Information processing device and sensing method

Abstract

To provide an information processing device for sensing an object.SOLUTION: An information processing device senses an object using an FMCW radar, the information processing device comprising: a data processing unit that performs an angle FFT on a received signal to obtain an angular power spectrum relating to an angle with the object; an acquisition unit that acquires an angle peak bin indicating an angle with respect to the object based on the angular power spectrum; an extracting section that extracts an output signal corresponding to the power spectrum signal; and a correcting section that corrects the phase of the output signal according to bin numbers of a plurality of peak bins, and the data processing unit applies a window function of a higher order than a rectangular window to the received signal, and the correction unit corrects the phase of the output signal so that a phase difference to be added or subtracted becomes (N-1) / 2×π×sin(θ(i)) [rad] according to the bin number of the angle peak bin.SELECTED DRAWING: Figure 5A

Description

[Technical field]

[0001] The present invention relates to an information processing device and a sensing method. [Background technology]

[0002] 2. Description of the Related Art Conventionally, information processing devices that sense objects using a Doppler radar are known (see, for example, Patent Document 1). Patent Document 1 Patent No. 6029108 Summary of the Invention

[0003] In a first aspect of the present invention, there is provided an information processing device for sensing an object using an FMCW radar, the information processing device including: a data processing unit that processes received signals of a plurality of antennas based on received waves of the FMCW radar to generate a power spectrum signal of a predetermined number of bins, the data processing unit performing an angle FFT on the received signals to obtain an angle power spectrum relating to an angle with the object; an acquisition unit that acquires a plurality of peak bins corresponding to the object based on the power spectrum signal, the acquisition unit acquiring angle peak bins indicating an angle with the object based on the angle power spectrum; the data processing unit applies a window function of a higher order than a rectangular window to the received signal, and the correction unit corrects the phase of the output signal according to the bin number of the angle peak bin so that a phase difference to be added or subtracted becomes (N-1) / 2×π×sin(θ(i)) [rad] according to the bin number of the angle peak bin, where N is the number of the plurality of antennas, i is the bin number of the angle peak bin, and θ(i) [rad] is an angle corresponding to the bin number of the angle peak bin.

[0004] In the information processing device, the plurality of peak bins may include a peak bin having a predetermined first bin number and a peak bin having a second bin number different from the first bin number, and the correction unit may correct a phase of the peak bin having the second bin number without correcting a phase of the peak bin having the first bin number.

[0005] In any of the above information processing devices, the data processing unit may acquire a distance power spectrum relating to a distance to the object by performing a distance FFT on the received signal, and the acquisition unit may acquire a distance peak bin indicating a distance to the object based on the distance power spectrum.

[0006] In any of the information processing devices described above, the acquisition section may acquire the angle peak bin of the angle FFT based on a data string corresponding to a position of the distance peak bin specified by the distance FFT.

[0007] In any of the information processing devices described above, the correction section may correct a phase of the output signal in accordance with the bin number of the distance peak bin.

[0008] In any of the above information processing devices, the data processing unit may obtain the distance power spectrum by performing the distance FFT on the received signal to which a first window function has been applied, and obtain the angle power spectrum by performing the angle FFT on the received signal to which a second window function different from the first window function has been applied. The second window function may be a window function of a higher order than the first window function.

[0009] In any of the above information processing devices, the data processing unit may acquire a velocity power spectrum relating to the velocity of the object by performing a velocity FFT on the received signal, and the acquisition unit may acquire a velocity peak bin indicating the velocity of the object based on the velocity power spectrum.

[0010] In any of the information processing devices described above, the acquisition section may acquire the velocity peak bin of the velocity FFT based on a data string corresponding to a distance peak bin position identified by a distance FFT.

[0011] Any of the information processing devices described above may include a data conversion unit that converts a plurality of data corresponding to the object into clustered data.

[0012] Any of the information processing devices described above may include a tracking unit that tracks the object based on past data of the object when data corresponding to the object is not obtained within a predetermined period of time.

[0013] In any of the information processing devices described above, the extraction section may extract IQ data including an in-phase component and a quadrature component orthogonal to the in-phase component, from the peak bin acquired by the acquisition section.

[0014] In any of the information processing devices described above, the data processing unit may process the received signal using a CAPON method or a compressed sensing algorithm.

[0015] In a second aspect of the present invention, there is provided a method for sensing an object using an FMCW radar, comprising the steps of: processing received signals of a plurality of antennas based on received waves of the FMCW radar to generate a power spectrum signal having a predetermined number of bins, the step performing an angle FFT on the received signals to generate an angle power spectrum relating to an angle with the object; and acquiring a plurality of peak bins corresponding to the object based on the power spectrum signal, the step acquiring an angle peak bin indicating an angle with the object based on the angle power spectrum. The present invention provides a sensing method comprising the steps of: extracting an output signal corresponding to the power spectrum signal; and correcting a phase of the output signal according to bin numbers of the plurality of peak bins, wherein the step of generating the power spectrum signal includes a step of applying a window function of a higher order than a rectangular window to the received signal, and the step of correcting the phase of the output signal includes a step of correcting the phase of the output signal such that a phase difference to be added or subtracted becomes (N-1) / 2×π×sin(θ(i)) [rad] according to the bin number of the angle peak bin, where N is the number of the plurality of antennas, i is the bin number of the angle peak bin, and θ(i) [rad] is an angle corresponding to the bin number of the angle peak bin.

[0016] The above summary of the invention does not list all of the features of the present invention. Also, subcombinations of these features may also be inventions. [Brief description of the drawings]

[0017] [Figure 1A] 6 shows an overview of the configuration of a system 600. [Figure 1B] 2 shows an example of a transmission wave 220 transmitted by the transmitting unit 120. [Figure 1C] 1 is a diagram for explaining an example of signal processing by the information processing device 500. FIG. [Figure 1D] 2 is a diagram for explaining a distance R, a velocity V, and an angle θ of an object 210. FIG. [Figure 1E]2 is a diagram for explaining a distance R, a velocity V, an angle θ1, and an angle θ2 of an object 210. FIG. [Diagram 2] FIG. 6 is a diagram for explaining the operating principle of the system 600. [Figure 3A] 1 is a diagram for explaining the principle of detecting a distance R to an object 210. FIG. [Figure 3B] 1 is a diagram for explaining the principle of detecting the velocity V of an object 210. FIG. [Figure 3C] 1 is a diagram for explaining the principle of detecting an angle θ of an object 210. FIG. [Figure 4A] 1 is a diagram for explaining a method of sensing an object 210 with body movement. [Figure 4B] 2 shows an example of the effect of body movement of the object 210. [Figure 5A] 2 shows an example of the configuration of the signal processing unit 400. [Figure 5B] An example of the operation of the signal processing unit 400 will be described. [Figure 5C] 4 is a diagram for explaining the configuration of a data processing unit 410 and an acquisition unit 420 in more detail. [Figure 5D] 13 is an example of a phase correction algorithm of the correction unit 440. [Figure 5E] 13 is a modified example of the phase correction algorithm of the correction unit 440. [Figure 6A] 1 shows an example of the relationship between signal intensity and phase with respect to bin number. [Figure 6B] 1 shows an example of the relationship between signal intensity and phase with respect to bin number. [Figure 6C] 1 shows an example of the relationship between signal intensity and phase with respect to bin number. [Figure 6D] 1 shows an example of the relationship between signal intensity and phase with respect to bin number. [Figure 7] 1 shows a modified example of the signal processing unit 400. [Figure 8A] 1 shows a modified example of the signal processing unit 400. [Figure 8B] 4 is a diagram for explaining the configuration of a data processing unit 410 and an acquisition unit 420 in more detail. [Figure 9A]1 shows a modified example of the signal processing unit 400. [Figure 9B] An example of the operation of the signal processing unit 400 will be described. [Figure 9C] An example of data conversion by the data control unit 450 will be described. [Figure 10] 1 shows an example of a result of phase compensation performed by the information processing device 500. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] The present invention will be described below through embodiments of the invention, but the following embodiments do not limit the scope of the invention according to the claims. Furthermore, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention.

[0019] 1A shows an outline of the configuration of a system 600. The system 600 includes a transmitting / receiving unit 100 and an information processing device 500. The transmitting / receiving unit 100 includes a transmitting unit 120 and a receiving unit 140.

[0020] The transmitter 120 transmits a frequency modulated continuous wave radar (FMCW radar) to the target 210 as a transmission wave 220. The FMCW radar is a continuous oscillation radar with a modulated frequency. For example, the FMCW radar has a burst wave including multiple chirps. In each chirp, the frequency is swept in time. The transmitter 120 may have multiple transmission antennas.

[0021] The receiving unit 140 receives the reflected wave of the FMCW radar reflected by the object 210. The receiving unit 140 in this example demodulates the received wave 230 to generate an analog beat signal 150. The receiving unit 140 may have multiple receiving antennas. By providing multiple receiving antennas, information regarding the angle θ of the object 210 as viewed from the transmitting / receiving unit 100 can be obtained.

[0022] The beat signal 150 is an example of an IF signal down-converted to an IF (Intermediate Frequency) frequency proportional to the TOF (Time of Flight) of the reflected wave. The TOF is the time it takes for the transmitted transmission wave 220 to be received as a reflected wave, and increases as the distance R between the information processing device 500 and the target object 210 increases. The frequency of the beat signal 150 is proportional to the TOF, and therefore changes in proportion to the distance R.

[0023] The transmission / reception control unit 160 controls the transmission and reception of signals by the transmission unit 120 and the reception unit 140. In one example, the transmission / reception control unit 160 controls the modulation width and period of the chirp frequency of the transmission wave 220.

[0024] The information processing device 500 calculates the distance R and the velocity V of the object 210 by AD converting the beat signal 150 and processing the signal. The information processing device 500 of this example can detect minute vibrations of several mm by calculating the distance R in terms of phase. The information processing device 500 has an input unit 300 and a signal processing unit 400.

[0025] The input unit 300 converts the analog beat signal 150 input from the receiving unit 140 into a digital received signal. The received signal may be received signals from a plurality of antennas included in the receiving unit 140. The input unit 300 may be an ADC configured by an integrated circuit such as an RFIC.

[0026] The signal processing unit 400 is a digital signal processor (DSP) that executes signal processing such as FFT based on the digital received signal output by the input unit 300. The signal processing unit 400 detects the object 210 by processing the digital received signal. In this specification, detecting the object 210 refers to acquiring the distance R, speed V, angle θ, etc. of the object 210 and detecting the presence of the object 210. The distance R, speed V, and angle θ of the object 210 will be described later.

[0027] Moreover, the signal processing unit 400 senses the object 210 based on the micro-vibration data of the object 210. In this specification, sensing the object 210 refers to acquiring a biosignal such as micro-vibration data of the object 210. A biosignal exists when the object 210 is a living organism, and is generated by breathing, heartbeat, and the like.

[0028] The micro-vibration data is data based on the heartbeat and respiration of the target object 210. In one example, the information processing device 500 can obtain, as the micro-vibration data, a resolution with the wavelength of the FMCW radar as the maximum. For example, the information processing device 500 can obtain a resolution 100 to 1000 times that of one wavelength of the millimeter wave band (frequency band of about 30 to 300 GHz) used in the FMCW radar.

[0029] The system 600 senses the object 210 by transmitting an FMCW radar to the object 210. The system 600 can sense the object 210 even when the relative speed between the system 600 and the object 210 is zero by appropriately processing a received signal based on the modulated frequency of the FMCW radar.

[0030] FIG. 1B shows an example of a transmission wave 220 transmitted by the transmitting unit 120. The transmission wave 220 includes m chirps in one burst, where m is an integer equal to or greater than 2. The transmitting / receiving unit 100 modulates the frequency of the chirps and calculates the distance R, speed V, and angle θ of the object 210 by analyzing the difference between the transmission wave 220 and the reception wave 230. The transmitting / receiving unit 100 may appropriately adjust the modulation width or period of the chirp frequency depending on the position of the object 210, etc. The transmission wave 220 in this example includes m chirps of the same waveform, but may also include chirps of different waveforms.

[0031] FMCW radar is a radar that detects the distance to a target and the relative speed by utilizing the time difference between the return of echoes from the object 210. For example, FMCW radar linearly increases and decreases the frequency in a cycle of several microseconds to several hundred microseconds, and uses only either the up or down frequency for detection. However, in the FMCW method, both the up and down frequencies may be used for detection.

[0032] FMCW radar can detect angle information simultaneously by arranging multiple channels. For example, FMCW radar can achieve long-distance detection in the 76G band (76-77GHz) and medium- or short-distance detection in the 79G band (77-81GHz). Note that FMCW radar may be a type that linearly raises and lowers the frequency in a cycle of several milliseconds to several hundred milliseconds.

[0033] FIG. 1C is a diagram for explaining an example of signal processing by information processing device 500. As shown in FIG.

[0034] The frequency of the chirp of the transmitted wave 220 increases linearly from a low frequency during the period from time T0 to time Tm. The received wave 230 is received after a delay time Td depending on the distance R to the object 210. The delay time Td varies depending on the distance R to the object 210.

[0035] The beat signal 150 is generated in the transmitting / receiving unit 100. The beat signal 150 is generated from the difference between the chirp of the transmitted wave 220 and the chirp of the received wave 230. The frequency of the beat signal 150 is proportional to the magnitude of the delay time Td.

[0036] The spectrum data is calculated by the signal processing unit 400 using the signal between time Td and time Tm when the frequency of the beat signal 150 is relatively stable. A specific method of calculating the spectrum data will be described later.

[0037] 1D is a diagram for explaining the distance R, the velocity V, and the angle θ of the object 210. In this example, for simplicity, the transmitting unit 120 and the receiving unit 140 are considered to be in the same position.

[0038] The object 210 moves at a speed V at a position at a distance R from the transceiver unit 100. The speed V is the relative speed between the transceiver unit 100 and the object 210. The angle θ is the angle of the object 210 as viewed from the transceiver unit 100. Specifically, if the direction in which the receivers 140 are arranged is the X-axis direction and the direction in which the FMCW radar is emitted is the Y-axis direction, then the angle θ is the angle between the Y-axis and the position of the object 210 on the XY plane.

[0039] FIG. 1E is a diagram for explaining the distance R, speed V, angle θ1, and angle θ2 of the object 210. The information processing device 500 can detect and sense the object 210 using a similar principle as a so-called 3D radar that detects a new axis (Z axis) perpendicular to the XY plane. In this case, the information processing device 500 acquires three-dimensional information by using the angle θ2 projected onto the YZ plane in addition to the angle θ1 projected onto the XY plane. In this specification, when simply referring to the angle θ, it may be interpreted as including both the angle θ1 and the angle θ2.

[0040] 2 is a diagram for explaining the operation principle of the system 600. The system 600 uses a data cube 38 to obtain information regarding the distance R, the velocity V, and the angle θ of the object 210.

[0041] The transceiver 100 has a plurality of channels. In one example, the transceiver 100 has one transmitting antenna and k receiving antennas, where k is an integer equal to or greater than 1. The transceiver 100 has a plurality of channels, and is therefore capable of detecting the angle θ. Reflected signals from the object 210a and the object 210b are input to the k receiving antennas, respectively.

[0042] The data cube 38 includes data strings relating to the distance R, the velocity V, and the angle θ. The data cube 38 includes a distance data string obtained by a distance FFT, a velocity data string obtained by a velocity FFT, and an angle data string obtained by an angle FFT.

[0043] A distance power spectrum with n / 2 bins is obtained by power transforming the distance data sequence. The distance power spectrum includes two distance peak bins corresponding to the distances to the object 210a and the object 210b. The distance peak bin may be a bin that has a peak corresponding to the distance to the object 210 in the distance power spectrum.

[0044] A velocity power spectrum with m bins is obtained by performing a velocity FFT on the data sequence corresponding to the distance peak bin position of the distance power spectrum and power converting the newly obtained velocity data sequence. The velocity power spectrum includes a velocity peak bin corresponding to the velocity of the object 210a or the object 210b. The velocity peak bin may be a bin in the velocity power spectrum having a peak corresponding to the velocity of the object 210. Which living organism the velocity corresponds to depends on the distance peak bin position of the selected distance power spectrum.

[0045] An angle power spectrum with k bins is obtained by performing an angle FFT on the data sequence corresponding to the distance peak bin position of the distance power spectrum and power converting the newly obtained angle data sequence. The angle power spectrum includes an angle peak bin corresponding to the angle of the object 210a or the object 210b. The angle peak bin may be a bin in the angle power spectrum having a peak corresponding to the angle of the object 210. Which living body angle it corresponds to depends on the distance peak bin position of the selected distance power spectrum.

[0046] The information processing device 500 calculates, in a time series, phase information obtained by performing phase conversion on the angle data sequence of the object 210 obtained by the angle FFT, and acquires biosignal data of the object 210. Note that the bin position of the phase information used as the biosignal information corresponds to the angle peak bin position of the angle power spectrum. The biosignal data is an example of an output signal of the information processing device 500.

[0047] 3A is a diagram for explaining the principle of detecting the distance R to the object 210. The distance R to the object 210 is calculated by the distance FFT of at least one chirp. The solid line in the graph indicates the transmitted wave 220, and the dashed line indicates the received wave 230. The vertical axis indicates frequency, and the horizontal axis indicates time.

[0048] The IF frequency is the frequency of the beat signal 150 obtained by mixing the transmission wave 220 and the reception wave 230 of the FMCW radar. The IF frequency becomes higher as the distance R between the information processing device 500 and the target 210 increases. The information processing device 500 can obtain an IF frequency proportional to the distance R to the target 210 for a certain period of time. That is, the information processing device 500 can obtain the distance R to the target 210 by analyzing the IF frequency.

[0049] The information processing device 500 executes distance FFT processing for each chirp. For example, when the number of FFT points is n, n / 2 points of real and imaginary data strings are obtained, and the number of bins is n / 2. The information processing device 500 calculates a distance peak bin by performing power conversion based on the result of the distance FFT. The information processing device 500 can calculate the distance R to the object 210 based on the frequency at which the peak appears.

[0050] FIG. 3B is a diagram for explaining the principle of detecting the velocity V of the object 210. For each data sequence obtained by the distance FFT for multiple chirps in a burst, a velocity FFT is performed on a new data sequence corresponding to the distance peak bin position of the distance power spectrum. One burst includes m chirps. The velocity V of the object 210 is calculated by performing a power conversion based on the result of the velocity FFT.

[0051] The information processing device 500 executes velocity FFT processing for one burst. For example, when the number of chirps in one burst is m, m points of real and imaginary data sequences are obtained, and the number of bins is m. The information processing device 500 calculates the velocity peak bin position by performing power conversion based on the result of the velocity FFT. The information processing device 500 can calculate the velocity V from the frequency of the velocity peak bin of the velocity FFT.

[0052] 3C is a diagram for explaining the principle of detecting the angle θ of the object 210. For each data sequence obtained by the distance FFT for the chirp of each channel, the angle θ is obtained by performing an angle FFT on a new data sequence corresponding to the distance peak bin position of the distance power spectrum. One chirp may include k chirps corresponding to k channels, or may include one chirp overall. As an example, when MIMO (Multi-Input and Multi-Output) technology is used, k chirps corresponding to k channels are included.

[0053] The information processing device 500 executes angle FFT processing for k channels. For example, in the case of k channels, k points of data strings of real parts and imaginary parts are obtained, and the number of bins is k. The information processing device 500 receives the received waves 230 by the receiving units 140 of k channels arranged in the X-axis direction. Since a phase difference corresponding to the angle θ of the object 210 occurs in the received waves 230 received by each receiving unit 140, the information processing device 500 can calculate the angle θ of the object 210 by analyzing the received signals of the k channels.

[0054] 4A is a diagram for explaining a method of sensing an object 210 with body movement. The body movement of the object 210 is a large movement compared to a biological signal such as a heartbeat. When the object 210 moves, the position of the angle peak bin of the angle power spectrum may move. When the object 210 is not in the same bin and the bin is changed, a phase discontinuity (i.e., a phase shift) may occur.

[0055] The information processing device 500 can obtain accurate angle information of the object 210 by tracking the angle peak bin of the angle FFT of the object 210 as a basic principle of biological sensing by FMCW radar. On the other hand, even if the angle peak bin is changed, if only the same bin is tracked, the accuracy of the phase information obtained may deteriorate compared to the angle peak bin. The information processing device 500 of this example can suppress such deterioration of the accuracy of the phase information by compensating for the phase shift.

[0056] FIG. 4B shows an example of the effect of motion of the object 210. To extract such motion, phase is acquired at multiple points of the chirp. However, when the angle peak bin from which the phase value is acquired changes, the phase signal becomes discontinuous and a phase shift of approximately π [rad] or less may occur. The value of the phase shift depends on the number of bins to be transitioned. In this example, several phase shifts occur when the angle peak bin changes between bin numbers 9 and 17 in the period from 0 seconds to 10 seconds.

[0057] 5A shows an example of the configuration of the signal processing unit 400. The signal processing unit 400 includes a data processing unit 410, an acquisition unit 420, an extraction unit 430, and a correction unit 440.

[0058] The data processing unit 410 processes the received signals of the multiple antennas based on the received wave 230 to generate a power spectrum signal with a predetermined number of bins. The data processing unit 410 of this example performs at least one of a distance FFT, a velocity FFT, or an angle FFT to generate a power spectrum signal. As an example, the data processing unit 410 performs an angle FFT on the received signal to obtain an angle power spectrum related to the angle with the object 210. As another example, the data processing unit 410 may perform a distance FFT on the received signal to obtain a distance power spectrum related to the distance with the object 210, and may perform a velocity FFT on the received signal to obtain a velocity power spectrum related to the velocity with the object 210. The data processing unit 410 may process the received signal using other algorithms such as the CAPON method or compressed sensing.

[0059] The acquiring unit 420 acquires a peak bin corresponding to the object 210 based on the power spectrum signal output by the data processing unit 410. When the bin number of the peak bin to which the object 210 belongs is changed, the acquiring unit 420 may change the peak bin to be acquired. As an example, the acquiring unit 420 acquires an angle peak bin indicating the angle with the object 210 based on the angle power spectrum. As another example, the acquiring unit 420 may acquire a distance peak bin indicating the distance with the object 210 based on the distance power spectrum, and may acquire a speed peak bin indicating the speed with the object 210 based on the speed power spectrum.

[0060] The extraction unit 430 extracts a predetermined output signal from the power spectrum signal. The extraction unit 430 of this example extracts IQ data including an in-phase component and a quadrature component orthogonal to the in-phase component based on the peak bin acquired by the acquisition unit 420. As an example, the extraction unit 430 may extract an output signal corresponding to an angle power spectrum. As another example, the extraction unit 430 may extract an output signal corresponding to a distance power spectrum, or an output signal corresponding to a velocity power spectrum.

[0061] The correction unit 440 executes a phase correction algorithm based on the IQ data extracted by the extraction unit 430 and the detection result of the acquisition unit 420. The correction unit 440 corrects the phase of the output signal according to the bin numbers of the multiple peak bins. This allows the correction unit 440 to suppress the influence of a phase shift that occurs when the bin number to which the object 210 belongs changes according to a time change. The specific operation of the correction unit 440 will be described later.

[0062] 5B shows an example of the operation of the signal processing unit 400. The signal processing unit 400 receives a digital reception signal that has been AD converted by the input unit 300.

[0063] The data processing unit 410 executes distance FFT processing. The data processing unit 410 executes window function processing and distance FFT on the received signal to obtain a distance power spectrum. The data processing unit 410 also executes angle FFT. The data processing unit 410 executes window function processing and angle FFT on the received signal to obtain an angle power spectrum. The data processing unit 410 outputs the generated distance power spectrum and angle power spectrum to the acquisition unit 420.

[0064] The acquisition unit 420 acquires a distance peak bin indicating the distance R to the object 210 based on the distance power spectrum. The acquisition unit 420 may acquire the peak position by a simple peak detection algorithm, or may acquire the peak position by using another algorithm such as a CFAR algorithm. The acquisition unit 420 outputs the acquired information regarding the distance R to the object 210 to the extraction unit 430. The acquisition unit 420 may also output the acquired information regarding the distance R to the object 210 to the correction unit 440.

[0065] The acquiring unit 420 acquires an angle peak bin indicating the angle θ with the object 210 based on the angle power spectrum. The acquiring unit 420 in this example acquires the angle peak bin of the angle FFT based on a data string corresponding to the distance peak bin position specified by the distance FFT. This makes it possible to acquire the angle θ of any object 210 selected according to the distance R.

[0066] The extraction unit 430 acquires IQ data corresponding to the peak bin acquired by the acquisition unit 420 from the results of the distance FFT and angle FFT by the data processing unit 410. The extraction unit 430 may directly extract the IQ data from the power spectrum signal, or may use other algorithms such as compressed sensing. The extraction unit 430 outputs the extraction result of the IQ data to the correction unit 440.

[0067] The correction unit 440 corrects the phase of the output signal of the data processing unit 410. The correction unit 440 of this example corrects the phase of the output signal based on the angle acquisition result after the angle FFT. For example, the correction unit 440 corrects the phase of the angle peak bin of the angle power spectrum according to the bin number of the angle peak bin acquired by the angle FFT.

[0068] 5C is a diagram for explaining a more detailed configuration of data processing unit 410 and acquisition unit 420. Data processing unit 410 has a window function execution unit 411, a distance FFT execution unit 412, a window function execution unit 413, and an angle FFT execution unit 414.

[0069] The window function execution unit 411 executes a predetermined window function on the digital received signal. The window function execution unit 411 may use any window function such as a rectangular window, a Hann window, a Hamming window, or a Blackman window. The window function execution unit 411 may extract the received signal using a window function of a higher order than the rectangular window. The Hann window, the Hamming window, and the Blackman window are examples of window functions of a higher order than the rectangular window.

[0070] In this example, after the window function execution unit 411 extracts the digital received signal, the distance FFT execution unit 412 performs distance FFT to convert it into a distance power spectrum in the frequency domain. However, the order of FFT processing and window function processing is not limited to this. The window function execution unit 411 may extract the signal after the distance FFT execution unit 412 performs distance FFT to convert it into a distance power spectrum in the frequency domain. Other algorithms such as the CAPON method or compressed sensing may be used for the spectrum conversion.

[0071] The window function execution unit 413 applies a predetermined window function to the digital received signal. The window function execution unit 413 may use any window function such as a rectangular window, a Hann window, a Hamming window, or a Blackman window. The window function execution unit 413 may extract the received signal using a window function of a higher order than the rectangular window. The Hann window, the Hamming window, and the Blackman window are examples of window functions of a higher order than the rectangular window.

[0072] In this example, after the window function execution unit 413 extracts the digital received signal, the angle FFT execution unit 414 performs angle FFT to convert it into an angle power spectrum in the frequency domain. However, the order of FFT processing and window function processing is not limited to this. The window function execution unit 413 may extract the signal after the angle FFT execution unit 414 performs angle FFT to convert it into an angle power spectrum in the frequency domain. Other algorithms such as the CAPON method or compressed sensing may be used for the spectrum conversion.

[0073] The window function executing unit 411 and the window function executing unit 413 may use window functions of different orders. That is, the data processing unit 410 may acquire a distance power spectrum by performing a distance FFT on a received signal to which a first window function is applied, and may acquire an angle power spectrum by performing an angle FFT on a received signal to which a second window function different from the first window function is applied. In this case, the second window function may be a window function of a higher order than the first window function. However, the orders of the first window function and the second window function may be the same, and the second window function may be a window function of a lower order than the first window function.

[0074] When the order of the window function applied to the received signal for performing the distance FFT is high, there is a risk that the phase signals of different objects 210 at the same distance may be modulated by each other. On the other hand, when the order of the window function applied to the received signal for performing the angle FFT is high, the number of phase shifts π described later can be reduced, and the phase error of the output signal can be reduced. Therefore, by making the second window function a window function of a higher order than the first window function, it is possible to reduce the phase error of the output signal while suppressing the modulation of the phase signals of different objects 210 at the same distance.

[0075] The acquisition unit 420 detects a peak position indicating the distance R or angle θ of the object 210 from the frequency domain power spectrum signal output by the distance FFT execution unit 412 or the angle FFT execution unit 414 using a predetermined detection algorithm. The acquisition unit 420 may acquire the peak position using a simple peak detection algorithm, or may acquire the peak position using other algorithms such as a CFAR algorithm. The acquisition unit 420 outputs peak bins corresponding to the acquired distance R and / or angle θ of the object 210 to the extraction unit 430 and the correction unit 440. The acquisition unit 420 in this example outputs the distance peak bin and the angle peak bin to the extraction unit 430, and outputs the angle peak bin to the correction unit 440.

[0076] 5D shows an example of a phase correction algorithm of the correction unit 440. The correction unit 440 includes a phase selection unit 441 and a rotation unit 442.

[0077] The phase selection unit 441 selects a phase having a magnitude according to (N-1) / 2×π×sin(θ(i)) [rad] based on the bin number of the angle peak bin of the angle power spectrum, where N is the number of antennas, i is the bin number of the angle peak bin, and θ(i) [rad] is the angle corresponding to the bin number i of the angle peak bin.

[0078] The rotation unit 442 rotates the phase of the IQ data by the phase selected by the phase selection unit 441. This enables the correction unit 440 to correct the phase of the angle peak bin in accordance with a change in the bin number of the angle peak bin.

[0079] 5E shows a modified example of the phase correction algorithm of the correction unit 440. The correction unit 440 has a phase selection unit 441, a phase conversion unit 443, and an addition / subtraction unit 444. The correction unit 440 of this example corrects the phase of the angle peak bin so that the phase difference to be added or subtracted becomes (N-1) / 2×π×sin(θ(i)) [rad] according to the bin number of the angle peak bin. N is the number of multiple antennas, i is the bin number of the angle peak bin, and θ(i) [rad] is the angle corresponding to the bin number i of the angle peak bin.

[0080] The phase conversion unit 443 converts the IQ data into phase data. In this example, the phase conversion unit 443 converts the IQ data into tan -1 The phase data is converted by calculating (Q / I).

[0081] The adding / subtracting unit 444 adds or subtracts a phase according to (N-1) / 2×π×sin(θ(i)) [rad] to the phase data converted by the phase conversion unit 443 based on the bin number acquired by the acquiring unit 420.

[0082] In addition, the data processing unit 410 may apply a window function of a higher order than a rectangular window to the received signal in order to remove the phase shift of approximately π. The relationship between the higher order window function and the removal of the phase shift of approximately π will be described later.

[0083] As described above, the information processing device 500 of this example corrects the phase of the output signal according to the bin number of the angle peak bin. The number of data used in the angle FFT depends on the number of antennas, and may be less than the number of data used in the distance FFT. For this reason, a zero-padding FFT or other complex transformation algorithm may be required to correct the phase of the output signal. The information processing device 500 of this example can correct the phase of the output signal without using a complex algorithm by correcting the phase of the output signal according to the bin number of the angle peak bin.

[0084] FIG. 6A shows an example of the relationship between signal strength and phase with respect to the bin number. The data processing unit 410 in this example uses a rectangular window as a window function. The phase error of the output signal follows a sine function according to the angle θ of the object 210. In addition, the phase error of the output signal may have a phase shift π according to the bin number of the angle peak bin. The number of phase shifts π depends on the number N of multiple antennas used. The total number of phase shifts π when a rectangular window is used is N-1. The phase error according to the sine function depends only on the angle θ of the object 210 and does not depend on the distance R of the object 210. On the other hand, the angle θ at which the phase shift π occurs depends on the distance R of the object 210.

[0085] 6B shows an example of the relationship between signal strength and phase with respect to the bin number. In this example, the data processing unit 410 uses a Hann window as a window function. When a Hann window is used as a window function, the phase shift π is shifted further away from the actual target position compared to when a rectangular window is used.

[0086] FIG. 6C shows an example of the relationship between signal strength and phase with respect to the bin number. The data processing unit 410 in this example uses a Hamming window as a window function. When a Hamming window is used as a window function, the phase shift π is shifted further away from the actual target position compared to when a Hann window is used. In addition, the number of phase shifts π is reduced by two.

[0087] FIG. 6D shows an example of the relationship between signal intensity and phase with respect to the bin number. The data processing unit 410 in this example uses a Blackman window as a window function. When a Blackman window is used as a window function, the phase shift π is shifted further away from the actual target position compared to when a Hamming window is used. In addition, the number of phase shifts π is reduced by two.

[0088] As described above, window functions other than the rectangular window shift the phase by π away from the actual target position more than the rectangular window does. Also, the higher the order of the window function, the less phase shift occurs.

[0089] In addition, by using a window function of a higher order than a rectangular window, the information processing device 500 can easily identify the position of the bin in which the object 210 exists even if the object 210 moves. When multiple objects 210 exist, an optimal window function may be selected as appropriate according to the distance between the objects.

[0090] FIG. 7 shows a modified example of the signal processing unit 400. In this example, differences from the embodiment of FIG. 5B will be particularly described. Other points may be the same as in FIG. 5B. The acquisition unit 420 of this example also outputs information on the acquired distance R from the object 210 to the correction unit 440. The correction unit 440 of this example may perform both correction based on the distance acquisition result and correction based on the angle acquisition result. The explanation of the correction based on the angle acquisition result is omitted in this example, as it is the same as that described in relation to FIG. 5B.

[0091] The correction unit 440 of this example corrects the phase of the output signal based on the distance acquisition result after the distance FFT. For example, the correction unit 440 corrects the phase of the distance peak bin of the distance power spectrum according to the bin number of the distance peak bin acquired by the distance FFT. As an example, the correction unit 440 may correct the phase of the output signal so that the difference between the phase to be added or subtracted to the odd bin number and the phase to be added or subtracted to the even bin number becomes (2×j+1)π[rad]. j may be any integer.

[0092] Fig. 8A shows a modified example of the signal processing unit 400. In this example, differences from the embodiment in Fig. 5B will be particularly described. Other points may be the same as in Fig. 5B. The data processing unit 410 in this example performs a velocity FFT in addition to a distance FFT and an angle FFT.

[0093] The data processing unit 410 acquires a velocity power spectrum by performing a velocity FFT on the received signal. The data processing unit 410 also acquires an angle power spectrum by performing an angle FFT on the received signal. After performing the distance FFT, the data processing unit 410 of this example performs a velocity FFT and an angle FFT based on a data string corresponding to the distance peak bin position identified by the distance FFT.

[0094] The acquiring unit 420 acquires a peak bin related to the velocity V of the object 210 in addition to the distance R and angle θ. The acquiring unit 420 acquires a velocity peak bin indicating the velocity V of the object 210 based on the velocity power spectrum. The acquiring unit 420 in this example acquires the velocity peak bin of the velocity FFT based on a data string corresponding to the distance peak bin position specified by the distance FFT. This makes it possible to acquire the velocity V of any object 210 selected according to the distance R.

[0095] The extraction unit 430 extracts each IQ data based on a peak bin indicating the distance R, velocity V, and angle θ of the object 210. The extraction unit 430 in this example extracts IQ data from the velocity V and angle θ of the bin corresponding to the distance peak bin of the distance R of the object 210. The extraction unit 430 may extract the IQ data directly, or may extract the IQ data using another algorithm such as compressed sensing.

[0096] The correction unit 440 executes a phase correction algorithm based on the IQ data extracted from at least one of the distance R, the velocity V, and the angle θ extracted by the extraction unit 430. For example, the correction unit 440 corrects the phase of each of the output signals of the distance R, the velocity V, and the angle θ based on the IQ data extracted from the distance R, the velocity V, and the angle θ of the object 210.

[0097] The information processing device 500 of this example can output output signals with the phase corrected for each of the distance R, the speed V, and the angle θ. The information processing device 500 may correct only the phase of the output signal for any one of the distance R, the speed V, or the angle θ.

[0098] 8B is a diagram for explaining a more detailed configuration of data processing unit 410 and acquisition unit 420. In this example, processing related to velocity FFT will be explained. Data processing unit 410 in this example has window function execution unit 415 and velocity FFT execution unit 416.

[0099] The window function execution unit 415 executes a predetermined window function to extract the input signal. The velocity FFT execution unit 416 performs velocity FFT processing on the signal extracted by the window function execution unit 415, thereby converting it into a velocity power spectrum in the frequency domain.

[0100] In this example, after the window function execution unit 415 extracts the digital received signal, the velocity FFT execution unit 416 performs velocity FFT and converts it into a velocity power spectrum in the frequency domain. However, the order of FFT processing and window function processing is not limited to this. After the velocity FFT execution unit 416 converts it into a velocity power spectrum, the window function execution unit 415 may extract the signal. For the spectrum conversion, the CAPON method or other algorithms such as compressed sensing may be used.

[0101] The acquisition unit 420 detects a peak position indicating the velocity V of the object 210 from the frequency domain power spectrum signal output by the velocity FFT execution unit 416 using a predetermined detection algorithm. The acquisition unit 420 may acquire the peak position using a simple peak detection algorithm, or may acquire the peak position using another algorithm such as a CFAR algorithm. The acquisition unit 420 outputs a velocity peak bin corresponding to the acquired velocity V of the object 210 to the extraction unit 430 and the correction unit 440.

[0102] 9A shows a modified example of the signal processing unit 400. In this example, differences from the embodiment of FIG. 8A will be particularly described. Other points may be the same as those in FIG. 8A. The signal processing unit 400 of this example has a data control unit 450. The data control unit 450 has a data conversion unit 452 and a tracking unit 454.

[0103] The data conversion unit 452 converts a plurality of pieces of data corresponding to the object 210 acquired by the acquisition unit 420 into clustered data. For example, when a plurality of detection points that may belong to the same object 210 occur, the data conversion unit 452 replaces the plurality of detection points corresponding to the object 210 with one detection point. This enables the information processing device 500 to simplify processing by clustering groups corresponding to the same object 210 into one detection point.

[0104] When data corresponding to the object 210 is not obtained within a predetermined period, the tracking unit 454 tracks the object 210 based on past data of the object 210. Even when the object 210 cannot be detected by processing within a predetermined time and data loss occurs in detecting the vibration of the object 210, the tracking unit 454 can complement the data of the object 210 by using a tracking algorithm. This allows the information processing device 500 to predict the position of the object 210 even when there is no measurement data within a predetermined time.

[0105] In this example, the data control unit 450 outputs the clustered data to the extraction unit 430 and the correction unit 440, which can simplify the processing in the extraction unit 430 and the correction unit 440. In addition, the data control unit 450 outputs the tracked data to the extraction unit 430 and the correction unit 440, which can prevent data loss in the extraction unit 430 and the correction unit 440.

[0106] 9B shows an example of the operation of the signal processing unit 400. The data control unit 450 may perform a clustering process or a tracking process on the detection points corresponding to at least one of the distance R, the velocity V, or the angle θ of the object 210. The data control unit 450 outputs the processed data to the extraction unit 430 and the correction unit 440, respectively.

[0107] 9C shows an example of data conversion by data control unit 450. The data conversion method in this example is merely an example and is not limited to this.

[0108] In step S100, the data control unit 450 clusters the acquired data acquired by the acquisition unit 420. For example, the acquired data is clustered into clusters according to the distribution of the acquired data. In this example, the data control unit 450 clusters the acquired data into three clusters. The data control unit 450 sets representative values ​​d1 to d3 representing the respective clusters. The representative values ​​may be values ​​near the center of the cluster distribution.

[0109] In step S102, the calculated representative value is associated with the past tracked representative values ​​t1-t3 to form a pair of the new representative value and the past representative value. In this example, the representative value d2 is associated with the past representative value t1, and the representative value d3 is associated with the past representative value t2.

[0110] In step S104, the data control unit 450 manages the target to link the correct ID to the associated data.

[0111] Since there is no corresponding representative value for the representative value d1, a new ID is assigned to the new object 210. Since there are no tracked past representative values ​​t1 and t2 for the representative values ​​d2 and d3, respectively, they are associated with the ID corresponding to the object 210. Since there is no corresponding representative value for the current representative value for the past representative value t3, it is removed.

[0112] In step S106, the representative values ​​t1, t2, and t4 are newly tracked. The representative values ​​t1 and t2 continue to exist as tracked representative values. The representative value t4 is tracked as a new representative value based on the representative value d1. The tracked representative values ​​t1, t2, and t4 may be used for target association in step S102.

[0113] Depending on the detection algorithm or tracking algorithm, adjacent bins may be selected. Even in such a case where adjacent bins are selected and a phase shift occurs, the information processing device 500 can compensate for the phase shift of the output signal.

[0114] 10 shows an example of a phase compensation result by the information processing device 500. By compensating the phase, the information processing device 500 can remove phase discontinuity compared to the case without phase compensation. The correction unit 440 of this example corrects the phase of some angle peak bins when the angle peak bin changes between bin numbers 9 to 17. Note that the bin numbers to be subjected to phase correction are not limited to this.

[0115] The information processing device 500 can compensate for the phase shift accompanying the movement of the object 210 and sense the object 210 more accurately over time. The information processing device 500 can detect small body movements of the object 210 in a non-contact manner and can be used in various fields. The information processing device 500 can be applied to the medical field to detect biological signals such as heartbeat or respiration. In addition, the information processing device 500 may be used to detect defects by sensing vibrations of structures such as buildings or bridges, or may be used to sense vibrations of motors, etc.

[0116] Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It is clear to those skilled in the art that various modifications and improvements can be made to the above embodiments. It is clear from the description of the claims that such modifications and improvements can also be included in the technical scope of the present invention.

[0117] It should be noted that the order of execution of each process, such as operations, procedures, steps, and stages, in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not specifically stated as "before," "prior to," etc., and may be realized in any order unless the output of a previous process is used in a later process. Even if the operational flow in the claims, specifications, and drawings is explained using "first," "next," etc. for convenience, it does not mean that it is essential to perform the process in this order. [Explanation of symbols]

[0118] 38... data cube, 100... transmitting / receiving unit, 120... transmitting unit, 140... receiving unit, 150... beat signal, 160... transmitting / receiving control unit, 210... object, 220... transmitting wave, 230... receiving wave, 300... input unit, 400... signal processing unit, 410... data processing unit, 411... window function execution unit, 412... distance FFT execution unit, 413... window function execution unit, 4 14... angle FFT execution unit, 415... window function execution unit, 416... velocity FFT execution unit, 420... acquisition unit, 430... extraction unit, 440... correction unit, 441... phase selection unit, 442... rotation unit, 443... phase conversion unit, 444... addition / subtraction unit, 450... data control unit, 452... data conversion unit, 454... tracking unit, 500... information processing device, 600... system

Claims

1. An information processing device that senses an object using an FMCW radar, a data processing unit that processes received signals of a plurality of antennas based on received waves of the FMCW radar to generate a power spectrum signal of a predetermined number of bins, and performs an angle FFT on the received signals to obtain an angle power spectrum relating to an angle with the object; an acquisition unit that acquires a plurality of peak bins corresponding to the object based on the power spectrum signal, the acquisition unit acquiring an angle peak bin indicating an angle with the object based on the angle power spectrum; an extraction unit that extracts an output signal corresponding to the power spectrum signal; a correction unit that corrects a phase of the output signal in accordance with bin numbers of the plurality of peak bins; Equipped with The data processing unit applies a window function having a higher order than a rectangular window to the received signal; the correction unit corrects the phase of the output signal so that a phase difference to be added or subtracted becomes (N-1) / 2 x π x sin(θ(i)) [rad] in accordance with the bin number of the angle peak bin; N is the number of the plurality of antennas, i is the bin number of the angle peak bin, θ(i) [rad] is the angle corresponding to the bin number of the angle peak bin Information processing device.

2. the plurality of peak bins include a peak bin having a predetermined first bin number and a peak bin having a second bin number different from the first bin number, The correction unit corrects a phase of a peak bin having the second bin number without correcting a phase of a peak bin having the first bin number. The information processing device according to claim 1 .

3. The data processing unit performs a distance FFT on the received signal to obtain a distance power spectrum related to the distance to the object; The acquisition unit acquires a distance peak bin indicating a distance to the object based on the distance power spectrum.

3. The information processing device according to claim 1 or 2.

4. The acquisition unit acquires the angle peak bin of the angle FFT based on a data string corresponding to a position of the distance peak bin specified by the distance FFT. The information processing device according to claim 3 .

5. The correction unit corrects a phase of the output signal in accordance with the bin number of the distance peak bin. The information processing device according to claim 3 .

6. The data processing unit includes: performing the distance FFT on the received signal to which a first window function has been applied to obtain the distance power spectrum; obtaining the angular power spectrum by performing the angular FFT on the received signal to which a second window function different from the first window function is applied; The second window function is a higher order window function than the first window function. The information processing device according to claim 3 .

7. The data processing unit performs a velocity FFT on the received signal to obtain a velocity power spectrum related to the velocity of the object; The acquisition unit acquires a velocity peak bin indicating a velocity of the object based on the velocity power spectrum.

3. The information processing device according to claim 1 or 2.

8. The acquisition unit acquires the velocity peak bin of the velocity FFT based on a data string corresponding to a distance peak bin position specified by the distance FFT. The information processing device according to claim 7.

9. a data conversion unit that converts a plurality of data corresponding to the object into clustered data; 3. The information processing device according to claim 1 or 2.

10. A tracking unit is provided for tracking the object based on past data of the object when data corresponding to the object is not obtained within a predetermined period of time.

3. The information processing device according to claim 1 or 2.

11. The extracting unit extracts IQ data including an in-phase component and a quadrature component orthogonal to the in-phase component from the peak bin acquired by the acquiring unit.

3. The information processing device according to claim 1 or 2.

12. The data processing unit processes the received signal using a CAPON method or a compressed sensing algorithm.

3. The information processing device according to claim 1 or 2.

13. A method for sensing an object using an FMCW radar, comprising: A step of processing received signals of a plurality of antennas based on received waves of the FMCW radar to generate a power spectrum signal of a predetermined number of bins, and performing an angle FFT on the received signals to generate an angle power spectrum related to an angle with the object; obtaining a plurality of peak bins corresponding to the object based on the power spectrum signal, and obtaining an angle peak bin indicating an angle with the object based on the angle power spectrum; extracting an output signal responsive to the power spectrum signal; correcting a phase of the output signal according to bin numbers of the plurality of peak bins; Equipped with generating the power spectrum signal includes applying a window function of a higher order than a rectangular window to the received signal; the step of correcting the phase of the output signal includes a step of correcting the phase of the output signal such that a phase difference to be added or subtracted becomes (N-1) / 2 x π x sin(θ(i)) [rad] in accordance with the bin number of the angle peak bin; N is the number of the plurality of antennas, i is the bin number of the angle peak bin, θ(i) [rad] is the angle corresponding to the bin number of the angle peak bin Sensing method.

Images