Sensing Fractional Power Spectrum of Nonstationary Signals with Coprime Filter Banks

Li, Xiaomin; Wang, Huali; Lv, Wanghan; Luo, Haichao

doi:https://doi.org/10.1155/2020/2608613

Complexity

On this page

Abstract Introduction Preliminaries Discussion Conclusion Appendix Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2020 | Article ID 2608613 | https://doi.org/10.1155/2020/2608613

Sensing Fractional Power Spectrum of Nonstationary Signals with Coprime Filter Banks

Xiaomin Li,^1,2Huali Wang ,³Wanghan Lv,⁴and Haichao Luo⁵

Academic Editor: Yan-Ling Wei

Received29 Feb 2020

Revised17 Jun 2020

Accepted29 Jun 2020

Published17 Jul 2020

Abstract

The coprime discrete Fourier transform (DFT) filter banks provide an effective scheme of spectral sensing for wide-sense stationary (WSS) signals in case that the sampling rate is far lower than the Nyquist sampling rate. And the resolution of the coprime DFT filter banks in the Fourier domain (FD) is , where and are coprime. In this work, a digital fractional Fourier transform- (DFrFT-) based coprime filter banks spectrum sensing method is suggested. Our proposed method has the same sampling principle as the coprime DFT filter banks but has some advantages compared to the coprime DFT filter banks. Firstly, the fractional power spectrum of the chirp-stationary signals which are nonstationary in the FD can be sensed effectively by the coprime DFrFT filter banks because of the linear time-invariant (LTI) property of the proposed system in discrete-time Fourier domain (DTFD), while the coprime DFT filter banks can only sense the power spectrum of the WSS signals. Secondly, the coprime DFrFT filter banks improve the resolution from to by combining the fractional filter banks theory with the coprime theory. Simulation results confirm the obtained analytical results.

1. Introduction

Power spectrum plays an important role in signal processing such as array processing [1–4], spectral estimation [5–7], signal detection and estimation [8–12], and so on. As the traditional methods of sensing power spectrum are associated with wide-sense stationary (WSS) signals whose second-order expectations remain unchanged over time, the conventional methods are under Fourier analysis [13–19], and all these developments have led to satisfactory results.

Compared with the WSS signals, the nonstationary signals are characterized by being oriented along an oblique axis in the time-frequency domain while the stationary signals are projected in a horizontal direction. And as a spot treatment in the frequency domain, the conventional Fourier analysis does not allow associating events in the time domain with the frequency domain, so Fourier analysis is not a useful tool for processing nonstationary signals.

As a generalized form of the Fourier transform (FT), the fractional Fourier transform (FrFT) [20, 21] allows the signals in the time-frequency domain to be projected onto a line of arbitrary angle. In contrast to the standard Fourier analysis, the fractional Fourier analysis has a notable potential in the treatment of nonstationary signals due to an additional degree of freedom. The analysis of nonstationary signals by means of the fractional Fourier transformation has been reported [22–24]. And many signal processing theories in the Fourier domain (FD) have been extended to the fractional Fourier domain (FrFD) based on the relationship between FrFT and FT, such as the filters theory [25–29], the correlation theory [30–32], and the power spectral density theory [33, 34].

Meanwhile, it is necessary to sense a wide band of power spectrum, leading to prohibitively high Nyquist rates which will exceed the specifications of best commercial analog-to-digital converters (ADCs). Therefore, designing a sub-Nyquist sampling scheme which can effectively sense the power spectrum of nonstationary signals is a challenging task. Coprime theory [35, 36], which is well suited for analyzing sparsely sampled signals in case that sampling rate is far lower than the Nyquist sampling rate, has gained increasing attention in recent years. Coprime theory can be well coupled with DFT filter banks theory in temporal domain to sense the power spectrum of WSS signals [35]. In [35], the WSS premise of the input signals ensures the decimate operation in cross-correlation that can be coupled with coprimality well (see (13)); therefore, when the inputs are nonstationary, i.e., the second-order expectations of the input sequence change over time [37–40], the decimate operation cannot be performed to introduce the coprime theory in, resulting in inefficient sensing of the power spectrum for nonstationary signals. Thus, the coprime DFT filter banks [35] is not suitable to process nonstationary signals.

Recently, several research efforts based on the coprime DFT filter banks [35] are developed from different aspects. Huang et al. [41] applied two simple operations on each low-rate data channel of the original coprime DFT filter banks to form a 0.5 frequency resolution shifted analyzer and synthesized it and the original coprime DFT filter banks to remove all the annoying spurious peaks. However, many decision thresholds were needed to set in the modified analyzer, so the modified analyzer is not suitable for engineering applications. In an attempt to completely remove the spurious-peak side effect, Xiangdong Huang and Yuxuan [42] proposed a high-performance coprime spectral analysis method based on paralleled all-phase point-pass filtering; the proposed method can not only completely remove the spurious-peak side effect, but also can be implemented easily in the real-world application. Vaidyanathan and Pal [43] extended one-dimensional coprime DFT filter banks to multiple dimensions. Wu and Liang [44] used short-time Fourier transform to segment the nonstationary signals into piecewise stationary signals. However, the window function in short-time Fourier transform has the trade-off between temporal resolution and frequency resolution, resulting in low estimation accuracy for the power spectrum.

This paper aims to develop a coprime digital fractional Fourier transform (DFrFT) filter banks theory for sensing the fractional power spectrum of the chirp stationary signals, which are nonstationary in the usual sense or in the FD [22, 33]. We describe how to construct a sub-Nyquist system by using two low-speed coprime sampling ADCs and two DFrFT filter banks and derive the mathematical expression of the output, which shows that the proposed system can effectively sense the fractional power spectrum of nonstationary signals without loss of any information.

First, following the coprime premise and the property of the low-pass filter in discrete-time fractional Fourier domain (DTFrFD), we construct the coprime pair of DFrFT filter banks and prove the uniqueness of its passband in corresponding DTFD (with its argument scaled by ) (see Section 3.3 for details). Second, we prove that the coprime pair of DFrFT filter banks is linear time invariant (LTI) in DTFD based on the low-pass filter theory in DTFrFD, as the corresponding chirp modulated forms of the nonstationary signals are stationary, the decimate operation can be performed in the cross-correlation analysis between the outputs of the coprime pair of DFrFT filter banks, and the conventional power spectrum of the stationary signals can be acquired based on the uniqueness of the passband (see section 3.4 for details). Third, in terms of the polyphase representation of the filters in DTSFrFD, the decimator at the output of the filters can be moved to the left of the polyphase subfilters, resulting in an efficient coprime DFrFT filter banks (see Section 3.5 for details). Further, based on the fractional power spectrum theory, the final fractional power spectrum of the nonstationary signals can be acquired by the proposed coprime DFrFT filter banks. Besides, when the input is deterministic signal, the cross-correlations give the sensing of the spectrum in DTFrFD based on the fractional convolution theory (see Section 4.1 for details). The modified architecture not only senses the nonstationary signals but also has an accurate resolution in DTFD.

The outline of this paper is as follows. In Section 2, the problem formulation is introduced based on the basic preliminaries. In Section 3, the coprime DFrFT filter banks for sensing the fractional power spectrum of nonstationary signals is elaborated. In Section 4, the special cases for deterministic signals are discussed, and the performance of the proposed method is analyzed in terms of the resolution and the fractional spectrum estimation.

2. Preliminaries

2.1. DTSFrFT

The FrFT is a generalization of the FT, which essentially allows the signals in the time-frequency domain to be projected onto a line of arbitrary angle [21, 22]. Simplified fractional Fourier transform (SFrFT) [45] has the same effect as FrFT of order for filter design, but for digital implementation, it is simpler than the original FrFT. In digital signal processing systems, the signals used are digital signals sampled from the analog signals; their representations in simplified fractional Fourier domain (SFrFD) should be obtained by discrete-time simplified fractional Fourier transform (DTSFrFT). The DTSFrFT of is defined as follows [45]:where is the digital frequency of , which is the variable in SFrFD, and is the sample spacing in temporal domain.

And we havewhere is the DTSFrFT of and is the DTFrFT of .

2.2. The Low-Pass Filter in DTSFrFD

Suppose is a low-pass filter with length , and its frequency response is , i.e.,where is the digital frequency in discrete-time Fourier domain (DTFD) and is the cutoff frequency. Let , and its fractional frequency response is , [26], i.e.,where is the digital frequency in DTSFrFD and is the frequency response of with its argument scaled by . is defined as the fractional frequency response of low-pass filter in DTSFrFD. Equation (4) shows that , the low-pass filter in DTSFrFD, is equivalent to the discrete-time Fourier transform (DTFT) of , with its argument scaled by . So, is linear time invariant (LTI) in DTFD. And this relation is the instrument for the signal polyphase representation and filter bank theorems in DTSFrFD as will be discussed later.

2.3. The Chirp-Stationary Signal

Definition 1. (the chirp-stationary signal). For a nonstationary random signal , if the chirp modulated form of is stationary, i.e., , where is the autocorrelation of , the signal is called the chirp-stationary signal.
And we have an important result for .

Theorem 1. (see [34]). The relationship between the fractional power spectrum of and the conventional power spectrum of iswhere is the conventional power spectrum of stationary signal and is the fractional power spectrum of nonstationary signal .

2.4. Convolution Theorem in DTSFrFD

For any discrete-time sequences and , the discrete-time fractional Fourier convolution is as follows [32]:where denotes the fractional convolution operator. denotes the traditional convolution operator. is the sample spacing in temporal domain. The discrete-time fractional Fourier transform (DTFrFT) of can be denoted as follows:where , , and are the DTFrFT of , , and . And the DTSFrFT of can be denoted as follows:where , , and are the DTSFrFT of , , and . Combining equations (6) and (8) illustrates that the discrete-time fractional Fourier convolution of and is corresponding to the product of and in DTSFrFD. Similarly, suppose .

Then,where , are the DTFrFT of , and is the DTSFrFT of .

2.5. Original Coprime DFT Filter Banks

The coprime pair of DFT filter banks introduced in [35] which is illustrated in Figure 1 has a unique passband with the width and the center in the Fourier domain (FD) (see [35] for details). As a result, any stimulus component near , , can be uniquely identified by the product filter banks.

In Figure 1, the cross-correlation between and iswhere is the power spectrum of which is the common WSS input of the two filter banks and and are the frequency response of filters and , respectively, which are acquired through applying zooming and shifting operations on the transfer curve of ideal lowpass filters (or ) using the shift parameter (or ). In terms of the properties of the unique passband in the FD, equation (10) can be approximately written aswhere is a constant which depends on the exact passband shapes and on the coprime pair and and . Further, based on the wide-sense stationary (WSS) premise of input , time domain averages can be used to estimate correlations:and since and are outputs of LTI systems with a common WSS input , they are jointly WSS, sowhere and are coprime. That is, we can simply multiply decimated versions of and and perform the averaging. Then, based on the polyphase forms of filters and , the decimator can be combined with and to realize the coprime decimate for input signal as shown in Figure 2. Besides, in Figure 2, and can be, respectively, acquired by employing two analog-to-digital converters (ADCs) with the sampling rates , to discretize a continuous signal .

3. The Proposed Architecture

3.1. Problem Formulation

From the above description in Section 2.5, the coprime DFT filter banks architecture [35] as shown in Figure 2 which can sense the power spectrum of stationary signals effectively is constructed based on two points: the passband uniqueness and WSS premise. The passband uniqueness ensures that any stimulus component near the passband can be uniquely identified by the product filter banks, i.e., the equivalence relationship in equation (11). And the WSS premise guarantees the decimated operation in equation (13) which can be combined with the polyphase representation of the filters to achieve the coprime sampling structure as shown in Figure 2. Accordingly, when the input signal is not WSS, the decimated operation cannot be performed so that the coprime theory cannot be coupled with DFT filtering bank theory to derive the architecture in Figure 2.

3.2. System Description

In this paper, we introduce the fractional Fourier analysis theory (the fractional filters theory and the fractional power spectrum theory) into the original coprime DFT filter banks to derive a modified architecture for sensing the fractional power spectrum of the chirp-stationary signals. First, we construct the coprime pair of DFrFT filter banks which is shown in Figure 3 and prove the uniqueness of their passband in DTFD (see Section 3.3 for details). Second, we implement the cross-correlation between the outputs of the two filter banks; then based on the properties of the low-pass filter in DTSFrFD introduced in Section 2.2 and the uniqueness of the passband, the cross-correlation analysis can be converted from DTFrFD into DTFD. As the coprime pair of DSFrFT filter banks is LTI in corresponding DTFD (see Section 3.4 for details) and the input signal is stationary when it is modulated by a corresponding chirp signal, the decimated operation can be performed successfully. Third, in terms of the polyphase representation of the filters in DTSFrFD (see Section 3.5 for details), the decimator at the output of the filters can be moved to the left of the polyphase subfilters, resulting in an efficient polyphase implementation. Further, based on Theorem 4 (see Section 3.6 for details), the final fractional power spectrum of the nonstationary signals can be acquired. And the proposed coprime filter banks is shown in Figure 4.

3.3. The Coprime Pair of DFrFT Filter Banks and the Uniqueness of Their Passband

We specifically take advantage of the coprimality of and to construct a coprime pair of DFrFT filter banks by combining an band DFrFT filter bank and an band DFrFT filter bank. And the proposal is illustrated in Figure 3. In this system, the two filter banks are composed of and subfilters , , respectively, which can be acquired through applying zooming and shifting operations on the transfer curve of the ideal low-pass filter , in DTSFrFD. And we will prove in the following that the product filter has a unique passband with the width and the center .

Assume that is the ideal low-pass filter in DTSFrFD.

And is the digital frequency in DTSFrFD; is the digital frequency in DTFD. Thus, has passbands in DTFD, with each passband having width . , , where , can be rewritten as

These are shifted version of , in increments of , for each in , has passbands, each passband has width , and the passbands are centered at .

Similarly, assume that is the ideal low-pass filter in DTSFrFD.

Thus, has passbands in DTFD, with each passband having width . , , where , can be rewritten asthese are shifted version of , in increments of , for each in , has passbands, each passband has width , and the passbands are centered at .

For the two filter banks , , , , now consider the product of the and responses and :and there are two important theorems as follows:

Theorem 2. Given any integer in , there is a unique , i.e., a unique pair, with passband centered at .

Theorem 3. has precisely one passband, and it has width . That is, there is only one overlapping band among the bands of and bands of . This overlapping passband is centered at for some integer in . The proofs are in Appendix A and Appendix B.
Figure 5 demonstrates how the filters work for and . As shown in Figure 5, the filter has three passbands, and each has width . There are four distinct shifted versions of , . The filter has four passbands, and each has width . There are three distinct shifted versions of , . Each shifted version overlaps with any shifted version in precisely one passband. Therefore, the product is bandpass with a single passband, having width . Furthermore, the twelve combinations of produce twelve distinct filters , covering . Therefore, we get the effect of an -band filter bank in DTFD by combining one -band filter bank with one -band filter bank.

3.4. The Cross-Correlation Analysis

In this part, we will analyze the cross-correlation between the outputs of the two filter banks and in Figure 3 to acquire their corresponding decimated versions, which will be combined with the polyphase representation of DSFrFT filters in Section 3.5 to form the structure in Figure 4.

Assume the two filter banks and in Figure 3 have a common chirp-stationary random sequence as the input, with sample spacing ; let ; as known in Definition 2 in Section 3.6, is the traditional WSS sequence with autocorrelation , . The two filter banks , , and , , have , outputs, respectively. Now take the output of the filter bank and the output of the filter bank . The cross-correlation between and is

Letting , we have , where is the autocorrelation of , . Since ,

Parseval’s relation yieldswhere is the power spectrum of . Since and are coprime, the product filter has a single passband, with width , and center at , for some integer in . For large , the product is a very narrow band filter, and soand is a constant which depends on the exact passband shapes, and . Assuming the WSS process of is ergodic, we use time domain averages to estimate correlations:where refers to the number of consumed snapshots. As can be seen from Section 2.2, for the low-pass filter in DTSFrFD, there is , where is the DTFT of . Thus, , the low-pass filter in DTSFrFD, is equivalent to , which is the low-pass filter in DTFD. Accordingly, is LTI in DTFD. And filter banks and are LTI in DTFD. Therefore, and are outputs of LTI systems with a common WSS input ; they are jointly WSS, sofor any integer . That is, we can simply multiply decimated versions of and and perform the averaging.

3.5. The Polyphase Representation of DSFrFT Filters

Suppose is the -type polyphase component of which has been defined in Section 2.2,i.e.,where .

Substituting (24) into (4),

Equation (25) is the equivalence polyphase representation of in DTFD. In most applications, the decimator is preceded by a low-pass digital filter called the decimation filter. And according to equation (25), the equivalent polyphase implementation of low-pass decimation filter in DTFD is illustrated in Figure 6.

Accordingly, since

So, we can represent the filters in appropriate polyphase forms:

Figure 7 shows the polyphase forms of the filters in DTFD. Similarly, the polyphase forms of the filters areand its polyphase forms in DTFD are shown in Figure 8. And we have the noble identities of decimator [46] which is shown in Figure 9. Such interconnections arise when we try to use the polyphase representation for decimation filters. And we can redraw Figure 6 as Figure 10. Accordingly, combining with the analysis in Section 3.4, when decimation ratio , an efficient polyphase implementation coprime DSFrFT filter banks under coprime sampling is shown in Figure 4. Besides, in Figure 4, and can be, respectively, acquired by employing two ADCs with the sampling rates , to discretize a continuous signal .

3.6. The Fractional Power Spectrum of Nonstationary Signals in DTSFrFD

In this section, we will prove that the output of the architecture in Figure 4 is the fractional power spectrum of nonstationary signals . To make it more clear, we first define the chirp stationarity random sequence.

Definition 2 (the chirp-stationary random sequence). For nonstationary random sequence , if is a stationary random sequence, we call the chirp-stationary random sequence.
And a useful conclusion associated with the fractional power spectrum of is shown here for convenience.

Theorem 4. The relationship between the fractional power spectrum of the chirp-stationary random sequence and the conventional power spectrum of the stationary random sequence isin which is the conventional power spectrum of in DTFD and is the fractional power spectrum of in DTSFrFD. See Appendix C for the certification process.

Therefore, as can be seen from equation (30), except for a constant coefficient, has the same spectral shape with . So, a traditional spectral estimate of the chirp-stationary sequence is transformed into the simplified fractional spectral estimation of the nonstationary sequence . So, equation (23) can be rewritten aswhere is a constant which depends on the exact passband shapes, and . That is, in DTSFrFD, the output of the system as shown in Figure 4 is the fractional power spectrum of multiplied by a constant.

3.7. Choice of the Transform Orders and

As shown in Figure 4, there are two transform orders and to be determined. The transform order is used to convert nonstationary signals into stationary signals, while decides the resolution of the proposed system which partially influences the sensing accuracy.

The received chirp-stationary signal is composed of a modulated signal and noise. Its model is given bywhere and are received signal and modulated signal, respectively. is assumed to be additive white Gaussian noise.

The modulated signal is given by , where is the amplitude, and and are the carrier frequency and the initial uniformly distributed random phase, respectively. is the Nyquist sampling rate. is the phase function, which determines the modulation type of the signal. For simplicity and without loss of generality, we assume that is an invariant constant.

In Definition 1, we know that for the chirp-stationary signal , its chirp modulated form is stationary, i.e., ; therefore, the transform order .

In the fractional Fourier domain (FrFD), support of signals’ change is associated with the transform order, and there exists an optimum transform order in which the energy of signals is maximally concentrated [30–32]. When a signal is transformed by FrFT at its optimum order, transform kernel acts as a matched filter. Therefore, the transform order is equal to the optimal transform which can maximize the absolute amplitude, i.e., .

Furthermore, since the energy of Gaussian noise signals cannot effectively concentrate in FrFD, the Gaussian noise can be suppressed effectively.

The optimal transformation order corresponding to maximum magnitude obtained from the FrFT is given by ,

Where is the FrFT of the stationary signal in the -order FrFD. And will directly affect the sensing performance; hence, the method of searching the optimal order is important. The traditional method to get the optimal order in the DTSFrFD is peak sweeping method [32, 33], which is an easy method to realize. And obviously, the search-based algorithms require numerous extra calculations and have the contradiction between estimation performance and complexity.

In this section, we introduce the normalized second-order central moment (NSOCM) calculation method [47] to directly obtain the optimal transformation order in DTSFrFD. Compared with the search-based algorithms, the NSOCM approach has higher computational efficiency because of its nonergodic search mechanism.

According to [47], the optimal order is normally given bywhere is the time-bandwidth product for , and the NSOCM of is defined bywhere is the normalized first-order origin moment of and is the normalized second-order origin moment of . The NSOCM represent the time width and frequency width of , respectively. Hence, equation (33) becomes

The NSOCM product is given bywhere is the mixed second-order moment. Setting the first derivative of with respect to the order equal to zero, we obtain

For this case where is equal to the extreme point , the product reaches the extremum values. This result demonstrates that when satisfies equation (38) as follows, the product reaches its minimum.

Based on the theoretical analysis above, the calculation process of the optimal transform order can be summarized into the specific procedures as follows:(1)Take the 0.5^th- and 1^st-order DTSFrFT of signal to obtain , (2)Calculate the normalized first-order origin moments and , the normalized second-order origin moments , and , the mixed second-order moment , and NSOCM and in accordance with the definition(3)Obtain the optimal order of by using (38) in the range of .

And the flowchart is shown in Figure 11.

4. Discussion

4.1. Special Cases for Deterministic Signals

The above system as shown in Figure 4 is suitable for the chirp-stationary random signals which have random characters. We have analyzed the output of the system and obtained the conclusion that the proposed system can sense the fractional power spectrum of the chirp-stationary random signals effectively. When the input signals of the system are deterministic signals which have compact support in the FrFD, the cross-correlation analysis is equivalent to the convolution theorem in the sense of FrFT. And special forms of the outputs occur.

When is a deterministic signal, suppose that is the output of filter bank , is the output of filter bank , , and the two filter banks have a common input . Then, the product of the multiband filters and is

According to the fractional convolution theorem in Section 2.4, the spectrum of in FrFD is

According to the convolution theorem in FrFD, the spectrum of in FrFD is

So, the final output of the system iswhere denotes the traditional convolution operator.

The spectrum of in FrFD is

According to (2), there isand since , we havethat is, in DTSFrFD, the system output which is shown in Figure 12 is the square of the fractional spectrum of deterministic signal . And this conclusion can be used to estimate the fractional spectrum .

4.2. Simulation

We demonstrate the simulations for three different cases: resolution performance, the sensing accuracy, and the influence of the order .

4.2.1. Resolution Performance

In this experiment, we will examine the resolution performance of our proposed method. The traditional DFT coprime filter banks structure [35] and the conventional polyphase filter banks [48] are given for comparison. The original multiband signal is denoted by . And the noisy signal is , where is white Gaussian noise. And is given by the following:

The choices of the parameters are listed in Table 1.

The optimal orders [34] of and are , ; and have the best energy concentration at orders and , respectively. For the system shown in Figure 4, we set and . And the sampling frequency is . The individual filter banks could only have resolved separations and , respectively. However, the coprime DFrFT filter banks could have resolved separations , which is more effective than the individual filter banks. The output of the original system [35] is shown in Figure 13(a); the fractional spectrum estimation in the FrFD is plotted in Figure 13(b). And it can be seen from Figure 13 that the outputs of the original system are overlapped in FD while the outputs of the proposed system are separated from each other in DTSFrFD with order .

(a)

(b)

Figure 14 shows the resolution for different numbers of channels. In Figure 14, the number of channels (i.e., ) varies from 10 to 50 with a step of 10. And appropriate and which are coprime are selected corresponding to every number of channels to reach the maximum resolution. For example, when the number of channels is 10, there are three coprime pairs: (1, 9), (3, 7), and (5, 5), and we set and to get the maximum resolution. The maximum resolution of the proposed system is while the coprime DFT filter banks is . It is observed that in Figure 14, the proposed system has the best performance because the parameter is introduced into the system by DTSFrFT as analyzed in Section 3.3. Figure 14 also shows that our method has better performance when the number of channels increases. Thus, in practical applications, the proposed system is more cost-effective than the traditional coprime DFT filter banks.

4.2.2. Spectrum Sensing Accuracy

In the second experiment, we test the spectrum sensing accuracy of our proposed coprime DFrFT filter banks, compared with traditional coprime DFT filter banks [35], spectrum reconstruction technique [49], and non-undersampling method [50].

To evaluate the sensing accuracy of the proposed system as shown in Figure 4, a chirp-stationary random signal with the initial uniformly distributed random phase is considered. And is given by the following:where is the amplitude of the signal which could be fixed. is the time scale factor which determines the signal duration. is the signal modulation rate. is the frequency carrier. is a rectangular time-window denoted by , if ; else, . The choices of the parameters for are listed in Table 2.

We specify two coprime integers and , and based on the definition of chirp-stationary signal, we can calculate the fractional order and the optimal transformation order . The SNR ranges from 0 dB to 20 dB with step size 2 dB. Here, we define the spectrum sensing accuracy aswhere and are the estimated lower boundary and upper boundary in the j^th Monte Carlo simulation, respectively. and are the true lower boundary and upper boundary. Here, 200 Monte Carlo simulations are used.

The original input signal in the temporal domain is shown in Figure 15(a), and the outputs in the FrFD are plotted in Figure 15(b). From the results in Figure 15, the fractional powers spectrum of the chirp-stationary signal can be sensed effectively by the proposed method. And the simulation results are consistent with the analysis in Section 3.

(a)

(b)

Figure 16 shows the sensing accuracy for different SNRs. In Figure 16, non-undersampling method has the best performance because it uses all Nyquist samples in signal processing. However, the results do not preclude the usefulness of the proposed method because the sampling rate of our method is far lower than the Nyquist sampling rate which will greatly reduce the pressure of ADC. Figure 16 also shows that our method has better performance for sensing chirp-stationary random signal than the traditional DFT coprime filter banks. This is because the traditional DFT coprime filter banks is suitable for sensing the power spectrum of wide-sense stationary (WSS) signals, while our proposed system has the ability of sensing the power spectrum of nonstationary signals. The sensing accuracy of the spectrum reconstruction technique improves gradually with the increase in SNR. However, As plotted in Figure 16, our proposed coprime DFrFT filter banks outperform the spectrum reconstruction technique. The explanation for this phenomenon is that cross-interference during reconstruction is so severe that the sensing accuracy is greatly affected by it.

4.2.3. Simulation of Influence of the Order

According to the analysis in Section 3.7, the -order SFrFT can be regarded as the projection on the rotated frequency axis ; consequently, the spectrum distribution of the signal directly depends on the transform order of the SFrFT. When the transform order is equal to the optimal transformation order , the energy of signals are concentrated maximally, and the fractional power spectrum is a pulse, which maximizes the sensing accuracy. Correspondingly, an inaccurate order may lead to changes in the spectral width. If the bandwidth of the real spectrum is wider than the bandwidth of the theoretical order, then the real maximum bandwidth will be bigger than the sampling rate. This will result in spectrum aliasing. The spectrum sensing accuracy will rapidly decrease.

In the last experiments, we demonstrate the influence of the order on the spectrum sensing accuracy. The tested signal is the same as that in Section 4.2.2, and we show the balance between the sensing accuracy and the transform order in the noise-free and noisy cases. Each simulation has 300 trials to ensure statistically stable results.

As the optimal transformation order of the tested signal is , we investigate the sensing accuracy under different where varies from to with a step of , the SNR is dB. It can be seen from Figure 17 that the closer to the transform order is, the smaller the estimation error is. Conversely, the closer to or , the larger the error is. Furthermore, the sensing accuracy in noise-free case is the best because that the signal is not polluted by the noise.

Figure 18 depicts the performance in terms of sensing accuracy with different SNRs and fractional order for the proposed method. The SNR varies from 0 dB to 20 dB with a step of 2 dB. The fractional order varies from to with a step of . It is observed that the sensing accuracy has the opposite trend as the SNR.

Furthermore, both Figures 17 and 18 indicate that if the transform order is close to the optimal order , the sensing accuracy will be improved particularly fast. Therefore, the proposed system is sensitive to the transform order , especially when the value of is near . So, how to enhance the robustness of the system is a problem that needs to be further studied.

5. Conclusion

This paper puts forward coprime DFrFT filter banks for sensing the fractional power spectrum of the nonstationary signals. We gave a deep analysis on the mechanism that the nonstationary signals cannot be effectively sensed by the original coprime DFT filter banks. By means of the relationship between FrFT theory and FT theory, the modified coprime DFrFT filter banks can sense the fractional power spectrum of the nonstationary signals precisely. Furthermore, the proposed coprime DFrFT filter banks obtain a better resolution scaled by than the original coprime DFT filter banks as well. And the special case for deterministic signals is also discussed based on the convolution theorem in DTSFrFD. Simulation results have shown that our proposed method outperforms traditional coprime DFT filter banks in sensing the fractional power spectrum of nonstationary signals. The further work should address the problem of the robustness of the system to the transform order .

Appendix

Proof of Theorem 2

Theorem 2. Given any integer in , there isaunique , i.e., a unique pair, with passband centered at .

Proof. The normalized band locations of and areand they can be written asGiven any integer in , there is a unique pair (quotient and remainder of modulo ) such that . Similarly, there is a unique pair such that . Thus, the filter has passband centered at .
Theorem 2 is proved.

Proof of Theorem 3

Theorem 3. has precisely one passband, and it has width . That is, there is only one overlapping band among the bands of and bands of . This overlapping passband is centered at for some integer in .

Proof. Assume the contrary. Thus, suppose has two passbands, centered at and . Thus,that is,However, since and , this contradicts the coprimality of and .
Thus, Theorem 3 is proved.

Proof of Theorem 4

Theorem 4. The relationship between the fractional power spectrum of the chirp-stationary random sequence and the conventional power spectrum of the stationary random sequence isin which is the conventional power spectrum of in discrete-time Fourier domain (DTFD) and is the fractional power spectrum of in DTSFrFD.

Proof. The autocorrelation of random sequence iswhich is stationary random sequence, sois just a function related to , and soSincewe haveThen, according to the definition of the power spectrum of the stationary discrete random process in DTFD, there isThus, we can observe from (C.8) that can be seen as a DTSFrFT of . According to the definition of discrete-time signal power spectrum at angle in FRFD,It can be seen that is the DTSFrFT of . Combining equations (C. 8) and (C. 9), for chirp-stationary random sequences at angle ,where is the power spectrum of and is the power spectrum of in DTSFrFD at angle . Substituting into equation (C.10),Thus, Theorem 4 is proved.

Nomenclature

FT:	Fourier transform
FD:	Fourier domain
DFT:	Discrete Fourier transform
DFrFT:	Digital fractional Fourier transform
DTFT:	Discrete-time Fourier transform
DTFD:	Discrete-time Fourier domain
FrFT:	Fractional fourier transform
FrFD:	Fractional fourier domain
SFrFT:	Simplified fractional fourier transform
SFrFD:	Simplified fractional fourier domain
DSFrFT:	Digital simplified fractional fourier transform
DTFrFT:	Discrete time fractional fourier transform
DTFrFD:	Discrete time fractional Fourier domain
DTSFrFT:	Discrete-time simplified fractional Fourier transform
DTSFrFD:	Discrete-time simplified fractional Fourier domain
WSS:	Wide-sense stationary
LTI:	Linear time invariant.

Data Availability

No data were used to support this study.

Conflicts of Interest

The authors declare that they have no conflicts of interest regarding the publication of this paper.

Acknowledgments

This study was supported by the National Natural Science Foundation of China (no. 61271354) and Henan Province Science and Technology Key Project (no. 142102210431).

References

Z. C. Yang, L. I. Xiang, and H. Q. Wang, “An overview of space-time adaptive processing technology based on sparsity of space-time power spectrum,” Acta Electronica Sinica, vol. 42, no. 6, pp. 1194–1204, 2014.
View at: Google Scholar
P. L. Shui and D. C. Li, “Floating small target detection in sea clutter via normalised doppler power spectrum,” IET Radar Sonar & Navigation, vol. 10, no. 4, pp. 699–706, 2016.
View at: Publisher Site | Google Scholar
G. Xiong, C. Xi, D. Li, and W. Yu, “Cross correlation singularity power spectrum theory and application in radar target detection within sea clutters,” IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 6, pp. 3753–3766, 2019.
View at: Publisher Site | Google Scholar
G. Xiong, C. Xi, J. He, and W. Yu, “Radar target detection method based on cross-correlation singularity power spectrum,” IET Radar, Sonar & Navigation, vol. 13, no. 5, pp. 730–739, 2019.
View at: Publisher Site | Google Scholar
D. A. Warde and S. M. Torres, “The autocorrelation spectral density for doppler-weather-radar signal analysis,” IEEE Transactions on Geoscience & Remote Sensing, vol. 52, no. 1, pp. 508–518, 2013.
View at: Publisher Site | Google Scholar
P. H. Y. Lee, J. D. Barter, K. L. Beach et al., “Power spectral lineshapes of microwave radiation backscattered from sea surfaces at small grazing angles,” IEE Proceedings—Radar, Sonar and Navigation, vol. 142, no. 5, pp. 252–258, 1995.
View at: Publisher Site | Google Scholar
M.-H. Ka, A. I. Baskakov, V. A. Terechov, and A. A. Kononov, “Estimation of the sea-surface slope variance based on the power spectrum width of a radar scatterometer,” IEEE Geoscience and Remote Sensing Letters, vol. 10, no. 3, pp. 593–597, 2013.
View at: Publisher Site | Google Scholar
C. M. Hoang, M. Kim, and S. H. Kong, “Automatic recognition of general lpi radar waveform using ssd and supplementary classifier,” IEEE Transactions on Signal Processing, vol. 67, no. 13, pp. 3516–3530, 2019.
View at: Publisher Site | Google Scholar
T. R. Kishore and K. D. Rao, “Automatic intrapulse modulation classification of advanced lpi radar waveforms,” IEEE Transactions on Aerospace & Electronic Systems, vol. 53, no. 2, pp. 901–914, 2017.
View at: Publisher Site | Google Scholar
M. A. Govoni, H. Li, and J. A. Kosinski, “Range-doppler resolution of the linear-fm noise radar waveform,” IEEE Transactions on Aerospace and Electronic Systems, vol. 49, no. 1, pp. 658–664, 2013.
View at: Publisher Site | Google Scholar
S. Liu, S. Tao, Y. D. Zhang, T. Ran, and F. Yuan, “A fast algorithm for multi-component lfm signal analysis exploiting segmented dpt and sdfrft,” in Proceedings of the 2015 IEEE Radar Conference, pp. 1139–1143, Arlington, VA, USA, May 2015.
View at: Publisher Site | Google Scholar
M. Daković, T. Thayaparan, and L. Stanković, “Time-frequency-based detection of fast manoeuvring targets,” IET Signal Processing, vol. 4, no. 3, pp. 287–297, 2010.
View at: Publisher Site | Google Scholar
E. G. Baxa, “On implementation of the discrete fourier transform-the stuse algorithm for spectral estimation,” IEEE Transactions on Acoustics Speech & Signal Processing, vol. 37, no. 11, pp. 1763–1765, 2002.
View at: Publisher Site | Google Scholar
P. Welch, “The use of fast fourier transform for the estimation of power spectra: a method based on time averaging over short, modified periodograms,” IEEE Transactions on Audio and Electroacoustics, vol. 15, no. 2, pp. 70–73, 2003.
View at: Publisher Site | Google Scholar
T. Lobos, Z. Leonowicz, J. Rezmer, and P. Schegner, “High-resolution spectrum-estimation methods for signal analysis in power systems,” IEEE Transactions on Instrumentation and Measurement, vol. 55, no. 1, pp. 219–225, 2006.
View at: Publisher Site | Google Scholar
F. Yao and T. C. M. Lee, “Spectral density estimation using sharpened periodograms,” IEEE Transactions on Signal Processing, vol. 55, no. 9, pp. 4711–4716, 2007.
View at: Publisher Site | Google Scholar
Y. Ephraim and W. J. J. Roberts, “On second-order statistics of log-periodogram with correlated components,” IEEE Signal Processing Letters, vol. 12, no. 9, pp. 625–628, 2005.
View at: Publisher Site | Google Scholar
S. M. Kay and S. L. Marple, “Spectrum analysis-a modern perspective,” Proceedings of the IEEE, vol. 69, no. 11, pp. 1380–1419, 1981.
View at: Publisher Site | Google Scholar
G. C. Carter and A. H. Nuttall, “Analysis of a generalised framework for spectral estimation. part 1: the technique and its mean value,” IEE Proceedings F Communications, Radar and Signal Processing, vol. 130, no. 3, p. 239, 1983.
View at: Publisher Site | Google Scholar
S. Liu, T. Shan, R. Tao et al., “Sparse discrete fractional fourier transform and its applications,” IEEE Transactions on Signal Processing, vol. 62, no. 24, pp. 6582–6595, 2014.
View at: Publisher Site | Google Scholar
L. B. Almeida, “The fractional fourier transform and time-frequency representations,” IEEE Transactions on Signal Processing, vol. 42, no. 11, pp. 3084–3091, 1994.
View at: Publisher Site | Google Scholar
R. Torres and E. Torres, “Fractional fourier analysis of random signals and the notion of $\alpha$-Stationarity of the wigner-ville distribution,” IEEE Transactions on Signal Processing, vol. 61, no. 6, pp. 1555–1560, 2013.
View at: Publisher Site | Google Scholar
B. Boashash and S. Ouelha, “Designing high-resolution time-frequency and time-scale distributions for the analysis and classification of non-stationary signals: a tutorial review with a comparison of features performance,” Digital Signal Processing, vol. 77, pp. 120–152, 2018.
View at: Publisher Site | Google Scholar
I. E. Sejdić and L. Stanković, “Fractional fourier transform as a signal processing tool: an overview of recent developments,” Signal Processing, vol. 91, no. 6, pp. 1351–1369, 2011.
View at: Publisher Site | Google Scholar
H. M. Ozaktas, D. Mendlovic, L. Onural, and B. Barshan, “Convolution, filtering, and multiplexing in fractional fourier domains and their relation to chirp and wavelet transforms,” Journal of the Optical Society of America A, vol. 11, no. 2, pp. 547–559, 1994.
View at: Publisher Site | Google Scholar
X. Y. Meng, R. Tao, and Y. Wang, “The fractional fourier domain analysis of two channel filter banks,” Acta Electronica Sinica, vol. 36, no. 5, pp. 919–927, 2008.
View at: Google Scholar
X. Y. Meng, R. Tao, and Y. Wang, “Generalized design method of multirate filter banks in the fractional fourier domain,” Acta Electronica Sinica, vol. 37, no. 9, pp. 2046–2051, 2009.
View at: Google Scholar
X. Meng, R. Tao, and Y. Wang, “Fractional fourier domain analysis of decimation and interpolation,” Science in China Series F: Information Sciences, vol. 50, no. 4, pp. 521–538, 2007.
View at: Publisher Site | Google Scholar
A. Kutay, H. M. Ozaktas, O. Ankan, and L. Onural, “Optimal filtering in fractional fourier domains,” IEEE Transactions on Signal Processing, vol. 45, no. 5, pp. 1129–1143, 1997.
View at: Publisher Site | Google Scholar
A. I. Zayed, “A convolution and product theorem for the fractional fourier transform,” IEEE Signal Processing Letters, vol. 5, no. 4, pp. 101–103, 1998.
View at: Publisher Site | Google Scholar
R. Tao, B. Deng, and Y. Wang, “Research progress of the fractional fourier transform in signal processing,” Science in China Series F, vol. 49, no. 1, pp. 1–25, 2006.
View at: Publisher Site | Google Scholar
Z. Feng and R. Tao, “Fmulti-sampling rate signal processing based on discrete-time fractional fourier transform,” Progress in Natural Science, vol. 18, no. 1, pp. 93–101, 2008.
View at: Google Scholar
R. Tao, B.-Z. Li, and Y. Wang, “Spectral analysis and reconstruction for periodic nonuniformly sampled signals in fractional fourier domain,” IEEE Transactions on Signal Processing, vol. 55, no. 7, pp. 3541–3547, 2007.
View at: Publisher Site | Google Scholar
T. Ran, Z. Feng, and W. Yue, “Fractional power spectrum,” IEEE Transactions on Signal Processing, vol. 56, no. 9, pp. 4199–4206, 2008.
View at: Publisher Site | Google Scholar
P. P. Vaidyanathan and P. Pal, “Sparse sensing with co-prime samplers and arrays,” IEEE Transactions on Signal Processing, vol. 59, no. 2, pp. 573–586, 2011.
View at: Publisher Site | Google Scholar
S. Liu, Y. Ma, and T. Shan, “Segmented discrete polynomial-phase transform with coprime sampling,” The Journal of Engineering, vol. 2019, no. 19, pp. 5619–5621, 2019.
View at: Publisher Site | Google Scholar
A. Shaker and E. Hüllermeier, “Recovery analysis for adaptive learning from non-stationary data streams: experimental design and case study,” Neurocomputing, vol. 150, pp. 250–264, 2015.
View at: Publisher Site | Google Scholar
C. Wang, J. Wang, and T. Zhang, “Operational modal analysis for slow linear time-varying structures based on moving window second order blind identification,” Signal Processing, vol. 133, pp. 169–186, 2017.
View at: Publisher Site | Google Scholar
B. Boashash, A. Aïssa-El-Bey, and A. Aïssa-El-Bey, “Robust multisensor time-frequency signal processing: a tutorial review with illustrations of performance enhancement in selected application areas,” Digital Signal Processing, vol. 77, pp. 153–186, 2018.
View at: Publisher Site | Google Scholar
C. Cheng, A. Sa-ngasoongsong, O. Beyca et al., “Time series forecasting for nonlinear and non-stationary processes: a review and comparative study,” IIE Transactions, vol. 47, no. 10, pp. 1053–1071, 2015.
View at: Publisher Site | Google Scholar
X. Huang, Y. Han, Z. Yan, H. Xian, and W. Lu, “Resolution doubled co-prime spectral analyzers for removing spurious peaks,” IEEE Transactions on Signal Processing, vol. 64, no. 10, pp. 2489–2498, 2016.
View at: Publisher Site | Google Scholar
J. W. Xiangdong Huang and S. H. A. N. Yuxuan, “High-performance co-prime spectral analysis method based on parallelled all-phase point-pass filtering,” Journal of Electronics & Information Technology, vol. 42, no. 3, pp. 613–620, 2020.
View at: Google Scholar
P. P. Vaidyanathan and P. Pal, “Theory of sparse coprime sensing in multiple dimensions,” IEEE Transactions on Signal Processing, vol. 59, no. 8, pp. 3592–3608, 2011.
View at: Publisher Site | Google Scholar
Q. Wu and Q. Liang, “Coprime sampling for nonstationary signal in radar signal processing,” Eurasip Journal onWireless Communications and Networking, vol. 58, pp. 1–11, 2013.
View at: Google Scholar
S.-C. Pei and J.-J. Ding, “Simplified fractional fourier transforms,” Journal of the Optical Society of America A, vol. 17, no. 12, pp. 2355–2367, 2000.
View at: Publisher Site | Google Scholar
P. P. Vaidyanathan, “Multirate digital filters, filter banks, polyphase networks, and applications: a tutorial,” Proceedings of the IEEE, vol. 78, no. 1, pp. 56–93, 2002.
View at: Publisher Site | Google Scholar
W. Yuqing and P. Zhenming, “The optimal fractional s transform of seismic signal based on the normalized second-order central moment,” Journal of Applied Geophysics, vol. 129, pp. 8–16, 2016.
View at: Publisher Site | Google Scholar
G. Lopez-Risueno, J. Grajal, and A. Sanz-Osorio, “Digital channelized receiver based on time-frequency analysis for signal interception,” IEEE Transactions on Aerospace and Electronic Systems, vol. 41, no. 3, pp. 879–898, 2005.
View at: Publisher Site | Google Scholar
C.-P. Yen, Y. Tsai, and X. Wang, “Wideband spectrum sensing based on sub-nyquist sampling,” IEEE Transactions on Signal Processing, vol. 61, no. 12, pp. 3028–3040, 2013.
View at: Publisher Site | Google Scholar
S. Khorshidi, M. Karimi, and A. R. Nematollahi, “New autoregressive (ar) order selection criteria based on the prediction error estimation,” Signal Processing, vol. 91, no. 10, pp. 2359–2370, 2011.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2020 Xiaomin Li et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

195

Downloads

734

Citations