Direction of arrival estimation method under non-ideal array element condition

Abstract

The invention relates to a direction of arrival estimation method under a non-ideal array element condition, and the method comprises the steps: arranging and constructing an asymmetric cross array based on a dipole antenna, and enabling the asymmetric cross array to serve as an array element; forming an array by a plurality of array elements, and performing frequency domain conventional beam forming processing on each frame of received signal in the time window to obtain beam output of each frame in different incoming wave directions; recording the maximum beam output direction of each frame, namely DOA estimation of each frame; counting DOA estimation of all frames to obtain a direction with the highest statistical frequency; and weighting the two received signals of each array element to obtain the received signals of the whole array. Compared with the prior art, the invention provides a brand new non-ideal array element receiving array, the DOA detection efficiency and precision are improved through a weighting method on the basis, and the target direction finding precision and reliability under the asymmetric cross array condition can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electromagnetic signal target direction finding, and particularly relates to a wave direction estimation method under non-ideal array element conditions. BACKGROUND

[0002] In the field of electromagnetic signal target direction finding, Direction of Arrival (DOA) estimation is to use a receiving array to perform beamforming on the received original signal, and then to obtain the angular position information of the target signal source through a matching filter algorithm processing method. Traditional DOA estimation algorithms are mostly based on an ideal array model, that is, each array element in the array is assumed to be a perfect electromagnetic point, without error and completely independent. Considering the ideal array element, each array element is regarded as an electromagnetic point, and the received signal of a single array element can be represented as: where τ m is the signal delay of the mth array element relative to the first array element. The received signal of the entire array is y(t) = a(θ) * s1(t). Where s1(t) = s(t)e jwt , and a(θ) is the direction vector of the array. After the received signal is subjected to beamforming, the power spectrum is: For any given angle, the direction vector a(θ) can be obtained, and by traversing all angles in space, the direction with the maximum power is obtained, which is the DOA direction of the signal.

[0003] However, in actual applications, the array elements are often affected by various non-ideal factors, such as mutual coupling error, amplitude and phase error, position error, etc. These factors can significantly reduce the accuracy and reliability of DOA estimation. In addition, different physical arrangement methods of array elements can also affect the performance of the direction finding algorithm, especially when the array elements are designed to be more complex structures. When the array elements cannot be idealized and abstracted as ideal electromagnetic points, due to the non-point source characteristics, array element coupling and polarization diversity caused by the complex structure of the array elements, the actual receiving conditions of the array elements under different receiving electromagnetic element non-ideal combination conditions need to be considered, and the DOA weighting method under the new non-ideal array element method needs to be researched.

[0004] At present, some progress has been made in the research of DOA estimation algorithm under non-ideal array elements, but most of the researches are still concentrated on uniform arrays or simple non-uniform arrays, and the research on the special structure of the non-symmetrical cross array is relatively less. Therefore, it is of great significance to develop a DOA estimation method suitable for non-symmetrical cross array for improving the accuracy and reliability of target direction finding. SUMMARY

[0005] The present application aims to overcome the defects of the prior art and provide a direction of arrival estimation method under non-ideal array element conditions, which can improve the target direction finding performance under non-symmetric cross array conditions.

[0006] The object of the present application can be achieved by the following technical solutions: a direction of arrival estimation method under non-ideal array element conditions, comprising the following steps:

[0007] S1, based on a dipole antenna, arranging and constructing a non-symmetric cross array as an array element;

[0008] S2, forming an array with multiple array elements, performing frequency domain conventional beamforming processing on each frame of received signals in a time window to obtain beam outputs of each frame in different directions of arrival;

[0009] S3, recording the maximum direction of each frame of beam outputs, i.e. the DOA estimation of each frame;

[0010] S4, statistically analyzing the DOA estimations of all frames to obtain the direction with the highest statistical frequency;

[0011] S5, performing weighted processing on the two received signals of each array element to obtain the received signal of the entire array.

[0012] Further, the non-symmetric cross array in step S1 includes two dipole antennas, which are placed vertically intersecting each other, and the intersection point is the feed point.

[0013] Further, the lengths of the conductors of the dipole antenna on both sides of the feed point are not equal, and the dipole antenna is in a non-symmetric form.

[0014] Further, the process of step S2 includes:

[0015] Forming an array with multiple array elements, and the received signal of each array element is the addition result of the received signals of the two dipole antennas;

[0016] Frame processing is performed on continuous signals in a certain time window;

[0017] Beamforming processing is performed on each frame of signals, the beamforming adjusts the phase and amplitude of each array element, so that the signal in a specific direction is enhanced, and the signal in other directions is suppressed, and the beam outputs of each frame in different directions of arrival P n (θ j ) are obtained, which represents the beam output of the nth frame in the direction of arrival θ j .

[0018] Further, the received original signal of the array is:

[0019]

[0020] y i = (y i,1 +y i,2 )

[0021] wherein y i is the received signal of the i-th array element in the array, M is the total number of array elements in the array, y i,1 , y i,2 correspond to the received signals of the two dipole antennas in the i-th array element, respectively.

[0022] Further, the step S3 is specifically to select the maximum beam output from the beam outputs corresponding to different directions of arrival of each frame, and record the direction corresponding to the maximum beam output, i.e. the DOA estimation of the signal of the frame.

[0023] Further, the step S4 is specifically to count the frequency of each direction θ j in all frames, and find the direction θ * with the highest frequency as the final DOA estimation value in the time window.

[0024] Further, the step S5 includes the following steps:

[0025] S51, for each array element, compare the direction with the highest statistical frequency with the included angle of the cross array pointing in the array element, to determine the weight of the two dipole antenna received signal components in the array element, for weighting processing of the two dipole antenna received signals in the array element;

[0026] S52, based on the weighted received signals of each array element, obtain the received signal of the entire array.

[0027] Further, the weight of the two dipole antenna received signal components in the array element in the step S51 is:

[0028]

[0029] wherein g m,1 , g m,2 correspond to the weight of the two dipole antenna received signal components in the m-th array element, respectively, k0 is a constant, θ is the direction with the highest statistical frequency, and θ is the included angle of the cross array pointing in the array element.

[0030] Further, the received signal of the entire array in the step S52 is:

[0031]

[0032] wherein y is the weighted received signal of the M-th array element, and gM,α is the weight of the signal component received by the alpha-th dipole antenna in the M-th array element, y M,α is the signal received by the alpha-th dipole antenna in the M-th array element.

[0033] Compared with the prior art, the present application has the following advantages:

[0034] The present application proposes a new array element arrangement mode with a cross array, and proposes a wave direction estimation method under the condition of the cross array, wherein the physical placement characteristics of the dipole antennas in the asymmetric cross array determine different receiving abilities of the array elements to signals in different directions, after a plurality of array elements are combined to form an array, first, the frequency domain conventional beamforming processing is performed on each frame of received signals in a time window to obtain beam outputs of each frame in different wave directions; then, the maximum direction of each frame of beam outputs, i.e., the DOA estimation of each frame, is recorded; the DOA estimations of all frames are counted to obtain the direction with the highest statistical frequency; finally, the two received signals of each array element are weighted, and the target direction finding performance under the condition of the cross array is improved through the weighting.

[0035] The present application designs an array element containing two dipole antennas, the two dipole antennas are placed perpendicularly and cross each other, the intersection point is the feed point, the conductor lengths of the dipole antennas on both sides of the feed point are different, and the dipole antennas are in an asymmetric form. Thus, an array element composed of asymmetric dipole antennas is realized, the array element itself has a certain directivity from the physical layer, the array composed of the array elements can receive wave signals in all directions, and the signal reception in a specific direction can be enhanced from the physical characteristics of the array elements. The array elements placed in the cross array have the physical characteristics of long axis and short axis, so that the direction near the long axis has a higher spatial resolution, and the direction change of the incoming wave is more sensitive. In the case of knowing the approximate wave direction of the signal in advance, the pre-placed direction of the cross array can have a higher direction finding efficiency than the conventional ideal array element.

[0036] The application firstly carries out frame processing on continuous signals in a certain time window, each frame signal is processed by beamforming, the beamforming is to adjust the phase and amplitude of each array element, so that the signal in a certain direction is enhanced, and the signal in other directions is suppressed; after completing the conventional beamforming of all frame signals, frequency statistics is carried out, the purpose is to extract the most possible signal arrival direction from the DOA estimation results of all frames, this frequency statistics method can effectively reduce the misjudgment caused by single frame signal noise or interference, and improve the robustness of DOA estimation; finally, the DOA direction obtained by frequency statistics is compared with the angle between the cross array pointing of the array element, if the angle between the preliminary estimation direction of DOA and the short axis is small, it indicates that the gain contribution of the long axis corresponding dipole antenna to the signal is large, at this time, the long axis corresponding receiving signal component should be given a larger weight; on the contrary, if the angle with the long axis is small, the short axis corresponding receiving signal component should be given a larger weight. By adjusting the weight of the two receiving signal components, the DOA estimation value can be effectively corrected, and the estimation accuracy can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 It is a method flowchart of the application;

[0038] Figure 2 It is a half-wave dipole array element schematic diagram based on the application;

[0039] Figure 3 It is a half-wave dipole direction Figure 3 schematic diagram;

[0040] Figure 4 It is a symmetric isometric long cross array element arrangement mode and directional diagram schematic diagram;

[0041] Figure 5 It is a symmetric non-isometric long cross array element arrangement mode and directional diagram schematic diagram;

[0042] Figure 6 It is a non-symmetric dipole antenna schematic diagram;

[0043] Figure 7 It is a non-symmetric cross array element arrangement mode and directional diagram schematic diagram;

[0044] Figure 8 It is a directional diagram change trend schematic diagram of the non-symmetric cross array element;

[0045] Figure 9 It is a receiving array schematic diagram based on the non-symmetric cross array element in the embodiment. DETAILED DESCRIPTION

[0046] The application will be described in detail below in combination with the drawings and specific embodiments.

[0047] Embodiment

[0048] As shown in Figure 1 , a wave direction estimation method under non-ideal array element conditions comprises the following steps:

[0049] S1, based on a dipole antenna, an asymmetric cross array is arranged and constructed as an array element;

[0050] S2, a plurality of array elements are combined into an array, and the frequency domain conventional beamforming processing is performed on each frame of received signals in a time window to obtain the beam output of each frame in different wave directions;

[0051] S3, the maximum direction of each frame of beam output, i.e. the DOA estimation of each frame, is recorded;

[0052] S4, the DOA estimations of all frames are counted to obtain the direction with the highest statistical frequency;

[0053] S5, the two received signals of each array element are weighted and processed to obtain the received signal of the entire array.

[0054] In step S1, the non-ideal array element proposed in the scheme is realized based on a dipole antenna. The Dipole half-wave dipole antenna is composed of two conductors with equal length, small diameter and opposite phase, which are connected through a feed point. The two end points in the middle are excited by equal amplitude and opposite phase voltage, and the distance between the middle end points is much smaller than the working wavelength and can be ignored. When the Dipole antenna is used as a receiving antenna, it captures electromagnetic waves from space. Electromagnetic waves induce currents in the antenna conductor, which are then transmitted to the receiver for processing, thereby restoring the original signal. Figure 2 is a schematic diagram of the basic unit half-wave dipole used by the receiving array element in the scheme. When the overall length of the dipole is equal to half the wavelength, the signal receiving effect is optimal. In engineering, the current in the dipole conductor is approximated as a sinusoidal distribution. The current distribution of the half-wave dipole can be expressed as: I(z) = I0sink(l-|z|). Where I0 is the wave crest current, is the wave number. The radiation field of the half-wave dipole can be calculated by the superposition principle through the current distribution as where is the directivity coefficient of the half-wave dipole antenna.

[0055] The Dipole antenna usually has omnidirectional characteristics in the horizontal direction (i.e. the plane perpendicular to the antenna axis), i.e. the signal is uniformly radiated or received within a range of 360 degrees. However, in the vertical direction (i.e. the direction parallel to the antenna axis), its radiation or reception ability is usually weak, forming a so-called "donut" shaped radiation pattern, and its three-dimensional directional diagram is as shown in Figure 3To adjust the directivity of Dipole antenna, a receiving array composed of multiple elements is usually adopted and directional receiving is realized through phase control and other techniques. The cross array arrangement proposed in the present scheme makes the elements themselves have certain directivity from the physical layer.

[0056] Firstly, consider the case where the two Dipole antennas corresponding to the cross array are equal in length and symmetrically arranged, as shown in FIG. 1. Figure 4 A rectangular coordinate system is established with the intersection of the cross array as the origin, the positive east direction as the x-axis and the positive north direction as the y-axis. The two axes of the cross array fall in the first and fourth quadrants, respectively. The antenna in the first quadrant is denoted as No. 1, and the angle between the axis and the positive east direction is The two Dipole antennas are independent of each other in the signal receiving process, and the received signals are y m,1 , y m,2 The received signal of the element is the sum of the two received components, i.e., y m = y m,1 +y m,2 . The signal gain of the cross array in the four directions is completely the same, and the directivity diagram is shown by the blue line. The element has the same signal receiving capability in the four directions, and although it can ensure that the array composed of the element can receive incoming wave signals in all directions, it cannot enhance the signal receiving in a specific direction from the physical characteristics of the element itself.

[0057] Secondly, consider the case where the lengths of the conductors on both sides of the feed points of the two vertical Dipole antennas are equal, but the lengths of the two axes are different. The specific case of the cross array corresponding to each element is shown in FIG. 2. Figure 5 The directivity diagram is shown by the blue line. Since increasing the length of the Dipole antenna within a certain range can improve the signal gain in the vertical direction, the directivity diagram of the element as a whole has four lobes, and the signal gain in the direction of the short axis is stronger than that in the direction of the long axis.

[0058] Further consider the case where the lengths of the conductors on both sides of the Dipole antenna are different, as shown in FIG. 3. Figure 6 The direction of the Dipole is the z-axis, and the lengths of the conductors on both sides of the feed point are l1 and l2, respectively. Then the current distribution in the Dipole antenna is:

[0059]

[0060] According to the Maxwell equations, we have:

[0061]

[0062] The directivity coefficient of the asymmetric Dipole antenna is obtained as follows:

[0063]

[0064] Consider the element composed of the asymmetric dipole antenna, according to the directivity coefficient, the directivity diagram can be made as shown in Figure 7 When the length of the conductor fixed at one end of the dipole antenna is four-tenths of the wavelength, and the length of the conductor at the other end is gradually lengthened within a certain range, the signal gain in the vertical direction of the antenna gradually increases, and the trend of the directivity diagram changes as shown in Figure 8 The blue line in the figure shows the directivity diagram of a conventional half-wave dipole, the red line shows the case where one end of the antenna is four-tenths of the wavelength and the other end is half a wavelength. The green line shows the case where the other end is one wavelength. When the length of the antenna conductor exceeds a certain range, the directivity diagram is no longer directed to the direction of the axis.

[0065] In this embodiment, the array composed of non-ideal cross arrays as elements is composed of the above asymmetric cross array, and the physical swing diagram of the whole array is as shown in Figure 9 Consider the case where the array has M elements, each element is the above asymmetric cross array, and the received original signal of the array is represented as:

[0066]

[0067] Among them, the signal received by the mth element is:

[0068]

[0069] y = [y1, y2, …, y M ]

[0070] In the formula, τ m is the time delay of the mth element relative to the first element, and τ m,α is the time delay of the two dipole antennas in the mth element relative to the center of the element. The signals received by the two dipole antennas of each element are independent of each other, and the received signal of the element is represented as the sum of the two sub-signals.

[0071] Consider the case of N signals, and the received signal of the whole array is: The array output is obtained by beamforming the received signal: z(t) = w H y.

[0072] First, according to the conventional wideband DOA estimation algorithm, the received signal is subjected to conventional beamforming, the DOA estimation direction corresponding to each frame is obtained, and the frequency statistics on the time axis are obtained. The highest frequency DOA estimation direction is obtained. Compare the highest frequency DOA estimation direction with the direction of the two axes of the cross array in the element, and give smaller weights to the two received signal components on the axis with smaller angle:

[0073]

[0074] In the formula, g m,1, g m,2 corresponding to the weight of the two dipole antenna received signal components in the mth array element, k0 is a constant, and θ is the direction with the highest statistical frequency, is the included angle of the cross array pointing in the array element.

[0075] The specific steps are as follows:

[0076] (1) For each frame of received signal in the time window, a conventional beamforming in the frequency domain is performed to obtain the beam output P j (θ n ) of the nth frame of incoming wave direction θ j .

[0077] (2) The maximum direction of each frame of beam output is recorded, i.e.

[0078] (3) The DOA estimates of all frames are counted to obtain the direction θ * with the highest statistical frequency.

[0079] (4) The two received signals on the cross array of the mth array element are weighted as described above.

[0080] The entire array received signal is weighted to obtain:

[0081]

[0082] In the formula, is the weighted received signal of the Mth array element, g M,α is the weight of the a th dipole antenna received signal component in the Mth array element, and y M,α is the a th dipole antenna received signal in the Mth array element.

[0083] As can be seen from the above, the present scheme proposes a non-ideal array element arrangement, the core of which is two vertically crossed dipole antennas. This special design makes the array element have an asymmetric structure, which is specifically manifested in that the lengths of the antenna conductors on both sides of the feed point are not equal. This asymmetric structure endows the array element with significantly enhanced receiving ability in the short axis direction (i.e. the direction of the shorter conductor) when placed horizontally. In actual arrangement, if the approximate direction of the signal source (for example, the position of a ship on a specific channel) is known in advance, the short axis direction of the array element can be adjusted to point to the signal source, thereby significantly improving the receiving performance.

[0084] In order to accurately estimate the direction of arrival (DOA) of the signal under the condition of non-ideal array element, the present scheme proposes an innovative correction algorithm. The algorithm consists of three main steps: conventional beamforming, frequency statistics, and secondary correction, which will be described one by one in detail.

[0085] In the conventional beamforming step, the traditional beamforming DOA method is used as the basis. First, the continuous signals in a certain time window are processed by frame. Each frame signal is processed by beamforming. Beamforming adjusts the phase and amplitude of each array element, so that the signal in a certain direction is enhanced, and the signal in other directions is suppressed. Specifically, for the nth frame signal, the beam output P j (θ n ) in each possible direction θ j is calculated by the beamforming algorithm. Then, the direction with the maximum beam output of the frame signal is recorded, that is This direction is considered as the possible DOA direction of the frame signal.

[0086] After completing the conventional beamforming of all frame signals, enter the frequency statistics step. The purpose of this step is to extract the most likely signal arrival direction from the DOA estimation results of all frames. By counting the frequency of each direction θ j in all frames, the direction θ * with the highest frequency is found as the final DOA estimation value under the time window length. This frequency statistics method can effectively reduce the misjudgment caused by single frame signal noise or interference, and improve the robustness of DOA estimation.

[0087] In the secondary correction stage, first compare the DOA direction obtained by frequency statistics with the angle between the cross array pointing direction and the array element. Since the main lobe of the directional diagram of the dipole antenna is perpendicular to the plane in which it is located, and the signal receiving ability of the longer axis is stronger, therefore under the condition of non-ideal array element, the signal gain of the longer axis corresponding to the cross array element is larger in the direction of the short axis pointing direction. If the DOA preliminary estimation direction is small with the short axis, it means that the gain contribution of the longer axis corresponding dipole antenna to the signal is larger, at this time the larger weight should be given to the receiving signal component corresponding to the longer axis; on the contrary, if the angle with the longer axis is small, the receiving signal component corresponding to the short axis should be given a larger weight. By adjusting the weight of the receiving signal component of the two dipole antennas in the array element, the DOA estimation value can be effectively corrected, and the estimation accuracy can be improved.

[0088] The present scheme proposes a new array model, and improves the DOA detection efficiency and accuracy through the weighting method based on the model. The array element placed by the cross array has the physical characteristics of the long axis and the short axis, which makes the long axis have higher spatial resolution near the direction, and is more sensitive to the direction change of the incoming wave. In the case of knowing the approximate direction of the incoming signal, the pre-placed direction of the cross array makes it have higher direction finding efficiency than the traditional ideal array element. In addition, the weighting method proposed in the present scheme improves the accuracy of the DOA method in the narrowband signal detection field by giving different weights to the two pairs of receiving electrodes corresponding to the axis in the incoming wave direction.

Claims

1. A method for direction of arrival estimation under non-ideal array element conditions, characterized in that, The method comprises the following steps: S1, arranging and constructing an asymmetric cross array as an array element based on a dipole antenna; S2, grouping a plurality of array elements into an array, performing frequency domain conventional beamforming processing on each frame of received signals in a time window to obtain beam outputs of each frame in different directions of arrival (DOA); S3, recording the maximum direction of each frame of beam outputs, i.e., the DOA estimation of each frame; S4, statistically estimating the DOA of all frames to obtain the direction with the highest statistical frequency; S5, performing weighted processing on two received signals of each array element to obtain the received signal of the entire array.

2. The method of claim 1, wherein, The asymmetric cross array in the step S1 comprises two dipole antennas, which are vertically crossed and placed, and the intersection point is the feed point.

3. The method of claim 2, wherein, The lengths of the conductors of the dipole antennas on both sides of the feed point are different, and the lengths are asymmetric.

4. The method of claim 3, wherein, The process of the step S2 comprises: Grouping a plurality of array elements into an array, and the received signal of each array element is the addition result of the received signals of the two dipole antennas; Performing frame processing on continuous signals in a certain time window; Beamforming is performed for each frame of signals, and the beamforming adjusts the phase and amplitude of each array element so that the signal in a specific direction is enhanced and the signal in other directions is suppressed, obtaining the beam output P n (θ j ) of each frame in different directions of arrival j .

5. The method of claim 4, wherein, The received original signal of the array is: y i = (y i,1 + y i,2 ) Wherein, y i is the received signal of the i th array element in the array, M is the total number of array elements in the array, y i,1 , y i,2 respectively correspond to the received signals of the two dipole antennas in the i th array element.

6. The method of claim 1, wherein, The step S3 specifically screens out the maximum beam output from the beam outputs corresponding to different directions of arrival of each frame, and records the direction corresponding to the maximum beam output, i.e., the DOA estimation of the frame signal.

7. The method of claim 6, wherein, The step S4 is specifically counting the number of occurrences of each direction θ j The direction θ with the highest number of occurrences is found * i.e. as the final DOA estimate for the time window.

8. The method of claim 3, wherein, The step S5 comprises the following steps: S51, for each array element, comparing the direction with the highest statistical frequency with the included angle of the cross array in the array element to determine the weight of the received signal components of the two dipole antennas in the array element, which is used for weighted processing on the received signals of the two dipole antennas in the array element; S52, obtaining the received signal of the entire array based on the weighted received signals of the array elements.

9. The method of claim 8, wherein, The weights of the received signal components of the two dipole antennas in the array element in the step S51 are respectively: wherein g m,1 , g m,2 correspond to the weight of the two dipole antenna signal components received in the mth array element, k0 is a constant, θ is the direction of the highest statistical frequency, is the included angle of the cross array in the array element.

10. The method of claim 9, wherein, The received signal of the entire array in the step S52 is: wherein, is the weighted received signal of the Mth array element, g M,α is the weight of the a-th dipole antenna received signal component of the Mth array element, y M,α is the a-th dipole antenna received signal of the Mth array element.

Images