Signal processing device, signal processing method, program, and storage medium

Abstract

This signal processing device comprises: a window function unit (21) that multiplies an input signal by a new window function obtained by multiplying a window function by a function that is a linear sum of trigonometric functions, and extracts a specific signal component present in the input signal; and a frequency converting unit (22) that frequency-converts the specific signal component present in the input signal extracted by the window function unit (21).

Description

Signal processing device, signal processing method, program, and storage medium

[0001] The present disclosure relates to a signal processing device, a signal processing method, a program, and a storage medium that extract specific signal components present in an input signal using a window function and perform frequency conversion on the extracted signal.

[0002] A technique for frequency-converting specific signal components present in an input signal to perform frequency analysis on the input signal is known. To suppress side lobe components in the frequency domain after frequency conversion, a technique for using a window function to extract specific signal components present in the input signal is known. Various research efforts are being conducted on window functions for reducing side lobe components. One window function for reducing side lobe components is shown in Non-Patent Document 1. Non-Patent Document 1 proposes a window function based on a new class of cosine-sum window functions.

[0003] T. Yamaoka and S. Kageme, “New class of cosine-sum windows,” IEEE Access, vol.11, 2023, pp.5296-5305

[0004] In the frequency domain, specific signal components present in the input signal shown in Non-Patent Document 1 are extracted using the proposed cosine sum window function and frequency converted, and although the side lobe components in the frequency domain are suppressed, further reduction is desired.

[0005] The present disclosure has been made in consideration of the above-mentioned points, and has an object to provide a signal processing device that further reduces side lobe components in the frequency domain.

[0006] The signal processing device according to the present disclosure includes a window function unit that multiplies an input signal by a new window function obtained by multiplying a window function by a function that is a linear sum of trigonometric functions, and extracts a specific signal component present in the input signal, and a frequency conversion unit that frequency converts the specific signal component present in the input signal extracted by the window function unit.

[0007] According to the present disclosure, specific signal components present in an input signal are extracted using a new window function obtained by multiplying a window function by a linear sum of trigonometric functions, thereby further reducing side lobe components in the frequency domain.

[0008] FIG. 1 is a block diagram showing an example of the configuration of a signal processing device according to a first embodiment. FIG. 2 is a diagram showing a frequency response w(f) obtained by performing a frequency conversion on a window function w(t) given by a Gaussian window. FIG. 3 is a diagram showing an average frequency response w(f)ave. obtained by combining and averaging frequency-shifted frequency responses w(f-0.5) and w(f+0.5). FIG. 4 is a diagram showing an example of the hardware configuration of a signal processing device according to a first embodiment. FIG. 5 is a flowchart showing an example of the operation of a signal processing device according to a first embodiment.

[0009] Embodiment 1. A signal processing device according to embodiment 1 will be described with reference to Figures 1 to 5. The signal processing device according to embodiment 1 is a signal processing device used, for example, in a radar system, in which an antenna element receives an incoming wave from a target, the received analog signal is converted into a digital signal by a signal receiving unit, the received signal is temporarily stored as an input signal, a window function is used to extract a specific signal component that is a desired signal component from the input signal consisting of the stored digital signal, the frequency of the extracted specific signal component is converted, and the signal is output as a signal to be subjected to frequency analysis.

[0010] The signal processing device according to the first embodiment is characterized by a window function used to extract a specific signal component. The window function used in the signal processing device according to the first embodiment is, for example, a new window function obtained by multiplying a window function given by a Gaussian window by a function that is a linear sum of trigonometric functions. Note that the window function is not limited to a window function given by a Gaussian window, and may also be a window function given by a Laplace window. In the following description, a case where a window function given by a Gaussian window is used will be described.

[0011] First, we will explain the concept of the new window function used in the signal processing device according to embodiment 1. The window function w(t) (-0.5≦t≦0.5) for obtaining the new window function is given by a Gaussian window shown in the following equation (1).

[0012] In the above formula (1), the window width is given as 1 second, but any time width can be used by scaling.

[0013] Figure 2 shows the frequency response w(f) obtained by performing a frequency conversion on the window function w(t) given by a Gaussian window. In Figure 2, the horizontal axis represents frequency, the vertical axis represents the frequency domain response, and w(f) represents the frequency response. As can be seen from the frequency response w(f) shown in Figure 2, the positive and negative signs of the amplitude of the side lobe components in the frequency domain change periodically in increments of 1 Hz. This is a phenomenon that occurs because the window function w(t) is given with a finite length.

[0014] Considering this phenomenon, by combining frequency-shifted frequency responses w(f), it is possible to reduce sidelobe components by causing them to interfere with each other. Figure 3 shows the average frequency response w(f) ave., obtained by overlaying the frequency responses w(f-0.5) and w(f+0.5), i.e., by combining and averaging them. In Figure 3, the horizontal axis represents frequency, and the vertical axis represents the frequency domain response. w(f-0.5) represents the frequency response obtained by shifting the frequency by -0.5, w(f+0.5) represents the frequency response obtained by shifting the frequency by +0.5, and w(f) ave. represents the average frequency response obtained by combining and averaging the frequency responses w(f-0.5) and w(f+0.5).

[0015] As can be seen from the frequency response shown in FIG. 3, the frequency response w(f-0.5) and the frequency response w(f+0.5) have side lobe component amplitudes that are positive or negative in the frequency domain, and therefore the average frequency response w(f)ave. obtained by averaging the frequency response w(f-0.5) and the frequency response w(f+0.5) has reduced side lobe components in the frequency domain. In the signal processing device according to the first embodiment, a new window function is set based on the idea of ​​performing shift processing and averaging processing on the frequency response. The new window function will be described below.

[0016] The frequency component of the average frequency response w(f) ave. obtained by averaging the frequency response w(f-0.5) and the frequency response w(f+0.5) is defined as Wpro1(f). The frequency component Wpro1(f) is expressed by the following equation (2): Wpro1(f) = (w(f-0.5) + w(f+0.5)) / 2 (2) The frequency component Wpro1(f) is transformed by the inverse frequency transform F －１ ｆ Applying [ ](t) and analyzing, the following equation (3) is obtained.

[0017]

[0018] As can be seen from equation (3) above, the frequency response obtained by performing frequency conversion using a new window function obtained by multiplying the window function w(t) by cos(jπt) is equivalent to the average frequency response w(f)ave. obtained by averaging the frequency response w(f-0.5) and the frequency response w(f+0.5), and side lobe components in the frequency domain are reduced. cos(jπt) is a trigonometric function with a half period, and is a function of the linear sum of trigonometric functions.

[0019] Furthermore, the frequency component Wpro2(f) obtained by repeating the side lobe reduction process is subjected to the inverse frequency transform F －１ ｆ Applying [ ](t) and analyzing, the following equation (4) is obtained.

[0020]

[0021] As can be seen from equation (4) above, the frequency response obtained by performing frequency conversion using a new window function obtained by multiplying the window function w(t) by (0.5 + 0.5 cos(j2πt)) further reduces side lobe components in the frequency domain. The (0.5 + 0.5 cos(j2πt)) by which the window function w(t) is multiplied is a function that is a linear sum of trigonometric functions, and corresponds to a Hann window. The new window function is a window function obtained by multiplying the window function w(t) by the Hann window (0.5 + 0.5 cos(j2πt)).

[0022] Furthermore, the frequency component Wpro3(f) obtained by repeating the side lobe reduction process is subjected to an inverse frequency transform F －１ ｆApplying [ ](t) and analyzing, the following equation (5) is obtained.

[0023]

[0024] As can be seen from equation (5) above, the frequency response obtained by performing frequency conversion using a new window function obtained by multiplying the window function w(t) by (0.75 cos(πt) + 0.25 cos(3πt)) further reduces side lobe components in the frequency domain. The (0.75 cos(πt) + 0.25 cos(3πt)) by which the window function w(t) is multiplied is a linear sum function of trigonometric functions, and corresponds to cosine-sum windows. The new window function is a window function obtained by multiplying the window function w(t) by the cosine-sum window (0.75 cos(πt) + 0.25 cos(3πt)).

[0025] The signal processing device according to the first embodiment is a signal processing device that extracts a specific signal component present in an input signal by using a new window function obtained by multiplying a window function w(t) by a function that is a linear sum of trigonometric functions, and reduces side lobe components in the frequency domain with respect to a frequency response obtained by performing frequency conversion on the extracted specific signal component.

[0026] As shown in Fig. 1, the signal processing device according to the first embodiment includes a storage device unit 10 and a signal processing unit 20. The storage device unit 10 has an input signal storage unit 11 and an output signal storage unit 12. The input signal storage unit 11 temporarily stores, as an input signal S(t), a received signal converted into a digital signal by the signal receiving unit 30. The input signal S(t) stored in the input signal storage unit 11 and provided to the signal processing unit 20 can have any number of dimensions, t, but in the first embodiment, it is assumed to be a one-dimensional signal, and the signal processing unit 20 performs frequency conversion on the dimension of t.

[0027] The signal receiving unit 30 is a commonly known signal receiver that converts an analog received signal from an antenna element that receives an incoming wave from a target into a digital signal and outputs the digital received signal, for example, in a radar system. The output signal storage unit 12 temporarily stores the output signal output from the signal processing unit 20, and the stored output signal is output to an external device 40. The external device 40 is, for example, a display device or other display device.

[0028] The signal processing unit 20 has a window function unit 21 and a frequency conversion unit 22. The window function unit 21 acquires the input signal S(t) stored in the input signal storage unit 11, and multiplies the acquired input signal S(t) by a new window function obtained by multiplying the window function by a function of the linear sum of trigonometric functions to extract a specific signal component S(t) present in the input signal S(t). ｗ The width of the signal to be extracted is set to "1", the window function is w(x) (-0.5≦x≦0.5), the function of the linear sum of the trigonometric function is set to a trigonometric function with a half period, and the signal component S ｗ The center of the extraction of (t) is t ０ Then, a specific signal component S ｗ (t) can be expressed by the following equation (6).

[0029]

[0030] The processing using the new window function by the window function unit 21 uses the same data length as the processing using the conventional window function w(x), and as a result, there is no problem such as an increased amount of calculation. Note that the width of the signal to be extracted is 1, and the range for designing the window function w(x) is set to -0.5≦x≦0.5. However, the range for designing the window function w(x) can be changed by changing the scale depending on the width of the signal to be extracted. Also, while x is given as a real value, it may also take a discrete value. Furthermore, while the function of the linear sum of trigonometric functions is set to a trigonometric function with a half period, as explained above, the function of the linear sum of trigonometric functions may also be a cosine sum window such as a Hann window.

[0031] The frequency conversion unit 22 converts a specific signal component S(t) present in the input signal S(t) extracted by the window function unit 21. ｗ(t) is frequency converted to frequency f, and the frequency component S ｗ (f) is obtained. The function after frequency transformation is called F ｔ [S ｗ (t)](f), the frequency component S ｗ (f) can be expressed by the following equation (7).

[0032] The frequency component S obtained by the frequency conversion unit 22 ｗ (f) is stored in the output signal storage unit 12. The frequency component S obtained by the frequency conversion unit 22 ｗ (f) is a specific signal component S present in the input signal S(t) extracted by processing using a new window function obtained by multiplying the window function by a linear sum function of trigonometric functions in the window function unit 21. ｗ Since (t) is frequency converted, the side lobe components in the frequency domain are reduced due to the effect of the new window function.

[0033] The signal processing device according to the first embodiment is realized by a computer hardware configuration, and as shown in FIG. 4, includes a CPU (Central Processing Unit) 1A, a large-capacity semiconductor memory (RAM: Random Access Memory) 1B, a storage device (ROM: Read only memory) 1C such as a non-volatile storage device such as a hard disk device or an SSD device, an input interface unit 1D, an output interface unit 1E, and a signal path (bus) 1F.

[0034] The CPU 1A controls and manages the RAM 1B, the ROM 1C, the input interface section 1D, and the output interface section 1E. The CPU 1A loads programs stored in the ROM 1C into the RAM 1B, and executes various processes based on the programs loaded into the RAM 1B.

[0035] The storage device 10 is configured with a RAM 1B. The input signal storage 11 configured with the RAM 1B acquires the input signal S(t) from the signal receiving unit 30 via the input interface 1D under the control and management of the CPU 1A, and stores the input signal S(t). The output signal storage 12 configured with the RAM 1B acquires the frequency component S(t) output from the signal processing unit 20 under the control and management of the CPU 1A. ｗ The output signal (f) is temporarily stored and output to the external device 40 via the output interface section 1E.

[0036] The signal processing unit 20 is composed of a CPU 1A and a RAM 1B, and the CPU 1A performs window function processing (window function unit 21) and frequency conversion processing (frequency conversion unit 22) on the input signal S(t) temporarily stored in the RAM 1B using a new window function stored in the RAM 1B based on a program loaded from the ROM 1C to the RAM 1B, thereby generating a frequency component S ｗ (f) is obtained and stored in RAM 1B.

[0037] Next, the operation of the signal processing device according to the first embodiment, that is, the signal processing method, will be described with reference to Fig. 5. In step ST1, an input signal S(t) from the signal receiving unit 30 is input to and stored in the input signal storage unit 11. Step ST1 is a signal input step.

[0038] In step ST2, a window function w(t−t ０ ) to the linear sum of trigonometric functions cos(π(t-t ０ )) and multiplying it by the new window function obtained by multiplying it by the specific signal component S ｗ (t) is extracted by the window function unit 21. Step ST2 is a window function processing step.

[0039] In step ST3, the frequency conversion unit 22 converts the specific signal component S extracted by the window function unit 21 into ｗ (t) as a function F ｔ [S ｗ (t)](f) to obtain the frequency component S ｗ Step ST3 is a frequency conversion step.

[0040] In step ST4, the frequency component S obtained by the frequency conversion unit 22 ｗ (f) is stored in the output signal storage unit 12. Step ST3 is a frequency component storage step. After that, the frequency components S ｗ (f) is read out to the external device 40.

[0041] In the signal processing device according to the first embodiment, the signal processing method in steps ST2 and ST3 is performed by the CPU 1A executing processing in accordance with a program stored in the RAM 1B (ROM 1C). That is, the signal processing program stored in the ROM 1C includes a procedure for multiplying an input signal by a new window function obtained by multiplying a window function by a function that is a linear sum of trigonometric functions, and extracting a specific signal component present in the input signal, and a procedure for frequency-converting the extracted specific signal component present in the input signal.

[0042] In the signal processing device according to the first embodiment, the window function unit 21 extracts a specific signal component present in the input signal using a new window function obtained by multiplying a window function by a linear sum of trigonometric functions, and the frequency conversion unit performs frequency conversion on the specific signal component extracted by the window function unit, thereby further reducing side lobe components in the frequency domain. Moreover, in the signal processing device according to the first embodiment, the data length used in the calculation processing for the extraction of the specific signal component and the frequency conversion is the same as the data length used for processing using the window function, and as a result, there is no problem such as an increase in the amount of calculation.

[0043] Any of the components of the embodiments may be modified or omitted.

[0044] The signal processing device according to the present disclosure is applied to, for example, a signal processing device used in a radar system or a signal processing device used in a radar image generation system.

[0045] 10 storage device unit, 11 input signal storage unit, 12 output signal storage unit, 20 signal processing unit, 21 window function unit, 22 frequency conversion unit

Claims

1. A signal processing device comprising: a window function unit that multiplies an input signal by a new window function obtained by multiplying a window function by a function that is a linear sum of trigonometric functions, and extracts a specific signal component present in the input signal; and a frequency conversion unit that frequency converts the specific signal component present in the input signal extracted by the window function unit.

2. A signal processing device according to claim 1, wherein said window function is a window function given by a Gaussian window.

3. A signal processing device according to claim 1 or 2, wherein the function of the linear sum of trigonometric functions is a trigonometric function with a half period.

4. A signal processing device according to claim 1 or 2, wherein the function of the linear sum of trigonometric functions is a Hann window.

5. A signal processing device according to claim 1 or 2, wherein the function of the linear sum of trigonometric functions is a cosine sum window.

6. A signal processing method for a signal processing device having a window function unit and a frequency conversion unit, comprising: a step in which the window function unit multiplies an input signal with a new window function obtained by multiplying a window function by a linear sum of trigonometric functions, and extracts a specific signal component present in the input signal; and a step in which the frequency conversion unit frequency-converts the extracted specific signal component present in the input signal.

7. A signal processing program that causes a computer to execute the steps of: multiplying an input signal by a new window function obtained by multiplying a window function by a linear sum of trigonometric functions, and extracting a specific signal component present in the input signal; and frequency converting the extracted specific signal component present in the input signal.

8. A recording medium storing a program that causes a computer to execute the steps of: multiplying an input signal by a new window function obtained by multiplying a window function by a linear sum of trigonometric functions, and extracting specific signal components present in the input signal; and frequency converting the extracted specific signal components present in the input signal.

Images