CN113341378B

CN113341378B - Self-adaptive channelized receiving method based on frequency spectrum differential entropy detection

Info

Publication number: CN113341378B
Application number: CN202110410504.0A
Authority: CN
Inventors: 傅雄军; 崔宗丁; 吴菲; 卢继华; 谢民; 赵硕; 邓博文; 高浩东; 姜嘉环
Original assignee: Beijing Institute of Technology BIT
Current assignee: Beijing Institute of Technology BIT
Priority date: 2021-04-13
Filing date: 2021-04-13
Publication date: 2022-08-16
Anticipated expiration: 2041-04-13
Also published as: CN113341378A

Abstract

The invention relates to a self-adaptive channelized receiving method based on frequency spectrum differential entropy detection, and belongs to the technical field of radar signal reconnaissance and signal detection. Aiming at receiving a broadband Chirp signal, a narrow-band Chirp signal and a two-phase coding signal, performing first-stage channelized reception by using a channelized structure based on an odd-type group and an even-type polyphase filter group, and performing smoothing processing, false signal elimination, differential entropy detection, wide-narrow signal judgment and parameter estimation on the frequency spectrum energy of the received signal to obtain the number of cross channels and signal parameters; performing channel matching according to the number of channels, and selecting a channelized structure with a small number of channels; and performing secondary channelized reception according to the signal parameters, performing channel combination on the broadband signals, performing adaptive frequency band subdivision on the narrowband signals, performing secondary channel detection, and extracting effective signals. The method has the advantages of full probability receiving, high real-time performance, high detection accuracy, capability of detecting complex signals with low signal-to-noise ratio and strong noise filtering capability.

Description

Self-adaptive channelized receiving method based on frequency spectrum differential entropy detection

Technical Field

The invention relates to a self-adaptive channelized receiving method based on frequency spectrum differential entropy detection, and belongs to the technical field of radar signal reconnaissance and signal detection.

Background

In modern high-tech electronic warfare, the number of unknown radiation sources in a complex electromagnetic space is large, the signal forms are various, the signal frequency distribution range is wide, and the noise is large, so that the work of collecting electromagnetic environment information in a battlefield is difficult. The traditional digital receiver needs to use a high-speed DSP to process high-speed ADC sampling data, and the current DSP has low data processing efficiency, can cause the phenomenon of data loss and cannot meet the real-time property. A digital channelized receiver based on a polyphase filter is an efficient data rate reduction technology, which not only can match the data processing speed of an ADC (analog to digital converter) with that of a DSP (digital signal processor), but also enhances the real-time performance of system operation after all the operation of the digital channelized receiver is placed after data extraction.

In recent years, researchers have proposed dynamic channelization reception techniques using analysis and synthesis filter banks and signal detection methods for channelization reception. However, most channelized receiving technologies are fixed channel structures, and are difficult to process narrowband and broadband signals simultaneously, so that the flexibility is insufficient; for narrow-band signals in a monitoring frequency band, a cross-channel condition exists during receiving, and the monitoring frequency band is difficult to further subdivide so as to improve the signal-to-noise ratio of the signals; under the condition of low signal-to-noise ratio, the current signal detection method is difficult to accurately detect the channel where the target signal is located, and is easy to output false signals.

The invention aims to solve the technical defects of the analysis and synthesis filter bank and the detection method, and provides an adaptive channelized receiving method based on spectrum differential entropy detection.

Disclosure of Invention

The invention aims to solve the technical problems that when the existing channelizing method based on analysis and comprehensive filter banks is used for receiving radar signals, narrow-band and wide-band signals cannot be processed simultaneously, the narrow-band signals are easy to cross channels and difficult to further subdivide, and the detection accuracy under low signal-to-noise ratio is low.

In order to achieve the purpose, the invention adopts the following technical scheme:

the self-adaptive channelized receiving system based on the self-adaptive channelized receiving method comprises the following steps: the device comprises a primary channelized receiving module, a primary channel detection and estimation module, a channel matching module, a secondary self-adaptive channelized receiving module and a secondary channel detection module;

the first-stage channelized receiving module comprises an odd-type polyphase filter bank and an even-type polyphase filter bank which are fixed and have the same bandwidth; the channel detection and estimation module comprises a channel detection submodule and a frequency spectrum sensing submodule; the secondary self-adaptive channelized receiving module comprises a multi-channel merging module and a dynamic channelized module;

the connection relationship of each module in the self-adaptive channelized receiving system is as follows:

the primary channelized receiving module is connected with the primary channel detecting and estimating module and the channel matching module, the primary channel detecting and estimating module is respectively connected with the primary channelized receiving module, the channel matching module and the secondary adaptive channelized receiving module, and the secondary adaptive channelized receiving module is connected with the secondary channel detecting module;

in order to effectively prevent the channel crossing phenomenon, the receiving blind area of the odd-type polyphase filter bank is exactly the center of the pass band of the even-type polyphase filter bank, and the receiving blind area of the even-type polyphase filter bank is exactly the center of the pass band of the odd-type polyphase filter bank; the channel detection submodule detects whether a channel has a signal by adopting a frequency spectrum differential entropy and eliminates false signals caused by a transition band of a filter by utilizing frequency spectrum mutation characteristics, so that the problem of high omission factor under the condition of low signal-to-noise ratio is solved; the frequency spectrum sensing submodule extracts the center frequency and bandwidth information of the signal by adopting a short-time Fourier transform technology, the time domain and frequency domain resolution is good, and the signal-to-noise ratio of a weak signal can be greatly improved;

the channel matching module selects a set of filter banks with a smaller number of cross channels in the first-stage channelization module according to the result of the channel detection submodule; the channel merging module merges the broadband signals by utilizing the comprehensive filter group to output of the first-stage channelizing module, and the dynamic channelizing module can adaptively adjust and analyze the structure of the filter bank according to the spectrum sensing result so as to maximally improve the signal-to-noise ratio of the narrowband signals; the secondary channel detection module also detects by using frequency spectrum differential entropy detection.

The self-adaptive channelized receiving method comprises the stages of primary channelized receiving, primary channel detection and estimation, channel matching, secondary channelized self-adaptive receiving and secondary channel detection, and specifically comprises the following steps:

step 1: the first-stage channelized receiving stage is that a first-stage channelized receiving module is constructed, and the method further comprises two sub-steps of constructing a complex electromagnetic environment and constructing a polyphase filter bank, and specifically comprises the following steps:

step 1.1: constructing a complex electromagnetic environment comprising different types of broadband and narrowband radar signals, wherein the carrier frequency of the radar signals is in the monitoring frequency range of the reconnaissance receiver;

step 1.2: constructing a polyphase filter bank;

the polyphase filter bank comprises an odd polyphase filter bank and an even polyphase filter bank, and the construction processes of the two polyphase filter banks are as follows:

step 1.2.1: d/2 time data rate reduction processing is carried out on the signals received by the reconnaissance receiver, namely the radar signals in the step 1.1 are sampled, continuous delay processing is carried out on the obtained digital serial signals, D-path parallel high-speed signals are obtained, D/2 time extraction operation is carried out on each path of signals, and parallel signals of the D-path data rate reduction are obtained;

wherein D is the number of channels and is an even number; when an odd-type polyphase filter is constructed, each path of data is multiplied by a factor

Wherein m is the label of each channel data sampling point, and j is an imaginary number unit;

step 1.2.2: constructing a prototype low-pass filter, selecting a digital FIR (finite impulse response) equiripple low-pass filter, wherein the stop-band cutoff frequency fs of the digital FIR equiripple low-pass filter is twice of the 3dB cutoff frequency fp of the pass band, and selecting the order of the prototype filter according to the 50% overlapping ratio of the multiphase filter group;

step 1.2.3: the prototype filter in step 1.2.2 is subjected to D-fold decimation and 2-fold zero value interpolation operations, with 1: d, deserializing the proportion to obtain D filter coefficients, and performing low-pass filtering on the parallel data obtained in the step 1.2.1;

wherein, when the odd-type polyphase filter is constructed, each path of data is multiplied by a factor

Wherein, p is a channel label, and the value range of p is 0 to D-1;

step 1.2.4: inputting the filtered parallel data into an IFFT module for D-point IFFT processing;

step 1.2.5: after IFFT processing, the odd numbered channels are multiplied by a factor (-1) ^m ；

Step 2: the first-stage channel detection and estimation stage is to construct a first-stage channel detection and estimation module, and comprises a channel detection stage, a wide-narrow signal judgment stage and a spectrum sensing stage, and specifically comprises the following substeps:

step 2.1: the channel frequency domain detection is used for extracting the channel with the effective signal, and the channel frequency domain detection further comprises the following substeps:

step 2.1.1: performing continuous sliding window processing on the frequency spectrum energy of the subchannel output data, and taking the obtained average value as the value of the central point of the window to eliminate frequency spectrum jitter and burrs caused by environmental interference and noise;

step 2.1.2: carrying out multistage differential processing on the smoothed frequency domain energy signal;

step 2.1.3: removing false signals caused by a transition band of a low-pass filter, specifically: the time at which the peak of the differential signal for each channel is detected, occurs at (0,

) Or (a)

L) in the time interval, the false signals are considered to be false signals and are removed;

wherein, L is the total number of sampling points of the single-channel signal;

step 2.1.4: quantizing the differential signal of each channel, counting the probability of the differential signal value appearing in each quantization interval, and calculating the differential entropy of each channel;

step 2.1.5: the difference entropy ratio is used for distinguishing a pure noise channel from an effective signal channel, the detection threshold is recorded as Th1, and H is recorded _min For the minimum value of the differential entropy, go through k if H _k /H _min ＞Th1，H _k If the difference entropy of the kth channel is the k channel, the kth channel is considered to have effective signals, otherwise, the kth channel is a pure noise channel, and the number of the detected channels with the effective signals is recorded as Num;

wherein, the range of Th1 is 1 to 2, and the value range of k is 0 to D-1;

step 2.2: performing double sliding window energy detection on a channel frequency spectrum, and judging the position of a signal edge according to the position of a signal ratio in two sliding windows exceeding a judgment threshold Th2 so as to distinguish a narrow-band signal from a wide-band signal;

wherein Th2 ranges from 1 to 3;

the method specifically comprises the following steps: for the k channel, the energy E1 of the front window of the ith sampling point is recorded _k (i) Energy of the rear window is E2 _k (i) If E1 _k (i)/E2 _k (i)>Th2, judging that the signal rising edge occurs at the moment; otherwise E1 _k (i)/E2 _k (i) Judging whether a signal falling edge occurs at the moment or not at Th2, pairwise matching the rising edge and the falling edge, judging the channel signal to be a narrow-band signal if the matching occurs in the same channel, and otherwise, judging the channel signal to be a wide-band signal;

step 2.3: inputting the narrow-band signal detected in the step 2.2 into an STFT module for parameter extraction to obtain signal carrier frequency and bandwidth information;

wherein, STFT, short-time Fourier transform, is called short-time Fourier transform in English;

so far, from step 2.1 to step 2.3, a first-stage channel detection and estimation stage is completed;

and step 3: in the channel matching stage, a channel matching module is constructed, specifically: according to the number of channels with effective signals detected in the step 2.1 being Num, comparing the channel number of the signals passing through the even type polyphase filter bank with the channel number of the signals passing through the odd type polyphase filter bank, and selecting the filter bank with less channel number as a channelized receiving structure;

and 4, step 4: a secondary adaptive channelization receiving stage, namely constructing a secondary adaptive channelization receiving module, which comprises multi-channel combination and adaptive channelization receiving, and specifically comprises the following steps:

step 4.1: reserving channel information of the broadband signal, setting other channels to be 0, and inputting the channel information to a dual structure of a multi-item filter for channel combination to obtain the broadband signal;

step 4.2: and (3) self-adaptive channelized reception, specifically, according to the spectrum sensing result in the step 2.2, constructing a self-adaptive dynamic channelized reception structure to subdivide a signal frequency domain, and specifically, the method comprises the following substeps:

step 4.2.1: calculating the number N of secondary channelized channels, constructing an even type multiphase filter bank and an odd type multiphase filter bank according to the number of the channels, namely firstly performing N/2 times extraction on the channels subjected to primary channelized, then performing filtering through the filter banks, and finally outputting N sub-channel data through N-point IFFT, wherein the specific design is the same as the step 1.2;

step 4.2.2: matching channels, inputting the narrow-band signals into two sets of multiphase filter banks constructed in the step 4.2.1 according to the carrier frequency and the bandwidth of the narrow-band signals extracted in the step 2.3, comparing the channel-crossing number of the two sets of filter banks, and selecting the set with the smaller channel-crossing number as a secondary channelized receiving structure;

and 5: in the second-level channel detection stage, a second-level channel detection module is constructed, specifically:

and performing secondary channel detection by using spectrum differential entropy detection, and outputting a channel where the effective signal is located.

Advantageous effects

The invention relates to a self-adaptive channelized receiving method based on frequency spectrum differential entropy detection, which has the following beneficial effects compared with the existing channelized receiving method:

1. the dual-arrangement channelized and odd-arrangement channelized parallel structure is adopted, the receiving blind area of one filter set is the passband of the other filter set, the cross-channel condition is reduced, and blind area-free receiving can be realized;

2. the system is suitable for complex electromagnetic environment, can process narrow-band signals and broadband signals simultaneously, can be packaged into an optional functional module according to specific environment requirements, and has strong flexibility;

3. by the aid of a static and dynamic combined multi-stage channelization structure, channel frequency bands where narrow-band signals are located are adaptively subdivided, and signal-to-noise ratio of received signals is effectively improved;

4. the frequency spectrum difference entropy detection method is adopted for channel detection, so that the operation amount is reduced, the influence of noise is reduced in frequency domain detection, and a complex pulse signal with a low signal-to-noise ratio can be detected;

5. the narrow-band signal and the wide-band signal are distinguished by adopting a double sliding window energy detection method, and the accuracy is high.

Drawings

Fig. 1 is a schematic flow chart of an adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention;

fig. 2 is a waveform diagram of an input radar signal in embodiment 1 of the adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention;

fig. 3 is a diagram of a channelized reception structure in embodiment 1 of the adaptive channelized reception method based on spectrum differential entropy detection according to the present invention;

fig. 4 is a diagram of a channel detection result in embodiment 1 of the adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention;

fig. 5 is a diagram of a channel merging structure in embodiment 1 of the adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention;

fig. 6 is a waveform diagram of an output signal 1 in embodiment 1 of the adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention;

fig. 7 is a waveform diagram of an output signal 2 in embodiment 1 of the adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention;

fig. 8 is a waveform diagram of the output signal 3 in embodiment 1 of the adaptive channelization receiving method based on spectrum differential entropy detection according to the present invention.

Detailed Description

For better explaining the objects and advantages of the method, the detailed description will be made in conjunction with the accompanying drawings and the specific embodiments of the present invention for the content of the implementation of the adaptive channelization receiving method based on spectrum differential entropy detection.

Example 1

In modern electronic warfare, the number of unknown radiation sources is more than 2, the signal form comprises wide and narrow linear frequency modulation signals and two-phase encoding signals, the signal to noise ratio of the signals is low, the radar reconnaissance receiver receives enemy transmitting signals in a monitoring frequency band as full as possible, the monitoring frequency band is usually over 1GHz, huge signal data need to be processed by a digital channelization structure to reduce the data rate, and therefore data of each sub-signal are obtained and are used as reference for generating corresponding interference signals.

This embodiment illustrates a specific implementation of the adaptive channelization receiving method based on spectrum differential entropy detection in the present application in receiving different radar signals, and an implementation flowchart of the present application is shown in fig. 1.

The input signal of the invention is a digital signal which carries out critical sampling on a zero intermediate frequency signal of 0-2GHz, the signal-to-noise ratio is 5dB, the sampling rate is 2.5GHz, as shown in figure 2, figure (2a) is an input signal time domain diagram, figure (2b) is an input signal frequency domain diagram, and the radar signal parameters and signal forms are shown in table 1:

radar signal parameters input in table 1

Radar serial number	Modulation system	Center frequency (MHz)	Bandwidth (MHz)	Pulse width (mus)
					1	Linear frequency modulated signal	940	800	10
2	Linear frequency modulated signal	300	150	10
					3	Two-phase encoded signal	1863	23	10

The self-adaptive channelization method mainly comprises the stages of primary channelization reception, primary channel detection and estimation, channel matching, secondary channelization self-adaptive reception and secondary channel detection. The specific implementation procedure is as follows.

Step 1: a first-stage channelized receiving stage, namely constructing a first-stage channelized receiving module, and performing first-stage channelized receiving on signals in a monitoring frequency band by adopting a channelized structure with odd and even arrangement of fixed uniform channels;

step 1.1: constructing a complex electromagnetic environment, and carrying out critical sampling on the 0-2GHz monitoring frequency band after down-conversion, wherein the complex electromagnetic environment comprises a broadband Chirp signal, a narrow-band Chirp signal and a two-phase coding signal based on a 13-bit Barker code, and specific parameters of the signal are shown in table 1;

step 1.2: a channelized structure based on even-type permutation polyphase filter banks is constructed, as shown in (3a) of fig. 3, the number of channels of the structure is represented as D, and is fixed to 8 in this example, and as shown in (3c), the implementation steps of the even-type permutation polyphase filter banks are as follows:

step 1.2.1: performing D/2-fold extraction, namely 4-fold extraction, on the input radar signal s (n), to obtain 8 extracted sub-channel signals: s ₀ (m)、s ₁ (m)、…s ₇ (m)；

Step 1.2.2: constructing a high-order prototype low-pass filter, wherein the 3dB cut-off frequency is pi/8, the filter coefficient is expressed as h (n), and then carrying out 8-time extraction and 2-time zero value interpolation to obtain 8 groups of filter coefficients: h is ₀ (m/2)、h ₁ (m/2)、…h ₇ (m/2)；

Step 1.2.3: filtering the extracted sub-channel signal, i.e. the k sub-channel signal s _k (m) and h _k (m/2) convolving to obtain the k-th sub-channel signal x _k (m)：

Step 1.2.4: inputting the filtered parallel data into an IFFT module for 8-point IFFT processing;

step 1.2.5: multiplying odd numbered subchannels by a factor (-1) ^m To obtain 8 channelized output signals y _k (m)(k＝0,1,…7)；

Step 1.3: a channelization structure based on an odd-type arrangement polyphase filter bank is constructed, as shown in fig. 3b, the number of channels of the structure is the same as that of even-type arrangement channels, and is fixed to 8, and as shown in fig. 3d, the odd-type arrangement filter bank has the following specific implementation steps:

step 1.3.1: modulating the signal extracted in the step 1.2.1 to obtain 8 paths of output signals

Expressed as:

step 1.3.2: filtering the modulated sub-channel signal, i.e. the k-th sub-channel signal

With h generated in step 1.2.2 _k (m/2) convolving and modulating to obtain the kth sub-channel signal

Step 1.3.4: inputting the filtered parallel data into an IFFT module for 8-point IFFT processing;

step 1.3.5: multiplying odd numbered subchannels by a factor (-1) ^m To obtain 8 channelized output signals

Step 2: in the stage of first-stage channel detection and estimation, a first-stage channel detection and estimation module is constructed, a sub-channel with an effective signal is detected, and radar signal parameters of the sub-channel are extracted, and the method specifically comprises the following steps: the parallel signal frequency spectrum after channelized reception in the step 1 is subjected to deburring, multistage differential entropy detection, wide and narrow signal detection and parameter extraction processing, and the method comprises the following substeps:

step 2.1: FFT processing is carried out on each path of signals output by the even type arrangement channelization in the step 1.2, and frequency domain energy data F of each path of signals are obtained after modulus square _k (i) (k is 0,1, … 7), as shown in (4a) of fig. 4, and then for F _k (i) Performing energy continuous sliding window processing, and taking the obtained average value as the value of the window center point, as shown in fig. 4 b;

wherein, for the k channel, the energy Y after the ith point smoothing processing _k (i) Comprises the following steps:

wherein, the window in the formula is the width of the smooth window (the window is generally 10 to 40, and is related to the noise magnitude);

step 2.2: performing multi-stage difference processing on the smoothed frequency domain energy signal, as shown in fig. 4 c; for the kth channel, the differential value after the ith-point multi-stage differential processing is represented as:

D _k (i)＝Y _k (i+m)-Y _k (i) (5)

wherein m is a multi-stage differential stage, and m is 20;

step 2.3: eliminating false signal, recording subchannel data length as L, detecting D _k (i) The location of the peak, if present at (0,

) Or (a)

L) is removed within the interval;

step 2.4: the effective channel is detected by differential entropy, firstly, the differential signal of each channel is quantized, and the probability P of the differential signal value appearing in each quantization interval is counted _k (m), calculating the differential entropy for each channel:

wherein, M is the number of quantization intervals (M is generally 5 to 15), and M is a quantization interval label;

distinguishing pure noise channels and effective signal channels by using differential entropy ratio, considering that the channel in which the minimum differential entropy is positioned is noise, recording detection threshold as Th1, traversing k, and if H is _k /H _min If the signal level is more than Th1, the k channel is considered to have effective signals, otherwise, the k channel is a pure noise channel;

wherein, the detection threshold is:

wherein H _min η is a predetermined constant (generally about 1), in this case η is 0.92;

the difference entropy ratio is [ 1.33901.97681.78061.71461.79331.20311.45790.9998 ], the detection threshold is Th1 ═ 1.4564, and the detection result is: if valid signals exist in

channels

1, 2, 3, 4 and 6, and the number of valid channels is Num1, Num1 is 5;

step 2.5: performing double sliding window energy detection on a channel frequency spectrum, and judging the position of a signal edge according to the position of a signal ratio in two sliding windows exceeding a judgment threshold Th2, so as to distinguish a narrow-band signal from a wide-band signal;

the method specifically comprises the following steps: for the k channel, the energy E1 of the ith point front window is recorded _k (i) Energy of the rear window is E2 _k (i) If E1 _k (i)/E2 _k (i)>Th2, judging that the signal rising edge occurs at the moment, if E1 _k (i)/E2 _k (i) Judging that a signal falling edge occurs at the moment when the current is less than or equal to Th2, wherein the value of Th2 is 1 to 3 generally, and the value of Th2 is 1.3 in the case of the scheme;

as shown in fig. 4d, the

channels

1, 2, and 6 detect the rising edge, the

channels

1, 4, and 6 detect the falling edge, pair the rising edge and the falling edge pairwise, if the pair appears in the same channel, the channel signal is determined to be a narrowband signal, otherwise, the channel signal is a wideband signal, that is, the channel 1 determines that there is a narrowband signal, the

channels

2, 3, and 4 determine that there is a wideband signal, and the channel 6 determines that there is a narrowband signal;

step 2.6: inputting the narrow-band signal detected in the step 2.5 into an STFT module for parameter extraction, wherein a hamming window is selected as a sliding window, the window length is 128, the number of FFT points in each section is 128, and the number of overlapped samples in each section is 127;

signal center frequency and bandwidth information is obtained as shown in table 2:

table 2 parameter extraction information

Number of signals	Channel numbering	Center frequency (MHz)	Bandwidth (MHz)
				1	1	300.3	151
2	6	1875	20

Step 2.7: performing first-stage channelization detection and estimation on each path of signals output by the odd-type arrangement channelization in the step 1.3, executing the step 2.1 to the step 2.6, and detecting to obtain a channel spanning number Num2 which is 6;

so far, from step 2.1 to step 2.7, the first-stage channel detection and estimation stage of the embodiment is completed;

and step 3: a channel matching stage, namely a channel matching module is constructed, according to the channel detection result in the step 2, the channel number of the even-type arrangement channelized structure is compared with the channel number of the received signal of the odd-type arrangement channelized structure, Num1 is less than Num2, so that the even-type arrangement channelized structure is selected as a first-stage channelized receiving structure of the radar signal input by the implementation case;

and 4, step 4: a secondary adaptive channelization receiving stage, namely constructing a secondary adaptive channelization receiving module, inputting effective signals detected in the primary channelization into the secondary channelization receiving module for processing, and including two stages of multi-channel combination of broadband signals and adaptive channelization reception of narrowband signals, specifically including the following steps:

step 4.1: the multi-channel combination, the channel combination structure is the inverse process of the channelization structure, as shown in fig. 5, the data of the

effective channels

2, 3, 4 detected in step 2 are retained, and the other channel data are set to zero and recorded as

Then the odd numbered subchannels are multiplied by the modulation factor (-1) ^m Inputting a modulation signal to FThe FT module performs 8-point FFT processing, then filters the processed 8 paths of parallel data, performs 4-fold interpolation processing on the filtered data, and finally performs delay addition on the interpolated 8 paths of parallel data to obtain a combined signal s1(n), as shown in fig. 6, where fig. 6a is a time domain diagram of an output combined signal and fig. 6b is a frequency domain diagram of the output combined signal;

step 4.2: and (3) self-adaptive channelized reception, and reconstructing a channelized reception structure according to the center frequency and bandwidth information of the narrow-band signal obtained in the step (2), specifically comprising the following substeps:

step 4.2.1: a two-stage channelized receiving structure is constructed for the signal 1, wherein B1 is a single-channel bandwidth after the first-stage channelization, B2 is a bandwidth of the narrowband signal 1 in step 2.6, and the number N1 of channels is represented as:

N1＝B1/B2 (8)

the number N1 of channels is an even number rounded downward, in this case, N1 is 4;

then, channel matching is performed, and the center frequency of the 4-channel even type arrangement polyphase filter is

The 4-channel odd-type polyphase filter has a center frequency of

The details are shown in table 3:

TABLE 3 center frequency of passband of two-stage polyphase Filter Bank

Polyphase filter arrangement	Center frequency (MHz)
		Even type arrangement	0，156.25，312.5，468.75
Odd type arrangement	78.125，234.375，390.625，546.827

According to the principle of minimum number of channels, matching the center frequency of the signal 1 with a table 3, and selecting an even type channelized structure for secondary channelized reception;

carrying out N1/2 times of extraction, filtering and N1 point IFFT processing on a channel where the signal 1 is located, and specifically constructing the steps as the step 1.2;

step 4.2.2: a two-level channelized receiving structure is constructed for the signal 2, note B3 is the bandwidth of the narrowband signal 2 in step 2.6, and the number of channels N2 is represented as:

N2＝B1/B3 (9)

the number N2 of channels is an even number rounded downward, in this case N2 is 32;

then, channel matching is performed, and the center frequency of the 32-channel even type arrangement polyphase filter is

The center frequency of the 32-channel odd-type polyphase filter is

Specifically, as shown in table 4:

TABLE 4 center frequency of passband of two-stage polyphase filter bank

Polyphase filter arrangement	Center frequency (MHz)
		Even type arrangement	0，19.53，39.06，…605.46
Odd type arrangement	9.76，29.30，48.83，…615.23

Matching the center frequency of the signal 2 with the table 3 according to the principle of minimum number of channels, and selecting an even type channelized structure for secondary channelized reception;

carrying out N2/2 times of extraction, filtering and N2 point IFFT processing on a channel where the signal 2 is located, wherein the specific construction steps are shown in step 1.2;

and 5: and a secondary channel detection stage, namely, a secondary channel detection module is constructed, firstly, frequency spectrum smoothing is carried out on the parallel data output in the step 4.2, jitter and burrs are eliminated, then, frequency spectrum difference entropy detection is carried out, false signals are eliminated, and effective signals 2 and effective signals 3 are output, as shown in fig. 7 and 8, noise components in time domain diagrams (7a) and (8a) are greatly reduced, and the diagrams (7b) and (8b) are frequency domain diagrams output by a single channel, and noises of other channels are filtered.

According to the introduction of the above embodiment, the invention is a self-adaptive channelization receiving method based on spectrum differential entropy detection, which adopts a parallel connection structure of even-type arrangement channelization and odd-type arrangement channelization, and the receiving blind area of one set of filter bank is the passband of the other set of filter bank, thereby reducing the cross-channel situation and being capable of receiving without blind area; the method is suitable for complex electromagnetic environment, can simultaneously process narrow-band signals and large-band signals with the bandwidth less than 300MHz, can be packaged into an optional functional module according to specific environment requirements, and has strong flexibility; by the aid of a multi-stage channelization structure combining static state and dynamic state, channel frequency bands where narrow-band signals are located are adaptively subdivided, signal-to-noise ratio of received signals is effectively improved, and the signal-to-noise ratio gain is about 22dB after multi-stage channelization combined processing; the frequency spectrum difference entropy detection method is adopted for channel detection, so that the operation amount is reduced, the influence of noise is reduced in frequency domain detection, and a complex pulse signal with a signal-to-noise ratio of-3 dB can be detected; the narrow-band signal and the wide-band signal are distinguished by adopting a double sliding window energy detection method, and the accuracy is high.

While the foregoing is directed to the preferred embodiment of the present invention, it is not intended that the invention be limited to the embodiment and the drawings disclosed herein. Equivalents and modifications may be made without departing from the spirit of the disclosure, which is to be considered as within the scope of the invention.

Claims

1. The self-adaptive channelized receiving method based on the frequency spectrum differential entropy detection is characterized by comprising the following steps of: a supported adaptive channelized reception system comprising: the device comprises a primary channelized receiving module, a primary channel detection and estimation module, a channel matching module, a secondary self-adaptive channelized receiving module and a secondary channel detection module; the first-stage channelized receiving module comprises an odd-type polyphase filter bank and an even-type polyphase filter bank which are fixed and have the same bandwidth; the primary channel detection and estimation module comprises a channel detection submodule and a frequency spectrum sensing submodule; the secondary self-adaptive channelized receiving module comprises a multi-channel merging module and a dynamic channelized module;

step 1: the first-stage channelized receiving stage is that a first-stage channelized receiving module is constructed, and the first-stage channelized receiving stage also comprises two sub-steps of constructing a complex electromagnetic environment and constructing a multiphase filter bank, and specifically comprises the following steps:

step 1.2: constructing a polyphase filter bank;

the multi-phase filter bank comprises an odd multi-phase filter bank and an even multi-phase filter bank;

step 2.1.3: eliminating false signals caused by a transition band of a low-pass filter, specifically: the time at which the peak of the differential signal of each channel is detected occurs

Or

In the time interval, the false signals are considered to be false signals and are removed; l is the total number of sampling points of the single-channel signal;

step 2.1.5: the difference entropy ratio is used for distinguishing a pure noise channel from an effective signal channel, the detection threshold is recorded as Th1, and H is recorded _min For the minimum value of the differential entropy, go through k if H _k /H _min ＞Th1，H _k Considering that the k channel has effective signal for the differential entropy of the k channel, otherwise, recording the detected signal for the pure noise channelThe number of channels with effective signals is Num;

wherein, the value range of k is 0 to D-1, D is the number of channels and is an even number;

step 2.2: the method comprises the following steps of carrying out double sliding window energy detection on a channel frequency spectrum, judging the position of a signal edge according to the position of a signal ratio exceeding a judgment threshold Th2 in two sliding windows, and further distinguishing a narrow-band signal and a wide-band signal, wherein the method specifically comprises the following steps: for the k channel, the energy E1 of the front window of the ith sampling point is recorded _k (i) Energy of the rear window is E2 _k (i) If E1 _k (i)/E2 _k (i)>Th2, judging that the signal rising edge occurs at the moment; otherwise E1 _k (i)/E2 _k (i) Judging whether a signal falling edge occurs at the moment or not at Th2, pairwise matching the rising edge and the falling edge, judging the channel signal to be a narrow-band signal if the matching occurs in the same channel, and otherwise, judging the channel signal to be a wide-band signal;

wherein, STFT is short-time Fourier transform, and is called short-time Fourier transform; so far, from step 2.1 to step 2.3, a first-stage channel detection and estimation stage is completed;

and 3, step 3: in the channel matching stage, a channel matching module is constructed, specifically: according to the number of channels with effective signals detected in the step 2.1 being Num, comparing the channel number of the signals passing through the even type polyphase filter bank with the channel number of the signals passing through the odd type polyphase filter bank, and selecting the filter bank with less channel number as a channelized receiving structure;

and 4, step 4: a secondary adaptive channelized receiving stage, namely constructing a secondary adaptive channelized receiving module which comprises multi-channel combination and adaptive channelized receiving, and specifically comprising the following steps of:

step 4.1: reserving channel information of the broadband signal, setting other channels to be 0, and inputting the channel information to a dual structure of the polyphase filter for channel combination to obtain the broadband signal;

step 4.2: and (3) self-adaptive channelized reception, specifically, according to the spectrum sensing result in the step 2.2, constructing a self-adaptive dynamic channelized reception structure to subdivide a signal frequency domain, specifically comprising the following substeps:

step 4.2.1: calculating the number N of secondary channelized channels, constructing an even type multiphase filter bank and an odd type multiphase filter bank according to the number of the channels, namely firstly performing N/2 times extraction on the channels subjected to primary channelized, then performing filtering through the filter bank, and finally outputting N sub-channel data through N-point IFFT;

step 4.2.2: matching channels, inputting the narrow-band signals into the two sets of multiphase filter banks constructed in the step 4.2.1 according to the carrier frequency and the bandwidth of the narrow-band signals extracted in the step 2.3, comparing the channel number of the two sets of multiphase filter banks, and selecting the set with the smaller channel number as a secondary channelized receiving structure;

and 5: in the second-level channel detection stage, a second-level channel detection module is constructed, specifically: and performing secondary channel detection by using spectrum differential entropy detection, and outputting a channel where the effective signal is located.

2. The adaptive channelized reception method based on spectral differential entropy detection according to claim 1, characterized in that: the receiving blind area of the odd-type polyphase filter bank in the supported self-adaptive channelized receiving system is exactly the center of the passband of the even-type polyphase filter bank, and the receiving blind area of the even-type polyphase filter bank is exactly the center of the passband of the odd-type polyphase filter bank.

3. The adaptive channelized reception method based on spectral differential entropy detection according to claim 2, characterized in that: a channel detection submodule in the self-adaptive channelized receiving system detects whether a channel has a signal or not by adopting a frequency spectrum differential entropy, and eliminates a false signal caused by a transition band of a filter by utilizing frequency spectrum mutation characteristics.

4. The adaptive channelized reception method based on spectral differential entropy detection according to claim 3, characterized in that: a frequency spectrum sensing submodule in the self-adaptive channelized receiving system extracts the center frequency and bandwidth information of a signal by adopting a short-time Fourier transform technology, the time domain and frequency domain resolution is good, and the signal-to-noise ratio of a weak signal can be greatly improved.

5. The adaptive channelized reception method based on spectral differential entropy detection according to claim 4, characterized in that: the channel matching module in the self-adaptive channelized receiving system selects the set of filter bank with less cross-channel number in the first-stage channelized receiving module according to the result of the channel detection submodule.

6. The adaptive channelized reception method based on spectral differential entropy detection according to claim 5, characterized in that: a multi-channel merging module in the supported self-adaptive channelized receiving system merges broadband signals by utilizing a comprehensive filter bank to output of a first-stage channelized receiving module.

7. The adaptive channelized reception method based on spectral differential entropy detection according to claim 6, characterized in that: the dynamic channelization module in the self-adaptive channelization receiving system can self-adaptively adjust and analyze the structure of the filter bank according to the spectrum sensing result, and the signal-to-noise ratio of the narrow-band signal is maximally improved.

8. The adaptive channelized reception method based on spectral differential entropy detection according to claim 7, characterized in that: a secondary channel detection module in the self-adaptive channelized receiving system is used for detecting by utilizing frequency spectrum differential entropy detection.

9. The adaptive channelized reception method based on spectral differential entropy detection according to claim 8, characterized in that: in step 1.2, the two polyphase filter banks are constructed as follows:

wherein D is the number of channels and is an even number; when an odd-type polyphase filter bank is constructed, each path of data is multiplied by a factor

m is the label of each channel data sampling point, and j is an imaginary number unit;

step 1.2.3: the prototype low pass filter in step 1.2.2 is subjected to a D-fold decimation and 2-fold zero value interpolation operation with 1: d, deserializing the proportion of D to obtain D filter coefficients, and performing low-pass filtering on the parallel data obtained in the step 1.2.1;

wherein, when constructing the odd-type polyphase filter bank, each path of data is multiplied by a factor

Wherein, p is a channel label, and the value range of p is 0 to D-1;

step 1.2.5: after IFFT processing, the odd numbered channels are multiplied by a factor (-1) ^m 。

10. The adaptive channelized reception method based on spectral differential entropy detection according to claim 9, characterized in that: th1 ranges from 1 to 2, Th2 ranges from 1 to 3.