Method and system for monitoring interference signals in aviation radio signal transmission

Abstract

The present invention relates to the field of radio transmission, specifically to a method and system for monitoring interference signals in aviation radio signal transmission, the method comprising: step 1, continuously performing radio frequency scanning spectrum data acquisition; step 2, using a signal-to-noise separation threshold line to perform signal extraction, extracting the signal of the monitoring frequency band from the acquired spectrum data, the monitoring frequency band including the FM broadcast frequency band and the aviation communication frequency band; step 3, using a signal dwell time monitor to monitor each signal in the aviation communication frequency band to find abnormal signals; step 4, if an abnormal signal in the aviation communication frequency band is found in step 3, then calculating the correlation between the abnormal signal and all other signals in the monitoring frequency band to detect the interference source. The present invention solves the problems of slow processing speed, high investment cost, long interference search time and untimely supervision of traditional aviation radio supervision, and provides a guarantee for the reliable transmission of aviation radio signals.

Description

Technical Field

[0001] The present invention relates to the field of radio signal transmission, and in particular to a method and system for monitoring interference signals in aviation radio signal transmission. Background Art

[0002] The rapid development of radio and related technologies has brought great convenience to people's lives, but at the same time, it has also brought many impacts on the radio services in related frequency bands. In particular, civil aviation frequency band communications are often affected and interfered by abnormal radio, which seriously threatens the flight safety of civil aircraft and passengers.

[0003] The safe use of civil aviation radio dedicated frequencies is an important prerequisite for ensuring civil aviation flight safety. It is directly related to the safety of life and property of civil aviation passengers and the general public, and is related to the overall social stability.

[0004] The radio system of civil aircraft includes communication system and navigation system. The communication system includes: VHF communication, HF communication, selective calling, passenger broadcasting, intercom system and voice recorder. The navigation system includes: instrument landing, beacon, VHF omnidirectional beacon, automatic direction finding, rangefinder, transponder, weather radar, Omega navigation and Doppler navigation.

[0005] The effective range of the VHF communication band is relatively short, only within the visual range, and the effective distance varies with the altitude. At an altitude of 300 meters, the distance is 74 kilometers. It is currently the main communication tool for civil aircraft, used for two-way voice communication between crew members and ground control personnel when the aircraft takes off, lands, or passes through controlled airspace. The take-off and landing period is the busiest time for pilots to deal with problems, and it is also the time when accidents are most likely to occur during flight. Therefore, the high reliability of VHF communication must be guaranteed. Civil aircraft are generally equipped with more than one backup system. However, the VHF communication band is the band most susceptible to interference from FM broadcast frequencies.

[0006] At present, when civil aviation passenger planes encounter radio interference, the civil aviation department files interference complaints with the radio management department. After the monitoring station, frequency station department, and law enforcement inspection department put forward handling opinions, the relevant radio monitoring stations will organize technical forces to handle them. For example, large-scale road tests can be used to eliminate interference, or aerial inspections can be applied for. However, this method has a long inspection cycle and slow processing speed. Radio management personnel use drones or their own experience to find interference, but they do not understand the flight status information of civil aviation passenger planes at the time and the electromagnetic environment at the time. The inspection will cost a huge amount of money, the reference time is long, and the best monitoring time is often missed.

[0007] Therefore, traditional civil aviation radio regulatory methods have problems such as slow processing speed, high investment cost, and long search time. Summary of the invention

[0008] Based on the above problems, the present invention provides a method and system for monitoring interference signals in aviation radio signal transmission. The present invention is used to assist radio regulatory departments and aviation safety departments in supervising the use of aviation radio, solving the problems of slow processing speed, high investment cost, long search time and untimely supervision in traditional radio regulatory processes, and ensuring the reliable transmission of civil aviation frequency band signals.

[0009] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0010] A method for monitoring interference signals in aviation radio signal transmission, comprising:

[0011] Step 1, continuously collecting radio frequency spectrum data;

[0012] Step 2, using the signal-to-noise separation threshold line to perform signal extraction, extracting the signal of the monitoring frequency band from the collected spectrum data, wherein the monitoring frequency band includes the FM broadcast frequency band and the aviation communication frequency band;

[0013] Step 3, using a signal dwell time monitor to monitor each signal in the aviation communication frequency band to find abnormal signals;

[0014] Step 4: If an abnormal signal in the aviation communication frequency band is found in step 3, the correlation between the abnormal signal and all other signals in the monitoring frequency band is calculated to detect the interference source.

[0015] The step 2 comprises:

[0016] Step 21, scanning the collected spectrum data in units of spectrum data collection windows with a time interval of T, and performing smoothing processing on the spectrum data of each collection window to obtain a frame of average spectrum;

[0017] Step 22, calculating the left rising edge height array and the right falling edge height array of the mean spectrum;

[0018] Step 23, calculating a discrimination value based on the left rising edge height array and the right falling edge height array;

[0019] Step 24: Calculate a signal-to-noise separation threshold line based on the discrimination value, and extract the signal of the monitoring frequency band based on the signal-to-noise separation threshold line.

[0020] In the step 21, the smoothing process includes time domain smoothing and frequency domain smoothing;

[0021] The time domain smoothing refers to smoothing n frames of spectrum data in a time window with a time interval of T: for all frequency points in the monitoring frequency band, firstly, the average value of n level values ​​of each frequency point in the n frames of spectrum data is calculated, and then the average value of all frequency points is used as the level value of a frame of time domain mean spectrum at the corresponding frequency point, where n is an integer greater than 1;

[0022] The frequency domain smoothing mentioned above refers to performing a one-dimensional smoothing convolution operation on the above time domain mean spectrum: for the level values ​​S of all three adjacent frequency points in a frame of mean spectrum, j-1 , S j and S j+1 The convolution operation is performed using the following formula:

[0023]

[0024] And perform two consecutive convolution smoothing operations to obtain the mean spectrum, S j is the level value at frequency point j.

[0025] The step 22 comprises:

[0026] Obtaining all peaks and troughs of the mean spectrum, including: scanning the mean spectrum obtained in step 21, when the level value of a certain frequency point is greater than or equal to the level values ​​of the left and right frequency points, the frequency point is a peak, when the level value of a certain frequency point is less than or equal to the level values ​​of the left and right frequency points, the frequency point is a trough, and the level values ​​of adjacent peaks are greater than the level values ​​of the troughs;

[0027] The method of obtaining all left rising edge heights and right falling edge heights in the mean spectrum includes: the left rising edge height is the difference between any wave crest and its adjacent left wave trough, and all left rising edge heights form a left rising edge height array; the right falling edge height is the difference between any wave crest and its adjacent right wave trough, and all right falling edge heights form a right falling edge height array; the left rising edge height array and the right falling edge height array are one-dimensional arrays in which all values ​​are positive.

[0028] The step 23 comprises:

[0029] Sort the left rising edge heights in the left rising edge height array to obtain the sorted left rising edge height array h′ L ; Sort the right falling edge heights in the right falling edge height array to obtain the sorted right falling edge height array h′ R ;

[0030] For h′ L All the left rising edge heights in the equation are fitted with a cubic polynomial to obtain a continuous left rising edge height curve F L ; for h′ RA cubic polynomial is fitted to all the right falling edge heights in the equation to obtain a continuous right falling edge height curve F R ; Where the cubic polynomial F = a×x 3 +b×x 2 +c×x+d; where a, b, c, and d are parameter values ​​obtained by fitting a cubic polynomial;

[0031] Calculate Δ = b 2 -3×a×c; if Δ>0, calculate F L The horizontal coordinate value L corresponding to the minimum value of min , get h′ L The index is L min The left rising edge height corresponding to the rounded integer is taken as the left discriminant value; calculate F R The horizontal coordinate value R corresponding to the minimum value of min , get h′ R The index is R min Round off the integer corresponding to the right falling edge height as the right discriminant value; if Δ≤0, calculate F L The horizontal coordinate value L corresponding to the point with the smallest slope min , get h′ L The index is L min Round off the integer corresponding to the left rising edge height as the left discriminant value; calculate F R The horizontal coordinate value R corresponding to the point with the smallest slope min , get h′ R The index is R min The height of the right falling edge corresponding to the rounded integer is taken as the right discriminant value;

[0032] Calculate the mean of the left and right discriminant values ​​as the discriminant value.

[0033] The step 24 comprises:

[0034] For all waves in the mean spectrum, use the discriminant value h obtained noise Perform spectrum scaling, that is, scale the level values ​​corresponding to all frequency points except the trough of any wave. For the frequency point between the peak and the adjacent left trough of any wave, the scaling formula is:

[0035]

[0036] For any frequency point between a wave crest and the adjacent right trough in a wave, the scaling formula is:

[0037]

[0038] Among them, in the mean spectrum, for any wave, L k1The frequency point f of the left trough k1 The level value, L k2 The frequency point f of the right trough k2 The level value, f ki is a frequency point in the wave, ki∈[k1+1,...,k2-1], L ki is the frequency point f ki The level value, h L is the height of the left rising edge, h R is the height of the right falling edge, k1 is the index corresponding to the left trough frequency point in the mean spectrum array, k2 is the index corresponding to the right trough frequency point in the mean spectrum array, and ki is the index corresponding to a frequency point between the left trough and the right trough in the mean spectrum array;

[0039] After the mean spectrum is scaled, the signal-noise separation threshold line is obtained; after the signal-noise separation threshold line is obtained, each time a frame of spectrum data is collected, the frame spectrum is compared with the signal-noise separation threshold line, and the frequency region in the spectrum data whose level value is higher than the signal-noise separation threshold line is identified as the signal frequency band.

[0040] The step 3 comprises:

[0041] When the step 1 acquires a radio signal in the aviation communication frequency band, a signal residence time listener is set at the corresponding signal position, and the monitoring period of the signal residence time listener will continue until the radio signal disappears; when the signal residence time is lower than the set residence time threshold, the radio signal is determined to be a normal communication signal; when the signal residence time exceeds the signal residence time threshold, the radio signal frequency band is determined to be an abnormal signal.

[0042] The step 4 comprises:

[0043] When an abnormal signal in the aviation communication frequency band is found, the correlation between the abnormal signal and other signals in the monitoring frequency band is calculated using the Pearson correlation coefficient:

[0044]

[0045] Among them, in the t-frame signal at a frequency point of the abnormal signal, X i is the level value of the i-th frame signal, X is the level value of the t-frame signal at a frequency point of the abnormal signal, is the mean of X, Y i is the level value of the i-th frame signal in the t-frame signal at a frequency point of a certain other signal extracted except the abnormal signal, Y is the level value of the t-frame signal at a frequency point of a certain other signal extracted except the abnormal signal, is the mean of Y;

[0046] When calculating the correlation between an abnormal signal and any other signal, the frequency point sequences of the two signal frequency bands are regarded as two arrays. First, the middle values ​​of the two arrays are aligned, and the Pearson correlation coefficients of all intersecting frequency points in the time domain are calculated. Then, the average value of all the obtained correlation coefficient values ​​is calculated, and the average value is used as the correlation coefficient value between the abnormal signal and the signal. Finally, the two signals with the largest correlation coefficient values ​​between the abnormal signal and all other signals are determined as interference sources.

[0047] The method also includes step 5: providing information feedback on the abnormal signal and the corresponding interference source.

[0048] A monitoring system for interference signals in aviation radio signal transmission, comprising:

[0049] A spectrum data acquisition module, used for continuous radio frequency sweep spectrum data acquisition;

[0050] A signal extraction module, used to extract signals using a signal-to-noise separation threshold line, and extract signals of a monitoring frequency band from the collected spectrum data, wherein the monitoring frequency band includes an FM broadcast frequency band and an aviation communication frequency band;

[0051] An abnormal signal monitoring module is used to monitor each signal in the aviation communication frequency band using a signal dwell time monitor to detect abnormal signals;

[0052] The interference signal identification module is used to calculate the correlation between the abnormal signal and all other signals in the monitoring frequency band if an abnormal signal is found in the aviation communication frequency band, so as to detect the interference source.

[0053] Compared with the prior art, the method and system of the present invention have the following beneficial effects:

[0054] The radio signal extraction method described in the present invention has a signal recognition rate of 97% in the very high frequency band of 87MHz to 137MHz, and in practical applications, the algorithm complexity is low, which can greatly save computing resource overhead. In addition, the present invention is not only applicable to steady-state signals, such as FM broadcast signals, but also to non-steady-state signals, such as aviation communication signals.

[0055] The method for monitoring interference signals in aviation radio signal transmission described in the present invention can realize the automatic integration and real-time completion of radio data acquisition, signal extraction, abnormal monitoring and interference source discovery, and can assist in solving the problems of slow processing speed, high investment cost, long search time and untimely supervision in traditional radio supervision processes. The present invention can obtain real-time data of interference sources in a timely manner, better solve the problem of untimely supervision of traditional aviation radio frequency use, ensure the reliable transmission of radio signals in the aviation frequency band, ensure communication security, and can effectively reduce the overall capital investment in aviation radio supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 Flow chart of the method for monitoring interference signals in aviation radio signal transmission in this embodiment;

[0057] Figure 2 is a schematic diagram of the mean spectrum;

[0058] Figure 3 Flowchart for discriminant value calculation;

[0059] FIG4(a) is a time domain mean spectrum diagram after time domain smoothing from 87 MHz to 108 MHz, and FIG4(b) is a time domain mean spectrum diagram after time domain smoothing from 108 MHz to 137 MHz;

[0060] Figure 5(a) is the average spectrum diagram after smoothing in the frequency domain of 87MHz to 108MHz, and Figure 5(b) is the average spectrum diagram after smoothing in the frequency domain of 108MHz to 137MHz;

[0061] Figure 6 This is a schematic diagram of the result after sorting the rising edge height array;

[0062] Figure 7 This is a schematic diagram of the results after cubic polynomial fitting;

[0063] FIG8(a) is a schematic diagram of a frame of spectrum data from 87 MHz to 108 MHz and a signal-to-noise separation threshold line, and FIG8(b) is a schematic diagram of a frame of spectrum data from 108 MHz to 137 MHz and a signal-to-noise separation threshold line;

[0064] Fig. 9 It is a schematic diagram of abnormal signals and aviation communication signals in the spectrum data waterfall diagram;

[0065] FIG10( a ) and FIG10( b ) are schematic diagrams showing level changes at two different frequency points in 7 consecutive frames of spectrum data;

[0066] FIG. 11( a ) is a spectrum diagram of 5 consecutive frames of spectrum data generated by an interference source signal, and FIG. 11( b ) is a spectrum diagram of 5 consecutive frames of spectrum data of an abnormal signal generated by the interference source signal in a corresponding time period;

[0067] Fig.12 A schematic diagram of the positions of abnormal signals and interference source signals in the spectrum data waterfall diagram;

[0068] Fig.13 It is a block diagram of the composition of the monitoring system of interference signals in aviation radio signal transmission in the embodiment. DETAILED DESCRIPTION

[0069] The present invention will be further described below in conjunction with the accompanying drawings. The embodiments of the present invention include but are not limited to the following examples.

[0070] See also Figure 1 The method for monitoring interference signals in aviation radio signal transmission provided in this embodiment includes the following steps:

[0071] Step 1: Continue to collect radio frequency spectrum data.

[0072] Step 2: Use the signal-to-noise separation threshold line to perform signal extraction, and extract the signal of the monitoring frequency band from the collected spectrum data. The monitoring frequency band includes the FM broadcast frequency band and the aviation communication frequency band.

[0073] Step 3: Use a signal dwell time monitor to monitor each signal in the aviation communication frequency band to detect abnormal signals.

[0074] Step 4: If an abnormal signal in the aviation communication frequency band is found in step 3, the correlation between the abnormal signal and all other signals in the monitoring frequency band is calculated to detect the interference source.

[0075] Step 5: Provide information feedback on abnormal signals and corresponding interference sources.

[0076] In this embodiment, the radio receiving antenna used in step 1 is the HE600 antenna, and the radio digital receiver is the R&SEM100 (Compact Digital Receiver). The digital computer configures the R&S EM100 parameters by sending TCP messages SCPI (Standard Commands for Programmable Instruments). After receiving the configuration instructions, the R&S EM100 sends the acquired radio spectrum data to the digital computer in the form of UDP messages, wherein the TCP message is only sent for the first time and when the configuration parameters are changed, and the UDP message is continuously sent during the device startup. In this embodiment, step 1 continuously collects spectrum data of the VHF 87MHz~137MHz frequency band, the sampling step is 12.5KHz, and the total number of sampling points is N, wherein 87MHz~108MHz is the FM broadcast frequency band, and 108MHz~137MHz is the VHF aviation communication frequency band. The object of abnormal signal monitoring described in this article is the VHF aviation communication frequency band. When an abnormal signal is found, the interference source is searched in the monitoring frequency band (87MHz~137MHz).

[0077] In this embodiment, the step 1 continuously acquires spectrum data and sends it to a digital computer PC via a UDP datagram. After receiving the radio spectrum data, the PC extracts radio signals in the VHF 87MHz-137MHz frequency band.

[0078] Specifically, in step 2, when extracting radio signals in the VHF 87MHz-137MHz frequency band, the signal extraction is based on a spectrum data acquisition window with a time interval of T. After each time interval T, the spectrum signal-to-noise separation threshold line is obtained using the n frames of spectrum data in the previous spectrum data acquisition window. The signal-to-noise separation threshold line will be used for signal extraction by the n frames of spectrum data in the next spectrum data acquisition window with a time interval of T. The above operation is performed cyclically. n is an integer greater than 1. In this example, the time interval of the spectrum data acquisition window is n=10, that is, every 10 frames of spectrum data are used as a spectrum data acquisition window for obtaining the signal-to-noise separation threshold line.

[0079] In this embodiment, each time the spectrum data of a spectrum data acquisition window is obtained, the following operations are performed in sequence: smoothing the spectrum data to obtain a frame of average spectrum, calculating the left rising edge height and the right falling edge height of the average spectrum, calculating the discrimination value, calculating the signal-to-noise separation threshold line, and obtaining the signal frequency band. Among them, the peak, trough, left rising edge and right falling edge height of the average spectrum are as follows: Figure 2 As shown in L is the height of the left rising edge of a wave, h R It is the height of the right falling edge of a wave.

[0080] The smoothing of the spectrum data to obtain the mean spectrum includes two steps: time domain smoothing and frequency domain smoothing. Radio noise is mainly thermal noise and spatial noise in the system. In actual radio monitoring, the main noise source is the thermal noise in the system, and thermal noise is a typical Gaussian white noise. Gaussian white noise is a stationary random noise with a zero mean, that is, from the time dimension analysis, a certain frequency noise and field strength are random within a certain range, and the mean is certain. In addition, during the spectrum data acquisition process, due to the influence of equipment and environment, some frequencies occasionally have acquisition errors, which will appear as burrs on the spectrum diagram, thereby affecting the calculation of the signal-to-noise separation threshold line and the results of signal-to-noise ratio separation. Therefore, before performing spectrum signal extraction, performing a smoothing operation can effectively reduce the impact of noise on the signal extraction effect.

[0081] In this embodiment, the time domain smoothing, for N sampling points (or frequency points) in the spectrum data acquisition bandwidth of 87MHz to 137MHz, each sampling point has a total of n level value data in n frames in the spectrum data acquisition window, firstly, the average value of the n level values ​​of each sampling point is calculated, and then the average value of all sampling points is used as the level value of a frame of time domain mean spectrum at the corresponding frequency point. After calculating the average value for all sampling points, a frame of time domain mean spectrum is obtained.

[0082] In this embodiment, the frequency domain smoothing refers to performing a one-dimensional smoothing convolution operation on the above-mentioned time domain mean spectrum: for the level values ​​S of all three adjacent frequency points in a frame of mean spectrum, j-1 , S j and S j+1 Perform convolution operation, the convolution kernel used here is [2, 1, 2], that is:

[0083]

[0084] S j is the level value at frequency point j, S j-1 is the level value at frequency point (j-1), S j+1 is the level value of the frequency point (j+1). The purpose of this one-dimensional convolution operation is to remove the singular points in the spectrum data caused by noise, so that the spectrum data becomes smoother. Here, it is necessary to perform a second-order one-dimensional convolution on the above time domain mean spectrum, that is, perform two consecutive convolution smoothing operations to obtain the mean spectrum.

[0085] The calculation of the left rising edge height array and the right falling edge height array of the mean spectrum includes the following steps: obtaining all the peaks and troughs of the mean spectrum, and obtaining all the left rising edge heights and the right falling edge heights in the mean spectrum.

[0086] In this embodiment, the operation of obtaining all peaks and troughs of the mean spectrum is: scanning the above-mentioned mean spectrum, when the level value of a certain frequency point is greater than or equal to the level values ​​of the left and right frequency points, the frequency point is a peak, when the level value of a certain frequency point is less than or equal to the level values ​​of the left and right frequency points, the frequency point is a trough, and the level values ​​of adjacent peaks are greater than the level values ​​of the troughs, and after the scanning is completed, all peaks and troughs of the mean spectrum are obtained.

[0087] In this embodiment, the method of obtaining all left rising edge heights and right falling edge heights in the mean spectrum includes: the left rising edge height is the difference between any peak and its adjacent left trough, and all left rising edge heights (data values) constitute a left rising edge height array; the right falling edge height is the difference between any peak and its adjacent right trough, and all right falling edge heights (data values) constitute a right falling edge height array; the left rising edge height array and the right falling edge height array are one-dimensional arrays in which all values ​​are positive.

[0088] In this embodiment, the left discrimination value and the right discrimination value are calculated for the left rising edge height array and the right falling edge height array respectively.

[0089] The discriminant value calculation process is as follows Figure 3As shown, the calculation method of the discriminant value for the left rising edge height array and the right falling edge height array is the same. Here, only the left rising edge height is taken as an example. First, the left rising edge heights in the left rising edge height array are sorted in ascending order to obtain the sorted left rising edge height array h′ L ; Then, for h′ L All the left rising edge heights in the equation are fitted with a cubic polynomial to obtain a continuous left rising edge height curve F L ; Finally, according to F L Compute the left discriminant.

[0090] Specifically, Figure 4(a) is the time domain mean spectrum diagram after time domain smoothing of 87MHz~108MHz, and Figure 4(b) is the time domain mean spectrum diagram after time domain smoothing of 108MHz~137MHz. At this time, there is more Gaussian white noise; Figure 5(a) is the mean spectrum diagram after frequency domain smoothing of 87MHz~108MHz, and Figure 5(b) is the mean spectrum diagram after frequency domain smoothing of 108MHz~137MHz. At this time, there is less Gaussian white noise; after obtaining the mean spectrum, all peaks and troughs in the mean spectrum are obtained, and all left rising edge heights and right falling edge heights in the mean spectrum are obtained; each left rising edge height in the left rising edge height array is sorted in ascending order and presented as follows Figure 6 Trend shown.

[0091] according to Figure 6 As shown in the figure, in the spectrum data, there is a more obvious intensity mutation between the waves with smaller energy intensity and the waves with larger energy intensity. In order to obtain the threshold of signal and noise in the spectrum data, a cubic polynomial fitting method is used, and the results are as follows Figure 7 As shown, a cubic polynomial is fitted to each left rising edge height in the sorted left rising edge height array, and the cubic polynomial F=a×x 3 +b×x 2 +c×x+d, x is a real number, and the value of F during fitting is the height of each left rising edge in the left rising edge height array (data value). The cubic polynomial fitting is to fit the discrete height values ​​in the array through a function to obtain a curve representation with continuous values. If Δ>0, Δ=b 2 -3×a×c, calculate F L The horizontal coordinate value L corresponding to the minimum value of min , get h′ L The index is L min The left rising edge height corresponding to the rounded integer is taken as the left discriminant value, where:

[0092]

[0093] If Δ≤0, calculate F L The horizontal coordinate value L corresponding to the point with the smallest slopemin , get h′ L The index is L min The left rising edge height corresponding to the rounded integer is taken as the left discriminant value, where:

[0094]

[0095] a, b, c, and d are parameter values ​​obtained by cubic polynomial fitting. In the test example:

[0096]

[0097] That is, h′ L The index is L min =81 (rounded to the nearest integer), the corresponding left rising edge height value 6 is the left discriminant value. Similarly, the right discriminant value of the mean spectrum is also 6. Therefore, the average of the left and right discriminant values ​​of the mean spectrum is 6.

[0098] In this embodiment, F is calculated L The horizontal coordinate value L corresponding to the minimum value (or minimum slope point) of min When F is not calculated first L The minimum value (or minimum slope point) of is found at , and then the corresponding horizontal coordinate value is obtained. Instead, it is directly calculated using the above two formulas, which can simplify the calculation process and improve the calculation efficiency.

[0099] The signal-to-noise separation threshold line is obtained by using the above-obtained discrimination value h for all waves in the mean spectrum. noise =6 for spectrum scaling; the spectrum scaling is to scale the level values ​​corresponding to all frequency points of any wave (including the peak, excluding the trough). For the level value of the frequency point between the peak and the adjacent left trough in any wave, the scaling formula is:

[0100]

[0101] For the level value of the frequency point between the peak and the adjacent right trough in any wave, the scaling formula is:

[0102]

[0103] Among them, in the mean spectrum, for any wave, in this wave, L k1 The frequency point f of the left trough k1 The level value, L k2 The frequency point f of the right trough k2 The level value, f ki is a frequency point in the wave, ki∈[k1+1,...,k2-1], L ki is the frequency point fki The level value, h L is the height of the left rising edge, h R is the height of the right falling edge, k1 is the index of the left trough frequency point in the mean spectrum array, k2 is the index of the right trough frequency point in the mean spectrum array, and ki is the index of a frequency point between the left trough and the right trough in the mean spectrum array. The meaning of this scaling formula is: in a wave, based on the connection line between the two troughs, for any frequency point in the wave, the level value of the frequency point is proportionally scaled according to the proportion of the corresponding left rising edge height or right falling edge height.

[0104] After the mean spectrum is scaled, the signal-noise separation threshold line is obtained; after the signal-noise separation threshold line is obtained, each time a frame of spectrum data is collected, the frame spectrum is compared with the signal-noise separation threshold line, and the frequency region in the spectrum data whose level value is higher than the signal-noise separation threshold line is identified as the signal frequency band. As shown in Figure 8(a) and Figure 8(b), the dotted line is the obtained signal-noise separation threshold line, and the solid line is a frame of spectrum data, where Figure 8(a) is a schematic diagram of a frame of spectrum data from 87MHz to 108MHz and the signal-noise separation threshold line, and Figure 8(b) is a schematic diagram of a frame of spectrum data from 108MHz to 137MHz and the signal-noise separation threshold line; the region where the level value of a frame of spectrum data is 5dB higher than the signal-noise separation threshold line will be identified as the signal frequency band.

[0105] In this embodiment, the experimental data collection location is a place in Chengdu, Sichuan Province. Therefore, in the collected radio spectrum data, the VHF aviation communication frequency band contains more abnormal signals, which also provides a good simulation experiment environment. After long-term monitoring and a large number of experiments, it is found that the signal recognition rate of the radio signal extraction method used in this embodiment is higher than 97%, and the false alarm rate is low.

[0106] In this embodiment, according to the above method, the spectrum data of the very high frequency band 87MHz~137MHz is continuously monitored and signals are extracted. For a frame of spectrum data, when a radio signal in the aviation communication frequency band 108MHz~137MHz is obtained, a signal residence time listener is placed at the corresponding signal frequency band position. The monitoring period of the signal residence time listener will continue until the signal disappears.

[0107] The signal residence time listener is a computer program implementation method. Specifically, in this embodiment, the multi-threaded concurrent processing technology in computer program design is adopted. For example, at a certain moment, multiple radio signals are found in the aviation communication frequency band. Then, multiple signal residence time listeners will be placed at the corresponding positions to perform monitoring operations. From a macro perspective, these listeners are independent of each other and executed in parallel. When the signal residence time exceeds the signal residence time threshold, the signal frequency band is determined to be an abnormal signal. The signal residence time threshold is obtained by artificial long-term observation of the aviation radio spectrum characteristics. In this embodiment, the signal residence time threshold is 5 seconds.

[0108] In this embodiment, the EM100 monitors the VHF aviation communication frequency band 108MHz to 137MHz for a period of time through the visualization of voice and spectrum data. Fig. 9 As shown in the figure, the mark ① is an abnormal signal, and the mark ② is an aviation communication signal. The voice of the tower control personnel communicating with the pilot of the aircraft presents intermittent and continuous characteristics, and the voice duration of each communication is usually 3 to 5 seconds. However, the waterfall chart shows many Fig. 9 The continuous signals marked with ① are found through voice demodulation monitoring to be intermodulation signals or harmonics from the broadcast frequency band, that is, black broadcasting.

[0109] In this embodiment, when an abnormal signal is found in an aviation frequency band, the interference source of the abnormal signal is sought in the monitoring frequency band. The method for judging the interference source is to determine the degree of correlation between the abnormal signal and other signals. Here, other signals refer to other signals in the monitoring frequency band except the abnormal signal, and the two signals with the highest correlation are regarded as interference sources.

[0110] In the radio spectrum, intermodulation interference refers to the situation when two or more frequency signals are simultaneously input into the transmitter and receiver, a third frequency component is generated due to the nonlinearity of the circuit. When the frequency component happens to fall into the working frequency band of a radio station, the station will be interfered. The characteristic of this interference is that the interference source and the interfered signal are synchronized. However, the level values ​​of the interference source and the interfered signal are usually in a certain proportional relationship, that is, their electromagnetic intensities are not equal. Therefore, this method uses the Pearson correlation coefficient to obtain the degree of correlation based on this characteristic:

[0111]

[0112] Among them, in the t-frame signal at a frequency point of the abnormal signal, X i is the level value of the i-th frame signal, X is the level value of the t-frame signal (i.e. all signals in frames 1 to t) at a frequency point of the abnormal signal (i.e. X is an array), is the mean of X, Y iis the level value of the i-th frame signal in the t-frame signal at a frequency point of a certain other signal extracted except the abnormal signal, Y is the level value of the t-frame signal at a frequency point of a certain other signal extracted except the abnormal signal (that is, Y is an array), is the mean value of Y. In this method, it is necessary to use any abnormal signal to calculate the correlation with all other abnormal signals except the abnormal signal obtained in step 1. When calculating the correlation between the abnormal signal and any other signal, the frequency point sequences of the two signal frequency bands are regarded as two arrays. First, the middle values ​​of the two arrays are aligned, and the Pearson correlation coefficients of all intersecting frequency points in the time domain are calculated respectively; then, the average value of all the obtained correlation coefficient values ​​is calculated as the correlation coefficient value between the abnormal signal and the signal, and finally, the two signals with the largest correlation coefficient values ​​between the abnormal signal and all other signals are determined as interference sources.

[0113] When the Pearson correlation coefficient is used to calculate the correlation between two frequency points in the time domain, Figure 10(a) and Figure 10(b) are schematic diagrams of the level value changes of two different frequency points in 7 consecutive frames of spectrum data. Although there is a large difference in the level value range of the two frequency points, it can be observed from the images that the two frequency points show a trend of synchronous change within the time range of these 7 moments, which just meets the properties of intermodulation and harmonics in radio signals.

[0114] When the Pearson correlation coefficient is used to measure the correlation between two signals in the time domain, as shown in Figures 11(a) and 11(b), Figure 11(a) is a spectrum diagram of 5 consecutive frames of spectrum data (i.e., T1, T2, T3, T4, and T5 moments) generated by an interference source signal, and Figure 11(b) is a spectrum diagram of 5 consecutive frames of spectrum data (i.e., T1, T2, T3, T4, and T5 moments) of an abnormal signal generated by the above interference source signal in the corresponding time period. In order to facilitate the observation of the change trend, the 5 consecutive frames of spectrum data are placed in the same figure here. Analysis of the signal change trends of all frequency points in the two signals shows that although the level values ​​of each frequency point are different, they all show a trend of synchronous change, that is, the correlation between the two signals is extremely high. In the two signal channels described here, the interference source is the interference of the VHF FM broadcast band 92.50MHz to the VHF aviation communication band 111.00MHz. In radio signal monitoring, the spectrum data waterfall chart can intuitively reflect the duration of the signal, such as Fig.12As shown in the figure, abnormal signals and interference source signals continue to appear during the monitoring period. After obtaining the frequency band through this method, it is found that the VHF FM broadcast band 92.50MHz is a black broadcast "releasing loan information" based on the voice demodulation monitoring of the signals of these two channels for a long time, and the VHF aviation communication band 111.00MHz monitors the same audio information at the same time. This also confirms the correctness of this method.

[0115] In this embodiment, when an interference source is detected, feedback can be provided to relevant personnel in a timely manner. When an interference source of an abnormal signal is obtained in step 4, the interfered aviation communication signal and the corresponding interference source signal are collected at the same time, and AM and FM audio demodulation is performed. Since the signal modulation method of the broadcast frequency band is FM and the modulation method of the aviation communication frequency band is AM, two demodulation methods need to be executed to obtain the corresponding audio information. The acquired audio information and spectrum data information need to be saved, and their corresponding time tags and frequency tags are saved. Among them, the voice demodulation uses intermediate frequency measurement to perform audio data analysis on abnormal signals and interference source signals, and the audio data is converted into text and keywords are extracted using voice recognition. This information will be fed back to the radio control personnel for playback confirmation to assist the supervisors in completing their supervision work.

[0116] like Fig.13 As shown, based on the same inventive concept, this embodiment also provides a monitoring system for interference signals in aviation radio signal transmission, including:

[0117] A spectrum data acquisition module, used for continuous radio frequency sweep spectrum data acquisition;

[0118] A signal extraction module, used to extract signals using a signal-to-noise separation threshold line, and extract signals of a monitoring frequency band from the collected spectrum data, wherein the monitoring frequency band includes an FM broadcast frequency band and an aviation communication frequency band;

[0119] An abnormal signal monitoring module is used to monitor each signal in the aviation communication frequency band using a signal dwell time monitor to detect abnormal signals;

[0120] The interference signal identification module is used to calculate the correlation between the abnormal signal and all other signals in the monitoring frequency band if an abnormal signal is found in the aviation communication frequency band, so as to detect the interference source.

[0121] The above system may also include an information feedback module, which is used to provide information feedback on abnormal signals and corresponding interference sources to relevant staff.

[0122] Since they are based on the same inventive concept, the technical means adopted by this system and the aforementioned method are the same. For the specific processing operations of each module, please refer to the aforementioned content and will not be repeated here.

Claims

1. A method for monitoring interference signals in aviation radio signal transmission, characterized in that: include: Step 1, continuously collecting radio frequency spectrum data; Step 2, using the signal-to-noise separation threshold line to perform signal extraction, extracting the signal of the monitoring frequency band from the collected spectrum data, wherein the monitoring frequency band includes the FM broadcast frequency band and the aviation communication frequency band; The step 2 comprises: Step 21, scanning the collected spectrum data in units of spectrum data collection windows with a time interval of T, and performing smoothing processing on the spectrum data of each collection window to obtain a frame of average spectrum; Step 22, calculating the left rising edge height array and the right falling edge height array of the mean spectrum; Step 23, calculating a discrimination value based on the left rising edge height array and the right falling edge height array; Step 24, calculating a signal-to-noise separation threshold line based on the discrimination value, and extracting a signal of the monitoring frequency band based on the signal-to-noise separation threshold line; The step 23 comprises: Sort the left rising edge heights in the left rising edge height array to obtain the sorted left rising edge height array h′ L ; Sort the right falling edge heights in the right falling edge height array to obtain the sorted right falling edge height array h′ R ; For h′ L All the left rising edge heights in the equation are fitted with a cubic polynomial to obtain a continuous left rising edge height curve F L ; for h′ R A cubic polynomial is fitted to all the right falling edge heights in the equation to obtain a continuous right falling edge height curve F R ; Where the cubic polynomial F = a×x 3 +b×x 2 +c×x+d; where a, b, c, and d are parameter values ​​obtained by fitting a cubic polynomial; Calculate Δ = b 2 -3×a×c; if Δ>0, calculate F L The horizontal coordinate value L corresponding to the minimum value of min , get h′ L The index is L min The left rising edge height corresponding to the rounded integer is taken as the left discriminant value; calculate F R The horizontal coordinate value R corresponding to the minimum value of min , get h′ R The index is R min Round off the integer corresponding to the right falling edge height as the right discriminant value; if Δ≤0, calculate F L The horizontal coordinate value L corresponding to the point with the smallest slope min , get h′ L The index is L min Round off the integer corresponding to the left rising edge height as the left discriminant value; calculate F R The horizontal coordinate value R corresponding to the point with the smallest slope min , get h′ R The index is R min The height of the right falling edge corresponding to the rounded integer is taken as the right discriminant value; Calculate the mean of the left discriminant value and the right discriminant value as the discriminant value; The step 24 comprises: For all waves in the mean spectrum, the discriminant value h is obtained using noise Perform spectrum scaling, that is, scale the level values ​​corresponding to all frequency points except the trough of any wave. For the level value of the frequency point between the peak and the adjacent left trough of any wave, the scaling formula is: For the level value of the frequency point between the peak and the adjacent right trough in any wave, the scaling formula is: Among them, in the mean spectrum, for any wave, L k1 The frequency point f of the left trough k1 The level value, L k2 The frequency point f of the right trough k2 The level value, f ki is a frequency point in the wave, ki∈[k1+1,...,k2-1], L ki is the frequency point f ki The level value, h L is the height of the left rising edge, h R is the height of the right falling edge, k1 is the index corresponding to the left trough frequency point in the mean spectrum array, k2 is the index corresponding to the right trough frequency point in the mean spectrum array, and ki is the index corresponding to a frequency point between the left trough and the right trough in the mean spectrum array; After the mean spectrum is scaled, the signal-noise separation threshold line is obtained; after the signal-noise separation threshold line is obtained, each time a frame of spectrum data is collected, the frame spectrum is compared with the signal-noise separation threshold line, and the frequency region in the spectrum data whose level value is higher than the signal-noise separation threshold line is identified as the signal frequency band; Step 3, using a signal dwell time monitor to monitor each signal in the aviation communication frequency band to find abnormal signals; Step 4: If an abnormal signal in the aviation communication frequency band is found in step 3, the correlation between the abnormal signal and all other signals in the monitoring frequency band is calculated to detect the interference source.

2. The method for monitoring interference signals in aviation radio signal transmission according to claim 1, characterized in that: In the step 21, the smoothing process includes time domain smoothing and frequency domain smoothing; The time domain smoothing refers to smoothing n frames of spectrum data in a time window with a time interval of T: for all frequency points in the monitoring frequency band, firstly, the average value of n level values ​​of each frequency point in the n frames of spectrum data is calculated, and then the average value of all frequency points is used as the level value of a frame of time domain mean spectrum at the corresponding frequency point, where n is an integer greater than 1; The frequency domain smoothing mentioned above refers to performing a one-dimensional smoothing convolution operation on the above time domain mean spectrum: for the level values ​​S of all three adjacent frequency points in a frame of mean spectrum, j-1 , S j and S j+1 The convolution operation is performed using the following formula: And perform two consecutive convolution smoothing operations to obtain the mean spectrum, S j is the level value at frequency point j.

3. The method for monitoring interference signals in aviation radio signal transmission according to claim 1, characterized in that: The step 22 comprises: Obtaining all peaks and troughs of the mean spectrum, including: scanning the mean spectrum obtained in step 21, when the level value of a certain frequency point is greater than or equal to the level values ​​of the left and right frequency points, the frequency point is a peak, when the level value of a certain frequency point is less than or equal to the level values ​​of the left and right frequency points, the frequency point is a trough, and the level values ​​of adjacent peaks are greater than the level values ​​of the troughs; The method of obtaining all left rising edge heights and right falling edge heights in the mean spectrum includes: the left rising edge height is the difference between any wave crest and its adjacent left wave trough, and all left rising edge heights form a left rising edge height array; the right falling edge height is the difference between any wave crest and its adjacent right wave trough, and all right falling edge heights form a right falling edge height array; the left rising edge height array and the right falling edge height array are one-dimensional arrays in which all values ​​are positive.

4. The method for monitoring interference signals in aviation radio signal transmission according to claim 1, characterized in that: The step 3 comprises: When the step 1 acquires a radio signal in the aviation communication frequency band, a signal residence time listener is set at the corresponding signal position, and the monitoring period of the signal residence time listener will continue until the radio signal disappears; when the signal residence time is lower than the set residence time threshold, the radio signal is determined to be a normal communication signal; when the signal residence time exceeds the signal residence time threshold, the radio signal frequency band is determined to be an abnormal signal.

5. The method for monitoring interference signals in aviation radio signal transmission according to claim 1, characterized in that: The step 4 comprises: When an abnormal signal in the aviation communication frequency band is found, the correlation between the abnormal signal and other signals in the monitoring frequency band is calculated using the Pearson correlation coefficient: Among them, in the t-frame signal at a frequency point of the abnormal signal, X i is the level value of the i-th frame signal, X is the level value of the t-frame signal at a frequency point of the abnormal signal, is the mean of X, Y i is the level value of the i-th frame signal in the t-frame signal at a frequency point of a certain other signal extracted except the abnormal signal, Y is the level value of the t-frame signal at a frequency point of a certain other signal extracted except the abnormal signal, is the mean of Y; When calculating the correlation between an abnormal signal and any other signal, the frequency point sequences of the two signal frequency bands are regarded as two arrays. First, the middle values ​​of the two arrays are aligned, and the Pearson correlation coefficients of all intersecting frequency points in the time domain are calculated. Then, the average value of all the obtained correlation coefficient values ​​is calculated, and the average value is used as the correlation coefficient value between the abnormal signal and the signal. Finally, the two signals with the largest correlation coefficient values ​​between the abnormal signal and all other signals are determined as interference sources.

6. The method for monitoring interference signals in aviation radio signal transmission according to claim 1, characterized in that: The method also includes step 5: providing information feedback on the abnormal signal and the corresponding interference source.

7. A monitoring system for interference signals in aviation radio signal transmission, used to implement the method described in any one of claims 1 to 6, characterized in that: include: A spectrum data acquisition module, used for continuous radio frequency sweep spectrum data acquisition; A signal extraction module, used to extract signals using a signal-to-noise separation threshold line, and extract signals of a monitoring frequency band from the collected spectrum data, wherein the monitoring frequency band includes an FM broadcast frequency band and an aviation communication frequency band; An abnormal signal monitoring module is used to monitor each signal in the aviation communication frequency band using a signal dwell time monitor to detect abnormal signals; The interference signal identification module is used to calculate the correlation between the abnormal signal and all other signals in the monitoring frequency band if an abnormal signal is found in the aviation communication frequency band, so as to detect the interference source.

Images