Frequency hopping detection result completion method based on distributed multi-station monitoring equipment

Abstract

The present invention is a method for completing frequency hopping detection results based on distributed multi-station monitoring equipment. A plurality of receiving stations are used to receive multi-frequency hopping signals respectively, and the multi-frequency hopping signals received by each receiving station are detected, and each hopping signal is separated and extracted; a multi-station coordinated frequency hopping pattern completion operation is performed according to each hopping signal of each receiving station, and finally the frequency hopping pattern detection result of each receiving station is obtained; the multi-station coordinated frequency hopping pattern completion operation includes missed detection completion and / or wrong detection correction; the present invention adopts a plurality of receiving stations with a distributed layout to receive multi-frequency hopping signals, and uses the frequency hopping patterns received by other receiving stations to complete the frequency hopping patterns not detected by a certain receiving station, and corrects the frequency hopping patterns detected by a certain receiving station by mistake, thereby solving the problem that the traditional single-station frequency hopping detection cannot be corrected when an error occurs, which has a serious impact on subsequent signal reconnaissance.

Description

Technical Field

[0001] The invention belongs to the field of communication technology, and in particular relates to a method for completing frequency hopping detection results based on distributed multi-station monitoring equipment. Background Art

[0002] In future information confrontation, electromagnetic threats are more dynamic and uncertain than threats in other dimensions, and require higher accuracy, real-time and dynamic performance of communication intelligence reconnaissance. Traditional fixed-frequency communication technology can no longer meet the requirements of reliable and secure communication to a certain extent. In order to ensure low interception, anti-interference and stability of communication signals, frequency hopping communication systems have emerged. Frequency hopping communication has been widely used in various information systems due to its advantages such as good confidentiality, strong anti-interference ability, low interception probability and flexible networking ability. Frequency hopping communication technology itself has strong anti-interference ability, which leads to severe challenges for communication reconnaissance under non-cooperative conditions. The reconnaissance of frequency hopping signals is to complete the tasks of interception, detection, parameter estimation, network station sorting, de-hopping and demodulation of frequency hopping signals, and then obtain the original baseband signal and analyze the communication information hidden in the baseband signal.

[0003] The detection of frequency hopping signals is to intercept unknown frequency hopping signals, including detecting frequency hopping signals mixed in various modulation modes and noise, and removing useless signals other than frequency hopping signals, such as fixed frequency, interference, etc. Most of the technologies used for frequency hopping signal detection are now mature, mainly including channelized reception, autocorrelation detection and time-frequency analysis. The existing frequency hopping signal detection technology mainly uses energy detection method, correlation detection method and detection method based on time-frequency analysis to detect the signal and obtain the start and end time and start and end frequency of the signal. In recent years, with the development of the artificial intelligence industry, many scholars at home and abroad have gradually applied deep learning to blind detection of communication signals. Among them, some scholars directly use the signal time domain sequence to train the network for detection, allowing the network to extract features from the signal by itself; some scholars first obtain the signal frequency domain information through time-frequency transformation and then send it to the neural network. These methods can achieve good blind detection results.

[0004] However, the existing traditional frequency hopping signal detection technology uses a single station to perform frequency hopping signal detection. During the detection process, it is very likely that missed detections will occur due to obstruction and interference in the signal transmission process, and false detections will occur in which two hopping signals with similar time and frequency are detected as one hopping signal. These missed and erroneous detection results will have a serious impact on the subsequent parameter estimation and signal sorting steps of frequency hopping reconnaissance. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a method for completing frequency hopping detection results based on distributed multi-station monitoring equipment, which adopts multiple receiving stations with a distributed layout to receive multi-frequency hopping signals, and uses the frequency hopping patterns received by other receiving stations to complete the frequency hopping patterns not detected by a certain receiving station, and corrects the frequency hopping patterns misdetected by a certain receiving station, thereby solving the problem that the traditional single-station frequency hopping detection cannot be corrected when errors occur, which has a serious impact on subsequent signal reconnaissance.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] A method for completing frequency hopping detection results based on distributed multi-station monitoring equipment comprises the following steps:

[0008] Step 1, using multiple receiving stations to respectively receive multi-frequency hopping signals, wherein the multiple receiving stations use a distributed architecture;

[0009] Step 2, detecting the multi-frequency hopping signal received by each receiving station, and separating and extracting each hopping signal;

[0010] Step 3, according to each hopping signal of each receiving station obtained, a multi-station coordinated frequency hopping pattern completion operation is performed to finally obtain the frequency hopping pattern detection result of each receiving station; the multi-station coordinated frequency hopping pattern completion operation includes missed detection completion and / or wrong detection correction; the missed detection completion is to judge the missed detection signal based on the similarity of the starting time, the similarity of the frequency and the correlation of the signal, and to complete the frequency information and the time information according to the signal time difference; the wrong detection correction is to judge the wrong detection signal based on the similarity of the starting time, the similarity of the ending time, the similarity of the starting frequency, the similarity of the ending frequency and the correlation of the signal, and use the time and frequency information obtained by the receiving station without wrong detection to correct the starting time and the ending time through the arrival time difference of the two station signals; or correct the starting time and the ending time through the arrival time difference of the two station signals.

[0011] In one embodiment, in step 1, the multiple receiving stations adopt a distributed architecture, which means that each receiving station is within the receiving range of the frequency hopping signal.

[0012] In one embodiment, the step 2 uses a frequency hopping signal detection algorithm to obtain the starting frequency f of each hopping signal received by the receiving station. start , stop frequency f end , take-off time t start and the end time t end ; First use the starting time t of each jump signal start and the end time t end The signal is intercepted in the time domain, and then according to the starting frequency fstart With the stop frequency f end The carrier frequency and bandwidth of each hop signal are calculated, and each hop signal is filtered in the frequency domain to separate the independence of each hop signal.

[0013] In one embodiment, the frequency domain filtering is to use a bandpass filter to perform frequency domain adaptive filtering on each hop signal in the frequency domain according to the relationship between bandwidth and sampling frequency; the design method of the bandpass filter is as follows:

[0014] First, according to the starting frequency f of each hopping signal start With the stop frequency f end , calculate the filter bandwidth B required for the hopping signal, B = (f end -f start )2, and then calculate the bandwidth B and the original sampling rate f s The ratio Bf s , generate the corresponding low-pass filter, and then perform spectrum shifting on the low-pass filter. By multiplying the filter by the signal carrier frequency, the filter is shifted from zero frequency to the frequency band where the hopping signal is located, that is, the frequency where the band-pass filter is located, and finally the required band-pass filter is generated.

[0015] In one embodiment, the missed detection and completion method is as follows:

[0016] s11, for each hop signal of each receiving station, the jump time t start , center frequency comparison, and set thresholds to judge similarity;

[0017] s12, for the frequency hopping signals that simultaneously satisfy the similarity of the take-off time and the similarity of the center frequency, correlation calculation is performed between the two signals to obtain the peak value of the cross-correlation result of the two signals and the corresponding time difference τ;

[0018] s13, setting a threshold to determine and screen the size of the cross-correlation peak of the two signals. The correlation result after cross-correlation has an obvious peak, and its size is significantly larger than the result obtained by performing correlation operation after different hopping signals arrive at the receiving station. In this way, the same hopping signal received by the two receiving stations is screened out, and it is determined that the hopping signal is not missed, and the time difference τ′ between the two receiving stations receiving the hopping signal is recorded; when the signal A1 detected by the receiving station A and all the signals detected by other receiving stations cannot simultaneously meet the standards of similarity of the starting time, similarity of the center frequency and signal correlation, it is determined that the other receiving stations have missed the detection of the signal A1;

[0019] s14, for the missed detection signal, the frequency information of the missed detection receiving station detection result is supplemented and the signal is extracted by using the receiving station A, the extracted signal is correlated with the signal A1 detected by the receiving station A, and the time difference between the extracted signal and the signal A1 is obtained, and the time difference is added to the original start time and end time in the detection result of the missed detection receiving station, that is, the time information of the missed detection signal of other receiving stations is supplemented according to the detection result of the receiving station A;

[0020] s15, traverse all receiving stations to detect all hopping signals, and complete the frequency information and time information of the missed detection signals of other receiving stations, and finally complete the completion of the frequency hopping patterns detected by all receiving stations.

[0021] In one embodiment, the error detection and correction method is as follows:

[0022] s21, for each hop signal of each receiving station, the jump time t start , end time t end Starting frequency f start and the stop frequency f end Compare and set thresholds for similarity judgment;

[0023] s22, for the frequency hopping signals that simultaneously meet the similarity of the starting time, the similarity of the ending time, the similarity of the starting frequency, and the similarity of the ending frequency, correlation calculation is performed between two of them to obtain the peak value of the cross-correlation result of the two signals and the corresponding time difference τ;

[0024] s23, setting a threshold K to determine and screen the size of the cross-correlation peak of the two signals. If the cross-correlation peak is greater than the given threshold, the two signals are determined to be received signals of the same hop transmission signal that are transmitted to the two receiving stations, and at the same time, it is determined that the hop signal does not cause false detection in the two receiving stations;

[0025] s24, for two signals that meet the similarity of the starting time and the similarity of the ending time, but only meet one of the similarity of the starting frequency and the similarity of the ending frequency, calculate the correlation, and obtain the cross-correlation peak value and the corresponding time difference τ″. If the cross-correlation peak value is greater than the threshold value K, it is determined that a false detection occurs. The false detection situation is: a certain receiving station mistakenly detects two frequency hopping signals with similar time and frequency as the same hopping signal; the starting time and the ending time obtained by the receiving station without false detection are added to the time difference τ″ to realize information correction and completion of the false detection signal;

[0026] s25, for two signals that satisfy the starting frequency similarity and the ending frequency similarity, but only satisfy one of the starting time similarity and the ending time similarity, the correlation calculation is performed to obtain the cross-correlation peak value and the corresponding time difference τ″′. If the cross-correlation peak value is greater than the threshold K, it is determined that a false detection occurs. The false detection situation is: a certain receiving station mistakenly detects two frequency-hopping signals with similar frequencies and close times as the same hopping signal; the starting time and ending time obtained by the receiving station where no false detection occurs are added to the time difference τ″′ to realize information correction and completion of the false detection signal.

[0027] In one embodiment, the correlation calculation is to calculate the similarity between the two signals, and the similarity is measured by cross-correlation. For the same source transmission signal s(t), the signals received by the two receiving stations are x(t) and y(t), respectively, then the cross-correlation function of x(t) and y(t) is expressed as:

[0028]

[0029] In the formula, A represents the ratio of the amplitudes of the two received signals, A = η 2 / η 1 , η 1 , η 2 They represent the amplitude gain of s(t) after passing through the two transmission channels, d represents the time difference between the two signals, and d = τ 2 -τ 1 , τ 1 and τ 2 They represent the time difference between the two signals arriving at the receiving station; τ represents the time difference between x(t) and y(t) (i.e., the time difference estimated at the receiving end after being affected by noise, etc.), R ss ( ) represents the autocorrelation function of the source transmitted signal s(t), represents the interference noise n of the first transmission line 1 The cross-correlation function of (t) and the source transmitted signal s(t), represents the interference noise n of the second transmission line 2 The cross-correlation function of (t) and the source transmitted signal s(t), Indicates n 1 (t) and n 2 (t) is the cross-correlation function of

[0030] In one embodiment, the interference noise n 1 (t) and n 2 (t) is a stationary and uncorrelated Gaussian white noise independent of s(t), then and are all 0; then:

[0031] When τ=d, the cross-correlation function can reach its maximum value, and τ at this time is the time difference between the two signals.

[0032] In one embodiment, after completing the frequency hopping pattern completion operation of multi-station coordination, independence extraction is performed on all signals for subsequent estimation of multi-station parameters and sorting of frequency hopping signals.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] For frequency hopping signal detection, the existing technical solutions may miss detection and misdetection of frequency hopping signals due to obstruction or interference signals during transmission when a single monitoring station conducts reconnaissance and detection of frequency hopping signals. In this case, it is difficult to improve the traditional single-station frequency hopping detection technology. The present invention is based on distributed multi-station collaboration and performs multi-station completion for the detection results. The frequency hopping patterns received by other receiving stations are used to complete the frequency hopping patterns not detected by a certain receiving station and to correct the frequency hopping patterns misdetected by a certain receiving station. Compared with the frequency hopping detection algorithm based on the time-frequency matrix with better detection performance in the current single-station detection, the present invention has better detection performance, can greatly improve the accuracy of frequency hopping signal detection, and has better detection effect under low signal-to-noise ratio, and the algorithm has better robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of the method of the present invention.

[0036] Figure 2 It is a design flow chart of the adaptive filter of the present invention.

[0037] Figure 3 It is a flow chart of the multi-station frequency hopping pattern missed detection completion method based on the detection results of the present invention.

[0038] Figure 4 Schematic diagram of the cross-correlation of signals at two receiving stations of the present invention, wherein: (a) is the correlation result after the same hop signal arrives at the two receiving stations; (b) is the correlation result after different hop signals arrive at the two receiving stations.

[0039] Figure 5 It is a flow chart of the multi-station frequency hopping pattern error detection correction method based on the detection results of the present invention.

[0040] Figure 6 It is a background image of the frequency hopping signal pattern completion and sorting experimental environment of the present invention.

[0041] Figure 7 This is a diagram showing the effect of detecting a single-station frequency hopping signal under missed detection conditions of the present invention.

[0042] Figure 8The figures are effect diagrams of multi-station frequency hopping signal detection and frequency hopping pattern completion of the present invention, wherein (a) is the detection effect diagram of monitoring station 1; (b) is the detection effect diagram of monitoring station 2; (c) is the missed detection completion effect diagram of monitoring station 1; (d) is the missed detection completion effect diagram of monitoring station 2; (e) is the comparison between the missed detection completion of monitoring station 1 and the real frame; (f) is the comparison between the missed detection completion of monitoring station 2 and the real frame; (g) is the comparison and enlargement of the missed detection completion of monitoring station 1 and the real frame; (h) is the comparison and enlargement of the missed detection completion of monitoring station 2 and the real frame.

[0043] Fig. 9 The figures are single-station frequency hopping signal detection effect diagrams under the condition of misdetection of the present invention, wherein (a) is a single-station detection effect diagram; (b) is an enlarged single-station detection effect diagram; and (c) is an enlarged single-station detection effect diagram.

[0044] Fig.10 1 is a diagram of the detection and correction of frequency hopping signals at multiple stations according to the present invention, wherein (a) is a diagram of the detection of monitoring station 1; (b) is a diagram of the detection of monitoring station 2; (c) is an enlarged diagram of the detection of monitoring station 1; (d) is an enlarged diagram of the detection of monitoring station 1; (e) is an enlarged diagram of the detection of monitoring station 2; (f) is an enlarged diagram of the detection of monitoring station 2; (g) is an error detection correction diagram of monitoring station 1; (h) is an error detection correction diagram of monitoring station 2; (i) is an enlarged diagram of the error detection correction diagram of monitoring station 1; (j) is (k) is an enlarged picture of the effect of the error detection correction of monitoring station 1; (l) is an enlarged picture of the effect of the error detection correction of monitoring station 2; (m) is the comparison between the error detection correction of monitoring station 1 and the real frame; (n) is the comparison between the error detection correction of monitoring station 2 and the real frame; (o) is an enlarged comparison between the error detection correction of monitoring station 1 and the real frame; (p) is an enlarged comparison between the error detection correction of monitoring station 1 and the real frame; (q) is an enlarged comparison between the error detection correction of monitoring station 2 and the real frame; (r) is an enlarged comparison between the error detection correction of monitoring station 2 and the real frame.

[0045] Fig.11 It is a comparison curve diagram of the multi-station collaborative completion detection performance of the present invention. DETAILED DESCRIPTION

[0046] The embodiments of the present invention are described in detail below with reference to the accompanying drawings and examples.

[0047] The reconnaissance and monitoring of frequency hopping signals are based on the detection results of frequency hopping signals. However, frequency hopping signals may be blocked or interfered with during transmission from the transmitting station to the receiving station. Using the traditional single-station signal detection algorithm, the receiving station may fail to detect a certain hopping signal, i.e., miss detection; or misdetect two hopping signals with similar hopping time and frequency as one hopping signal, i.e., misdetection. Missed and erroneous detection results will have a serious impact on the subsequent parameter estimation and signal sorting of frequency hopping reconnaissance.

[0048] Therefore, the present invention proposes a method for completing frequency hopping detection results based on distributed multi-station monitoring equipment. The method uses the broadband frequency hopping signal detection results based on the time-frequency matrix to extract the independence of each detected frequency hopping signal, and matches, completes and corrects multi-station detection signals based on time similarity, frequency similarity and signal correlation.

[0049] refer to Figure 1 The present invention provides a method for completing frequency hopping detection results based on a distributed multi-station monitoring device, comprising the following steps:

[0050] Step 1: Use multiple receiving stations to receive multi-frequency hopping signals respectively.

[0051] The multiple receiving stations described in the present invention adopt a distributed architecture, that is, each receiving station is within the receiving range of the frequency hopping signal.

[0052] In the embodiment of the present invention, two receiving stations are used for the convenience of description, namely Figure 1 The receiving stations shown are receiving station 1 and receiving station 2. In fact, the present invention can use any number of receiving stations greater than 2, and obviously the more the number, the better the effect may be, but the computational complexity will also increase accordingly.

[0053] Step 2: Detect the multi-frequency hopping signal received by each receiving station, and separate and extract each hopping signal.

[0054] For example, in this step, a frequency hopping signal detection algorithm, such as a traditional energy detection method and a deep learning frequency hopping signal detection algorithm, is used to obtain the starting frequency f of each hopping signal received by the receiving station. start , stop frequency f end , take-off time t start and the end time t end Based on these time-frequency information, the independence of each hop signal received by multiple stations can be extracted. First, the time parameter of each hop signal detected (i.e., the starting time t start and the end time t end ) intercepts the signal in the time domain, and then gives the starting frequency f according to the detection result start With the stop frequency f end The carrier frequency and bandwidth of each hop signal are calculated, and each hop signal is filtered in the frequency domain to separate the independence of each hop signal.

[0055] The above-mentioned frequency domain filtering uses a bandpass filter to perform frequency domain adaptive filtering on each hop signal in the frequency domain according to the relationship between bandwidth and sampling frequency; that is, the bandpass filter of the present invention should be an adaptive filter, which has the following three characteristics:

[0056] 1) The filter bandwidth can be adaptively changed as the bandwidth of the target signal to be filtered changes.

[0057] 2) The filter is a bandpass filter, and the passband frequency position is determined by the frequency range occupied by the target signal.

[0058] 3) The filter is adaptive to the sampling rate of the original signal.

[0059] The embodiment of the present invention provides a design method of the bandpass filter, such as Figure 2 As shown, the process is as follows:

[0060] First, according to the starting frequency f of each hopping signal in the broadband frequency hopping signal detection result start With the stop frequency f end , calculate the filter bandwidth B required for the hopping signal, B = (f end -f start )2, and then calculate the bandwidth B and the original sampling rate f s The ratio Bf s , generate the corresponding low-pass filter, and then perform spectrum shifting on the low-pass filter. By multiplying the filter by the signal carrier frequency, the filter is shifted from zero frequency to the frequency band where the hopping signal is located, that is, the frequency where the band-pass filter is located, and finally the required band-pass filter is generated.

[0061] After completing the independence extraction of the signal in time domain interception and frequency domain filtering, the separated signals of each hop are obtained.

[0062] Step 3: Perform a multi-station coordinated frequency hopping pattern completion operation according to each hopping signal of each receiving station, and finally obtain the frequency hopping pattern detection result of each receiving station.

[0063] In the present invention, the multi-station coordinated frequency hopping pattern completion operation includes missed detection completion or wrong detection correction, and both can be used. In order to perform missed detection completion and wrong detection correction of multi-station detection signals, the present invention compares the similarity of time domain information and frequency domain information and calculates the signal correlation for each hopping signal received by each receiving station, so as to find out the same signal received by multiple receiving stations, the signal missed by a certain receiving station but detected by other receiving stations, and the wrong detection signal that the receiving station detects a signal with similar time or frequency as a hopping signal, thereby compensating and correcting the signal detection result and time-frequency domain parameter estimation result of each receiving station, laying a foundation for the subsequent multi-station parameter estimation of the frequency hopping signal and the sorting of the frequency hopping signal.

[0064] For the calculation of signal correlation, a cross-correlation algorithm model is needed. Cross-correlation is a measure used to represent the similarity between two signals. The ultimate goal is to calculate the similarity between two signals. The basic idea of ​​the basic cross-correlation algorithm is to perform cross-correlation operations on two signals, obtain the position of the function peak, and solve the time difference.

[0065] Specifically, for the same source transmitting signal s(t), after being transmitted through two different channels, the attenuation and mixed noise generated are different. Assuming that the signals received by the two receiving stations are x(t) and y(t), the expressions are as follows:

[0066]

[0067] Where η 1 , η 2 They represent the amplitude gain of the source transmission signal s(t) after passing through the two transmission channels, n 1 (t) represents the interference noise of the first transmission path, n 2 (t) represents the interference noise of the second transmission path, τ 1 and τ 2 They respectively represent the time difference between the two signals arriving at the receiving station.

[0068] For the convenience of expression, normalization is performed using x(t) as the standard, and the above formula can be expressed as:

[0069]

[0070] In the formula, A represents the ratio of the amplitudes of the two received signals, A = η 2 η 1 , d represents the time difference between the two signals, d = τ 2 -τ 1 . From this we can find the cross-correlation function of x(t) and y(t):

[0071]

[0072] In the formula, τ represents the time difference between x(t) and y(t) (i.e., the time difference estimated at the receiving end after being affected by noise, etc.), R ss ( ) represents the autocorrelation function of the source transmitted signal s(t), represents the interference noise n of the first transmission line 1 The cross-correlation function of (t) and the source transmitted signal s(t), represents the interference noise n of the second transmission line 2 The cross-correlation function of (t) and the source transmitted signal s(t), Indicates n 1 (t) and n 2(t) is the cross-correlation function of

[0073] Assume that n 1 (t) and n 2 (t) is a stationary and uncorrelated Gaussian white noise independent of s(t), then and are all 0. Therefore, the above formula can be expressed as:

[0074] R xy (τ) = AR ss (τ-d)

[0075] From the properties of the correlation function, we can get R ss (τ)≤R ss (0), therefore, when τ = d, the cross-correlation function can reach its maximum value, where τ is the time difference between the two signals.

[0076] The missed detection completion of the present invention is based on the similarity of the start time, the similarity of the frequency and the correlation of the signal, to judge the missed detection signal, and to complete the frequency information and time information according to the signal time difference. The process is referred to Figure 3 ,as follows:

[0077] s11, for each hop signal of each receiving station, the jump time t start , center frequency comparison, and set thresholds to judge the similarity.

[0078] Specifically, taking two receiving stations as an example, for each hopping signal detected by the two receiving stations, the starting time t is extracted according to the detection result given by the broadband frequency hopping signal detection algorithm. start And center frequency parameters. Afterwards, the take-off time of each jump signal detected by receiving station 1 is traversed, and the take-off time is compared with the signal detection result of receiving station 2. At the same time, a threshold is set to judge the similarity of the take-off time of the two signals. The setting of the threshold is proportional to the jump period of the signal. In this embodiment, the threshold is set to 0.01 times the signal jump period. When the signals of the two receiving stations meet the similarity of the take-off time, the center frequencies of the two signals are compared for similarity. Similarly, a threshold is set to judge whether the two signals meet the similarity standard of the center frequency. The setting of the threshold is proportional to the bandwidth of the signal. In this embodiment, the threshold is set to 0.005 times the signal bandwidth. When the take-off time and center frequency of the two signals meet the similarity standard, the correlation calculation can be performed on the two signals separated and extracted by the two receiving stations. When multiple receiving stations are used, the principle and method are the same.

[0079] s12, for the frequency hopping signals that simultaneously meet the similarity of the starting time and the similarity of the center frequency, correlation calculation is performed between the two signals to obtain the peak value of the cross-correlation result of the two signals and the corresponding time difference τ.

[0080] S13, also set the threshold, and judge and screen the size of the cross-correlation peak of the two signals. If the cross-correlation operation is performed on the same hop signal that arrives at the two receiving stations after transmission, the correlation result after the cross-correlation has an obvious peak, and its size is significantly larger than the result of the correlation operation after different hop signals arrive at the receiving station. Therefore, the same hop signal received by the two receiving stations can be screened out by setting the threshold. Figure 4 As shown in FIG. 1 , it is a schematic diagram of cross-correlation operation of the signals detected and separated by two receiving stations. Figure 4 (a) is the result of cross-correlation of the received signals of the same hop signal arriving at two receiving stations after transmission. The horizontal axis value of the peak in the figure is the time difference τ between the two received signals. Figure 4 (b) is the result of cross-correlation of the received signals of different hop signals arriving at the two receiving stations. It can be seen from the figure that there is no obvious peak.

[0081] When two signals simultaneously meet the similarity standards of time domain information and frequency domain information and the signal correlation is greater than a given threshold, the two signals are determined to be the received signals of the same hopping transmission signal that are transmitted to the two receiving stations. At the same time, it is determined that the hopping signal is not missed in the broadband signal detection algorithm of other receiving stations (when two receiving stations are selected, it is set to receiving station 2), and the time difference τ′ when the two receiving stations receive the hopping signal is recorded.

[0082] When a certain hopping signal detected by receiving station A (when two receiving stations are selected, set as receiving station 1) and all signals detected by other receiving stations (when two receiving stations are selected, set as receiving station 2) cannot simultaneously meet the standards of similarity of starting time, similarity of center frequency and signal correlation, it can be judged that (when two receiving stations are selected, set as receiving station 2) the hopping signal is missed, and for the sake of convenience of description, it is recorded as signal A1.

[0083] s14, for the missed detection signal, the receiving station A is used to complete the frequency information and extract the signal of the detection result of the missed detection receiving station, and the correlation between the extracted signal and the signal A1 detected by the receiving station A is calculated to obtain the time difference between the extracted signal and the signal A1. The time difference is used to correct the start time and end time in the detection result of the receiving station A, that is, the start time and end time are added to the original start time and end time in the detection result of the missed detection receiving station, thereby realizing the completion of the time information of the missed detection signals of other receiving stations according to the detection result of the receiving station A.

[0084] s15, traverse all receiving stations to detect all hopping signals, and complete the frequency information and time information of the missed detection signals of other receiving stations, and finally complete the completion of the frequency hopping patterns detected by all receiving stations.

[0085] After completing the detection results of multiple stations, the independence of all signals is extracted for subsequent estimation of multi-station parameters and sorting of frequency hopping signals.

[0086] The error detection correction of the present invention is based on the similarity of the starting time, the similarity of the ending time, the similarity of the starting frequency, the similarity of the ending frequency and the correlation of the signals, to judge the error detection signal, and to use the time and frequency information obtained by the detection of the receiving station where no error detection occurs, to correct the starting time and the ending time through the time difference of the signals arriving at the two stations; or to correct the starting time and the ending time through the time difference of the signals arriving at the two stations.

[0087] refer to Figure 5 , the specific method is as follows:

[0088] s21, for each hop signal of each receiving station, the jump time t start , end time t end Starting frequency f start and the stop frequency f end Compare and set thresholds for similarity judgment.

[0089] Similarly, taking two receiving stations as an example, for each hopping signal detected by the two receiving stations, the starting time t is extracted according to the detection result given by the broadband frequency hopping signal detection algorithm. start , end time t end Starting frequency f start and the stop frequency f end After that, each hop signal detected by receiving station 1 and receiving station 2 is traversed and its starting time t is calculated. start , end time t end Starting frequency f start and the stop frequency f end By comparing these four parameters, thresholds are set for the four parameters of the two signals to judge the similarity of the parameters. The setting of the time parameter threshold is proportional to the signal hopping period, and the setting of the frequency parameter threshold is proportional to the bandwidth of each hopping signal.

[0090] s22, for the frequency hopping signals that simultaneously meet the similarity of the starting time, the similarity of the ending time, the similarity of the starting frequency, and the similarity of the ending frequency, correlation calculation is performed between the two signals to obtain the peak value of the cross-correlation result of the two signals and the corresponding time difference τ.

[0091] Specifically, taking two receiving stations as an example, when the two signals simultaneously meet the standards of similarity of the take-off time, similarity of the end time, similarity of the start frequency, and similarity of the end frequency, the correlation of the two signals extracted by the two receiving stations is calculated. After the correlation of the two signals is calculated by the cross-correlation algorithm, the peak value of the cross-correlation result of the two signals and the corresponding time difference τ can be obtained.

[0092] s23, set the threshold K to determine the size of the cross-correlation peak of the two signals. If the cross-correlation peak is greater than the given threshold K, the two signals are determined to be the same hop transmission signal that is transmitted to the two receiving stations, and the hop signal is determined to have no false detection in the broadband signal detection algorithm of the two receiving stations. The threshold K will also be used to determine the signal similarity in the following cases.

[0093] s24, for two signals that satisfy the similarity of the starting time and the similarity of the ending time, but only satisfy one of the similarity of the starting frequency and the similarity of the ending frequency, the correlation calculation is performed to obtain the cross-correlation peak value and the corresponding time difference τ″. If the cross-correlation peak value is greater than the threshold value K, it is determined that a false detection occurs. The false detection situation is that a certain receiving station mistakenly detects two frequency-hopping signals with similar times and frequencies as the same hopping signal. The time information and frequency information detected by the receiving station where no false detection occurs are used to correct the starting time and the ending time through the arrival time difference τ of the two station signals. That is, the starting time and the ending time detected by the receiving station where no false detection occurs are added to the time difference τ″, so as to realize the information correction and completion of the false detection signal.

[0094] s25, for two signals that satisfy the similarity of the starting frequency and the similarity of the ending frequency, but only satisfy one of the similarity of the starting time and the similarity of the ending time, the correlation calculation is performed to obtain the cross-correlation peak value and the corresponding time difference τ″′. If the cross-correlation peak value is greater than the threshold value K, it is determined that a misdetection occurs. The misdetection situation is that a certain receiving station misdetects two frequency-hopping signals with similar frequencies and close times as the same hopping signal. The frequency information detected by the receiving station where the misdetection occurs is used, and the time information detected by the receiving station where the misdetection does not occur is combined. The starting time and the ending time are corrected by the time difference τ″′, that is, the starting time and the ending time detected by the receiving station where the misdetection does not occur are added to the time difference τ″′, so as to realize the information correction and completion of the misdetected signal.

[0095] After completing the error correction and completion of the detection results of multiple stations, the independence of all signals is extracted for subsequent estimation of multi-station parameters and sorting of frequency hopping signals.

[0096] The effect of the present invention can be further demonstrated by the following simulation experiment:

[0097] (1) Experimental environment and experimental data

[0098] The actual environmental background considered by the present invention is as follows Figure 6 As shown, the four frequency hopping stations in the non-cooperative synchronous orthogonal network transmit frequency hopping signals at the same time. Each hopping signal has the same optional frequency range but will not cause frequency collision, and the starting time is synchronized. Two monitoring stations are deployed by the friendly party to monitor and detect the frequency hopping signals emitted by the non-cooperative transmitting station, including the detection of frequency hopping signals, and after detection, the multi-station coordinated completion of the frequency hopping pattern, multi-station frequency hopping signal parameter estimation, and multi-station frequency hopping signal sorting.

[0099] The frequency hopping signal simulation parameters in this experiment are as follows:

[0100] Table 1 Frequency hopping signal simulation parameters

[0101] Parameter name Parameter Value Sampling frequency Fs 250MHz Frequency hopping signal symbol rate Fd 50KHz <![CDATA[Hopping rate R hop > 100 hops <![CDATA[Frequency hopping bandwidth W FH > 100MHz <![CDATA[Hopping interval W hop > 1MHz Modulation BPSK

[0102] The bit signal-to-noise ratio range of the frequency hopping signal is 5dB, 7dB,...19dB, the noise is Gaussian white noise, the distance between each transmitting station is greater than 10km, the distance between two monitoring receiving stations is greater than 10km, and the distance between each transmitting station and the monitoring receiving station is greater than 20km.

[0103] (2) Simulation analysis of missed detection and completion of multi-station cooperative frequency hopping patterns

[0104] In this experiment, the transmitting end is a synchronous orthogonal network station composed of four frequency hopping transmitting devices. The receiving station performs frequency hopping signal detection on signal test data with different signal-to-noise ratios according to the detection process. When the receiving station misses the detection due to obstruction during transmission or the aliasing of a hopping signal with the interference signal, the effect of single receiving station detection and multiple receiving stations collaborative detection and completion is simulated.

[0105] like Figure 7 The figure shows the detection effect when the traditional single-station frequency hopping signal detection misses detection. It can be seen from the figure that one hop signal is missed, and the other signals are accurately selected and detected. In the case of traditional single-station monitoring, the parameter information of the missed hop signal is lost, and it is difficult to recover from the missed hop.

[0106] like Figure 8 As shown in (a) and (b), it is a detection effect diagram when a missed detection occurs in the distributed multi-station frequency hopping signal detection. It can be seen from the figure that two hopping signals of both receiving stations are missed, and other signals are accurately selected and detected. The effect diagram after multi-station completion of the detection results using the multi-station cooperative frequency hopping pattern missed detection method of the present invention is as follows Figure 8As shown in (c)-(h), it can be seen from the figure that the signals missed by the two receiving stations are filled in, and the detection frame after filling has a high degree of overlap with the real frame of the signal, that is, each hop of the frequency hopping signal can be accurately selected to obtain its time and frequency range parameter information.

[0107] (3) Simulation analysis of error detection and correction of multi-station cooperative frequency hopping patterns

[0108] In this experiment, the transmitting end is a synchronous orthogonal network station composed of four frequency hopping transmitting devices. The receiving station performs frequency hopping signal detection on signal test data with different signal-to-noise ratios according to the detection process. When the receiving station detects two hopping signals with similar frequencies and similar hopping times or similar hopping times and similar frequencies as one hopping signal, resulting in false detection, the single receiving station detection and the multiple receiving stations collaboratively detect and correct the false detection results.

[0109] like Fig. 9 As shown in FIG. 1 , it is a detection effect diagram of the traditional single-station frequency hopping signal detection when a misdetection occurs. From the figure, it can be seen that some signals with the same frequency range and adjacent time range are misdetected as one-hop signals, and at the same time, some signals with the same time range and adjacent frequency are misdetected as one-hop signals. Fig. 9 The middle figure (b) shows an enlarged view of a signal with the same time range and adjacent frequencies being misdetected as a one-hop signal; the figure (c) shows an enlarged view of a signal with the same frequency range and adjacent time range being misdetected as a one-hop signal. In the case of traditional single-station monitoring, these misdetected signals may cause errors in subsequent parameter estimation and signal sorting, and it is difficult to correct the misdetection.

[0110] like Fig.10 As shown in (a) and (b), it is a detection effect diagram when the distributed multi-station frequency hopping signal detection has an error detection situation. It can be seen from the figure that in monitoring station 1, there are signals with the same frequency range and adjacent time range that are mistakenly detected as one-hop signals, and there are also signals with the same time range and adjacent frequencies that are mistakenly detected as one-hop signals; the signals in monitoring station 2 are all accurately selected and detected. The effect diagram after the detection results are supplemented by multiple stations using the multi-station cooperative frequency hopping pattern error detection correction method of the present invention is as follows Fig.10 As shown in (c)-(r), it can be seen from the figure that the misdetected signal in the receiving station 1 is successfully corrected and re-filled, and the detection frame after filling has a high degree of overlap with the real frame of the signal, that is, each hop of the frequency hopping signal can be accurately framed to obtain its time and frequency range parameter information.

[0111] like Fig.11As shown, it is the detection performance curve of the traditional single-station detection algorithm and the distributed multi-station cooperative frequency hopping signal completion method proposed in the present invention under different bit signal-to-noise ratios. From the figure, it can be analyzed that: when the bit signal-to-noise ratio is between 6dB and 12dB, the distributed multi-station cooperative frequency hopping signal missed detection completion method and the wrong detection correction method proposed in the present invention have a gain of about 1dB compared with the traditional single-station detection algorithm; the distributed multi-station cooperative frequency hopping signal completion method has a gain of about 1-2dB compared with the traditional single-station detection algorithm. In addition, when the bit signal-to-noise ratio of the distributed multi-station cooperative frequency hopping signal completion method proposed in the present invention is 7dB, the detection accuracy rate reaches more than 90%, which shows that the present invention has good performance and can greatly improve the accuracy of frequency hopping signal detection.

[0112] In summary, the distributed multi-station cooperative frequency hopping signal completion method of the present invention can obtain a higher detection accuracy rate when the traditional detection algorithm has missed detection or false detection, has good performance, and has a better detection effect under low signal-to-noise ratio. The algorithm has better robustness and can also accurately obtain the time-frequency information of the signal, providing a relatively complete and accurate data basis for the subsequent process of frequency hopping signal reconnaissance.

Claims

1. A method for completing frequency hopping detection results based on distributed multi-station monitoring equipment, characterized in that: The steps include: Step 1, using multiple receiving stations to respectively receive multi-frequency hopping signals, wherein the multiple receiving stations use a distributed architecture; Step 2, detecting the multi-frequency hopping signal received by each receiving station, and separating and extracting each hopping signal; Step 3, according to each hopping signal of each receiving station obtained, a multi-station coordinated frequency hopping pattern completion operation is performed to finally obtain the frequency hopping pattern detection result of each receiving station; the multi-station coordinated frequency hopping pattern completion operation includes missed detection completion and / or wrong detection correction; the missed detection completion is based on the similarity of the starting time, the similarity of the frequency and the correlation of the signal, to judge the missed detection signal, and to complete the frequency information and time information according to the arrival time difference of the signals of the two receiving stations; the wrong detection correction is based on the similarity of the starting time, the similarity of the ending time, the similarity of the starting frequency, the similarity of the ending frequency and the correlation of the signal, to judge the wrong detection signal, and to use the time information and frequency information obtained by the receiving station without wrong detection to correct the starting time and the ending time according to the arrival time difference of the signals of the two receiving stations; The method for missing detection and completion is as follows: s11, for each hop signal of each receiving station, the jump time t start , center frequency comparison, and set thresholds to judge similarity; s12, for the frequency hopping signals that simultaneously satisfy the similarity of the take-off time and the similarity of the center frequency, correlation calculation is performed between the two signals to obtain the peak value of the cross-correlation result of the two signals and the corresponding time difference τ; s13, setting a threshold to determine and screen the size of the cross-correlation peak of the two signals. The correlation result after cross-correlation has an obvious peak, and its size is significantly larger than the result obtained by performing correlation operation after different hopping signals arrive at the receiving station. In this way, the same hopping signal received by the two receiving stations is screened out, and it is determined that the hopping signal is not missed, and the time difference τ′ between the two receiving stations receiving the hopping signal is recorded; when the signal A1 detected by the receiving station A and all the signals detected by other receiving stations cannot simultaneously meet the standards of similarity of the starting time, similarity of the center frequency and signal correlation, it is determined that the other receiving stations have missed the detection of the signal A1; s14, for the missed detection signal, the frequency information of the missed detection receiving station detection result is supplemented and the signal is extracted by using the receiving station A, the extracted signal is correlated with the signal A1 detected by the receiving station A, and the time difference between the extracted signal and the signal A1 is obtained, and the time difference is added to the original start time and end time in the detection result of the missed detection receiving station, that is, the time information of the missed detection signal of other receiving stations is supplemented according to the detection result of the receiving station A; s15, traverse all receiving stations to detect all hopping signals, and complete the frequency information and time information of the missed detection signals of other receiving stations, and finally complete the completion of the frequency hopping patterns detected by all receiving stations; The method of error detection and correction is as follows: s21, for each hop signal of each receiving station, the jump time t start , end time t end Starting frequency f start and the stop frequency f end Compare and set thresholds for similarity judgment; s22, for the frequency hopping signals that simultaneously meet the similarity of the starting time, the similarity of the ending time, the similarity of the starting frequency, and the similarity of the ending frequency, correlation calculation is performed between two of them to obtain the peak value of the cross-correlation result of the two signals and the corresponding time difference τ; s23, setting a threshold K to determine and screen the size of the cross-correlation peak of the two signals. If the cross-correlation peak is greater than the given threshold, the two signals are determined to be received signals of the same hop transmission signal that are transmitted to the two receiving stations, and at the same time, it is determined that the hop signal does not cause false detection in the two receiving stations; s24, for two signals that meet the similarity of the starting time and the similarity of the ending time, but only meet one of the similarity of the starting frequency and the similarity of the ending frequency, calculate the correlation, and obtain the cross-correlation peak value and the corresponding time difference τ″. If the cross-correlation peak value is greater than the threshold value K, it is determined that a false detection occurs. The false detection situation is: a certain receiving station mistakenly detects two frequency hopping signals with similar time and frequency as the same hopping signal; the starting time and the ending time obtained by the receiving station without false detection are added to the time difference τ″ to realize information correction and completion of the false detection signal; s25, for two signals that satisfy the starting frequency similarity and the ending frequency similarity, but only satisfy one of the starting time similarity and the ending time similarity, the correlation calculation is performed to obtain the cross-correlation peak value and the corresponding time difference τ″′. If the cross-correlation peak value is greater than the threshold K, it is determined that a false detection occurs. The false detection situation is: a certain receiving station mistakenly detects two frequency-hopping signals with similar frequencies and close times as the same hopping signal; the starting time and ending time obtained by the receiving station where no false detection occurs are added to the time difference τ″′ to realize information correction and completion of the false detection signal.

2. The method for completing frequency hopping detection results based on distributed multi-station monitoring equipment according to claim 1, characterized in that: In the step 1, the multiple receiving stations adopt a distributed architecture, which means that each receiving station is within the receiving range of the frequency hopping signal.

3. The method for completing frequency hopping detection results based on distributed multi-station monitoring equipment according to claim 1, characterized in that: In step 2, a frequency hopping signal detection algorithm is used to obtain the starting frequency f of each hopping signal received by the receiving station. start , stop frequency f end , take-off time t start and the end time t end ; First use the starting time t of each jump signal start and the end time t end The signal is intercepted in the time domain, and then according to the starting frequency f start With the stop frequency f end The carrier frequency and bandwidth of each hop signal are calculated, and each hop signal is filtered in the frequency domain to separate the independence of each hop signal.

4. The method for completing frequency hopping detection results based on distributed multi-station monitoring equipment according to claim 3 is characterized in that: The frequency domain filtering is to use a bandpass filter to perform frequency domain adaptive filtering on each hop signal in the frequency domain according to the relationship between bandwidth and sampling frequency; the design method of the bandpass filter is as follows: First, according to the starting frequency f of each hopping signal start With the stop frequency f end , calculate the filter bandwidth B required for the hopping signal, B = (f end -f start ) / 2, and then calculate the bandwidth B and the original sampling rate f s The ratio B / f s , generate the corresponding low-pass filter, and then perform spectrum shifting on the low-pass filter. By multiplying the filter by the signal carrier frequency, the filter is shifted from zero frequency to the frequency band where the hopping signal is located, that is, the frequency where the band-pass filter is located, and finally the required band-pass filter is generated.

5. The method for completing frequency hopping detection results based on distributed multi-station monitoring equipment according to claim 1, characterized in that: The correlation calculation is to calculate the similarity between two signals, and the similarity is measured by cross-correlation. For the same source transmission signal s(t), the signals received by two receiving stations are x(t) and y(t), respectively. Then the cross-correlation function of x(t) and y(t) is expressed as: Where A represents the amplitude ratio of the two received signals, A = η2 / η1, η1, η2 represent the amplitude gain of s(t) after passing through the two transmission channels, d represents the time difference between the two signals, d = τ2-τ1, τ1 and τ2 represent the time difference between the two signals arriving at the receiving station, τ represents the time difference between x(t) and y(t), R ss () represents the autocorrelation function of the source transmitted signal s(t), represents the cross-correlation function between the interference noise n1(t) of the first transmission path and the source transmission signal s(t), represents the cross-correlation function between the interference noise n2(t) of the second transmission path and the source transmission signal s(t), represents the cross-correlation function of n1(t) and n2(t).

6. The method for completing frequency hopping detection results based on distributed multi-station monitoring equipment according to claim 5, characterized in that: The interference noises n1(t) and n2(t) are stationary and uncorrelated Gaussian white noises independent of s(t), then and are all 0; then: R xy (τ) = AR ss (τ-d) When τ=d, the cross-correlation function can reach its maximum value, and τ at this time is the time difference between the two signals.

7. The method for completing frequency hopping detection results based on distributed multi-station monitoring equipment according to claim 1, characterized in that: After completing the multi-station coordinated frequency hopping pattern completion operation, all signals are extracted independently for subsequent multi-station parameter estimation and frequency hopping signal sorting.

Images