JP7749882B2 - Speed estimation device, target classification device, speed estimation method, and target classification method - Google Patents

Description

The present disclosure relates to a speed estimation device, a target classification device, a speed estimation method, and a target classification method.

There is a technology for detecting unknown radio wave sources using received signals based on radio waves emitted from the radio wave source, relayed by a satellite, and received by an antenna. For example, Non-Patent Document 1 describes a technology for locating the position of a radio wave source by performing correlation processing between received signals and then using information indicating the time difference of arrival (TDOA) between the received radio waves and the frequency difference of arrival (FDOA), which indicates the Doppler frequency difference between the received radio waves.

d. p. Haworth, “Interference localization for eutelsat satellites-the first European transmitter location system,” International journal of satellite communications, vol. 15, 155-183 (1997).

The conventional technology described in Non-Patent Document 1 can estimate the position of an unknown radio wave source (hereinafter referred to as a target), but has the problem of not being able to estimate the target's movement information. For this reason, the conventional technology was unable to obtain target movement information in addition to target position information and use this information to classify targets.

The present disclosure aims to solve the above problem and provide a speed estimation device that can estimate target speed information for use in target classification.

The speed estimation device disclosed herein receives radio waves transmitted from a target, which is a radio wave source, and arrives via multiple satellites. It sequentially acquires control points, which are estimated positions of the target, based on the arrival time difference and arrival frequency difference between the received signals determined by correlation processing between the received signals. It includes: a frequency distribution calculation unit that calculates the frequency distribution of the control points converted to latitude and longitude; a control point extraction unit that extracts control points included in the area with the highest frequency from the frequency distribution; an error ellipse calculation unit that calculates, for each pair of satellites, an error ellipse, which is an area where the control point extracted from the frequency distribution may exist; a theoretical value calculation unit that calculates the theoretical value of the axis length of the error ellipse in which the control point exists, based on the signal-to-noise ratio of the received signal; a variance value calculation unit that calculates the variance value of the FDOA between received signals accompanying the movement of the target, based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the control point extracted from the frequency distribution exists; and a target velocity component calculation unit that calculates the velocity component of the target based on the variance value.

According to the present disclosure, a control point included in the region with the highest frequency is extracted from the frequency distribution of the control points of the coordinate-transformed radio wave source, and the variance value of the FDOA between received signals associated with the movement of the target is calculated based on the difference between the axis length of the error ellipse in which the control point extracted from the frequency distribution exists and the theoretical value of the axis length of the error ellipse calculated based on the signal-to-noise ratio of the received signal. The target's velocity component is then calculated based on this variance value. This allows the speed estimation device disclosed herein to estimate target speed information for use in target classification.

1 is a block diagram showing an example of the configuration of a target classification device according to a first embodiment; FIG. 10 is a diagram illustrating an example of a frequency distribution of target orientation points. FIG. 10 is a diagram illustrating an outline of a process for extracting orientation points from a frequency distribution. FIG. 10 is a diagram showing information on orientation points stored in an orientation point storage unit. FIG. 10 is a diagram illustrating the relationship between the value of γ ² and the probability that the target orientation point is within the error ellipse. FIG. 10 is a diagram showing an error ellipse in which orientation points exist. FIG. 1 is a diagram illustrating the definition of a velocity vector in a ground coordinate system. 3 is a flowchart illustrating a target classification method according to the first embodiment. 3 is a flowchart showing a speed estimation method according to the first embodiment. 10A and 10B are block diagrams showing a hardware configuration for realizing the functions of the speed estimation device according to the first embodiment. FIG. 10 is a block diagram showing a configuration example of a target classification device according to a second embodiment. 10 is a flowchart showing a speed estimation method according to the second embodiment; FIG. 10 is a block diagram showing a configuration example of a target classification device according to a third embodiment. 10 is a flowchart showing a speed estimation method according to the third embodiment.

Embodiment 1.
Fig. 1 is a block diagram showing an example of the configuration of a target classification device 1 according to embodiment 1. In Fig. 1, the target classification device 1 is connected to ground station antennas #1A, #2A, and #3A, which receive radio waves from satellites #1, #2, and #3, respectively. In addition to the target classification device 1, Fig. 1 also shows target A, which is the radio wave source and is the classification target. Target A is a radio wave source whose position is unknown.

The number of earth station antennas provided corresponds to the number of satellites. In Fig. 1, three satellites #1, #2, and #3 are provided as satellites, and three earth station antennas #1A, #2A, and #3A are provided as earth station antennas.
Satellites #1, #2, and #3 each receive radio waves from target A and transmit the received radio wave signals to ground station antennas #1A, #2A, and #3A, which in turn receive the radio waves that have arrived via satellites #1, #2, and #3, respectively.

The target classification device 1 comprises a speed estimation device 2 and a target classification unit 3. The speed estimation device 2 comprises signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, control point calculation units 6-1, 6-2, and 6-3, control point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, and reception SNR estimation units 10-1, 10-2, and 10-3. These components make up the positioning processing unit. Note that the signal receiving units 4-1, 4-2, and 4-3 may also be provided in ground station antennas #1A, #2A, and #3A.

Radio waves transmitted from target A and arriving via satellites #1, #2, and #3 are received by ground station antennas #1A, #2A, and #3A. The positioning processing unit performs correlation processing between the signals received by ground station antennas #1A, #2A, and #3A, calculates the TDOA (time difference of arrival) and FDOA (frequency difference of arrival) between the received signals based on the results of the correlation processing, and sequentially acquires the ground point of target A by estimating the estimated position of target A based on the TDOA and FDOA.

The positioning processing unit will now be described in detail.
The number of signal receiving units provided corresponds to the number of satellites. Figure 1 shows a case where three signal receiving units 4-1, 4-2 and 4-3 are provided as signal receiving units.
The signal receiving units 4-1, 4-2 and 4-3 generate analog signals by performing various signal processing such as amplification, band pass processing (filter processing) and frequency conversion processing on the RF (radio frequency) output from the ground station antennas #1A, #2A and #3A.

These analog signals are complex signals having in-phase and quadrature components. The signal receivers 4-1, 4-2, and 4-3 then convert the analog signals into received signals, which are digital complex signals, to obtain a complex signal vector x(t). In the case of FIG. 1 , there are three earth station antennas, so the signal receivers 4-1, 4-2, and 4-3 obtain a total of three received signals: complex signal vectors x ₁ (t), x ₂ (t), and x ₃ (t). These three received signals are output to correlation processors 5-1, 5-2, and 5-3, respectively, and then to received SNR estimation units 10-1, 10-2, and 10-3.

The correlation processing units 5-1, 5-2, and 5-3 calculate TDOA and FDOA information by performing correlation processing between the received signals acquired by the signal receiving units 4-1, 4-2, and 4-3, respectively. For example, the correlation processing units 5-1, 5-2, and 5-3 calculate TDOA and FDOA information by extracting peak values of the CAF (Cross Ambiguity Function) according to the method described in Non-Patent Document 1 as the correlation processing between the above complex signal vectors.

When target A is a radar wave source, there will be multiple peak values of the CAF, and as a result, multiple TDOA and FDOA information will be calculated. Signals indicating the TDOA and FDOA information calculated by correlation processing units 5-1, 5-2, and 5-3 are output to control point calculation units 6-1, 6-2, and 6-3. The following explanation assumes that target A is a radar wave source.

The calculated TDOA for each set of satellites #1, #2, and #3 are denoted as τ _12(i) , τ _23(j) , and τ _13(k) , respectively, and the FDOA are denoted as f _12(i) , f _23(j) , and f _13(k) , respectively, where i, j, and k are positive integers.
The numbers of CAF peaks calculated from the received signals between satellites #1, #2, and #3 are _N12 , _N23 , and _N13 , respectively, where i, j, and k satisfy the ranges 1≦i≦ _N12 , 1≦j≦ _N23 , and 1≦k≦ _N13 , respectively.
TDOAτ _12(i) and FDOAf _12(i) are output to a control point calculation unit 6-1, TDOAτ _23(j) and FDOAf _23(j) are output to a control point calculation unit 6-2, and TDOAτ _13(k) and FDOAf _13(k) are output to a control point calculation unit 6-3.

The control point calculation units 6-1, 6-2, and 6-3 use the TDOA and FDOA information to solve the simultaneous equations of the following formulas (1) to (7) to calculate control points based on the TDOA and FDOA, respectively. In the following formulas (1) to (7), the speed of light is c, the position vectors of satellites #1, #2, and #3 are p _s1 , p _s2 , and p _s3 , the velocity vectors of satellites #1, #2, and #3 are v _s1 , v _s2 , and v _s3 , respectively, and the radius of the Earth when the Earth is a sphere is R _E. Also, the center frequency of the received signal is f ₀ .
Specifically, the orientation point calculation unit 6-1 calculates the orientation point p _12(i) according to the following formulas (1), (2), and (7). The orientation point calculation unit 6-2 calculates the orientation point p _23(j) according to the following formulas (3), (4), and (7). The orientation point calculation unit 6-3 calculates the orientation point p _13(k) according to the following formulas (5), (6), and (7).
Thereafter, a signal indicating the orientation point p _12(i) calculated by the orientation point calculation unit 6-1 is output to the orientation point accumulation unit 7. A signal indicating the orientation point p _23(j) calculated by the orientation point calculation unit 6-2 is output to the orientation point accumulation unit 7. A signal indicating the orientation point p _13(k) calculated by the orientation point calculation unit 6-3 is output to the orientation point accumulation unit 7.

Since the information on the orientation point of target A is calculated corresponding to the set of TDOA and FDOA output from the correlation processing units 5-1, 5-2, and 5-3, the orientation point calculation units 6-1, 6-2, and 6-3 calculate N ₁₂ , N ₂₃ , and N ₁₃ orientation points, respectively.
When solving the above simultaneous equations, either a method of finding a solution through iterative calculations such as Newton's method or steepest descent method, or a method of finding the roots of a polynomial, as described in the reference literature, may be used.
(Reference) K.C. Ho and Y. T. Chan, “Geolocation of a known altitude object from TDOA and FDOA measurements” in IEEE Transactions on Aerospace and Electronic Systems, vol. 33, no. 3, pp. 770-783, July 1997.

The orientation points calculated by the orientation point calculation units 6-1, 6-2, and 6-3 are collected into a single set and stored in the orientation point storage unit 7. For example, the grouped set of orientation points is stored in the orientation point storage unit 7 as p _all = [p _12(i) , p _23(j) , p _13(k) ].

The accumulation of the control points in the control point accumulation unit 7 is controlled by an accumulation time determination unit 8. When the time when the control point accumulation unit 7 starts to acquire information about the control points is set to 0, the accumulation time determination unit 8 measures the elapsed time t from that point and determines whether the elapsed time t has reached a preset accumulation time T. If t<T and the elapsed accumulation time t has not reached the accumulation time T, the accumulation time determination unit 8 causes the control point accumulation unit 7 to continue accumulating p _all , creates a new p _all-T , and stores it in the control point accumulation unit 7. If t≧T and the elapsed accumulation time t has reached the accumulation time T, the accumulation time determination unit 8 outputs information indicating p _all-T stored in the control point accumulation unit 7 to the coordinate conversion unit 9.

The coordinate conversion unit 9 converts the three-dimensional vector information of each control point stored in p _all-T output from the control point accumulation unit 7 into information of latitude φ and longitude θ. The information of the control point converted by the coordinate conversion unit 9 is output to the frequency distribution calculation unit 11 as p _all-T-latlon .

The reception SNR estimation units 10-1, 10-2, and 10-3 perform a Fourier transform on the complex signal vectors received by the signal reception units 4-1, 4-2, and 4-3, and estimate the SNR of the received signal based on the difference between the peak value of the frequency spectrum and the noise floor, etc. The SNRs estimated by the reception SNR estimation units 10-1, 10-2, and 10-3, respectively, are output to the error ellipse axis length theoretical value calculation unit 14.

In addition to the positioning processing unit, the speed estimation device 2 also includes a frequency distribution calculation unit 11, a control point extraction unit 12, an error ellipse calculation unit 13, an error ellipse axis length theoretical value calculation unit 14, an FDOA movement variance value calculation unit 15, and a target speed component calculation unit 16. These components generate information for estimating the speed of target A.

The frequency distribution calculation unit 11 calculates the frequency distribution of the control points based on the latitude and longitude information obtained by the coordinate conversion unit 9. The frequency distribution of the control points is a distribution for evaluating the density of the control points. FIG. 2 is a diagram showing an example of the frequency distribution of the control points of target A. The frequency distribution calculation unit 11 creates the frequency distribution shown in FIG. 2 for p _all-T-latlon output from the coordinate conversion unit 9. For example, the frequency distribution calculation unit 11 generates table data in which the direction of latitude φ is divided into M grids and the direction of longitude θ is divided into N grids, and calculates frequency distribution data in which the number of control points present in area B (hereinafter referred to as cells) of each grid in the table data is associated with the number of control points present in area B. The frequency distribution data calculated by the frequency distribution calculation unit 11 is output to the control point extraction unit 12.

The control point extraction unit 12 extracts control points included in the area with the maximum frequency from the frequency distribution calculated by the frequency distribution calculation unit 11. Fig. 3 is a diagram showing an overview of the process of extracting control points from the frequency distribution, showing frequency distribution data. In Fig. 3, cell C of (Δθ ₄ , Δφ ₆ ) in the frequency distribution is the cell where the most control points exist, i.e., the frequency of control points is the largest. As shown in Fig. 3, the control point extraction unit 12 extracts control points existing in cell C with the maximum frequency in the frequency distribution. The set of control points extracted from cell C can be divided into three control points calculated for each set of satellites #1, #2, and #3.
The control point extraction unit 12 extracts control points p _ext-12 , p _ext-23 and p _ext-13 corresponding to the set of satellites #1, #2 and #3 from cell C and outputs them to the error ellipse calculation unit 13. Furthermore, the control point extraction unit 12 calculates these average values, p ₁₂ bar, p ₂₃ bar and p ₁₃ bar, and outputs them to the error ellipse calculation unit 13.

Here, the reason why the orientation point extraction unit 12 extracts the orientation point included in the area with the maximum frequency will be explained. FIG. 4 is a diagram showing information on the orientation points stored in the orientation point storage unit 7.
When viewed at a single time, it is difficult to determine whether a location point calculated based on a signal acquired by a radar wave source or the like is derived from target A, which is the true radio wave source, or is an ambiguity.
However, for example, as shown in FIG. 4, when the accumulated results of control points over multiple time periods are viewed, the ambiguity fluctuates in accordance with the movements of satellites #1, #2, and #3, while the true target A actually exists at that location, and therefore the control point appears in the same location regardless of the movements of satellites #1, #2, and #3.
Furthermore, even if target A is moving, the speed is sufficiently small compared to the speed of orbiting satellites #1, #2, and #3, so that as a result, when viewed over multiple time periods, the behavior of the position fluctuation differs depending on whether the control point is derived from target A or an ambiguity, and if it is derived from target A, it exhibits characteristics that are relatively close to a fixed point. In Fig. 4, symbol D indicates a control point derived from target A, and symbol E indicates an ambiguity control point. By using this characteristic, the speed estimation device 2 can extract the control point derived from target A without using azimuth information.

The error ellipse calculation unit 13 calculates an error ellipse, which is an area where the orientation points extracted from the frequency distribution may exist, for each set of satellites #1, #2, and #3. For example, the error ellipse calculation unit 13 calculates an error ellipse indicating an area where each orientation point may exist for each of orientation points p _ext-12 , p _ext-23 , and p _ext-13 extracted from the frequency distribution. Specifically, the error ellipse calculation unit 13 calculates error covariance matrices R ₁₂ , R ₂₃ , and R ₁₃ from each orientation point according to the following equations (8), (9), and (10). However, in the following equations (8), (9), and (10), the centers of gravity of p _ext-12 , p _ext-23 , and p _ext-13 , respectively, are set to p _ext-12bar , p _ext-23bar , and p _ext-13bar .

5 is a diagram showing the relationship between the value of ^γ2 and the probability that the orientation point of target A exists within the error ellipse (hereinafter referred to as the existence probability of the orientation point). In the above equations (8), (9), and (10), ^γ2 is determined by a ^chi -squared distribution with two degrees of freedom, and as shown in FIG. 5, the existence probability of each orientation point changes depending on its value.

Next, the error ellipse calculation unit 13 calculates the axis lengths _σφ12 , σθ12, σφ23, _σθ23 , σφ13, and _σθ13 of the error ellipse and the slopes _φ12 , φ23, and φ13 from the origin according to the following equations (11), (12), ( ₁₃ ), (14), (15), (16), ( ₁₇ ), ( ₁₈ ), and (19) based on the components of the error covariance matrices _R12 , _{R23 , and R13} . The axis length parameters of the error ellipse calculated by the error ellipse calculation unit 13 are output to the FDOA shift variance calculation unit 15. Fig. 6 is a diagram showing an error ellipse in which an orientation point exists. In Figure 6, symbols E1, E2, and E3 are error ellipses in which orientation points exist, and the error ellipse calculation unit 13 calculates the error ellipses E1, E2, and E3 using information on the orientation points extracted from the frequency distribution by the orientation point extraction unit 12.

The theoretical error ellipse axis length calculation unit 14 is a theoretical value calculation unit that calculates the theoretical value of the axis length of the error ellipse in which the orientation point exists, based on the SNR of the received signal. The axis length of the error ellipse is a parameter that represents the spread of the ellipse. Furthermore, as described in the above-mentioned reference document, for example, the theoretical error ellipse axis length calculation unit 14 calculates the theoretical value σ ² _τ of the TDOA variance value and the theoretical value σ ² _f of the FDOA variance value using information on the SNR of the received signal according to the following equations (20) and (21). In the following equations (20) and (21), B is the signal bandwidth, and T is the signal observation time.

The theoretical TDOA variance value σ ² _τ and the theoretical FDOA variance value σ ² _f calculated by the error ellipse axis length theoretical value calculation unit 14 are output to the FDOA moving variance value calculation unit 15 .
When there are a plurality of observed values, the error ellipse axis length theoretical value calculation unit 14 outputs the arithmetic mean value of these values to the FDOA moving variance value calculation unit 15 .

The FDOA moving variance calculation unit 15 is a variance calculation unit that calculates the variance of the FDOA between received signals accompanying the movement of target A based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists.
First, the FDOA moving variance calculation unit 15 estimates the FDOA variance value that occurs with the movement of target A, using the error ellipse axis length parameters calculated by the error ellipse calculation unit 13 and the theoretical TDOA variance value σ ² _τ and the theoretical FDOA variance value σ ² _f calculated by the error ellipse axis length theoretical value calculation unit 14. Here, the relationship between the theoretical TDOA variance value σ ² _τ and the theoretical FDOA variance value σ ² _f and the variances σ ² _φ and σ ² _θ in the space of latitude φ and longitude θ is expressed by the following equation (22). Note that P _SN is the error covariance matrix in the space of latitude φ and longitude θ.

In the above equation (22), D is a 2×2 matrix expressed as D=NE, and N and E are expressed by the following equations (23) and (24), respectively, where Δ _p represents the gradient at the position vector p.

In the above equations (23) and (24), each component of p is expressed by the following equation (25) using information on the latitude φ and longitude θ.

Assuming that D ⁻¹ = D _inv and the (a, b) component of the matrix D _inv is D _inv (a, b), the components P _SN (1, 1), P _SN (1, 2) and P _SN (2, 2) of P _SN in the above equation (22) can be expressed by the following equations (26), (27) and (28).

When target A is moving, the variance value σ ² _f-move associated with the movement of target A is added to the variance value of the FDOA. On the other hand, when the amount of movement of target A is sufficiently small compared to the momentum of the satellite, the variance value of the TDOA does not depend on the movement of target A. Taking this into consideration, the error covariance matrices P _SN-v (1,1), P _{SN-v (1,2), and P SN -v} (2,2) when target A is moving can be expressed by the following equations (29), (30), and (31).

By using the above equations (26), (27), (29) and (31), the variance value σ ² _f-move occurring in the FDOA due to the movement of target A can be estimated according to the following equation (32).

The first term P _SN+v (1,1) and the second term P _{SN+v (2,2) of the numerator of the above equation (32) are calculated based on the axis lengths of the error ellipse calculated by the error ellipse calculation unit 13. Furthermore, the third term -P SN} (1,1) and the fourth term -P SN (2,2) of the numerator of the above equation (32) are calculated using the theoretical errors σ _φ and σ _θ in the space of latitude φ and longitude _θ calculated based on information on the _SNR of the received signal. When there is information on a plurality of SNRs of the received signal, σ _φ and σ _θ are average values of the results calculated for each SNR.

The FDOA moving variance calculation unit 15 calculates the variance σ ² _f-move(12) occurring in the FDOA due to the movement of target A for the received signals from satellites #1 and #2 according to the following equation (33).

The FDOA moving dispersion value calculation unit 15 calculates the dispersion value σ ² _f-move(23) occurring in the FDOA due to the movement of target A for the received signals from satellites #2 and #3 according to the following equation (34). Furthermore, the FDOA moving dispersion value calculation unit 15 calculates the dispersion value σ ² _f-move(13 ) occurring in the FDOA due to the movement of target A for the received signals from satellites #1 and #3 according to the following equation (35).

The target velocity component calculation unit 16 calculates the velocity components of the target A in each direction based on the FDOA variance value calculated by the FDOA moving variance value calculation unit 15 .
Fig. 7 is a diagram showing the definition of a velocity vector in the Earth's surface coordinate system. In Fig. 7, the velocity components in the north, east, and zenith directions (vertical directions) on the Earth's surface are defined as _vN , _vE , and _vU . The relationship between the velocity _vSurf = [ _vN , _vE , _vU ] in the Earth's surface coordinate system and the velocity _vtarget = [ _vx , _vy , _vz ] in the ECEF coordinate system can be expressed by the following equations (36) and (37) using a transformation matrix B for performing coordinate transformation.

The difference δFDOA in FDOA that occurs as the target A moves can be formulated as shown in the following equation (38) for the received signals from the satellites #1 and #2.

The variance value σ ² _FDOA,mv of the FDOA occurring with the movement of target A for the received signals from satellites #1 and #2 can be expressed by the following equation (39).
The following equation (39) relates to the position and velocity of target A at a specific time.

When target A is observed multiple times, if the average position vector of target A during the observation time is p ₁₂ bar and the averages of the velocity components of target A in each direction are v _N bar, v _E bar, and v _U bar, then the above equation (39) can be expressed by the following equation (40): In the following equation (40), the average values p ₁₂ bar, p ₂₃ bar, and p _{13 bar of the position vector of target A during multiple time observations for satellite #1 and satellite #2, satellite #2 and satellite #3, and satellite #3 and satellite #1} are output from the control point extraction unit 12.

The above formula (40) can be expressed by the following formulas (41) and (42).

Let transformation matrix B be B = [b ₁ , b ₂ , b ₃ ]. When the variance calculated by FDOA moving variance value calculation unit 15 using the above equations (33), (34), and (35) is the left-hand side σ ² _FDOA,mv of the above equation (40), the following three equations (43) _{, (44), and (45) are obtained, which are related to the FDOA variance values σ 2 FDOA,mv} ⁽ ₁₂₎ , σ ² FDOA,mv(23) , and σ ² _FDOA,mv (13) associated with the movement of target A.

The target velocity component calculation unit 16 calculates information v ² _{N bar, v} 2 E bar, and v ² U bar related to the components of the velocity vector of target A in each direction according to the following equations (46), (47), and (48), using the variance values σ 2 FDOA _,mv(12) bar, σ ² _FDOA,mv ( ²³ ) bar, and σ 2 _FDOA ,mv( ¹³ ₎ bar of the FDOA associated with the movement of target A obtained by the above equations (43), (44), and ( ⁴⁵ ₎ , respectively.

In the above equations (46), (47), and (48), M is a matrix expressed by the following equation (49), and its (a, b) component is defined as M(a, b). The target velocity component calculation unit 16 outputs information v ² _N bar, v ² _E bar, and v ² _U bar relating to the calculated components of the velocity vector in each direction, together with average values p ₁₂ bar, p ₂₃ bar, and p ₁₃ bar of the position vector of target A, to the target classifier 3.

The target classification unit 3 estimates the speed of target A from information v ² _N bar, v ² _E bar, and v ² _U bar about the components of each direction of the velocity vector of target A, and estimates position information of target A from average values p ₁₂ bar, p ₂₃ bar, and p ₁₃ bar of the position vector of target A. Then, the target classification unit 3 classifies target A based on the estimated speed information and position information of target A.

For example, the target classifying unit 3 calculates the estimated position p _est bar of the target A using the following equation (50).

The target classification unit 3 calculates the absolute value (speed) v _est of the speed of target A using the following equation (51). Although the case where the target classification unit 3 calculates the speed v _est of target A has been described, the speed estimation device 2 may calculate the speed v _est of target A. In this case, the target speed component calculation unit 16 estimates the speed of target A from information v ² _N bar, v ² _E bar, and v ² _U bar related to the components of the speed vector in each direction.

Note that while Figure 1 shows a speed estimation device 2 equipped with a positioning processing unit, the positioning processing unit may be equipped in a positioning device provided separately from the speed estimation device 2. In other words, the speed estimation device 2 is capable of acquiring control point information from the positioning device, and is equipped with a frequency distribution calculation unit 11, a control point extraction unit 12, an error ellipse calculation unit 13, an error ellipse axis length theoretical value calculation unit 14, an FDOA moving variance value calculation unit 15, and a target speed component calculation unit 16.

Next, a target classification method using the target classification device 1 will be described.
8 is a flowchart showing the target classification method according to embodiment 1, where the parameters _v0 , _v1 , _v2 , and _v3 shown in Fig. 8 are constants having the magnitude relationship _v0 < _v1 < _v2 < _v3 . The target classification unit 3 estimates the estimated position _pestbar of target A and the absolute value (speed) vestbar of the velocity of target A according to the above equations (50) and ( ₅₁ ) (step ST1).

Next, the target classification unit 3 compares the target speed v _est bar with the constant v ₁ to determine whether v _est bar < v ₁ (step ST2). If _{v est} bar < v ₁ (step ST2; YES), the target classification unit 3 determines whether the estimated position p _est bar of target A is on land (step ST3). If the estimated position p _est bar of target A is not on land (step ST3; NO), the target classification unit 3 classifies target A as a ship (step ST4).

If the estimated position _{p_est} of target A is on land (step ST3; YES), the target classification unit 3 compares the target's speed _{v_est} with a constant _{v_0} to determine whether _{v_est} < _{v_0} (step ST5). If _{v_est} < _{v_0} (step ST5; YES), the target classification unit 3 classifies target A as a fixed station or a stationary vehicle (step ST6). If _{v_est} < _{v_1} and _{v_est} ≧ _{v_0} , that is, _{v_0} ≦ _{v_est} < _{v_1} (step ST5; NO), the target classification unit 3 classifies target A as a vehicle (step ST7).

On the other hand, if v _est bar ≧ _v1 (step ST2; NO), the target classification unit 3 determines whether _v1 < v _est bar < _v2 (step ST8). If _v1 < v _est bar < _v2 (step ST8; YES), the target classification unit 3 classifies target A as an unmanned aircraft or a rotary-wing aircraft (step ST9).

If _v1 < _vest bar < _v2 is not true (step ST8; YES), the target classification unit 3 determines whether _v2 < _vest bar < _v3 is true (step ST10). If _v2 < _vest bar < _v3 is true (step ST10; YES), the target classification unit 3 classifies target A as a civil aircraft (step ST11). If _v2 < _vest bar < _v3 is not true (step ST10; NO), the target classification unit 3 classifies target A as a military aircraft (step ST12).

Next, a speed estimation method performed by the speed estimation device 2 will be described.
FIG. 9 is a flowchart showing the speed estimation method according to the first embodiment, illustrating a series of operations performed by the speed estimation device 2.
The signal receiving units 4-1, 4-2, and 4-3 perform AD conversion processing on the signals received from satellites #1, #2, and #3 by the earth station antennas #1A, #2A, and #3A, and then acquire complex baseband signals in digital format (step ST1A).
Specifically, the signal receivers 4-1, 4-2, and 4-3 first perform various signal processing, such as amplification, band-pass filtering, and frequency conversion, on the RF (radio frequency) outputs of the ground station antennas #1A, #2A, and #3 to generate analog signals. These analog signals are complex signals having in-phase and quadrature components. The signal receivers 4-1, 4-2, and 4-3 then convert these analog signals into received signals, which are digital complex signals, to obtain complex signal vectors.
Signals indicating the complex signal vectors acquired by the signal receiving units 4-1, 4-2, and 4-3 are output to correlation processing units 5-1, 5-2, and 5-3, and then output to reception SNR estimating units 10-1, 10-2, and 10-3.

Next, the reception SNR estimation units 10-1, 10-2, and 10-3 estimate the SNRs of the reception signals acquired by the signal reception units 4-1, 4-2, and 4-3 (step ST2A).
For example, the reception SNR estimation units 10-1, 10-2, and 10-3 perform a Fourier transform on the complex signal vectors received by the signal reception units 4-1, 4-2, and 4-3, and estimate the SNRs of the received signals based on the difference between the peak value of the frequency spectrum and the noise floor, etc. The SNRs estimated by the reception SNR estimation units 10-1, 10-2, and 10-3, respectively, are output to the error ellipse axis length theoretical value calculation unit 14.

The correlation processing units 5-1, 5-2 and 5-3 calculate the information of TDOA and FDOA by correlation processing between the received signals respectively acquired by the signal receiving units 4-1, 4-2 and 4-3, and the orientation point calculation units 6-1, 6-2 and 6-3 calculate the orientation points based on the TDOA and FDOA respectively using the information of TDOA and FDOA (step ST3A). For example, the correlation processing units 5-1, 5-2 and 5-3 calculate the information of TDOA and FDOA by extracting the peak value of CAF as a correlation processing between complex signal vectors.
The orientation point calculation units 6-1, 6-2, and 6-3 use the information on the TDOA and FDOA to solve the simultaneous equations (1) to (7) above, thereby calculating orientation points based on the TDOA and FDOA, respectively.

The accumulation time determination unit 8 accumulates the information on the control points calculated by the control point calculation units 6-1, 6-2, and 6-3, respectively, and the SNRs estimated by the reception SNR estimation units 10-1, 10-2, and 10-3, respectively, in the control point accumulation unit 7 for a fixed time (accumulation time T) (step ST4A). When the elapsed time t reaches the accumulation time T, the accumulation time determination unit 8 outputs the information on the control points and the SNRs held in the control point accumulation unit 7 to the coordinate conversion unit 9. The coordinate conversion unit 9 converts the three-dimensional vector information of each control point output from the control point accumulation unit 7 into latitude and longitude information. The control point information converted by the coordinate conversion unit 9 is output to the frequency distribution calculation unit 11.

The frequency distribution calculation unit 11 calculates the frequency distribution of the control point based on the information on the latitude φ and longitude θ of the control point that has been coordinate-converted by the coordinate conversion unit 9 (step ST5A). The frequency distribution data calculated by the frequency distribution calculation unit 11 is output to the control point extraction unit 12.

The control point extraction unit 12 extracts control points that exist in the area with the highest frequency from the frequency distribution calculated by the frequency distribution calculation unit 11 (step ST6A). For example, the control point extraction unit 12 extracts control points corresponding to the set of satellites #1, #2, and #3, and outputs these together with their average values to the error ellipse calculation unit 13.

The error ellipse calculation unit 13 calculates the center of gravity of the distribution of the control points extracted by the control point extraction unit 12, and calculates three error ellipses, which are areas where these control points may exist, for each pair of satellites #1, #2, and #3 (step ST7A). A signal indicating the error ellipse calculated by the error ellipse calculation unit 13 is output to the FDOA moving variance value calculation unit 15.

The error ellipse axis length theoretical value calculation unit 14 calculates the theoretical value of the axis length of the error ellipse in which the orientation point exists based on the SNR of the received signal (step ST8A). Information indicating the theoretical value of the axis length of the error ellipse calculated by the error ellipse axis length theoretical value calculation unit 14 is output to the FDOA moving variance value calculation unit 15.

The FDOA movement variance calculation unit 15 calculates the variance value of the FDOA between the received signals accompanying the movement of target A based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists (step ST9A).
The target velocity component calculation unit 16 calculates the velocity components of the target A in each direction based on the FDOA variance calculated by the FDOA moving variance calculation unit 15 (step ST10A).
The target classification unit 3 classifies the target A based on the average value of the orientation points obtained in step ST6A and the speed information of the target A calculated using the speed component of the target A obtained in step ST10A.

Next, a hardware configuration for realizing the functions of the speed estimation device 2 will be described.
The functions of the signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimators 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance calculation unit 15, and target speed component calculation unit 16 provided in the speed estimation device 2 are realized by processing circuits. That is, the speed estimation device 2 includes a processing circuit for executing the processes from step ST1A to step ST10A shown in FIG. 9 . The processing circuit may be dedicated hardware, or may be a CPU (Central Processing Unit) that executes a program stored in memory.

Figure 10A is a block diagram showing the hardware configuration that realizes the functions of the speed estimation device 2. Figure 10B is a block diagram showing the hardware configuration that executes software that realizes the functions of the speed estimation device 2. In Figures 10A and 10B, the input interface 100 is an interface that relays signals that the speed estimation device 2 acquires from ground station antennas #1A, #2A, and #3A. The output interface 101 is an interface that relays information indicating the speed components in each direction of target A that is output from the speed estimation device 2 to the target classification unit 3.

When the processing circuit is the dedicated hardware processing circuit 102 shown in Figure 10A, the processing circuit 102 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof. The functions of the signal receiving units 4-1, 4-2 and 4-3, correlation processing units 5-1, 5-2 and 5-3, orientation point calculation units 6-1, 6-2 and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimation units 10-1, 10-2 and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance value calculation unit 15 and target speed component calculation unit 16 provided in the speed estimation device 2 may be realized by separate processing circuits, or these functions may be realized together by a single processing circuit.

When the processing circuit is the processor 103 shown in FIG. 10B, the functions of the signal receivers 4-1, 4-2, and 4-3, correlation processors 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimation units 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, theoretical error ellipse axis length calculation unit 14, FDOA moving variance calculation unit 15, and target speed component calculation unit 16 provided in the speed estimation device 2 are realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in memory 104.

By reading and executing the programs stored in memory 104, processor 103 realizes the functions of signal receivers 4-1, 4-2, and 4-3, correlation processors 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimation units 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, theoretical error ellipse axis length calculation unit 14, FDOA moving variance calculation unit 15, and target speed component calculation unit 16, which are provided in speed estimation device 2. For example, speed estimation device 2 includes memory 104 for storing a program that, when executed by processor 103, results in the processing of steps ST1A to ST10A shown in FIG. 9 being performed. These programs cause a computer to execute the processing procedures or methods performed by the signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimation units 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance value calculation unit 15, and target velocity component calculation unit 16. The memory 104 may be a computer-readable storage medium that stores a program for causing the computer to function as signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point accumulation unit 7, accumulation time determination unit 8, coordinate conversion unit 9, reception SNR estimation units 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance value calculation unit 15, and target velocity component calculation unit 16.

Memory 104 may be, for example, non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically-EPROM) (registered trademark), magnetic disk, flexible disk, optical disk, compact disk, mini disk, DVD, etc.

Some of the functions of the signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, control point calculation units 6-1, 6-2, and 6-3, control point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimation units 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, control point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance value calculation unit 15, and target speed component calculation unit 16 provided in the speed estimation device 2 may be realized by dedicated hardware, and other functions may be realized by software or firmware.

For example, the functions of the signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, and reception SNR estimation units 10-1, 10-2, and 10-3 may be realized by the processing circuit 102, which is dedicated hardware, and the functions of the frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance value calculation unit 15, and target velocity component calculation unit 16 may be realized by the processor 103 reading and executing a program stored in memory 104. In this way, the processing circuit can realize the above functions using hardware, software, firmware, or a combination of these.

Note that while Figure 10B shows a configuration using a single processor 103, this is not limited to this. For example, the speed estimation device 2 may be realized using multiple processors that operate in conjunction with each other.

As described above, the speed estimation device 2 according to the first embodiment includes a frequency distribution calculation unit 11 that calculates the frequency distribution of the control points, a control point extraction unit 12 that extracts control points included in an area where the frequency is maximum from the frequency distribution, an error ellipse calculation unit 13 that calculates, for each pair of satellites, an error ellipse which is an area in which the control point extracted from the frequency distribution may exist, an error ellipse axis length theoretical value calculation unit 14 that calculates a theoretical value of the axis length of the error ellipse in which the control point exists based on the SNR of the received signal, an FDOA moving variance value calculation unit 15 that calculates a variance value of the FDOA between received signals accompanying the movement of target A based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the control point extracted from the frequency distribution exists, and a target velocity component calculation unit 16 that calculates a velocity component of the target based on the variance value.
Based on the difference between the theoretical value of the axis length of the error ellipse in which the orientation point exists, estimated based on the SNR of the received signal, and the axis length of the error ellipse in which the orientation point exists, extracted from the frequency distribution of the orientation point, the variance value of the FDOA between the received signals accompanying the movement of target A is calculated, and the speed component of target A is calculated based on this variance value. In this way, the speed estimation device 2 can estimate speed information of target A to be used for classifying target A.

The target classification device 1 according to embodiment 1 includes a speed estimation device 2 and a target classification unit 3 that classifies target A based on the speed information of target A estimated from the speed component of target A calculated by the target speed component calculation unit 16 and the position information of target A. This allows the target classification device 1 to classify target A using the speed information of target A in addition to the position information of target A.

The speed estimation method according to the first embodiment includes step ST5A in which the frequency distribution calculation unit 11 calculates the frequency distribution; step ST6A in which the control point extraction unit 12 extracts control points that are included in the region with the highest frequency from the frequency distribution; step ST7A in which the error ellipse calculation unit 13 calculates an error ellipse, which is a region in which the control points extracted from the frequency distribution may exist, for each pair of satellites; step ST8A in which the error ellipse axis length theoretical value calculation unit 14 calculates the theoretical value of the axis length of the error ellipse in which the control points exist, based on the signal-to-noise ratio of the received signal; step ST9A in which the FDOA moving variance value calculation unit 15 calculates the variance of the FDOA based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the control points extracted from the frequency distribution exist; and step ST10A in which the target velocity component calculation unit 16 calculates the velocity component of target A based on the variance of the FDOA. By executing this method, the speed estimation device 2 can estimate the speed information of the target A to be used for categorizing the target A.

The target classification method according to the first embodiment includes step ST1, in which the speed estimation device 2 calculates the speed components of target A, and steps ST2 to ST12, in which the target classification unit 3 classifies target A based on the speed information of target A estimated from the speed components of target A and the position information of target A. By executing this method, the target classification device 1 can classify target A using the speed information of target A in addition to the position information of target A.

Embodiment 2.
Fig. 11 is a block diagram showing an example of the configuration of a target classifying device 1A according to embodiment 2. In Fig. 11, the target classifying device 1A is connected to ground station antennas #1A, #2A, and #3A, which receive radio waves from satellites #1, #2, and #3, respectively, in the same way as the target classifying device 1. In addition to the target classifying device 1A, Fig. 11 also shows target A, which is the source of the radio waves and is the target to be classified.

The number of ground station antennas provided corresponds to the number of satellites. In Fig. 11, three satellites #1, #2, and #3 are provided as satellites, and three ground station antennas #1A, #2A, and #3A are provided as ground station antennas.
Satellites #1, #2, and #3 each receive radio waves from target A and transmit the received radio wave signals to ground station antennas #1A, #2A, and #3A, which in turn receive the radio waves that have arrived via satellites #1, #2, and #3, respectively.

The target classification device 1A includes a speed estimation device 2A and a target classification unit 3A. Like the speed estimation device 2, the speed estimation device 2A includes a positioning processing unit that is composed of signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, control point calculation units 6-1, 6-2, and 6-3, control point accumulation unit 7, accumulation time determination unit 8, coordinate conversion unit 9, and reception SNR estimation units 10-1, 10-2, and 10-3. Note that the signal receiving units 4-1, 4-2, and 4-3 may be included in the ground station antennas #1A, #2A, and #3A.
The operation of the positioning processing unit is the same as in the first embodiment.

In addition to the positioning processing unit described above, the speed estimation device 2A also includes a frequency distribution calculation unit 11, a control point extraction unit 12, an error ellipse calculation unit 13, a theoretical error ellipse axis length calculation unit 14, an FDOA moving variance calculation unit 15, and a target speed calculation unit 17. These components generate information for estimating the speed of target A. The frequency distribution calculation unit 11, control point extraction unit 12, error ellipse calculation unit 13, theoretical error ellipse axis length calculation unit 14, and FDOA moving variance calculation unit 15 in the speed estimation device 2A function in the same way as in the speed estimation device 2.

In the speed estimation device 2 according to embodiment 1, the target velocity component calculation unit 16 determines the velocity components of target A in each direction, and then the target classification unit 3 calculates the speed of target A. This has the advantage of providing three-dimensional motion information for target A. However, depending on the positional relationship between the satellite and target A, the matrix M expressed by the above equation (49) may become singular, which may result in an error in the calculated velocity vector and therefore in an error in the speed information used for target classification.

In contrast, in the speed estimation device 2A, the target speed calculation unit 17 calculates multiple speeds of target A based on the FDOA variance value calculated by the FDOA moving variance value calculation unit 15, and calculates the arithmetic mean of the multiple speeds of target A as the speed of target A. This enables the speed estimation device 2A to calculate the speed of target A more robustly.

Specifically, the target speed calculation unit 17 obtains speed information of target A using each of the above equations (43), (44), and (45), and sets the average value of these to be the speed of target A. When target A is moving along the ground surface, it is considered that the vertical velocity component _vU of target A is sufficiently small compared with the northward velocity component _vN and the eastward velocity component _vE . In this case, the above equations (43), (44), and (45) can be transformed into the following equations (52), (53), and (54).

Furthermore, assuming that there is no significant difference between the northward velocity component _vN and the eastward velocity component _vE of target A, and assuming that _vN ≈ _vE = vbar, the above equations (52), (53), and (54) can be simplified to the following equations (55), (56), and (57).

By independently using the above equations (55), (56), and (57), the speed v _est bar of the target A is expressed by the following equations (58), (59), and (60).
The target speed calculation unit 17 calculates three speeds v est bar of target A according to the following equations (58), (59) and (60) by using the variance values σ ² _F-move-12 , σ ² _F-move-23 and σ ² _F-move-13 of the _FDOA between the received signals accompanying the movement of target A.

The target speed calculation unit 17 calculates the arithmetic mean of the multiple speeds of the target A. For example, the target speed calculation unit 17 calculates the arithmetic mean of the three speeds v _est bar of the target A obtained from the above equations (58), (59), and (60) according to the following equation (61), as the final speed v _est bar of the target A.

By finding the speed of target A from the above equation (61), the noise component can be reduced due to the averaging effect, and the inverse matrix calculation shown in the above equation (49) becomes unnecessary.
This makes it possible to avoid cases where the matrix M becomes singular depending on the positional relationship between the target A and the satellite, making it difficult to perform the inverse matrix calculation. This allows the speed estimation device 2A to more robustly calculate the speed of the target A.

Next, a speed estimation method according to the second embodiment will be described.
12 is a flowchart showing a speed estimation method according to embodiment 2, illustrating a series of operations performed by speed estimation device 2 A. The processes from step ST1 B to step ST9 B in FIG. 12 are the same as the processes from step ST1 A to step ST9 A in FIG. 9.

The target speed calculation unit 17 calculates multiple speeds of target A based on the FDOA variance value calculated by the FDOA moving variance value calculation unit 15, and calculates the arithmetic mean of the multiple speeds of target A as the speed of target A (step ST10B).
Next, the target classification unit 3A classifies the target A based on the average value p _est bar of the orientation points calculated in step ST6B and the velocity information v _est bar of the target A calculated in step ST10B. For example, the target classification unit 3A classifies the target A by executing the target classification method shown in FIG.

Next, a hardware configuration for realizing the functions of the speed estimation device 2A will be described.
The functions of the signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimators 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance calculation unit 15, and target speed calculation unit 17 provided in the speed estimation device 2A are realized by processing circuits. That is, the speed estimation device 2A includes a processing circuit for executing the processes from step ST1B to step ST10B shown in FIG. 12. The processing circuit may be dedicated hardware, or may be a CPU that executes a program stored in memory.

Note that while Figure 10B shows a configuration using a single processor 103, this is not limited to this. For example, the speed estimation device 2A may be realized using multiple processors that operate in conjunction with each other.

As described above, the speed estimation device 2A according to embodiment 2 comprises a frequency distribution calculation unit 11 that calculates the frequency distribution of the control points; a control point extraction unit 12 that extracts control points included in the region with the highest frequency from the frequency distribution; an error ellipse calculation unit 13 that calculates an error ellipse for each pair of satellites, which is the region in which the control points extracted from the frequency distribution may exist; an error ellipse axis length theoretical value calculation unit 14 that calculates the theoretical value of the axis length of the error ellipse in which the control points exist based on the SNR of the received signals; an FDOA movement variance calculation unit 15 that calculates the variance value of the FDOA between received signals associated with the movement of target A based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the control points extracted from the frequency distribution exist; and a target speed calculation unit 17 that calculates multiple speeds of target A based on the variance values and calculates the arithmetic average of the multiple speeds of target A as the speed of target A. The speed estimation device 2A calculates a variance value of the FDOA between received signals accompanying the movement of target A based on the difference between the theoretical value of the axis length of the error ellipse in which the orientation point exists, estimated based on the SNR of the received signal, and the axis length of the error ellipse in which the orientation point exists, extracted from the frequency distribution of the orientation point, and calculates the speed of target A based on this variance value. This allows the speed estimation device 2A to estimate speed information of target A to be used for classifying target A. Furthermore, the speed estimation device 2A can calculate the speed of target A more robustly than the speed estimation device 2.

The target classification device 1A according to embodiment 2 includes a speed estimation device 2A and a target classification unit 3A that classifies target A based on the speed information of target A estimated from the velocity components of target A calculated by the target speed calculation unit 17, and the position information of target A. This allows the target classification device 1A to classify target A using the speed information of target A in addition to the position information of target A. Furthermore, because the speed estimation device 2A can calculate the speed of target A more robustly, the target classification device 1A is ultimately able to stably output classification results for target A.

The speed estimation method according to the second embodiment includes step ST5B in which the frequency distribution calculation unit 11 calculates the frequency distribution, step ST6B in which the control point extraction unit 12 extracts control points included in the area with the highest frequency from the frequency distribution, step ST7B in which the error ellipse calculation unit 13 calculates an error ellipse for each pair of satellites, which is an area where the control points extracted from the frequency distribution may exist, and step ST7B in which the error ellipse theoretical axis length calculation unit 14 calculates an area where the control points may exist based on the signal-to-noise ratio of the received signal. The method includes step ST8B in which the FDOA moving variance calculation unit 15 calculates a theoretical value of the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists, step ST9B in which the FDOA moving variance calculation unit 15 calculates a variance value of the FDOA based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists, and step ST10B in which the target speed calculation unit 17 calculates a plurality of speeds of the target A based on the variance value of the FDOA and calculates the arithmetic mean of the plurality of velocities of the target A as the speed of the target A. By executing this method, the speed estimation device 2A can estimate speed information of the target A to be used for classifying the target A. Furthermore, the speed estimation device 2A can calculate the speed of the target A more robustly than the speed estimation device 2.

The target classification method according to embodiment 2 includes step ST1, in which the speed estimation device 2A calculates speed information for target A, and steps ST2 to ST12, in which the target classification unit 3A classifies target A based on the speed information and position information for target A. By executing this method, the target classification device 1A can classify target A using the speed information for target A in addition to the position information for target A. Furthermore, because the speed estimation device 2A can calculate the speed of target A more robustly, the target classification device 1A can ultimately output the classification results for target A stably.

Embodiment 3.
Fig. 13 is a block diagram showing an example of the configuration of a target classifying device 1B according to embodiment 3. In Fig. 11, target classifying device 1B is connected to ground station antennas #1A, #2A, and #3A, respectively, which receive radio waves from satellites #1, #2, and #3, similar to target classifying device 1. In addition to target classifying device 1B, Fig. 13 also shows target A, the source of the radio waves, as the target to be classified.

The number of ground station antennas provided corresponds to the number of satellites. In Fig. 13, three satellites #1, #2, and #3 are provided as satellites, and three ground station antennas #1A, #2A, and #3A are provided as ground station antennas.
Satellites #1, #2, and #3 each receive radio waves from target A and transmit the received radio wave signals to ground station antennas #1A, #2A, and #3A, which in turn receive the radio waves that have arrived via satellites #1, #2, and #3, respectively.

The target classification device 1B includes a speed estimation device 2B and a target classification unit 3B. Like the speed estimation device 2, the speed estimation device 2B includes a positioning processing unit consisting of signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, control point calculation units 6-1, 6-2, and 6-3, control point accumulation unit 7, accumulation time determination unit 8, coordinate conversion unit 9, and reception SNR estimation units 10-1, 10-2, and 10-3. Note that the signal receiving units 4-1, 4-2, and 4-3 may be included in the ground station antennas #1A, #2A, and #3A.
The operation of the positioning processing unit is the same as in the first embodiment.

In addition to the positioning processing unit described above, the speed estimation device 2B also includes a frequency distribution calculation unit 11, a control point extraction unit 12, an error ellipse calculation unit 13, a theoretical error ellipse axis length calculation unit 14, an FDOA moving variance calculation unit 15, and a target plane direction velocity component calculation unit 18. These components generate information for estimating the speed of target A. The frequency distribution calculation unit 11, control point extraction unit 12, error ellipse calculation unit 13, theoretical error ellipse axis length calculation unit 14, and FDOA moving variance calculation unit 15 in the speed estimation device 2B function in the same way as in the speed estimation device 2.

In the speed estimation device 2B, a target planar direction velocity component calculation unit 18 calculates the planar direction velocity components of target A based on the FDOA variance value calculated by the FDOA moving variance value calculation unit 15. The planar direction velocity components of target A are velocity components in the direction of the earth's surface plane, for example, a northward velocity component _vN and an eastward velocity component _vE . Note that the planar direction velocity components of target A may also be southward velocity components and westward velocity components. By using the planar direction velocity components of target A, the speed estimation device 2B can more robustly calculate the speed of target A while retaining information on the velocity vector of target A.

The target plane direction velocity component calculation unit 18 calculates _only the north direction velocity component _vN and the east direction velocity component _vE under the assumption that the vertical direction velocity component vU is sufficiently smaller than the north direction velocity component _vN and the east direction velocity component _vE . Specifically, by using the above equations (43), (44), and (45), under the condition that the vertical direction velocity component _vU ≈ 0, the north direction velocity component _vN and the east direction velocity component _vE have three relationships shown in the following equations (62), (63), and (64).

By using the above equations (62), (63), and (64), the northward velocity component _vN and the eastward velocity component _vE can be expressed as arithmetic means shown in the following equations (65) and (66). The target plane direction velocity component calculation unit 18 calculates the northward velocity component _vN and the eastward velocity component _vE in accordance with the following equations (65) and (66).

Depending on the positional relationship between the satellite and target A, the denominator terms in the above equations (55), (56), and (57) become extremely small. In other words, when the conditions shown in the following equations (67), (68), and (69) hold for a certain minute constant ε, the term is omitted and calculation averaging is performed.
This makes it possible to exclude singular values from the calculated v _N ² and v _E ² , and therefore the speed estimation device 2B can calculate the speed of target A more robustly than the speed estimation device 2. Furthermore, the speed estimation device 2B can obtain more information regarding the movement of target A than the speed estimation device 2A.

Next, a speed estimation method according to the third embodiment will be described.
14 is a flowchart showing a speed estimation method according to embodiment 3, illustrating a series of operations performed by the speed estimation device 2 B. The processes from step ST1 C to step ST9 C in FIG. 14 are similar to the processes from step ST1 A to step ST9 A in FIG. 9.

The target planar direction velocity component calculation unit 18 calculates the planar direction velocity component of the target A based on the FDOA variance value calculated by the FDOA moving variance value calculation unit 15 (step ST10C).
The target classification unit 3B classifies the target A based on the average value p _estbar of the control points calculated in step ST6C and the north velocity component _vN and the east velocity component _vE , which are the planar velocity components of the target A calculated in step ST10C. For example, the target classification unit 3B classifies the target A by executing the target classification method shown in FIG.

Next, a hardware configuration for realizing the functions of the speed estimation device 2B will be described.
The functions of the signal receiving units 4-1, 4-2, and 4-3, correlation processing units 5-1, 5-2, and 5-3, orientation point calculation units 6-1, 6-2, and 6-3, orientation point storage unit 7, storage time determination unit 8, coordinate conversion unit 9, reception SNR estimators 10-1, 10-2, and 10-3, frequency distribution calculation unit 11, orientation point extraction unit 12, error ellipse calculation unit 13, error ellipse axis length theoretical value calculation unit 14, FDOA moving variance calculation unit 15, and target plane direction velocity component calculation unit 18 provided in the speed estimation device 2B are realized by processing circuits. That is, the speed estimation device 2B includes a processing circuit for executing the processes from step ST1C to step ST10C shown in FIG. 14. The processing circuit may be dedicated hardware, or may be a CPU that executes a program stored in memory.

Note that while Figure 10B shows a configuration using a single processor 103, this is not limited to this. For example, the speed estimation device 2B may be realized using multiple processors that operate in conjunction with each other.

As described above, the speed estimation device 2B according to the third embodiment includes a frequency distribution calculation unit 11 that calculates the frequency distribution of the control points, a control point extraction unit 12 that extracts control points included in an area where the frequency is maximum from the frequency distribution, an error ellipse calculation unit 13 that calculates, for each pair of satellites, an error ellipse which is an area in which the control point extracted from the frequency distribution may exist, an error ellipse axis length theoretical value calculation unit 14 that calculates a theoretical value of the axis length of the error ellipse in which the control point exists based on the SNR of the received signal, an FDOA movement variance value calculation unit 15 that calculates a variance value of the FDOA between received signals accompanying the movement of the target A based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the control point extracted from the frequency distribution exists, and a target planar direction velocity component calculation unit 18 that calculates a velocity component in the planar direction of the target A based on the variance value.
The speed estimation device 2B calculates the FDOA variance value between the received signals accompanying the movement of the target A based on the difference between the theoretical value of the axis length of the error ellipse in which the orientation point exists, estimated based on the SNR of the received signal, and the axis length of the error ellipse in which the orientation point exists, extracted from the frequency distribution of the orientation point, and calculates the velocity component of the target A in the planar direction based on the FDOA variance value. This allows the speed estimation device 2B to estimate the speed information of the target A to be used for classifying the target A.
Furthermore, the speed estimation device 2B can robustly calculate the magnitude of the velocity components of the target A in the north-south and east-west directions with respect to the positional relationship between the satellite and the target A. In other words, the speed estimation device 2B can estimate the velocity of the target A more robustly than the speed estimation device 2, and can obtain more information regarding the movement of the target A than the speed estimation device 2A.

The target classification device 1B according to the third embodiment includes a speed estimation device 2B and a target classification unit 3 that classifies target A based on the planar direction velocity component information of target A calculated by a target planar direction velocity component calculation unit 18 and the position information of target A. As a result, the target classification device 1B can classify target A using the speed information of target A in addition to the position information of target A.
Furthermore, since the speed estimation device 2B can calculate the speed of the target A more robustly, the target classification device 1B can consequently output the classification result of the target A stably.

The speed estimation method according to the third embodiment includes step ST5C in which the frequency distribution calculation unit 11 calculates the frequency distribution, step ST6C in which the control point extraction unit 12 extracts, from the frequency distribution, a control point included in an area with the maximum frequency, step ST6C in which the error ellipse calculation unit 13 calculates, for each pair of satellites, an error ellipse which is an area in which the control point extracted from the frequency distribution may exist, step ST7C in which the error ellipse axis length theoretical value calculation unit 14 calculates, based on the signal-to-noise ratio of the received signal, a theoretical value of the axis length of the error ellipse in which the control point exists, step ST8C in which the FDOA moving variance value calculation unit 15 calculates a variance value of the FDOA based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the control point extracted from the frequency distribution exists, and step ST10C in which the target planar direction velocity component calculation unit 18 calculates a velocity component in the planar direction of the target A based on the variance value of the FDOA.
By executing this method, the speed estimation device 2B can estimate speed information of the target A to be used for classifying the target A. Furthermore, the speed estimation device 2B can estimate the speed of the target A more robustly than the speed estimation device 2, and can obtain more information regarding the movement of the target A than the speed estimation device 2A.

The target classification method according to the third embodiment includes step ST1 in which the speed estimation device 2B calculates speed component information of the target A in a planar direction, and steps ST2 to ST12 in which the target classification unit 3B classifies the target A based on the speed component information of the target A in the planar direction and position information of the target A. By executing this method, the target classification device 1B can classify the target A using the speed information of the target A in addition to the position information of the target A.
Furthermore, since the speed estimation device 2B can calculate the speed of the target A more robustly, the target classification device 1B can consequently output the classification result of the target A stably.

In addition, it is possible to combine each embodiment, modify any component of each embodiment, or omit any component of each embodiment.

The speed estimation device disclosed herein can be used, for example, in surveillance systems using satellite constellations.

1, 1A, 1B Target classification device, 2, 2A, 2B Speed estimation device, 3, 3A, 3B Target classification unit, 4-1 to 4-3 Signal receiving unit, 5-1 to 5-3 Correlation processing unit, 6-1 to 6-3 Control point calculation unit, 7 Control point accumulation unit, 8 Accumulation time determination unit, 9 Coordinate conversion unit, 10-1 to 10-3 Received SNR estimation unit, 11 Frequency distribution calculation unit, 12 Control point extraction unit, 13 Error ellipse calculation unit, 14 Error ellipse axis length theoretical value calculation unit, 15 FDOA moving variance value calculation unit, 16 Target velocity component calculation unit, 17 Target speed calculation unit, 18 Target planar direction velocity component calculation unit, 100 Input interface, 101 Output interface, 102 Processing circuit, 103 Processor, 104 Memory.

Claims (12)

A speed estimation device that receives radio waves transmitted from a target that is a radio wave source and arrives via a plurality of satellites, and sequentially acquires a control point that is an estimated position of the target, the control point being estimated based on the arrival time difference and arrival frequency difference between the received signals obtained by correlation processing between the received signals,
a frequency distribution calculation unit that calculates a frequency distribution of the control points that have been converted into latitude and longitude coordinates;
a control point extraction unit that extracts a control point that exists in an area where the frequency is maximum from the frequency distribution;
an error ellipse calculation unit that calculates an error ellipse, which is an area where the orientation point extracted from the frequency distribution may exist, for each pair of satellites;
a theoretical value calculation unit that calculates a theoretical value of the axis length of an error ellipse in which the orientation point exists based on the signal-to-noise ratio of the received signal;
a variance value calculation unit that calculates a variance value of the arrival frequency difference between the received signals accompanying movement of the target based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists;
a target velocity component calculation unit that calculates a velocity component of the target based on the variance value. A speed estimation device that receives radio waves transmitted from a target that is a radio wave source and arrives via a plurality of satellites, and sequentially acquires a control point that is an estimated position of the target, the control point being estimated based on the arrival time difference and arrival frequency difference between the received signals obtained by correlation processing between the received signals,
a frequency distribution calculation unit that calculates a frequency distribution of the control points that have been converted into latitude and longitude coordinates;
a control point extraction unit that extracts a control point that exists in an area where the frequency is maximum from the frequency distribution;
an error ellipse calculation unit that calculates an error ellipse, which is an area where the orientation point extracted from the frequency distribution may exist, for each pair of satellites;
a theoretical value calculation unit that calculates a theoretical value of the axis length of an error ellipse in which the orientation point exists based on the signal-to-noise ratio of the received signal;
a variance value calculation unit that calculates a variance value of the arrival frequency difference between the received signals accompanying movement of the target based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists;
a target speed calculation unit that calculates a plurality of speeds of the target based on the variance value and calculates an arithmetic average of the plurality of speeds of the target as the speed of the target. A speed estimation device that receives radio waves transmitted from a target that is a radio wave source and arrives via a plurality of satellites, and sequentially acquires a control point that is an estimated position of the target, the control point being estimated based on the arrival time difference and arrival frequency difference between the received signals obtained by correlation processing between the received signals,
a frequency distribution calculation unit that calculates a frequency distribution of the control points that have been converted into latitude and longitude coordinates;
a control point extraction unit that extracts a control point that exists in an area where the frequency is maximum from the frequency distribution;
an error ellipse calculation unit that calculates an error ellipse, which is an area where the orientation point extracted from the frequency distribution may exist, for each pair of satellites;
a theoretical value calculation unit that calculates a theoretical value of the axis length of an error ellipse in which the orientation point exists based on the signal-to-noise ratio of the received signal;
a variance value calculation unit that calculates a variance value of the arrival frequency difference between the received signals accompanying movement of the target based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists;
a target planar direction velocity component calculation unit that calculates a velocity component in a planar direction of the target based on the variance value. The speed estimation device according to claim 1;
a target classification unit that classifies the target based on speed information of the target estimated from the velocity component of the target calculated by the target velocity component calculation unit and position information of the target. The speed estimation device according to claim 2;
a target classification unit that classifies the target based on the speed information of the target calculated by the target speed calculation unit and position information of the target. A speed estimation device according to claim 3;
a target classification unit that classifies the target based on the velocity component information in the planar direction of the target calculated by the target planar direction velocity component calculation unit and position information of the target. A speed estimation method executed by the speed estimation device according to claim 1, comprising:
a step in which the frequency distribution calculation unit calculates the frequency distribution;
a step in which the orientation point extraction unit extracts orientation points that exist in an area with the maximum frequency from the frequency distribution;
a step in which the error ellipse calculation unit calculates an error ellipse, which is an area where the orientation point extracted from the frequency distribution may exist, for each pair of satellites;
The theoretical value calculation unit calculates a theoretical value of an axis length of an error ellipse in which the orientation point exists based on the signal-to-noise ratio of the received signal;
The variance value calculation unit calculates the variance value based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists;
a step in which the target velocity component calculation unit calculates a velocity component of the target based on the variance value. A speed estimation method executed by the speed estimation device according to claim 2, comprising:
a step in which the frequency distribution calculation unit calculates the frequency distribution;
a step in which the orientation point extraction unit extracts orientation points that exist in an area with the maximum frequency from the frequency distribution;
a step in which the error ellipse calculation unit calculates an error ellipse, which is an area where the orientation point extracted from the frequency distribution may exist, for each pair of satellites;
The theoretical value calculation unit calculates a theoretical value of an axis length of an error ellipse in which the orientation point exists based on the signal-to-noise ratio of the received signal;
The variance value calculation unit calculates the variance value based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists;
the target speed calculation unit calculates a plurality of speeds of the target based on the variance value, and calculates an arithmetic average of the plurality of speeds of the target as the speed of the target. A speed estimation method executed by the speed estimation device according to claim 3, comprising:
a step in which the frequency distribution calculation unit calculates the frequency distribution;
a step in which the orientation point extraction unit extracts orientation points that exist in an area with the maximum frequency from the frequency distribution;
a step in which the error ellipse calculation unit calculates an error ellipse, which is an area where the orientation point extracted from the frequency distribution may exist, for each pair of satellites;
The theoretical value calculation unit calculates a theoretical value of an axis length of an error ellipse in which the orientation point exists based on the signal-to-noise ratio of the received signal;
The variance value calculation unit calculates the variance value based on the difference between the theoretical value of the axis length of the error ellipse and the axis length of the error ellipse in which the orientation point extracted from the frequency distribution exists;
a step in which the target planar direction velocity component calculation unit calculates a planar direction velocity component of the target based on the variance value. A target classification method performed by the target classification device of claim 4, comprising:
a step of calculating a velocity component of the target by the velocity estimation device according to claim 1;
a step in which the target classification unit classifies the target based on speed information of the target estimated from the velocity component of the target and position information of the target. A target classification method performed by the target classification device of claim 5, comprising:
a step of calculating speed information of the target by the speed estimation device according to claim 2;
a step in which the target classification unit classifies the target based on speed information of the target and position information of the target. A target classification method performed by the target classification device of claim 6, comprising:
a step of calculating velocity component information in a planar direction of the target by the velocity estimation device according to claim 3;
a step in which the target classification unit classifies the target based on velocity component information in a planar direction of the target and position information of the target.