Signal processing device and signal processing method

Abstract

Provided is a signal processing device comprising a signal acquisition unit (1) that acquires a discretized signal in which a target signal for estimating the time variation of a frequency component is discretized, an autocorrelation calculation unit (2) that calculates a symmetric instantaneous autocorrelation with respect to the discretization signal acquired by the signal acquisition unit (1), and a map calculation unit (3) that calculates a time frequency map from the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit (2). In the signal processing device, when the sampling interval of the discretized signal is represented by ΔT, the sampling time of the discretized signal is represented by pΔT (p = 1, ..., K (where K is an integer equal to or greater than 2)), and the symmetric instantaneous autocorrelation is represented by C(p, u), the symmetric instantaneous autocorrelation when u, which is a variable of the the symmetric instantaneous autocorrelation, is 2q (where q is an integer) is different from the symmetric instantaneous autocorrelation when u is 2q − 1, and the sampling interval in the u direction of the symmetric instantaneous autocorrelation is narrower than the sampling interval in the q direction when the symmetric instantaneous autocorrelation is C(p, q).

Description

Signal processing device and signal processing method

[0001] The present disclosure relates to a signal processing device and a signal processing method.

[0002] There is a signal processing method for estimating the time fluctuation of frequency components contained in a signal. For example, Non-Patent Document 1 discloses an example of such a signal processing method, which calculates a symmetric instantaneous autocorrelation for a discretized signal obtained by discretizing a signal to be estimated, and then calculates a Wigner-Ville distribution by performing an FFT (Fast Fourier Transform) on the symmetric instantaneous autocorrelation. The Wigner-Ville distribution is a time-frequency map showing the relationship between time and frequency components, and thus allows the time fluctuation of frequency components contained in a signal to be determined. When the above symmetric instantaneous autocorrelation is C(p,q), the variable q of C(p,q) changes positively and negatively simultaneously, so C(p,q) is symmetric. Since C(p,q) does not take an expected value, it is an instantaneous autocorrelation, which is different from so-called autocorrelation.

[0003] S. Mopuri, and A. Acharyya, “Low complexity VLSI architecture design methodology for Wigner Ville distribution,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol.67, no.12, pp.3532-3536, Dec. 2020.

[0004] The signal processing method disclosed in Non-Patent Document 1 performs an FFT on the symmetric instantaneous autocorrelation of a discretized signal. Therefore, the amount of calculation required to calculate a time-frequency map is reduced compared to when a Fourier transform is performed on the symmetric instantaneous autocorrelation of a signal to be estimated. However, there is a problem in that the observation width of the frequency components is halved.

[0005] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a signal processing device that can reduce the amount of calculation required to calculate a time-frequency map while ensuring a frequency component observation width similar to that obtained when a symmetric instantaneous autocorrelation of a signal to be estimated is Fourier transformed.

[0006] A signal processing device according to the present disclosure includes a signal acquisition unit that acquires a discretized signal obtained by discretizing a signal whose time variation of frequency components is to be estimated, an autocorrelation calculation unit that calculates a symmetric instantaneous autocorrelation for the discretized signal acquired by the signal acquisition unit, and a map calculation unit that calculates a time-frequency map from the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit. In the signal processing device, the sampling interval of the discretized signal is ΔT, the sampling time of the discretized signal is pΔT (p = 1, ..., K: K is an integer of 2 or greater), the symmetric instantaneous autocorrelation is C(p, u), the symmetric instantaneous autocorrelation when a variable of the symmetric instantaneous autocorrelation is 2q (q is an integer) is different from the symmetric instantaneous autocorrelation when u is 2q-1, and the sampling interval in the u direction of the symmetric instantaneous autocorrelation is narrower than the sampling interval in the q direction when the symmetric instantaneous autocorrelation is C(p, q).

[0007] According to the present disclosure, it is possible to reduce the amount of calculation required to calculate a time-frequency map while ensuring the same frequency component observation width as when Fourier transforming the symmetric instantaneous autocorrelation of a signal to be estimated.

[0008] FIG. 1 is a configuration diagram showing a signal processing device according to a first embodiment. FIG. 2 is a hardware configuration diagram showing hardware of the signal processing device according to the first embodiment. FIG. 3 is a hardware configuration diagram of a computer in the case where the signal processing device is realized by software, firmware, or the like. FIG. 4 is a flowchart showing a signal processing method which is a processing procedure of the signal processing device. FIG. 5 is an explanatory diagram showing an example of a time-frequency map P(p, f) shown in equation (4). FIG. 6 is an explanatory diagram showing an example of a time-frequency map 4P(p, f) calculated by a map calculation unit 3. FIG. 7 is a configuration diagram showing a signal processing device according to a second embodiment. FIG. 8 is a hardware configuration diagram showing hardware of the signal processing device according to the second embodiment. FIG. 9 is a configuration diagram showing a signal processing device according to a third embodiment. FIG. 10 is a hardware configuration diagram showing hardware of the signal processing device according to the third embodiment.

[0009] In order to explain the present disclosure in more detail, embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0010] Embodiment 1 Fig. 1 is a configuration diagram showing a signal processing device according to embodiment 1. Fig. 2 is a hardware configuration diagram showing the hardware of the signal processing device according to embodiment 1. The signal processing device shown in Fig. 1 includes a signal acquisition unit 1, an autocorrelation calculation unit 2, and a map calculation unit 3.

[0011] The signal acquirer 1 is realized by, for example, a signal acquisition circuit 11 shown in Fig. 2. The signal acquirer 1 acquires a discretized signal obtained by discretizing a target signal for which time fluctuations in frequency components are to be estimated. The signal acquirer 1 outputs the discretized signal to the autocorrelation calculation unit 2. In the signal processing device shown in Fig. 1, the signal acquirer 1 acquires the discretized signal. However, this is merely an example, and the signal acquirer 1 may also acquire a target signal for which time fluctuations in frequency components are to be estimated, and discretize the target signal.

[0012] The autocorrelation calculation unit 2 is realized by, for example, the autocorrelation calculation circuit 12 shown in FIG. 2 . The autocorrelation calculation unit 2 acquires a discretized signal from the signal acquisition unit 1. The autocorrelation calculation unit 2 calculates a symmetric instantaneous autocorrelation for the discretized signal. The autocorrelation calculation unit 2 outputs the symmetric instantaneous autocorrelation to the map calculation unit 3. Specifically, when the sampling interval of the discretized signal is ΔT and the sampling time of the discretized signal is pΔT (p = 1, ..., K: K is an integer equal to or greater than 2), the autocorrelation calculation unit 2 calculates C(p, u) shown in the following formula (1) as the symmetric instantaneous autocorrelation. In formula (1), when u, which is a variable of the symmetric instantaneous autocorrelation C(p, u), is 2q (q is an integer), C(p, u) is different from C(p, u) when u is 2q-1. The sampling interval in the u direction of C(p,u) is narrower than the sampling interval in the q direction when the symmetric instantaneous autocorrelation is C(p,q).

[0013] In equation (1), * denotes a complex conjugate.

[0014] The map calculation unit 3 is realized by, for example, the map calculation circuit 13 shown in Fig. 2. The map calculation unit 3 acquires the symmetric instantaneous autocorrelation C(p, u) from the autocorrelation calculation unit 2. The map calculation unit 3 calculates a time-frequency map from the symmetric instantaneous autocorrelation C(p, u). Specifically, the map calculation unit 3 calculates the time-frequency map by Fourier transforming the symmetric instantaneous autocorrelation C(p, u) in the u direction.

[0015] 1, it is assumed that each of the components of the signal processing device, that is, a signal acquisition unit 1, an autocorrelation calculation unit 2, and a map calculation unit 3, is realized by dedicated hardware as shown in Fig. 2. That is, it is assumed that the signal processing device is realized by a signal acquisition circuit 11, an autocorrelation calculation circuit 12, and a map calculation circuit 13. Each of the signal acquisition circuit 11, the autocorrelation calculation circuit 12, and the map calculation circuit 13 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof.

[0016] The components of the signal processing device are not limited to those realized by dedicated hardware, and the signal processing device may be realized by software, firmware, or a combination of software and firmware. The software or firmware is stored as a program in the memory of a computer. The computer refers to hardware that executes a program, and includes, for example, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, a processor, or a DSP (Digital Signal Processor).

[0017] 3 is a hardware configuration diagram of a computer when the signal processing device is realized by software, firmware, etc. When the signal processing device is realized by software, firmware, etc., a program for causing the computer to execute the respective processing procedures of the signal acquisition unit 1, the autocorrelation calculation unit 2, and the map calculation unit 3 is stored in a memory 21. Then, a processor 22 of the computer executes the program stored in the memory 21.

[0018] 2 shows an example in which each of the components of the signal processing device is realized by dedicated hardware, while Fig. 3 shows an example in which the signal processing device is realized by software, firmware, etc. However, this is merely an example, and some of the components in the signal processing device may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, etc.

[0019] The Wigner-Ville distribution is a time-frequency map P(t, f) calculated by Fourier transforming the symmetric instantaneous autocorrelation of the signal s(t) to be estimated, as shown in the following equation (2):

[0020] In equation (2), * denotes a complex conjugate, t denotes time, f denotes frequency, and τ denotes time.

[0021] In contrast, as shown in the following equation (3), a symmetric instantaneous autocorrelation C(p,q) for a discretized signal s(p) (p is a natural number representing time) is calculated, and the Wigner-Ville distribution calculated by performing FFT on the discretized signal s(p) becomes a time-frequency map P(p,f) as shown in the following equation (4). The discretized signal s(p) is a signal obtained by sampling the estimation target signal s(t) at a sampling interval Dt.

[0022] In equations (3) and (4), q is an integer and FFT[. ] ｑ represents the FFT for q.

[0023] In this case, since q changes positively and negatively at the same time, the sampling interval in the q direction of the symmetric instantaneous autocorrelation C(p,q) is twice the sampling interval of the symmetric instantaneous autocorrelation for the signal s(t). As a result, the frequency component observation width of the time-frequency map P(p,f) shown in equation (4) is half the frequency component observation width of the time-frequency map P(t,f) shown in equation (2).

[0024] Next, the operation of the signal processing device shown in Fig. 1 will be described. Fig. 4 is a flowchart showing a signal processing method, which is a processing procedure of the signal processing device. The signal acquisition unit 1 acquires a discretized signal s(p) obtained by discretizing a signal s(t) for which time fluctuations of frequency components are to be estimated (step ST1 in Fig. 4). The signal acquisition unit 1 outputs the discretized signal s(p) to the autocorrelation calculation unit 2.

[0025] The autocorrelation calculation unit 2 acquires the discretized signal s(p) from the signal acquisition unit 1. The autocorrelation calculation unit 2 calculates the symmetric instantaneous autocorrelation C(p, u) for the discretized signal s(p) as shown in equation (1) (step ST2 in FIG. 4). The symmetric instantaneous autocorrelation C(p, u) differs when u = 2q and when u = 2q - 1. Therefore, the sampling interval ΔT in the u direction of the symmetric instantaneous autocorrelation C(p, u) is half the sampling interval in the q direction of the symmetric instantaneous autocorrelation C(p, q) shown in equation (3). In other words, the sampling interval ΔT in the u direction of the symmetric instantaneous autocorrelation C(p, u) remains narrow. The autocorrelation calculation unit 2 outputs the symmetric instantaneous autocorrelation (p, u) to the map calculation unit 3.

[0026] The map calculation unit 3 acquires the symmetric instantaneous autocorrelation (p, u) from the autocorrelation calculation unit 2. The map calculation unit 3 performs a Fourier transform on the symmetric instantaneous autocorrelation C(p, u) in the u direction to calculate a Wigner-Ville distribution as a time-frequency map 4P(p, f) (step ST3 in FIG. 4). The Fourier transform may be, for example, an FFT or a DFT (Discrete Fourier Transform). Because the sampling interval ΔT in the u direction of the symmetric instantaneous autocorrelation C(p, u) remains narrow, the frequency component observation width of the Wigner-Ville distribution does not narrow. In other words, the frequency component observation width of the time-frequency map 4P(p, f) is the same as the frequency component observation width of the time-frequency map P(t, f) shown in Equation (2).

[0027] Fig. 5 is an explanatory diagram showing an example of the time-frequency map P(p, f) shown in equation (4). Fig. 6 is an explanatory diagram showing an example of the time-frequency map 4P(p, f) calculated by the map calculation unit 3. As shown in Figs. 5 and 6, the frequency component observation width of the time-frequency map 4P(p, f) is twice the frequency component observation width of the time-frequency map P(p, f) shown in equation (4). Frequency aliasing occurs in the time-frequency map P(p, f) shown in equation (4), but frequency aliasing does not occur in the time-frequency map 4P(p, f).

[0028] In the first embodiment described above, the signal processing device is configured to include a signal acquisition unit 1 that acquires a discretized signal obtained by discretizing a target signal for estimating time fluctuations of frequency components, an autocorrelation calculation unit 2 that calculates a symmetric instantaneous autocorrelation for the discretized signal acquired by the signal acquisition unit 1, and a map calculation unit 3 that calculates a time-frequency map from the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit 2. In the signal processing device, the sampling interval of the discretized signal is ΔT, the sampling time of the discretized signal is pΔT (p = 1, ..., K: K is an integer of 2 or more), the symmetric instantaneous autocorrelation is C(p, u), the symmetric instantaneous autocorrelation when u, which is a variable of the symmetric instantaneous autocorrelation, is 2q (q is an integer) is different from the symmetric instantaneous autocorrelation when u is 2q-1, and the sampling interval in the u direction of the symmetric instantaneous autocorrelation is narrower than the sampling interval in the q direction when the symmetric instantaneous autocorrelation is C(p, q). Therefore, the signal processing device can reduce the amount of calculation for calculating the time-frequency map while ensuring the same frequency component observation width as when Fourier transforming the symmetric instantaneous autocorrelation of the signal to be estimated.

[0029] 1, the autocorrelation calculation unit 2 calculates C(p, u) shown in equation (1) as the symmetric instantaneous autocorrelation. However, this is merely an example, and the autocorrelation calculation unit 2 may also calculate, for example, C(p, u) shown in equation (5) below or C(p, u) shown in equation (6) below as the symmetric instantaneous autocorrelation.

[0030] In equation (5), α is a real number, for example, 0<α< 1. In equation (6), a is a constant.

[0031] Furthermore, when 1≦p≦K, the autocorrelation calculation unit 2 may calculate C(p,u) as shown in the following equation (7) by fixing p, such as p=K / 2+1. In equation (7), C(p,u) is calculated to match the notation of equation (5) or equation (6). However, since the right-hand side of equation (7) uses K / 2+1 instead of p and p is not fixed, equation (7) may be written as C(u). In this case, the FFT on C(p,u) obtains a frequency distribution of the signal s(p) with a fixed time. However, even if the signal s(p) contains a chirp component, the frequency component is imaged without blurring due to the chirp component. In this way, even if a chirp component is contained, the frequency component can be estimated with high accuracy without narrowing the frequency observation width.

[0032]

[0033] Second Embodiment In a second embodiment, a signal processing device including a filtering unit 4 that performs filtering to reduce interference components contained in the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit 2 will be described.

[0034] Fig. 7 is a configuration diagram showing a signal processing device according to embodiment 2. In Fig. 7, the same reference numerals as in Fig. 1 indicate the same or corresponding parts, and detailed description thereof will be omitted. Fig. 8 is a hardware configuration diagram showing hardware of the signal processing device according to embodiment 2. In Fig. 8, the same reference numerals as in Fig. 2 indicate the same or corresponding parts, and detailed description thereof will be omitted. The signal processing device shown in Fig. 7 includes a signal acquisition unit 1, an autocorrelation calculation unit 2, a filtering unit 4, and a map calculation unit 3.

[0035] The filtering unit 4 is realized by, for example, the filtering circuit 14 shown in FIG. 8 . The filtering unit 4 acquires the symmetric instantaneous autocorrelation C(p, u) from the autocorrelation calculation unit 2. The filtering unit 4 performs filtering to reduce interference components contained in the symmetric instantaneous autocorrelation C(p, u). The filtering unit 4 outputs the filtered symmetric instantaneous autocorrelation C(p, u) to the map calculation unit 3.

[0036] 7, it is assumed that each of the components of the signal processing device, that is, the signal acquisition unit 1, the autocorrelation calculation unit 2, the filtering unit 4, and the map calculation unit 3, is realized by dedicated hardware as shown in Fig. 8. That is, it is assumed that the signal processing device is realized by a signal acquisition circuit 11, an autocorrelation calculation circuit 12, a filtering circuit 14, and a map calculation circuit 13. Each of the signal acquisition circuit 11, the autocorrelation calculation circuit 12, the filtering circuit 14, and the map calculation circuit 13 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination thereof.

[0037] The components of the signal processing device are not limited to those realized by dedicated hardware, and the signal processing device may be realized by software, firmware, or a combination of software and firmware. When the signal processing device is realized by software, firmware, or the like, a program for causing a computer to execute the respective processing procedures of the signal acquisition unit 1, the autocorrelation calculation unit 2, the filtering unit 4, and the map calculation unit 3 is stored in a memory 21 shown in Fig. 3. Then, a processor 22 shown in Fig. 3 executes the program stored in the memory 21.

[0038] 8 shows an example in which each of the components of the signal processing device is realized by dedicated hardware, while Fig. 3 shows an example in which the signal processing device is realized by software, firmware, etc. However, this is merely an example, and some of the components in the signal processing device may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, etc.

[0039] Next, the operation of the signal processing device shown in Fig. 7 will be described. However, apart from the filtering unit 4, the signal processing device is the same as that shown in Fig. 1. Therefore, only the operation of the filtering unit 4 will be described here. The filtering unit 4 acquires the symmetric instantaneous autocorrelation C(p, u) from the autocorrelation calculation unit 2. The symmetric instantaneous autocorrelation C(p, u) may contain an interference component. For example, when calculating the symmetric instantaneous autocorrelation C(p, u), a signal from another wave source or the signal itself at another time may be included in the symmetric instantaneous autocorrelation C(p, u) as an interference component.

[0040] The filtering unit 4 performs filtering to reduce interference components contained in the symmetric instantaneous autocorrelation C(p, u). For example, a two-dimensional low-pass filter in the time-frequency domain can be used as the filtering. The filtering unit 4 outputs the filtered symmetric instantaneous autocorrelation C(p, u) to the map calculation unit 3. In this case, the map calculation unit 3 calculates a time-frequency map 4P(p, f) from the filtered symmetric instantaneous autocorrelation C(p, u) by the filtering unit 4.

[0041] In the above-described second embodiment, the signal processing device shown in Fig. 7 is configured to include a filtering unit 4 that performs filtering to reduce interference components included in the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit 2, and to have the map calculation unit 3 calculate a time-frequency map from the symmetric instantaneous autocorrelation after filtering by the filtering unit 4. Therefore, the signal processing device shown in Fig. 7 can improve the calculation accuracy of the time-frequency map compared to the signal processing device shown in Fig. 1.

[0042] Third Embodiment In a third embodiment, a signal processing device including a scaling unit 6 that performs scaling on the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit 2 will be described.

[0043] Fig. 9 is a configuration diagram showing a signal processing device according to embodiment 3. In Fig. 9, the same reference numerals as in Figs. 1 and 7 indicate the same or corresponding parts, and detailed description thereof will be omitted. Fig. 10 is a hardware configuration diagram showing hardware of the signal processing device according to embodiment 3. In Fig. 10, the same reference numerals as in Figs. 2 and 8 indicate the same or corresponding parts, and detailed description thereof will be omitted. The signal processing device shown in Fig. 9 includes a signal acquisition unit 1, an autocorrelation calculation unit 5, a scaling unit 6, and a map calculation unit 7.

[0044] The autocorrelation calculation unit 5 is realized by, for example, the autocorrelation calculation circuit 15 shown in FIG. 10 . The autocorrelation calculation unit 5 acquires the discretized signal from the signal acquisition unit 1. The autocorrelation calculation unit 5 calculates, for example, the parametric symmetric instantaneous autocorrelation R(p, u) as shown in the following equation (8) as the symmetric instantaneous autocorrelation of the discretized signal. The autocorrelation calculation unit 5 outputs the parametric symmetric instantaneous autocorrelation R(p, u) to the scaling unit 6.

[0045]

[0046] The scaling unit 6 is realized by, for example, a scaling circuit 16 shown in FIG. 10 . The scaling unit 6 acquires the parametric symmetric instantaneous autocorrelation R(p, u) from the autocorrelation calculation unit 5. The scaling unit 6 performs scaling on the parametric symmetric instantaneous autocorrelation R(p, u). The scaling unit 6 outputs the scaled parametric symmetric instantaneous autocorrelation R(p, u) to the map calculation unit 7.

[0047] The map calculation unit 7 is realized by, for example, a map calculation circuit 17 shown in FIG. 10 . The map calculation unit 7 acquires the scaled parametric symmetric instantaneous autocorrelation R(p, u) from the scaling unit 6. The map calculation unit 7 calculates a two-dimensional frequency map P(c, f) as a time-frequency map from the scaled parametric symmetric instantaneous autocorrelation R(p, u). Specifically, the map calculation unit 3 performs a two-dimensional FFT on the scaled parametric symmetric instantaneous autocorrelation R(p, u) in the p and u directions to calculate an LV distribution (Lv's Distribution) as the two-dimensional frequency map P(c, f).

[0048] In the signal processing device shown in Fig. 9, the autocorrelation calculation unit 5, the scaling unit 6, and the map calculation unit 7 are applied to the signal processing device shown in Fig. 1. However, this is merely an example, and the autocorrelation calculation unit 5, the scaling unit 6, and the map calculation unit 7 may also be applied to the signal processing device shown in Fig. 7.

[0049] 9, it is assumed that each of the components of the signal processing device, that is, the signal acquisition unit 1, the autocorrelation calculation unit 5, the scaling unit 6, and the map calculation unit 7, is realized by dedicated hardware as shown in Fig. 10. That is, it is assumed that the signal processing device is realized by a signal acquisition circuit 11, an autocorrelation calculation circuit 15, a scaling circuit 16, and a map calculation circuit 17. Each of the signal acquisition circuit 11, the autocorrelation calculation circuit 15, the scaling circuit 16, and the map calculation circuit 17 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination thereof.

[0050] The components of the signal processing device are not limited to those realized by dedicated hardware, and the signal processing device may be realized by software, firmware, or a combination of software and firmware. When the signal processing device is realized by software, firmware, or the like, a program for causing a computer to execute the respective processing procedures of the signal acquisition unit 1, the autocorrelation calculation unit 5, the scaling unit 6, and the map calculation unit 7 is stored in a memory 21 shown in Fig. 3. Then, a processor 22 shown in Fig. 3 executes the program stored in the memory 21.

[0051] 10 shows an example in which each of the components of the signal processing device is realized by dedicated hardware, while Fig. 3 shows an example in which the signal processing device is realized by software, firmware, etc. However, this is merely an example, and some of the components in the signal processing device may be realized by dedicated hardware, and the remaining components may be realized by software, firmware, etc.

[0052] Next, the operation of the signal processing device shown in Fig. 9 will be described. The signal acquisition unit 1 acquires a discretized signal s(p) obtained by discretizing a signal s(t) for which time fluctuations of frequency components are to be estimated. The signal acquisition unit 1 outputs the discretized signal s(p) to the autocorrelation calculation unit 5.

[0053] The autocorrelation calculation unit 5 acquires the discretized signal s(p) from the signal acquisition unit 1. The autocorrelation calculation unit 5 calculates the parametric symmetric instantaneous autocorrelation R(p, u) as the symmetric instantaneous autocorrelation for the discretized signal s(p) as shown in equation (8). The parametric symmetric instantaneous autocorrelation R(p, u) differs when u = 2q and when u = 2q - 1, similar to the symmetric instantaneous autocorrelation C(p, u) shown in equation (1). The autocorrelation calculation unit 5 outputs the parametric symmetric instantaneous autocorrelation R(p, u) to the scaling unit 6.

[0054] The scaling unit 6 acquires the parametric symmetric instantaneous autocorrelation R(p, u) from the autocorrelation calculation unit 5. The scaling unit 6 performs scaling on the parametric symmetric instantaneous autocorrelation R(p, u). Scaling on the parametric symmetric instantaneous autocorrelation R(p, u) is performed, for example, by multiplying the parametric symmetric instantaneous autocorrelation R(p, u) by a known scaling function. Since scaling itself is a known technique, detailed description thereof will be omitted. The scaling unit 6 outputs the scaled parametric symmetric instantaneous autocorrelation R(p, u) to the map calculation unit 7.

[0055] The map calculation unit 7 acquires the scaled parametric symmetric instantaneous autocorrelation R(p, u) from the scaling unit 6. The map calculation unit 7 calculates a two-dimensional frequency map P(c, f) from the scaled parametric symmetric instantaneous autocorrelation R(p, u). Specifically, the map calculation unit 3 performs two-dimensional FFT on the scaled parametric symmetric instantaneous autocorrelation R(p, u) in the p and u directions to calculate the LV distribution as the two-dimensional frequency map P(c, f).

[0056] In the above-described third embodiment, the signal processing device shown in Fig. 9 is configured to include a scaling unit 6 that performs scaling on the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit 5, and to have a map calculation unit 7 calculate a time-frequency map from the symmetric instantaneous autocorrelation after scaling by the scaling unit 6. Therefore, the signal processing device shown in Fig. 9 can improve the calculation accuracy of the time-frequency map compared to the signal processing device shown in Fig. 1 .

[0057] In addition, the present disclosure allows for free combination of the respective embodiments, modification of any of the components of the respective embodiments, or omission of any of the components of the respective embodiments.

[0058] The present disclosure can reduce the amount of calculation required to calculate a time-frequency map while ensuring a frequency component observation width similar to that required when performing a Fourier transform on a symmetric instantaneous autocorrelation of a signal to be estimated, and can be used in signal processing devices and signal processing methods.

[0059] REFERENCE SIGNS LIST 1 signal acquisition unit, 2 autocorrelation calculation unit, 3 map calculation unit, 4 filtering unit, 5 autocorrelation calculation unit, 6 scaling unit, 7 map calculation unit, 11 signal acquisition circuit, 12 autocorrelation calculation circuit, 13 map calculation circuit, 14 filtering circuit, 15 autocorrelation calculation circuit, 16 scaling circuit, 17 map calculation circuit, 21 memory 22 processor.

Claims

1. A signal processing device comprising: a signal acquisition unit that acquires a discretized signal, the discretized signal being a target of estimating the time variation of frequency components; an autocorrelation calculation unit that calculates a symmetric instantaneous autocorrelation for the discretized signal acquired by the signal acquisition unit; and a map calculation unit that calculates a time-frequency map from the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit, wherein the sampling interval of the discretized signal is ΔT, the sampling time of the discretized signal is pΔT (p = 1, ..., K: K is an integer of 2 or more), the symmetric instantaneous autocorrelation is C(p, u), the symmetric instantaneous autocorrelation when u, a variable of the symmetric instantaneous autocorrelation, is 2q (q is an integer) is different from the symmetric instantaneous autocorrelation when u is 2q-1, and the sampling interval in the u direction of the symmetric instantaneous autocorrelation is narrower than the sampling interval in the q direction when the symmetric instantaneous autocorrelation is C(p, q).

2. The signal processing device according to claim 1, characterized in that the map calculation unit calculates the time-frequency map by Fourier transforming C(p, u), which is the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit, in the u direction.

3. The signal processing device according to claim 1 or 2, wherein p in C(p, u) is 1 or more and K or less, and p is fixed to p=K / 2+1.

4. A signal processing device as claimed in any one of claims 1 to 3, characterized in that it comprises a filtering unit that performs filtering to reduce interference components contained in the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit, and the map calculation unit calculates a time-frequency map from the symmetric instantaneous autocorrelation after filtering by the filtering unit.

5. A signal processing device according to any one of claims 1 to 4, characterized in that it comprises a scaling unit that performs scaling on the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit, and the map calculation unit calculates a time-frequency map from the symmetric instantaneous autocorrelation after scaling by the scaling unit.

6. A signal processing method comprising: a signal acquisition unit acquiring a discretized signal obtained by discretizing a signal for which time fluctuations of frequency components are to be estimated; an autocorrelation calculation unit calculating a symmetric instantaneous autocorrelation for the discretized signal acquired by the signal acquisition unit; and a map calculation unit calculating a time-frequency map from the symmetric instantaneous autocorrelation calculated by the autocorrelation calculation unit; wherein the sampling interval of the discretized signal is ΔT, the sampling time of the discretized signal is pΔT (p = 1, ..., K: K is an integer of 2 or more), the symmetric instantaneous autocorrelation is C(p, u), the symmetric instantaneous autocorrelation when u, a variable of the symmetric instantaneous autocorrelation, is 2q (q is an integer) is different from the symmetric instantaneous autocorrelation when u is 2q-1, and the sampling interval in the u direction of the symmetric instantaneous autocorrelation is narrower than the sampling interval in the q direction when the symmetric instantaneous autocorrelation is C(p, q).