Sensing receiver entity and terminal entity, and sensing transmitter entity

Abstract

The present disclosure implements a concrete AI-based sensing operation that enables a sensing receiver entity to perform a high-performance AI-based sensing operation by combining AI technology and ISAC technology to minimize the influence due to the difference in AI computing and / or ISAC capability between sensing entities.

Description

Signal receiving device and terminal device, signal transmitting device

[0001] The present invention relates to ISAC (Integrated Sensing and Communication) technology applied to a mobile communication network.

[0002] This application claims the benefit of Korean Application No. 10-2024-0121567, filed September 6, 2024, the entire contents of which are incorporated herein by reference for all purposes.

[0003] When explaining the positioning technology applied to 5G, it is possible to measure / position the location of a terminal based on the transmission and reception operation of a reference signal for positioning.

[0004] That is, according to existing positioning technology, base station / terminal capability to transmit and receive reference signals (e.g. PRS, SRS, etc.) for positioning purposes is essential.

[0005] Therefore, in the existing positioning technology, location measurement / positioning is possible for terminals such as smartphones and tablets because they have the capability to transmit and receive reference signals (e.g., PRS, SRS, etc.), but location measurement / positioning is not possible for objects such as people, objects, and buildings (hereinafter, Unconnected objects) that do not have the capability to transmit and receive positioning reference signals.

[0006] In this regard, discussions have begun on a technology that can support location measurement / positioning for objects that do not have the capability to transmit and receive positioning reference signals, i.e., unconnected objects that cannot communicate with themselves from the base station's perspective. This technology is ISAC (Integrated Sensing and Communication).

[0007] ISAC technology is a technology that expands from the concept of Positioning, which simply measures location, to include sensing of surrounding objects and environments based on the concept of Sensing, and can support positioning of unconnected objects that do not have the capability of transmitting and receiving positioning reference signals.

[0008] However, the current ISAC technology only defines the concept of the technology, and there is no specific discussion or definition of how to increase the accuracy and sophistication of the sensing operation.

[0009] Meanwhile, learning-based AI models are being applied to technologies that compute / process various types of data, and even on-device AI implemented on the user (terminal) side has emerged.

[0010] Creating AI models and utilizing them for computation or application requires data learning and inference processes. However, the types, amounts, and cycles of data used for learning and inference are not specifically defined at the conceptual level. In particular, in the case of on-device AI, the AI ​​models capable of inference or learning can vary depending on the hardware and software specifications of each terminal. Therefore, AI application methods that consider the performance of different terminals are necessary.

[0011] Accordingly, the present invention proposes a new technology method that can improve and enhance the accuracy of sensing results / data according to the sensing operation of the ISAC technology by combining AI technology and ISAC technology.

[0012] The purpose of the present invention is to provide a new technology solution that can improve and enhance the accuracy of sensing results / data according to the sensing operation of the ISAC technology by combining AI technology and ISAC technology.

[0013] A signal receiving device according to one embodiment of the present invention comprises: a memory including a command; and a processor that performs an AI-based sensing operation by executing the command, inferring a sensing result according to the performance of the sensing operation through an AI model based on an AI parameter obtained through transmission and reception of a preset specific signal, in relation to a sensing operation performed based on a radio wave measurement of the signal serviced in mobile communication.

[0014] Specifically, the specific signal may be at least one reference signal included in an AI Parameter Set set for obtaining the AI ​​Parameter through a setup procedure with a signal transmission device that mutually performs the sensing operation.

[0015] Specifically, the AI ​​model may be a pre-trained model that uses AI parameters obtained through transmission and reception of the specific signal as input, performs the sensing operation based on the radio wave measurement value for the specific signal, and utilizes the derived sensing result as output, thereby inferring the output for the input.

[0016] Specifically, the processor can perform a verification procedure for the sensing result inferred from the AI ​​model by utilizing past sensing result information stored by performing the AI-based sensing operation, or location estimation information derived by performing a separate sensing operation.

[0017] Specifically, when the signal list to be used in the AI ​​model is confirmed when setting the AI ​​Parameter Set, the processor can train the AI ​​model or infer the sensing result by using only the signals in the signal list among the signals in the AI ​​Parameter Set.

[0018] In one embodiment of the present invention, a terminal device includes a memory including a command; and a processor that performs an AI-based sensing operation by executing the command, inferring a sensing result according to the performance of the sensing operation through an AI model based on an AI parameter obtained through transmission and reception of a preset specific signal, in relation to a sensing operation performed based on measurement of a radio wave of a signal serviced in mobile communication.

[0019] In one embodiment of the present invention, a signal transmission device includes a memory including a command; and a processor that, by executing the command, sets an AI Parameter Set for acquiring an AI Parameter in a signal reception device that mutually performs the Sensing operation, which is performed based on a radio wave measurement of the signal serviced in mobile communication, so that the signal reception device can perform an AI-based Sensing operation by acquiring and using the AI ​​Parameter through transmission and reception of a specific signal according to the AI ​​Parameter Set setting.

[0020] In one embodiment of the present invention, an AI-based sensing operation method performed in a signal receiving device includes, in relation to a sensing operation performed based on a radio wave measurement of the signal serviced in mobile communication, a step of setting an AI Parameter Set for obtaining AI Parameters through a setting procedure with a signal transmitting device that mutually performs the sensing operation; a step of obtaining AI Parameters through transmission and reception of a specific signal according to the setting of the AI ​​Parameter Set; and a step of performing an AI-based sensing operation for inferring a sensing result according to the performance of the sensing operation through an AI model based on the obtained AI Parameters.

[0021] According to embodiments of the present invention, a new type of AI-based sensing operation is implemented through a specific technical configuration that can improve and enhance the accuracy of sensing results / data according to the sensing operation of the ISAC technology by combining AI technology and ISAC technology.

[0022] Accordingly, according to the present invention, it is possible to infer and learn an AI model corresponding to the performance of individual devices due to differences in AI Computing and / or ISAC Capability between devices, thereby having scalability that can be quickly applied to various terminals, and can produce the effect of performing high-performance AI-based sensing operations in a signal receiving device (sensing entity).

[0023] Figure 1 is an example diagram illustrating a scenario of ISAC technology.

[0024] FIG. 2 is a drawing explaining the configuration / function of a signal receiving device and a signal transmitting device according to one embodiment of the present invention.

[0025] Figures 3 and 4 are exemplary flowcharts explaining the operation flow of the AI-based sensing operation method proposed in the present invention.

[0026] Hereinafter, various embodiments of the present invention will be described with reference to the attached drawings.

[0027] The present invention relates to ISAC (Integrated Sensing and Communication) technology applied to a mobile communication network.

[0028] When explaining the positioning technology applied to 5G, it is possible to measure / position the location of a terminal based on the transmission and reception operation of a reference signal for positioning.

[0029] That is, according to existing positioning technology, base station / terminal capability to transmit and receive reference signals (e.g. PRS, SRS, etc.) for positioning is essential.

[0030] And the existing positioning technology method can be explained as a method of extracting information such as TDoA (Time Difference of Arrival) through transmission and reception of reference signals for positioning purposes such as tactics, and then estimating the location by utilizing geometric operations such as triangulation, and a fingerprinting method (a method of estimating the location based on previously stored information) using RSRP, RSSI, CQI, etc. transmitted from terminals or anchor nodes.

[0031] Therefore, in the existing positioning technology, location measurement / positioning is possible for terminals such as smartphones and tablets because they have the capability to transmit and receive reference signals (e.g., PRS, SRS, etc.), but location measurement / positioning is not possible for objects such as people, objects, and buildings (hereinafter, Unconnected objects) that do not have the capability to transmit and receive positioning reference signals.

[0032] In this regard, discussions have begun on a technology that can support location measurement / positioning for objects that do not have the capability to transmit and receive positioning reference signals, i.e., unconnected objects that cannot communicate with themselves from the base station's perspective. This technology is ISAC (Integrated Sensing and Communication).

[0033] ISAC technology is a technology that expands from the concept of Positioning, which simply measures location, to include sensing of surrounding objects and environments based on the concept of Sensing, and can support positioning of unconnected objects that do not have the capability of transmitting and receiving positioning reference signals.

[0034] Figure 1 illustrates the basic scenario of ISAC technology.

[0035] As illustrated in Fig. 1, sensing in ISAC technology means detecting the presence of an unconnected object (e.g., a person) that is not connected to a network (e.g., a base station), and further measuring the shape, location, and moving speed of the unconnected object.

[0036] In addition, as can be seen in Fig. 1, the core concept of ISAC technology applied to mobile communication networks is based on wireless sensing technology, and is sensing using radio waves of signals serviced by mobile communication, rather than sensing using various sensors related to existing motion, location, environment, etc.

[0037] In particular, as the available bandwidth increases and high-resolution sensing becomes possible as we move to higher frequency bands, it may be appropriate to apply ISAC technology to next-generation mobile communication networks, such as 5G and even 6G, that use higher frequency bands.

[0038] However, the current ISAC technology only defines the concept of the technology, and there is no specific discussion or definition of how to increase the accuracy and sophistication of the sensing operation.

[0039] Meanwhile, learning-based AI models are being applied to technologies that compute / process various types of data, and even on-device AI implemented on the user (terminal) side has emerged.

[0040] Accordingly, the present invention proposes a new technology method that can improve and enhance the accuracy of sensing results / data according to the sensing operation of the ISAC technology by combining AI technology and ISAC technology.

[0041] More specifically, the present invention proposes an AI-based sensing operation that infers the sensing result / data of an ISAC operation (hereinafter, sensing operation) through AI by combining AI technology and ISAC technology.

[0042] At this time, in the present invention, when utilizing AI technology, it is possible to select information to be utilized as input for AI learning / inference (e.g., some or all of the information utilized in existing location estimation (e.g., TDoA, RSRP, RSSI, CQI, etc.)), and it is also possible to utilize the information by classifying its validity, and therefore, a new type of AI-based sensing operation is proposed.

[0043] Additionally, there may be differences in AI Computing and / or ISAC Capability between base stations and terminals, and in particular, it may be difficult to perform AI learning / inference due to limitations in computational capacity on terminals.

[0044] In the present invention, by implementing a configuration for transmitting AI parameters for AI learning / inference between a base station (or edge server) and a terminal in an ISAC scenario, we propose an AI-based sensing operation that can be performed even on a terminal with limited computational capacity by minimizing the impact due to differences in AI computing and / or ISAC capabilities between devices.

[0045] Hereinafter, with reference to FIG. 2, the configuration / function of a signal receiving device and a signal transmitting device that realize the proposed technology of the present invention will be specifically described.

[0046] Before explanation, the signal receiving device (100) of the present invention illustrated in FIG. 2 refers to a device that operates as a Sensing Entity that measures the radio wave of a signal in an ISAC scenario, and may be a base station (BS) or a terminal (UE).

[0047] In relation to this, more specifically, a base station (BS) and a terminal (UE) can operate as a sensing entity that measures the radio wave of a signal provided in mobile communication.

[0048] In ISAC technology, the sensing mode in which BS and UE operate as sensing entities can be defined as the following 1st to 6th modes, depending on the measurement topology that measures the propagation of a signal (the radio wave reflected by the target).

[0049] Mode 1 (BS->BS Bi-static): A mode in which, depending on the measurement topology, a BS transmits a sensing signal and another BS measures the propagation of the sensing signal to generate sensing data.

[0050] Second mode (BS Mono-static): A mode in which the BS transmits a sensing signal and the BS itself measures the propagation of the sensing signal to generate sensing data.

[0051] Mode 3 (BS->UE Bi-static): A mode in which the BS transmits a sensing signal and the UE measures the propagation of the sensing signal to generate sensing data.

[0052] Mode 4 (UE->BS Bi-static): A mode in which the UE transmits a sensing signal and the BS measures the propagation of the sensing signal to generate sensing data.

[0053] Mode 5 (UE->UE Bi-static): A mode in which a UE transmits a sensing signal and another UE measures the propagation of the sensing signal to generate sensing data.

[0054] Mode 6 (UE Mono-static): A mode in which the UE transmits a sensing signal and the UE itself measures the propagation of the sensing signal to generate sensing data.

[0055] That is, in the present invention, when the UE is a signal receiving device (100), the signal transmitting device (200) may be another UE or BS, and when the BS is a signal receiving device (100), the signal transmitting device (200) may be another BS or UE.

[0056] However, for the convenience of explanation, in the following, it will be assumed that a UE having relatively limited computational capacity, such as AI Computing and / or ISAC Capability, is a signal receiving device (100), and a BS (or Edge server) is a signal transmitting device (200) (e.g., BS->UE Bi-static).

[0057] As illustrated in FIG. 2, a signal receiving device (100, UE) according to an embodiment of the present invention may be configured with a memory (not shown) including a command; a processor (hereinafter, AI Parameter Set setting unit (110), Sensing operation unit (130)) that performs an AI-based Sensing operation by inferring a Sensing result according to the performance of the Sensing operation through an AI model based on AI Parameters obtained through transmission and reception of a preset specific signal in relation to a Sensing operation performed based on measurement of a radio wave of the signal serviced in mobile communication by executing the command.

[0058] Here, the sensing operation means all operations that measure the radio waves of signals (hereinafter, sensing signals) provided by mobile communication and generate sensing results / data.

[0059] In the signal receiving device (100, UE) of the present invention, the sensing operation unit (130) can perform both an operation of providing sensing results / data generated by measuring radio waves of a sensing signal according to a general ISAC sensing operation (hereinafter, general sensing operation), and an operation of providing sensing results / data inferred through an AI model (120) according to the AI-based sensing operation of the present invention, i.e., an AI-based sensing operation.

[0060] To explain specifically, the AI-based sensing operation proposed in the present invention is an operation of inferring sensing results / data according to the performance of the sensing operation through an AI model (120) based on AI parameters acquired through transmission and reception of a preset specific signal.

[0061] At this time, a specific signal means at least one signal included in the AI ​​Parameter Set set for obtaining the AI ​​Parameter through a setup procedure with a signal transmission device (200, BS) that mutually performs the sensing operation.

[0062] Specifically, in the present invention, the signal receiving device (100, UE) / signal transmitting device (200, BS) goes through a setting process of setting parameters for mutual AI learning / inference, and the parameters set at this time are defined as “AI parameters.”

[0063] For example, as AI parameters, information utilized in existing location estimation (e.g., TDoA, RSRP, RSSI, CQI, RI, etc.) can be utilized, and in the present invention, some or all of these pieces of information that can be utilized as AI parameters can be selected and set as an AI parameter set in the signal receiving device (100).

[0064] For example, based on the current 5G, uplink reference signals include DMRS (Demodulation Reference Signal), PTRS (Phase-Tracking Reference Sign), and SRS (Sounding Reference Signal), and downlink reference signals include DMRS, PTRS, PRS (Positioning Reference Signal), CSI-RS (Channel State Information Reference Signal), PSS (Primary Synchronization Signal), and SSS (Secondary Synchronization Signal).

[0065] Here, the process of setting the AI ​​Parameter Set performed in the present invention will be described as an example. A signal transmission device (200, BS) can transmit an AI Parameter Set for mutual AI learning / inference to a signal reception device (100, UE) that mutually performs a sensing operation.

[0066] The AI ​​Parameter Set transmitted at this time may include information (hereinafter, AI Parameter information) that selects / specifies some or all of the information used in existing location estimation (e.g., TDoA, RSRP, RSSI, CQI, RI, etc.) as AI Parameters to be used in AI learning / inference.

[0067] As a more specific example, the AI ​​Parameter Set may further include a list of signals to be utilized when learning / inferring an AI model (120), and for example, a list of signals specifying reference signals (e.g., DMRS, PTRS, PRS, CSI-RS, etc.) to be utilized in obtaining selected / specific AI parameters may be included.

[0068] And, the signal transmission device (200, BS) can transmit this AI Parameter Set to the signal reception device (100, UE) through RRC (Radio Resource Control) or MAC CE (MAC Control Element), DCI transmitted through PDCCH, UCI.

[0069] Furthermore, in the present invention, a control frame that allocates resources for transmitting an AI Parameter Set can be transmitted separately from an existing control frame.

[0070] For example, in the present invention, a Control frame that allocates resources for transmitting an AI Parameter Set can be distinguished by using an Indicator for an AI Parameter Set to be used for AI learning, an Indicator for an AI Parameter Set to be used for AI inference, or by modifying and using an existing Control frame.

[0071] Continuing the explanation again, in a signal receiving device (100, UE) that has received an AI Parameter Set, the AI ​​Parameter Set setting unit (110) can transmit its own Capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability) for information included in the transmitted AI Parameter Set in the form of an Ack.

[0072] Of course, the AI ​​Parameter Set setting unit (110) in the signal receiving device (100, UE) can perform AI learning / inference according to the AI ​​Parameter Set setting thereafter by setting the information included in the transmitted AI Parameter Set if it is acceptable in its Capability.

[0073] Afterwards, when the signal transmission device (200, BS) receives capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability) and setting completion in the form of an Ack from the signal reception device (100, UE), the setting process of the mutual AI Parameter Set can be completed.

[0074] Meanwhile, if the previously transmitted AI Parameter Set is not set in the signal receiving device (100, UE), the signal transmitting device (200, BS) may re-perform the AI ​​Parameter Set setting process to reset the AI ​​Parameter Set with a low amount of computation to suit the signal receiving device (100, UE) based on the Capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability) confirmed in the Ack.

[0075] In this regard, FIG. 2 illustrates a signal transmission device (100, BS) according to one embodiment of the present invention.

[0076] A signal transmission device (100, BS) according to an embodiment of the present invention may be configured with a memory (not shown) including a command; a processor (hereinafter, AI Parameter Set setting unit (210), ISAC control unit (220)) that sets an AI Parameter Set for AI Parameter acquisition in a signal reception device (100, UE) that mutually performs the Sensing operation in relation to the Sensing operation of the ISAC by executing the command, so that the signal reception device (100, UE) can perform an AI-based Sensing operation by acquiring and using the AI ​​Parameter through transmission and reception of a specific signal according to the AI ​​Parameter Set setting.

[0077] Here, in the process of setting the AI ​​Parameter Set, the aforementioned operations of the signal transmission device (100, BS) can be handled by the AI ​​Parameter Set setting unit (210) in the signal transmission device (100, BS).

[0078] As described above, in the present invention, a configuration is implemented in which a specific AI Parameter Set is selected / classified to suit the computational capacity, such as AI Computing and / or ISAC Capability of a signal receiving device (100), for AI learning / inference in an ISAC scenario, and set between signal transmitting and receiving devices (base station or edge server / terminal).

[0079] In particular, considering that each signal receiving device (100, e.g., UE) may have different sensing capabilities (e.g., TDoA, RSRP, RSSI, CQI, etc.) and different sensing purposes, the present invention implements a configuration in which a selection / classification / specific AI Parameter Set is set between signal transmitting and receiving devices (base station or edge server / terminal) in consideration of the sensing capabilities / purpose of each signal receiving device (100, e.g., UE).

[0080] Meanwhile, the AI ​​model (120) mounted on the signal receiving device (100, UE) of the present invention can be learned according to the AI ​​Parameter Set set for each signal receiving device (100, UE).

[0081] Specifically, the AI ​​model (120) can be defined as a pre-trained model that uses AI parameters acquired through transmission and reception of a specific signal preset by the AI ​​Parameter Set as input, performs the sensing operation of ISAC based on the radio wave measurement value for the specific signal, and utilizes the derived sensing result as output, thereby inferring the output for the input.

[0082] To explain a specific embodiment, the signal receiving device (100, UE) of the present invention can perform AI learning / inference according to the AI ​​Parameter Set setting set internally through the aforementioned AI Parameter Set setting process.

[0083] First, the AI ​​learning process performed according to the AI ​​Parameter Set settings is as follows.

[0084] The signal receiving device (100, UE) of the present invention can obtain an AI parameter (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) selected / specified as AI parameter information for transmission / reception of a specific signal set in advance, for example, a reference signal (single or multiple) according to a signal list, according to an AI parameter set setting.

[0085] This can be obtained through the sensing capability (e.g., TDoA, RSRP, RSSI, CQI, etc.) that the signal receiving device (100, UE) already possesses, and the AI ​​parameters (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) obtained in this way are used as inputs during learning of the AI ​​model (120) and can be used as inputs during inference of the AI ​​model (120) later.

[0086] Meanwhile, the signal receiving device (100, UE) of the present invention can generate sensing results / data by measuring the radio wave of the reference signal (single or multiple) used to obtain the AI ​​parameter (e.g., one or multiple of TDoA, RSRP, RSSI, CQI, RI, etc.) and performing a sensing operation based on the radio wave measurement value.

[0087] This can be obtained by performing a general sensing operation provided by the sensing operation unit (130), and the sensing result / data generated in this way can be used as an output during learning of the AI ​​model (120).

[0088] Accordingly, in the signal receiving device (100, UE) of the present invention, the AI ​​model (120) can be trained to infer the sensing result / data of the same reference signal (sensing signal) as an output for input (Input, AI Parameter (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.)) obtained from transmission and reception of reference signals (single or multiple) for a certain period of time (e.g., a set period of time, or until a condition satisfying AI model performance is satisfied).

[0089] As mentioned in the description, in the present invention, a signal list may be included in the AI ​​Parameter Set.

[0090] In this case, the signal receiving device (100, UE) can check the signal list to be used in the AI ​​model (120) when setting the aforementioned AI Parameter Set, and can train the AI ​​model (120) or infer the sensing result by using only the signals in the signal list among the signals in the AI ​​Parameter Set.

[0091] In this case, among the various reference signals transmitted and received between the signal receiving device (100, UE) / signal transmitting device (200, BS), only some of them specified as the signal list are utilized, so the accuracy of the sensing result of the AI ​​model (120) or the inference accuracy of the AI ​​model (120) may decrease, but there is an advantage in that the amount of computation for storing and processing various AI data related to the AI-based sensing operation of the present invention can be reduced.

[0092] In addition, since the signal transmission device (200, ISAC control unit (220)) can know the signal list of the reference signal used by the signal reception device (100) for AI-based sensing operation, data management for each signal list can be useful by attaching a label to the corresponding data set.

[0093] In addition, there is an advantage in that the accuracy of the sensing results / inference of the AI ​​model (120) can be improved or the amount of computation for storing and processing various AI data can be adjusted through an operation of increasing or decreasing the number of reference signals between the signal receiving device (100, UE) / signal transmitting device (200, BS) when necessary (e.g., when the computational capacity such as AI Computing and / or ISAC Capability changes).

[0094] Furthermore, in the present invention, the time / frequency / period, etc. of the frame for transmitting / receiving a reference signal (single or multiple) transmitted / received between a signal receiving device (100, UE) / signal transmitting device (200, BS) for learning / inference of an AI model (120) can be adjusted.

[0095] For example, in the present invention, the transmission (transmission / reception) time of a reference signal (single or multiple) for learning / inference of an AI model (120) can be allocated by the signal transmission device (200, BS) at a desired time when setting the AI ​​Parameter Set in units of Frame, sub-Frame, Slot, etc.

[0096] Alternatively, in the present invention, if the time / frequency / period resources allocated to transmission (transmission / reception) of a reference signal (single or multiple) for learning / inference of an AI model (120) are not determined, a default value may be allocated and used in the system, and the signal transmission device (200, BS) may transmit this to the signal reception device (100, UE) when setting the AI ​​Parameter Set, or may implicitly transmit that the default value has been used.

[0097] Next, the AI ​​inference process performed according to the AI ​​Parameter Set settings is described as follows.

[0098] In the signal receiving device (100, UE) of the present invention, the sensing operation unit (130) can obtain an AI parameter (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) selected / specified as AI parameter information for transmission / reception of a specific signal set in advance, for example, a reference signal (single or multiple) according to a signal list, according to the AI ​​Parameter Set setting.

[0099] As explained above, this can be obtained through sensing capabilities (e.g., TDoA, RSRP, RSSI, CQI, etc.) possessed by the signal receiving device (100, UE).

[0100] And, the sensing operation unit (130) inputs AI parameters (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) obtained from transmission and reception of the tactical reference signal (single or multiple) into the AI ​​model (120), and can obtain sensing results / data according to the performance of the sensing operation inferred for the current input from the AI ​​model (120) as output.

[0101] As described above, in the present invention, in the ISAC scenario, an AI Parameter Set for AI learning / inference, particularly an AI Computing and / or ISAC Capability of a signal receiving device (100, Sensing Entity), is selected / classified to suit the computational capacity and the sensing capability / purpose it possesses, and a configuration is implemented to set a specific AI Parameter Set between signal transmitting and receiving devices (e.g., a base station or an Edge server / terminal).

[0102] That is, in the present invention, a device-to-device operation for setting an AI Parameter Set is defined, AI learning / inference is implemented according to the AI ​​Parameter Set set through the operation, and a series of AI-based Sensing operations are implemented to support ISAC-related Sensing operations through AI inference based on this.

[0103] Through this, the present invention enables learning / inference of an AI model (120) for ISAC / Sensing operation according to the situation of each signal receiving device (Sensing Entity), thereby minimizing the impact due to differences in AI Computing and / or ISAC Capability between devices, thereby enabling the signal receiving device (Sensing Entity) to perform high-performance AI-based Sensing operation.

[0104] In particular, in the present invention, it is expected that the reference signal, which was previously only used for channel estimation, signal demodulation, phase tracking, etc., can be utilized as composite data by collecting single or multiple reference signals during AI-based sensing operation through AI Parameter Set setting in addition to the original use.

[0105] In addition, in the present invention, the signal receiving device (100, UE) can transmit the inferred sensing result / data by performing an AI-based sensing operation to the signal transmitting device (200, BS) through a desired time / frequency slot. At this time, the transmission method can be DCI, UCI through RRC, MAC CE, PDCCH, and when a resource for ISAC is separately allocated, the data channel of the corresponding resource can be utilized for transmission.

[0106] To explain a more specific embodiment, the signal receiving device (100, UE) of the present invention can perform a verification procedure for the sensing result / data inferred from the AI ​​model (120) by utilizing past sensing result information stored by performing an AI-based sensing operation, or location estimation information derived by performing a separate sensing operation.

[0107] For example, the signal receiving device (100, UE) of the present invention can store the inferred sensing results / data by performing AI-based sensing operations for a certain period of time and manage them as past sensing result information.

[0108] Accordingly, the signal receiving device (100, UE) performs an AI-based sensing operation and compares the newly inferred sensing result / data with past sensing result information (particularly information matching the input condition), thereby performing a verification procedure on the sensing result / data inferred from the AI ​​model (120), thereby verifying the performance / accuracy / reliability of the AI ​​model (120).

[0109] As another example, the signal receiving device (100, UE) of the present invention can derive location estimation information by performing a separate sensing operation each time it performs an AI-based sensing operation to infer sensing results / data, or periodically during inference.

[0110] For example, the signal receiving device (100, UE) can derive the existing sensing (or positioning) operation through the sensing capability (e.g., TDoA, RSRP, RSSI, CQI, etc.).

[0111] Accordingly, the signal receiving device (100, UE) can verify the performance / accuracy / reliability of the AI ​​model (120) by performing a verification procedure on the sensing results / data inferred from the AI ​​model (120) by comparing the newly inferred sensing results / data by performing an AI-based sensing operation with the position estimation information derived from the existing sensing (or positioning) operation.

[0112] In addition, the signal receiving device (100, UE) can update the AI ​​model (120) by performing a subsequent procedure to improve the performance / accuracy / reliability of the AI ​​model (120) when the performance / accuracy / reliability of the AI ​​model (120) falls below a reference value.

[0113] In this way, the verification procedure for the performance / accuracy / reliability of the AI ​​model (120) performed in the signal receiving device (100) of the present invention may be performed independently in the signal receiving device (100) or may be performed in conjunction with the signal transmitting device (200).

[0114] For example, the signal transmission device (200, BS, ISAC control unit (220)) can manage past sensing result information for each signal reception device (100, UE).

[0115] In addition, the signal transmission device (200, BS, ISAC control unit (220)) performs an AI-based sensing operation from each signal reception device (100, UE) to receive newly inferred sensing results / data, and by utilizing the past sensing result information, the performance / accuracy / reliability of the AI ​​model (120) of the signal reception device (100, UE) can be verified.

[0116] Accordingly, when a signal receiving device (100, UE) is identified in which the performance / accuracy / reliability of the AI ​​model (120) falls below a reference value, the signal transmitting device (200, BS, ISAC control unit (220)) may update the AI ​​model (120) of the signal receiving device (100, UE) by performing a follow-up procedure to improve the performance / accuracy / reliability of the AI ​​model (120) for the signal receiving device (100, UE).

[0117] At this time, a follow-up procedure for improving the performance / accuracy / reliability of the AI ​​model (120) may be performed by re-performing the process of setting the AI ​​Parameter Set for the signal receiving device (100) and increasing the number of reference signals through the signal list.

[0118] Of course, the subsequent procedure for improving the performance / accuracy / reliability of the AI ​​model (120) may be carried out by increasing the learning period of the AI ​​model (120) and retraining it, in addition to the tactical example.

[0119] As described above through various embodiments, according to the present invention, by combining AI technology and ISAC technology, a new type of AI-based sensing operation is implemented through a specific technology configuration that can increase and improve the accuracy of sensing results / data according to the sensing operation of the ISAC technology.

[0120] Accordingly, according to the present invention, the effect of performing high-performance AI-based sensing operations in a signal receiving device (sensing entity) can be achieved by minimizing the impact caused by differences in AI Computing and / or ISAC Capability between devices.

[0121] Meanwhile, in the tactic, it was explained assuming that the UE is a signal receiving device (100) and the BS (or Edge server) is a signal transmitting device (200) (e.g., BS->UE Bi-static), but the present invention is not limited thereto.

[0122] That is, the AI-based sensing operation of the present invention can be applied to all of the first to sixth modes described above, depending on the Sensing Entity that measures the propagation of the signal and generates the sensing result / data.

[0123] Below, with reference to FIGS. 3 and 4, is an exemplary flowchart explaining the operation flow of the AI-based sensing operation method proposed in the present invention.

[0124] Before a specific description, the signal receiving device (100) of the present invention refers to a device that operates as a Sensing Entity that measures the radio wave of a signal in an ISAC scenario, and may be a base station (BS) or a terminal (UE).

[0125] However, for the convenience of explanation, in the following, it will be assumed that a UE having relatively limited computational capacity, such as AI Computing and / or ISAC Capability, is a signal receiving device (100), and a BS (or Edge server) is a signal transmitting device (200) (e.g., BS->UE Bi-static).

[0126] Referring to FIG. 3, according to the AI-based sensing operation method of the present invention, the BS (200) as a signal transmission device (200) can perform an AI Parameter Set setting process for transmitting / setting an AI Parameter Set for mutual AI learning / inference to the UE (100) with which the sensing operation is mutually performed (S10).

[0127] For example, BS (200) transmits an AI Parameter Set for mutual AI learning / inference to UE (100).

[0128] The AI ​​Parameter Set transmitted at this time may include information (hereinafter, AI Parameter information) that selects / specifies some or all of the information used in existing location estimation (e.g., TDoA, RSRP, RSSI, CQI, RI, etc.) as AI Parameters to be used in AI learning / inference.

[0129] As a more specific example, the AI ​​Parameter Set may further include a list of signals to be utilized when learning / inferring an AI model (120) of a UE (100), and for example, a list of signals specifying reference signals (e.g., DMRS, PTRS, PRS, CSI-RS, etc.) to be utilized in obtaining selected / specific AI parameters may be included.

[0130] The UE (100) can transmit its capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability) for the information included in the transmitted AI Parameter Set in the form of an Ack.

[0131] Of course, the UE (100) can perform AI learning / inference according to the AI ​​Parameter Set settings thereafter by setting the information included in the transmitted AI Parameter Set, if acceptable in its Capability.

[0132] According to the AI-based sensing operation method of the present invention, when the BS (200) receives capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability) and setting completion in the form of an Ack from the UE (100), the setting process of the mutual AI Parameter Set can be completed.

[0133] Meanwhile, if the previously transmitted AI Parameter Set is not set to the UE (100), the BS (200) may re-perform the AI ​​Parameter Set setting process to reset the AI ​​Parameter Set with a lower computational amount to suit the UE (100) based on the Capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability) confirmed in the Ack.

[0134] According to the AI-based sensing operation method of the present invention, the UE (100) as a signal receiving device (100) can perform AI learning / inference according to the AI ​​Parameter Set setting set internally through the aforementioned AI Parameter Set setting process.

[0135] First, the AI ​​learning process performed according to the AI ​​Parameter Set settings is as follows.

[0136] The UE (100) can obtain an AI parameter (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) selected / specified as AI parameter information (S30) for transmission / reception of a specific signal set in advance, for example, a reference signal (single or multiple) according to a signal list (S20, S25).

[0137] Meanwhile, the UE (100) can generate sensing results / data by measuring the radio wave of the reference signal (sensing signal) (single or multiple) used to obtain the AI ​​parameter (e.g., one or multiple of TDoA, RSRP, RSSI, CQI, RI, etc.) and performing a sensing operation based on the radio wave measurement value.

[0138] Accordingly, in the UE (100), for a certain period of time (e.g., a set period, or until a condition that satisfies AI model performance is satisfied), the AI ​​model (120) can be trained to infer the sensing result / data of the same reference signal (sensing signal) as an output (Output) for the input (Input, AI Parameter (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.)) obtained from transmission and reception of a reference signal (single or multiple) (S30).

[0139] Next, the AI ​​inference process performed according to the AI ​​Parameter Set settings is described as follows.

[0140] The UE (100) can obtain an AI Parameter (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) selected / specified as AI Parameter information (S50) for transmission / reception of a specific signal set in advance, for example, a reference signal (single or multiple) according to a signal list (S40, S45).

[0141] And, the UE (100) inputs AI parameters (e.g., one or more of TDoA, RSRP, RSSI, CQI, RI, etc.) obtained from transmission and reception of a tactical reference signal (single or multiple) into the AI ​​model (120), and obtains sensing results / data according to the performance of the sensing operation inferred for the current input from the AI ​​model (120) as output, thereby performing the AI-based sensing operation of the present invention (S50).

[0142] As described above, in the present invention, in the ISAC scenario, an AI Parameter Set for AI learning / inference, particularly an AI Computing and / or ISAC Capability of a signal receiving device (100, Sensing Entity), is selected / classified to suit the computational capacity and the sensing capability / purpose it possesses, and a configuration is implemented to set a specific AI Parameter Set between signal transmitting and receiving devices (e.g., a base station or an Edge server / terminal).

[0143] That is, in the present invention, an operation between devices for setting an AI Parameter Set is defined, AI learning / inference is implemented according to the AI ​​Parameter Set set through the operation, and a series of AI-based Sensing operations are implemented based on this through AI inference of the Sensing operation of ISAC.

[0144] Through this, the present invention enables learning / inference of an AI model (120) for ISAC / Sensing operation according to the situation of each signal receiving device (Sensing Entity), thereby minimizing the impact due to differences in AI Computing and / or ISAC Capability between devices, thereby enabling the signal receiving device (Sensing Entity) to perform high-performance AI-based Sensing operation.

[0145] In particular, in the present invention, it is expected that the reference signal, which was previously only used for channel estimation, signal demodulation, phase tracking, etc., can be utilized as composite data by collecting single or multiple reference signals during AI-based sensing operation through AI Parameter Set setting in addition to the original use.

[0146] Referring to the following FIG. 4, an additional embodiment will be described. According to the AI-based sensing operation method of the present invention, the UE (100) as a signal receiving device (100) performs the AI-based sensing operation of the present invention by obtaining the sensing result / data inferred through the AI ​​model (120) as an output, as in step S50 referring to the aforementioned FIG. 3, and then transmits the data related thereto to the BS (200) (S70).

[0147] Data related to AI-based sensing operations transmitted in this way may include sensing results / data as a result of AI-based sensing operations, internally conducted verification results, and labeling of whether AI is used.

[0148] As an example illustrated in FIG. 4 illustrates, the UE (100) can verify the reliability of the AI ​​model (120) by performing an AI-based sensing operation in step S50 and comparing the newly inferred sensing result / data with internally stored sensing result information or position estimation information derived from an existing sensing (or positioning) operation (S60).

[0149] In addition, the UE (100) can transmit the AI-based sensing operation results performed this time, i.e., the sensing results / data, the verification results, and the AI ​​usage labeling indicating whether AI was used in the sensing results / data transmitted this time, to the BS (200) (S70).

[0150] In this way, the BS (200) as a signal transmission device (200) can manage ISAC operation data based on data transmitted from each signal reception device (100), i.e., each UE (100), and by utilizing this, various operation procedures related to ISAC can be made possible thereafter (S80).

[0151] Here, the ISAC operation data includes, for each signal receiving device (100), i.e., each UE (100), past sensing result information, verification result, and labeling of whether AI is used for each sensing result / data, and may further include previously transmitted capability information (e.g., computational capacity such as AI Computing and / or ISAC Capability).

[0152] As described above, according to the present invention, by combining AI technology and ISAC technology, a new type of AI-based sensing operation is implemented through a specific technology configuration that can increase and improve the accuracy of sensing results / data according to the sensing operation of the ISAC technology.

[0153] Accordingly, according to the present invention, the effect of performing high-performance AI-based sensing operations in a signal receiving device (sensing entity) can be achieved by minimizing the impact caused by differences in AI Computing and / or ISAC Capability between devices.

[0154] An AI-based sensing operation method according to an embodiment of the present invention may be implemented in the form of program commands that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program commands, data files, data structures, etc., alone or in combination. The program commands recorded on the medium may be those specially designed and configured for the present invention or may be those known to and usable by those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program commands, such as ROMs, RAMs, and flash memories. Examples of program commands include not only machine language codes generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. The above hardware devices may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

[0155] Although the present invention has been described in detail with reference to preferred embodiments, the present invention is not limited to the above-described embodiments, and it will be understood that the technical idea of ​​the present invention extends to a range in which various modifications or changes can be made by anyone having ordinary skill in the art to which the present invention pertains without departing from the gist of the present invention as claimed in the following claims.

Claims

1. In a signal receiving device that receives a signal, memory containing instructions; and A signal receiving device characterized by comprising a processor that performs an AI-based sensing operation, inferring a sensing result according to the performance of the sensing operation through an AI model based on an AI parameter obtained through transmission and reception of a preset specific signal, in relation to a sensing operation performed based on a radio wave measurement of the signal by executing the command; 2. In paragraph 1, The above specific signal is, A signal receiving device characterized in that it is at least one reference signal included in an AI Parameter Set set for obtaining the AI ​​Parameter, through a setting procedure with a signal transmitting device that mutually performs the sensing operation.

3. In paragraph 1, The above AI model is, The AI ​​parameter obtained through transmission and reception of the specific signal is used as input, and the sensing result derived by performing the sensing operation based on the radio wave measurement value for the specific signal is used as output. A signal receiving device characterized in that it is a model trained to infer the output for the input.

4. In paragraph 1, The above processor, A signal receiving device characterized in that it performs a verification procedure for the sensing result inferred from the AI ​​model by utilizing past sensing result information stored by performing the AI-based sensing operation or location estimation information derived by performing a separate sensing operation.

5. In paragraph 2, The above processor, A signal receiving device characterized in that, when a signal list to be used in the AI ​​model is confirmed when setting the AI ​​Parameter Set, only signals in the signal list among the signals in the AI ​​Parameter Set are used to train the AI ​​model or infer the sensing result.

6. In the terminal device, memory containing instructions; and A terminal device characterized by comprising a processor that performs an AI-based sensing operation, inferring a sensing result according to the performance of the sensing operation through an AI model based on an AI parameter obtained through transmission and reception of a preset specific signal, in relation to a sensing operation performed based on measurement of a radio wave of a signal serviced in mobile communication by executing the above command.

7. In a signal transmitting device that receives a signal, memory containing instructions; and By executing the above command, in relation to the sensing operation performed based on the measurement of the radio wave propagation of the above signal, A signal transmission device characterized by including a processor that sets an AI Parameter Set for acquiring AI Parameters in a signal receiving device that mutually performs the sensing operation, so that the signal receiving device can acquire and use the AI ​​Parameters through transmission and reception of a specific signal according to the AI ​​Parameter Set settings to perform an AI-based sensing operation.

8. In an AI-based sensing operation method performed in a signal receiving device that receives a signal, In relation to the sensing operation performed based on the radio wave measurement of the above signal, a step of setting an AI Parameter Set for obtaining AI Parameters through a setting procedure with a signal transmission device that mutually performs the sensing operation; A step of obtaining AI parameters through transmission and reception of specific signals according to the above AI Parameter Set settings; An AI-based sensing operation method characterized by including a step of performing an AI-based sensing operation to infer a sensing result according to the performance of the sensing operation through an AI model based on the acquired AI parameter.

Images