A dynamic digital channelization receiving method

Abstract

The present invention provides a dynamic digital channelization receiving method, including: establishing a mathematical model of a WOLA analysis filter bank; establishing a mathematical model of a WOLA synthesis filter bank; determining the passband cut-off frequency and the stopband start frequency of a prototype filter according to a channel division method; designing the prototype filter using the FRM technique; performing spectrum detection on the output signal of the analysis filter bank using energy detection, and inputting the detected cross-channel signal into the synthesis filter bank for reconstruction; The present invention considers the entire implementation process of dynamic channelization. By improving the FRM technique, a novel FRM is proposed for designing filters with power complementary characteristics, which are then used as the prototype filters of the dynamic channelization structure of the WOLA filter bank, reducing the computational complexity of directly designing such filters and overlapping channels to reduce signal distortion caused by blind spots.

Description

Technical Field

[0001] The present invention relates to the field of military communication technologies, and particularly to a dynamic digital channelization receiving method. Background Art

[0002] With the continuous development of technology, the era of traditional warfare has gradually passed. Modern warfare relies more on the competition of technologies, and electronic warfare is one of the main combat methods. Electronic warfare mainly includes electronic reconnaissance, electronic jamming, etc. Among them, electronic reconnaissance intercepts and analyzes the signals sent by the enemy, enabling our side to have a better understanding of the enemy's dynamics, facilitating the formulation of targeted strategies, and playing a crucial role. In electronic reconnaissance, it is necessary to intercept the enemy's signals first, so digital receivers are essential. For signals in a relatively wide frequency band range, broadband digital receivers are required.

[0003] Channelization is one of the key technologies in broadband digital receivers. Its implementation principle is to divide broadband digital signals into multiple sub-channels, and then process each sub-channel separately to achieve the processing of signals that overlap in time domain but do not overlap in frequency domain, and the amplitude information and phase information of the original signal can be retained. With the increasingly complex electromagnetic environment, the bandwidth, center frequency, etc. of the received signals are constantly changing. It is essential to study non-uniform digital channelization receivers that can be dynamically adjusted according to the received signals.

[0004] Non-uniform digital channelization receivers include direct type and indirect type. The direct type divides the channel into sub-channels with different bandwidths and dynamically adjusts the parameters of the filter according to the input signal. The design complexity is relatively high, and the real-time performance is poor. The indirect type reconstructs the signals that cross channels through a reconstruction method to achieve the purpose of restoring the original signal. The reconstruction method requires an analysis-synthesis filter bank. The analysis part is used to uniformly divide the signal, and the synthesis part reconstructs the input cross-channel signals. However, this method has high requirements for the filter bank and can have good reconstruction characteristics only when the accurate reconstruction conditions are met. Therefore, it is necessary to conduct research on the reconstruction filter bank.

[0005] The commonly used structure for dynamic channelization is the polyphase DFT filter bank, but this structure has the limitation that the decimation factor and the number of channels must be integer multiples, which cannot meet the flexibility requirements of the interception system. The WOLA filter bank does not have such a limitation. In addition, due to the non-ideal characteristics of the filter, there will be a blind zone phenomenon during the channelization process, resulting in signal distortion. Therefore, it is of great practical significance to carry out the design of the prototype filter based on the WOLA filter bank and reduce the distortion caused by the blind zone. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the present invention proposes a dynamic digital channelization receiving method, which uses a WOLA filter bank dynamic channelization structure based on FRM. By improving the FRM technology, the structure proposes a new type of FRM for designing filters with power complementary characteristics, which is then used as the prototype filter of the WOLA filter bank dynamic channelization structure, reducing the computational complexity of directly designing the filter and overlapping the channels to reduce signal distortion caused by blind spots.

[0007] In a first aspect, the present invention provides a dynamic digital channelization receiving method, the method comprising:

[0008] S1: Establish a mathematical model of the WOLA analysis filter bank;

[0009] S2: Establish a mathematical model of the WOLA synthesis filter bank;

[0010] S3: Determine the passband cut-off frequency and stopband start frequency of the prototype filter according to the channel division method;

[0011] S4: Design the prototype filter using the FRM technology;

[0012] S5: Use energy detection to perform spectrum detection on the subband signals output by the analysis filter bank, and input the detected cross-channel signals into the synthesis filter bank for reconstruction.

[0013] The specific steps of S1 include:

[0014] In the WOLA filter bank dynamic channelization structure, the analysis filter bank model is:

[0015]

[0016] where the output signal y k (m) represents the signal output from the k-th subchannel, 1 ≤ k ≤ K, K represents the number of channels, w k represents the center frequency of the k-th subchannel, D represents the decimation factor, m is an integer, represents the Fourier transform of h(-r)x(r + mD), as follows:

[0017]

[0018] where h(-r) represents the filter, x(r + mD) represents the input signal, and for further derivation, let x m (r) = h(-r)x(r + mD), and perform superposition operation on x m (r) every K points to obtain

[0019]

[0020] Let \(r = p + lK\), then:

[0021]

[0022] The specific steps of step S2 include:

[0023] In the dynamic channelization structure of the WOLA filter bank, the comprehensive filter bank model is:

[0024]

[0025] Wherein, represents the output signal, with every Dth output, D represents the decimation factor, M is an integer, \(r = 0, 1, \cdots, D - 1\), K represents the number of channels, h represents the filter, and U m represents the inverse Fourier transform of the input signal after complex modulation, as shown below:

[0026]

[0027]

[0028] Where \(w\) k represents the center frequency of the \(k\)th sub-channel, and \(X\) K (m) represents the signal output by the analysis filter.

[0029] The specific steps of step S3 include:

[0030] Set the number of channels to K, the decimation factor to D, the bandwidth of each sub-channel to \(2\pi / K\), the interpolation factor of the prototype filter in the FRM technology to L, and the interpolation factor of the masking filter to N. In order to prevent cross-channel signals from being distorted, partial overlap is performed between each sub-channel to obtain the normalized passband cut-off frequency \(\omega\) p1 and the normalized stopband start frequency \(\omega\) s1 , and the transition bandwidth is \(f\) b ;

[0031] \(\omega\) p1 =\(\pi / K - f\) b / 2

[0032] \(\omega\) s1 =\(\pi / K - f\) b / 2

[0033] Through the improved FRM technology, filters with power complementary characteristics can be designed. However, due to the non-ideal characteristics of the filters, the passband of the filters is not an ideal broken line, which will cause partial distortion of the cross-channel signals. Therefore, this paper proposes to extend the passband of the low-pass filter by \(f\) t, obtain the normalized passband cut-off frequency ω of the low-pass filter p and the normalized stopband start frequency ω s ;

[0034] ω p = π / K - f b / 2 + f t

[0035] ω s = π / K - f b / 2 + f t

[0036] The specific steps of step S4 include:

[0037] The expression of the filter designed by FRM is:

[0038] H(z) = {H a (z L )H Ma (z N ) + H c (z L )H Mc (z N )}G(z)

[0039] where H a (z L ) and H c (z L ) represent the interpolated prototype filter and complementary filter, L represents the interpolation multiple, H ma (z N ) and H mc (z N ) represent the interpolated masking filter, N represents the interpolation multiple, G(z) represents the low-pass filter, and H(z) represents the filter designed by FRM;

[0040] The relevant parameters of the prototype filter are:

[0041] θ = Lω p - 2lπ

[0042] φ = Lω s - 2lπ

[0043] l = floor(ω p L / 2π)

[0044] where θ represents the passband cut-off frequency of the prototype filter, represents the stopband start frequency of the prototype filter, ω p represents the passband cut-off frequency of H(z), ω s represents the stopband start frequency of H(z), and floor() represents rounding down;

[0045] The relevant parameters of the two shielded filters are as follows:

[0046]

[0047]

[0048] Among them, ω Map represents the passband cut-off frequency of the shielded filter 1, ω Mas represents the stopband start frequency of the shielded filter 1, ω Mcp represents the passband cut-off frequency of the shielded filter 2, ω Mcs represents the stopband start frequency of the shielded filter 2. The value of N needs to satisfy that both ω Mas and ω Mcs are less than π;

[0049] The relevant parameters of the low-pass filter are:

[0050]

[0051] Among them, ω Gp represents the passband cut-off frequency of the low-pass filter, ω Gs represents the stopband start frequency of the low-pass filter.

[0052] The specific steps of step S5 include:

[0053] Taking the prototype filter designed by the FRM technology as the prototype filter of the WOLA filter bank dynamic channelization structure, performing energy detection on the sub-channel signals output by the analysis filter bank, setting a decision threshold α, and calculating the energies of K sub-channels; if the energy is greater than the threshold α, it is determined that there are sub-band signals or partial sub-band signals in the sub-channel; if the energy is less than the threshold α, it is determined that there is no signal;

[0054] Determining whether cross-channel occurs for the sub-channels with signals, inputting the cross-channel sub-channel signals into the synthesis filter bank for reconstruction to obtain sub-signals; the signals without cross-channel are directly output by the analysis filter bank.

[0055] As can be seen from the above technical solutions, the present invention provides a dynamic digital channelization receiving method, using a WOLA filter bank dynamic channelization structure based on FRM. This structure improves the FRM technology, proposes a new type of FRM for designing filters with power complementary characteristics, and then uses it as the prototype filter of the WOLA filter bank dynamic channelization structure, reducing the computational complexity of directly designing this filter. While reducing the filter design complexity, it also overlaps the channels to reduce signal distortion caused by blind spots. Description of the Drawings

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following briefly introduces the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0057] Figure 1 It is a block diagram of the implementation of the analysis part of the WOLA filter bank;

[0058] Figure 2 It is a block diagram of the implementation of the synthesis part of the WOLA filter bank;

[0059] Figure 3 It is a block diagram of the implementation of the FRM technology;

[0060] Figure 4 It is a schematic diagram of the process of synthesizing the final narrowband filter of the relevant filter of the FRM technology implementation;

[0061] Figure 5 It is a block diagram of implementing the analysis part and the synthesis part of the WOLA filter bank using the prototype filter designed by the FRM;

[0062] Figure 6 It is the amplitude-frequency response of the narrow transition band filter designed by the FRM technology;

[0063] Figure 7 It is the signal spectrum of the input channelized receiver and each divided sub-channel. The input signals include QPSK signals, single-carrier signals, and LFM signals;

[0064] Figure 8 It is the spectrum of each sub-channel output by the analysis part of the dynamic channelized receiver proposed by the present invention for the input signal;

[0065] Figure 9 It is the spectrum of the signal reconstructed by the synthesis part of the dynamic channelized receiver proposed by the present invention for the cross-channel QPSK signal;

[0066] Figure 10 It is the spectrum of the signal reconstructed by the synthesis part of the dynamic channelized receiver proposed by the present invention for the cross-channel LFM signal;

[0067] Figure 11 It is to compare the bit error rate of the reconstructed QPSK signal and the theoretical bit error rate with the original QPSK signal. It can be seen that the bit error rate of the original QPSK signal and the bit error rate of the reconstructed QPSK signal are almost the same and are close to the theoretical bit error rate, which can meet the system requirements;

[0068] Figure 12 It is a schematic flow chart of a dynamic channelized receiver of a WOLA filter bank based on FRM provided by the present invention. Specific implementation manners

[0069] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0070] As Figure 12 shown, a dynamic digital channelization receiving method provided by an embodiment of the present invention includes the following steps:

[0071] S1: Establish a mathematical model of a WOLA analysis filter bank;

[0072] S2: Establish a mathematical model of a WOLA synthesis filter bank;

[0073] S3: Determine the passband cut-off frequency and stopband start frequency of the prototype filter according to the channel division method;

[0074] S4: Design the prototype filter using the FRM technology;

[0075] S5: Use energy detection to perform spectrum detection on the subband signals output by the analysis filter bank, and input the detected cross-channel signals into the synthesis filter bank for reconstruction.

[0076] In this embodiment, the specific process of step S1 is as follows:

[0077] In the dynamic channelized structure of the WOLA filter bank, the analysis filter bank model is:

[0078]

[0079] where the output signal y k (m) represents the signal output from the k-th subchannel, 1 ≤ k ≤ K, K represents the number of channels, w k represents the center frequency of the k-th subchannel, D represents the decimation factor, m is an integer, represents the Fourier transform of h(-r)x(r + mD), as follows:

[0080]

[0081] where h(-r) represents the filter, x(r + mD) represents the input signal, and for Further derivation, let x m (r) = h(-r)x(r + mD), perform superposition operation on x m (r) every K points, and obtain

[0082]

[0083] Let r = p + lK, then:

[0084]

[0085] As Figure 1 shown, the implementation process of the WOLA analysis filter bank described in this embodiment is as follows:

[0086] S11: Define two shift registers Rs1 and Rs2 with the same length as the filter h(-r), with a length of N and all set to 0;

[0087] S12: Then input D data into Rs1 in sequence, multiply the data in the shift register Rs1 by the filter h(-r), and store it in another shift register Rs2;

[0088] S13: Divide the obtained data into N / K groups and perform superposition in order to obtain K data, and perform FFT operation on the K data;

[0089] S14: Perform a circular shift operation on the data obtained after the FFT operation, and the circular shift bit number is the modulus value of -mD and K;

[0090] S15: Return to S11, continue to input data, and repeat the subsequent operations until all data is output from the shift register Rs1.

[0091] In this embodiment, the specific process of step S2 is as follows:

[0092] In the dynamic channelization structure of the WOLA filter bank, the comprehensive filter bank model is:

[0093]

[0094] Among them, represents the output signal, output every D, D represents the decimation factor, M is an integer, r = 0, 1,..., D - 1, K represents the number of channels, h represents the filter, U m represents the inverse Fourier transform of the input signal after complex modulation, as follows:

[0095]

[0096] Among them, w k represents the center frequency of the kth sub-channel, XK (m) represents the signal output by the analysis filter.

[0097] As Figure 2 shown, the implementation process of the WOLA synthesis filter bank described in this embodiment is as follows:

[0098] S21: First, define two shift registers Rs3 and Rs4 with the same length as the filter, and set their initial states to 0;

[0099] S22: Perform IFFT operation on K input data and perform circular shift. The shift bit number is the modulus of mD and K to obtain U m ;

[0100] S23: Fill U m sequentially into register Rs3, and then multiply the data in Rs3 by the filter h(r);

[0101] S24: Add the data obtained in S23 to the data in Rs4 and store the result in Rs4. Then, shift Rs4, take D data at one end as the output, and input D data with value 0 at the other end;

[0102] S25: Return to S22, continue to input data and repeat the subsequent operations until all data is input.

[0103] In this embodiment, step S3 specifically includes:

[0104] Set the number of channels to K, the decimation factor to D, the bandwidth of each sub-channel to 2π / K, the interpolation factor of the prototype filter in FRM technology to L, and the interpolation factor of the masking filter to N. In order to prevent cross-channel signals from being distorted, partial overlap is performed between each sub-channel to obtain the normalized passband cut-off frequency ω p1 and the normalized stopband start frequency ω s1 , and the transition bandwidth is f b ;

[0105] ω p1 = π / K - f b / 2 (8)

[0106] ω s1 = π / K - f b / 2 (9)

[0107] Through the improved FRM technology, filters with power complementary characteristics can be designed. However, due to the non-ideal characteristics of the filters, the passbands of the filters are not ideal broken lines, which will cause partial distortion of cross-channel signals. Therefore, this paper proposes to extend the passband of the low-pass filter by f t , to obtain the normalized passband cut-off frequency ω pand the normalized stopband starting frequency ω s ;

[0108] ω p = π / K - f b / 2 + f t (10)

[0109] ω s = π / K - f b / 2 + f t (11)

[0110] In this embodiment, step S4 specifically includes:

[0111] The expression of the filter designed by FRM is:

[0112] H(z) = {H a (z L )H Ma (z N ) + H c (z L )H Mc (z N )}G(z) (12)

[0113] where H a (z L ) and H c (z L ) represent the interpolated prototype filter and complementary filter, L represents the interpolation multiple, H ma (z N ) and H mc (z N ) represent the interpolated shield filter, N represents the interpolation multiple, G(z) represents the low-pass filter, and H(z) represents the filter designed by FRM;

[0114] Figure 3 is the implementation block diagram of the filter designed by FRM technology;

[0115] S41: According to the normalized passband cut-off frequency ω p and the normalized stopband starting frequency ω s obtained in step S3 and the interpolation multiple, determine the passband cut-off frequency and stopband starting frequency of the prototype filter;

[0116] The relevant parameters of the prototype filter are:

[0117] θ = Lω p - 2lπ (13)

[0118] φ = Lω s - 2lπ (14)

[0119] l = floor(ω p L / 2π) (15)

[0120] where θ represents the passband cutoff frequency of the prototype filter, represents the stopband start frequency of the prototype filter, ω p represents the passband cutoff frequency of H(z), ω s represents the stopband start frequency of H(z), and floor() represents rounding down.

[0121] S42: Determine the passband cutoff frequency and stopband start frequency of the two shielding filters according to the passband cutoff frequency and stopband start frequency of the prototype filter obtained in S41 and the shielding filter interpolation multiple obtained in step S3;

[0122] The relevant parameters of the two shielding filters are:

[0123]

[0124] where ω Map represents the passband cutoff frequency of shielding filter 1, ω Mas represents the stopband start frequency of shielding filter 1, ω Mcp represents the passband cutoff frequency of shielding filter 2, ω Mcs represents the stopband start frequency of shielding filter 2, and the value of N needs to be such that ω Mas and ω Mcs are both less than π.

[0125] S43: Determine the passband cutoff frequency and stopband start frequency of the low-pass filter according to the passband cutoff frequency and stopband start frequency of the prototype filter obtained in S41 and the prototype filter interpolation multiple and shielding filter interpolation multiple obtained in step S3;

[0126] The relevant parameters of the low-pass filter are:

[0127]

[0128] where ω Gp represents the passband cutoff frequency of the low-pass filter, ω Gs represents the stopband start frequency of the low-pass filter.

[0129] Figure 4 is the specific implementation process of obtaining a narrow transition band filter through the FRM technique. Among them, Figure 4 (a) represents the prototype filter H a (z) and the complementary filter H c (z). Figure 4 (b) is to interpolate them to obtain H a (z L) and H c (z L ), the interpolation multiple L = 4. Figure 4 (c) are two shield filters. Figure 4 (d) is the shield filter H obtained after interpolation ma (z N ) and H mc (z N ), the interpolation multiple N = 2. Figure 4 (e) and Figure 4 (f) are respectively H ma (z N ) filtering H a (z L ) and H mc (z N ) filtering H c (z L ). Figure 4 (g) represents the low - pass filter G(z). Figure 4 (h) is Figure 4 (e) and Figure 4 (f) finally obtaining the filter H(z) after passing through the low - pass filter, whose transition band is provided by the prototype filter and is 1 / L of the transition band of the prototype filter;

[0130] In this embodiment, step S6 specifically includes:

[0131] Taking the prototype filter designed by the FRM technology as the prototype filter of the WOLA filter bank dynamic structure, performing energy detection on the sub - channel signals output by the analysis filter bank, setting the decision threshold α, and calculating the energy of K sub - channels; if its energy is greater than the threshold α, it is determined that there is a sub - band signal or a partial sub - band signal in the sub - channel; if its energy is less than the threshold α, it is determined that there is no signal.

[0132] Judging whether cross - channel occurs for the sub - channels with signals, inputting the cross - channel sub - channel signals into the synthesis filter bank for reconstruction to obtain sub - signals; the signals without cross - channel are directly output in the analysis filter bank.

[0133] Figure 5 is the overall implementation process of the entire dynamic channelization. The input signal first passes through the analysis part, dividing the broadband signal into each sub - channel, then using energy detection to detect each sub - channel to judge whether there is signal cross - channel, and finally inputting the cross - channel sub - channel signals into the synthesis part for reconstruction;

[0134] As can be seen from the above technical solutions, the present invention provides a dynamic digital channelization receiving method, which uses a dynamic channelization structure of a WOLA filter bank based on the FRM. This structure improves the FRM technology, making the filters designed by FRM have power complementary characteristics, and then using them as the prototype filters of the dynamic channelization structure of the WOLA filter bank, reducing the complexity of directly designing such filters. While reducing the high complexity of filter design, by overlapping the channels, the signal distortion caused by blind spots is reduced.

[0135] The WOLA filter bank is not restricted by the number of channels and the decimation factor, and can realize a dynamically flexible designed channelization receiver; the FRM technology can design narrow transition band filters with relatively low computational complexity, and can greatly reduce the complexity in scenarios with a large number of channels. Through improvement, a new type of FRM is proposed, which can be used to design filters with power complementary characteristics; therefore, the dynamic channelization structure of the WOLA filter bank based on FRM proposed by the present invention applies the new type of FRM technology to the dynamic channelization structure of the WOLA filter bank, overlaps the channels, reduces the complexity of prototype filter design and reduces the signal distortion caused by blind spots.

[0136]

[0137] Thus, a dynamic digital channelization receiving method proposed in this embodiment, which uses a dynamic channelization structure of a WOLA filter bank based on FRM, considers the entire implementation process of dynamic channelization. This structure improves the FRM technology, proposes a new type of FRM for designing filters with power complementary characteristics, and then uses them as the prototype filters of the dynamic channelization structure of the WOLA filter bank, reducing the complexity of directly designing the filters, overlapping the channels, and reducing the signal distortion caused by blind spots.

[0138] Set the number of channels K = 16, the decimation factor D = 8, the interpolation factors L = 40, N = 10. To prevent the cross-channel signals from being distorted, partial overlap is performed between each sub-channel, such as Figure 9 , and the normalized passband cut-off frequency of the filter generated by FRM is obtained as 0.0614 and the normalized stopband start frequency is 0.0644. The input signals include QPSK signals, single-carrier signals, and LFM signals, which overlap in the time domain and do not overlap in the frequency domain.

[0139] Figure 6 is the spectrum of the prototype filter designed by the FRM technology, and it can be seen that the transition band of the filter is very narrow.

[0140] Figure 7 represents the input signal spectrum diagram and channel division, showing the positions where the input signals theoretically are in the channels.

[0141] Figure 8 It is the spectrum of the signals output from each sub-channel after the input signal passes through the analysis part of the dynamic channelization structure. It can be seen that the QPSK signals are located in channels 1-6, the single-carrier signal is located in channel 7, and the LFM signals are located in channels 7-9, and the effect is good.

[0142] Figure 9 It is the reconstructed signal obtained after the cross-channel QPSK signal is input into the comprehensive filter bank. It can be seen that the spectrum of the reconstructed signal is basically the same as that of the input signal, and the effect is good.

[0143] Figure 10 It is the reconstructed signal obtained after the cross-channel LFM signal is input into the comprehensive filter bank. It can be seen that the spectrum of the reconstructed signal is basically the same as that of the input signal, and the effect is good.

[0144] Figure 11 To verify the reconstruction effect, the bit error rates of the reconstructed QPSK signal and the original QPSK signal are compared. It can be seen that the two are basically the same and are close to the theoretical bit error rate, which can meet the system requirements.

[0145] As described above, it is only the specific implementation manner of the present invention. However, the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A dynamic digital channelization receiving method, characterized in that: S1: Establish a mathematical model of the WOLA analysis filter bank; S2: Establish a mathematical model of the WOLA synthesis filter bank; S3: Determine the passband cut-off frequency and the stopband start frequency of the low-pass filter according to the channel division method, including: setting the number of channels as K, the decimation factor as D, the bandwidth of each sub-channel as 2π / K, the interpolation factor of the prototype filter in the FRM technology as L, the interpolation factor of the shielding filter as N, and partially overlapping between each sub-channel to obtain the normalized passband cut-off frequency ω p1 and the normalized stopband start frequency ω s1 , and the transition bandwidth is f b ; ω p1 = π / K - f b / 2 ω s1 = π / K + f b / 2 Extend the passband of the low-pass filter by f t , to obtain the normalized passband cut-off frequency ω p and the normalized stopband start frequency ω s ; ω p = π / K - f b / 2 + f t ω s = π / K + f b / 2 + f t S4: Use the FRM technique to design a prototype filter; wherein, the expression of the filter designed by FRM is: H(z) = {H a (z L )H Ma (z N ) + H c (z L )H Mc (z N )}G(z) Where H a (z L ) and H c (z L ) represent the interpolated prototype filter and complementary filter respectively, L represents the interpolation factor, H ma (z N ) and H mc (z N ) represent the interpolated masking filter, N represents the interpolation factor, G(z) represents the low-pass filter, and H(z) represents the filter designed by FRM; S5: Calculate the energy value of each sub-band signal, compare it with a preset threshold, and determine a cross-channel signal if it is greater than the threshold. Input the detected cross-channel signal into the synthesis filter bank for reconstruction, specifically including: Use the prototype filter designed by the FRM technique as the prototype filter of the dynamic channelization structure of the WOLA filter bank, perform energy detection on the sub-channel signals output by the analysis filter bank, set a decision threshold α, and calculate the energy of K sub-channels; if its energy is greater than the threshold α, it is determined that there is a sub-band signal or a partial sub-band signal in the sub-channel; if its energy is less than the threshold α, it is determined that there is no signal. Determine whether cross-channel occurs for the sub-channels with signals, input the cross-channel sub-channel signals into the synthesis filter bank for reconstruction to obtain sub-signals; the signals that do not generate cross-channel are directly output in the analysis filter bank.

2. The dynamic digital channelization receiving method according to claim 1, wherein The specific content of step S1 includes: In the dynamic channelization structure of the WOLA filter bank, the analysis filter bank model is: where the output signal y k (m) represents the signal output on the k-th subchannel, where 1 ≤ k ≤ K and K represents the number of channels, and w k represents the center frequency of the k-th subchannel, D represents the decimation factor, m is an integer representing the input at D data instants, represents the Fourier transform of h(−r)x(r + mD), as shown below: where h(-r) represents the filter, and x(r + mD) represents the input signal. For Further derivation, let x m (r) = h(-r)x(r + mD), and perform superposition operation on x m (r) every K points to obtain 1 ≤ p ≤ K, let r = p + lK, then:

3. A dynamic digital channelization receiving method according to claim 1, characterized in that The specific content of step S2 includes: In the dynamic channelization structure of the WOLA filter bank, the synthesis filter bank model is: Among them, represents the output signal. Each time, D signals are output. D represents the decimation factor. M and m are integers, r = 0, 1, ..., D - 1. K represents the number of channels, h represents the filter, and U m represents the inverse Fourier transform of the input signal after complex modulation, as follows: where w k represents the center frequency of the k-th sub-channel, and X K (m) represents the signal output from the analysis filter.

4. A dynamic digital channelization receiving method according to claim 1, characterized in that In step S4, The relevant parameters of the prototype filter are: θ = Lω p -2lπ φ = Lω s -2lπ l = floor(ω p L / 2π) where θ represents the passband cutoff frequency of the prototype filter, represents the stopband start frequency of the prototype filter, ω p represents the passband cutoff frequency of H(z), ω s represents the stopband start frequency of H(z), l represents an integer, and floor() represents rounding down; The relevant parameters of the two masking filters are: where ω Map represents the passband cut-off frequency of the shielded filter 1, ω Mas represents the stopband start frequency of the shielded filter 1, ω Mcp represents the passband cut-off frequency of the shielded filter 2, ω Mcs represents the stopband start frequency of the shielded filter 2, and the magnitude of N needs to satisfy that ω Mas and ω Mcs are both less than π; The relevant parameters of the low-pass filter are: where ω Gp represents the passband cut-off frequency of the low-pass filter, and ω Gs represents the stopband start frequency of the low-pass filter.

Images