Adaptive Channelization Detection Method Based on Spectrum Sensing

Abstract

The present invention discloses a spectrum-sensing-based adaptive channelization detection method, belonging to the field of electronic reconnaissance, which performs spectrum sensing, spectrum analysis, channel matching, detection, and parameter estimation on electromagnetic signals and outputs detection results. First, the electromagnetic signals in the monitored frequency band are successively subjected to AD sampling and short-time Fourier transform (STFT) to obtain spectrum-sensing data; the spectrum-sensing data is input for real-time energy spectral density estimation to generate a mean spectrum and a maximum spectrum; the frequency bands where the signals are distributed are subjected to spectrum analysis through the mean spectrum and the maximum spectrum to generate channel adaptation parameters; the digital channelization structure is adaptively adjusted according to the channel adaptation parameters to achieve the optimal matching between the signal bandwidth and the channelization bandwidth, and the input electromagnetic signals are subjected to adaptive channelization detection and parameter estimation to complete the output of the final detection results.

Description

Technical Field

[0001] The present invention belongs to the field of electronic reconnaissance, and particularly relates to a spectrum sensing adaptive channelization detection method. Background Art

[0002] With the rapid development of various radio technologies, complex, flexible, and agile signals, as well as radiation source signals of various new and unknown threats, continue to emerge; there are a large number of radio devices with complex waveforms and rich signal styles, and various electromagnetic environments are intertwined, resulting in crowded and overlapping spectra and fluctuating energy; this makes the electromagnetic environment in the electronic battlefield increasingly complex. The traditional signal processing technology systems and processing processes are relatively fixed, and gradually begin to be difficult to meet the requirements of radar countermeasure reconnaissance processing, with deficiencies such as low detection accuracy, weak anti-noise performance, and poor adaptability.

[0003] Developed countries such as the United States are leading in the research on radar countermeasure systems with adaptive and cognitive functions, and some results can be preliminarily applied. Domestically, the processing architectures and processing methods of radar reconnaissance systems with adaptive analysis capabilities are in the initial stage and exploration stage. In the article "Brief Discussion on Spectrum Sensing Technology Research in Cognitive Radio" by Aierken Aizezi, several main spectrum sensing technologies are introduced, the basic principle of energy detection is analyzed in detail, and a cooperative sensing technology based on energy sensing is proposed; this article mainly stays in theoretical analysis, without specific solutions and technical approaches, and has poor adaptability to non-cooperative signals; it is of great significance to study a spectrum sensing adaptive channelization detection method to achieve non-cooperative high-sensitivity signal detection in complex electromagnetic environments. Summary of the Invention

[0004] The present invention proposes a spectrum sensing adaptive channelization detection method, which realizes spectrum sensing adaptive channelization detection in complex electromagnetic environments; through spectrum sensing and spectrum analysis, channel adaptation parameters are obtained to achieve optimal adaptive channelization detection and improve the detection probability; by adopting a combination of composite channels and multi-level channels, the detection, separation, and sensing analysis capabilities of multiple simultaneously arriving communication and radar signals are effectively improved, the accuracy of signal detection is increased, and thus the correct rate of radar modulation recognition is improved.

[0005] The technical solution for realizing the present invention is as follows: A spectrum sensing adaptive channelization detection method, the steps are as follows:

[0006] Step 1: Perform AD sampling on the electromagnetic signals in the monitored frequency band to obtain AD data, and respectively cache and perform short-time Fourier transform on the AD data to obtain cached data and spectrum sensing data; then proceed to Step 2.

[0007] Step 2: Estimate the energy spectral density of the spectrum sensing data to generate the mean spectrum and the maximum value spectrum, and then proceed to Step 3.

[0008] Step 3: Conduct spectrum analysis on the frequency band of the signal distribution through the mean spectrum and the maximum value spectrum to generate channel adaptation parameters, and then proceed to Step 4.

[0009] Step 4: Adaptively adjust the digital channelization structure according to the channel adaptation parameters to obtain an adaptive digital channelization filter structure, and achieve the optimal matching between the signal bandwidth and the channelization bandwidth.

[0010] Step 5: Input the buffered data in Step 1 into the adaptive digital channelization filter structure for adaptive channelization detection and parameter estimation, and complete the output of the final detection result.

[0011] Compared with the prior art, the remarkable advantages of the present invention are as follows:

[0012] 1) The present invention realizes spectrum sensing adaptive channelization detection in a complex electromagnetic environment, and improves the adaptability to the complex electromagnetic environment.

[0013] 2) Through spectrum sensing and spectrum analysis, channel adaptation parameters are obtained to achieve optimal adaptive channelization detection and improve the detection probability.

[0014] 3) A combination method of using composite channels and multi-level channels is proposed, which effectively improves the detection, separation, and sensing analysis capabilities of multiple simultaneously arriving communication and radar signals, improves the accuracy of signal detection, and further improves the correct rate of radar modulation recognition.

[0015] 4) It is suitable for processing on an FPGA, realizes high-speed, real-time, and fully pipelined signal detection, and has high engineering applicability. Description of the Drawings

[0016] Figure 1 It is a flow chart of the spectrum sensing adaptive channelization detection method of the present invention.

[0017] Figure 2 It is a block diagram of FFT equivalent operation processing.

[0018] Figure 3 It is a sliding diagram of the time domain window.

[0019] Figure 4 It is a schematic diagram of channel merging.

[0020] Figure 5 It is a schematic diagram of channel splitting. Detailed Embodiment

[0021] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0022] In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions conflicts with each other or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0023] Next, the specific implementation manners, as well as the technical difficulties and inventive points of the present invention, will be further introduced in combination with the design examples.

[0024] Combined with Figures 1 to 5 , a spectrum sensing adaptive channelization detection method is as follows:

[0025] Step 1: Perform AD sampling on the electromagnetic signals in the monitored frequency band to obtain AD data. After performing short-time Fourier transform (STFT) on the AD data, spectrum sensing data is obtained. At the same time, the AD data is cached to obtain cached data, specifically as follows:

[0026] Step 11: Perform sliding update on the AD data to ensure that the input data has no overlap. Divide consecutive M points into one frame of AD sub-data, where M = 256, 512, 1024, 2048...; when performing short-time Fourier transform, the smaller the number of sliding points, the smaller the probability of signal loss and the higher the time resolution. Therefore, if each time an FFT is performed, the data only slides by one point, then it can be ensured that the signal data is completely not lost, which is called 100% data overlap; the number of sliding points M is N / 2, where N is the number of FFT points, that is, 50% overlap; the number of sliding points M is N, where N is the number of FFT points, that is, no overlap. When there is no data overlap, M = N.

[0027] Step 12: Perform 8-way FFT parallel operation on each frame of AD sub-data:

[0028] Perform continuous 8-way N / 8-point FFT parallel operation on one frame of AD sub-data, where N is the number of FFT points. Perform frequency-domain accumulation on the outputs of 8 consecutive FFT parallel operations to generate one frame of spectrum measurement results, and output it as the spectrum sensing result of one frame of AD sub-data; define the AD data sampling rate as Fs, then the frequency resolution of each sampling point in the spectrum sensing result of one frame of AD sub-data is Fs / N.

[0029] When \(X(k)\), \(k = 0, 1, \ldots, N / 8 - 1\), \(X(k)\) represents the result of an \(N / 8\)-point FFT, and \(k\) represents the sequence number. The expressions of \(X(k + N / 8)\), \(X(k + 2N / 8)\), \(X(k + 3N / 8)\), \(X(k + 4N / 8)\), \(X(k + 5N / 8)\), \(X(k + 6N / 8)\), and \(X(k + 7N / 8)\) are derived from the calculation of \(X(k)\) to achieve the \(N\)-point spectrum operation of 8-channel data.

[0030] Let \(N = 4096\), and use continuous 8 parallel 512-point FFT operations. Then, the output effect of a 4096-point FFT is equivalently obtained by deriving from the outputs of continuous 8 FFT operations (as Figure 2 shown). When \(X(k)\), \(k = 0, 1, \ldots, 511\), \(k\) represents the sequence number, and the calculation and derivation process of \(X(k)\) is as follows:

[0031]

[0032] Among them, \(n\), \(m\), and \(k\) represent the sequence number, and \(j\) represents the imaginary number.

[0033] When \(X(k + 512)\), \(k = 0, 1, \ldots, 511\), that is, replace \(k\) in the above \(X(k)\) with \(k + 512\), and after simplification, the expression of \(X(k + 512)\) can be obtained. The calculation and derivation process of \(X(k + 512)\) is as follows:

[0034]

[0035] When \(X(k + 1024)\), \(k = 0, 1, \ldots, 511\), that is, replace \(k\) in the above \(X(k)\) with \(k + 1024\), and after simplification, the expression of \(X(k + 1024)\) can be obtained. The calculation and derivation process of \(X(k + 1024)\) is as follows:

[0036]

[0037] And so on, the expressions of \(X(k + 1536)\), \(X(k + 2048)\), \(X(k + 2560)\), \(X(k + 3072)\), and \(X(k + 3584)\) can be obtained. In this way, the \(N\)-point spectrum operation of 8-channel data can be achieved.

[0038] As shown in Table 1, it is the 4096-point distribution form after the butterfly calculation. The numbers in the first column represent the arrangement order of 8 channels, and each row represents the sequence of spectrum points.

[0039] Table 1 Distribution Table of FFT Operation Results

[0040] 0 0 1 2 ... 511 4 2048+0 2048+1 2048+2 ... 2048+511 2 1024+0 1024+1 1024+2 ... 1024+511 6 3072+0 3072+1 3072+2 ... 3072+511 1 512+0 512+1 512+2 ... 512+511 5 2560+0 2560+1 2560+2 ... 2560+511 3 1536+0 1536+1 1536+2 ... 1536+511 7 3584+0 3584+1 3584+2 ... 3584+511

[0041] After performing frequency-domain accumulation processing on the outputs of 8 consecutive 512-point FFT operations, a frame of 4096-point spectrum measurement results is generated and output as the spectrum sensing result of a frame of AD data. Given that the sampling rate of the AD data is 2400 MHz, the frequency resolution of each spectrum sensing result output is 0.5859375 MHz.

[0042] Step 13: For the next frame of AD sub-data, repeat Step 12 and correspondingly output the spectrum sensing data of this frame of AD sub-data.

[0043] Step 14: Cache the AD data to obtain cached data. According to the synchronization of AD data processing, caching the input AD data needs to be aligned with the output of the spectrum analysis adaptive digital channelization filter structure to achieve synchronous analysis processing.

[0044] Step 2: Estimate the energy spectral density of the spectrum sensing data to obtain the energy values of each spectrum point. Calculate the mean and maximum values of the corresponding frequency points in multiple frames of spectrum sensing data to obtain the mean spectrum and the maximum value spectrum.

[0045] In Step 3, perform spectrum analysis on the frequency bands where the signals are distributed through the mean spectrum and the maximum value spectrum to generate channel adaptation parameters. Compared with the prior art, it can achieve adaptive channel bandwidth adjustment and optimal matched filtering, improving the detection probability. Specifically, it includes the following steps:

[0046] Step 31: Sort the data in the mean spectrum by size, take the mean of the smallest 256 data points, and raise the above mean by X1 dB, where X1 takes an empirical value of 3 - 6, as the spectral detection noise threshold.

[0047] Step 32: Perform a moving average on the data in the maximum value spectrum for X2 consecutive points to obtain the smoothed result of the maximum value spectrum, and perform a maximum value expansion on the smoothed result, with the expansion points being X3; according to the envelope of the maximum value spectrum after the maximum value expansion, multiply it by the real-time threshold adjustment coefficient as the real-time spectral signal envelope threshold; where the range of X2 is 1 - N / 16, and the range of X3 is 2 - N / 8.

[0048] Step 33: Compare the real-time spectral signal envelope threshold with the spectral detection noise threshold, and take the larger value of the two as the spectral detection threshold.

[0049] Step 34: Compare the maximum value spectrum with the spectral detection threshold to obtain the detected VP; a detected VP of '1' represents a signal bandwidth of Fs / N, and a continuous D-point VP value of '1' indicates a signal bandwidth of Fs / N * D; generate channel adaptation parameters according to the distribution of the detected VP, including signal frequency distribution and signal bandwidth information;

[0050] Step 4: Perform adaptive adjustment on the digital channelization structure according to the channel adaptation parameters to obtain an adaptive digital channelization filter structure, achieving the optimal matching between the signal bandwidth and the channelization bandwidth. The specific steps are as follows:

[0051] Step 41: According to the channel adaptation parameters, determine the signal frequency distribution, signal bandwidth, and adjacent signal frequency interval, and perform channel merging according to the following criteria:

[0052] 1). N / 512 is the minimum interval of a single channel, that is, 2 n The minimum value is N / 512; when N is 4096, first merge 8 adjacent channels into 1 channel (channel interval 4.6875 MHz), then 4096 channels are merged into 512 channels;

[0053] 2). Within 256 channels, the merged channel interval L is at intervals of powers of 2 (2 n ); merge according to 1 channel, 2 channels, 4 channels... 2 n channels.

[0054] 3). Within 512 channels, the merged channel interval is L, and the remainder of the starting channel n / L is 1.

[0055] 4). The maximum interval of a single channel is N / 16, that is, 2 n The maximum value is N / 16; when N is 4096, at most merge the results of 256 consecutive adjacent channels into 1 channel, corresponding to a channel interval of 150 MHz, then 4096 channels are merged into 16 channels.

[0056] Design the digital channelization distribution according to the above 4 criteria.

[0057] As Figure 4 shown, channels 3 and 4 are merged into a new channel 3, channels 5 and 6 are merged into a new channel 5, channels 7 and 8 are merged into a new channel 7, and channels 13, 14, 15, and 16 are merged into a new channel 13.

[0058] Step 42: According to the channel merging result, automatically select the matching digital channelization polyphase filter to form a hybrid channel polyphase filter mode, ensuring the optimal matching between the signal bandwidth and the channelization bandwidth.

[0059] Step 43: For the channel adaptation parameters, if there are multiple signals within a minimum channel interval of 4.6875 MHz, select the filtering data of this channel, split the channel according to the signal frequency and bandwidth, and use the method of a two-stage tracking filter to form a multi-stage channel filter mode; as Figure 5 shown, there are 2 signals in channel 5, and channel 5 is split into two secondary sub-channels, channel 5-1 and channel 5-2.

[0060] Step 44: Form an adaptive digital channelization filter structure according to the filtering architecture combining a composite channel and a multi-level channel.

[0061] Step 5: Perform adaptive channelization detection and parameter estimation on the input buffered data to complete the output of the final detection result, which specifically includes the following steps:

[0062] Step 51: Input the buffered data into the adaptive digital channelization filter to obtain the filtered data of each channel; estimate the background noise of the filtered data to form a noise threshold.

[0063] Step 52: Perform time-domain energy accumulation of the filtered data with different numbers of points to generate a signal accumulation envelope; generate an adaptive detection threshold according to the noise threshold and the signal accumulation envelope.

[0064] Step 53: Obtain the adaptive detection VP by comparing the signal accumulation envelope with the adaptive detection threshold; eliminate false signals with small pulse widths according to the combined edge characteristic information to obtain the adaptive detection result; determine the starting frequency F1 and the ending frequency F2 of the channel according to the bandwidth and the position of the channel where the adaptive detection result is located, and at the same time perform FFT processing on the filtered data of this channel, and take the FFT peak position and the 3dB bandwidth to obtain the precise frequency and bandwidth of the signal.

[0065] Step 54: Obtain the precise pulse arrival time TOA and pulse width PW according to the adaptive detection VP and the combined edge characteristic information to complete the output of the final detection result.

Claims

1. A spectrum sensing-based adaptive channelization detection method, characterized in that It includes the following steps: Step 1: Perform AD sampling on the electromagnetic signals of the monitored video band to obtain AD data, and respectively cache and perform short-time Fourier transform on the AD data, corresponding to obtaining cached data and spectrum sensing data; specifically as follows: Step 11: Perform sliding update on the AD data to ensure that the input data has no overlap, and divide consecutive M points into one frame of AD sub-data, where M = 256, 512, 1024, 2048...; Step 12: Perform 8-way FFT parallel operation on each frame of AD sub-data: Perform continuous 8-way N / 8-point FFT parallel operation on one frame of AD sub-data, where N is the FFT point number. Perform frequency-domain accumulation on the outputs of continuous 8 FFT parallel operations to generate one frame of spectrum measurement results, and output it as the spectrum sensing result of one frame of AD sub-data; define the AD data sampling rate as Fs, then the frequency resolution of each sampling point in the spectrum sensing result of one frame of AD sub-data is Fs / N; When is represents the N / 8-point FFT result, k represents the sequence number, and through the calculation and derivation of and the expressions of , , , , , , , the N-point spectrum operation of 8-channel data is realized; Step 13: For the next frame of AD sub-data, repeat Step 12, and correspondingly output the spectrum sensing data of this frame of AD sub-data; Step 14: Cache the AD data to obtain cached data; Proceed to Step 2; Step 2: Perform energy spectral density estimation on the spectrum sensing data to generate a mean spectrum and a maximum spectrum, and then proceed to Step 3; Step 3: Perform spectrum analysis on the frequency bands of signal distribution through the mean spectrum and the maximum spectrum to generate channel adaptation parameters, specifically including the following steps: Step 31: Sort the data in the mean spectrum by size, take the mean of the smallest 256 data points, and raise the above mean by X1 dB as the spectral detection noise threshold; Step 32: Perform sliding average on the data in the maximum spectrum for consecutive X2 points to obtain the smoothed result of the maximum spectrum. Perform maximum value expansion on the smoothed result, and the expansion point number is X3; According to the envelope of the maximum spectrum after maximum value expansion, multiply it by the real-time threshold adjustment coefficient as the real-time spectral signal envelope threshold; where the range of X2 is 1~N / 16, and the range of X3 is 2~N / 8; Step 33: Compare the real-time spectral signal envelope threshold with the spectral detection noise threshold, and take the larger value of the two as the spectral detection threshold; Step 34: Perform a threshold-crossing comparison between the maximum spectrum and the spectral detection threshold to obtain the detected VP; When the detected VP is '1', it represents the signal bandwidth Fs / N, and when the VP value of consecutive D points is '1', it indicates that the signal bandwidth is Fs / N*D; Generate channel adaptation parameters according to the distribution of the detected VP, including signal frequency distribution and signal bandwidth information; Proceed to Step 4; Step 4: Perform adaptive adjustment on the digital channelization structure according to the channel adaptation parameters to obtain an adaptive digital channelization filter structure, and realize the optimal matching of the signal bandwidth and the channelization bandwidth; Step 5: Input the cached data in Step 1 into the adaptive digital channelization filter structure to perform adaptive channelization detection and parameter estimation, and complete the output of the final detection result.

2. The method for spectrum-sensing-based adaptive channelization detection according to claim 1, wherein: When the data has no overlap, M = N.

3. The method for spectrum-sensing adaptive channelization detection according to claim 1, wherein In Step 2, perform energy spectral density estimation on the spectrum sensing data to obtain the energy values of each spectrum point, and calculate the mean and maximum values of the corresponding frequency points in multiple frames of spectrum sensing data to obtain the mean spectrum and the maximum spectrum.

4. The method for spectrum sensing-based adaptive channelization detection according to claim 1, characterized in that, In step 4, the digital channelization structure is adaptively adjusted according to the channel adaptation parameters to obtain an adaptive digital channelization filter structure, realizing the optimal matching of the signal bandwidth and the channelization bandwidth. The specific steps are as follows: Step 41: According to the channel adaptation parameters, determine the signal frequency distribution, signal bandwidth, and adjacent signal frequency interval, and merge channels according to the following criteria: 1). N / 512 is the minimum interval of a single channel, that is, 2 n The minimum value is N / 512; first, N / 512 adjacent channels are merged into 1 channel in sequence, and the channel interval is Fs / 512 MHz, then N channels are merged into 512 channels; 2), within 256 channels, the combined channel interval L is an exponent of 2 n as the interval: combine according to 1 channel, 2 channels, 4 channels... 2 n channels for combination; 3) Within 512 channels, the merged channel interval is L, and the remainder of the starting channel n / L is 1; 4) The maximum interval of a single channel is N / 16, i.e., 2 n The maximum value is N / 16. At most, the results of N / 16 consecutive adjacent channels are merged into 1 channel. The channel interval is Fs / 16 MHz, so N channels are merged into 16 channels; The digital channelization distribution is designed according to the above 4 criteria; Channels 3 and 4 are merged into a new channel 3, channels 5 and 6 are merged into a new channel 5, channels 7 and 8 are merged into a new channel 7, and channels 13, 14, 15, and 16 are merged into a new channel 13; Step 42: According to the channel merging result, automatically select the matching digital channelization polyphase filter to form a hybrid channel polyphase filter mode, ensuring the optimal matching of the signal bandwidth and the channelization bandwidth; Step 43: For the channel adaptation parameters, if there are multiple signals within a minimum channel interval N / 512, select the filtered data of this channel, generate a secondary tracking filter according to the signal frequency and bandwidth, and form a multi-level channel filter mode; Step 44: According to the filtering architecture combining the composite channel and the multi-level channel, form an adaptive digital channelization filter structure.

5. The method for spectrum-sensing-based adaptive channelization detection according to claim 4, wherein In step 5, the buffered data in step 1 is input into the adaptive digital channelization filter structure for adaptive channelization detection and parameter estimation, and the final detection result is output. The specific steps are as follows: Step 51: Input the buffered data into the adaptive digital channelization filter structure to obtain the filtered data of each channel; estimate the background noise of the filtered data to form a noise threshold; Step 52: Accumulate the time-domain energy of the filtered data with different numbers of points to generate a signal accumulation envelope; generate an adaptive detection threshold according to the noise threshold and the signal accumulation envelope; Step 53: Compare the signal accumulation envelope with the adaptive detection threshold to obtain the adaptive detection VP; eliminate false signals with small pulse widths according to the combined edge characteristic information to obtain the adaptive detection result; determine the starting frequency F1 and the ending frequency F2 of the channel according to the bandwidth and position of the channel where the adaptive detection result is located, and at the same time perform FFT processing on the filtered data of this channel, and take the FFT peak position and 3dB bandwidth to obtain the accurate signal frequency and bandwidth; Step 54: According to the adaptive detection VP, combine the edge characteristic information to obtain the accurate pulse arrival time TOA and pulse width PW, and complete the output of the final detection result.

Images