value every 2 dBm from -20 dBm to 23 dBm;

[1029] Sequence information, including sequence type information (ZC sequence, PN sequence, etc.), sequence generation method, sequence length, etc.;

[1030] Signal direction, the angle information or beam information of signal transmission;

[1031] QCL relationship. For example, the sensing signal includes multiple resources, and each resource is QCL with an SSB. QCL includes Type A, B, C, or D;

[1032] Antenna port information, such as the maximum number of antenna ports, antenna port index;

[1033] Cyclic prefix (CP) information, including CP type (such as Normal Cyclic Prefix (NCP), Extended Cyclic Prefix (ECP), or a newly designed CP dedicated to sensing measurement, etc.), CP length, etc.

[1034] Explanation 4: Sensing requirements

[1035] The sensing requirement information includes at least one of the following:

[1036] Perception services or perception service types. The perception services can be, for example, detecting the presence of a target, positioning, speed detection, distance detection, angle detection, acceleration detection, material analysis, component analysis, shape detection, classification, radar cross section (RCS) detection, polarization scattering characteristic detection, fall detection, intrusion detection, quantity statistics, indoor positioning, gesture recognition, lip reading recognition, gait recognition, facial expression recognition, facial recognition, breathing monitoring, heart rate monitoring, pulse monitoring, humidity / brightness / temperature / atmospheric pressure monitoring, air quality monitoring, weather condition monitoring, environment reconstruction, terrain and landform, building / vegetation distribution detection, pedestrian or vehicle flow detection, crowd density, vehicle density detection, etc.; The perception service types can classify multiple different perception services according to certain characteristics. For example, they can be classified into detection-type perception services (such as intrusion detection, fall detection) according to function, parameter estimation-type perception services (distance, angle, speed calculation), recognition-type perception services (action recognition, identity recognition), etc., or they can be, for example, target detection and tracking-type perception services (including the presence of a target, target ranging / ranging / angle measurement / localization / trajectory tracking), environment monitoring-type perception services (including rainfall detection, flood monitoring), action detection-type perception services (including gesture / action recognition, breathing / heartbeat detection, fall detection), etc. They can also be classified according to the perception range (close-range perception, medium-range perception, long-range perception), according to the perception fineness (coarse-grained perception, fine-grained perception, etc.), according to power consumption / energy consumption, according to resource occupancy, etc.

[1037] Perceived target area: It refers to the area where the perceived object may exist, or the area where imaging or environment reconstruction needs to be carried out.

[1038] Perceived object type: Classify the perceived objects according to the possible motion characteristics of the perceived objects. Each perceived object type contains information such as the motion speed, motion acceleration, and typical RCS of typical perceived objects.

[1039] Perceived QoS: Performance indicators for perceiving the perceived target area or the perceived object.

[1040] The perceived QoS includes at least one of the following:

[1041] Perceived resolution (which can be divided into: ranging resolution, angle measurement resolution, speed measurement resolution, imaging resolution), etc.

[1042] Perceived accuracy (which can be divided into: ranging accuracy, angle measurement accuracy, speed measurement accuracy, positioning accuracy, etc.);

[1043] Perceived range (which can be divided into: ranging range, speed measurement range, angle measurement range, imaging range, etc.);

[1044] Perception latency (the time interval from the transmission of the perception signal to the acquisition of the perception result, or the time interval from the initiation of the perception requirement to the acquisition of the perception result);

[1045] Perception update rate (the time interval between two adjacent executions of perception and the acquisition of the perception result);

[1046] Detection probability (the probability of being correctly detected when the perceived object exists);

[1047] False alarm probability (the probability of erroneously detecting a perceived target when the perceived object does not exist);

[1048] The maximum number of perceivable targets.

[1049] In summary, in the embodiments of the present application, by proposing a method for collecting model training data assisted by perception, network nodes can expand the input information for model training by obtaining perception information, thereby improving the inference accuracy of the AI model.

[1050] For the dataset collection method provided in the embodiments of the present application, the execution subject may be a dataset collection device. For the model training method provided in the embodiments of the present application, the execution subject may be a model training device. For the model storage method provided in the embodiments of the present application, the execution subject may be a model storage device. In the embodiments of the present application, taking the dataset collection device executing the dataset collection method as an example, the dataset collection device provided in the embodiments of the present application is described. In the embodiments of the present application, taking the model training device executing the model training method as an example, the model training device provided in the embodiments of the present application is described. In the embodiments of the present application, taking the model storage device executing the model training method as an example, the model storage device provided in the embodiments of the present application is described.

[1051] Figure 23 The structural diagram of the dataset collection device provided in the embodiments of the present application is shown. The dataset collection device can be applied to the first device. As Figure 23 shown, the dataset collection device 910 includes:

[1052] A first processing module 911, configured to obtain communication measurement-related data and perception measurement-related data;

[1053] A second processing module 912, configured to determine a target dataset for model training based on the communication measurement-related data and the perception measurement-related data.

[1054] Optionally, the perception measurement-related data includes at least one of the following information:

[1055] Perception measurement quantity;

[1056] Indicator of the perception measurement quantity;

[1057] Time stamp of the sensed measurement quantity;

[1058] Information of the sending device of the sensed measurement quantity;

[1059] Information of the receiving device of the sensed measurement quantity;

[1060] Coordinate information of the sensed measurement quantity;

[1061] Information for indicating the performance index of the sensed measurement quantity;

[1062] Information for indicating the source of the sensed measurement quantity;

[1063] Information for indicating the type of the sensed measurement quantity;

[1064] Information for indicating the sensing mode of the sensed measurement quantity;

[1065] Configuration information of the target signal;

[1066] Information of the sending device of the target signal;

[1067] Information of the receiving device of the target signal;

[1068] Wherein, the target signal is a signal used for sensing.

[1069] Optionally, the device further includes:

[1070] A first sending module, configured to send a first request to a second device, where the first request is used to request sensed measurement-related data required for model training.

[1071] Optionally, the first request includes at least one of the following information:

[1072] Information for indicating that the model is in the training stage;

[1073] A dataset identifier, and the dataset corresponding to the dataset identifier includes sensed measurement-related data;

[1074] A configuration identifier of the sensed measurement quantity;

[1075] An identifier of the sensed measurement quantity;

[1076] Configuration information of the sensed measurement quantity required for model training;

[1077] Information for indicating the number threshold of the sensed measurement quantity required for model training;

[1078] Information for indicating the source of the sensed measurement quantity required for model training;

[1079] Information for indicating the type of the sensed measurement quantity required for model training;

[1080] Information on the sensing mode for indicating the amount of sensing measurement required for model training;

[1081] Information for indicating sensing requirements;

[1082] Configuration information of the target signal;

[1083] Information on the sending device of the target signal;

[1084] Information on the receiving device of the target signal;

[1085] Wherein, the target signal is a signal used for sensing.

[1086] Optionally, the sensing measurement-related data obtained by the first device includes a first sensing measurement amount;

[1087] The second processing module is specifically used for at least one of the following:

[1088] Generating a first training sample based on the target timestamp, the timestamp of the first sensing measurement amount, and the valid duration of the first sensing measurement amount, and determining the target data set based on the first training sample;

[1089] Generating a first training sample based on the first indication and the target timestamp, and determining the target data set based on the first training sample;

[1090] Generating a first training sample based on whether a second indication is received, and determining the target data set based on the first training sample;

[1091] Wherein, the target timestamp is the measurement timestamp of the first communication measurement amount, and the first communication measurement amount is the last communication measurement amount corresponding to the true value or label in the communication measurement-related data;

[1092] The first indication is used to indicate the failure time or failure time difference of the first sensing measurement amount;

[1093] The second indication is used to indicate the failure of the first sensing measurement amount.

[1094] Optionally, the determination method of the valid duration of the first sensing measurement amount includes at least one of the following:

[1095] When a first valid duration is pre-configured or defined, the valid duration of the first sensing measurement amount is the first valid duration;

[1096] When a second valid duration is pre-configured or defined and the first target signal carries a time adjustment value, the valid duration of the first sensing measurement amount is determined according to the second valid duration and the time adjustment value;

[1097] When the first target signal carries a third effective duration, the effective duration of the first perception measurement quantity is the third effective duration;

[1098] Wherein, the first target signal is the target signal corresponding to the first perception measurement quantity, and the target signal is a signal used for perception.

[1099] Optionally, the second processing module is specifically configured to perform at least one of the following:

[1100] When the time corresponding to the target timestamp is earlier than the first time, generate a first training sample including the communication measurement related data and the first perception measurement quantity; the first time is the time corresponding to the timestamp of the first perception measurement quantity plus the effective duration of the first perception measurement quantity;

[1101] When the time corresponding to the target timestamp is earlier than the second time, generate a first training sample including the communication measurement related data and the first perception measurement quantity; the second time is the expiration time of the first perception measurement quantity determined based on the first indication;

[1102] When the first device does not receive the second indication, generate a first training sample including the communication measurement related data and the first perception measurement quantity.

[1103] Optionally, the second processing module includes:

[1104] A receiving unit, configured to receive first information from a third device, where the first information includes at least one of communication measurement related data and perception measurement related data;

[1105] A processing unit, configured to generate a training sample based on the first information, and determine the target data set based on the training sample; or, generate a training sample based on the first information and stored historical data, and determine the target data set based on the training sample;

[1106] Wherein, the historical data includes at least one of communication measurement related data and perception measurement related data.

[1107] Optionally, the first information includes at least one of the following:

[1108] A data set identifier, where the data set corresponding to the data set identifier includes perception measurement related data;

[1109] A configuration identifier of the perception measurement quantity;

[1110] An identifier of the perception measurement quantity;

[1111] Identification of the sensing measurement link;

[1112] Identification of the sensing service;

[1113] Identification of the sensing service type;

[1114] Sensing measurement quantity;

[1115] Timestamp of the sensing measurement quantity;

[1116] Information of the sending device of the sensing measurement quantity;

[1117] Coordinate information of the sensing measurement quantity;

[1118] Information for indicating the performance index of the sensing measurement quantity;

[1119] Information for indicating the source of the sensing measurement quantity;

[1120] Information for indicating the type of the sensing measurement quantity;

[1121] Information for indicating the use of the sensing measurement quantity;

[1122] Information for indicating the sensing mode of the sensing measurement quantity;

[1123] Communication measurement quantity;

[1124] Communication measurement resource identifier;

[1125] Timestamp of the communication measurement quantity;

[1126] Information for indicating the first time window;

[1127] Information for indicating the first valid time;

[1128] Configuration identifier of the target signal;

[1129] Configuration information of the target signal;

[1130] Information of the sending device of the target signal;

[1131] Information of the receiving device of the target signal;

[1132] Wherein, the target signal is a signal used for sensing.

[1133] Optionally, the first information includes a second sensing measurement quantity, a timestamp of the second sensing measurement quantity, a second communication measurement quantity, and a timestamp of the second communication measurement quantity;

[1134] The processing unit is specifically configured to perform at least one of the following:

[1135] When the time difference between the timestamp of the second sensing measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to the first time window indicated by the first information, generate a second training sample including the second communication measurement quantity and the second sensing measurement quantity;

[1136] When the time difference between the timestamp of the second sensing measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to a predefined second time window, generate a second training sample including the second communication measurement quantity and the second sensing measurement quantity.

[1137] Optionally, the first information includes a third sensing measurement quantity and the timestamp of the third sensing measurement quantity;

[1138] The processing unit is specifically configured to:

[1139] Generate a third training sample including a third communication measurement quantity and the third sensing measurement quantity;

[1140] Wherein, the third communication measurement quantity includes at least one of the following:

[1141] The communication measurement quantity whose time difference between the timestamp of the historical data and the timestamp of the third sensing measurement quantity is less than or equal to the first time window indicated by the first information;

[1142] The communication measurement quantity whose time difference between the timestamp of the historical data and the timestamp of the third sensing measurement quantity is less than or equal to a predefined second time window.

[1143] Optionally, the first information includes a fourth communication measurement quantity and the timestamp of the fourth communication measurement quantity;

[1144] The processing unit is specifically configured to:

[1145] Generate a fourth training sample including the fourth communication measurement quantity and a fourth sensing measurement quantity;

[1146] Wherein, the fourth communication measurement quantity includes at least one of the following:

[1147] The sensing measurement quantity whose time difference between the timestamp of the historical data and the timestamp of the fourth communication measurement quantity is less than or equal to the first time window indicated by the first information;

[1148] The sensing measurement quantity whose timestamp of the historical data is less than or equal to the first valid time indicated by the first information;

[1149] The sensing measurement quantity whose timestamp of the historical data is less than or equal to a predefined second valid time.

[1150] Optionally, the processing unit is specifically configured to perform at least one of the following:

[1151] When the first information includes a fifth communication measurement and a fifth sensing measurement, generating a fifth training sample including the fifth communication measurement and the fifth sensing measurement;

[1152] When the first information includes a fifth communication measurement and a third indication, generating a fifth training sample including the fifth communication measurement and a fifth sensing measurement, where the fifth sensing measurement is the sensing measurement corresponding to the third indication;

[1153] When the first information includes a fifth communication measurement, generating a fifth training sample including the fifth communication measurement and a fifth sensing measurement, where the fifth sensing measurement is the sensing measurement of the most recent training sample.

[1154] Optionally, the third indication includes at least one of the following:

[1155] The identifier of the training sample;

[1156] The identifier of the resource used for the communication measurement result;

[1157] The reporting identifier of the communication measurement result;

[1158] The identifier of the resource used for the sensing measurement;

[1159] The reporting identifier of the sensing measurement;

[1160] The identifier of the sensing measurement.

[1161] Optionally, the device further includes:

[1162] A second sending module, configured to send second information to a fourth device, where the second information includes the target data set and the identifier of the target data set, the fourth device is a device for model training, and the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set.

[1163] Optionally, the device further includes:

[1164] A third sending module, configured to send third information to a fifth device, where the third information includes the target data set and the identifier of the target data set, and the fifth device is a device for storing the model training data set.

[1165] Optionally, the second information or the third information further includes target information;

[1166] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[1167] Optionally, the target information is used to indicate at least one of the following:

[1168] Whether a perception measurement quantity is used for a sample in the target data set;

[1169] The quantity of perception measurement quantities used for a sample in the target data set;

[1170] The proportion of perception measurement quantities used for a sample in the target data set;

[1171] Whether the quantity of perception measurement quantities used for a sample in the target data set meets a minimum quantity threshold;

[1172] Whether the proportion of perception measurement quantities used for a sample in the target data set meets a minimum proportion threshold;

[1173] The proportion of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for a sample in the target data set;

[1174] The quantity of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for a sample in the target data set;

[1175] A first threshold, where the first threshold is a proportion threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in a sample;

[1176] A second threshold, where the second threshold is a proportion threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in a sample.

[1177] Optionally, the target information is used to indicate at least one of the following:

[1178] A data set identifier, where the data set corresponding to the data set identifier includes perception measurement-related data;

[1179] A configuration identifier of a perception measurement quantity;

[1180] An identifier of a perception measurement quantity;

[1181] An identifier of a perception measurement link;

[1182] An identifier of a perception service;

[1183] An identifier of a perception service type;

[1184] The parameter items of the perception measurement quantities used for a sample in the target data set, where the parameter items include at least one of Doppler of a path, time delay, power, and angle;

[1185] A sending device for the perception measurement quantity used by the samples in the target dataset;

[1186] A receiving device for the perception measurement quantity used by the samples in the target dataset;

[1187] The coordinates of the perception measurement quantity used by the samples in the target dataset;

[1188] The source of the perception measurement quantity used by the samples in the target dataset;

[1189] The perception mode of the perception measurement quantity used by the samples in the target dataset;

[1190] The processing level of the perception measurement quantity used by the samples in the target dataset;

[1191] The perception service of the perception measurement quantity used by the samples in the target dataset;

[1192] The use of the perception measurement quantity used by the samples in the target dataset;

[1193] The performance index of the perception measurement quantity used by the samples in the target dataset;

[1194] The perception measurement quantity used by the samples in the target dataset is a fixed perception measurement quantity or a combination of perception measurement quantities;

[1195] The perception measurement quantity used by the samples in the target dataset is a variable perception measurement quantity or a combination of perception measurement quantities;

[1196] The number of perception measurement quantities used by the samples in the target dataset is variable;

[1197] Configuration information of the second target signal;

[1198] Wherein, the second target signal is the target signal corresponding to the perception measurement quantity used by the samples in the target dataset, and the target signal is a signal used for perception.

[1199] Optionally, the device further includes:

[1200] A third processing module, configured to use the target dataset to train a target AI unit.

[1201] Figure 24 The structural diagram of the dataset collection device provided by the embodiment of the present application is shown. The dataset collection device can be applied to a second device. As Figure 24 shown, the dataset collection device 920 includes:

[1202] A receiving module 921, configured to receive a first request from a first device, where the first request is used to request perception measurement-related data required for model training;

[1203] A processing module 922, configured to perform a first operation, where the first operation includes any one of the following:

[1204] Based on the first request, send a target signal;

[1205] Based on the first request, determine configuration information of the target signal;

[1206] Based on the first request, send the configuration information of the target signal;

[1207] Based on the first request, send perception measurement-related data;

[1208] Wherein, the target signal is a signal used for perception.

[1209] Optionally, the first request includes at least one of the following information:

[1210] Information for indicating that the model is in the training phase;

[1211] A dataset identifier, where the dataset corresponding to the dataset identifier includes perception measurement-related data;

[1212] A configuration identifier of the perception measurement quantity;

[1213] An identifier of the perception measurement quantity;

[1214] Configuration information of the perception measurement quantity required for model training;

[1215] Information for indicating a quantity threshold of the perception measurement quantity required for model training;

[1216] Information for indicating the source of the perception measurement quantity required for model training;

[1217] Information for indicating the type of the perception measurement quantity required for model training;

[1218] Information for indicating the perception mode of the perception measurement quantity required for model training;

[1219] Information for indicating the perception requirement;

[1220] Configuration information of the target signal;

[1221] Information about the sending device of the target signal;

[1222] Information about the receiving device of the target signal;

[1223] Wherein, the target signal is a signal used for perception.

[1224] Figure 25 shows the structural diagram of the dataset collection device provided by an embodiment of the present application. The dataset collection device can be applied to a third device. As Figure 25 shown, the dataset collection device 930 includes:

[1225] A sending module 931, configured to send first information to a first device, where the first information includes at least one of communication measurement related data and sensing measurement related data;

[1226] The first information includes at least one of the following:

[1227] Communication measurement quantities and sensing measurement quantities, where the communication measurement quantities and the sensing measurement quantities are determined to belong to the same training sample;

[1228] Communication measurement quantities and a third indication, where the communication measurement quantities and the sensing measurement quantities indicated by the third indication are determined to belong to the same training sample;

[1229] Communication measurement quantities, where the communication measurement quantities and the sensing measurement quantities of the most recent training sample are determined to belong to the same training sample;

[1230] Or,

[1231] The first information includes at least one of the following:

[1232] A dataset identifier, where the dataset corresponding to the dataset identifier includes sensing measurement related data;

[1233] A configuration identifier of the sensing measurement quantity;

[1234] An identifier of the sensing measurement quantity;

[1235] An identifier of the sensing measurement link;

[1236] An identifier of the sensing service;

[1237] An identifier of the sensing service type;

[1238] Sensing measurement quantities;

[1239] A timestamp of the sensing measurement quantity;

[1240] Information about the sending device of the sensing measurement quantity;

[1241] Coordinate information of the sensing measurement quantity;

[1242] Information for indicating a performance metric of the sensing measurement quantity;

[1243] Information for indicating the source of the sensing measurement quantity;

[1244] Information for indicating the type of the sensed measurement quantity;

[1245] Information for indicating the use of the sensed measurement quantity;

[1246] Information for indicating the sensing mode of the sensed measurement quantity;

[1247] Communication measurement quantity;

[1248] Communication measurement resource identifier;

[1249] Timestamp of the communication measurement quantity;

[1250] Information for indicating the first time window;

[1251] Information for indicating the first valid time;

[1252] Configuration identifier of the target signal;

[1253] Configuration information of the target signal;

[1254] Information about the sending device of the target signal;

[1255] Information about the receiving device of the target signal;

[1256] Wherein, the target signal is a signal used for sensing.

[1257] Optionally, the third indication includes at least one of the following:

[1258] Identifier of the training sample;

[1259] Identifier of the resource used for the communication measurement result;

[1260] Reporting identifier of the communication measurement result;

[1261] Identifier of the resource used for the sensed measurement quantity;

[1262] Reporting identifier of the sensed measurement quantity;

[1263] Identifier of the sensed measurement quantity.

[1264] Figure 26 The structural diagram of the model training device provided by the embodiment of the present application is shown. The model training device can be applied to the fourth device. As Figure 26 shown, the model training device 940 includes:

[1265] A receiving module 941, configured to receive second information from a first device, where the second information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensed measurement-related data;

[1266] A processing module 942, configured to use the target data set to train a target AI unit.

[1267] Optionally, the perception measurement related data includes at least one of the following information:

[1268] Perception measurement quantity;

[1269] Indicator of the perception measurement quantity;

[1270] Timestamp of the perception measurement quantity;

[1271] Information of the sending device of the perception measurement quantity;

[1272] Information of the receiving device of the perception measurement quantity;

[1273] Coordinate information of the perception measurement quantity;

[1274] Information for indicating a performance metric of the perception measurement quantity;

[1275] Information for indicating the source of the perception measurement quantity;

[1276] Information for indicating the type of the perception measurement quantity;

[1277] Information for indicating the perception mode of the perception measurement quantity;

[1278] Configuration information of the target signal;

[1279] Information of the sending device of the target signal;

[1280] Information of the receiving device of the target signal;

[1281] Wherein, the target signal is a signal used for perception.

[1282] Optionally, the device further includes:

[1283] A first sending module, configured to send fourth information to a fifth device, where the fourth information includes the target data set and an identifier of the target data set, and the fifth device is a device storing a model training data set.

[1284] Optionally, the device further includes:

[1285] A second sending module, configured to send fifth information to a sixth device, where the fifth information includes the target AI unit, an identifier of the target AI unit, and an identifier of the target data set, and the sixth device is a device storing the target AI unit.

[1286] Optionally, the second information or the fourth information or the fifth information further includes target information;

[1287] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[1288] Optionally, the target information is used to indicate at least one of the following:

[1289] Whether the samples in the target data set use perception measurement quantities;

[1290] The number of perception measurement quantities used by the samples in the target data set;

[1291] The proportion of perception measurement quantities used by the samples in the target data set;

[1292] Whether the number of perception measurement quantities used by the samples in the target data set meets the minimum number threshold;

[1293] Whether the proportion of perception measurement quantities used by the samples in the target data set meets the minimum proportion threshold;

[1294] The proportion of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used by the samples in the target data set;

[1295] The number of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used by the samples in the target data set;

[1296] A first threshold, which is the proportion threshold between the number of valid perception measurement quantities and the total number of perception measurement quantities in a sample;

[1297] A second threshold, which is the proportion threshold between the number of actually used perception measurement quantities and the total number of the maximum supportable perception measurement quantities in a sample.

[1298] Optionally, the target information is used to indicate at least one of the following:

[1299] A data set identifier, and the data set corresponding to the data set identifier includes perception measurement related data;

[1300] A configuration identifier of the perception measurement quantity;

[1301] An identifier of the perception measurement quantity;

[1302] An identifier of the perception measurement link;

[1303] An identifier of the perception service;

[1304] An identifier of the perception service type;

[1305] The parameter items of the perception measurement quantities used by the samples in the target data set, and the parameter items include at least one of Doppler of the path, time delay, power, and angle;

[1306] A transmitting device for the perception measurement quantity used by the samples in the target data set;

[1307] A receiving device for the perception measurement quantity used by the samples in the target data set;

[1308] The coordinates of the perception measurement quantity used by the samples in the target data set;

[1309] The source of the perception measurement quantity used by the samples in the target data set;

[1310] The perception mode of the perception measurement quantity used by the samples in the target data set;

[1311] The processing level of the perception measurement quantity used by the samples in the target data set;

[1312] The perception service of the perception measurement quantity used by the samples in the target data set;

[1313] The use of the perception measurement quantity used by the samples in the target data set;

[1314] The performance index of the perception measurement quantity used by the samples in the target data set;

[1315] The perception measurement quantity used by the samples in the target data set is a fixed perception measurement quantity or a combination of perception measurement quantities;

[1316] The perception measurement quantity used by the samples in the target data set is a variable perception measurement quantity or a combination of perception measurement quantities;

[1317] The number of perception measurement quantities used by the samples in the target data set is variable;

[1318] Configuration information of the second target signal;

[1319] Wherein, the second target signal is the target signal corresponding to the perception measurement quantity used by the samples in the target data set, and the target signal is a signal used for perception.

[1320] Figure 27 The structure diagram of the data set collection device provided by the embodiment of the present application is shown. The data set collection device can be applied to the fifth device. As Figure 27 shown, the data set collection device 950 includes:

[1321] A receiving module 951, configured to receive third information from a first device, or receive fourth information from a fourth device, where the third information or the fourth information includes a target data set and an identifier of the target data set; the target data set includes communication measurement related data and perception measurement related data;

[1322] A processing module 952, configured to store the target data set and an identifier of the target data set;

[1323] Wherein, the first device is a device for collecting the target data set, the fourth device is a device for performing model training, and the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set.

[1324] Optionally, the sensing measurement related data includes at least one of the following information:

[1325] Sensing measurement quantity;

[1326] Indicator of the sensing measurement quantity;

[1327] Timestamp of the sensing measurement quantity;

[1328] Information of the sending device of the sensing measurement quantity;

[1329] Information of the receiving device of the sensing measurement quantity;

[1330] Coordinate information of the sensing measurement quantity;

[1331] Information for indicating a performance metric of the sensing measurement quantity;

[1332] Information for indicating the source of the sensing measurement quantity;

[1333] Information for indicating the type of the sensing measurement quantity;

[1334] Information for indicating the sensing mode of the sensing measurement quantity;

[1335] Configuration information of the target signal;

[1336] Information of the sending device of the target signal;

[1337] Information of the receiving device of the target signal;

[1338] Wherein, the target signal is a signal used for sensing.

[1339] Optionally, the third information or the fourth information further includes target information;

[1340] The target information is used to indicate the sensing characteristics of the target data set, or the target information is used to indicate the sensing characteristics and communication characteristics of the target data set.

[1341] Optionally, the target information is used to indicate at least one of the following:

[1342] Whether the samples in the target data set use sensing measurement quantities;

[1343] The number of perception measurement quantities used by the samples in the target dataset;

[1344] The proportion of the perception measurement quantities used by the samples in the target dataset;

[1345] Whether the number of the perception measurement quantities used by the samples in the target dataset meets the minimum quantity threshold;

[1346] Whether the proportion of the perception measurement quantities used by the samples in the target dataset meets the minimum proportion threshold;

[1347] The proportion of the perception measurement quantities that meet the timeliness among the perception measurement quantities used by the samples in the target dataset;

[1348] The number of the perception measurement quantities that meet the timeliness among the perception measurement quantities used by the samples in the target dataset;

[1349] The first threshold, which is the proportion threshold between the number of valid perception measurement quantities and the total number of perception measurement quantities in a sample;

[1350] The second threshold, which is the proportion threshold between the number of actually used perception measurement quantities and the total number of the maximum supportable perception measurement quantities in a sample.

[1351] Optionally, the target information is used to indicate at least one of the following:

[1352] The dataset identifier, and the dataset corresponding to the dataset identifier includes perception measurement related data;

[1353] The configuration identifier of the perception measurement quantity;

[1354] The identifier of the perception measurement quantity;

[1355] The identifier of the perception measurement link;

[1356] The identifier of the perception service;

[1357] The identifier of the perception service type;

[1358] The parameter items of the perception measurement quantities used by the samples in the target dataset, and the parameter items include at least one of Doppler of the diameter, time delay, power, and angle;

[1359] The sending device of the perception measurement quantities used by the samples in the target dataset;

[1360] The receiving device of the perception measurement quantities used by the samples in the target dataset;

[1361] The coordinates of the perception measurement quantities used by the samples in the target dataset;

[1362] The source of the perception measurement quantities used by the samples in the target dataset;

[1363] The perception mode of the perception measurement quantities used by the samples in the target dataset;

[1364] The processing level of the perception measurement quantities used by the samples in the target dataset;

[1365] The perception service of the perception measurement quantities used by the samples in the target dataset;

[1366] The use of the perception measurement quantities used by the samples in the target dataset;

[1367] The performance indicators of the perception measurement quantities used by the samples in the target dataset;

[1368] The perception measurement quantity used by the samples in the target dataset is a fixed perception measurement quantity or a combination of perception measurement quantities;

[1369] The perception measurement quantity used by the samples in the target dataset is a variable perception measurement quantity or a combination of perception measurement quantities;

[1370] The number of the perception measurement quantities used by the samples in the target dataset is variable;

[1371] The configuration information of the second target signal;

[1372] Wherein, the second target signal is the target signal corresponding to the perception measurement quantity used by the samples in the target dataset, and the target signal is the signal used for perception.

[1373] Figure 28 The structural diagram of the model storage device provided by the embodiment of the present application is shown. The model storage device can be applied to the sixth device. As Figure 28 shown, the model storage device 960 includes:

[1374] A receiving module 961, configured to receive fifth information from a fourth device, where the fifth information includes a target AI unit, an identifier of the target AI unit, and an identifier of a target dataset, and the dataset corresponding to the identifier of the target dataset includes communication measurement-related data and perception measurement-related data;

[1375] A first processing module 962, configured to store the fifth information.

[1376] Optionally, the perception measurement-related data includes at least one of the following information:

[1377] Perception measurement quantity;

[1378] The indication of the perception measurement quantity;

[1379] Timestamp of the sensed measurement quantity;

[1380] Information of the sending device of the sensed measurement quantity;

[1381] Information of the receiving device of the sensed measurement quantity;

[1382] Coordinate information of the sensed measurement quantity;

[1383] Information for indicating the performance index of the sensed measurement quantity;

[1384] Information for indicating the source of the sensed measurement quantity;

[1385] Information for indicating the type of the sensed measurement quantity;

[1386] Information for indicating the sensing mode of the sensed measurement quantity;

[1387] Configuration information of the target signal;

[1388] Information of the sending device of the target signal;

[1389] Information of the receiving device of the target signal;

[1390] Wherein, the target signal is a signal used for sensing.

[1391] Optionally, the device further includes:

[1392] A second processing module, configured to associate the identifier of the target AI unit with the identifier of the target data set.

[1393] Optionally, the fifth information further includes target information;

[1394] The target information is used to indicate the sensing characteristics of the target data set, or the target information is used to indicate the sensing characteristics and communication characteristics of the target data set.

[1395] Optionally, the target information is used to indicate at least one of the following:

[1396] Whether the samples in the target data set use the sensed measurement quantity;

[1397] The number of the sensed measurement quantities used by the samples in the target data set;

[1398] The proportion of the sensed measurement quantities used by the samples in the target data set;

[1399] Whether the number of the sensed measurement quantities used by the samples in the target data set meets the minimum number threshold;

[1400] Whether the proportion of the sensed measurement quantities used by the samples in the target data set meets the minimum proportion threshold;

[1401] The proportion of the perception measurement quantities used in the samples of the target dataset that meet timeliness requirements;

[1402] The quantity of the perception measurement quantities used in the samples of the target dataset that meet timeliness requirements;

[1403] A first threshold, where the first threshold is a ratio threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in a sample;

[1404] A second threshold, where the second threshold is a ratio threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in a sample.

[1405] Optionally, the target information is used to indicate at least one of the following:

[1406] A dataset identifier, where the dataset corresponding to the dataset identifier includes perception measurement-related data;

[1407] A configuration identifier of the perception measurement quantity;

[1408] An identifier of the perception measurement quantity;

[1409] An identifier of the perception measurement link;

[1410] An identifier of the perception service;

[1411] An identifier of the perception service type;

[1412] The parameter items of the perception measurement quantities used in the samples of the target dataset, where the parameter items include at least one of Doppler of the path, time delay, power, and angle;

[1413] The sending device of the perception measurement quantities used in the samples of the target dataset;

[1414] The receiving device of the perception measurement quantities used in the samples of the target dataset;

[1415] The coordinates of the perception measurement quantities used in the samples of the target dataset;

[1416] The source of the perception measurement quantities used in the samples of the target dataset;

[1417] The perception mode of the perception measurement quantities used in the samples of the target dataset;

[1418] The processing level of the perception measurement quantities used in the samples of the target dataset;

[1419] The perception service of the perception measurement quantities used in the samples of the target dataset;

[1420] The use of the perception measurement quantity used by the samples in the target dataset;

[1421] The performance indicators of the perception measurement quantity used by the samples in the target dataset;

[1422] The perception measurement quantity used by the samples in the target dataset is a fixed perception measurement quantity or a combination of perception measurement quantities;

[1423] The perception measurement quantity used by the samples in the target dataset is a variable perception measurement quantity or a combination of perception measurement quantities;

[1424] The number of perception measurement quantities used by the samples in the target dataset is variable;

[1425] The configuration information of the second target signal;

[1426] Wherein, the second target signal is the target signal corresponding to the perception measurement quantity used by the samples in the target dataset, and the target signal is the signal used for perception.

[1427] In summary, in the embodiments of the present application, by proposing a method for collecting model training data assisted by perception, the network node can expand the input information for model training by obtaining perception information, thereby improving the inference accuracy of the AI model.

[1428] The above device in the embodiments of the present application can be an electronic device, such as an electronic device with an operating system, or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices other than the terminal. Exemplarily, the terminal can include, but is not limited to, the types of terminals listed above, and other devices can be a server, a Network Attached Storage (NAS), etc., which are not specifically limited in the embodiments of the present application.

[1429] The above device provided by the embodiments of the present application can implement Figures 3 to 22 each process implemented by the method embodiment and achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[1430] Such as Figure 29As shown in the figure, an embodiment of the present application further provides a communication device 1100, including a processor 1101 and a memory 1102. A program or instruction that can run on the processor 1101 is stored on the memory 1102. When the program or instruction is executed by the processor 1101, it implements each step of the above-mentioned embodiment of the data set collection method on the first device side, or implements each step of the above-mentioned embodiment of the data set collection method on the second device side, or implements each step of the above-mentioned embodiment of the data set collection method on the third device side, or implements each step of the above-mentioned embodiment of the model training method on the fourth device side, or implements each step of the above-mentioned embodiment of the data set collection method on the fifth device side, or implements each step of the above-mentioned embodiment of the model storage method on the sixth device side, and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[1431] An embodiment of the present application further provides a communication device, including a processor and a communication interface. The communication interface is coupled to the processor. Wherein, the processor is used for: the first device acquires communication measurement-related data and perception measurement-related data; the first device determines a target data set for model training based on the communication measurement-related data and the perception measurement-related data.

[1432] An embodiment of the present application further provides a communication device, including a processor and a communication interface. The communication interface is coupled to the processor. Wherein, the communication interface is used for: receiving a first request from a first device, the first request being used to request perception measurement-related data required for model training; the processor is used for: performing a first operation, where the first operation includes any one of the following:

[1433] Based on the first request, sending a target signal;

[1434] Based on the first request, determining configuration information of the target signal;

[1435] Based on the first request, sending configuration information of the target signal;

[1436] Based on the first request, sending perception measurement-related data;

[1437] Wherein, the target signal is a signal used for perception.

[1438] An embodiment of the present application further provides a communication device, including a processor and a communication interface. The communication interface is coupled to the processor. Wherein, the communication interface is used for: sending first information to a first device, the first information including at least one of communication measurement-related data and perception measurement-related data;

[1439] The first information includes at least one of the following:

[1440] Communication measurement quantities and sensing measurement quantities, where the communication measurement quantities and the sensing measurement quantities are determined to belong to the same training sample;

[1441] Communication measurement quantities and a third indication, where the communication measurement quantities and the sensing measurement quantities indicated by the third indication are determined to belong to the same training sample;

[1442] Communication measurement quantities, where the communication measurement quantities and the sensing measurement quantities of the most recent training sample are determined to belong to the same training sample;

[1443] Or,

[1444] The first information includes at least one of the following:

[1445] A dataset identifier, where the dataset corresponding to the dataset identifier includes sensing measurement-related data;

[1446] A configuration identifier of sensing measurement quantities;

[1447] An identifier of sensing measurement quantities;

[1448] An identifier of a sensing measurement link;

[1449] An identifier of a sensing service;

[1450] An identifier of a sensing service type;

[1451] Sensing measurement quantities;

[1452] A timestamp of sensing measurement quantities;

[1453] Information about the sending device of sensing measurement quantities;

[1454] Coordinate information of sensing measurement quantities;

[1455] Information for indicating a performance metric of sensing measurement quantities;

[1456] Information for indicating the source of sensing measurement quantities;

[1457] Information for indicating the type of sensing measurement quantities;

[1458] Information for indicating the use of sensing measurement quantities;

[1459] Information for indicating the sensing mode of sensing measurement quantities;

[1460] Communication measurement quantities;

[1461] A communication measurement resource identifier;

[1462] A timestamp of communication measurement quantities;

[1463] Information for indicating a first time window;

[1464] Information for indicating a first valid time;

[1465] Configuration identifier of the target signal;

[1466] Configuration information of the target signal;

[1467] Information of the sending device of the target signal;

[1468] Information of the receiving device of the target signal;

[1469] Wherein, the target signal is a signal used for sensing.

[1470] An embodiment of the present application further provides a communication device, including a processor and a communication interface, the communication interface is coupled to the processor, wherein, the communication interface is configured to: receive second information from a first device, the second information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensing measurement-related data; the processor is configured to: use the target data set to train a target AI unit.

[1471] An embodiment of the present application further provides a communication device, including a processor and a communication interface, the communication interface is coupled to the processor, wherein, the communication interface is configured to: receive third information from a first device, or receive fourth information from a fourth device, the third information or the fourth information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and sensing measurement-related data; the processor is configured to: store the target data set and the identifier of the target data set; wherein, the first device is a device that collects the target data set, the fourth device is a device for model training, the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set.

[1472] An embodiment of the present application further provides a communication device, including a processor and a communication interface, the communication interface is coupled to the processor, wherein, the communication interface is configured to: receive fifth information from a fourth device, the fifth information includes a target AI unit, an identifier of the target AI unit and an identifier of a target data set, the data set corresponding to the identifier of the target data set includes communication measurement-related data and sensing measurement-related data; the processor is configured to: store the fifth information.

[1473] The processor of the above communication device is configured to run a program or an instruction to implement the steps in the method embodiment as shown in Figures 3 to 22 and can achieve the same technical effect.

[1474] The above communication device may be a terminal or a network-side device.

[1475] Specifically, Figure 30 FIG. is a schematic diagram of the hardware structure of a terminal for implementing an embodiment of the present application.

[1476] The terminal 1200 includes, but is not limited to, at least some components such as a radio frequency unit 1201, a network module 1202, an audio output unit 1203, an input unit 1204, a sensor 1205, a display unit 1206, a user input unit 1207, an interface unit 1208, a memory 1209, and a processor 1210.

[1477] Those skilled in the art can understand that the terminal 1200 may further include a power source (such as a battery) for supplying power to each component. The power source can be logically connected to the processor 1210 through a power management system, so as to implement functions such as management of charging, discharging, and power consumption management through the power management system. Figure 30 The terminal structure shown in does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements, which will not be elaborated here.

[1478] It should be understood that in the embodiments of the present application, the input unit 1204 may include a graphics processing unit (GPU) 12041 and a microphone 12042. The graphics processor 12041 processes the image data of static pictures or videos obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The display unit 1206 may include a display panel 12061, and the display panel 12061 may be configured in the form of a liquid crystal display, an organic light-emitting diode, etc. The user input unit 1207 includes at least one of a touch panel 12071 and other input devices 12072. The touch panel 12071 is also called a touch screen. The touch panel 12071 may include two parts: a touch detection device and a touch controller. The other input devices 12072 may include, but are not limited to, a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, and a joystick, which will not be elaborated here.

[1479] In the embodiments of the present application, after receiving downlink data from a network-side device, the radio frequency unit 1201 may transmit it to the processor 1210 for processing; in addition, the radio frequency unit 1201 may send uplink data to the network-side device. Generally, the radio frequency unit 1201 includes, but is not limited to, an antenna, an amplifier, a transceiver, a coupler, a low-noise amplifier, a duplexer, etc.

[1480] The memory 1209 can be used to store software programs or instructions as well as various data. The memory 1209 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data. Among them, the first storage area may store an operating system, application programs or instructions required for at least one function (such as a sound playback function, an image playback function, etc.). In addition, the memory 1209 may include a volatile memory or a non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDR SDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM), and a direct rambus random access memory (DRRAM). The memory 1209 in the embodiments of the present application includes but is not limited to these and any other suitable types of memories.

[1481] The processor 1210 may include one or more processing units; optionally, the processor 1210 integrates an application processor and a modem processor. Among them, the application processor mainly processes operations related to the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communication signals, such as a baseband processor. It can be understood that the above-mentioned modem processor may not be integrated into the processor 1210 either.

[1482] Among them, the processor 1210 is used for:

[1483] The first device obtains communication measurement-related data and perception measurement-related data;

[1484] The first device determines a target data set for model training based on the communication measurement-related data and the perception measurement-related data.

[1485] Optionally, the perception measurement related data includes at least one of the following information:

[1486] Perception measurement quantity;

[1487] Indicator of the perception measurement quantity;

[1488] Timestamp of the perception measurement quantity;

[1489] Information of the sending device of the perception measurement quantity;

[1490] Information of the receiving device of the perception measurement quantity;

[1491] Coordinate information of the perception measurement quantity;

[1492] Information for indicating the performance index of the perception measurement quantity;

[1493] Information for indicating the source of the perception measurement quantity;

[1494] Information for indicating the type of the perception measurement quantity;

[1495] Information for indicating the perception mode of the perception measurement quantity;

[1496] Configuration information of the target signal;

[1497] Information of the sending device of the target signal;

[1498] Information of the receiving device of the target signal;

[1499] Wherein, the target signal is a signal used for perception.

[1500] Optionally, the radio frequency unit 1201 is used for:

[1501] Sending a first request to a second device, where the first request is used to request perception measurement related data required for model training.

[1502] Optionally, the first request includes at least one of the following information:

[1503] Information for indicating that the model is in the training stage;

[1504] Dataset identifier, and the dataset corresponding to the dataset identifier includes perception measurement related data;

[1505] Configuration identifier of the perception measurement quantity;

[1506] Identifier of the perception measurement quantity;

[1507] Configuration information of the perception measurement quantity required for model training;

[1508] Information for indicating the number threshold of the perception measurement quantity required for model training;

[1509] Information for indicating the source of the perceptual measurement quantity required for model training;

[1510] Information for indicating the type of the perceptual measurement quantity required for model training;

[1511] Information for indicating the perceptual mode of the perceptual measurement quantity required for model training;

[1512] Information for indicating the perceptual requirement;

[1513] Configuration information of the target signal;

[1514] Information of the sending device of the target signal;

[1515] Information of the receiving device of the target signal;

[1516] Wherein, the target signal is a signal used for perception.

[1517] Optionally, the processor 1210 is further configured to:

[1518] Generate training samples according to the obtained communication measurement-related data and perceptual measurement-related data, and the target data set includes the training samples.

[1519] Optionally, the obtained perceptual measurement-related data includes a first perceptual measurement quantity;

[1520] The processor 1210 is further configured to perform at least one of the following:

[1521] Generate a first training sample based on the target timestamp, the timestamp of the first perceptual measurement quantity, and the effective duration of the first perceptual measurement quantity, and determine the target data set based on the first training sample;

[1522] Generate a first training sample based on the first indication and the target timestamp, and determine the target data set based on the first training sample;

[1523] Generate a first training sample based on whether a second indication is received, and determine the target data set based on the first training sample;

[1524] Wherein, the target timestamp is the measurement timestamp of a first communication measurement quantity, and the first communication measurement quantity is the last communication measurement quantity corresponding to the true value or label in the communication measurement-related data;

[1525] The first indication is used to indicate the failure time or failure time difference of the first perceptual measurement quantity;

[1526] The second indication is used to indicate the failure of the first perceptual measurement quantity.

[1527] Optionally, the method for determining the effective duration of the first sensed measurement quantity includes at least one of the following:

[1528] When a first effective duration is pre-configured or defined, the effective duration of the first sensed measurement quantity is the first effective duration;

[1529] When a second effective duration is pre-configured or defined and the first target signal carries a time adjustment value, the effective duration of the first sensed measurement quantity is determined according to the second effective duration and the time adjustment value;

[1530] When the first target signal carries a third effective duration, the effective duration of the first sensed measurement quantity is the third effective duration;

[1531] Wherein, the first target signal is the target signal corresponding to the first sensed measurement quantity, and the target signal is a signal used for sensing.

[1532] Optionally, the processor 1210 is further configured to perform at least one of the following:

[1533] When the time corresponding to the target timestamp is earlier than the first time, generate a first training sample including the communication measurement related data and the first sensed measurement quantity; the first time is the time corresponding to the timestamp of the first sensed measurement quantity plus the effective duration of the first sensed measurement quantity;

[1534] When the time corresponding to the target timestamp is earlier than the second time, generate a first training sample including the communication measurement related data and the first sensed measurement quantity; the second time is the expiration time of the first sensed measurement quantity determined based on the first indication;

[1535] When the first device does not receive the second indication, generate a first training sample including the communication measurement related data and the first sensed measurement quantity.

[1536] Optionally, the radio frequency unit 1201 is further configured to:

[1537] Receive a first piece of information from a third device, where the first piece of information includes at least one of communication measurement related data and sensed measurement related data;

[1538] The processor 1210 is further configured to:

[1539] Generate a training sample based on the first piece of information, and determine the target data set based on the training sample; or, generate a training sample based on the first piece of information and the stored historical data, and determine the target data set based on the training sample;

[1540] Among them, the historical data includes at least one of communication measurement-related data and sensing measurement-related data.

[1541] Optionally, the first information includes at least one of the following:

[1542] Dataset identifier, and the dataset corresponding to the dataset identifier includes sensing measurement-related data;

[1543] Configuration identifier of the sensing measurement quantity;

[1544] Identifier of the sensing measurement quantity;

[1545] Identifier of the sensing measurement link;

[1546] Identifier of the sensing service;

[1547] Identifier of the sensing service type;

[1548] Sensing measurement quantity;

[1549] Timestamp of the sensing measurement quantity;

[1550] Information of the sending device of the sensing measurement quantity;

[1551] Coordinate information of the sensing measurement quantity;

[1552] Information for indicating the performance index of the sensing measurement quantity;

[1553] Information for indicating the source of the sensing measurement quantity;

[1554] Information for indicating the type of the sensing measurement quantity;

[1555] Information for indicating the use of the sensing measurement quantity;

[1556] Information for indicating the sensing mode of the sensing measurement quantity;

[1557] Communication measurement quantity;

[1558] Communication measurement resource identifier;

[1559] Timestamp of the communication measurement quantity;

[1560] Information for indicating the first time window;

[1561] Information for indicating the first valid time;

[1562] Configuration identifier of the target signal;

[1563] Configuration information of the target signal;

[1564] Information of the sending device of the target signal;

[1565] Information of the receiving device of the target signal;

[1566] Wherein, the target signal is a signal used for sensing.

[1567] Optionally, the first information includes a second sensing measurement, a timestamp of the second sensing measurement, a second communication measurement, and a timestamp of the second communication measurement;

[1568] The processor 1210 is further configured to perform at least one of the following:

[1569] When a time difference between the timestamp of the second sensing measurement and the timestamp of the second communication measurement is less than or equal to a first time window indicated by the first information, generate a second training sample including the second communication measurement and the second sensing measurement;

[1570] When a time difference between the timestamp of the second sensing measurement and the timestamp of the second communication measurement is less than or equal to a predefined second time window, generate a second training sample including the second communication measurement and the second sensing measurement.

[1571] Optionally, the first information includes a third sensing measurement and a timestamp of the third sensing measurement;

[1572] The processor 1210 is further configured to:

[1573] Generate a third training sample including a third communication measurement and the third sensing measurement;

[1574] Wherein, the third communication measurement includes at least one of the following:

[1575] A communication measurement when a time difference between a timestamp of the historical data and a timestamp of the third sensing measurement is less than or equal to the first time window indicated by the first information;

[1576] A communication measurement when a time difference between a timestamp of the historical data and a timestamp of the third sensing measurement is less than or equal to the predefined second time window.

[1577] Optionally, the first information includes a fourth communication measurement and a timestamp of the fourth communication measurement;

[1578] The processor 1210 is further configured to:

[1579] Generate a fourth training sample including the fourth communication measurement and a fourth sensing measurement;

[1580] Wherein, the fourth communication measurement includes at least one of the following:

[1581] The time difference between the timestamp of the historical data and the timestamp of the fourth communication measurement quantity is less than or equal to the perception measurement quantity of the first time window indicated by the first information;

[1582] The timestamp of the historical data is less than or equal to the perception measurement quantity of the first valid time indicated by the first information;

[1583] The timestamp of the historical data is less than or equal to the perception measurement quantity of a predefined second valid time.

[1584] Optionally, the processor 1210 is further configured to perform at least one of the following:

[1585] When the first information includes a fifth communication measurement quantity and a fifth perception measurement quantity, generate a fifth training sample including the fifth communication measurement quantity and the fifth perception measurement quantity;

[1586] When the first information includes a fifth communication measurement quantity and a third indication, generate a fifth training sample including the fifth communication measurement quantity and a fifth perception measurement quantity, where the fifth perception measurement quantity is the perception measurement quantity corresponding to the third indication;

[1587] When the first information includes a fifth communication measurement quantity, generate a fifth training sample including the fifth communication measurement quantity and a fifth perception measurement quantity, where the fifth perception measurement quantity is the perception measurement quantity of the most recent training sample.

[1588] Optionally, the third indication includes at least one of the following:

[1589] The identifier of the training sample;

[1590] The identifier of the resource used for the communication measurement result;

[1591] The reporting identifier of the communication measurement result;

[1592] The identifier of the resource used for the perception measurement quantity;

[1593] The reporting identifier of the perception measurement quantity;

[1594] The identifier of the perception measurement quantity.

[1595] Optionally, the radio frequency unit 1201 is further configured to:

[1596] Send second information to a fourth device, where the second information includes the target data set and the identifier of the target data set, the fourth device is a device for model training, and the identifier of the target data set is associated with the perception characteristics of the target data set, or the identifier of the target data set is associated with the perception characteristics and communication characteristics of the target data set.

[1597] Optionally, the radio frequency unit 1201 is further configured to:

[1598] Send third information to a fifth device, where the third information includes the target data set and an identifier of the target data set, and the fifth device is a device storing a model training data set.

[1599] Optionally, the second information or the third information further includes target information;

[1600] The target information is used to indicate the perception characteristics of the target data set, or the target information is used to indicate the perception characteristics and communication characteristics of the target data set.

[1601] Optionally, the target information is used to indicate at least one of the following:

[1602] Whether a perception measurement quantity is used for a sample in the target data set;

[1603] The quantity of perception measurement quantities used for a sample in the target data set;

[1604] The proportion of perception measurement quantities used for a sample in the target data set;

[1605] Whether the quantity of perception measurement quantities used for a sample in the target data set meets a minimum quantity threshold;

[1606] Whether the proportion of perception measurement quantities used for a sample in the target data set meets a minimum proportion threshold;

[1607] The proportion of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for a sample in the target data set;

[1608] The quantity of perception measurement quantities that meet timeliness requirements among the perception measurement quantities used for a sample in the target data set;

[1609] A first threshold, where the first threshold is a proportion threshold between the quantity of valid perception measurement quantities and the total quantity of perception measurement quantities in a sample;

[1610] A second threshold, where the second threshold is a proportion threshold between the quantity of actually used perception measurement quantities and the total quantity of the maximum supportable perception measurement quantities in a sample.

[1611] Optionally, the target information is used to indicate at least one of the following:

[1612] A data set identifier, where the data set corresponding to the data set identifier includes perception measurement related data;

[1613] A configuration identifier of the perception measurement quantity;

[1614] An identifier of the perception measurement quantity;

[1615] Identification of the sensing measurement link;

[1616] Identification of the sensing service;

[1617] Identification of the sensing service type;

[1618] Parameter items of the sensing measurement quantity used by the samples in the target dataset, where the parameter items include at least one of Doppler of the diameter, time delay, power, and angle;

[1619] Transmitting device of the sensing measurement quantity used by the samples in the target dataset;

[1620] Receiving device of the sensing measurement quantity used by the samples in the target dataset;

[1621] Coordinates of the sensing measurement quantity used by the samples in the target dataset;

[1622] Source of the sensing measurement quantity used by the samples in the target dataset;

[1623] Sensing mode of the sensing measurement quantity used by the samples in the target dataset;

[1624] Processing level of the sensing measurement quantity used by the samples in the target dataset;

[1625] Sensing service of the sensing measurement quantity used by the samples in the target dataset;

[1626] Use of the sensing measurement quantity used by the samples in the target dataset;

[1627] Performance indicators of the sensing measurement quantity used by the samples in the target dataset;

[1628] The sensing measurement quantity used by the samples in the target dataset is a fixed sensing measurement quantity or a combination of sensing measurement quantities;

[1629] The sensing measurement quantity used by the samples in the target dataset is a variable sensing measurement quantity or a combination of sensing measurement quantities;

[1630] The number of sensing measurement quantities used by the samples in the target dataset is variable;

[1631] Configuration information of the second target signal;

[1632] Wherein, the second target signal is the target signal corresponding to the sensing measurement quantity used by the samples in the target dataset, and the target signal is the signal used for sensing.

[1633] In summary, in the embodiments of the present application, by proposing a method for collecting model training data assisted by sensing, network nodes can obtain sensing information to expand the input information for model training, thereby improving the inference accuracy of the AI model.

[1634] It can be understood that the implementation processes of the various implementation manners mentioned in this embodiment can Figures 3 to 22 be described in the relevant descriptions of the method embodiments and achieve the same or corresponding technical effects. To avoid repetition, they are not elaborated herein.

[1635] Specifically, the embodiments of the present application further provide a network-side device. As Figure 31 shown, the network-side device 1300 includes: an antenna 131, a radio frequency device 132, a baseband device 133, a processor 134, and a memory 135. The antenna 131 is connected to the radio frequency device 132. In the uplink direction, the radio frequency device 132 receives information through the antenna 131 and sends the received information to the baseband device 133 for processing. In the downlink direction, the baseband device 133 processes the information to be sent and sends it to the radio frequency device 132. After processing the received information, the radio frequency device 132 sends it out through the antenna 131.

[1636] The methods executed by the first device, the second device, the third device, and the fourth device in the above embodiments can all be implemented in the baseband device 133, and the baseband device 133 includes a baseband processor.

[1637] The baseband device 133 may include, for example, at least one baseband board, and multiple chips are provided on the baseband board. As Figure 31 shown, one of the chips is, for example, a baseband processor, which is connected to the memory 135 through a bus interface to call the program in the memory 135 and execute the operations of the network device shown in the above method embodiments.

[1638] The network-side device may further include a network interface 136, and this interface is, for example, a Common Public Radio Interface (CPRI).

[1639] Specifically, the network-side device 1300 in the embodiments of the present application further includes: instructions or programs stored on the memory 135 and executable on the processor 134. The processor 134 calls the instructions or programs in the memory 135 to execute Figures 21 to 24 the methods executed by the respective modules shown, and achieve the same technical effects. To avoid repetition, they are not elaborated herein.

[1640] Specifically, the embodiments of the present application further provide a network-side device. As Figure 32As shown, the network-side device 1400 includes: a processor 1401, a network interface 1402, and a memory 1403. Among them, the network interface 1402 is, for example, a common public radio interface (CPRI).

[1641] Specifically, the network-side device 1400 in the embodiment of the present application further includes: instructions or programs stored on the memory 1403 and executable on the processor 1401. The processor 1401 calls the instructions or programs in the memory 1403 to execute Figures 21 to 24 the methods executed by the modules shown, and achieve the same technical effects. To avoid repetition, they will not be elaborated here.

[1642] The embodiment of the present application further provides a readable storage medium. Programs or instructions are stored on the readable storage medium. When the programs or instructions are executed by a processor, they implement Figures 3 to 22 each process of the method embodiment and can achieve the same technical effects. To avoid repetition, they will not be elaborated here.

[1643] Among them, the processor is the processor in the terminal described in the above embodiment. The readable storage medium includes computer-readable storage media, such as computer read-only memory ROM, random access memory RAM, magnetic disks, or optical discs, etc. In some examples, the readable storage medium may be a non-transitory readable storage medium.

[1644] The embodiment of the present application further provides a chip. The chip includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement Figures 3 to 22 each process of the method embodiment and can achieve the same technical effects. To avoid repetition, they will not be elaborated here.

[1645] It should be understood that the chip mentioned in the embodiment of the present application may also be referred to as a system-on-chip, system chip, chip system, or system-on-chip, etc.

[1646] The embodiment of the present application further provides a computer program / program product. The computer program / program product is stored in a storage medium. The computer program / program product is executed by at least one processor to implement Figures 3 to 22 each process of the method embodiment and can achieve the same technical effects. To avoid repetition, they will not be elaborated here.

[1647] The embodiment of the present application further provides a communication system, including: a first device and a second device. The first device can be used to execute the steps of the data set collection method on the first device side as above, and the second device can be used to execute the steps of the data set collection method on the second device side as above.

[1648] Optionally, the communication system further includes a third device, which can be used to perform the steps of the data set collection method on the third device side as described above.

[1649] Optionally, the communication system further includes a fourth device, which can be used to perform the steps of the model training method on the fourth device side as described above.

[1650] Optionally, the communication system further includes a fifth device, which can be used to perform the steps of the data set collection method on the fifth device side as described above.

[1651] Optionally, the communication system further includes a sixth device, which can be used to perform the steps of the model storage method on the sixth device side as described above.

[1652] It should be noted that in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or device including that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in a reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted or combined. Additionally, the features described with reference to certain examples may be combined in other examples.

[1653] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of a computer software product plus a necessary general hardware platform, and of course, they can also be implemented by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, magnetic disk, optical disk, etc.) and includes several instructions for causing a terminal or a network-side device to perform the methods described in various embodiments of the present application.

[1654] The embodiments of the present application have been described above in conjunction with the accompanying drawings, but the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms of embodiments without departing from the purpose of the present application and the scope protected by the claims. These embodiments are all within the protection scope of the present application.

Claims

1. A method for collecting a dataset, characterized in that, Including: A first device obtains communication measurement-related data and sensing measurement-related data; The first device determines a target data set for model training based on the communication measurement-related data and the sensing measurement-related data.

2. The method according to claim 1, characterized in that The sensing measurement-related data includes at least one of the following information: Sensing measurement quantity; Indicator of the sensing measurement quantity; Timestamp of the sensing measurement quantity; Information of the sending device of the sensing measurement quantity; Information of the receiving device of the sensing measurement quantity; Coordinate information of the sensing measurement quantity; Information for indicating a performance metric of the sensing measurement quantity; Information for indicating the source of the sensing measurement quantity; Information for indicating the type of the sensing measurement quantity; Information for indicating the sensing mode of the sensing measurement quantity; Configuration information of the target signal; Information of the sending device of the target signal; Information of the receiving device of the target signal; Wherein, the target signal is a signal used for sensing.

3. The method according to claim 1 or 2, characterized in that, The method further includes: The first device sends a first request to a second device, and the first request is used to request sensing measurement-related data required for model training.

4. The method according to claim 3, characterized in that, The first request includes at least one of the following information: Information for indicating that the model is in the training phase; Data set identifier, and the data set corresponding to the data set identifier includes sensing measurement-related data; Configuration identifier of the sensing measurement quantity; Identifier of the sensing measurement quantity; Configuration information of the sensing measurement quantity required for model training; Information for indicating a number threshold of the sensing measurement quantity required for model training; Information for indicating the source of the sensing measurement quantity required for model training; Information for indicating the type of the sensing measurement quantity required for model training; Information for indicating the sensing mode of the sensing measurement quantity required for model training; Information for indicating sensing requirements; Configuration information of the target signal; Information of the sending device of the target signal; Information of the receiving device of the target signal; Wherein, the target signal is a signal used for sensing.

5. The method according to claim 1, wherein The sensing measurement-related data includes a first sensing measurement quantity; The first device determines a target data set for model training based on the communication measurement-related data and the sensing measurement-related data, including at least one of the following: The first device generates a first training sample based on a target timestamp, the timestamp of the first sensing measurement quantity, and the effective duration of the first sensing measurement quantity, and determines the target data set based on the first training sample; The first device generates a first training sample based on a first indication and a target timestamp, and determines the target data set based on the first training sample; The first device generates a first training sample based on whether a second indication is received, and determines the target data set based on the first training sample; Wherein, the target timestamp is the measurement timestamp of a first communication measurement quantity, and the first communication measurement quantity is the last communication measurement quantity corresponding to the true value or label in the communication measurement-related data; The first indication is used to indicate the failure time or the failure time difference of the first sensing measurement quantity; The second indication is used to indicate that the first sensing measurement quantity fails.

6. The method according to claim 5, wherein The determination method of the effective duration of the first sensing measurement quantity includes at least one of the following: When a first effective duration is pre-configured or defined, the effective duration of the first perception measurement quantity is the first effective duration; When a second effective duration is pre-configured or defined and the first target signal carries a time adjustment value, the effective duration of the first perception measurement quantity is determined according to the second effective duration and the time adjustment value; When the first target signal carries a third effective duration, the effective duration of the first perception measurement quantity is the third effective duration; Wherein, the first target signal is the target signal corresponding to the first perception measurement quantity, and the target signal is a signal used for perception.

7. The method according to claim 5 or 6, characterized in that The first device generates a first training sample based on a target timestamp, the timestamp of the first perception measurement quantity, and the effective duration of the first perception measurement quantity, including: When the time corresponding to the target timestamp is earlier than a first time, the first device generates a first training sample including the communication measurement related data and the first perception measurement quantity; the first time is the time corresponding to the timestamp of the first perception measurement quantity plus the effective duration of the first perception measurement quantity; The first device generates a first training sample based on a first indication and a target timestamp, including: When the time corresponding to the target timestamp is earlier than a second time, the first device generates a first training sample including the communication measurement related data and the first perception measurement quantity; the second time is the expiration time of the first perception measurement quantity determined based on the first indication; The first device generates a first training sample based on whether a second indication is received, including: When the first device does not receive the second indication, the first device generates a first training sample including the communication measurement related data and the first perception measurement quantity.

8. The method according to claim 1, characterized in that, The first device determines a target data set for model training based on the communication measurement related data and the perception measurement related data, including: The first device receives first information from a third device, and the first information includes at least one of communication measurement related data and perception measurement related data; The first device generates a training sample based on the first information and determines the target data set based on the training sample; or, the first device generates a training sample based on the first information and stored historical data and determines the target data set based on the training sample; Wherein, the historical data includes at least one of communication measurement related data and perception measurement related data.

9. The method according to claim 8, characterized in that, The first information includes at least one of the following: A data set identifier, and the data set corresponding to the data set identifier includes perception measurement related data; A configuration identifier of a perception measurement quantity; An identifier of a perception measurement quantity; An identifier of a perception measurement link; An identifier of a perception service; An identifier of a perception service type; A perception measurement quantity; A timestamp of a perception measurement quantity; Information about the sending device of a perception measurement quantity; Coordinate information of a perception measurement quantity; Information for indicating a performance index of a perception measurement quantity; Information for indicating the source of a perception measurement quantity; Information for indicating the type of a perception measurement quantity; Information for indicating the use of a sensed measurement quantity; Information for indicating the sensing mode of a sensed measurement quantity; Communication measurement quantity; Communication measurement resource identifier; Timestamp of the communication measurement quantity; Information for indicating a first time window; Information for indicating a first valid time; Configuration identifier of a target signal; Configuration information of a target signal; Information of the transmitting device of the target signal; Information of the receiving device of the target signal; Wherein, the target signal is a signal used for sensing.

10. The method according to claim 8 or 9, characterized in that The first information includes a second sensed measurement quantity, a timestamp of the second sensed measurement quantity, a second communication measurement quantity, and a timestamp of the second communication measurement quantity; Based on the first information, the first device generates a training sample, including at least one of the following: When a time difference between the timestamp of the second sensed measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to a first time window indicated by the first information, the first device generates a second training sample including the second communication measurement quantity and the second sensed measurement quantity; When a time difference between the timestamp of the second sensed measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to a predefined second time window, the first device generates a second training sample including the second communication measurement quantity and the second sensed measurement quantity.

11. The method according to claim 8 or 9, characterized in that The first information includes a third sensed measurement quantity and a timestamp of the third sensed measurement quantity; Based on the first information and stored historical data, the first device generates a training sample, including: The first device generates a third training sample including a third communication measurement quantity and the third sensed measurement quantity; Wherein, the third communication measurement quantity includes at least one of the following: A communication measurement quantity when a time difference between a timestamp of the historical data and a timestamp of the third sensed measurement quantity is less than or equal to a first time window indicated by the first information; A communication measurement quantity when a time difference between a timestamp of the historical data and a timestamp of the third sensed measurement quantity is less than or equal to a predefined second time window.

12. The method according to claim 8 or 9, characterized in that, The first information includes a fourth communication measurement quantity and a timestamp of the fourth communication measurement quantity; Based on the first information and stored historical data, the first device generates a training sample, including: The first device generates a fourth training sample including the fourth communication measurement quantity and a fourth sensed measurement quantity; Wherein, the fourth communication measurement quantity includes at least one of the following: A sensed measurement quantity when a time difference between a timestamp of the historical data and a timestamp of the fourth communication measurement quantity is less than or equal to a first time window indicated by the first information; A sensed measurement quantity when a timestamp of the historical data is less than or equal to a first valid time indicated by the first information; A sensed measurement quantity when a timestamp of the historical data is less than or equal to a predefined second valid time.

13. The method according to claim 8 or 9, characterized in that Based on the first information, the first device generates a training sample, including at least one of the following: When the first information includes a fifth communication measurement quantity and a fifth sensed measurement quantity, the first device generates a fifth training sample including the fifth communication measurement quantity and the fifth sensed measurement quantity; When the first information includes a fifth communication measurement quantity and a third indication, the first device generates a fifth training sample including the fifth communication measurement quantity and a fifth sensing measurement quantity, where the fifth sensing measurement quantity is the sensing measurement quantity corresponding to the third indication; When the first information includes a fifth communication measurement quantity, the first device generates a fifth training sample including the fifth communication measurement quantity and a fifth sensing measurement quantity, where the fifth sensing measurement quantity is the sensing measurement quantity of the most recent training sample.

14. The method according to claim 13, wherein The third indication includes at least one of the following: The identifier of the training sample; The identifier of the resource used for the communication measurement result; The reporting identifier of the communication measurement result; The identifier of the resource used for the sensing measurement quantity; The reporting identifier of the sensing measurement quantity; The identifier of the sensing measurement quantity.

15. The method according to any one of claims 1 to 14, characterized in that, The method further includes: The first device sends second information to a fourth device, where the second information includes the target data set and the identifier of the target data set. The fourth device is a device for model training, and the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set.

16. The method according to any one of claims 1 to 15, characterized in that, The method further includes: The first device sends third information to a fifth device, where the third information includes the target data set and the identifier of the target data set. The fifth device is a device for storing the model training data set.

17. The method according to claim 15 or 16, characterized in that, The second information or the third information further includes target information; The target information is used to indicate the sensing characteristics of the target data set, or the target information is used to indicate the sensing characteristics and communication characteristics of the target data set.

18. The method according to claim 17, characterized in that, The target information is used to indicate at least one of the following: Whether the samples in the target data set use sensing measurement quantities; The quantity of sensing measurement quantities used by the samples in the target data set; The proportion of sensing measurement quantities used by the samples in the target data set; Whether the quantity of sensing measurement quantities used by the samples in the target data set meets a minimum quantity threshold; Whether the proportion of sensing measurement quantities used by the samples in the target data set meets a minimum proportion threshold; The proportion of sensing measurement quantities that meet timeliness requirements among the sensing measurement quantities used by the samples in the target data set; The quantity of sensing measurement quantities that meet timeliness requirements among the sensing measurement quantities used by the samples in the target data set; A first threshold, where the first threshold is a proportion threshold between the quantity of valid sensing measurement quantities and the total quantity of sensing measurement quantities in a sample; A second threshold, where the second threshold is a proportion threshold between the quantity of actually used sensing measurement quantities and the total quantity of the maximum supportable sensing measurement quantities in a sample.

19. The method according to claim 17 or 18, characterized in that, The target information is used to indicate at least one of the following: A data set identifier, where the data set corresponding to the data set identifier includes sensing measurement-related data; The configuration identifier of the sensing measurement quantity; The identifier of the sensing measurement quantity; The identifier of the sensing measurement link; The identifier of the sensing service; The identifier of the sensing service type; The parameter items of the sensing measurement quantities used by the samples in the target data set, where the parameter items include at least one of Doppler, delay, power, and angle of the path; The sending device of the perception measurement quantity used by the samples in the target dataset; The receiving device of the perception measurement quantity used by the samples in the target dataset; The coordinates of the perception measurement quantity used by the samples in the target dataset; The source of the perception measurement quantity used by the samples in the target dataset; The perception mode of the perception measurement quantity used by the samples in the target dataset; The processing level of the perception measurement quantity used by the samples in the target dataset; The perception service of the perception measurement quantity used by the samples in the target dataset; The use of the perception measurement quantity used by the samples in the target dataset; The performance indicators of the perception measurement quantity used by the samples in the target dataset; The perception measurement quantity used by the samples in the target dataset is a fixed perception measurement quantity or a combination of perception measurement quantities; The perception measurement quantity used by the samples in the target dataset is a variable perception measurement quantity or a combination of perception measurement quantities; The number of the perception measurement quantities used by the samples in the target dataset is variable; The configuration information of the second target signal; Wherein, the second target signal is the target signal corresponding to the perception measurement quantity used by the samples in the target dataset, and the target signal is the signal used for perception.

20. The method according to any one of claims 1 to 19, characterized in that The method further includes: The first device uses the target dataset to train the target AI unit.

21. A model training method, characterized in that, Including: The fourth device receives the second information from the first device, and the second information includes the target dataset and the identifier of the target dataset; The target dataset includes communication measurement related data and perception measurement related data; The fourth device uses the target dataset to train the target AI unit.

22. The method according to claim 21, wherein The method further includes: The fourth device sends the fourth information to the fifth device, and the fourth information includes the target dataset and the identifier of the target dataset, and the fifth device is the device storing the model training dataset.

23. The method according to claim 21 or 22, characterized in that, The method further includes: The fourth device sends the fifth information to the sixth device, and the fifth information includes the target AI unit, the identifier of the target AI unit and the identifier of the target dataset, and the sixth device is the device storing the target AI unit.

24. The method according to any one of claims 21 to 23, characterized in that, The second information or the fourth information or the fifth information further includes target information; The target information is used to indicate the perception characteristics of the target dataset, or the target information is used to indicate the perception characteristics and communication characteristics of the target dataset.

25. A dataset collection device, characterized in that, Applied to the first device, the device includes: The first processing module is used to obtain communication measurement related data and perception measurement related data; The second processing module is used to determine the target dataset for model training based on the communication measurement related data and the perception measurement related data.

26. The device according to claim 25, characterized in that, The perception measurement related data includes the first perception measurement quantity; The second processing module is specifically used for at least one of the following: Generating a first training sample based on the target timestamp, the timestamp of the first perception measurement quantity and the effective duration of the first perception measurement quantity, and determining the target dataset based on the first training sample; Generate a first training sample based on the first indication and the target timestamp, and determine the target data set based on the first training sample; Generate a first training sample based on whether the second indication is received, and determine the target data set based on the first training sample; Wherein, the target timestamp is the measurement timestamp of the first communication measurement quantity, and the first communication measurement quantity is the last communication measurement quantity corresponding to the true value or label in the communication measurement related data; The first indication is used to indicate the failure time or the failure time difference of the first sensing measurement quantity; The second indication is used to indicate the failure of the first sensing measurement quantity.

27. The device according to claim 26, wherein, The determination method of the effective duration of the first sensing measurement quantity includes at least one of the following: When a first effective duration is pre-configured or defined, the effective duration of the first sensing measurement quantity is the first effective duration; When a second effective duration is pre-configured or defined and the first target signal carries a time adjustment value, the effective duration of the first sensing measurement quantity is determined according to the second effective duration and the time adjustment value; When the first target signal carries a third effective duration, the effective duration of the first sensing measurement quantity is the third effective duration; Wherein, the first target signal is the target signal corresponding to the first sensing measurement quantity, and the target signal is the signal used for sensing.

28. The device according to claim 26 or 27, characterized in that, The second processing module is specifically used for at least one of the following: When the time corresponding to the target timestamp is earlier than the first time, generate a first training sample including the communication measurement related data and the first sensing measurement quantity; the first time is the time corresponding to the timestamp of the first sensing measurement quantity plus the effective duration of the first sensing measurement quantity; When the time corresponding to the target timestamp is earlier than the second time, generate a first training sample including the communication measurement related data and the first sensing measurement quantity; The second time is the failure time of the first sensing measurement quantity determined based on the first indication; When the first device does not receive the second indication, generate a first training sample including the communication measurement related data and the first sensing measurement quantity.

29. The device according to claim 25, wherein The second processing module includes: A receiving unit, configured to receive first information from a third device, where the first information includes at least one of communication measurement related data and sensing measurement related data; A processing unit, configured to generate a training sample based on the first information and determine the target data set based on the training sample; or, generate a training sample based on the first information and the stored historical data, and determine the target data set based on the training sample; Wherein, the historical data includes at least one of communication measurement related data and sensing measurement related data.

30. The device according to claim 29, characterized in that, The first information includes a second sensing measurement quantity, the timestamp of the second sensing measurement quantity, a second communication measurement quantity, and the timestamp of the second communication measurement quantity; The processing unit is specifically used for at least one of the following: When the time difference between the timestamp of the second sensing measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to the first time window indicated by the first information, generate a second training sample including the second communication measurement quantity and the second sensing measurement quantity; When the time difference between the timestamp of the second sensing measurement quantity and the timestamp of the second communication measurement quantity is less than or equal to a predefined second time window, generate a second training sample including the second communication measurement quantity and the second sensing measurement quantity.

31. The device according to claim 29, wherein The first information includes a third sensing measurement quantity and the timestamp of the third sensing measurement quantity; The processing unit is specifically configured to: Generate a third training sample including a third communication measurement quantity and the third sensing measurement quantity; Wherein, the third communication measurement quantity includes at least one of the following: The communication measurement quantity whose time difference between the timestamp of the historical data and the timestamp of the third sensing measurement quantity is less than or equal to the first time window indicated by the first information; The communication measurement quantity whose time difference between the timestamp of the historical data and the timestamp of the third sensing measurement quantity is less than or equal to a predefined second time window.

32. The device according to claim 29, characterized in that, The first information includes a fourth communication measurement quantity and the timestamp of the fourth communication measurement quantity; The processing unit is specifically configured to: Generate a fourth training sample including the fourth communication measurement quantity and a fourth sensing measurement quantity; Wherein, the fourth communication measurement quantity includes at least one of the following: The sensing measurement quantity whose time difference between the timestamp of the historical data and the timestamp of the fourth communication measurement quantity is less than or equal to the first time window indicated by the first information; The sensing measurement quantity whose timestamp of the historical data is less than or equal to the first effective time indicated by the first information; The sensing measurement quantity whose timestamp of the historical data is less than or equal to a predefined second effective time.

33. The device according to claim 29, wherein The processing unit is specifically configured to at least one of the following: When the first information includes a fifth communication measurement quantity and a fifth sensing measurement quantity, generate a fifth training sample including the fifth communication measurement quantity and the fifth sensing measurement quantity; When the first information includes a fifth communication measurement quantity and a third indication, generate a fifth training sample including the fifth communication measurement quantity and a fifth sensing measurement quantity, and the fifth sensing measurement quantity is the sensing measurement quantity corresponding to the third indication; When the first information includes a fifth communication measurement quantity, generate a fifth training sample including the fifth communication measurement quantity and a fifth sensing measurement quantity, and the fifth sensing measurement quantity is the sensing measurement quantity of the most recent training sample.

34. The device according to any one of claims 25 to 33, characterized in that It further includes at least one of the following: A second sending module, configured to send second information to a fourth device, where the second information includes the target data set and the identifier of the target data set, the fourth device is a device for model training, and the identifier of the target data set is associated with the sensing characteristics of the target data set, or the identifier of the target data set is associated with the sensing characteristics and communication characteristics of the target data set; A third processing module, configured to use the target data set to train a target AI unit.

35. A model training device, characterized in that, Applied to a fourth device, the apparatus includes: A receiving module, configured to receive second information from a first device, where the second information includes a target data set and an identifier of the target data set; the target data set includes communication measurement-related data and perception measurement-related data; A processing module, configured to use the target data set to perform training of a target AI unit.

36. A communication device, characterized in that, It includes a processor and a memory, and the memory stores programs or instructions that can run on the processor. When the programs or instructions are executed by the processor, the steps of the method according to any one of claims 1 to 24 are implemented.

37. A readable storage medium, characterized in that, Programs or instructions are stored on the readable storage medium. When the programs or instructions are executed by a processor, the steps of the method according to any one of claims 1 to 24 are implemented.

Images