Method and apparatus for mitigating GNSS signal interference using an adaptive notch filter

Abstract

A method and apparatus for mitigating GNSS signal interference using an adaptive notch filter (ANF) operates based on signals received from one or more satellites of a global navigation satellite system (GNSS) such as GPS, GLONASS. The apparatus in one embodiment consists of a notch filter having a tunable transfer function zero frequency, receiving an input signal and generating an output signal. A bandpass filter connected to the output of the notch filter receives the output signal. An adaptation block is connected to the bandpass filter and adjusts the parameters of the notch filter to minimize a particular cost function.

Description

Technical Field

[0001] The present disclosure generally relates to a method of filtering, and more specifically, to a method and apparatus for mitigating GNSS signal interference using an adaptive notch filter.

Background Art

[0002] Signal transmission is often subject to interference from multiple sources. The sources of interference can be natural or artificial, and can also be intentional or unintentional. Interference and noise may interfere with operations that require the reception of useful signals. For example, signal loss may occur due to thermal noise or interference jamming (such as signal blocking), which may interfere with the operations necessary to receive the transmitted signal. GNSS signals from artificial satellites of the Global Navigation Satellite System (GNSS) are generally weak in intensity and thus are susceptible to interference. What is needed here is a method for mitigating signal interference.

Summary of the Invention

[0003] The apparatus according to one embodiment has a frequency tunable to the zeros of the transfer function (also referred to as the "zero frequency") and consists of a notch filter configured to receive an input signal and generate an output signal. The band-pass filter is connected to the output of the notch filter and is configured to receive the output signal. The adaptation block is connected to the band-pass filter and is configured to adjust the parameters of the notch filter to minimize a specific cost function. In one embodiment, the notch filter is a first-order digital complex filter, the input of the digital complex filter is connected to the input of the apparatus, and at the time when the adaptation is completed, the frequency of the zeros of the transfer function of the digital complex filter is equal to the interference frequency. In one embodiment, the band-pass filter is a first-order digital complex filter having a pole frequency that coincides with the zero frequency of the notch filter. The digital complex filter has the transfer coefficient of the band-pass filter, and the input of the digital complex filter is connected to the output of the apparatus. In one embodiment, the adaptation block is configured to track the interference frequency and adjust the zeros of the filter to achieve minimization of the cost function, and the input of the adaptation block is connected to the output of the band-pass filter. The apparatus in one embodiment is configured to mitigate multi-spectrum interference by a notch filter having a specific transfer function. The high-pass filter in one embodiment is shifted using, for real coefficients, those real coefficients multiplied by a power function of a complex power exponent. Further, a method is also described that includes the step of receiving an input signal from one or more artificial satellites of a global positioning satellite system. The input signal is filtered by a notch filter and input to a band-pass filter, and the output signal thereof is input to an adaptation block, where the parameters of the notch filter are adjusted according to the least mean squares (LMS) algorithm. Further, an apparatus consisting of a processor and a memory connected to the processor is also described. The memory stores computer program instructions that cause the processor to perform operations at runtime.

Brief Description of the Drawings

[0004]

Figure 1

[0005]

Figure 2

[0006]

Figure 3

[0007]

Figure 4

[0008]

Figure 5

[0009]

Figure 6

[0010]

Figure 7

[0011]

Figure 8

[0012]

Figure 9

[0013] A method and apparatus for mitigating GNSS signal interference using an adaptive notch filter (ANF) operates based on signals received from one or more satellites of a Global Navigation Satellite System (GNSS), such as GPS, GLONASS, etc.

[0014] FIG. 1 shows an interference mitigation system 100 of a GNSS receiver. The interference mitigation system 100 consists of a notch filter 102 and an adaptive block 104. In one embodiment, the interference mitigation system 100 is used to mitigate intentionally generated interference. For example, the mitigation system 100 can be used as a countermeasure for signals that are being jammed (or blocked by the use of interference). The signal y(t) at the input of the GNSS receiver may contain both thermal noise and interference. (For example, filtering by an RF bandpass filter), downconversion, and sampling frequency f S =(T S ) -1 After analog-to-digital conversion (ADC) using, the resulting GNSS digital signal y[n] at the input side of the notch filter 102 of the interference mitigation system 100 of the GNSS receiver can be expressed as follows. y[n]=x[n]+i[n]+η[n] (1) where TIFF2025517890000002.tif10159 is a combination of GNSS signals from different multiple satellites, i[n]- is an interference signal, and η[n]- is thermal noise having a spectral density N 0 .

[0015] Mitigating GNSS signal interference in the time domain may involve removing the interference signal i[n] from the signal y[n]. TIFF2025517890000003.tif14159

[0016] Narrowband harmonic interference and chirp signals can be expressed in the following single-component form. i[n]=A[n]exp{jφ[n]} (3)

[0017] In this situation, the instantaneous value of the interference signal frequency (Equation (3)) has the following form. TIFF2025517890000004.tif14159

[0018] For the case of a single-component single equation (3), the following can be used for predicting the form of the interference signal i[n] from the past sample i[n - 1].[[]END]] i[n]=a[n]i[n-1] (5)

[0019] For adjacent samples, considering that the noise amplitude is considered not to change, and taking into account that the coefficient of the linear prediction of the error signal is determined by estimating only the instantaneous value of the interference frequency f i from Equations (3) and (5), the coefficient a[n] can be expressed as follows. TIFF2025517890000005.tif17159

[0020] In one embodiment, the interference mitigation system of the GNSS receiver uses a digital notch filter having a first-order complex transfer function (also referred to as a first-order digital complex filter) with the following transfer function. TIFF2025517890000006.tif17159Here z 0 [n]- is the complex zero of the transfer function (Equation (7)), and k α <1- is the real coefficient at the pole of the transfer function of the filter.

[0021] The numerator of the transfer function of Equation (7) is referred to as the moving average (MA) part. The MA component 106 shown in FIG. 1 is used to execute the function numerator of Equation (7) in one embodiment. In one embodiment, the MA part 106 is used for mitigating the interference signal. The filter transfer function H MA (z)=1-z 0 [n]z -1The output signal after passing through the MA part having [it] can be expressed as follows. TIFF2025517890000007.tif18159

[0022] According to Equation (8), when the interference signal estimation (Equation (6)) coincides with the zero point value of the notch filter that is in synchronization with the interference of the filter transfer function (Equation (7)), the single-component interference signal is relaxed, which is expressed as follows. TIFF2025517890000008.tif13159

[0023] Filter transfer function The autoregressive (AR) part of TIFF2025517890000009.tif13159 reduces the influence of MA on the distortion of the desired (i.e., useful) signal by narrowing the stopband width of the filter. The AR component 108 shown in FIG. 1 is used to execute the AR part of the filter transfer function in one embodiment. Coefficient k α As the value of approaches 1, the frequency distortion of the desired signal becomes smaller accordingly.

[0024] In accordance with the least mean squares (LMS) algorithm, the search for the optimal value (also referred to as the target value) of the zero point frequency of the notch filter is repeatedly executed at each sample. The search direction and the numerical value of the correction addition itself are related to the value of the gradient of the cost function, which is equal to the power of the output signal represented below. J[n]=E{|x f [n]| 2} (10) Here, x f [n]- is the output signal of the MA block in the notch filter. When the adaptation is successful, that is, when the cost function reaches the minimum value, the zero point frequency of the notch filter coincides with the interference frequency with a predetermined accuracy.

[0025] Restricting the increase in the interference relaxation depth means that the interference signal power J on the output side of the notch filter and the noise power To reduce the ratio to 14159, where N 0 - is the thermal noise spectral density and B- is the bandwidth of the useful GNSS signal.

[0026] For the two-sided spectrum of the digital signal with the Nyquist frequency f S / 2, the bandwidth frequency is B = 2·(f S / 2) = f S That is. In the case of a narrowband interference signal, the bandwidth of the noise Compared with TIFF2025517890000011.tif12159, the bandwidth Δω of the interference signal i is so small that it is generally impossible to reduce the spectral density of the interference signal to the level of the spectral density of the thermal noise.

[0027] Figure 2 shows an adaptive notch filter 200 of a GNSS receiver according to an embodiment. A bandpass filter 206 is arranged on the input side of the adaptive block 204. By using a bandpass filter on the input side of the adaptive block, the mitigation depth of harmonic interference (for example, continuous wave (CW) interference) and chirp interference can be further increased by an additional 20 dB compared with an ANF-based interference mitigation system that can otherwise be used for interference countermeasures. The center frequency of the bandpass filter 206 coincides with the zero frequency of the notch filter 202. In one embodiment, the notch filter 202 has a transfer function with a tunable frequency of the zero. In one embodiment, the bandpass filter 206 is a digital complex filter. After the signal is processed by the bandpass filter 206, when the adaptive block 204 calculates the gradient of the cost function, the contribution due to the power of the thermal noise signal present in the output signal of the notch filter 202 decreases in the adaptive block 204.

[0028] Note that, since a band-pass filter is used in the arithmetic circuit of the adaptive notch filter 200 of the GNSS receiver, it should be noted that the band-pass filter 206 arranged on the input side of the adaptive block 204 does not change the signal in the direct transmission channel.

[0029] The width of the rejection / blocking band of the notch filter 202 with the transfer function in the form shown in Equation (7) is B -3dB =(1 - k α )f S π / 10. In one embodiment, it is presumed that the blocking band of the notch filter 202 coincides with the frequency band of the interference signal (B -3dB = Δω i ). When the bandwidth of the notch filter 202 is related to the bandwidth of the interference signal, the reduction of the noise level on the input side of the adaptive block 204 and the gain in the ratio J / N are TIFF2025517890000012.tif12159.

[0030] Figure 3 shows a graph 300 of the reduction of the noise spectral density by the band-pass filter 202. As shown in Figure 3, the input noise signal 302 has a uniform power spectral density (PSD) distribution over the frequency range from -80 MHz to 80 MHz. The band-pass output signal 304 has a resonance response centered around 23 MHz.

[0031] The band-pass filter on the input side of the adaptive block should be tunable because, in the adaptation of the notch filter, its pole frequencies change in synchronization with the frequency of the interference signal. However, the adjustment of the digital filter requires real-time recalculation of the transfer function coefficients. For example, according to one embodiment, as the resonance frequency ω i = 2πf i , a biquadratic filter is used as the band-pass filter with a Q value of Q. After the bilinear frequency transformation of the biquadratic, which is an analog embodiment TIFF2025517890000013.tif12159, the digital biquadratic transfer function in the z domain has the following form. TIFF2025517890000014.tif13159 Here TIFF2025517890000015.tif13159 it is.

[0032] To adjust the resonance frequency of the digital filter by the transfer function (Equation (11)), the coefficients a 1 and a 2 are changed. In one embodiment, the zero frequency of the transfer function of the digital complex filter is equal to the interference frequency when the adaptation is completed.

[0033] Furthermore, after frequency conversion, distortion occurs in the frequency characteristics, and the level thereof increases as the resonance frequency approaches the Nyquist frequency. The value of the resonance frequency shifts from the calculated value. To ensure the accuracy of the digital filter at the frequency ω, predistortion (see Equation (8)) according to TIFF2025517890000016.tif12159 is introduced. In this case, the coefficient β in Expression (11) should take the form of TIFF2025517890000017.tif12159.

[0034] As a result, when synchronizing the zero frequency of the notch filter in the adaptation process, it is required to calculate the coefficients of the transfer function of the digital biquadratic filter according to the sampling frequency by the following equation. TIFF2025517890000018.tif15159

[0035] When implementing interference mitigation by a field programmable gate array (FPGA), relatively large computing resources are required for Equation (12).

[0036] A transfer function with real coefficients (see Equation (11)) generally has complex conjugate zeros and poles. Regarding the bilateral spectrum, the frequency response of a filter with real coefficients has a symmetric response in both the positive frequency region and the conjugate negative part. However, in principle, the interference signal is located in one of the two conjugate parts of the spectrum. Here, the band-pass filter with a real transfer function will have an excessive bandwidth in the conjugate part of the spectrum where the interference signal does not exist, and the value of the ratio J / N will be degraded by at least 3 dB.

[0037] FIG. 4 shows a filter section 400 of a GNSS receiver having a digital complex first-order band-pass filter according to an embodiment. The filter section 400 has a band-pass filter 406 on the input side of an adaptation block 404. The band-pass filter 406 has a complex transfer function, and the complex transfer function provides a selection in only one of the two conjugate parts of the bilateral spectrum.

[0038] In one embodiment, the band-pass filter 406 operates specifically as the following transfer function. TIFF2025517890000019.tif13159

[0039] In one embodiment, the band-pass filter 406 has the following characteristics.

[0040] The transfer function shown in Equation (13) corresponds to a digital complex first-order band-pass filter having complex poles TIFF2025517890000020.tif10159, where TIFF2025517890000021.tif10159 is the zero frequency of the transfer function of the notch filter 402 (see Equation (7)), that is, the value of the resonance frequency of the proposed band-pass filter coincides with the zero frequency of the transfer function of the notch filter 402.

[0041] The real coefficient k b <1 determines the bandwidth of the band-pass filter, and its numerical value is given by the formula B -3dB =(1 - kb )f S It can also be estimated by π / 10. If the value of the coefficient k b is closer to 1, the bandwidth of the predetermined band-pass filter becomes narrower accordingly.

[0042] Resonance frequency In TIFF2025517890000022.tif9159, the transfer coefficient of the proposed filter is as follows.

[0043] TIFF2025517890000023.tif11159 This is because TIFF2025517890000024.tif8159.

[0044] Finally, it is guaranteed that only one of the (two) conjugate parts of the bilateral frequency spectrum is selected by the band-pass complex filter operating according to the transfer function shown in Equation (13).

[0045] The band-pass filter in one embodiment consists of a digital complex first-order filter having a pole frequency that coincides with the zero frequency of the notch filter, and this digital complex filter has the transfer function of TIFF2025517890000025.tif13159.

[0046] FIG. 5 shows the spectrum 500 of the band-pass filter 406 that operates according to the transfer function shown in Equation (13) and has the characteristics described above.

[0047] The method for mitigating interference is generally limited to the method for mitigating only one type of interference. For example, the PB (pulse blanker) method works well for wideband pulse interference signals but is not designed to handle continuous interference. The multi-spectrum method using the discrete Fourier transform (DFT), discrete wavelet transform (DWT), or Karhunen–Loeve transform (KLT) provides a wide range of flexibility for mitigating interference but requires extremely large computational resources.

[0048] The adaptive notch filtering (ANF) method described herein provides favorable results for quasi-harmonic interference and chirp signals, but may be less effective in mitigating multi-spectrum or impulse interference signals. However, the effectiveness of the adaptive notch filter method can be increased as described below.

[0049] The notch-type digital complex first-order bandpass filter in one embodiment operates according to the transfer function of Equation (7). The digital complex first-order bandpass filter described herein does not require additional arithmetic resources for recomputing the transfer function and provides tracking and mitigation of chirp interference at a frequency rate up to 10 MHz / μs. The adaptive block in one embodiment is configured to track the interference frequency and adjust the zeros of the filter for the purpose of achieving minimization of the cost function. In the operation of the digital complex first-order filter, the real coefficient k α simultaneously affects the depth of signal mitigation and the stopband width. To mitigate wideband (e.g., multi-spectrum type OFDM signal) interference, the stopband width of the notch filter increases, i.e., k α decreases. However, in this case, the degree of signal attenuation in the stopband decreases. To resolve this contradiction, the order of the filter is increased to expand the stopband of the notch filter. One of two possible solutions can be implemented to resolve this contradiction.

[0050] In the first approach, a digital complex third-order filter composed of three notch filters connected in multiple stages is used, and as a result, transfer coefficients in the following form are obtained. TIFF2025517890000026.tif14159 Here TIFF2025517890000027.tif11159.

[0051] FIG. 6 shows a filter section 600 of a GNSS receiver including the above-described digital complex cubic filter having three notch filters 602A, 602B, and 603C connected in multiple stages. The GNSS receiver 600 further has an adaptive block 604 with a bandpass filter 606 on the input side.

[0052] FIG. 7 shows a graph 700 depicting the frequency response of the digital complex cubic filter of FIG. 6.

[0053] Note that when changing (i.e., tuning) the zero frequency f 0 of the notch filter during adaptation, it should be noted that there is no need to change the value of Δf 1 which determines the stop band. The values of the complex poles of these additional two first-order filters TIFF2025517890000028.tif9159 do not require complex calculations and are determined by the sum of the arguments of two complex numbers.

[0054] In the case of multi-spectrum interference having a wider bandwidth, generally, it is necessary to increase the order of the notch filter. Further, the positions of the zeros in the approximation function need to be symmetric around the center frequency. The transfer function of such a filter can be expressed as follows. TIFF2025517890000029.tif28159

[0055] The number and positions of the zeros (Δf 1 <Δf 2 <…Δf k ) are determined by the required depth of interference mitigation in the stop band of a 2(k + 1)-order notch filter. The frequency value f 0 corresponds to the center frequency of the interference spectrum, and it is required that the numerical value 2·Δf k - corresponds to the signal width of this interference.

[0056] In the second approach, for interference mitigation, a notch-type high-order complex filter is implemented in a high-pass filter with some real coefficients by shifting those real coefficients by multiplying them with a power function of a complex power exponent. The real coefficients of the impulse response A finite impulse response (FIR) filter with TIFF2025517890000030.tif12159, or a transfer function Either an infinite impulse response (IIR) filter with TIFF2025517890000031.tif16159 can be used as a high-pass filter. A high-pass filter with real coefficients provides signal attenuation for the bilateral spectrum within the band [-ω c , +ω c .

[0057] FIG. 8 shows a graph 800 of the frequency response of a digital complex notch filter by shifting a high-pass filter with some real coefficients.

[0058] These coefficients of the complex filter are further formed by multiplying some coefficients of the original high-pass filter with a power function of a complex power exponent in TIFF2025517890000032.tif11159 for the FIR filter, or with a power function of a complex power exponent in TIFF2025517890000033.tif11159 for the IIR filter (for all m = 0..(M 1(2) - 1)).

[0059] As a result, the transfer function of the complex IIR filter is obtained as TIFF2025517890000034.tif16159, or the impulse response characteristics of the complex FIR filter are obtained as TIFF2025517890000035.tif11159.

[0060] When the notch filter is shifted in the adaptation process, TIFF2025517890000036.tif9159 is used as a complex power exponent, where ω 0 - is the center frequency of the complex filter. The notch filter synthesized by shifting the frequency characteristic to the center frequency ω 0 follows the approximation function of the original high-pass filter and guarantees attenuation within the band [ω c - ω c , ω 0 + ω c . By selecting the high-pass filter, the width of the selection required for the frequency response is determined. If the value of the width of the transition region of the frequency response becomes smaller, the distortion of the useful signal is thereby reduced.

[0061] The method in one embodiment can be used for the detection and mitigation of multi-spectrum interference (orthogonal frequency division multiplexing (OFDM) signals) when using a high-order notch filter in an ANF structure.

[0062] The approaches and techniques described here provide not only the detection and mitigation of single-component interference (such as CW and chirp), but also the detection and mitigation of multi-spectrum interference, thereby having significant advantages over other studies focused on ANF-based interference mitigation systems.

[0063] The techniques described here can be used not only for the detection and mitigation of single-component interference (such as CW and chirp), but also for countermeasures against multi-spectrum wideband interference.

[0064] According to one embodiment, a computer is used to implement the operations and mathematical formulas of the components described herein, for example, as shown in FIGS. 1, 2, 4, and 6. These components may include, for example, notch filters, bandpass filters, and adaptive blocks. A high-level block diagram of this computer is shown in FIG. 9. Computer 902 includes a processor 904 that controls the overall operation by executing computer program instructions that define the overall operation of computer 902. The computer program instructions are stored in a storage device 912 or other computer-readable medium (such as a magnetic disk, CD ROM, etc.) and are loaded into memory 910 when it is desired to execute the computer program instructions. Thus, the components and mathematical formulas described herein are defined by computer program instructions stored in memory 910 and / or storage device 912 and can be controlled by processor 904 that executes the computer program instructions. For example, the computer program instructions can be implemented as computer-executable code programmed by one skilled in the art to execute the algorithms defined by the components and mathematical formulas described herein. In this way, processor 904 executes the algorithms defined by the components and mathematical formulas described herein, such as the components shown in FIGS. 1, 2, 4, and 6 herein. Computer 902 further has one or more network interfaces 906 for communicating with other devices via a network. Computer 902 further has an input / output device 708 (such as a display, keyboard, mouse, speaker, button, etc.) that enables two-way operation of the user with computer 902. Those skilled in the art will recognize that actual computer implementations may include other components, and FIG. 9 represents a high-level view of some of the components of such a computer for illustrative purposes.

[0065] The above detailed description is to be understood as illustrative and not restrictive in all respects, and the scope of the inventive concept disclosed herein should be construed by the broadest width permitted by each patent law. Each embodiment shown and described herein is merely illustrative of each principle of the inventive concept, and those skilled in the art will understand that various modifications can be implemented without departing from the scope and gist of the inventive concept. Those skilled in the art will consider that various other combinations of features can be implemented without departing from the scope and gist of the inventive concept.

Claims

1. A notch filter having a zero frequency of a transfer function capable of being synchronized, configured to receive an input signal and generate an output signal; A band-pass filter connected to the output of the notch filter and configured to receive the output signal; An adaptive block connected to the band-pass filter and configured to adjust parameters of the notch filter to minimize a specific cost function; An apparatus comprising the above.

2. The notch filter is a first-order digital complex filter, The input of the digital complex filter is connected to the input of the apparatus, The apparatus according to claim 1, wherein the frequency of the zero of the transfer function of the digital complex filter is equal to the interference frequency when adaptation is completed.

3. The band-pass filter is a first-order digital complex filter having a pole frequency that coincides with the zero frequency of the notch filter and a transfer function The apparatus according to claim 1, wherein the input of the digital complex band-pass filter is connected to the output of the apparatus.

4. The adaptive block is configured to track the interference frequency and adjust the zero of the filter to achieve minimization of the cost function, The apparatus according to claim 1, wherein the input of the adaptive block is connected to the output of the band-pass filter.

5. Configured to mitigate multi-spectrum interference, The apparatus according to claim 1, wherein the notch filter has a transfer function The apparatus according to claim 1, wherein the notch filter is implemented by multiplying the real coefficients of its transfer function by a power function of a complex exponential and then shifting a high-pass filter.

6. Configured to mitigate multi-spectrum interference, The apparatus according to claim 1, wherein the notch filter is implemented by multiplying the real coefficients of its transfer function by a power function of a complex exponential and then shifting a high-pass filter.

7. Receiving an input signal from one or more artificial satellites of a global positioning satellite system; Filtering the input signal using a first transfer function of a notch filter to generate a filtered signal; Filtering the filtered signal using a second transfer function of a band-pass filter to generate a band-pass filtered signal; and Tracking the interference frequency of the band-pass filtered signal. A method comprising the above.

8. The notch filter is a first-order digital complex filter, ​ The input of the digital complex filter is connected to the input of a device that receives the input signal, The method according to claim 7, characterized in that the frequency of the transmission function zero of the digital complex filter is equal to the interference frequency when adaptation is completed.

9. The band-pass filter is a first-order digital complex filter, The digital complex filter has a pole frequency that coincides with the zero frequency of the notch filter, and the digital complex filter The method according to claim 7, characterized by having a transfer function of.

10. The method according to claim 7, further comprising tracking an interference frequency and adjusting the zeros of the filter to achieve minimization of a cost function.

11. The notch filter has the following transfer function The method according to claim 7, characterized by having.

12. The method according to claim 7, characterized in that the notch filter is a high-pass filter shifted using a real coefficient multiplied by a power function of a complex power exponent for real coefficients.

13. A device comprising a processor and a memory connected to the processor, wherein the memory causes the processor to perform the following operations during execution, namely receiving an input signal from one or more artificial satellites of a global positioning satellite system, filtering the input signal using a first transfer function of a notch filter to generate a filtered signal, filtering the filtered signal using a second transfer function of a band-pass filter to generate a band-pass filtered signal, and tracking the interference frequency of the band-pass filtered signal, A device characterized by storing computer program instructions for causing the above to be carried out.

14. The notch filter is a first-order digital complex filter, The device according to claim 13, characterized in that the zero frequency of the transfer function of the digital complex filter is equal to the interference frequency when adaptation is completed.

15. The band-pass filter is a first-order digital complex filter, The digital complex filter has a pole frequency that coincides with the zero frequency of the notch filter, and the digital complex filter The device according to claim 13, characterized by having a transfer function of.

16. The apparatus according to claim 13, wherein the operation further comprises tracking an interference frequency and adjusting a zero frequency of the notch filter to achieve minimization of a cost function.

17. The notch filter has the following transfer function The apparatus according to claim 13, characterized in that it has the following transfer function.

18. The apparatus according to claim 13, wherein the notch filter is characterized in that real coefficients of the transfer function of the band-pass filter are shifted by an interference frequency multiplied by a power function of a complex power exponent.

Images