JP2014169962A - Weight calculation method, weight calculation device, adaptive array antenna and radar system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance an unwanted signal suppression performance in a nonstationary unwanted signal environment with computational complexity corresponding to a range segmented system in a STAP system.SOLUTION: In a weight calculation circuit for calculating an adaptive weight by computing a covariance matrix from data of a cell estimated as formed only from unwanted waves, a plurality of range groups are generated by dividing received data on a certain fixed interval basis, and a weight application group is selected in order from among the plurality of range groups (S1). Regarding each of selected weight application groups, a number of ranges in the vicinity of a weight application range are extracted at random as training samples from any other range than the weight application range (S3) until achieving the predetermined number of training samples (S2). A weight is calculated from the extracted training samples by using the covariance matrix (S4), and the calculated weight is applied to the selected weight application range group (S5).

Description

  The present embodiment is a weight calculation method suitable for detecting a reflected signal from a target by suppressing unnecessary waves by weight control, a weight calculation device using the weight calculation method, and an adaptive method employing the weight calculation device. The present invention relates to an array antenna and a radar apparatus incorporating the adaptive array antenna.

  In recent years, in order to further improve the target detection accuracy, pulse radar apparatuses have incorporated an adaptive array antenna to perform so-called adaptive null steering. This adaptive null steering is a process that forms a received combined beam so that the directivity in the direction in which the unwanted wave arrives becomes zero (null) by performing weight control on the phase and amplitude of the received signal in the adaptive array antenna. is there. An adaptive array antenna used for such a purpose has a weight so that the above-mentioned received combined beam can be properly formed even in an environment where a large number of delayed signals arrive or an environment where clutter and unnecessary waves exist. There is a demand for control.

  Therefore, a weight control method that employs a space-time adaptive signal processing (STAP) method is attracting attention in an adaptive array antenna. This STAP system has the characteristics that SINR (Signal to Interference plus Noise Ratio) can be further improved and the directivity in the arrival direction of the unwanted wave can be made close to zero (null). Processing like this is performed. First, a target reflected signal is received by a plurality (N) of antennas (element antennas, that is, channels) arranged in an array, and the received signal is converted into a range (distance) cell (range) having a width corresponding to the received pulse width. cell) is stored in the corresponding cell positions of all the processing range cells formed so as to be continuous with a predetermined length on the time axis. Then, from the stored data, the covariance matrix is calculated from the data of the range cell excluding the range cell that is assumed to contain the target signal (referred to as the processing applied range cell), that is, the cell that is assumed to be formed only from unnecessary waves. To do. Finally, the beam combining circuit performs weight control on the antenna reception signal using the adaptive weight calculated based on the covariance matrix.

  By the way, in the conventional STAP system, for example, when calculating the weight for each range cell with respect to the received data, from the data (cell formed only from unnecessary waves) excluding the processing range cell (cell assumed to include a signal). A sliding window method in which training is performed and the covariance matrix is calculated, or the received data is divided into certain intervals, the covariance matrix is calculated using the data of all the intervals, and one weight is calculated within the interval A range segmented method is used.

Japanese Patent No. 4823261 JP 2007-003325 A

  However, in the sliding window method employed in the conventional STAP method, the covariance matrix and the weight are calculated for each range cell for all received data, so that a relatively large suppression performance is exhibited. Is enormous. The range segmented method, on the other hand, greatly reduces the amount of computation compared to the sliding window method, but because the range data is divided into a plurality of range groups, there is no unnecessary signal in the received data used for covariance matrix calculation. In some cases, or when the unnecessary signal suppression signal varies depending on the range, the suppression performance may be inferior.

  The present embodiment has been made to solve the above-described problem. In the STAP method, a weight calculation method, a weight calculation device, and an adaptive array that can realize suppression of unnecessary signals with a calculation amount equivalent to the range segmented method. An object is to provide an antenna and a radar device.

  In order to solve the above-described problem, according to the present embodiment, a plurality of processes having a length corresponding to a predetermined distance on a time axis are processed on reception data obtained by receiving a target reflected signal of a radar pulse by an antenna. Store in the corresponding cell position along the reception timing with respect to the range cell, and use the value stored in the plurality of processing range cells so that the arrival direction of the unnecessary wave becomes zero with respect to the arrival direction of the target reflected signal When calculating weights for the phase and amplitude of the received data to form a received composite beam, the received data is divided into certain intervals to generate a plurality of range groups, and weights are applied from the plurality of range groups Select groups in order, and for each selected weight application group, until the number of training samples determined in advance is reached, A mode in which training samples are randomly extracted from outside the weight application range, weights are calculated from the extracted training samples using a covariance matrix, and the calculated weights are applied to the selected weight application range group; To do.

The block diagram which shows one Embodiment of the weight calculation apparatus which concerns on this embodiment. The schematic block block diagram of the radar apparatus incorporating the weight calculation apparatus in the space-time adaptive signal processing concerning this embodiment, and an adaptive array antenna. FIG. 7 is a conceptual diagram of received data and weight application for sliding window type covariance matrix calculation. FIG. 6 is a conceptual diagram of received data and weight application for range-segmented covariance matrix calculation. FIG. 4 is a conceptual diagram of application of a target and an unnecessary signal, reception data for calculating a covariance matrix, and weight in the range segmented method. FIG. 5 is a conceptual diagram of target and unnecessary signal and received data for covariance matrix calculation and weight application according to the present embodiment. The flowchart which shows the flow of the weight application process by this embodiment.

  Hereinafter, embodiments will be described with reference to the drawings.

  FIG. 1 is a block diagram showing an embodiment of a weight calculation apparatus according to this embodiment. In FIG. 1, reference numeral 11 denotes a CPU (arithmetic processing unit), which is connected to a program storage ROM 13, a data input / output interface (I / O) 14, and a temporary data storage RAM 15 through a bus 12. The ROM 13 stores a weight calculation program according to the present embodiment. When an instruction to start processing is given, the CPU 11 loads the program from the ROM 13, loads data via the data input / output interface 14, and temporarily stores it in the RAM 15. The data is stored, data is appropriately read from the RAM 15, weight calculation processing is executed, and the obtained weight calculation result is output from the interface 14.

  The weight calculation apparatus according to the present embodiment having the above-described configuration uses the weight calculation method according to the present embodiment that suppresses SINR degradation with respect to the target Doppler frequency. Therefore, it is possible to obtain good SINR characteristics. Therefore, this weight calculation device is employed in an adaptive array antenna to calculate weights for input / output of individual antenna elements. According to this, it is possible to form a combined beam having good SINR characteristics.

  Incidentally, adaptive array antennas are employed in radar devices such as a synthetic aperture radar device for capturing a target. Therefore, by adopting the weight calculation device of the present embodiment for the adaptive array antenna as described above, it is possible to form a combined beam having good SINR characteristics. Therefore, in a radar device using this antenna, Will be able to better capture the goal.

  As an example of the radar apparatus, FIG. 2 shows a schematic block configuration diagram of a radar apparatus in which a weight calculation apparatus in space-time adaptive signal processing to which the present embodiment is applied is incorporated. In FIG. 2, reference numeral 21 denotes an adaptive array antenna that receives a target reflected signal of a radar pulse with N antenna elements. Each element output of the antenna 21 is received and detected by the receiving unit 22 and sent to the data storage unit 23. In the data storage unit 23, a storage area corresponding to a processing range cell having a length corresponding to a predetermined distance is prepared in advance, and input data is sequentially stored in a storage area at a corresponding cell position according to the reception timing.

  Here, a part of the antenna element output is sent to the reference signal estimation unit 24 and used as a reference for the amplitude and phase of the received signal. The excitation unit 26 periodically excites the reference signal estimation unit 24 and the reference signal generation unit 25 to estimate and generate a reference signal for weight calculation of each range cell corresponding to a predetermined distance.

  The accumulated data in the data accumulation unit 23 is sent to the spatiotemporal adaptive signal processing unit 27. The spatio-temporal adaptive signal processing unit 27 uses the weight calculation circuit 271 to generate a covariance matrix from data of a range cell excluding a range cell that is assumed to include a target signal, that is, a cell that is assumed to be formed only from unnecessary waves. Calculate. Finally, the beam combining circuit 272 performs weight control on the antenna reception signal with the adaptive weight calculated based on the covariance matrix to obtain output data.

  In the weight control in the spatio-temporal adaptive signal processing system having the above configuration, the weight calculation circuit 271 performs weight calculation for each range cell in order to calculate the adaptive weight. In this weight calculation circuit 271, the amount of change in the range direction of the unnecessary signal is determined, and the reception data larger than the amount of change in the unnecessary signal is calculated using the reception data randomly extracted from other ranges. Therefore, it is possible to improve the suppression performance for unnecessary signals, thereby obtaining good SINR characteristics.

  Next, the adaptive weight calculation method according to the present embodiment in the spatio-temporal adaptive signal processing will be described in comparison with the conventional sliding window method and range segmented method.

  FIG. 3 is a conceptual diagram of received data and weight application for calculating a sliding window type covariance matrix. As can be seen from FIG. 3, in the sliding window method, it is necessary to perform weight calculation for each range, and a huge amount of information processing is required. Not suitable.

  FIG. 4 is a conceptual diagram of received data and weight application for range-segmented covariance matrix calculation. FIG. 4A shows a state in which the received data is divided into certain intervals and range groups 1 to G are generated. When (b) selects range groups 1 to 3 in (a) and selects range groups 1 and 3 as covariance matrix calculation data to be applied to range group 2, (c) A case is shown in which range groups 2 to 4 are selected and range groups 2 and 4 are selected as covariance matrix calculation data to be applied to range group 2. In the range segmented method, as can be seen from FIG. 4, the received data is divided into a plurality of range groups (1 to G in the figure), and the range groups to which weights are applied are shifted in groups. For this reason, it can be seen that the processing load can be significantly improved as compared with the sliding window method.

  However, in the range segmented method, for example, when the relationship between the target and unnecessary signal in the range group 4 is as shown in FIG. An unnecessary signal does not exist in the range (range group 3 and range group 5) for calculating the covariance matrix, and even if a weight is applied to the range group 4, the unnecessary signal is not suppressed.

  FIG. 6 is a conceptual diagram of target and unnecessary signal, reception data for covariance matrix calculation, and weight application according to the present embodiment. The color shading indicates range extraction, and the darker the color, the higher the range extraction probability. For example, when the relationship between the target and the unnecessary signal in the range group 4 is as shown in FIG. 6, there are many ranges in which the unnecessary signal does not exist in the data on both sides of the range group 4, but the covariance matrix from other ranges. If the data used for is acquired, an unnecessary signal can be obtained, and the unnecessary signal can be suppressed.

  In general, if the weight application range is close to the range extracted as a training sample, the unnecessary signal suppression effect is high. Therefore, a random number is generated by a function such as a normal distribution to extract many ranges near the weight application range, and a covariance matrix is derived. By using this method, the suppression of unnecessary signals becomes high, and even when there are ranges in which the properties of the unnecessary signals are different, it is possible to sufficiently cope with them.

  FIG. 7 is a flowchart showing the flow of weight application processing according to this embodiment. First, the weight application group i (first time i = 1) is selected (step S1), and the training sample number K is determined (step S2). Next, until the total number of training samples reaches K, training samples are randomly extracted from outside the weight application range (step S3). Then, the weight is calculated from the training sample using the covariance matrix (step S4), and the weight is applied to the weight application range group 1 (step S5). Subsequently, it is determined whether the weight application group i is i = G (G is the number of all weight application range groups) (step S6). If i = G, the i-th (i = i + 1) process is performed as steps i to S5. The series of processing up to is repeated, and when i = G and the weight to the weight application group G is applied, the series of processing ends.

  As described above, according to the present embodiment, as a range for calculating the covariance matrix from the received data of the weight application range group, a random number is generated by a function such as a normal distribution to extract many ranges near the weight application range. Therefore, a training sample in which clutter is present can be extracted almost certainly, and unnecessary signals can be suppressed with extremely high accuracy.

  Note that the present embodiment is not limited to the above-described embodiment as it is, and can be embodied by modifying the components without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  DESCRIPTION OF SYMBOLS 11 ... CPU (arithmetic processing unit), 12 ... Bus, 13 ... ROM for program storage, 14 ... Data input / output interface (I / O), 15 ... RAM for temporary data storage, 21 ... Adaptive array antenna, 22 ... Receiver , 23... Data storage unit, 24... Reference signal estimation unit, 25 .. reference signal generation unit, 26... Excitation unit, 27 .. spatiotemporal adaptive signal processing unit, 271.

Claims (4)

The reception data obtained by receiving the target reflected signal of the radar pulse by the antenna is stored in the corresponding cell position according to the reception timing with respect to a plurality of processing range cells having a length corresponding to the predetermined distance on the time axis,
Using the values stored in the plurality of processing range cells, with respect to the phase and amplitude of the reception data for forming a reception combined beam so that the arrival direction of the unwanted wave becomes zero with respect to the arrival direction of the target reflected signal A method for calculating weights,
Dividing the received data in a certain interval to generate a plurality of range groups,
Select a weight application group from the plurality of range groups in order,
For each selected weight application group, a large number of ranges near the weight application range are randomly extracted as training samples from outside the weight application range until the number of training samples determined in advance is reached, and covariance is obtained from the extracted training samples. A weight calculation method comprising calculating a weight using a matrix and applying the calculated weight to the selected weight application range group. Stores received data obtained by receiving a target reflected signal of a radar pulse by an antenna at a corresponding cell position in accordance with the reception timing for a plurality of processing range cells having a length corresponding to a predetermined distance on the time axis Means,
Using the values stored in the plurality of processing range cells, with respect to the phase and amplitude of the reception data for forming a reception combined beam so that the arrival direction of the unwanted wave becomes zero with respect to the arrival direction of the target reflected signal Calculating means for calculating weights,
The calculation means divides the received data into a certain interval to generate a plurality of range groups, sequentially selects weight application groups from the plurality of range groups, and is determined in advance for each selected weight application group. Until the number of training samples reaches, it is randomly extracted so that the range near the weight application range increases as a training sample from other than the weight application range, and the weight is calculated using the covariance matrix from the extracted training sample, A weight calculation apparatus that applies the calculated weight to the selected weight application range group. An adaptive array antenna in which a plurality of element antennas are arranged in an array, the direction of which is controlled in an arbitrary direction and a target reflected signal of a radar pulse is received,
Storage means for storing received data obtained by receiving the target reflected signal at a corresponding cell position along the reception timing with respect to a plurality of processing range cells having a length corresponding to a predetermined distance on the time axis,
Using the values stored in the plurality of processing range cells, with respect to the phase and amplitude of the reception data for forming a reception combined beam so that the arrival direction of the unwanted wave becomes zero with respect to the arrival direction of the target reflected signal A calculating means for calculating a weight;
Beam forming means for performing weight control on the target reflected signal by the calculated weight to form a received combined beam;
The calculation means divides the received data into a certain interval to generate a plurality of range groups, sequentially selects weight application groups from the plurality of range groups, and is determined in advance for each selected weight application group. Until the number of training samples reaches the number of training samples, a number of ranges near the weight application range are randomly extracted as training samples from outside the weight application range, and weights are calculated from the extracted training samples using a covariance matrix. An adaptive array antenna, wherein the weight is applied to the selected weight application range group. A plurality of element antennas are arranged in an array, the target reflection signal of the radar pulse is received by being controlled in an arbitrary direction, and the target reflection signal is received by the given adaptive weight with respect to the reception data obtained by reception. An adaptive array antenna that forms a received combined beam by performing weight control on
The reception data of the target reflection signal is stored in the corresponding cell positions along the reception timing with respect to a plurality of processing range cells having a length corresponding to a predetermined distance on the time axis, and values stored in the plurality of processing range cells A weight calculation device for calculating a weight for the phase and amplitude of the reception data for forming a reception combined beam so that the arrival direction of the unwanted wave becomes zero with respect to the arrival direction of the target reflected signal,
A signal processing device for detecting a target from a target reflection signal subjected to weight control by the adaptive array antenna,
The weight calculation apparatus generates a plurality of range groups by dividing the received data in a certain interval, sequentially selects weight application groups from the plurality of range groups, and determines each of the selected weight application groups in advance. Until the number of training samples is reached, a lot of ranges near the weight application range are randomly extracted as training samples from other than the weight application range, and weights are calculated from the extracted training samples using a covariance matrix. And applying the selected weight to the selected weight application range group.

Images