Measurement value processing system and measurement value processing method

Abstract

This process allows for the suppression of the influence of objects outside the monitoring space when measuring the propagation path characteristics of wireless signals passing through the monitoring space, while also taking into account fluctuations in the static environment within the monitoring space. [Solution] The measurement value processing system according to one embodiment of the present invention uses the processor 41 of the motion determination device 30 to process pre-set dimensional information with respect to the monitoring space, the sampling frequency of the wireless signal, and static environmental measurement values ​​(h) based on measurements obtained in an environment where no moving objects are present in the monitoring space. baseline ) is used to determine the time threshold (integrated threshold τ) corresponding to the boundary of the monitoring space. int The system is configured to perform a threshold calculation process to determine the threshold value, and a spatial separation process to suppress signal components with propagation delays exceeding a time threshold among the signal components included in the measured values ​​obtained during system operation. Furthermore, if a change in static environmental measurement values ​​is detected during system operation, the threshold calculation process is re-executed.

Description

[Technical Field]

[0001] The present invention relates to a measurement value processing system for processing measurement values ​​that represent the propagation path characteristics of a radio signal that has reached a receiver from a transmitter through multiple propagation paths. [Background technology]

[0002] In recent years, development has progressed on a technology called WLAN sensing, which uses Wi-Fi® wireless signals to detect the movement and position of targets (people and objects), and it has been standardized as "IEEE 802.11bf". WLAN sensing detects the movement and position of objects in the communication environment by analyzing the minute changes in phase and amplitude that occur when wireless signals transmitted and received by Wi-Fi-enabled wireless devices are reflected and attenuated by objects in the communication environment (people, walls, furniture, etc.) based on CSI (Channel State Information). With WLAN sensing, it is possible to detect the movement of objects without using video from cameras, etc., making it possible to understand the state of a person while respecting privacy. In addition, development is also progressing on "BLE Channel Sounding" (hereinafter referred to as "BLE CS") technology, which measures the distance between two devices using Bluetooth® wireless signals.

[0003] Prior art related to the present invention includes the following. For example, Patent Document 1 discloses an invention that causes a computer to perform the following actions: to obtain the initial position of an object before it moves; to obtain at least one radio signal from a multipath channel affected by the object's movement; to extract a time-series CSI of the multipath channel from at least one radio signal; to determine the distance of the object's movement based on the time-series CSI; to estimate the direction of the object's movement; and to determine the new position of the object after the movement based on the distance, direction, and initial position.

[0004] The technology described in Patent Document 1 is an approach based on statistical analysis in Euclidean space, and uses time reversal theory to analyze statistical quantities such as the correlation of multipath signals, RSSI (Received Signal Strength Indicator), and the variance and standard deviation of CSI to track objects. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Special Publication No. 2019-518202 [Overview of the project] [Problems that the invention aims to solve]

[0006] In an indoor environment, radio signals transmitted from a transmitting radio reach the receiving radio via multiple paths, including reflections from walls, furniture, people, and objects outside the monitoring space (such as people in adjacent rooms or corridors, or passersby outdoors). This phenomenon is known as multipath propagation. When using CSI data for purposes such as detecting target activity within a monitoring space, multipath components outside the monitoring space become inherent noise and are a major cause of false detections.

[0007] Conventional WLAN sensing technologies, as shown in Patent Document 1 and others, primarily employ the following methods as countermeasures against multipath interference. (1) Time-domain noise averaging by statistical analysis (autocorrelation, variance, mean, etc.) (2) Fitting known patterns by training the classifier (3) Directional filtering by multiple antenna beamforming

[0008] These methods are all limited to statistical processing and signal quality improvement, and do not take into account multipath components outside the surveillance space. Furthermore, in statistical processing, if a noise source outside the surveillance space has the same features and behavioral patterns as the target inside the surveillance space (for example, if the target is a person in one room and another person is in the next room), the signal components of the noise source cannot be fundamentally removed, leading to false detections. Conventional BLE CS technologies also focus on improving attack resistance and ranging accuracy, and do not take into account multipath components outside the surveillance space.

[0009] This invention has been made in view of the above-mentioned conventional circumstances, and aims to enable a process to suppress the influence of objects outside the monitoring space on measured values ​​representing the propagation path characteristics of wireless signals passing through the monitoring space, while taking into account fluctuations in the static environment within the monitoring space. [Means for solving the problem]

[0010] A measurement value processing system according to one aspect of the present invention is a measurement value processing system that processes measurement values ​​representing propagation path characteristics of a wireless signal propagating in a monitoring space, comprising one or more processors, the processor performing a threshold calculation process to calculate a time threshold corresponding to the boundary of the monitoring space using pre-set dimensional information with respect to the monitoring space, the sampling frequency of the wireless signal, and static environmental measurement values ​​based on measurement values ​​obtained in an environment where no moving objects exist in the monitoring space, and a spatial separation process to suppress signal components among the measurement values ​​obtained during system operation that have a propagation delay exceeding the time threshold, and further configured to re-execute the threshold calculation process when a change in the static environmental measurement values ​​is detected during system operation.

[0011] In the above-described measurement processing system, the threshold calculation process may be configured to calculate a reference threshold that serves as the basis for the time threshold using the dimensional information and the sampling frequency, calculate an effective threshold corresponding to the effective boundary of the monitoring space using the static environmental measurement values, and integrate the reference threshold and the effective threshold to calculate the time threshold.

[0012] Furthermore, in the measurement value processing system described above, the threshold calculation process can selectively use one of a plurality of calculation methods that integrate the reference threshold and the effective threshold, and the plurality of calculation methods may be configured to include a method in which the smaller of the reference threshold and the effective threshold is used as the time threshold, and a method in which the result of weighted addition of the reference threshold and the effective threshold is used as the time threshold.

[0013] Furthermore, in the above-described measurement processing system, the threshold calculation process may be configured to adjust the time threshold using the estimated maximum delay calculated by environmental measurements in each predetermined number of frames.

[0014] Furthermore, in the measurement data processing system described above, the spatial separation process may be configured to perform a calculation that suppresses signal components having propagation delays exceeding the time threshold, using a window function with the time threshold as the center value.

[0015] Furthermore, the above-described measurement processing system may be configured to have presets for each room type in which dimensional information is set in advance, and when a room type relating to the monitoring space is selected by the user, the dimensional information set in the preset for that room type may be used as the dimensional information set in advance for the monitoring space.

[0016] Furthermore, in the above-described measurement value processing system, the processor may be configured to further perform a baseline subtraction process, which subtracts the static environmental measurement value from the measurement value obtained after the spatial separation process.

[0017] Furthermore, in the measurement value processing system described above, the processor may be configured to individually perform the spatial separation process on a first measurement value acquired by a first wireless standard and a second measurement value acquired by a second wireless standard that uses a communication method different from the first wireless standard, and to integrate the results of the spatial separation process on the first measurement value and the results of the spatial separation process on the second measurement value.

[0018] A measurement value processing method according to one aspect of the present invention is a measurement value processing method that processes measurement values ​​representing propagation path characteristics of a wireless signal propagating in a monitoring space using one or more processors, comprising the steps of: performing a threshold calculation process to calculate a time threshold corresponding to the boundary of the monitoring space using pre-set dimensional information with respect to the monitoring space, the sampling frequency of the wireless signal, and static environmental measurement values ​​based on measurement values ​​obtained in an environment where no moving objects exist in the monitoring space; and performing a spatial separation process to suppress signal components among the signal components included in the measurement values ​​obtained during system operation that have a propagation delay exceeding the time threshold, wherein the method is configured to re-execute the threshold calculation process when a change in the static environmental measurement values ​​is detected during system operation.

[0019] In the above measurement processing method, the threshold calculation process may be configured to calculate a reference threshold that serves as the basis for the time threshold using the dimensional information and the sampling frequency, calculate an effective threshold corresponding to the effective boundary of the monitoring space using the static environmental measurement values, and integrate the reference threshold and the effective threshold to calculate the time threshold.

[0020] Furthermore, in the measurement value processing method described above, the threshold calculation process can selectively use one of a plurality of calculation methods that integrate the reference threshold and the effective threshold, and the plurality of calculation methods may be configured to include a method in which the smaller of the reference threshold and the effective threshold is used as the time threshold, and a method in which the result of weighted addition of the reference threshold and the effective threshold is used as the time threshold.

[0021] Furthermore, in the above measurement processing method, the threshold calculation process may be configured to adjust the time threshold using the estimated maximum delay calculated by environmental measurements in each frame for every predetermined number of frames.

[0022] Furthermore, in the measurement data processing method described above, the spatial separation process may be configured to perform an operation to suppress signal components having propagation delays exceeding the time threshold, using a window function with the time threshold as the center value.

[0023] Furthermore, the above measurement processing method may be configured to have presets for each room type in which dimensional information is set in advance, and when a room type relating to the monitoring space is selected by the user, the dimensional information set in the preset for that room type may be used as the dimensional information set in advance for the monitoring space.

[0024] Furthermore, the above-described measurement processing method may be configured to include a step of further performing a baseline subtraction process, which subtracts the static environmental measurement value from the measurement value obtained after the spatial separation process.

[0025] Furthermore, in the above-described measurement value processing method, the spatial separation processing may be performed individually on a first measurement value obtained by a first wireless standard and a second measurement value obtained by a second wireless standard that uses a communication method different from the first wireless standard, and the results of the spatial separation processing for the first measurement value and the results of the spatial separation processing for the second measurement value may be integrated. [Effects of the Invention]

[0026] According to the present invention, it is possible to perform a process that suppresses the influence of objects outside the monitoring space on measured values ​​representing the propagation path characteristics of wireless signals passing through the monitoring space, while taking into account fluctuations in the static environment within the monitoring space. [Brief explanation of the drawing]

[0027] [Figure 1]This figure shows an example configuration of an operation determination system according to one embodiment of the present invention. [Figure 2] This figure shows an example of the hardware that makes up the operation determination device shown in Figure 1. [Figure 3] This figure shows an example of the functions of the operation determination device shown in Figure 1. [Figure 4] This figure shows an example of the processing flow using the operation determination device shown in Figure 1. [Figure 5] This figure shows an example of the processing flow when implementing a system that integrates spatial separation and baseline reduction. [Figure 6] This figure shows an example of the processing flow for real-time processing during system operation related to the integration of spatial separation and baseline reduction. [Figure 7] This figure shows an example of the processing flow for background processing during system operation related to the integration of spatial separation and baseline reduction. [Figure 8] This figure shows the measurement results from experiments related to spatial separation processing. [Modes for carrying out the invention]

[0028] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Figure 1 shows an example of the configuration of an action determination system according to an embodiment of the present invention. The action determination system shown in Figure 1 comprises a plurality of wireless devices 20 (here, two wireless devices 20A and 20B) that perform wireless communication in a communication environment including a monitoring space 10, and an action determination device 30 that determines the action of a target present in the space 10 based on measurement data related to wireless communication. The monitoring space 10 is, for example, the main living space of a person 12 who is the target of action determination. Static structures 16 such as walls, furniture, beds, and wheelchairs may exist inside the monitoring space 10. Hereinafter, an example will be given of determining whether a target present in the monitoring space 10 has fallen over.

[0029] Radio transceivers 20A and 20B are wireless communication devices compliant with the Wi-Fi standard. Radio transceivers 20A and 20B may be dedicated devices with the functions described later, or they may be general-purpose devices. Examples of general-purpose devices include wireless routers, smartphones, or home appliances with wireless communication capabilities such as air conditioners and lighting devices (so-called "smart appliances"). For the sake of explanation, in the following, radio transceiver 20A will be referred to as the transmitter (transmitting station) and radio transceiver 20B as the receiver (receiving station), but each radio transceiver may be configured to have both transmitting and receiving functions.

[0030] Upon receiving a radio signal transmitted from radio 20A, radio 20B calculates CSI, which represents the propagation path characteristics related to multiple propagation paths (multipath), as measurement data related to wireless communication. In Wi-Fi, methods such as OFDM (Orthogonal Frequency Division Multiplexing) and OFDMA (Orthogonal Frequency Division Multiple Access) are used, and radio 20A divides the transmission data into multiple subcarriers and transmits it as a radio signal. The radio signal transmitted from radio 20A reaches radio 20B after being affected by its phase and amplitude as it passes through multiple propagation paths. Radio 20B estimates the phase and amplitude for each subcarrier of the received radio signal and outputs CSI as measurement data.

[0031] Furthermore, CSI measurement is not limited to a configuration based on radio signals transmitted and received between multiple radio units 20. For example, radio unit 20B may send a radio signal towards the monitoring space 10, and radio unit 20B itself may receive the radio signal reflected by an object in the monitoring space 10 to perform CSI measurement. This configuration is effective for monitoring relatively small spaces (for example, the interior of a car).

[0032] The motion determination device 30 determines the motion of a target present in the monitoring space 10 based on time-series data of the CSI obtained with respect to the wireless signal propagating in the monitoring space 10. When a target present in the monitoring space 10 performs any motion, fluctuations in phase and amplitude occur in the wireless signal propagating near the target, and a fluctuation pattern specific to the motion appears in the time-series data of the CSI. Therefore, the motion of the target can be determined by analyzing the temporal fluctuations of the CSI.

[0033] The motion detection device 30 may be a device deployed inside or near the monitoring space 10, or it may be an external server (including a cloud server) connected via the Internet. Also, although the motion detection device 30 is represented as a separate device in Figure 1, it may be integrated into the wireless device 20B.

[0034] As shown in Figure 2, the operation determination device 30 is composed of a computer equipped with hardware resources such as a processor 41, memory 42, and storage 43. For example, the processor 41 loads a predetermined program stored in the storage 43 onto the memory 42 and executes it, thereby realizing each of the functions described later (see Figure 3). The processor 41 can be freely selected from existing processors such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), and two or more can be used in any combination. The storage 43 can be freely selected from existing devices such as an HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory, and two or more can be used in any combination.

[0035] Figure 3 shows an example of the functions of the motion determination device 30. The motion determination device 30 shown has a CSI acquisition function 51, a spatial separation function 52, a motion determination function 53, and a notification function 54. Each of these functions will be described in detail below.

[0036] The CSI acquisition function 51 performs a CSI acquisition process to acquire CSI data relating to radio signals propagating within the monitoring space 10 from radio devices 20 (e.g., radio device 20B) deployed inside or near the monitoring space 10. If CSI is measured by multiple radio devices 20, CSI data may be acquired from each radio device 20.

[0037] The spatial separation function 52 applies spatial separation processing (Spatial Shutter) to the CSI data acquired by the CSI acquisition function 51 to suppress the influence (external noise) of objects outside the monitoring space (for example, the person 14 shown in Figure 1). The details of the spatial separation processing will be explained in detail later.

[0038] The motion determination function 53 performs motion determination processing to determine the motion of a target present in the monitoring space 10 (for example, whether or not the target has fallen over) based on the CSI data that has undergone spatial separation processing. For target motion determination, for example, known WLAN sensing technology can be used. Alternatively, target motion determination may be implemented using the technology described in Patent Document 1 mentioned above. Furthermore, complex feature quantities representing the phase and amplitude included in the CSI may be mapped to a hyperbolic space with negative curvature characteristics, and target motion may be determined based on geometric indicators in the hyperbolic space.

[0039] The notification function 54 performs notification processing to inform a user, such as a monitor, of the target's fall if, as a result of the motion detection processing, it is determined that a target located within the monitoring space 10 has fallen over. The notification processing can be performed through the device itself or another device using any method such as screen display, audio output, lamp illumination, or a combination thereof.

[0040] Next, the details of the spatial separation process will be explained. The spatial separation process performs calculations on the propagation delay of each propagation path in the CSI data acquired by the CSI acquisition function 51, separating it into a first signal component (hereinafter referred to as the "desired component") which is estimated to be unaffected by objects outside the monitoring space, and a second signal component (hereinafter referred to as the "external noise component") which is estimated to be affected by objects outside the monitoring space. Processed CSI data with the external noise component suppressed is then generated and output.

[0041] Spatial separation can be achieved using various methods. For example, spatial separation can be achieved by the following procedure. (Procedure A1) A first conversion process is performed to convert CSI data, which is the frequency domain representation of the propagation path characteristics of wireless signals propagating within the monitoring space 10, into CIR (Channel Impulse Response) data, which is the time domain representation. (Procedure A2) By applying time-domain gating to the CIR data, signal components with propagation delays exceeding a predetermined time threshold are removed as external noise components. (Procedure A3) A second conversion process is applied to the CIR data from which external noise components have been removed, converting it back into CSI data, which is a frequency domain representation. The result is then output as processed CSI data.

[0042] The first transformation process is achieved, for example, by performing an IFFT (Inverse Fast Fourier Transform) on the CSI data. If CSI is H(k) and CIR is h(τ), then h(τ) can be expressed by the following equation (1).

[0043]

number

[0044] Here, k is the index of a subcarrier, and H(k) represents the complex channel response of subcarrier k. Further, τ is a time delay index (tap), which physically corresponds to the signal propagation delay from the transmitting station (radio device 20A) to the receiving station (radio device 20B). That is, a tap with a larger τ represents a signal component that has passed through a longer path length (a far multipath).

[0045] In the time-domain gating process, for example, τ which is the threshold of the delay time corresponding to the boundary of the monitoring space 10 th is used to perform a hard truncation process shown in the following equation (2). That is, a process of removing a signal component exceeding τ th from h(τ) as an external noise component is performed, and the result is output as h^(τ).

[0046] [Equation]

[0047] Here, when the maximum distance of the monitoring space is d max , the speed of light is c (= 3 × 10 8 m / s), and the sampling period is T s (= 1 / bandwidth), the threshold τ of the delay time corresponding to the boundary of the monitoring space th can be expressed by the following equation (3). For example, when using T s = 50 ns (802.11n HT20, 20 MHz bandwidth), in the case of d max = 15 m, τ th = 1, and in the case of d max = 30 m, τ th = 2. When a wider bandwidth (for example, 80 MHz, 160 MHz, 320 MHz of 802.11bf-2025) is used, finer delay resolution can be obtained.

[0048] [Equation]

[0049] The second transformation is achieved by applying an FFT (Fast Fourier Transform) to h^(τ), which is the result of the time-domain gating process. If H^(k) is the CSI after the FFT, then H^(k) can be expressed by equation (4) below. In this way, the second transformation is performed as the inverse transformation of the first transformation.

[0050]

number

[0051] Note that the above threshold τ th τ is a fixed value, but it is an adaptive threshold that is adjusted according to the environment. ad It is also possible to use (t). τ ad (t) can be calculated, for example, by the following equation (5). Here, τ base λ is the estimated maximum delay calculated from environmental measurements, and λ∈(0,1) is a coefficient that determines the adaptation speed.

[0052]

number

[0053] Furthermore, time-domain gating is not limited to the hard truncation process described above; it is also possible to use the soft gating function shown in equation (6) below. Here, w(τ) is the sigmoid function and κ is the steepness parameter.

[0054]

number

[0055] Soft gating has the advantage of suppressing frequency domain leakage compared to hard truncation. Alternatively, the Hanning window, Gaussian window, Tukey window, etc., may be applied instead of the sigmoid function.

[0056] Here, we will explain why the sampling frequency of the wireless signal is used when calculating the threshold. Using multiple rooms of different sizes as monitoring spaces, we conducted experiments on spatial separation processing in three cases: (S0) when both the room and the corridor are empty, (S2) when one person is in the room and the corridor is empty, and (S3) when the room is empty and one person is in the corridor. The measurement results shown in Figure 8 were obtained. In this figure, IPR represents the Indoor Power Ratio, and SNR represents the Signal-to-Power Ratio.

[0057] As is clear from the measurement results in Figure 8, there is a clear difference in the spatial separation processing results between rooms of 4-6m size and rooms of 8-10m size. Detailed analysis of these measurement results revealed that the reason for the large change in performance depending on the room size is the limitation of the delay resolution of the CIR caused by the subcarrier spacing used in CSI measurement. Therefore, in this invention, when calculating the threshold, a configuration is introduced that uses not only the maximum distance of the monitored space and the speed of light, but also the sampling frequency of the wireless signal.

[0058] Figure 4 shows an example of the processing flow by the operation determination device 30. The operation determination device 30 acquires CSI data obtained with respect to the wireless signal propagating in the monitoring space 10 at each time frame of a predetermined length (step S11). Next, it performs IFFT processing on the CSI data, which is a frequency domain representation, to convert it into CIR data, which is a time domain representation (step S12). Next, it performs time domain gating processing on the CIR data to remove signal components with propagation delays exceeding a predetermined time threshold as external noise components (step S13). Next, it performs FFT processing on the CIR data from which the external noise components have been removed to convert it into CSI data, which is a frequency domain representation (step S14). Next, it provides the CSI data from which the external noise components have been removed to the operation determination process (step S15).

[0059] As described above, in the operation determination system of this example, the processor 41 of the operation determination device 30 performs the following operations: a CSI acquisition process to acquire CSI data representing the propagation path characteristics of a radio signal that has reached the receiving station (radio 20B) from the transmitting station (radio 20A) through multiple propagation paths; a spatial separation process which performs calculations on the acquired CSI data regarding the propagation delay of each propagation path to separate it into a desired component which is estimated to have no influence from objects outside the monitoring space and an external noise component which is estimated to have influence from objects outside the monitoring space, and generates and outputs processed CSI data with the external noise component suppressed; and an operation determination process which determines the operation of a target present in the monitoring space 10 based on the processed CSI data.

[0060] With this configuration, target movement within the monitored space can be determined using processed CSI data from which external noise components have been removed, resulting in more accurate determination than conventional methods. While it is ideal to completely remove external noise components, suppressing them to a certain level is sufficient to improve the accuracy of target movement determination. In this system, processed CSI data is used to determine target movement, but the use of processed CSI data is not limited to this. It can be used for various processes such as people counting, indoor positioning, indoor environment monitoring, gait recognition, and vital signal extraction.

[0061] The method described above is a direct method of mutually converting frequency-domain CSI data and time-domain CIR data using IFFT / FFT, but propagation delay separation can also be performed using any alternative method, as exemplified below. Furthermore, while it is desirable to completely remove external noise components, even suppressing them to some extent can improve the accuracy of target operation determination.

[0062] (Alternative method a) Wavelet Transform By applying continuous small wave conversion (CWT) or discrete small wave conversion (DWT) to CSI data H(k), a two-dimensional representation of scale and frequency W(s,k) can be obtained. Physically, the scale parameter s corresponds to the propagation delay τ. Then, the scale threshold s corresponds to the boundary of the monitoring space. th Using s>s th A gating process is performed to determine that the coefficient corresponding to s > s is an external noise component and set to 0 (remove it). Alternatively, s > s th The coefficients corresponding to this are attenuated using a soft gating function such as a Hanning window. Then, an inverse continuous small wave transform (ICWT) or inverse discrete small wave transform (IDWT) is applied to the gated coefficients, and the result is output as processed CSI data. This allows for spatial selection effects equivalent to IFFT + gating to be obtained.

[0063] (Alternative method b) Chirp-Z Transform By applying chirp Z transformation (CZT) to CSI data H(k), the delay range corresponding to the boundary of the monitoring space [0,τ th It is possible to obtain h'(τ), which is the time-domain representation CIR data calculated with high resolution for only [0,τ]. Due to the characteristics of CZT, the delay range to be calculated is [0,τ]. th By limiting it to ], it becomes implicitly gated. If necessary, τ th Hard truncation may be performed as an additional step. Then, the inverse transform of CZT (ICZT) or a normal IFFT is applied to h'(τ), and the result is output as processed CSI data. By using CZT, CIR can be calculated at any frequency and resolution, and only the delay range corresponding to the monitoring space can be calculated with high resolution.

[0064] (Alternative method c) Subspace estimation (MUSIC / ESPRIT) Apply eigenvalue decomposition such as MUSIC or ESPRIT to the covariance matrix of time-series CSI data (CSI data for multiple frames). For example, CSI data for multiple frames H (t)Based on (k), we construct the covariance matrix R shown in equation (7) below. Then, we perform the eigenvalue decomposition of the covariance matrix R shown in equation (8) below. Here, the superscript H represents the Hermitian transpose, and Λ represents the diagonal eigenvalue matrix.

[0065]

number

[0066]

number

[0067] Next, the group of eigenvectors corresponding to the large eigenvalues ​​is the signal subspace U. s (This corresponds to a short-range propagation path passing through the interior of the monitoring space), and the remaining eigenvectors form the external noise subspace U. n It is separated as (corresponding to a long-distance propagation path passing outside the monitoring space). And the signal subspace U s It retains only the external noise subspace U n Remove it. Then, use equation (9) below to create the signal subspace U s Projection is performed onto the subspace, and the result is output as processed CSI data. By performing eigenvalue decomposition using methods such as MUSIC or ESPRIT, super-resolution estimation is possible. Subspace estimation is a particularly advantageous method for sparse sampling (discrete channels) like BLE CS.

[0068]

number

[0069] (Alternative method d) Subspace separation (SVD / PCA) Singular Value Decomposition (SVD) or Principal Component Analysis (PCA) is applied to a matrix composed of time-series CSI data (CSI data for multiple frames). For example, a matrix X = [H] composed of CSI data for T frames. (1) ,H (2) ,…,H (T)For ](N×T), we perform the singular value decomposition shown in equation (10) below. Here, the superscript T represents the transpose.

[0070]

number

[0071] Next, the top r principal components (signal components corresponding to short-distance propagation paths passing through the inside of the monitoring space) and the residual components (signal components corresponding to long-distance propagation paths passing outside the monitoring space) are separated. Then, only the principal components are retained, and the residual components are set to 0 (removed). That is, matrix X is obtained using equation (11) below. ~ Calculate the matrix X. ~ Each column represents the processed CSI data for each frame (=H^ (t) (k))

[0072]

number

[0073] (Alternative method e) Learning filter (machine learning based) This machine learning model takes CSI data as input and is trained to extract signal components unaffected by external objects in the monitored space and output them as processed CSI data. It extracts only the signal components of the propagation path passing through the monitored space. For example, CSI data H(k) is input to an encoder layer (e.g., a 1D-CNN encoder, a Transformer encoder, etc.) and converted to a latent representation z. Next, a filter layer, trained using true CSI data without external noise as training data, suppresses external noise components in the latent space z. Finally, the suppressed latent representation z is converted to the frequency domain by a decoder layer (e.g., a 1D-CNN decoder, etc.), and the processed CSI data H^(k) is output. Note that the three-stage explanation using encoder, filter, and decoder layers is for convenience, and in practice, it can be implemented as a single neural network.

[0074] The explanation so far concerns spatial separation processing for measured values ​​(CSI) representing the propagation path characteristics of Wi-Fi wireless signals. However, spatial separation processing can also be applied to measured values ​​representing the propagation path characteristics of other wireless signals, such as Bluetooth. Below, we will explain BLE CS as an example.

[0075] In BLE CS, the following measurements are performed across multiple CS channels between the initiator (transmitter) and the reflector (responder). [Mode-0] Frequency / Timing Calibration [Mode-1] Round-trip time (RTT) measurement [Mode-2] Phase-based Ranging (PCT) measurement using channel tone. [Mode-3]RTT+PCT mixed

[0076] The PCT measurements in Mode-2 are obtained by transmitting and receiving single tones to acquire the phase and amplitude of each channel, and for each CS channel k, the amplitude A(k) and phase φ(k) are obtained as a pair. The PCT measurements are expressed in complex form H using the following equation (12). BLE It is possible to convert this to (k). Mathematically, this is completely isomorphic to H(k), which is the CSI of the Wi-Fi subcarrier.

[0077]

number

[0078] Spatial separation processing can also be applied to BLE CS measurements using the following procedure. (Procedure B1) Using Mode-2 or Mode-3 of BLE CS, set the CS channel set {k1, k2, ..., k M Complex channel response H in} BLE (k m ) obtain. (Procedure B2) If the CS channels are not equally spaced, resample them onto an equally spaced frequency grid by interpolation (e.g., linear interpolation, Lagrangian interpolation, interpolation with phase unwrapping, etc.). Alternatively, non-equally spaced handling techniques such as MUSIC or ESPRIT may be applied directly. (Procedure B3) Apply IFFT to the equally spaced frequency domain channel response and convert it to CIR. BLE Obtain (τ). (Step B4)h BLE The aforementioned time-domain gating process is performed on (τ). Here, d max Ya T s This can be calculated based on the BLE CS operating bandwidth. (Procedure B5) h after time-domain gating BLE The corrected channel response H is obtained by applying FFT to ^(τ) and extracting only the signal components within the monitored space. BLE Get ^(k).

[0079] Note that standard BLE CS selectively uses only a portion of the 79 channels, resulting in sparser period-domain sampling compared to Wi-Fi CSI. In this case, the aforementioned "(alternative method c) subspace estimation" or "(alternative method b) chirp Z transform" may be more suitable than IFFT. Also, while one-to-one measurement of initiator and reflector is typical in a BLE environment, it is possible to construct pseudo-multi-input multiple-output (pseudo-MIMO) measurement by using multiple reflectors, thereby improving spatial resolution. Furthermore, when using Mode-3, the rough distance obtained from RTT and the precise phase information obtained from the phase are combined to determine the threshold τ used in time-domain gating. th This can improve accuracy.

[0080] In this case, if both Wi-Fi and BLE CS measurements can be acquired in parallel using the same wireless device, spatial separation processing can be applied to each and the results of these processes can be integrated and output. In other words, when measurements can be acquired using multiple wireless standards that use different communication methods (specifically, communication methods that differ in at least one of the physical layer modulation method or the frequency band used), it is possible to apply spatial separation processing to each measurement individually and then integrate and output the results.

[0081] The integration of multiple measurements obtained from multiple wireless standards can be carried out using the following procedure. (Integration Procedure 1) H is the data series of the Wi-Fi measurement values. WiFi Apply the spatial selection process described above (Steps A1) to (Steps A3) to (k), and process the data series H WiFi Get ^(k). (Integration Procedure 2) The data series of the BLE CS measurements is H BLE Apply the spatial selection process described above (Steps B1) to (Steps B5) to (k), and process the data series H BLE Get ^(k). (Integration Procedure 3) H WiFi ^ and H BLE ^ and are combined to form the integrated measurement H fused Get the ^ symbol.

[0082] The integration described in (Integration Procedure 3) above can be carried out by any of the following methods or a combination thereof. (Integration Method a) The data is integrated by weighted synthesis, which involves adding the result of multiplying the processed Wi-Fi data sequence by a weight value α and the result of multiplying the processed BLE CS data sequence by a weight value β. Here, α + β = 1 and can be dynamically adjusted according to the signal quality (RSSI, CV, correlation, etc.). (Integration method b) The processed data sequence from Wi-Fi and the processed data sequence from BLE CS are integrated by alternating their placement at each sampling timing, thereby improving the effective sampling rate. (Integration method c) A logical AND operation is used to determine whether an operation event was detected simultaneously in both the Wi-Fi processed data sequence and the BLE CS processed data sequence. The method is only valid if the event is detected simultaneously in both, thereby reducing the false detection rate. (Integration method d) A two-input neural network (e.g., Siamese CNN, Cross-Attention Transformer, etc.) is used to learn and supplement the complementarity between standards and integrate them.

[0083] Since Wi-Fi (2.4GHz / 5GHz / 6GHz) and BLE (2.4GHz) partially overlap or are close in frequency, they are sensitive to the same physical multipath structure. Therefore, the outputs of spatial separation processing for both are correlated but contain independent noise components, and it can be expected that the SNR will improve through integrated processing. In particular, a synergistic effect can be obtained by complementing the narrowband characteristics of BLE with the wideband characteristics of Wi-Fi.

[0084] Next, we will describe baseline subtraction, an extended technique related to the present invention. The spatial separation process described above suppresses multipath components outside the monitoring space based on propagation delay. However, multipath components from static reflectors such as walls and static structures 16 (furniture, beds, wheelchairs, etc.) within the monitoring space cannot be suppressed by the spatial separation process because the propagation delay falls within the monitoring space. In particular, there are many large pieces of furniture in the rooms of nursing care facilities, and static reflections can be large enough to bury weak signals such as respiration detection.

[0085] Therefore, baseline subtraction is performed on the CIR data from which external noise components have been suppressed by the time-domain gating process described in (Procedure A2), by subtracting static environmental measurement values ​​(indoor static reflection components) that were previously acquired in an environment where no moving objects exist in the monitored space. Then, (Procedure A3) is performed on the measurement values ​​after baseline subtraction. In other words, the result of converting the measurement values ​​after baseline subtraction from a time-domain representation to a frequency-domain representation is output as processed CSI data. This makes it possible to extract only the dynamic components caused by the movement of moving objects in the monitored space. Physically, this is equivalent to constructing a "spatial model" of each room, which is the monitored space, in advance.

[0086] Static environmental measurement values ​​can be calculated using the following formula (13) after installing devices such as the transmitting station (radio 20A) and receiving station (radio 20B), acquiring M frames of CSI data in a state where no human body is present (or stationary) in the monitored space, converting it to CIR data, and then calculating the static environmental measurement value h baseline (τ) is calculated. Here, M is the number of calibration frames, and a value of approximately 100 to 500 is used.

[0087]

number

[0088] In baseline subtraction, the subtraction shown in equation (14) below is applied to each frame in real time. Here, h dynamic (τ,t) is the CIR fluctuation component caused solely by human body movement. Static reflections from walls, furniture, floors, etc. are h baseline It is included in (τ) and is canceled out by subtraction.

[0089]

number

[0090] By combining baseline subtraction with time-domain gating, a double level of noise reduction can be achieved. Specifically, by processing in the following order: (1) IFFT (CIR acquisition), (2) time-domain gating (external noise reduction), (3) baseline subtraction (indoor static reflection reduction), and (4) FFT (processed CSI acquisition), clean CSI data can be obtained with both external noise components and indoor static reflection components removed.

[0091] In actual operating environments, the static reflection component of the room changes gradually due to factors such as furniture movement and temperature changes. Therefore, adaptive updating using an exponential moving average (EMA), as shown in equation (15) below, may be introduced. Here, α∈(0,1) is the updating coefficient, and a value of approximately 0.001 to 0.01 is used. Adaptive updating is performed only in "quiet intervals" where the energy of the dynamic component is below a predetermined threshold (for example, a state where no human movement is detected).

[0092]

number

[0093] Alternatively, the drift value D(t) calculated by the following formula (16) is a predetermined threshold D th If the value exceeds a certain threshold, the system may be configured to automatically recalculate the static environmental measurement.

[0094]

number

[0095] Next, we will describe a measurement data processing system that integrates spatial separation processing and baseline reduction processing. The basic system configuration is the same as the operation determination system described using Figure 1, etc., and the main differences will be explained below. The measurement data processing system in this example not only integrates spatial separation processing and baseline reduction processing, but also automatically adjusts the spatial boundary (threshold) used in spatial separation processing using the static environmental measurement values ​​described in relation to baseline reduction processing.

[0096] As a prerequisite, the dimensional information of the monitoring space should be set when the system is introduced (for example, when the wireless device 20 is installed in the monitoring space 10). The dimensional information of the monitoring space represents the dimensions of the room that will be the monitoring space, and for example, the maximum distance of the monitoring space (specifically, the diagonal length of the room) can be used. The dimensional information of the monitoring space may be input by the user through direct operation of the wireless device 20 or the motion detection device 30, or it may be input via an application such as a smartphone that can access these devices. Alternatively, presets for different room types (6 tatami mats / 8 tatami mats / 12 tatami mats, etc.) may be prepared, and the system may be configured to use the dimensional information set in the preset for that room type as the dimensional information of the monitoring space when the user selects the room type for the monitoring space, or the system may be configured to accept direct input of the dimensions (numerical values) of the monitoring space from the user.

[0097] Once the dimensional information of the monitoring space is set, the reference threshold τ0 of the delay time corresponding to the boundary of the monitoring space can be calculated using the following equation (17). The unit of the reference threshold τ0 is the number of delay taps (index of the time domain CIR). Also, d max is the maximum distance [m] of the monitoring space, and d margin is a predefined boundary margin [m] (e.g., 0 to 2m) that takes into account the round-trip path of wall reflections, and T s is the sampling period [s] (=1 / effective bandwidth), and c is the speed of light (3 × 10⁻¹⁰). 8 The value is m / s. Equation (17) for calculating the reference threshold τ0 is the threshold τ mentioned above. th Based on formula (3) for calculating the boundary margin d, margin This is an expanded version with added features.

[0098]

number

[0099] Simply setting the dimensions of the monitoring space does not reflect the spread of multipath delay caused by the actual static reflective structure (furniture, wall materials, etc.) of the monitoring space. Therefore, static environmental measurement values ​​h baselineBased on the cumulative energy distribution of the CIR (acquired when no moving objects are present), the effective spatial boundary is estimated using the following equation (18). Here, η is the cumulative energy ratio threshold, which is a value in the range of 0.90 to 0.98, for example, with 0.95 used as the default value.

[0100]

number

[0101] Equation (18) above shows that the effective threshold τ is the delay tap at which the energy of static reflections due to furniture, wall materials, etc., in the monitored space falls within η of the total. env This is how it is calculated. Even if the dimensions of the monitored space set in advance are the same, the actual multipath distribution will differ depending on the furniture arrangement, wall material, etc., so the effective threshold τ is calculated using static environmental measurements. env This represents a spatial boundary that reflects the reality of the surveillance space.

[0102] Next, the reference threshold τ0 and the effective threshold τ env These are integrated by the following equation (19), and the integrated threshold τ int We calculate τ. floor This is the lower limit number of taps for direct wave protection, and for example, 1 to 2 is used.

[0103]

number

[0104] Equation (19) sets the reference threshold τ0 as the upper limit (i.e., anything outside the monitoring space is always removed), and the effective threshold τ env The integrated threshold τ is adjusted so that the boundary becomes narrower when the value is small (the boundary is narrowed in rooms with little furniture and short reflections). int This calculates the dimensional information of the monitored space (maximum distance d). max ) and the sampling frequency of the wireless signal (=1 / T s The reference threshold τ0 calculated using ) and static environmental measurement value (h baseline The effective threshold τ calculated using ) envThe smaller of the two is the integrated threshold τ int They will hire them.

[0105] In addition, the integrated threshold τ is calculated using equation (20) below, instead of equation (19). int It is also possible to calculate the following: Here, α∈[0,1] are the reference threshold τ0 and the effective threshold τ env These are coefficients that represent the weighting of the reference threshold τ0 and the effective threshold τ. env By weighting and summing these, the integrated threshold τ int Perform the calculation.

[0106]

number

[0107] The integrated threshold τ calculated by the above calculation int By using this method, filtering of signals from outside the monitoring space (external noise) can be performed in a manner that is adapted to the actual environmental conditions of the monitoring space. The reference threshold τ0 and the effective threshold τ env Integrated threshold τ based on int The calculation of is not limited to the calculation methods shown in equations (19) and (20), but can also be calculated using other calculation methods.

[0108] Furthermore, during the operation of this system, the integrated threshold τ in the current frame t int (t) is the estimated maximum delay τ calculated from environmental measurements in the current frame t. base Using (t), smoothing is continuously adjusted according to the following equation (21). Here, equation (21) is obtained by applying the above-mentioned equation (5) to the integrated threshold τ int This is an extension of the smoothing adjustment for (t). Estimated maximum delay τ base (t) is estimated sequentially in the same manner as in equation (18) based on the CIR energy distribution of the current frame group. f(·) is calculated by selectively using the conservative integration shown in equation (19), the weighted integration shown in equation (20), or other computational methods, with a reference threshold τ0 and an effective threshold τ env The estimated maximum delay τ corresponds to baseThis is a function that integrates with (t). λ∈(0,1) is a coefficient that determines the adaptation speed, and for example, a value in the range of 0.001 to 0.1 is used.

[0109]

Equation

[0110] Accordingly, the integration threshold τ int (t) can be adjusted to follow gradual environmental changes (changes in propagation characteristics due to temperature, humidity, etc.). The adjustment of the integration threshold τ int (t) can be performed every predetermined number of frames, that is, every 1 frame or every plurality of frames. In addition, rapid environmental changes (such as movement of furniture) are handled by the automatic reset described below. That is, the roles for gradual environmental changes and rapid environmental changes are divided based on the time constant of the change.

[0111] In the automatic reset, first, the drift value D(t) of the static reflection component in the current frame t is calculated by the following equation (22). Equation (22) is a redefinition of equation (16) used in baseline shift detection as a trigger for automatic re-determination of spatial boundaries. Here, h baseline is the M-frame average of the "quiet interval" in which the energy of the dynamic component is equal to or less than a predetermined threshold (for example, a state in which human body motion is not detected). Also, h quiet (t) is the CIR obtained when the current frame t is in the quiet interval.

[0112]

Equation

[0113] Thereafter, it is determined whether or not the drift value D(t) calculated by equation (22) exceeds a predetermined threshold D th . As a result, when the drift value D(t) exceeds the threshold D th , this is detected as a change in the static environment measurement value (h baseline ), h is recalibrated, and the effective threshold τ baseline is recalibrated, and the effective threshold τenv and the calculation of the integrated threshold τ int are re-performed. Accordingly, when an environmental change such as furniture movement or layout rearrangement occurs, the spatial boundary (threshold) can be automatically updated without manual intervention.

[0114] The integrated threshold τ calculated and updated as described above int can be used in the spatial separation process through the operation expressed by the following equation (23). Here, the integrated threshold τ int is used as the center value of the window function related to spatial separation. w(τ) is a sigmoid function, and κ is a boundary steepness parameter (for example, 1 to 10). Note that a Hanning window, a Gaussian window, a Tukey window, or the like may be applied instead of the sigmoid function.

[0115]

Equation

[0116] In addition, the integrated threshold τ int can also be used in the baseline subtraction process for the result of the spatial separation process (CIR data after external noise component removal), as expressed in the following equation (24). Here, h^(τ) is the CIR data after external noise component removal by the spatial separation process.

[0117]

Equation

[0118] Next, the flow of integrally performing the spatial separation process and the baseline subtraction process will be described. Figure 5 shows an example of a processing flow at the time of system introduction (calibration stage). At the time of system introduction, the motion determination device 30 first obtains the dimension information of the monitoring space (here, the maximum distance d of the monitoring space maxThe settings are accepted from the user (step S21). Based on the dimensional information of this monitoring space, the reference threshold τ0 is calculated by the calculation of equation (17). Then, based on the CIR obtained when no moving objects are present in the monitoring space, the static environmental measurement value h baseline Obtain (step S22). Then, static environmental measurement h baseline Based on this, the effective threshold τ is calculated by the operation of equation (18). env Calculate (step S23). Then, static environmental measurement h baseline and effective threshold τ env Based on this, the integrated threshold τ is calculated by the operation of equation (19) or equation (20). int The integrated threshold τ is calculated (step S24). int This is used to perform spatial separation processing and baseline reduction processing during system operation.

[0119] Figure 6 shows an example of the processing flow for real-time processing during system operation. During system operation, the operation determination device 30 acquires CSI data obtained with respect to the wireless signal propagating in the monitoring space at each time frame of a predetermined length (step S31). Next, it performs IFFT processing on the CSI data, which is a frequency domain representation, to convert it into CIR data, which is a time domain representation (step S32). Next, it performs spatial separation processing on the CIR data using the calculation of equation (23) to remove signal components with propagation delays exceeding a predetermined time threshold as external noise components (step S33). Next, it performs baseline subtraction processing on the CIR data after the removal of external noise components using the calculation of equation (24) to remove indoor static reflection components (step S34). Next, it performs FFT processing on the CIR data from which external noise components and indoor static reflection components have been removed to convert it into CSI data, which is a frequency domain representation (step S35). Next, it provides the CSI data from which external noise components have been removed to the operation determination process (step S36). Furthermore, at any time, such as at the start or end of each frame's processing (steps S31-36), the calculation of equation (21) is performed to obtain the integrated threshold τ int Adjust (t) to follow gradual environmental changes (changes in propagation characteristics due to temperature, humidity, etc.).

[0120] Figure 7 shows an example of the processing flow of background processing during system operation. During system operation, the motion determination device 30 calculates the drift value D(t) by calculation of equation (22) for each silent interval in which the energy of the dynamic component is below a predetermined threshold (step S41). Next, when the drift value D(t) is below the threshold D th It is determined whether or not the value exceeds the threshold (step S42). As a result, if the drift value D(t) exceeds the threshold D th If it is determined that the limit has been exceeded, step S22(h) described above will be executed. baseline (Calculation of), step S23 (effective threshold τ) env (Calculation of), step S24 (integrated threshold τ) int Each process of the calculation of is re-executed (step S43). This allows the integrated threshold τ, which is the spatial boundary, to be used when environmental changes such as moving furniture or rearranging occur. int (t) can be automatically updated.

[0121] As described above, the measurement value processing system in this example uses the processor 41 of the motion determination device 30 to process pre-set dimensional information regarding the monitoring space, the sampling frequency of the wireless signal, and static environmental measurement values ​​(h) based on measurements obtained in an environment where no moving objects are present in the monitoring space. baseline ) is used to determine the time threshold (integrated threshold τ) corresponding to the boundary of the monitoring space. int The system is configured to perform a threshold calculation process to determine the threshold value, and a spatial separation process to suppress signal components with propagation delays exceeding a time threshold among the signal components included in the measured values ​​obtained during system operation. Furthermore, if a change in static environmental measurement values ​​is detected during system operation, the threshold calculation process is re-executed.

[0122] This configuration makes it possible to suppress the influence of objects outside the monitoring space on measured values ​​representing the propagation path characteristics of wireless signals passing through the monitoring space, while also considering fluctuations in the static environment within the monitoring space. Furthermore, by performing a baseline subtraction process that subtracts static environment measurements from the measured values ​​after spatial separation processing, reflection components due to static reflective structures within the monitoring space can also be removed. As a result, subsequent processing of this system (such as target operation determination) can be performed with greater accuracy.

[0123] Although embodiments of the present invention have been described above, these embodiments are merely illustrative and do not limit the technical scope of the present invention. The present invention can take many other forms, and various modifications such as omissions and substitutions can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention as described herein, and are included in the scope of the invention as described in the claims or its equivalents.

[0124] Furthermore, the present invention can be provided not only as the devices described above or as systems composed of such devices, but also as methods executed by these devices, programs for a processor to realize the functions of these devices, and storage media for storing such programs in a computer-readable manner. [Industrial applicability]

[0125] This invention can be used in a motion determination system that determines the behavior of a target present in a monitored space. [Explanation of symbols]

[0126] 10: Surveillance space, 12, 14: Person, 16: Static structure, 20A, 20B: Wireless device, 30: Motion detection device, 41: Processor, 42: Memory, 43: Storage, 51: CSI acquisition function, 52: Spatial isolation function, 53: Motion detection function, 54: Notification function

Claims

1. A measurement value processing system that processes measurement values ​​representing propagation path characteristics of wireless signals propagating in a monitoring space, Equipped with one or more processors, The aforementioned processor, A threshold calculation process that calculates a time threshold corresponding to the boundary of the monitoring space using pre-set dimensional information regarding the monitoring space, the sampling frequency of the wireless signal, and static environmental measurement values ​​based on measurements obtained in an environment where no moving objects exist within the monitoring space; During system operation, a spatial separation process is performed to suppress signal components in the measured values ​​that have a propagation delay exceeding the time threshold. Furthermore, the measurement value processing system re-executes the threshold calculation process when a change in the static environmental measurement value is detected during system operation.

2. In the measurement value processing system according to claim 1, The threshold calculation process is a measurement value processing system that calculates a reference threshold that serves as the basis for the time threshold using the dimensional information and the sampling frequency, calculates an effective threshold corresponding to the effective boundary of the monitoring space using the static environmental measurement values, and calculates the time threshold by integrating the reference threshold and the effective threshold.

3. In the measurement data processing system according to claim 2, The threshold calculation process can selectively use one of several calculation methods that integrate the reference threshold and the effective threshold. A measurement value processing system in which the plurality of calculation methods include a method in which the smaller of the reference threshold and the effective threshold is used as the time threshold, and a method in which the result of weighted addition of the reference threshold and the effective threshold is used as the time threshold.

4. In the measurement value processing system according to claim 1, The threshold calculation process is a measurement value processing system that adjusts the time threshold using the estimated maximum delay calculated by environmental measurements in each predetermined number of frames.

5. In the measurement value processing system according to claim 1, The spatial separation process is a measurement value processing system that performs a calculation to suppress signal components having propagation delays exceeding the time threshold, using a window function with the time threshold as the center value.

6. In the measurement value processing system according to claim 1, A measurement processing system that has presets for each room type with pre-configured dimensional information, and in response to a room type selected by the user for the monitoring space, uses the dimensional information set in the preset for that room type as the pre-configured dimensional information for the monitoring space.

7. In the measurement data processing system according to any one of claims 1 to 6, The processor further performs a baseline subtraction process, which subtracts the static environmental measurement value from the measurement value obtained through the spatial separation process, in a measurement value processing system.

8. In the measurement data processing system according to any one of claims 1 to 6, A measurement value processing system comprising: a processor that individually performs the spatial separation process on a first measurement value obtained by a first wireless standard and a second measurement value obtained by a second wireless standard using a communication method different from the first wireless standard, and integrates the results of the spatial separation process on the first measurement value and the results of the spatial separation process on the second measurement value.

9. A method for processing measured values ​​that represent propagation path characteristics of wireless signals propagating in a monitoring space, using one or more processors, The steps include: performing a threshold calculation process to calculate a time threshold corresponding to the boundary of the monitoring space using pre-set dimensional information regarding the monitoring space, the sampling frequency of the wireless signal, and static environmental measurement values ​​based on measurements obtained in an environment where no moving objects exist within the monitoring space; The system includes the step of performing a spatial separation process to suppress signal components among the measured values ​​obtained during system operation that have a propagation delay exceeding the time threshold, A measurement value processing method that re-executes the threshold calculation process when a change in the static environmental measurement value is detected during system operation.

10. In the measurement data processing method described in claim 9, The threshold calculation process is a measurement value processing method which calculates a reference threshold that serves as the basis for the time threshold using the dimensional information and the sampling frequency, calculates an effective threshold corresponding to the effective boundary of the monitoring space using the static environmental measurement value, and calculates the time threshold by integrating the reference threshold and the effective threshold.

11. In the measurement data processing method according to claim 10, The threshold calculation process can selectively use one of several calculation methods that integrate the reference threshold and the effective threshold. The measurement processing method includes a method in which the smaller of the reference threshold and the effective threshold is used as the time threshold, and a method in which the result of weighted addition of the reference threshold and the effective threshold is used as the time threshold.

12. In the measurement data processing method described in claim 9, The threshold calculation process is a measurement value processing method that adjusts the time threshold using the estimated maximum delay calculated by environmental measurements in each predetermined number of frames.

13. In the measurement data processing method described in claim 9, The spatial separation process is a measurement value processing method that performs a calculation to suppress signal components having propagation delays exceeding the time threshold, using a window function with the time threshold as its center value.

14. In the measurement data processing method described in claim 9, A measurement processing method that has presets for each room type with pre-set dimension information, and in response to a room type selected by the user for the monitoring space, uses the dimension information set in the preset for that room type as the dimension information set in advance for the monitoring space.

15. In the measurement processing method according to any one of claims 9 to 14, A method for processing measurement values, further comprising the step of performing a baseline subtraction process in which a static environmental measurement value is subtracted from the measurement value obtained after the spatial separation process.

16. In the measurement processing method according to any one of claims 9 to 14, A measurement value processing method comprising: individually performing the spatial separation process on a first measurement value obtained by a first wireless standard and a second measurement value obtained by a second wireless standard using a communication method different from the first wireless standard; and integrating the results of the spatial separation process on the first measurement value and the results of the spatial separation process on the second measurement value.

Images