Device sensing methods, apparatus and storage medium

Abstract

Provided in the present disclosure are device sensing methods, an apparatus and a storage medium. A method comprises: determining a first signal, the first signal being a signal obtained upon receiving a sensing signal reflected by a third device, or the first signal being a signal obtained upon receiving a sensing signal reflected by a fourth device, wherein the sensing signal is sent by a second device; on the basis of the first signal, performing channel estimation to determine a channel state information (CSI) matrix; inputting the CSI matrix into a first artificial intelligence (AI) model to acquire the number formula (I) of third devices output by the first AI model (I); and, on the basis of the CSI matrix, formula (I) and at least one second AI model, determining sensing results obtained by separately sensing the formula (I) third devices. The present disclosure can reduce effects of clutter, noise, etc. on sensing results during device sensing, thus improving the accuracy and reliability of the device sensing, reducing the degree of complexity of computations for determining the sensing results, and improving the availability of ISAC technology.

Description

Device perception method and device, and storage medium Technical Field

[0001] The present disclosure relates to the field of communications, and in particular to a device perception method and apparatus, and a storage medium. Background Art

[0002] Integrated Sensing and Communication (ISAC) is a crucial research area in wireless communications. ISAC technology enables sensing services based on existing mobile communications infrastructure, leveraging the advantages of mobile communication networks to meet the sensing needs of various service scenarios. This technology facilitates numerous application services, including detection, positioning and tracking, environmental reconstruction, target imaging, and gesture and posture recognition.

[0003] Summary of the Invention

[0004] To improve the reliability of device perception and the availability of ISAC technology, embodiments of the present disclosure provide a device perception method and apparatus, and a storage medium.

[0005] According to a first aspect of an embodiment of the present disclosure, a device perception method is provided. The method is performed by a first device and includes:

[0006] Determining a first signal, where the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device;

[0007] Perform channel estimation based on the first signal to determine a channel state information CSI matrix;

[0008] Input the CSI matrix into the first artificial intelligence AI model and obtain the number of third devices output by the first AI model

[0009] Based on the CSI matrix, and at least one second AI model, determining The perception results obtained by the third devices are respectively perceived.

[0010] According to a second aspect of an embodiment of the present disclosure, a device perception method is provided. The method is performed by a second device and includes:

[0011] A perception signal is sent to the third device or the fourth device.

[0012] According to a third aspect of an embodiment of the present disclosure, a device perception method is provided. The method is performed by a third device and includes:

[0013] The sensing signal is reflected to the first device, where the sensing signal is sent by the second device to the third device.

[0014] According to a fourth aspect of an embodiment of the present disclosure, a device perception method is provided. The method is performed by a fourth device and includes:

[0015] The sensing signal is reflected to the first device, where the sensing signal is sent by the second device to the fourth device.

[0016] According to a fifth aspect of an embodiment of the present disclosure, there is provided a first device, including:

[0017] a processing module configured to determine a first signal, where the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device;

[0018] The processing module is further configured to perform channel estimation based on the first signal and determine a channel state information CSI matrix;

[0019] The processing module is further configured to input the CSI matrix into the first artificial intelligence AI model, obtain the number of the third device output by the first AI model

[0020] The processing module is further configured to: and at least one second AI model, determining The perception results obtained by the third devices are respectively perceived.

[0021] According to a sixth aspect of an embodiment of the present disclosure, a second device is provided, including:

[0022] The transceiver module is configured to send a perception signal to the third device or the fourth device.

[0023] According to a seventh aspect of an embodiment of the present disclosure, a third device is provided, including:

[0024] The transceiver module is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the third device.

[0025] According to an eighth aspect of the embodiments of the present disclosure, a fourth device is provided, including:

[0026] The transceiver module is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the fourth device.

[0027] According to a ninth aspect of an embodiment of the present disclosure, a first device is provided, including:

[0028] one or more processors;

[0029] The processor is used to execute the device perception method of any one of the first aspects.

[0030] According to a tenth aspect of an embodiment of the present disclosure, a second device is provided, including:

[0031] one or more processors;

[0032] The processor is used to execute the device perception method of any one of the second aspects.

[0033] According to an eleventh aspect of the embodiments of the present disclosure, a third device is provided, including:

[0034] one or more processors;

[0035] The processor is used to execute the device perception method of any one of the third aspects.

[0036] According to a twelfth aspect of the embodiments of the present disclosure, a fourth device is provided, including:

[0037] one or more processors;

[0038] Among them, the processor is used to execute the device perception method of any one of the fourth aspects.

[0039] According to a thirteenth aspect of an embodiment of the present disclosure, there is provided a communication system, including:

[0040] A first device, the first device being configured to implement the device perception method according to any one of the first aspects;

[0041] a second device, the second device being configured to implement the device perception method according to any one of the second aspects;

[0042] a third device, the third device being configured to implement the device perception method according to any one of the third aspects;

[0043] A fourth device, the fourth device is configured to implement the device perception method of any one of the fourth aspects.

[0044] According to the fourteenth aspect of an embodiment of the present disclosure, a storage medium is provided, which stores instructions. When the instructions are executed on a communication device, the communication device executes a device perception method as described in any one of the first, second, third or fourth aspects.

[0045] According to the fifteenth aspect of an embodiment of the present disclosure, a computer program product is provided, comprising a computer program, which, when executed by a processor, is used to implement the device perception method of any one of the first aspect, the second aspect, the third aspect or the fourth aspect.

[0046] In the embodiment of the present disclosure, the first device may perform channel estimation based on the received first signal to determine the CSI matrix. Further, the number of the third device may be inferred by the first AI model. And inferred by the second AI model The perception results corresponding to each third device are respectively obtained, which improves the accuracy and reliability of device perception and the usability of ISAC technology.

[0047] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0049] FIG1A is an exemplary schematic diagram of the architecture of a communication system provided according to an embodiment of the present disclosure.

[0050] FIG1B is another exemplary schematic diagram of the architecture of a communication system provided according to an embodiment of the present disclosure.

[0051] FIG1C is a schematic diagram of an exemplary scenario of a perception mode provided according to an embodiment of the present disclosure.

[0052] FIG1D is a schematic diagram of an exemplary scenario of signal types in a synaesthesia channel model provided according to an embodiment of the present disclosure.

[0053] FIG1E is an exemplary schematic diagram of determining a perception result based on a perception algorithm according to an embodiment of the present disclosure.

[0054] FIG2A is an exemplary interaction diagram of a device perception method provided according to an embodiment of the present disclosure.

[0055] FIG2B is an exemplary interaction diagram of a device perception method provided according to an embodiment of the present disclosure.

[0056] FIG2C is a flowchart illustrating an exemplary model training method according to an embodiment of the present disclosure.

[0057] FIG2D is a flowchart illustrating an exemplary model training method according to an embodiment of the present disclosure.

[0058] FIG2E is a flowchart illustrating an exemplary model training method according to an embodiment of the present disclosure.

[0059] FIG3A is a schematic diagram of an exemplary flow chart of a device perception method provided according to an embodiment of the present disclosure.

[0060] FIG3B is a schematic diagram of an exemplary flow chart of a device perception method provided according to an embodiment of the present disclosure.

[0061] FIG3C is a schematic diagram of an exemplary flow chart of a device perception method provided according to an embodiment of the present disclosure.

[0062] FIG3D is a schematic diagram of an exemplary flow chart of a device perception method provided according to an embodiment of the present disclosure.

[0063] FIG4A is a schematic diagram of an exemplary flow chart of an AI-based target perception method provided according to an embodiment of the present disclosure.

[0064] FIG4B is a flowchart illustrating an exemplary method for AI-based target perception according to an embodiment of the present disclosure.

[0065] FIG5A is a schematic diagram of an exemplary interaction of a first device provided according to an embodiment of the present disclosure.

[0066] FIG5B is a schematic diagram of an exemplary interaction of a second device according to an embodiment of the present disclosure.

[0067] FIG5C is a schematic diagram of an exemplary interaction of a third device according to an embodiment of the present disclosure.

[0068] FIG5D is a schematic diagram of an exemplary interaction of a fourth device according to an embodiment of the present disclosure.

[0069] FIG6A is a schematic diagram of an exemplary interaction of a communication device according to an embodiment of the present disclosure.

[0070] FIG6B is an exemplary interaction diagram of a chip provided according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0071] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0072] The embodiments of the present disclosure provide a device perception method, apparatus, and storage medium.

[0073] In a first aspect, an embodiment of the present disclosure provides a device perception method, which is performed by a first device and includes:

[0074] Determining a first signal, where the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device;

[0075] Perform channel estimation based on the first signal to determine a channel state information CSI matrix;

[0076] Input the CSI matrix into the first artificial intelligence AI model and obtain the number of third devices output by the first AI model

[0077] Based on the CSI matrix, and at least one second AI model, determining The perception results obtained by the third devices are respectively perceived.

[0078] In the above embodiment, the first device can infer the number of third devices and the perception results corresponding to the third devices through the AI ​​model, reduce the impact of clutter, noise, etc. on the perception results during device perception, improve the accuracy and reliability of device perception, and reduce the computational complexity of determining the perception results, thereby improving the usability of ISAC technology.

[0079] In conjunction with some embodiments of the first aspect, in some embodiments, the method further includes:

[0080] Obtaining a first sample CSI matrix and a true number of third devices corresponding to each first sample CSI matrix;

[0081] Inputting the first sample CSI matrix into the first initial AI model, and obtaining an estimated number of third devices output by the first initial AI model;

[0082] Determine a first loss function based on the difference between the estimated number and the actual number;

[0083] Based on the first loss function, the first initial AI model is trained until the first stopping condition is met, thereby obtaining a first AI model.

[0084] In the above embodiment, the above method can be used to train the first AI model, which is simple to implement and has high usability.

[0085] In conjunction with some embodiments of the first aspect, in some embodiments, the first stop condition includes at least one of the following:

[0086] Reach the first training cycle number;

[0087] The first loss function falls within the first tolerance range;

[0088] The accuracy of the first initial AI model reaches a first value.

[0089] In the above embodiment, training can be stopped when the above first stopping condition is met, thereby obtaining a first AI model, thereby improving the reliability of the first AI model training.

[0090] In combination with some embodiments of the first aspect, in some embodiments, the number of second AI models is 1.

[0091] In the above embodiment, the number of the second AI model can be 1, which avoids occupying the memory of the first device and has high availability.

[0092] In combination with some embodiments of the first aspect, in some embodiments, based on the CSI matrix, and the second AI model to determine The perception results obtained by each third device include:

[0093] Input the CSI matrix into the second AI model to obtain a first vector output by the second AI model, where the dimension of the first vector is N×M, where N is the maximum number of third devices that the first device can sense, and M is the number of parameter types included in each sensed result.

[0094] Put the first vector in front The value of the position is determined as The perception results corresponding to the third devices are respectively obtained.

[0095] In the above embodiment, the first device can obtain based on the second AI model The perception results corresponding to each third device are respectively obtained, which improves the accuracy and reliability of device perception and the usability of ISAC technology.

[0096] In conjunction with some embodiments of the first aspect, in some embodiments, the method further includes:

[0097] Obtaining a second sample CSI matrix and a first true perception result corresponding to each second sample CSI matrix;

[0098] Convert the first real perception result into a first truth value vector, where the dimension of the first truth value vector is N×M;

[0099] Input the second sample CSI matrix into the second initial AI model to obtain a first estimated vector output by the second initial AI model, where the dimension of the first estimated vector is N×M;

[0100] Determining a second loss function based on a difference between the first estimated vector and the first true value vector;

[0101] Based on the second loss function, the second initial AI model is trained until the second stopping condition is met, thereby obtaining a second AI model.

[0102] In the above embodiment, the first device can be trained in the above manner to obtain the second AI model, which is simple to implement and has high usability.

[0103] In combination with some embodiments of the first aspect, in some embodiments, the number of second AI models is N, where N is the maximum number of third devices that the first device can perceive; wherein the number of third devices corresponding to the nth second AI model is n; wherein 1≤n≤N.

[0104] In the above embodiment, the number of the second AI model can be N, thereby reducing the determination The complexity of the perception results corresponding to the third devices is high, and the availability is high.

[0105] In combination with some embodiments of the first aspect, in some embodiments, based on the CSI matrix, the first number and N second AI models, determining The perception results obtained by each third device include:

[0106] In N second AI models, the number of corresponding third devices is determined to be The second AI model;

[0107] The number of third devices corresponding to the CSI matrix input is The second AI model is used to obtain the second vector output by the second AI model. The dimension of the second vector is Where M is the number of parameter types included in each perception result;

[0108] Based on the second vector, determine The perception results corresponding to the third devices are respectively obtained.

[0109] In the above embodiment, the number of corresponding third devices can be directly based on The second AI model determines The perception results corresponding to each third device are respectively obtained, which improves the accuracy and reliability of device perception and the usability of ISAC technology.

[0110] In conjunction with some embodiments of the first aspect, in some embodiments, the method further includes:

[0111] Obtain n third sample CSI matrices of corresponding third devices and a second real perception result corresponding to each third sample CSI matrix;

[0112] Convert the second real perception result into a second truth value vector, where the dimension of the second truth value vector is n×M;

[0113] Input the third sample CSI matrix into the second initial AI model to obtain a second estimated vector output by the second initial AI model, where the dimension of the second estimated vector is n×M;

[0114] Determining a third loss function based on a difference between the second estimated vector and the second true value vector;

[0115] Based on the third loss function, the second initial AI model is trained until the second stopping condition is met, and the training is stopped to obtain a second AI model corresponding to the number n of third devices.

[0116] In the above embodiment, N second AI models can be trained using the above method, which is simple to implement and has high availability.

[0117] In conjunction with some embodiments of the first aspect, in some embodiments, the second stop condition includes at least one of the following:

[0118] Reach the second training cycle number;

[0119] The second loss function is reduced to the second fault tolerance range;

[0120] The accuracy of the second initial AI model reaches a second value.

[0121] In the above embodiment, training can be stopped when the above second stopping condition is met, thereby obtaining one or N second AI models, thereby improving the reliability of the second AI model training.

[0122] In conjunction with some embodiments of the first aspect, in some embodiments, each perception result includes at least one of the following:

[0123] a distance value between the third device and the first device;

[0124] The speed value of the third device;

[0125] The horizontal angle value and / or the zenith angle value between the third device and the first device.

[0126] In the above embodiment, each perception result may include but is not limited to at least one of the above items, thereby improving the reliability and availability of device perception.

[0127] In conjunction with some embodiments of the first aspect, in some embodiments, the third device is any one of the following devices:

[0128] vehicle;

[0129] User equipment;

[0130] Unmanned vehicles;

[0131] Internet of Things (IoT) devices;

[0132] Passive Ambient IoT devices.

[0133] In the above embodiment, the third device may be any of the above devices, which improves the reliability and availability of the perception of the third device.

[0134] In conjunction with some embodiments of the first aspect, in some embodiments, the first device, the second device, and the third device are the same device; or

[0135] The first device, the second device, and the third device are different devices.

[0136] In the above embodiments, the device perception solution provided by the present disclosure can be applied to scenarios of self-transmission and self-reception or transmission and reception at different stations, thereby expanding the application scenarios of device perception and having high availability.

[0137] In a second aspect, an embodiment of the present disclosure provides a device perception method, which is performed by a second device and includes:

[0138] A perception signal is sent to the third device or the fourth device.

[0139] In the above embodiment, the second device may send a perception signal so that the first device can perceive the third device, thereby improving the accuracy and reliability of device perception and improving the availability of the ISAC technology.

[0140] In conjunction with some embodiments of the second aspect, in some embodiments, the third device is any one of the following devices:

[0141] vehicle;

[0142] User equipment;

[0143] Unmanned vehicles;

[0144] Internet of Things (IoT) devices;

[0145] Passive Ambient IoT devices.

[0146] In combination with some embodiments of the second aspect, in some embodiments, the second device and the third device are the same device or different devices.

[0147] In a third aspect, an embodiment of the present disclosure provides a device perception method, which is performed by a third device and includes:

[0148] The sensing signal is reflected to the first device, where the sensing signal is sent by the second device to the third device.

[0149] In the above embodiment, the third device can reflect the perception signal sent by the second device so that the first device obtains the first signal and perceives the third device based on the first signal, thereby improving the accuracy and reliability of device perception and improving the availability of ISAC technology.

[0150] In conjunction with some embodiments of the third aspect, in some embodiments, the third device is any one of the following devices:

[0151] vehicle;

[0152] User equipment;

[0153] Unmanned vehicles;

[0154] Internet of Things (IoT) devices;

[0155] Passive Ambient IoT devices.

[0156] In combination with some embodiments of the third aspect, in some embodiments, the first device, the second device, and the third device are different devices.

[0157] In a fourth aspect, an embodiment of the present disclosure provides a device perception method, which is performed by a fourth device and includes:

[0158] The sensing signal is reflected to the first device, where the sensing signal is sent by the second device to the fourth device.

[0159] In the above embodiment, the fourth device can reflect the perception signal sent by the second device so that the first device obtains the first signal and perceives the third device based on the first signal, thereby improving the accuracy and reliability of device perception and improving the availability of ISAC technology.

[0160] In combination with some embodiments of the fourth aspect, in some embodiments, the first device and the second device are the same device.

[0161] In a fifth aspect, an embodiment of the present disclosure provides a first device, including:

[0162] a processing module configured to determine a first signal, where the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device;

[0163] The processing module is further configured to perform channel estimation based on the first signal and determine a channel state information CSI matrix;

[0164] The processing module is further configured to input the CSI matrix into the first artificial intelligence AI model, obtain the number of the third device output by the first AI model

[0165] The processing module is further configured to: and at least one second AI model, determining The perception results obtained by the third devices are respectively perceived.

[0166] In a sixth aspect, an embodiment of the present disclosure provides a second device, including:

[0167] The transceiver module is configured to send a perception signal to the third device or the fourth device.

[0168] In a seventh aspect, an embodiment of the present disclosure provides a third device, including:

[0169] The transceiver module is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the third device.

[0170] In an eighth aspect, an embodiment of the present disclosure provides a fourth device, including:

[0171] The transceiver module is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the fourth device.

[0172] In a ninth aspect, an embodiment of the present disclosure provides a first device, including:

[0173] one or more processors;

[0174] The processor is used to execute the device perception method of any one of the first aspects.

[0175] In a tenth aspect, an embodiment of the present disclosure provides a second device, including:

[0176] one or more processors;

[0177] The processor is used to execute the device perception method of any one of the second aspects.

[0178] In an eleventh aspect, an embodiment of the present disclosure provides a third device, including:

[0179] one or more processors;

[0180] The processor is used to execute the device perception method of any one of the third aspects.

[0181] In a twelfth aspect, an embodiment of the present disclosure provides a fourth device, including:

[0182] one or more processors;

[0183] Among them, the processor is used to execute the device perception method of any one of the fourth aspects.

[0184] In a thirteenth aspect, an embodiment of the present disclosure provides a communication system, including:

[0185] A first device, the first device being configured to implement the device perception method according to any one of the first aspects;

[0186] a second device, the second device being configured to implement the device perception method according to any one of the second aspects;

[0187] a third device, the third device being configured to implement the device perception method according to any one of the third aspects;

[0188] A fourth device, the fourth device is configured to implement the device perception method of any one of the fourth aspects.

[0189] In the fourteenth aspect, an embodiment of the present disclosure proposes a storage medium storing instructions, which, when executed on a communication device, enables the communication device to execute a device perception method such as any one of the first, second, third or fourth aspects.

[0190] In a fifteenth aspect, an embodiment of the present disclosure proposes a computer program product, comprising a computer program, which, when executed by a processor, is used to implement the device perception method of any one of the first aspect, the second aspect, the third aspect or the fourth aspect.

[0191] It is understandable that the first device, the second device, the third device, the communication system, the storage medium, and the computer program are all used to execute the method proposed in the embodiment of the present disclosure. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects of the corresponding method and will not be repeated here.

[0192] The present disclosure provides a device perception method, device, and storage medium. In some embodiments, the terms device perception method, perception processing method, target perception method, and communication method are interchangeable; the terms device perception device, perception processing device, target perception device, and communication device are interchangeable; and the terms device perception system, perception processing system, target perception system, and communication system are interchangeable.

[0193] The embodiments of the present disclosure are not exhaustive and are merely illustrative of some embodiments, and are not intended to be a specific limitation on the scope of protection of the present disclosure. In the absence of contradiction, each step in a certain embodiment can be implemented as an independent embodiment, and the steps can be arbitrarily combined. For example, a solution after removing some steps in a certain embodiment can also be implemented as an independent embodiment, and the order of the steps in a certain embodiment can be arbitrarily exchanged. In addition, the optional implementation methods in a certain embodiment can be arbitrarily combined; in addition, the embodiments can be arbitrarily combined. For example, some or all steps of different embodiments can be arbitrarily combined, and a certain embodiment can be arbitrarily combined with the optional implementation methods of other embodiments.

[0194] In each embodiment of the present disclosure, unless otherwise specified or provided for by logic, the terms and / or descriptions between the embodiments are consistent and can be referenced by each other. The technical features in different embodiments can be combined to form a new embodiment based on their inherent logical relationships.

[0195] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure.

[0196] In the embodiments of the present disclosure, unless otherwise specified, elements expressed in the singular, such as "a", "an", "the", "above", "the", "the", etc., may mean "one and only one", or "one or more", "at least one", etc. For example, when articles such as "a", "an", "the" in English are used in translation, the noun following the article may be understood as a singular expression or a plural expression.

[0197] In the embodiments of the present disclosure, “plurality” refers to two or more.

[0198] In some embodiments, the terms "at least one," "one or more," "a plurality of," "multiple," etc. may be used interchangeably.

[0199] In some embodiments, descriptions such as "at least one of A and B," "A and / or B," "A in one case, B in another case," or "in response to one case A, in response to another case B" may include the following technical solutions depending on the situation: in some embodiments, A (A is executed independently of B); in some embodiments, B (B is executed independently of A); in some embodiments, execution is selected from A and B (A and B are selectively executed); and in some embodiments, A and B (both A and B are executed). The above is also applicable when there are more branches such as A, B, and C.

[0200] In some embodiments, "A or B" and other descriptions may include the following technical solutions depending on the situation: in some embodiments, A (A is executed independently of B); in some embodiments, B (B is executed independently of A); in some embodiments, execution is selected from A and B (A and B are selectively executed). The above is also applicable when there are more branches such as A, B, C, etc.

[0201] The prefixes such as "first" and "second" in the embodiments of the present disclosure are only used to distinguish different description objects and do not constitute any restriction on the position, order, priority, quantity or content of the description objects. For the statement of the description object, please refer to the description in the context of the claims or embodiments, and no unnecessary restriction should be constituted due to the use of prefixes. For example, if the description object is a "field", the ordinal number before the "field" in the "first field" and the "second field" does not limit the position or order between the "fields". "First" and "second" do not limit whether the "fields" they modify are in the same message, nor do they limit the order of the "first field" and the "second field". For another example, if the description object is a "level", the ordinal number before the "level" in the "first level" and the "second level" does not limit the priority between the "levels". For another example, the number of description objects is not limited by the ordinal number and can be one or more. Taking "first device" as an example, the number of "devices" can be one or more. In addition, the objects modified by different prefixes can be the same or different. For example, if the description object is "device", then the "first device" and the "second device" can be the same device or different devices, and their types can be the same or different; for another example, if the description object is "information", then the "first information" and the "second information" can be the same information or different information, and their contents can be the same or different.

[0202] In some embodiments, “including A,” “comprising A,” “used to indicate A,” and “carrying A” can be interpreted as directly carrying A or indirectly indicating A.

[0203] In some embodiments, devices and equipment can be interpreted as physical or virtual, and their names are not limited to the names recorded in the embodiments. In some cases, they can also be understood as "equipment", "device", "circuit", "network element", "node", "function", "unit", "section", "system", "network", "entity", "subject", etc.

[0204] In some embodiments, obtaining data, information, etc. may comply with the laws and regulations of the country where the data is obtained.

[0205] In some embodiments, data, information, etc. may be obtained with the user's consent.

[0206] In addition, each element, each row, or each column in the table of the embodiment of the present disclosure can be implemented as an independent embodiment, and the combination of any elements, any rows, and any columns can also be implemented as an independent embodiment.

[0207] FIG1A and FIG1B are schematic diagrams showing the architecture of a communication system according to embodiments of the present disclosure.

[0208] As shown in Figure 1A, a communication system 100 may include a first device 101, a second device 102, a third device 103, and a fourth device 104. The first device 101, the second device 102, and the third device 103 are the same device.

[0209] As shown in Figure 1B, a communication system 100' may include a first device 101, a second device 102, and a third device 103. The first device 101, the second device 102, and the third device 103 are different devices.

[0210] In some embodiments, the first device 101 may be a device for determining a perception result corresponding to each third device 103, which may include but is not limited to any of the following:

[0211] Network equipment, such as access network equipment;

[0212] User Equipment (UE);

[0213] vehicle.

[0214] In some embodiments, the second device 102 may be a device for sending a perception signal, which may include but is not limited to any of the following:

[0215] Network equipment, such as access network equipment;

[0216] UE;

[0217] vehicle.

[0218] In some examples, there may be one or more third devices 103. For example, the third device 103 may be a device that needs to be sensed, and the third device 103 may also be referred to as a "sensing target."

[0219] For example, the third device 103 may include but is not limited to any of the following devices:

[0220] vehicle;

[0221] User equipment UE;

[0222] Unmanned vehicles;

[0223] Internet of Things (IoT) devices;

[0224] Passive Ambient Internet of Things (Ambient IoT) devices.

[0225] Among them, unmanned driving equipment may include but is not limited to drones, unmanned vehicles, etc.

[0226] Among them, IoT devices may include but are not limited to smart wearable devices, such as smart watches, smart bracelets, etc., smart home devices, such as smart light bulbs, smart air conditioners, smart rice cookers, etc., smart electricity meters, smart water meters, etc.

[0227] Ambient IoT devices include, but are not limited to, devices that send data and / or signals after being triggered by other devices, such as terminals or network devices. They can be equipped with Radio Frequency Identification (RFID) tags, allowing other devices, such as terminals or network devices, to act as readers for operations such as tag inventory and data reporting.

[0228] In some embodiments, the fourth device 104 can be used to reflect the perception signal sent by the second device 102 in a self-transmitting and self-receiving scenario, which may include but is not limited to any of the following: network equipment, such as access network equipment; UE; vehicle.

[0229] In some embodiments, in FIG1A , the first device 101 , the second device 102 , and the third device 103 are the same device, and in this case, they can realize self-perception through self-transmission and self-reception.

[0230] In some instances, in FIG. 1B , the first device 101 , the second device 102 , and the third device 103 may be different devices. In this case, the third device 103 may be perceived through different-station transmission and reception.

[0231] In some embodiments, the above-mentioned user equipment UE includes, for example, a mobile phone, a wearable device, an Internet of Things device, a car with communication function, a smart car, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, a wireless terminal device in self-driving, a wireless terminal device in remote medical surgery, a wireless terminal device in a smart grid, a wireless terminal device in transportation safety, a wireless terminal device in a smart city, and at least one of a wireless terminal device in a smart home, but is not limited thereto.

[0232] In some embodiments, the above-mentioned access network device is, for example, a node or device that accesses the terminal to the wireless network. The access network device may include an evolved NodeB (eNB), a next generation evolved NodeB (ng-eNB), a next generation NodeB (gNB), a node B (NB), a home node B (HNB), a home evolved nodeB (HeNB), a wireless backhaul device, a radio network controller (RNC), a base station controller (BSC), a base transceiver station (BTS), a base band unit (BBU), a mobile switching center, a base station in a 6G communication system, an open base station (Open RAN), a cloud base station (Cloud RAN), a base station in other communication systems, and at least one of an access node in a Wi-Fi system, but is not limited thereto.

[0233] In some embodiments, the access network device can be composed of a centralized unit (CU) and a distributed unit (DU), where the CU can also be called a control unit. The CU-DU structure can be used to split the protocol layer of the access network device, with the functions of some protocol layers centrally controlled by the CU, and the functions of the remaining part or all of the protocol layers distributed in the DU, which is centrally controlled by the CU, but is not limited to this.

[0234] In some embodiments, the network device may also be a core network device. The core network device may be a device including one or more network elements, or may be multiple devices or a group of devices. The network element may be virtual or physical. The core network may include, for example, at least one of an Evolved Packet Core (EPC), a 5G Core Network (5GCN), and a Next Generation Core (NGC).

[0235] In some embodiments, the technical solution of the present disclosure can be applied to the Open RAN architecture. In this case, the interfaces between or within the access network devices involved in the embodiments of the present disclosure can be transformed into internal interfaces of the Open RAN, and the processes and information interactions between these internal interfaces can be implemented through software or programs.

[0236] It can be understood that the communication system described in the embodiment of the present disclosure is for the purpose of more clearly illustrating the technical solution of the embodiment of the present disclosure, and does not constitute a limitation on the technical solution proposed in the embodiment of the present disclosure. Ordinary technicians in this field can know that with the evolution of the system architecture and the emergence of new business scenarios, the technical solution proposed in the embodiment of the present disclosure is also applicable to similar technical problems.

[0237] The following embodiments of the present disclosure may be applied to the communication system shown in FIG1A or FIG1B , or part of the subject, but are not limited thereto. The subjects shown in FIG1A or FIG1B are examples. The communication system may include all or part of the subjects in FIG1A , or may include other subjects other than those in FIG1A or FIG1B . The number and form of each subject are arbitrary. Each subject may be physical or virtual. The connection relationship between the subjects is an example. The subjects may be connected or disconnected. The connection may be in any manner, directly or indirectly, and wired or wireless.

[0238] The embodiments of the present disclosure can be applied to Long Term Evolution (LTE), LTE-Advanced (LTE-A), LTE-Beyond (LTE-B), SUPER 3G, IMT-Advanced, 4th generation mobile communication system (4G), 5th generation mobile communication system (5G), 5G new radio (NR), future radio access (FRA), new radio access technology (RAT), new radio (NR), new radio access (NX), future generation radio access (FX), Global System for Mobile communications (GSM (registered trademark)), CDMA2000, Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi (registered trademark)), IEEE 802.16 (WiMAX (registered trademark)), IEEE 802.20, Ultra-WideBand (UWB), Bluetooth (registered trademark), Public Land Mobile Network (PLMN) networks, systems utilizing other communication methods, and next-generation systems based on these. Furthermore, a combination of multiple systems (for example, a combination of LTE or LTE-A with 5G) may also be used.

[0239] In the disclosed embodiment, the ISAC may include but is not limited to the following six sensing modes, as shown in Figure 1C. Mono-static means self-transmission and self-reception, i.e., the transmitting end and the receiving end are the same; bi-static means inter-station transmission and reception, i.e., the transmitting end and the receiving end are different.

[0240] Sensing Mode 1: Self-transmission and self-reception. Both the transmitter and receiver are user equipment (UE), such as a person. The transmitter sends a sensing signal to another UE (car) and receives the reflected echo from the other UE (car), thereby sensing the UE's own (person's) distance, speed, angle, and other information.

[0241] Sensing mode 2: autonomous transmission and reception. Both the transmitter and receiver are user equipment (UE), such as a person. The transmitter sends a sensing signal to an access network device, such as a gNB, and receives the reflected echo from the gNB to sense the user's distance, speed, angle, and other information.

[0242] Perception mode 3: inter-site transmission and reception. The transmitter is the gNB and the receiver is the UE (car). For example, the gNB sends a perception signal, and the UE (car) receives the signal forwarded by the user (person). The UE (car) calculates the channel matrix based on the received signal to perceive information such as the distance, speed, and angle of the user (person) in the environment.

[0243] Perception mode 4: inter-site transmission and reception. The transmitter is the UE (car) and the receiver is the gNB. For example, the UE (car) sends a perception signal, and the gNB receives the signal forwarded by the user (person). The gNB calculates the channel matrix based on the received signal to perceive information such as the distance, speed, and angle of the target in the environment, such as the user (person).

[0244] Sensing mode 5, with inter-site transmission and reception, is used by gNB#1 as the transmitter and gNB#2 as the receiver. For example, gNB#1 transmits a sensing signal, and gNB#2 receives the signal forwarded by the user (human). gNB#2 calculates the channel matrix based on the received signal to sense the distance, speed, angle, and other information of the target in the environment, such as the user (human).

[0245] Perception mode 6: inter-station transmission and reception. The transmitter is UE#1 (car #1) and the receiver is UE#2 (car #2). For example, UE#1 (car #1) sends a perception signal, and UE#2 (car #2) receives the signal forwarded by the user (person). UE#2 (car #2) calculates the channel matrix based on the received signal to perceive information such as the distance, speed, and angle of targets in the environment, such as the user (person).

[0246] It is understood that the aforementioned sensing target, namely the third device 103, may include drones, people in indoor and outdoor scenarios, cars on highways in outdoor scenarios, automated guided vehicles in factories in indoor scenarios, and dangerous targets on roadways or railways. Synaesthesia integration primarily involves the receiver calculating information such as the distance, speed, and angle of the sensing target in the environment based on received signals. The sensing target can be the receiver itself or another object.

[0247] In some embodiments, taking self-transmission and self-reception as an example, first device 101 and second device 102 are the same device, both assumed to be gNBs, and third device 103 is assumed to be a vehicle. For example, as shown in Figure 1C, in this scenario, there are a vehicle, a gNB, and an obstruction, which may be another UE or debris. Therefore, four types of signals can be considered in the synaesthesia channel model, namely:

[0248] Perceived Line Of Sight (LOS) path; Perceived Non-Line Of Sight (NLOS) path; Clutter LOS path; Clutter NLOS path.

[0249] In addition to the above four types of signals, the influence of channel noise also needs to be considered.

[0250] For example, as shown in FIG1D , the four types of signals are described as follows:

[0251] Perception LOS path: The path between the sensing signal reflector and the sensing target is the perception LOS path, and the path between the sensing target and the sensing signal receiver is also the perception LOS path (for example, the black solid line path in Figure 1D).

[0252] Perception NLOS path: The path between the sensing signal transmitter and the sensing target that needs to pass through debris is the perception NLOS path. The path between the sensing target and the sensing signal receiver that needs to pass through debris is also the perception NLOS path (for example, the blue solid line path in Figure 1D).

[0253] Clutter LOS path: The path between the sensing signal transmitter and the scattering clusters in the environment (such as debris) is the clutter LOS path. The path between the scattering clusters in the environment and the sensing signal receiver is also the clutter LOS path (such as the red dashed path in Figure 1D).

[0254] Clutter NLOS path: The path between the sensing signal transmitter and the scattering clusters (such as debris) in the environment that still needs to pass through the debris is the sensing NLOS path. The path between the scattering clusters (such as debris) in the environment and the sensing signal receiver that still needs to pass through the debris is also a sensing NLOS path (for example, the green dashed path in Figure 1D).

[0255] In the embodiment of the present disclosure, the main information for sensing the target is the channel state information (CSI) matrix H of the receiving end. The CSI matrix H is the superposition of the sensing target channel, the clutter channel and the noise, and can reflect the channel environment in which the sensing signal is located and the state information of the sensing target in the environment. The dimension of the CSI matrix H is related to the specific parameter settings in the actual synaesthesia scenario, and the information of the data in its different dimensions can reflect different types of state information of the sensing target. In the embodiment of the present disclosure, one possible situation is given:

[0256] The CSI matrix H usually includes three dimensions: subcarrier, Orthogonal Frequency Division Multiplexing (OFDM) symbol, and receiving antenna port.

[0257] Specifically, the channel matrix H represents the channel state information CSI, which can be a complex matrix of M×S×P, where M is the number of OFDM symbols, S is the number of subcarriers, and P is the number of receive antenna ports. The element H at the position of the mth OFDM symbol, the sth subcarrier, and the pth receive antenna port in the channel matrix H is m,s,p , which can be a complex number Z m,s,p , as shown in Formula 1, which represents the impact on the amplitude and phase of the signal during its propagation from the transmitting antenna to the receiving antenna:

[0258] H m,s,p =Z m,s,p =a m,s,p +i×b m,s,p Formula 1

[0259] Among them, a m,s,p 、b m,s,p Used to represent the complex number Z m,s,p The value of the real part and the value of the imaginary part. m,s,p 、b m,s,p Can be a real number.

[0260] According to the received signal model, the phase shifts in the three dimensions of the channel matrix do not affect each other, and different types of perception results of the target can be obtained based on the phase changes in the three dimensions. That is, the distance of the target can be estimated based on the phase difference caused by the time delay between subcarriers, the speed of the target can be estimated based on the phase difference caused by the Doppler effect between OFDM symbols, and the angle of the target can be estimated based on the phase change between the received signals of multiple antenna ports.

[0261] In some embodiments, the perception results of the three parameters of the distance, speed, and angle of the perception target can be obtained based on the CSI matrix through the perception algorithm, as shown in Figure 1E.

[0262] When determining perception results through perception algorithms, the applicable scenarios will be limited. For example, it will be affected by factors such as clutter and noise in the scene. In addition, the calculation of perception results determined by perception algorithms has high negative clutter and its accuracy needs to be improved.

[0263] Therefore, the present disclosure provides the following device perception method and apparatus, and storage medium.

[0264] FIG2A is an interactive diagram of a device perception method according to an embodiment of the present disclosure. As shown in FIG2A , the present disclosure embodiment relates to a device perception method, wherein the method is used in the scenario shown in FIG1A , i.e., a self-transmitting and self-receiving scenario, and the method includes:

[0265] In step S2101 , the second device 102 (ie, the first device 101 ) sends a perception signal to the fourth device 104 .

[0266] In some embodiments, the second device 102 is a device that sends a sensing signal.

[0267] In some embodiments, the first device 101 is a device that determines a perception result.

[0268] In some embodiments, the fourth device 104 is a device for reflecting a sensing signal in a self-transmitting and self-receiving scenario.

[0269] In some embodiments, the third device 103 refers to a device that needs to be sensed, and may also be referred to as a "sensing target."

[0270] In the self-transmitting and self-receiving scenario, the first device 101 , the second device 102 , and the third device 103 are the same device.

[0271] In some embodiments, the perception signal is a signal that enables the first device 101 to perceive the third device 103. Exemplarily, the perception signal may include but is not limited to a Channel State Information-Reference Signal (CSI-RS).

[0272] In some embodiments, the second device 102, the first device 101, and the third device 103 are the same device, i.e., they use self-transmission and self-reception to achieve self-awareness. The sensing signal sent by the first device 101 needs to be reflected by the fourth device 104 before being received by the first device 101, so that the first device 101 can sense itself.

[0273] Illustratively, in the above-mentioned sensing mode 1 and sensing mode 2, the first device 101 itself sends a sensing signal and receives a reflected echo of the sensing signal passing through the fourth device 104 .

[0274] Exemplarily, in the above-mentioned perception mode 1 and perception mode 2, the second device 102, the first device 101, and the third device 103 are the same user (person).

[0275] For example, in sensing mode 1, the fourth device 104 may be a vehicle.

[0276] Exemplarily, in perception mode 2, the fourth device 104 may be an access network device, such as a gNB.

[0277] For example, the fourth device 104 may also be a UE or other devices, which is not detailed in the present disclosure.

[0278] In some embodiments, the specific descriptions of the first device 101, the second device 102, the third device 103, and the fourth device 104 have been introduced in the embodiment of Figure 1A and will not be repeated here.

[0279] In step S2102 , the fourth device 104 reflects a sensing signal to the first device 101 .

[0280] In some embodiments, in perception mode 1 and perception mode 2, the first device 101, the second device 102, and the third device 103 are the same device, the first device 101 sends a perception signal to the fourth device 104, and the fourth device 104 reflects the perception signal to the first device 101. At this time, the signal received by the first device 101 is the first signal.

[0281] In step S2103 , the first device 101 performs channel estimation based on the first signal and determines a CSI matrix H.

[0282] In some embodiments, the first device 101 performs channel estimation based on the received first signal to determine the CSI matrix H. Assume that M is the number of OFDM symbols, S is the number of subcarriers, and P is the number of receiving antenna ports. The element H at the position of the mth OFDM symbol, the sth subcarrier, and the pth receiving antenna port in the CSI channel matrix H is m,s,p , which can be a complex number Z m,s,p , for example, expressed by Formula 1: H m,s,p＝ Z m,s,p =a m,s,p +i×b m,s,p Formula 1

[0283] Among them, a m,s,p 、b m,s,p Used to represent the complex number Z m,s,p The value of the real part and the value of the imaginary part. m,s,p 、b m,s,p Step S2104: The first device 101 inputs the CSI matrix H into the first AI model and obtains the number of the third device 103 output by the first AI model.

[0284] In some embodiments, the name of the first artificial intelligence (AI) model is not limited and can be interchangeable with an AI multi-target quantity estimation model, a target quantity estimation model, etc.

[0285] In some embodiments, the first AI model may be an AI model for inferring the number of third devices 103 .

[0286] In some embodiments, the input value of the first AI model is the CSI matrix H, and the output value is the number of third devices 103

[0287] In some embodiments, Can be a positive integer.

[0288] For example, N is the maximum number of third devices 103 that the first device 101 can perceive.

[0289] For example, when N is 3, the number of third devices 103 output by the first AI model is Can be 1, 2 or 3.

[0290] In some embodiments, the training process of the first AI model will be introduced in subsequent embodiments and will not be introduced here.

[0291] Step S2105: The first device 101 calculates the CSI matrix H based on the CSI matrix H. and at least one second AI model, determining The third devices 103 respectively perform perception to obtain the perception results.

[0292] In some embodiments, the second AI model may be an AI model used to infer a perception result corresponding to the third device.

[0293] In some embodiments, the perception result corresponding to each third device 103 may include but is not limited to at least one of the following:

[0294] a distance value between the third device 103 and the reference device;

[0295] a speed value of the third device 103;

[0296] The horizontal angle value and / or the zenith angle value between the third device 103 and the reference device.

[0297] The reference device may be the fourth device 104 .

[0298] The distance value may be expressed as d, which may be used to indicate a distance value of the third device 103 relative to the reference device (ie, the fourth device 104 ).

[0299] The speed value of the third device 103 may be expressed as v, which may be used to indicate the running speed of the third device 103 relative to the stationary object.

[0300] The horizontal angle value and / or zenith angle value between the third device 103 and the reference device can be expressed as θ. The horizontal angle value refers to the angle value obtained by projecting the lines connecting the third device 103 and the reference device (i.e., the fourth device 104) onto the horizontal plane from a reference point, such as the center of the earth or other specified location. The zenith angle value refers to the angle value of the line connecting the third device 103 and the reference device (i.e., the fourth device 104) relative to the ground normal.

[0301] In some embodiments, the number of second AI models is 1, and the second AI model outputs a vector having a dimension of N×M, where N is the maximum number of third devices 103 that the first device 101 can sense, and M is the number of parameter types included in each sensed result.

[0302] In an embodiment of the present disclosure, the first device 101 may input the CSI matrix into the second AI model and obtain a first vector output by the second AI model, where the dimension of the first vector is N×M.

[0303] For example, the maximum number of third devices 103 that the first device 101 can sense is N, and each sensed result includes a distance value d i , speed value v i , angle value θ i For three types of parameters, M is 3 and the dimension of the first vector is N×3.

[0304] Furthermore, the first device 101 may The value of the position is determined as The perception results corresponding to the third devices 103 are respectively obtained.

[0305] For example, the number of the third device 103 output by the first AI model is 2, M is 3, and the first vector is The first device 101 may determine that the perception result corresponding to the third device 103 - 1 is {d1, v1, θ1}, and the perception result corresponding to the third device 103 - 2 is {d2, v2, θ2}.

[0306] In some embodiments, the training process of the second AI model will be introduced in subsequent embodiments and will not be repeated here.

[0307] In some embodiments, the number of second AI models is N, where N is the maximum number of third devices 103 that the first device 101 can sense. The number of third devices corresponding to the nth second AI model is n, where 1 ≤ n ≤ N. It is understood that the dimension of the vector output by the nth second AI model is n×M.

[0308] In the embodiment of the present disclosure, the first device 101 has determined the actual number of the third device 103 based on the first AI model. It is possible to determine the N second AI models A corresponding second AI model is generated, and the CSI matrix is ​​input into the second AI model to obtain a second vector. The dimension of the second vector is The first device 101 may directly determine the perception result corresponding to each third device 103 based on the second vector.

[0309] For example, N is 3, M is also 3, the number of third devices 103 corresponding to the second AI model #1 is 1, and the output vector dimension is 1×3; the number of third devices 103 corresponding to the second AI model #2 is 2, and the output vector dimension is 2×3; the number of third devices 103 corresponding to the second AI model #3 is 3, and the output vector dimension is 3×3.

[0310] In the embodiment of the present disclosure, the first device 101 has determined the actual number of the third device 103 based on the first AI model. It is possible to determine the N second AI models Corresponding to a second AI model, assuming If it is 2, then the second AI model #2 is determined among the three second AI models mentioned above.

[0311] The first device 101 can input the CSI matrix into the determined second AI model and obtain a second vector output by the second AI model. The dimension of the second vector is For example, 2×3, assuming the second vector is The first device 101 may determine that the perception result corresponding to the third device 103 - 1 is {d1, v1, θ1}, and the perception result corresponding to the third device 103 - 2 is {d2, v2, θ2}.

[0312] In some embodiments, the training process of N second AI models will be introduced in subsequent embodiments and will not be repeated here.

[0313] The above description is merely an exemplary description, and the present disclosure does not limit the solution in which the first device 101 obtains the perception result corresponding to each third device 103 based on the CSI matrix H and uses the AI ​​model to infer.

[0314] In some embodiments, the names of information, etc. are not limited to the names described in the embodiments, and terms such as "information", "message", "signal", "signaling", "report", "configuration", "indication", "instruction", "command", "channel", "parameter", "domain", "field", "symbol", "symbol", "codeword", "codebook", "codeword", "codepoint", "bit", "data", "program", and "chip" can be used interchangeably.

[0315] In some embodiments, terms such as "send", "reflect", "report", "send", "transmit", "bidirectional transmission", "send and / or receive" can be used interchangeably.

[0316] In some embodiments, "obtain", "get", "get", "receive", "transmit", "bidirectional transmission", "send and / or receive" can be interchangeable, and can be interpreted as receiving from other entities, obtaining from protocols, obtaining from higher layers, obtaining by self-processing, autonomous implementation, etc.

[0317] In some embodiments, terms such as "certain", "preseted", "preset", "setting", "indicated", "a certain", "any", "first", and "designated" can be interchangeable. "Specific A", "preset A", "preset A", "setting A", "indicated A", "a certain A", "any A", and "first A" can be interpreted as A pre-specified in a protocol, etc., or as A obtained through setting, configuration, or indication, etc., or as specific A, a certain A, any A, or first A, etc., but not limited to this.

[0318] In some embodiments, the device perception method involved in the embodiments of the present disclosure may include at least one of steps S2101 to S2105. For example, step S2101 can be implemented as an independent embodiment, step S2102 can be implemented as an independent embodiment, steps S2101+S2102 can be implemented as an independent embodiment, step S2103 can be implemented as an independent embodiment, step S2104 can be implemented as an independent embodiment, step S2105 can be implemented as an independent embodiment, steps S2104+S2105 can be implemented as an independent embodiment, steps S2103+S2104+S2105 can be implemented as an independent embodiment, and steps S2101 to S2105 can be implemented as independent embodiments, but are not limited thereto.

[0319] In some embodiments, step S2101 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when other execution entities send a perception signal, step S2101 may not be executed.

[0320] In some embodiments, step S2102 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, if the first device 101 receives a sensing signal reflected by another execution entity, such as the third device 103, and thus obtains the first signal, step S2102 may not be performed.

[0321] In some embodiments, step S2103 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when the first device 101 determines the CSI matrix using other methods, step S2103 may not be performed.

[0322] In some embodiments, step S2104 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when the first device 101 does not need to determine the number of the third devices 103, step S2104 may not be performed.

[0323] In some embodiments, step S2105 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when the first device 101 determines the perception result corresponding to the third device 103 based on a non-AI perception algorithm, step S2105 may not be performed.

[0324] In some embodiments, steps S2101 to S2105 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0325] In some embodiments, the execution order of steps S2101 to S2105 is not limited.

[0326] In the above embodiment, the first device can send a sensing signal to the fourth device, the fourth device reflects it, and the first device performs channel estimation based on the received first signal to determine the CSI matrix. Furthermore, the number of the third device can be inferred by the first AI model. And inferred by the second AI model In the scenario of self-transmission and self-reception, the accuracy and reliability of device perception are improved, and the usability of ISAC technology is improved.

[0327] Figure 2B is an interactive diagram of a device perception method according to an embodiment of the present disclosure. As shown in Figure 2B, the embodiment of the present disclosure relates to a device perception method, wherein the method is used in the scenario shown in Figure 1B, that is, in a different station transmission and reception scenario, and the above method includes:

[0328] Step S2201: The second device 102 sends a perception signal to the third device 103.

[0329] In some embodiments, the second device 102 is a device that sends a sensing signal.

[0330] In some embodiments, the first device 101 is a device that determines a perception result.

[0331] In some embodiments, the third device 103 refers to a device that needs to be sensed, and may also be referred to as a "sensing target."

[0332] In the inter-station transmission and reception scenario, the first device 101 , the second device 102 , and the third device 103 are different devices.

[0333] In some embodiments, the perception signal is a signal that enables the first device 101 to perceive the third device 103. Exemplarily, the perception signal may include but is not limited to a CSI-RS.

[0334] In some embodiments, in an inter-station transmission and reception scenario, the second device 102 sends a perception signal to the third device 103 , and the third device 103 reflects the perception signal to the first device 101 , so that the first device 101 perceives the third device 103 .

[0335] In some embodiments, in the above-mentioned sensing modes 3 to 6, the second device 102 sends a sensing signal to the third device 103.

[0336] In one example, in perception mode 3, the second device 102 may be an access network device, such as a gNB, the third device 103 may be a user (person), and the first device 101 may be a vehicle.

[0337] In one example, in perception mode 4, the second device 102 may be a vehicle, the third device 103 may be a user (person), and the first device 101 may be an access network device, such as a gNB.

[0338] In one example, in perception mode 5, the second device 102 may be an access network device, such as gNB#1, the third device 103 may be a user (person), and the first device 101 may be an access network device, such as gNB#2.

[0339] In one example, in perception mode 6, the second device 102 may be vehicle #1, the third device 103 may be a user (person), and the first device 101 may be vehicle #2.

[0340] In some embodiments, the specific descriptions of the first device 101, the second device 102, and the third device 103 have been introduced in the embodiment of Figure 1B and will not be repeated here.

[0341] In step S2202 , the third device 103 reflects a sensing signal to the first device 101 .

[0342] In some embodiments, in perception mode 3 to perception mode 6, the second device 102 sends a perception signal to the third device 103, and the third device 103 reflects the perception signal to the first device 101. At this time, the signal received by the first device 101 is the first signal.

[0343] In step S2203 , the first device 101 performs channel estimation based on the first signal and determines a CSI matrix H.

[0344] In some embodiments, the implementation of step S2203 is similar to the aforementioned step S2103 and will not be repeated here.

[0345] Step S2204: The first device 101 inputs the CSI matrix H into the first AI model and obtains the number of the third device 103 output by the first AI model.

[0346] In some embodiments, the implementation of step S2204 is similar to the aforementioned step S2104 and will not be repeated here.

[0347] Step S2205: The first device 101 calculates the CSI matrix H based on the CSI matrix H. and at least one second AI model, determining The third devices 103 respectively perform perception to obtain the perception results.

[0348] In some embodiments, the second AI model may be an AI model used to infer a perception result corresponding to the third device.

[0349] In some embodiments, the perception result corresponding to each third device 103 may include but is not limited to at least one of the following:

[0350] a distance value between the third device 103 and the reference device;

[0351] a speed value of the third device 103;

[0352] The horizontal angle value and / or the zenith angle value between the third device 103 and the reference device.

[0353] The reference device may be the first device 101 .

[0354] The distance value may be expressed as d, which may be used to indicate a distance value of the third device 103 relative to the reference device (ie, the first device 101 ).

[0355] The speed value of the third device 103 may be expressed as v, which may be used to indicate the running speed of the third device 103 relative to the stationary object.

[0356] The horizontal angle value and / or zenith angle value between the third device 103 and the reference device can be expressed as θ, where the horizontal angle value refers to the angle value obtained by projecting the lines connecting the reference point, such as the center of the earth or other specified location, to the third device 103 and the reference device (i.e., the first device 101) onto the horizontal plane. The zenith angle value refers to the angle value of the line connecting the third device 103 and the reference device (i.e., the first device 101) relative to the ground normal.

[0357] In some embodiments, when the number of the second AI models is 1 or N, the implementation of step S2205 is similar to the aforementioned step S2105 and will not be repeated here.

[0358] In some embodiments, the device perception method involved in the embodiments of the present disclosure may include at least one of steps S2201 to S2205. For example, step S2201 can be implemented as an independent embodiment, step S2202 can be implemented as an independent embodiment, steps S2201+S2202 can be implemented as an independent embodiment, step S2203 can be implemented as an independent embodiment, step S2204 can be implemented as an independent embodiment, step S2205 can be implemented as an independent embodiment, steps S2204+S2205 can be implemented as an independent embodiment, steps S2203+S2204+S2205 can be implemented as an independent embodiment, and steps S2201 to S2205 can be implemented as independent embodiments, but are not limited thereto.

[0359] In some embodiments, step S2201 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when other execution entities send a perception signal, step S2201 may not be executed.

[0360] In some embodiments, step S2202 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, if the first device 101 receives a sensing signal reflected by another execution entity, such as the fourth device 104, and thus obtains the first signal, step S2202 may not be performed.

[0361] In some embodiments, step S2203 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when the first device 101 determines the CSI matrix using other methods, step S2203 may not be performed.

[0362] In some embodiments, step S2204 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when the first device 101 does not need to determine the number of the third devices 103, step S2204 may not be performed.

[0363] In some embodiments, step S2205 is optional, and one or more of these steps may be omitted or replaced in different embodiments. For example, when the first device 101 determines the perception result corresponding to the third device 103 based on a non-AI perception algorithm, step S2205 may not be performed.

[0364] In some embodiments, steps S2201 to S2205 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0365] In some embodiments, the execution order of steps S2201 to S2205 is not limited.

[0366] In the above embodiment, the second device can send a sensing signal to the third device, the third device reflects it, and the first device performs channel estimation based on the received first signal to determine the CSI matrix. Furthermore, the number of the third device can be inferred by the first AI model. And inferred by the second AI model The sensor results corresponding to each third device are respectively obtained. In the scenario of receiving and sending data at different stations, the accuracy and reliability of device perception are improved, and the usability of ISAC technology is improved.

[0367] FIG2C is a flow chart of a model training method according to an embodiment of the present disclosure. As shown in FIG2C , the embodiment of the present disclosure relates to a model training method for training a first AI model. The method can be executed by the first device 101 and specifically includes:

[0368] Step S2301 : Acquire first sample CSI matrices and the actual number of third devices 103 corresponding to each first sample CSI matrix.

[0369] In some embodiments, the first device 101 can obtain a first training set, which can be used to train the first initial AI model, and may include one or more first sample CSI matrices and the actual number of third devices 103 corresponding to each first sample CSI matrix.

[0370] In some embodiments, the first device 101 may obtain a first validation set, which may be used to verify the accuracy of the first initial AI model. The first validation set may include one or more fourth sample CSI matrices and the actual number of third devices 103 corresponding to each fourth sample CSI matrix.

[0371] In one example, the first training set and the first validation set may be the same set, or the first training set may be a subset of the first validation set, or the first validation set may be a subset of the first training set, or the first training set and the first validation set may not have the same sample CSI matrix, or may have partially the same sample CSI matrix, which is not limited in the present disclosure.

[0372] Step S2302: Input the first sample CSI matrix into the first initial AI model to obtain the estimated number of third devices output by the first initial AI model.

[0373] In some embodiments, the first initial AI model may use a residual network (RESNET) or a visual geometry group (VGG) network as its backbone network, and may include, but is not limited to, at least one of the following network layers: an input layer; a convolutional layer; a pooling layer; an activation function layer; a connection layer; and an output layer. This disclosure does not limit the architecture of the first initial AI model.

[0374] In some embodiments, the first device 101 may input the first sample CSI matrix into the first initial AI model, thereby obtaining the estimated number of the third device 103 output by the first initial AI model.

[0375] Step S2303: Determine a first loss function based on the difference between the estimated number and the actual number.

[0376] In some embodiments, the first loss function may be determined based on a difference between the estimated number and the actual number.

[0377] Step S2304: Based on the first loss function, the first initial AI model is trained until the first stopping condition is met, thereby obtaining a first AI model.

[0378] In some embodiments, the first device 101 reduces the first loss function based on the first loss function using, for example, a stochastic gradient descent (SGD) algorithm.

[0379] In some embodiments, the first stop condition may include but is not limited to at least one of the following:

[0380] Reach the first training cycle number;

[0381] On the first training set and / or the first validation set, the first loss function falls within a first error tolerance range;

[0382] The accuracy of the first initial AI model reaches a first value on the first training set and / or the first validation set.

[0383] In some embodiments, the first number of training cycles may be determined by a protocol or configured by a network device, which is not limited in this disclosure. To reduce the first loss function, the first device 101 increases the number of training cycles by 1 each time it adjusts the network layer parameters of the first initial AI model. The first device 101 may stop training when the number of training cycles for the first initial AI model reaches the first number of training cycles, at which point the first AI model may be obtained.

[0384] In some embodiments, the first fault tolerance range may be determined by a protocol or a network device configuration, which is not limited in this disclosure. The first device 101 may determine that when the first loss function on the first training set or the first validation set falls within the first fault tolerance range, training is stopped to obtain the first AI model.

[0385] In one example, the first verification set may be a set pre-acquired by the first device 101 for verifying the performance of the AI ​​model, which may include one or more fourth sample CSI matrices and the actual number of third devices 103 corresponding to each fourth sample CSI matrix.

[0386] In some embodiments, the first value may be determined by a protocol agreement or a network device configuration, which is not limited in this disclosure. The first device 101 may determine the accuracy of the first initial AI model on a first training set or a first validation set, and when the first value is reached, training is stopped to obtain the first AI model. The description of the first training set and the first validation set has been introduced in the previous embodiment and will not be repeated here.

[0387] In some embodiments, steps S2301 to S2304 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0388] In some embodiments, the execution order of steps S2301 to S2304 is not limited.

[0389] In the above embodiment, the first device can be trained in the above manner to obtain a first AI model, and the first AI model can be used to infer the number of third devices, which is simple to implement and has high usability.

[0390] FIG2D is a flow chart of a model training method according to an embodiment of the present disclosure. As shown in FIG2D , the present disclosure relates to a model training method, wherein the method is used to train a second AI model, and the number of second AI models is 1. The method can be executed by the first device 101 and specifically includes:

[0391] Step S2401: Acquire a second sample CSI matrix and a first real perception result corresponding to each second sample CSI matrix.

[0392] In some embodiments, the first device 101 can obtain a second training set, which can be used to train a second initial AI model, which may include one or more second sample CSI matrices and a first real perception result of the third device 103 corresponding to each second sample CSI matrix.

[0393] In some embodiments, the first device 101 can obtain a second verification set, which can be used to verify the accuracy of the second initial AI model, and may include one or more fifth sample CSI matrices, and a first real perception result of the third device 103 corresponding to each fifth sample CSI matrix.

[0394] In some embodiments, the second training set and the second validation set may be the same set, or the second training set may be a subset of the second validation set, or the second validation set may be a subset of the second training set, or the second training set and the second validation set may not have the same sample CSI matrix, or may have partially the same sample CSI matrix, which is not limited in the present disclosure.

[0395] Step S2402: convert the first real perception result into a first truth value vector.

[0396] In some embodiments, the dimension of the first truth vector is N×M.

[0397] Since the number of second AI models is 1 and the output vector dimension is N×M, if the first real perception result is directly converted into a vector, the dimension of the resulting vector is n'×M, where n' is the actual number of third devices 103 corresponding to the second sample CSI matrix. n' is less than or equal to N. Therefore, when converting the first real perception result into a first true value vector, if n' is less than N, it is necessary to pad the last (N-n')×M positions of the vector converted from the first real perception result with zeros to obtain the first true value vector.

[0398] For example, if the vector dimension corresponding to the first true result is 2×3 and N is 3, then the last 1×3 positions in the vector need to be padded with zeros to obtain a first true value vector with a dimension of 3×3.

[0399] Step S2403: Input the second sample CSI matrix into the second initial AI model to obtain a first estimated vector output by the second initial AI model.

[0400] In some embodiments, the second initial AI model can use a residual network (RESNET) or a visual geometry group (VGG) network as its backbone network, and can include, but is not limited to, at least one of the following network layers: input layer; convolution layer; pooling layer; activation function layer; connection layer; and output layer. This disclosure does not limit the architecture of the second initial AI model.

[0401] In some embodiments, the dimension of the first prediction vector is N×M.

[0402] Step S2404: Determine a second loss function based on the difference between the first estimated vector and the first true value vector.

[0403] In some embodiments, the dimensions of the first estimation vector and the first true value vector are both N×M.

[0404] In some embodiments, the first device 101 may determine a second loss function based on a difference between the first estimated vector and the first true value vector.

[0405] Step S2405: Based on the second loss function, the second initial AI model is trained until the second stopping condition is met, thereby obtaining a second AI model.

[0406] In some embodiments, the second stop condition may include but is not limited to at least one of the following:

[0407] Reach the second training cycle number;

[0408] On the second training set and / or the second validation set, the second loss function falls within a second error tolerance range;

[0409] The accuracy of the second initial AI model reaches a second value on the second training set and / or the second validation set.

[0410] In some embodiments, the implementation of step S2405 is similar to that of step S2404 and will not be repeated here.

[0411] In some embodiments, steps S2401 to S2405 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0412] In some embodiments, the execution order of steps S2401 to S2405 is not limited.

[0413] In the above embodiment, the first device can be trained in the above manner to obtain a second AI model, and the second AI model can be used to infer the perception results corresponding to the third device, which is simple to implement and has high usability.

[0414] FIG2E is a flow chart of a model training method according to an embodiment of the present disclosure. As shown in FIG2E , the embodiment of the present disclosure relates to a model training method, wherein the method is used to train a second AI model, and the number of second AI models is N. The method can be executed by the first device 101 and specifically includes:

[0415] Step S2501: Obtain n third sample CSI matrices of corresponding third devices and a second real perception result corresponding to each third sample CSI matrix.

[0416] In some embodiments, 1≤n≤N.

[0417] In some embodiments, assuming that N is 3, the first device 101 can obtain a third training set when the number of third devices is 1, 2, and 3, respectively, which may include a third sample CSI matrix and a second real perception result corresponding to each third sample CSI matrix.

[0418] In some embodiments, assuming that N is 3, the first device 101 may obtain a third verification set when the number of third devices is 1, 2, and 3, respectively, which may include a sixth sample CSI matrix and a second real perception result corresponding to each sixth sample CSI matrix.

[0419] In some embodiments, when the number of third devices is the same, the third training set and the third verification set can be the same set, or the third training set is a subset of the third verification set, or the third verification set is a subset of the third training set, or the third training set and the third verification set do not have the same sample CSI matrix, or have partially the same sample CSI matrix, and this is not limited in the present disclosure.

[0420] Step S2502: convert the second real perception result into a second true value vector.

[0421] In some embodiments, the first device 101 may directly convert the second real perception result into a vector to obtain a second truth value vector, where the dimension of the second truth value vector is n×M.

[0422] For example, assuming that M is 3 and the number of third devices is 1, the dimension of the second truth vector is 1×3. When the number of third devices is 2, the dimension of the second truth vector is 2×3. When the number of third devices is 3, the dimension of the second truth vector is 3×3, and so on. When the number of third devices is N, the dimension of the second truth vector is N×3.

[0423] Step S2503: Input the third sample CSI matrix into the second initial AI model to obtain a second estimated vector output by the second initial AI model.

[0424] In some embodiments, the structure of the second initial AI model has been introduced in the aforementioned embodiments and will not be repeated here.

[0425] In some embodiments, the first device 101 may input the third sample CSI matrix of the corresponding third device, number n, into the corresponding second initial AI model, thereby obtaining a second estimated vector output by the second initial AI model. The dimension of the second estimated vector is n×M.

[0426] Step S2504: Determine a third loss function based on the difference between the second estimated vector and the second true value vector.

[0427] In some embodiments, the dimension of the second true value vector is the same as the dimension of the second estimated vector, both being n×M.

[0428] In some embodiments, the first device 101 may determine a third loss function based on a difference between a second estimated vector having a dimension of n×M and a second true value vector having a dimension of n×M.

[0429] Step S2505: Based on the third loss function, the second initial AI model is trained until the second stopping condition is met, and the training is stopped to obtain a second AI model corresponding to the number of third devices being n.

[0430] In some embodiments, the second stop condition includes at least one of the following:

[0431] Reach the second training cycle number;

[0432] On the third training set and / or the third validation set, the second loss function is reduced to a second error tolerance range;

[0433] The accuracy of the second initial AI model reaches a second value on the third training set and / or the third validation set.

[0434] In some embodiments, the process of the first device 101 training the second AI model is similar to the process of training the first AI model, and will not be repeated here.

[0435] In some embodiments, the first device 101 trains corresponding second AI initial models for different numbers n of third devices 103, thereby obtaining a total of N second AI models.

[0436] In some embodiments, steps S2501 to S2505 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0437] In some embodiments, the execution order of steps S2501 to S2505 is not limited.

[0438] In the above embodiment, the first device can be trained in the above manner to obtain N second AI models, and the second AI model can be used to infer the perception results corresponding to the third device, which is simple to implement and has high usability.

[0439] FIG3A is a flow chart of a device perception method according to an embodiment of the present disclosure. As shown in FIG3A , the present disclosure embodiment relates to a device perception method, which can be performed by a first device 101 and includes:

[0440] Step S3101, obtain a first signal.

[0441] In some embodiments, the first device 101 may receive a perception signal reflected by the third device 103 or the fourth device 104 to obtain the first signal, but is not limited thereto and may also receive a perception signal reflected by other execution entities to obtain the first signal.

[0442] In some embodiments, the first device 101 obtains a first signal determined according to a predefined rule.

[0443] In some embodiments, the first device 101 performs processing to obtain the first signal.

[0444] In some embodiments, step S3101 is omitted, the first device 101 autonomously implements the function indicated by the first signal, or the first device 101 obtains the first signal based on predefined rules or protocol agreements, or the above function is default or default.

[0445] In some embodiments, the optional implementation method of step S3101 can refer to the optional implementation method of step S2102 in Figure 2A and other related parts of the embodiment involved in Figure 2A, or can refer to the optional implementation method of step S2202 in Figure 2B and other related parts of the embodiment involved in Figure 2B, which will not be repeated here.

[0446] Step S3102: Determine the CSI matrix.

[0447] In some embodiments, the optional implementation of step S3102 can refer to the optional implementation of step S2103 in Figure 2A and other related parts of the embodiment involved in Figure 2A, or can refer to the optional implementation of step S2203 in Figure 2B and other related parts of the embodiment involved in Figure 2B, which will not be repeated here.

[0448] Step S3103: Determine the number of third devices 103

[0449] In some embodiments, the optional implementation method of step S3103 can refer to the optional implementation method of step S2104 in Figure 2A and other related parts of the embodiment involved in Figure 2A, or can refer to the optional implementation method of step S2204 in Figure 2B and other related parts of the embodiment involved in Figure 2B, which will not be repeated here.

[0450] Step S3104: Determine the perception result corresponding to the third device 103.

[0451] In some embodiments, the optional implementation method of step S3104 can refer to the optional implementation method of step S2105 in Figure 2A and other related parts of the embodiment involved in Figure 2A, or can refer to the optional implementation method of step S2205 in Figure 2B and other related parts of the embodiment involved in Figure 2B, which will not be repeated here.

[0452] In some embodiments, steps S3101 to S3104 are optional, and one or more of these steps may be omitted or replaced in different embodiments.

[0453] In some embodiments, the execution order of steps S3101 to S3104 is not limited.

[0454] In the above embodiment, the first device can perform channel estimation based on the received first signal to determine the CSI matrix. Furthermore, the number of third devices can be inferred by the first AI model. And inferred by the second AI model The perception results corresponding to each third device are respectively obtained, which improves the accuracy and reliability of device perception and the usability of ISAC technology.

[0455] FIG3B is a flow chart of a device perception method according to an embodiment of the present disclosure. As shown in FIG3B , the present disclosure embodiment relates to a device perception method, which can be performed by the second device 102 and includes:

[0456] Step S3201: Send a perception signal.

[0457] In some embodiments, the second device 102 sends a perception signal to the third device 103 or the fourth device 104 .

[0458] In some embodiments, in a self-transmitting and self-receiving scenario, the first device 101 , the second device 102 , and the third device 103 are the same device.

[0459] In some embodiments, in a different-site transmission and reception scenario, the first device 101 , the second device 102 , and the third device 103 are different devices.

[0460] In some embodiments, the optional implementation of step S3201 can refer to the optional implementation of step S2101 in Figure 2A and other related parts of the embodiment involved in Figure 2A, or can refer to the optional implementation of step S2201 in Figure 2B and other related parts of the embodiment involved in Figure 2B, which will not be repeated here.

[0461] In the above embodiment, the second device can send a perception signal so that the first device can perceive the third device and determine the perception result, thereby improving the accuracy and reliability of device perception and improving the availability of ISAC technology.

[0462] FIG3C is a flow chart of a device perception method according to an embodiment of the present disclosure. As shown in FIG3C , the present disclosure embodiment relates to a device perception method, which can be performed by a third device 103 and includes:

[0463] Step S3301: Reflect the sensing signal.

[0464] In some embodiments, in a different-site transmission and reception scenario, the first device 101 , the second device 102 , and the third device 103 are different devices.

[0465] In some embodiments, the third device 103 may reflect the sensing signal sent by the second device 102 .

[0466] In some embodiments, the third device 103 reflects the sensing signal toward the first device 101 .

[0467] In some embodiments, the first device 101 obtains the first signal based on the sensing signal reflected by the third device 103 .

[0468] In some embodiments, the optional implementation of step S3301 can refer to the optional implementation of step S2202 in Figure 2B and other related parts of the embodiment involved in Figure 2B, which will not be repeated here.

[0469] In the above embodiment, the third device can reflect the sensing signal to the first device so that the first device can obtain the first signal, thereby performing channel estimation based on the received first signal and determining the CSI matrix. Furthermore, the number of the third device can be inferred by the first AI model. And inferred by the second AI model The sensor results corresponding to each third device are respectively obtained. In the scenario of receiving and sending data at different stations, the accuracy and reliability of device perception are improved, and the usability of ISAC technology is improved.

[0470] FIG3D is a flow chart of a device perception method according to an embodiment of the present disclosure. As shown in FIG3D , the present disclosure embodiment relates to a device perception method, which can be performed by the fourth device 104 and includes:

[0471] Step S3401: transmit a sensing signal.

[0472] In some embodiments, in a self-transmitting and self-receiving scenario, the first device 101 , the second device 102 , and the third device 103 are the same device.

[0473] In some embodiments, in a self-transmitting and self-receiving scenario, the fourth device 104 may reflect the sensing signal sent by the second device 102 .

[0474] In some embodiments, the fourth device 104 reflects the sensing signal toward the first device 101 .

[0475] In some embodiments, the first device 101 obtains the first signal based on the sensing signal reflected by the fourth device 104 .

[0476] In some embodiments, the optional implementation of step S3401 can refer to the optional implementation of step S2102 in Figure 2A and other related parts of the embodiment involved in Figure 2A, which will not be repeated here.

[0477] In the above embodiment, the fourth device can reflect the sensing signal to the first device so that the first device can obtain the first signal, thereby performing channel estimation based on the received first signal and determining the CSI matrix. Furthermore, the number of the third device can be inferred by the first AI model. And inferred by the second AI model In the scenario of self-transmission and self-reception, the accuracy and reliability of device perception are improved, and the usability of ISAC technology is enhanced.

[0478] The above process is further illustrated below with examples.

[0479] The AI-based integrated synaesthesia method proposed in this paper primarily detects the distance, speed, and angle of a target. The type of information required for perception may vary depending on the application scenario and service. While distance, speed, and angle information can meet the needs of most scenarios, the method proposed in this paper can also be applied to situations requiring the perception of other types of information.

[0480] The AI-based multi-target perception method proposed in the present invention includes the training and application methods of two AI models: multi-target quantity estimation and multi-target perception.

[0481] To train an AI perception model, you first need to collect a training dataset, which contains a certain number of data samples. Each data sample consists of two parts: the channel state information matrix H (model input data), the actual number of perceived targets n (perception target number estimation model output data), and the actual distance d, speed v, and angle θ information of each perceived target (multi-target perception model output data).

[0482] (1) Multi-target quantity estimation model (i.e., the aforementioned first AI model)

[0483] The multi-target number estimation model estimates the number of perceived targets n based on the input channel state information matrix H. Therefore, during the model training process, the model output result, i.e., the estimated number of targets, is used as the output. The error between the labels in the training set, that is, the actual number of perceived targets n, is used as the loss function. Model training is performed through methods such as stochastic gradient descent (SGD) to reduce the loss function, so that the model learns the mapping relationship between the input channel matrix and the output perception results. Model training is completed after reaching a specific number of training cycles or the model accuracy meets certain requirements.

[0484] (2) Multi-target perception model (i.e., the aforementioned second AI model, with a number of 1)

[0485] The multi-target perception model obtains the distances of multiple perception targets based on the input channel state information matrix H. speed angle In multi-target perception scenarios, the specific number of perceived targets is unknown, while the training and application of AI models require the input and output data dimensions to be certain. Therefore, this solution proposes an AI multi-target perception model design and training method to address this problem:

[0486] a. Model Structure: The multi-target perception model outputs the distance, velocity, and angle perception results for N targets, where N is a specific, large value representing the maximum number of perceived targets in most cases of synaesthesia integration. For example, in a real-world scenario, there may be one, two, or three perceived targets, so N = 3.

[0487] b. Model training: The label portion of each data sample in the training set is processed and converted into a vector of length N × 3. The distance d, velocity v, and angle θ of n perceived targets are arranged in sequence. Since n ≤ N, the (Nn) × 3 positions at the end of the vector that may exist are set to zero. During model training, the error between the N × 3 vector output by the model and the label vector in the training set is used as the loss function. Model training is performed using methods such as stochastic gradient descent (SGD) to reduce the loss function. Model training is completed after a specific number of training cycles or when the model accuracy meets certain requirements.

[0488] c. Model application (inference): The multi-target perception model obtains a vector of length N×3 based on the input channel state information matrix H. The output of the model is the estimated number of targets obtained according to the number of multi-target estimations. Process the output vector and take the front The value of each position is The perception results of a target.

[0489] The AI-based multi-target perception method proposed in the present invention is mainly implemented through the above two models, and its main application process is shown in Figure 4A. Among them, the specific model structure and implementation of the multi-target quantity estimation model and the multi-target perception model depend on the design and deployment in the actual application process, but the training data processing method, model input and output data type and dimension, model training and application process provided by the present invention are also applicable. The vector output by the AI ​​multi-target perception model in Figure 4A may include multiple parameter values ​​that are approximately 0, which are represented by "~0" in Figure 4A.

[0490] To train an AI perception model, you first need to collect a training dataset, which contains a certain number of data samples. Each data sample consists of two parts: the channel state information matrix H (model input data), the actual number of perceived targets n (perception target number estimation model output data), and the actual distance d, speed v, and angle θ information of each perceived target (multi-target perception model output data).

[0491] (1) Multi-target quantity estimation model (i.e., the aforementioned first AI model)

[0492] The multi-target number estimation model estimates the number of perceived targets n based on the input channel state information matrix H. Therefore, during the model training process, the model output result, i.e., the estimated number of targets, is used as the output. The error between the labels in the training set, that is, the actual number of perceived targets n, is used as the loss function. Model training is performed through methods such as stochastic gradient descent (SGD) to reduce the loss function, so that the model learns the mapping relationship between the input channel matrix and the output perception results. Model training is completed after reaching a specific number of training cycles or the model accuracy meets certain requirements.

[0493] (2) Multi-target perception model (i.e., the aforementioned second AI model, with a number of N)

[0494] The multi-target perception model obtains the distances of multiple perception targets based on the input channel state information matrix H. speed angle For multi-target perception scenarios, the specific number of perceived targets is unknown, while the training and application of AI models require that the input and output data dimensions are certain. Therefore, this solution proposes an AI multi-target perception model design and training method to address this problem, and train multiple AI perception models to be applied to situations with different numbers of perceived targets.

[0495] a. Model Structure: The multi-target perception model outputs the distance, velocity, and angle perception results for 1 to N targets, where N is a specific, large value representing the maximum number of perceived targets in most cases of synaesthesia integration. For example, in a real-world scenario, there may be 1, 2, or 3 perceived targets, so N = 3.

[0496] b. Model training: N AI models are trained using data from a dataset with 1 to N perceived targets. For n targets, the AI ​​model outputs an n×3 vector representing the distance, speed, and angle of the n perceived targets. During model training, the error between the model output vector and the label vector in the training set is used as the loss function. Model training is performed using methods such as stochastic gradient descent (SGD) to reduce the loss function. Model training is completed after a specific number of training cycles is reached or the model accuracy meets certain requirements.

[0497] c. Model application (inference): target quantity estimation value based on the output of the multi-target quantity estimation model Select the AI ​​multi-target perception model with the corresponding number of perception targets. The multi-target perception model is based on the input channel state information matrix H with a length of vector, as The perception results of a target.

[0498] The AI-based multi-target perception method proposed in this invention is primarily implemented through the two models described above, with its primary application process shown in Figure 4B . While the specific model structures and implementations of the multi-target quantity estimation model and the multi-target perception model depend on the design and deployment in the actual application, the training data processing method, model input and output data types and dimensions, and model training and application processes provided in this invention are equally applicable.

[0499] Aiming at the application scenario of integrated communication and perception, the present invention proposes an AI-based multi-target perception method, which uses a neural network model to respectively realize the estimation of the number of targets and the perception of the distance, speed, angle and other information of multiple targets. In the application scenario of integrated multi-target synaesthesia, based on the channel state information matrix of the receiving end and the data set composed of the actual number of perceived targets and the actual distance, speed, angle and other parameter information of each perceived target, the AI ​​perception model is trained and acquired, which can realize the perception of the distance, speed and angle of multiple targets with high accuracy. The AI-based multi-target perception solution proposed in the present invention has a large scope of application, which is conducive to solving the shortcomings of the traditional perception algorithm in the case of multiple targets with poor perception accuracy, thereby promoting the development and application of integrated communication and perception technology.

[0500] The embodiments of the present disclosure also propose an apparatus for implementing any of the above methods. For example, an apparatus is proposed, which includes units or modules for implementing each step performed by each device (such as the first device, the second device, the third device, the fourth device, etc.) in any of the above methods.

[0501] It should be understood that the division of the various units or modules in the above device is merely a division of logical functions. In actual implementation, they may be fully or partially integrated into a physical entity, or they may be physically separated. In addition, the units or modules in the device may be implemented in the form of a processor calling software: for example, the device includes a processor, the processor is connected to a memory, and the memory stores instructions. The processor calls the instructions stored in the memory to implement any of the above methods or implement the functions of the various units or modules of the above device, wherein the processor is, for example, a general-purpose processor, such as a central processing unit (CPU) or a microprocessor, and the memory is a memory within the device or a memory outside the device. Alternatively, the units or modules in the device can be implemented in the form of hardware circuits, and the functions of some or all of the units or modules can be realized by designing the hardware circuits. The above-mentioned hardware circuits can be understood as one or more processors; for example, in one implementation, the above-mentioned hardware circuit is an application-specific integrated circuit (ASIC), which realizes the functions of some or all of the above units or modules by designing the logical relationship of the components in the circuit; for example, in another implementation, the above-mentioned hardware circuit can be realized by a programmable logic device (PLD). Taking a field programmable gate array (FPGA) as an example, it can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by configuring the configuration file, thereby realizing the functions of some or all of the above units or modules. All units or modules of the above devices can be realized in the form of software called by the processor, or in the form of hardware circuits, or in part by the form of software called by the processor, and the rest by hardware circuits.

[0502] In the embodiments of the present disclosure, the processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a microprocessor), or a digital signal processor (DSP). In another implementation, the processor can implement certain functions through the logical relationship of the hardware circuit. The logical relationship of the above-mentioned hardware circuit is fixed or reconfigurable. For example, the processor is a hardware circuit implemented by an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and implementing the hardware circuit configuration can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units or modules. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), etc.

[0503] FIG5A is a schematic diagram of the structure of a first device according to an embodiment of the present disclosure. As shown in FIG5A , the first device 5100 may include: a processing module 5101 .

[0504] In some embodiments, the processing module 5101 is configured to determine a first signal, wherein the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device.

[0505] The processing module 5101 is further configured to perform channel estimation based on the first signal to determine a channel state information CSI matrix; input the CSI matrix into a first artificial intelligence AI model to obtain the number of third devices output by the first AI model; Based on the CSI matrix, the and at least one second AI model, determining The perception results obtained by each of the third devices are perceived respectively.

[0506] Optionally, the above-mentioned processing module 5101 is used to execute at least one of the other steps (such as step S2103, step S2104, step S2105, step S2203, step S2204, step S2205, but not limited to these) performed by the first device 5100 in any of the above methods, which are not repeated here.

[0507] FIG5B is a schematic diagram of the structure of a second device according to an embodiment of the present disclosure. As shown in FIG5B , the second device 5200 may include a transceiver module 5201 .

[0508] In some embodiments, the transceiver module 5201 is configured to send a perception signal to a third device or a fourth device.

[0509] Optionally, the above-mentioned transceiver module 5201 is used to execute at least one of the communication steps such as sending and / or receiving (for example, step S2101, step S2201, but not limited to this) performed by the second device 5200 in any of the above methods, which will not be repeated here.

[0510] FIG5C is a schematic diagram of the structure of a third device proposed in an embodiment of the present disclosure. As shown in FIG5C , the third device 5300 may include: a transceiver module 5301 .

[0511] In some embodiments, the processing module 5301 is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the third device.

[0512] Optionally, the above-mentioned transceiver module 5301 is used to execute at least one of the communication steps such as sending and / or receiving (for example, step S2202, but not limited to this) performed by the third device 5300 in any of the above methods, which will not be repeated here.

[0513] FIG5D is a schematic diagram of the structure of a fourth device proposed in an embodiment of the present disclosure. As shown in FIG5D , the fourth device 5400 may include: a transceiver module 5401 .

[0514] In some embodiments, the processing module 5401 is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the fourth device.

[0515] Optionally, the above-mentioned transceiver module 5401 is used to execute at least one of the communication steps such as sending and / or receiving (for example, step S2102, but not limited to this) performed by the fourth device 5400 in any of the above methods, which will not be repeated here.

[0516] In some embodiments, the transceiver module may include a transmitting module and / or a receiving module, and the transmitting module and the receiving module may be separate or integrated. Optionally, the transceiver module may be interchangeable with the transceiver.

[0517] In some embodiments, the processing module may be a single module or may include multiple submodules. Optionally, the multiple submodules each execute all or part of the steps required by the processing module. Optionally, the processing module and the processor may be interchangeable.

[0518] Figure 6A is a schematic diagram of the structure of a communication device 6100 proposed in an embodiment of the present disclosure. Communication device 6100 can be any device (e.g., a first device, a second device, a third device, a fourth device, etc.), or a chip, a chip system, or a processor that supports each device in implementing any of the above methods. It can also be a chip, a chip system, or a processor that supports a terminal in implementing any of the above methods. Communication device 6100 can be used to implement the methods described in the above method embodiments. For details, please refer to the description of the above method embodiments.

[0519] As shown in Figure 6A, the communication device 6100 includes one or more processors 6101. The processor 6101 can be a general-purpose processor or a dedicated processor, for example, a baseband processor or a central processing unit. The baseband processor can be used to process communication protocols and communication data, and the central processing unit can be used to control communication devices (such as base stations, baseband chips, terminal devices, terminal device chips, DUs or CUs, vehicles, unmanned driving devices, IoT devices, Amibent IoT devices, etc.), execute programs, and process program data. The communication device 6100 is used to perform any of the above methods.

[0520] In some embodiments, the communication device 6100 further includes one or more memories 6102 for storing instructions. Optionally, all or part of the memories 6102 may be located outside the communication device 6100.

[0521] In some embodiments, the communication device 6100 also includes one or more transceivers 6103 . When the communication device 6100 includes one or more transceivers 6103, the transceiver 6103 performs at least one of the communication steps such as sending and / or receiving in the above method (for example, step S2101, step S2102, step S2201, step S2202, but not limited to this), and the processor 6101 performs at least one of the other steps (for example, step S2103, step S2104, step S2105, step S2203, step S2204, step S2205, step S2301, step S2302, step S2303, step S2304, step S2401, step S2402, step S2403, step S2404, step S2405, step S2501, step S2502, step S2503, step S2504, but not limited to this).

[0522] In some embodiments, a transceiver may include a receiver and / or a transmitter. The receiver and transmitter may be separate or integrated. Optionally, the terms transceiver, transceiver unit, transceiver, and transceiver circuit may be used interchangeably; the terms transmitter, transmitting unit, transmitter, and transmitting circuit may be used interchangeably; and the terms receiver, receiving unit, receiver, and receiving circuit may be used interchangeably.

[0523] In some embodiments, the communication device 6100 may include one or more interface circuits 6104. Optionally, the interface circuit 6104 is connected to the memory 6102. The interface circuit 6104 may be configured to receive signals from the memory 6102 or other devices, and may be configured to send signals to the memory 6102 or other devices. For example, the interface circuit 6104 may read instructions stored in the memory 6102 and send the instructions to the processor 6101.

[0524] The communication device 6100 described in the above embodiments may be any device, but the scope of the communication device 6100 described in the present disclosure is not limited thereto, and the structure of the communication device 6100 may not be limited to FIG. 6A. The communication device may be an independent device or may be part of a larger device. For example, the communication device may be: 1) an independent integrated circuit IC, or a chip, or a chip system or subsystem; (2) a collection of one or more ICs, optionally, the above IC collection may also include a storage component for storing data or programs; (3) an ASIC, such as a modem; (4) a module that can be embedded in other devices; (5) a receiver, a terminal device, an intelligent terminal device, a cellular phone, a wireless device, a handheld device, a mobile unit, an in-vehicle device, a network device, a cloud device, an artificial intelligence device, etc.; (6) others, etc.

[0525] 6B is a schematic diagram of the structure of a chip 6200 according to an embodiment of the present disclosure. If the communication device 6200 can be a chip or a chip system, please refer to the schematic diagram of the structure of the chip 6200 shown in FIG6B , but the present disclosure is not limited thereto.

[0526] The chip 6200 includes one or more processors 6201 , and the chip 6200 is configured to execute any of the above methods.

[0527] In some embodiments, the chip 6200 further includes one or more interface circuits 6202. Optionally, the interface circuit 6202 is connected to the memory 6203. The interface circuit 6202 can be used to receive signals from the memory 6203 or other devices, and can be used to send signals to the memory 6203 or other devices. For example, the interface circuit 6202 can read instructions stored in the memory 6203 and send the instructions to the processor 6201.

[0528] In some embodiments, the interface circuit 6202 performs at least one of the communication steps such as sending and / or receiving in the above method (for example, step S2101, step S2102, step S2201, step S2202, but not limited to this), and the processor 6201 performs at least one of the other steps (for example, step S2103, step S2104, step S2105, step S2203, step S2204, step S2205, step S2301, step S2302, step S2303, step S2304, step S2401, step S2402, step S2403, step S2404, step S2405, step S2501, step S2502, step S2503, step S2504, but not limited to this).

[0529] In some embodiments, terms such as interface circuit, interface, transceiver pin, and transceiver may be used interchangeably.

[0530] In some embodiments, the chip 6200 further includes one or more memories 6203 for storing instructions. Alternatively, all or part of the memories 6203 may be located outside the chip 6200.

[0531] The present disclosure also proposes a storage medium having instructions stored thereon. When the instructions are executed on the communication device 6100, the communication device 6100 executes any of the above methods. Optionally, the storage medium is an electronic storage medium. Optionally, the storage medium is a computer-readable storage medium, but is not limited thereto and may also be a storage medium readable by other devices. Optionally, the storage medium may be a non-transitory storage medium, but is not limited thereto and may also be a transient storage medium.

[0532] The present disclosure also provides a program product, which, when executed by the communication device 6100, enables the communication device 6100 to perform any of the above methods. Optionally, the program product is a computer program product.

[0533] The present disclosure also provides a computer program that, when executed on a computer, causes the computer to perform any of the above methods. Other embodiments of the present disclosure will be readily apparent to those skilled in the art after considering the specification and practicing the invention disclosed herein.

[0534] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0535] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A device perception method, characterized in that: The method is performed by a first device and includes: determining a first signal, where the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device; Perform channel estimation based on the first signal to determine a channel state information CSI matrix; Input the CSI matrix into the first artificial intelligence AI model and obtain the number of the third device output by the first AI model Based on the CSI matrix, the and at least one second AI model, determining The perception results obtained by each of the third devices are perceived respectively.

2. The method according to claim 1, characterized in that The method further comprises: Acquire a first sample CSI matrix and a real number of the third devices corresponding to each first sample CSI matrix; Inputting the first sample CSI matrix into a first initial AI model, and obtaining an estimated number of the third device output by the first initial AI model; Determining a first loss function based on the difference between the estimated number and the actual number; Based on the first loss function, the first initial AI model is trained until a first stopping condition is met, thereby obtaining the first AI model.

3. The method according to claim 2, characterized in that The first stop condition includes at least one of the following: Reach the first training cycle number; The first loss function falls within a first fault tolerance range; The accuracy of the first initial AI model reaches a first value.

4. The method according to any one of claims 1 to 3, characterized in that The number of the second AI model is 1.

5. The method according to claim 4, characterized in that Based on the CSI matrix, the And the second AI model determines The perception results obtained by each of the third devices respectively performing perception include: Inputting the CSI matrix into the second AI model to obtain a first vector output by the second AI model, where the dimension of the first vector is N×M; where N is the maximum number of third devices that can be sensed by the first device, and M is the number of parameter types included in each of the sensed results; The first vector is placed before The value of the position is determined as The perception results corresponding to each of the third devices.

6. The method according to claim 5, characterized in that The method further comprises: Obtaining a second sample CSI matrix and a first true perception result corresponding to each of the second sample CSI matrices; Converting the first real perception result into a first truth value vector, where the dimension of the first truth value vector is N×M; Inputting the second sample CSI matrix into a second initial AI model to obtain a first estimated vector output by the second initial AI model, where the dimension of the first estimated vector is N×M; Determining a second loss function based on a difference between the first estimated vector and the first true value vector; Based on the second loss function, the second initial AI model is trained until the second stopping condition is met to obtain the second AI model.

7. The method according to any one of claims 1 to 3, characterized in that The number of the second AI models is N, where N is the maximum number of the third devices that the first device can perceive; wherein the number of the third devices corresponding to the nth second AI model is n; wherein 1≤n≤N.

8. The method according to claim 7, characterized in that Based on the CSI matrix, the first number and N second AI models, determine The perception results obtained by each of the third devices respectively performing perception include: In the N second AI models, the number of corresponding third devices is determined to be The second AI model; The number of the third device corresponding to the CSI matrix input is The second AI model is used to obtain a second vector output by the second AI model, where the dimension of the second vector is Wherein, M is the number of parameter types included in each of the perception results; Based on the second vector, determine The perception results corresponding to each of the third devices.

9. The method according to claim 8, characterized in that The method further comprises: Obtaining n third sample CSI matrices corresponding to the third device and a second real perception result corresponding to each third sample CSI matrix; Converting the second real perception result into a second truth value vector, where the dimension of the second truth value vector is n×M; Inputting the third sample CSI matrix into a second initial AI model to obtain a second estimated vector output by the second initial AI model, where the dimension of the second estimated vector is n×M; Determining a third loss function based on a difference between the second estimated vector and the second true value vector; Based on the third loss function, the second initial AI model is trained until the second stopping condition is met, and the training is stopped to obtain the second AI model corresponding to the number n of the third devices.

10. The method according to claim 6 or 9, characterized in that The second stop condition includes at least one of the following: Reach the second training cycle number; The second loss function is reduced to a second fault tolerance range; The accuracy of the second initial AI model reaches a second value.

11. The method according to any one of claims 1 to 10, characterized in that Each of the perception results includes at least one of the following: a distance value between the third device and a reference device; wherein the reference device is the first device or the fourth device; a speed value of the third device; The horizontal angle value and / or zenith angle value between the third device and a reference device; wherein the reference device is the first device or the fourth device.

12. The method according to any one of claims 1 to 11, characterized in that The third device is any one of the following devices: vehicle; User equipment; Unmanned vehicles; Internet of Things (IoT) devices; Passive Ambient IoT devices.

13. The method according to any one of claims 1 to 12, characterized in that The first device, the second device, and the third device are the same device; or The first device, the second device, and the third device are different devices.

14. A device perception method, characterized in that: The method is performed by a second device and includes: A perception signal is sent to the third device or the fourth device.

15. The method according to claim 14, characterized in that The third device is any one of the following devices: vehicle; User equipment; Unmanned vehicles; Internet of Things (IoT) devices; Passive Ambient IoT devices.

16. The method according to claim 14 or 15, characterized in that The second device and the third device are the same device or different devices.

17. A device perception method, characterized in that: The method is performed by a third device and includes: A perception signal is reflected to the first device, where the perception signal is sent by the second device to the third device.

18. The method according to claim 17, characterized in that The third device is any one of the following devices: vehicle; User equipment; Unmanned vehicles; Internet of Things (IoT) devices; Passive Ambient IoT devices.

19. The method according to claim 17 or 18, characterized in that The first device, the second device, and the third device are different devices.

20. A device perception method, characterized in that: The method is performed by a fourth device, including: A perception signal is reflected to the first device, where the perception signal is sent by the second device to the fourth device.

21. The method according to claim 20, characterized in that The first device and the second device are the same device.

22. A first device, characterized in that: include: a processing module configured to determine a first signal, where the first signal is a signal obtained when receiving a perception signal reflected by a third device, or the first signal is a signal obtained when receiving a perception signal reflected by a fourth device; wherein the perception signal is sent by the second device; The processing module is further configured to perform channel estimation based on the first signal and determine a channel state information CSI matrix; The processing module is further configured to input the CSI matrix into a first artificial intelligence (AI) model, and obtain the number of third devices output by the first AI model. The processing module is further configured to: and at least one second AI model, determining described The third device performs perception respectively and obtains the perception results.

23. A second device, characterized in that: include: The transceiver module is configured to send a perception signal to the third device or the fourth device.

24. A third device, characterized in that: include: The transceiver module is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the third device.

25. A fourth device, characterized in that: include: The transceiver module is configured to reflect a perception signal to the first device, where the perception signal is sent by the second device to the fourth device.

26. A first device, characterized in that: include: one or more processors; The processor is used to execute the device perception method according to any one of claims 1 to 13.

27. A second device, characterized in that: include: one or more processors; The processor is used to execute the device perception method described in any one of claims 14-16.

28. A third device, characterized in that: include: one or more processors; The processor is used to execute the device perception method described in any one of claims 17-19.

29. A fourth device, characterized in that: include: one or more processors; Wherein, the processor is used to execute the device perception method described in claim 20 or 21.

30. A communication system, characterized in that: include: A first device, wherein the first device is configured to implement the device perception method according to any one of claims 1 to 13; a second device, wherein the second device is configured to implement the device perception method according to any one of claims 14 to 16; a third device, wherein the third device is configured to implement the device perception method according to any one of claims 17 to 19; A fourth device, wherein the fourth device is configured to implement the device perception method described in claim 20 or 21.

31. A storage medium storing instructions, characterized in that: When the instruction is executed on the communication device, the communication device is caused to execute the device perception method according to any one of claims 1-13, 14-16, 17-19 or 20-21.

32. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, it is used to implement the device perception method described in any one of claims 1-13, 14-16, 17-19 or 20-21.

Images