JP2006208172A - Arrival time estimating device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To estimate arrival time of radio waves, even if the frequencies of both received signals and reference signals are not in agreement. <P>SOLUTION: In the arrival time estimating device arranging a received signal covariance matrix computing section 2 generating a covariance matrix from temporal waveform of the signals received at a received signal sampling section 1 and a mode vector computing section 3 computing a mode vector which is a function of both arrival time and frequency shift from temporal waveform of the signals transmitted periodically to the space, using a covariance matrix computed at the a mode vector computed at the received signal covariance matrix computing section 2 and a mode vector computing section 3, a MUSIC processing is implemented to estimate the arrival time of signals received at the received signal sampling section 1. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an arrival time estimation device that estimates an arrival time of a signal indicating a distance to a target existing in space, for example.

The arrival time estimation device is mounted on, for example, a radar or a GPS receiver, and estimates the arrival time estimation of radio waves when detecting a distance to a target existing in space, for example.
Conventional arrival time estimation devices estimate the arrival time estimate of a received signal by performing a correlation operation between the transmitted signal (or a reference signal correlated with the transmitted signal) and a received signal to obtain a delay profile. This method is generally adopted.

However, the method of estimating the arrival time by performing the correlation calculation has a problem that the time resolution is limited to the reciprocal of the bandwidth of the transmission / reception signal.
For this reason, when the radar employs a method of estimating the arrival time by performing the correlation calculation, the distance resolution is limited by the bandwidth of the transmission / reception signal. In addition, if there are a plurality of incoming waves that are closer than the time resolution, an error occurs in the estimation of the arrival time.
Therefore, in order to solve the problem that the distance resolution is limited by the bandwidth of the transmission / reception signal, the time is higher than the case where the super-resolution processing such as the MUSIC processing or the ESPRIT processing is performed and the correlation calculation is performed. A method for obtaining resolution has been proposed (see, for example, Patent Document 1).
The details of the method for estimating the arrival time of radio waves by performing super-resolution processing are disclosed in Non-Patent Document 1 below.

However, in a method for estimating the arrival time of radio waves by performing super-resolution processing such as MUSIC processing or ESPRIT processing, the frequency of the received signal and the reference signal must match. For example, in a GPS receiver, It cannot be applied when the Doppler frequency due to the movement of the satellite is unknown.
In addition, in the radio wave source position measurement device, the radio wave source and the receiver move, and the Doppler frequency is not known.
Moreover, in a radar, when a detection target moves and the Doppler frequency is unknown, it cannot apply.
Hereinafter, in order to explain the reason why it is not possible to apply when the Doppler frequency is unknown, an arrival time estimation device that employs a method of estimating the arrival time of radio waves by performing super-resolution processing will be described. To do.

First, the first Fourier transform means Fourier transforms the received signal, and the second Fourier transform means Fourier transforms the reference signal.
If the received signal is a signal obtained by delaying the reference signal by time τ, the Fourier-transformed received signal [X ₁ , X ₂ ,..., X _L ] and the Fourier-transformed reference signal [S ₁ , The relationship between S ₂ ,..., S _L ] is as follows.
[X ₁ , X ₂ ,..., X _L ]
= [S ₁ e ⁰ , S ₂ e ^−2πjfsτ ,..., S _L ^{−2πj (L−1) fsτ} ] (1)
Here, L is the number of samples of the received signal, and fs is the sampling frequency.

Next, the division means divides the received signal subjected to the Fourier transform by the reference signal for each element to obtain the following phase linearity.
[X ₁ / S ₁ , X ₂ / S ₂ ,..., X _L / S _L ]
= [E ⁰ , e ^−2πjfsτ ,..., E ^{−2πj (L−1) fsτ} ] (2)
Next, the covariance matrix calculation means calculates the covariance matrix of the division result of the division means, that is, the covariance matrix of Expression (2).
Next, the MUSIC processing means sets the geometric series as shown in the following formula (3) as the mode vector a _conv (τ), and the mode vector a _conv (τ) and the covariance matrix calculation means calculated by the covariance matrix calculating means. The delay time τ is estimated by performing the MUSIC process using the variance matrix.
a _conv (τ) = [e ⁰ , e ^−2πjfsτ ,..., e ^{−2πj (L−1) fsτ} ] ^T (3)

Here, consider a case where the frequency of the received signal is shifted with respect to the reference signal.
That is, the Fourier transformed received signal _{[X 1, X 2, ···} , X L] and the reference signal subjected to Fourier transform _{[S 1, S 2, ···} , S L] relationship of the following equation Assume that this is the case.
[X ₁ , X ₂ ,..., X _L ]
= [S _1-n e ⁰ , S _2-n e ^−2πjfsτ ,..., S _Ln ^{−2πj (L−1) fsτ} ] (4)
Here, n is the number of indexes for frequency shift.

When this frequency shift is ignored and the received signal is divided by the reference signal, the following equation is obtained.
[X ₁ / S ₁ , X ₂ / S ₂ ,..., X _L / S _L ]
= [(S _1-n / S ₁ ) e ⁰ , (S _2−n / S ₂ ) e ^{−2πjfsτ,.}
(S _Ln / S _L ) e ^{−2πj (L−1) fsτ} ] (5)
However, generally, since S _l and S _ln are irrelevant, the right side of Equation (5) has a random phase, and phase linearity cannot be obtained.
Therefore, in the conventional arrival time estimation method using the linear phase as the mode vector, the arrival time of the received signal is estimated when the frequency of the received signal is shifted with respect to the reference signal and the shift amount n is unknown. I can't.

Japanese translation of PCT publication No. 2003-532314 (page 15, FIG. 1) Nobuyoshi Kikuma, "Adaptive signal processing by array antenna", Science and Technology Publication, 1998, P269-P292

  Since the conventional arrival time estimation apparatus is configured as described above, the arrival time of radio waves can be estimated if the frequency of the received signal and the reference signal match. However, if the frequency of the received signal does not match the frequency of the reference signal, the arrival time of the radio wave cannot be estimated. For example, it can be applied to a GPS receiver or the like whose Doppler frequency due to satellite movement is unknown. There were issues such as being unable to do so.

  The present invention has been made to solve the above-described problems, and provides an arrival time estimation device that can estimate the arrival time of radio waves even when the frequency of the received signal and the reference signal do not match. With the goal.

  An arrival time estimation apparatus according to the present invention includes a covariance matrix generation unit that generates a covariance matrix from a time waveform of a signal received by a signal reception unit, and an arrival time from a time waveform of a signal periodically transmitted to space. And a mode vector generating means for generating a mode vector that is a function of the frequency shift amount, and using the covariance matrix generated by the covariance matrix generating means and the mode vector generated by the mode vector generating means, the super-resolution By executing the processing, the arrival time of the signal received by the signal receiving means is estimated.

  According to this invention, the covariance matrix generating means for generating the covariance matrix from the time waveform of the signal received by the signal receiving means, and the arrival time and the frequency shift amount from the time waveform of the signal periodically transmitted to the space Mode vector generating means for generating a mode vector that is a function of the above, and performing super-resolution processing using the covariance matrix generated by the covariance matrix generating means and the mode vector generated by the mode vector generating means As a result, the arrival time of the signal received by the signal receiving means is estimated, so that the frequency of the signal received by the signal receiving means and the signal periodically transmitted to the space does not match. Even in this case, there is an effect that the arrival time of the radio wave can be estimated.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing an arrival time estimation apparatus according to Embodiment 1 of the present invention. In FIG. 1, a received signal sampling unit 1 is transmitted to a space when a signal transmitter (not shown) periodically transmits the signal to the space. The periodically transmitted signals are sampled sequentially. The received signal sampling unit 1 constitutes a signal receiving unit.
The reception signal covariance matrix calculation unit 2 calculates a covariance matrix R _xx from one cycle of the signal sequence x (k) that is a time waveform of the signal sampled by the reception signal sampling unit 1. The received signal covariance matrix calculation unit 2 constitutes a covariance matrix generation unit.

The mode vector calculation unit 3 calculates a mode vector a (τ, Δf) that is a function of the arrival time and the frequency shift amount from a signal sequence s (t) that is a time waveform of a signal periodically transmitted to space. To do.
In the first embodiment, a description will be given of a case in which a signal periodically transmitted to a known space in the arrival time estimation apparatus is input to the mode vector calculation unit 3, but a signal periodically transmitted to the space. And a reference signal having a strong correlation with the mode vector calculator 3 may be input. Note that the mode vector calculation unit 3 constitutes a mode vector generation means.

The MUSIC processing unit 4 uses the covariance matrix R _xx calculated by the received signal covariance matrix calculation unit 2 and the mode vector a (τ, Δf) calculated by the mode vector calculation unit 3 to perform super resolution processing. By performing the above, the arrival time τ of the signal sampled by the reception signal sampling unit 1 is estimated.
In the first embodiment, the MUSIC processor 4 performs the MUSIC process that is a super-resolution process. However, the present invention is not limited to this. For example, by performing the ESPRIT process that is a super-resolution process, The arrival time τ of the signal sampled by the reception signal sampling unit 1 may be estimated. The MUSIC processing unit 4 constitutes arrival time estimating means.

FIG. 2 is a block diagram showing the MUSIC processing unit of the arrival time estimation apparatus according to Embodiment 1 of the present invention. In the figure, the eigenvalue / eigenvector calculation unit 11 is the covariance calculated by the received signal covariance matrix calculation unit 2. The eigenvalue λ _l (l = 1, 2,..., L) and the eigenvector e _l (l = 1, 2,..., L) of the matrix R _xx are calculated.
The noise eigenvalue determining unit 12 specifies an eigenvalue equal to the noise power σ ² (hereinafter referred to as a noise eigenvalue) among the eigenvalues λ _l of the covariance matrix R _xx calculated by the eigenvalue / eigenvector calculation unit 11 and also covariance The propagation path number M is estimated by comparing the eigenvalue λ _l of the matrix R _xx with the noise power σ ² .
The MUSIC spectrum calculation unit 13 relates to the arrival time τ and the frequency shift amount Δf from the eigenvector corresponding to the noise eigenvalue specified by the noise eigenvalue determination unit 12 and the mode vector a (τ, Δf) calculated by the mode vector calculation unit 3. The two-dimensional MUSIC spectrum P _mu (τ, Δf) is calculated, and the arrival time τ of the signal sampled by the received signal sampling unit 1 from the peak position of the two-dimensional MUSIC spectrum P _mu (τ, Δf) and the frequency shift amount Δf Is estimated.

Next, the operation will be described.
First, a signal transmitter (not shown) periodically transmits a signal to a space as shown in FIG.
However, it is assumed that this signal is, for example, a pulse signal or a pseudo noise code, and the pulse waveform or modulation code is known.
Such an assumption is common in radars and GPS receivers. Also in a portable terminal or the like, it is common when the communication method is a spread spectrum method. Even if the communication method of the portable terminal or the like is not a spread spectrum method, the arrival time estimation device of the first embodiment can be applied if the communication preamble part is used if the communication method is a normal digital communication method. .

When a signal transmitter (not shown) periodically transmits a signal to the space, the reception signal sampling unit 1 sequentially samples the signals periodically transmitted to the space.
That is, as shown in FIG. 3, the received signal sampling unit 1 receives a predetermined number (for example, L) for each repetition period (hereinafter referred to as “signal period”) of a signal transmitted from the signal transmitter. Sampling the signal.
Hereinafter, a signal sequence of L received signals sampled from the kth signal cycle is referred to as x (k), and this signal sequence x (k) is referred to as a received signal vector.
x (k) = [X ₁ , X ₂ ,..., X _L ] ^T (6)
However, T represents transposition of a matrix or a vector.

The received signal arrives via a plurality of paths and can be considered as a combination of waveforms of transmission signals to which various delays and frequency shifts are given, and therefore can be expressed as the following equation.
x (k) = Σa (τ _m , Δf _m ) c _m (k) + n (k) (7)
However, a (τ _m , Δf _m ) is a signal waveform vector when the transmission signal is delayed by time τ _m and the frequency is shifted by Δf _m .
delay time tau _m is the m-th path, frequency shift amount of Delta] f _m is the m-th path, c _m (k) is the complex amplitude of the m-th path in the k-th signal period, n (k) is k th It is a receiver noise vector in the signal period.

ただし、Ｅ［］は期待値演算を表し、Ｈは行列やベクトルの複素共役転置を表し、＊は複素共役を表している。 When the reception signal sampling unit 1 samples L reception signals for each signal period, the reception signal covariance matrix calculation unit 2 performs a signal sequence of the L reception signals as shown in the following equation (8). A covariance matrix R _xx is calculated from the received signal vector x (k).

However, E [] represents an expected value calculation, H represents a complex conjugate transpose of a matrix or a vector, and * represents a complex conjugate.

ただし、σ²は雑音電力であり、ｎ₁（ｋ）は雑音ベクトルｎ（ｋ）のｌ番目の要素である。なお、雑音電力はｌによらないことを仮定している。 If Expression (7) is substituted into Expression (8) and it is assumed that the transmission signal and noise are uncorrelated, Expression (8) can be expressed as Expression (9) below.

Where σ ² is noise power and n ₁ (k) is the l-th element of the noise vector n (k). It is assumed that the noise power does not depend on l.

As described above, since Expression (8) is expressed in the form of Expression (9), it can be seen that this is a case where the MUSIC algorithm can be applied.
However, in the actual calculation, the received signal covariance matrix calculation unit 2 substitutes the expected value calculation of Equation (8) with a time average calculation of K signal periods as shown below.

The mode vector calculation unit 3 uses the following equation (14) to calculate the arrival time from a signal sequence s (t) of one cycle that is a time waveform of a signal periodically transmitted to a space by a signal transmitter (not shown). A mode vector a (τ, Δf) that is a function of the frequency shift amount is calculated.
That is, a signal vector periodically transmitted to the space by one period is delayed within a predetermined range, and a frequency-shifted mode vector a (τ, Δf) is generated.
a (τ, Δf)
= [S (−τ) e ⁰ , s (T−τ) e ^−2πjΔfT , s (2T−τ) e ^−2πj2ΔfT
,..., S ((L-1) T-.tau.) E- ^{2.pi.j (L-1) .DELTA.fT} ] ^T
(14)
However, T is a sampling period.
Here, the mode vector calculator 3 calculates the mode vector a (τ, Δf) from a signal transmitted from a signal transmitter (not shown), but has a strong correlation with the signal transmitted from the signal transmitter. The mode vector a (τ, Δf) may be calculated from the related reference signal.

When the received signal covariance matrix calculation unit 2 calculates the covariance matrix R _xx and the mode vector calculation unit 3 calculates the mode vector a (τ, Δf), the MUSIC processing unit 4 calculates the covariance matrix R _xx and the mode. By performing MUSIC processing which is super-resolution processing using the vector a (τ, Δf), the arrival time τ of the signal sampled by the received signal sampling unit 1 and the frequency shift amount Δf are estimated, and the number of propagation paths Estimate M.
Details of MUSIC processing (Multiple Signal Classification) are described in detail on pages 191 to 202 and pages 269 to 282 of Nobuyoshi Kikuma "Adaptive signal processing by array antenna" Science and Technology Publication (published in 1998). The processing contents of the MUSIC processing unit 4 will be specifically described.

When the reception signal covariance matrix calculation unit 2 calculates the covariance matrix R _xx , the eigenvalue / eigenvector calculation unit 11 of the MUSIC processing unit 4 calculates the eigenvalue λ _l (l = 1, 2,...) Of the covariance matrix R _xx. .., L) and eigenvectors e _l (l = 1, 2,..., L).
Noise eigenvalue determination portion 12 of the MUSIC processing unit 4, the eigenvalues and eigenvectors calculation unit 11 calculates the eigenvalues lambda _l of the covariance matrix R _xx, as shown below, descending eigenvalues lambda _l of the covariance matrix R _xx , Identify the eigenvalues equal to the noise power σ ² of the received signal (hereinafter referred to as noise eigenvalues), and input the noise eigenvalues λ _{M + 1} , λ _{M + 2} ,..., Λ _L to the MUSIC spectrum calculation unit 13. Output.
λ ₁ ≧ λ ₂ ≧ ・ ・ ・ ≧ λ _M ≧ λ _{M + 1} = ... = λ _L = σ ²
(15)

In addition, the noise eigenvalue determining unit 12 compares the eigenvalue λ _l of the covariance matrix R _xx with the noise power σ ² to estimate the number of propagation paths M.
That is, among the eigenvalues lambda _l of the covariance matrix R _xx, identifies the noise power sigma ² larger eigenvalue lambda _l, estimates the number of the noise power sigma ² larger eigenvalue lambda _l and propagation path number M.
In the example of Expression (15), since λ ₁ , λ ₂ ,..., Λ _M is larger than the noise power σ ² , it is estimated that the number of propagation paths is M.
FIG. 4 is an explanatory diagram showing the eigenvalue λ _l of the covariance matrix R _xx , but in the example of FIG. 4, the number of eigenvalues having a value significantly greater than or equal to 0 is 3, so the number of propagation paths M Is estimated to be 3.

In the MUSIC spectrum calculation unit 13 of the MUSIC processing unit 4, the noise eigenvalue determination unit 12 specifies the noise eigenvalues λ _{M + 1} , λ _{M + 2} ,..., Λ _L , and the mode vector calculation unit 3 uses the mode vector a ( tau, when calculating the Delta] f), as shown in the following equation (16), the noise eigenvalues _{λ M, λ M + 1,} ···, eigenvector corresponding to _{λ L e M + 1, e} M + 2, .., E _L and the mode vector a (τ, Δf), the two-dimensional MUSIC spectrum P _mu (τ, Δf) relating to the arrival time τ and the frequency shift amount Δf is calculated.
P _mu (τ, Δf)
^{= (A H (τ, Δf} ) a (τ, Δf)) / (a H (τ, Δf) E N E H N a (τ, Δf))
(16)
E _N = [e _{M + 1} , e _{M + 2} ,..., E _L ] (17)

When the MUSIC spectrum calculation unit 13 calculates the two-dimensional MUSIC spectrum P _mu (τ, Δf) as described above, the received signal sampling unit 1 determines the peak position of the two-dimensional MUSIC spectrum P _mu (τ, Δf). The arrival time τ and the frequency shift amount Δf of the sampled signal are estimated.
FIG. 5 is an explanatory view showing an example of the calculation result of the two-dimensional MUSIC spectrum P _mu (τ, Δf). FIG. 5 shows an example in which the number M of propagation paths is 3, so that three peaks appear. Yes.

  As apparent from the above, according to the first embodiment, the received signal covariance matrix calculating unit 2 that generates the covariance matrix from the time waveform of the signal received by the received signal sampling unit 1, and the periodicity in the space A mode vector calculation unit 3 for calculating a mode vector that is a function of the arrival time and the frequency shift amount from the time waveform of the signal transmitted to the covariance matrix calculated by the received signal covariance matrix calculation unit 2; Since the MUSIC process is performed using the mode vector calculated by the mode vector calculation unit 3, the arrival time of the signal received by the reception signal sampling unit 1 is estimated. Estimate the arrival time of radio waves even if the frequency of the received signal and the signal periodically transmitted to space do not match Achieves the effect can be.

Thereby, for example, the range in which the super-resolution processing can be applied, such as a radar or a positioning device, is greatly expanded. When the arrival time estimation apparatus according to the first embodiment is mounted on a radar, a higher distance resolution can be obtained than when a conventional arrival time estimation method using correlation calculation is employed. Moreover, higher positioning accuracy can be obtained also when mounted on a positioning device.
In addition, the arrival time can be estimated even when the SNR is lower than when the arrival time estimation method based on the conventional correlation calculation is adopted.

Embodiment 2. FIG.
FIG. 6 is a block diagram showing a target detection apparatus (a radar equipped with an arrival time estimation apparatus) according to Embodiment 2 of the present invention. In the figure, the transmission signal source 21 is modulated by a pulse signal or a pseudo noise code. Periodic signals and so on.
The signal combiner 22 outputs a signal transmitted from the transmission signal source 21 to the circulator 23, while distributing a part of the signal to the A / D converter 27.
The circulator 23 supplies the signal output from the signal coupler 22 to the transmission / reception antenna 24 and radiates radio waves into the space, while the radio waves radiated into the space are reflected by the target reflector and received by the transmission / reception antenna 24. Then, the radio wave is output to the receiver 25.

The receiver 25 performs processing such as amplifying the radio wave output from the circulator 23 and converting the frequency of the amplified radio wave.
The A / D converter 26 converts the output signal of the receiver 25 from an analog signal to a digital signal, and outputs the digital signal to the arrival time estimation device 28 as a received signal.
The A / D converter 27 converts the signal distributed by the signal combiner 22 from an analog signal to a digital signal, and outputs the digital signal as a reference signal to the arrival time estimation device 28.
The arrival time estimation device 28 is the arrival time estimation device of FIG. 1, and receives the reception signal output from the A / D converter 26 and the reference signal output from the A / D converter 27, and Estimate arrival time, frequency shift amount and number of propagation paths.

Next, the operation will be described.
As shown in FIG. 3, the transmission signal source 21 periodically transmits a pulse signal, a signal modulated by a pseudo noise code, or the like.
When receiving the signal transmitted from the transmission signal source 21, the signal combiner 22 outputs the signal to the circulator 23 and distributes a part of the signal to the A / D converter 27.
When the circulator 23 receives a signal from the signal coupler 22, the circulator 23 supplies the signal to the transmission / reception antenna 24 to radiate radio waves into the space.
The circulator 23 outputs the radio wave to the receiver 25 when the radio wave radiated from the transmission / reception antenna 24 is reflected by the target reflector and received by the transmission / reception antenna 24.

When receiving the radio wave received by the transmission / reception antenna 24 from the circulator 23, the receiver 25 amplifies the radio wave and performs processing such as converting the frequency of the amplified radio wave.
The A / D converter 26 converts the output signal of the receiver 25 from an analog signal to a digital signal, and outputs the digital signal to the arrival time estimation device 28 as a received signal.
The A / D converter 27 converts the signal distributed by the signal combiner 22 from an analog signal into a digital signal, and outputs the digital signal as a reference signal to the arrival time estimation device 28.

When the arrival time estimation device 28 receives the reception signal from the A / D converter 26 and receives the reference signal from the A / D converter 27, the arrival time τ of the reception signal is received in the same manner as in the first embodiment. The frequency shift amount Δf and the number of propagation paths M are estimated.
The arrival time τ estimated by the arrival time estimation device 28 corresponds to the distance to the target reflector.
The frequency shift amount Δf estimated by the arrival time estimation device 28 corresponds to the line-of-sight speed of the target reflector, and the number of propagation paths M estimated by the arrival time estimation device 28 corresponds to the number of target reflection objects. To do.

  According to the second embodiment, the radar can estimate the number of target reflectors, the distance to the target reflector, and the line-of-sight speed of the target reflector. A plurality of close targets can be discriminated with high distance resolution. In addition, it is possible to discriminate a plurality of targets having closer line-of-sight speeds. In addition, a target can be detected even when the SNR is lower than that of a radar that performs conventional correlation calculation, and a smaller target can be detected at a longer distance.

In the second embodiment, the circulator 23 is mounted on the radar. However, a transmission / reception changeover switch or the like may be mounted instead of the circulator 23.
In addition, although the radar mounts the transmission / reception antenna 24, the transmission antenna and the reception antenna may be mounted separately instead of the transmission / reception antenna 24.
A bistatic configuration in which the transmitting and receiving stations are in different positions and a multistatic configuration in which there are a plurality of receiving stations are also conceivable, but the arrival time estimation device in FIG. 1 can be applied regardless of these configurations.

In the second embodiment, the transmission / reception antenna 24 transmits and receives radio waves. However, a sonar can be configured by transmitting and receiving sound waves instead of the radio waves.
In addition, a light wave radar can be configured by transmitting and receiving light instead of the radio wave.

Embodiment 3 FIG.
FIG. 7 is a block diagram showing a transmitter position positioning device (positioning device equipped with an arrival time estimation device) according to Embodiment 3 of the present invention. In the figure, the same reference numerals as those in FIG. Since it shows, description is abbreviate | omitted.
The radio wave transmitter 31 is a device that transmits radio waves, such as a portable terminal.
The pseudo noise code generator 32 of the radio wave transmitter 31 generates a periodic pseudo noise code, and the transmission signal generator 33 generates a transmission signal.
The modulation unit 34 of the radio wave transmitter 31 modulates the transmission signal generated from the transmission signal generation unit 33 with the periodic pseudo noise code generated from the pseudo noise code generation unit 32.
The transmission antenna 35 of the radio wave transmitter 31 transmits the radio wave modulated by the modulation unit 34.

The receiving antenna 41 is, for example, an antenna of a transmitter positioning device installed in a base station, and receives radio waves transmitted from the radio wave transmitter 31.
The receiver 42 performs quadrature detection by converting the frequency of the radio wave received by the receiving antenna 41, and outputs the quadrature detection signal as a reception signal to the arrival time estimation device 43 similar to FIG.
The arrival time estimation device 43 is an arrival time estimation device similar to that shown in FIG. 1, and receives the reception signal output from the receiver 42 and the pseudo noise code (reference signal) generated from the pseudo noise code generation unit 44. The arrival time τ of the received signal is estimated.
The pseudo noise code generator 44 generates the same pseudo noise code as the pseudo noise code generated from the pseudo noise code generator 32 of the radio wave transmitter 31.
The direct wave selection unit 45 selects the earliest arrival wave among the arrival times of the direct wave and multipath wave estimated by the arrival time estimation device 43 as a direct wave, and outputs the arrival time of the direct wave.
The positioning calculation unit 46 calculates the position of the radio wave transmitter 31 from the arrival times of the direct waves output from the plurality of direct wave selection units 45.

Next, the operation will be described.
The pseudo noise code generator 32 of the radio wave transmitter 31 generates a periodic pseudo noise code, and the transmission signal generator 33 generates a transmission signal.
When receiving the transmission signal from the transmission signal generation unit 33, the modulation unit 34 of the radio wave transmitter 31 modulates the transmission signal with the periodic pseudo noise code generated from the pseudo noise code generation unit 32.
The transmission antenna 35 of the radio wave transmitter 31 radiates the radio wave modulated by the modulation unit 34 into space.

Thereby, the receiving antenna 41 receives the radio wave transmitted from the radio wave transmitter 31.
When the receiving antenna 41 receives a radio wave, the receiver 42 converts the frequency of the radio wave to perform quadrature detection, and outputs the quadrature detection signal as a reception signal to the arrival time estimation device 43 similar to FIG. .
When the arrival time estimating device 43 receives the received signal from the receiver 42 and receives the pseudo noise code as the reference signal from the pseudo noise code generating unit 44, the arrival time estimating device 43 receives the signal from the radio wave transmitter 31 as in the first embodiment. Estimate the arrival time τ of a direct wave or multipath wave.
When the arrival time estimation device 43 estimates the arrival time τ of the direct wave or the multipath wave, the direct wave selection unit 45 selects the earliest arrival wave among the arrival times τ as the direct wave, and The arrival time τ is output to the positioning calculation unit 46.

When receiving the direct wave arrival time τ from the plurality of direct wave selection units 45, the positioning calculation unit 46 substitutes the arrival time τ into the following simultaneous equations and solves the simultaneous equations, whereby the radio wave transmitter 31. Calculate the position of.
cτ _n = ((x−X _n ) ² + (y−Y _n ) ² + (z−Z _n ) ² )) ^1/2 + ct _s
(18)
However, c is the speed of light, (x, y, z) position of the radio transmitter _{31, (X n, Y n} , Z n) the position of the n-th receive antenna 41, the tau _n n-th receive antenna 41 direct wave arrival time of the, t _s is the time when the radio wave is originated.

Equation (18), by solving the (x, y, z) because an equation with four variables with t _s, if the receiving antenna 41 is a four, a simultaneous equation Equation (18), The position (x, y, z) of the radio wave transmitter 31 can be measured.
When there are three receiving antennas 41, for example, assuming the height position z, the two-dimensional position (x, y) can be measured.
When there are five or more receiving antennas 41, a more accurate positioning result can be obtained by solving the equation (18) by the least square method.

  According to the third embodiment, since the direct wave and the multipath wave can be separated with higher resolution than when the arrival time estimation method for performing the conventional correlation calculation is adopted, the arrival time of the direct wave can be accurately determined. It is possible to obtain an effect that a highly accurate positioning result can be obtained. Even when the SNR is low, the position (x, y, z) of the radio wave transmitter 31 can be measured.

In the third embodiment, the radio wave transmitter 31 transmits a radio wave and the reception antenna 41 receives the radio wave. However, if the radio wave is transmitted and received instead of the radio wave, the sound wave is transmitted. A machine positioning device can be configured.
Further, if light is transmitted and received instead of the radio wave, a light wave transmitter position positioning device can be configured.

Embodiment 4 FIG.
8 is a block diagram showing a receiver positioning device (a positioning device in which an arrival time estimation device is mounted) according to Embodiment 4 of the present invention. In the figure, the same reference numerals as those in FIG. Since it shows, description is abbreviate | omitted.
The radio wave transmission station 51 is a radio wave transmission device such as a base station, and the signal receiver 52 is a radio wave reception device such as a mobile terminal.
The receiving antenna 53 receives the radio wave transmitted from the radio wave transmitting station 51. The receiver 54 converts the frequency of the radio wave received by the receiving antenna 53 to perform quadrature detection, and outputs the quadrature detection signal to the A / D converter 55.
The A / D converter 55 samples the quadrature detection signal output from the receiver 54, performs analog / digital conversion, and outputs the digital signal to the transmitter discriminating unit 56 as a received signal.
The transmitter discriminating unit 56 performs reception such as frequency division multiple access (FDMA), time division multiple access (TDMA), code division multiple access (CDMA), and the like, thereby receiving the output from the A / D converter 55. The signals are discriminated for each radio wave transmitting station 51 and distributed to different arrival time estimation devices 57, respectively.

The arrival time estimation device 57 is an arrival time estimation device similar to that shown in FIG. 1, and receives the received signal distributed by the transmitter discriminating unit 56 and the pseudo noise code (reference signal) generated from the pseudo noise code generation unit 58. Thus, the arrival time τ of the received signal is estimated.
The pseudo noise code generation unit 58 generates the same pseudo noise code as the code transmitted from each radio wave transmission station 51.
The direct wave selector 59 selects the earliest arrival wave among the arrival times of the direct wave and multipath wave estimated by the arrival time estimation device 57 as a direct wave, and outputs the arrival time of the direct wave.
The positioning calculation unit 60 calculates the position of the signal receiver 52 from the arrival times of the direct waves output from the plurality of direct wave selection units 59.

Next, the operation will be described.
The plurality of radio wave transmitting stations 51 modulate a transmission signal with a known pseudo noise code, and radiate the radio wave that is the modulation signal into space.
When the receiving antenna 53 receives radio waves transmitted from a plurality of radio wave transmission stations 51, the receiver 54 of the signal receiver 52 converts the frequency of the radio waves to perform quadrature detection, and the quadrature detection signal is converted to A / A. Output to the D converter 55.

Upon receiving the quadrature detection signal output from the receiver 54, the A / D converter 55 samples the quadrature detection signal, performs analog / digital conversion, and outputs the digital signal to the transmitter discriminating unit 56 as a received signal. To do.
When receiving the received signal from the A / D converter 55, the transmitter discriminating unit 56 performs a method such as frequency division multiple access (FDMA), time division multiple access (TDMA), or code division multiple access (CDMA). Thus, the received signal is discriminated for each radio wave transmitting station 51 and distributed to different arrival time estimation devices 57.

When the plurality of arrival time estimation devices 57 receive the received signal from the transmitter discriminating unit 56 and the pseudo noise code which is the reference signal from the pseudo noise code generating unit 58, the plurality of arrival time estimating devices 57 transmit radio waves in the same manner as in the first embodiment. The arrival time τ of a direct wave or multipath wave from the station 51 is estimated.
When the arrival time estimation device 57 estimates the arrival time τ of the direct wave or the multipath wave, the direct wave selection unit 59 selects the earliest arrival wave among the arrival times τ as the direct wave, and The arrival time τ is output to the positioning calculation unit 60.

When the positioning calculation unit 60 receives the arrival times τ of the direct waves from the plurality of direct wave selection units 59, the positioning calculation unit 60 substitutes the arrival times τ into the following simultaneous equations and solves the simultaneous equations, thereby obtaining the signal receiver 52. Calculate the position of.
cτ _n = ((x−X _n ) ² + (y−Y _n ) ² + (z−Z _n ) ² )) ^1/2 + ct _err
(19)
However, c is the speed of light, (x, y, z) is the position of the signal receiver 52, (X _n , Y _n , Z _n ) is the position of the _nth radio wave transmission station 51, and τ _n is the nth radio wave transmission. The arrival time of the direct wave from the station 51, t _err is the clock error of the signal receiver 52.

Since the equation (19) is an equation having four variables (x, y, z) and t _err , the equation (19) is a simultaneous equation if the receivable radio wave transmitting stations 51 are four stations. And the position (x, y, z) of the signal receiver 52 can be measured.
When the receivable radio wave transmitting stations 51 are three stations, for example, if the height position z is assumed, the two-dimensional position (x, y) can be measured.
When there are five or more receivable radio wave transmitting stations 51, a more accurate positioning result can be obtained by solving the equation (19) by the least square method.

  According to the fourth embodiment, since the direct wave and the multipath wave can be separated with higher resolution than when the arrival time estimation method for performing the conventional correlation calculation is adopted, the arrival time of the direct wave can be accurately determined. It is possible to obtain an effect that a highly accurate positioning result can be obtained. Further, even when the SNR is low, the position (x, y, z) of the signal receiver 52 can be measured.

In the fourth embodiment, the radio wave transmitting station 51 transmits radio waves and the receiving antenna 53 receives radio waves. However, if the radio waves are transmitted and received instead of the radio waves, the sound wave reception is performed. A machine positioning device can be configured.
In addition, if a light is transmitted and received instead of the radio wave, a light wave receiver position positioning device can be configured.

Embodiment 5. FIG.
FIG. 9 is a block diagram showing a GPS receiver (positioning device equipped with an arrival time estimation device) according to Embodiment 5 of the present invention. In the figure, the same reference numerals as those in FIG. 1 indicate the same or corresponding parts. Description is omitted.
The GPS satellite 71 orbits the earth and periodically transmits unique pseudo-noise codes at 1-millisecond intervals.
The receiving antenna 73 of the GPS receiver 72 receives a signal periodically transmitted from the GPS satellite 71.
The receiver 74 converts the frequency of the signal received by the receiving antenna 73 to perform quadrature detection, and outputs the quadrature detection signal to the A / D converter 75.
The A / D converter 75 samples the quadrature detection signal output from the receiver 74 L times every millisecond, performs analog / digital conversion, and uses the digital signal as a reception signal to generate a reception signal covariance matrix calculation unit. Output to 2. The reception signal covariance matrix calculation unit 2 is a calculation unit similar to the reception signal covariance matrix calculation unit 2 in FIG. 1, and calculates a covariance matrix from data for K milliseconds.

The arrival time estimation device 76 is the same arrival time estimation device as in FIG. However, the reception signal covariance matrix calculator 2 is not provided and is provided in the preceding stage.
The satellite number selection unit 77 selects the satellite number of the visible satellite estimated from the orbit information of the GPS satellite 71 and the predicted position of the GPS receiver 72. However, if there is no such preliminary information, a satellite number is selected at random.
The pseudo noise code generator 78 generates a unique pseudo noise code corresponding to the satellite number selected by the satellite number selector 77. The pseudo-noise code generator 78 stores a unique pseudo-noise code for each GPS satellite 71 in advance.
The direct wave selector 79 selects the earliest arrival wave among the arrival times of the direct wave and multipath wave estimated by the arrival time estimation device 76 as a direct wave, and outputs the arrival time of the direct wave.
The positioning calculation unit 80 calculates the position of the GPS receiver 72 from the arrival times of the direct waves output from the plurality of direct wave selection units 79.

Next, the operation will be described.
The plurality of GPS satellites 71 orbiting the earth periodically transmit unique pseudo noise codes at intervals of 1 millisecond.
The receiving antenna 73 of the GPS receiver 72 receives a signal periodically transmitted from the GPS satellite 71.

When the receiving antenna 73 receives a signal transmitted from the GPS satellite 71, the receiver 74 converts the frequency of the signal to perform quadrature detection, and outputs the quadrature detection signal to the A / D converter 75.
When the A / D converter 75 receives the quadrature detection signal from the receiver 74, the A / D converter 75 samples the quadrature detection signal L times every millisecond, performs analog / digital conversion, and receives the digital signal as a reception signal. The signal is output to the signal covariance matrix calculation unit 2.
When the reception signal covariance matrix calculation unit 2 receives the reception signal from the A / D converter 75, the covariance matrix is calculated from the data for K milliseconds using Equation (13) as in the first embodiment. R _xx is calculated.
The received signal covariance matrix calculation unit 2 calculates the covariance matrix R _xx in parallel or sequentially for all GPS satellites 71 in the visible range. This makes it possible to calculate the arrival times τ of radio waves from all GPS satellites 71.

On the other hand, the satellite number selection unit 77 selects the satellite number of the visible satellite estimated from the orbit information of the GPS satellite 71 and the predicted position of the GPS receiver 72. However, if there is no such preliminary information, a satellite number is selected at random.
When receiving the satellite number selected by the satellite number selection unit 77, the pseudo noise code generation unit 78 generates a unique pseudo noise code corresponding to the satellite number.

When arrival time estimation apparatus 76 receives covariance matrix R _xx from received signal covariance matrix calculation section 2 and receives a pseudo noise code that is a reference signal from pseudo noise code generation section 78, it performs the same operation as in the first embodiment. Thus, the arrival time τ of the direct wave or multipath wave from the GPS satellite 71 is estimated.
When the arrival time estimation device 76 estimates the arrival time τ of the direct wave or multipath wave, the direct wave selection unit 79 selects the earliest arrival wave among the arrival times τ as the direct wave, and The arrival time τ is output to the positioning calculation unit 80.

When the positioning calculation unit 80 receives the arrival times τ of direct waves from the plurality of direct wave selection units 79, the positioning calculation unit 80 substitutes the arrival times τ into the following simultaneous equations and solves the simultaneous equations, thereby obtaining the GPS receiver 72. Calculate the position of.
cτ _n = ((x−X _n ) ² + (y−Y _n ) ² + (z−Z _n ) ² )) ^1/2 + ct _err
(20)
Where c is the speed of light, (x, y, z) is the position of the GPS receiver 72, (X _n , Y _n , Z _n ) is the position of the _nth GPS satellite 71, and τ _n is the nth GPS satellite 71. The arrival time of the direct wave from, t _err is the clock error of the GPS receiver 72.

Since Equation (20) is an equation having four variables (x, y, z) and t _err , if there are four visible GPS satellites 71, Equation (20), which is a simultaneous equation, is solved. Thus, the position (x, y, z) of the GPS receiver 72 can be measured.
When the number of visible GPS satellites 71 is three, for example, assuming a height position z, the two-dimensional position (x, y) can be measured.
When there are five or more visible GPS satellites 71, a more accurate positioning result can be obtained by solving the equation (20) by the least square method.

Here, the eigenvalues and eigenvectors calculated inside the MUSIC processing unit 5 are calculated from the same covariance matrix R _xx in all the GPS satellites 71, and need only be calculated once. By doing so, the amount of calculation can be greatly reduced.
According to the fifth embodiment, since the direct wave and the multipath wave can be separated with higher resolution than when the arrival time estimation method for performing the conventional correlation calculation is adopted, the arrival time of the direct wave can be accurately determined. It is possible to obtain an effect that a highly accurate positioning result can be obtained.

Further, according to the fifth embodiment, another advantage is that the GPS receiver 72 can be highly sensitive.
The transmission signal from the GPS satellite 71 is BPSK modulated with navigation data of 50 bps, and since this navigation data is unknown, it is difficult to integrate over 20 milliseconds with the conventional correlation calculation method, and there is a limit to high sensitivity. However, in the arrival time estimation method of the fifth embodiment, since the navigation data is canceled when the covariance matrix is calculated, the covariance matrix is calculated using an infinitely long received signal in principle. Thus, the GPS receiver 72 can be made highly sensitive.
The reason will be described below.

First, it is assumed that the reception signal of the GPS receiver 72 is represented by the following equation (21).
x (k) = b (k) pn (k) + n (k) (21)
Here, k is an index representing a pseudo-noise code period of 1 millisecond, pn (k) is a pseudo-noise code including delay and Doppler frequency shift, and b (k) is “+1” or “−1” for BPSK modulation. Navigation data taking a value, n (k) is receiver noise.

ただし、擬似雑音符号ｐｎ（ｋ）と受信機雑音ｎ（ｋ）は、無相関であると仮定している。この仮定は、共分散行列を計算する平均数を十分にとれば通常成立する。
式（２２）に示すように、この実施の形態５による到来時刻推定法を用いれば、共分散行列の計算過程で航法データがキャンセルされるため、航法データが不明であっても長時間に亘って共分散行列の平均演算が可能となり、高感度なＧＰＳ受信機を得ることができる。 In this case, the covariance matrix R _xx of the received signal is expressed by the following equation (22).

However, it is assumed that the pseudo noise code pn (k) and the receiver noise n (k) are uncorrelated. This assumption usually holds if the average number for calculating the covariance matrix is sufficient.
As shown in the equation (22), when the arrival time estimation method according to the fifth embodiment is used, the navigation data is canceled in the process of calculating the covariance matrix. Therefore, even if the navigation data is unknown, it takes a long time. Thus, the covariance matrix can be averaged, and a highly sensitive GPS receiver can be obtained.

It is a block diagram which shows the arrival time estimation apparatus by Embodiment 1 of this invention. It is a block diagram which shows the MUSIC process part of the arrival time estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the transmission period of a signal. It is explanatory drawing which shows the eigenvalue (lambda) _l of covariance matrix _Rxx . It is explanatory drawing which shows the example of a calculation result of two-dimensional MUSIC spectrum _Pmu ((tau), (DELTA) f). It is a block diagram which shows the target detection apparatus (radar which mounts the arrival time estimation apparatus) by Embodiment 2 of this invention. It is a block diagram which shows the transmitter position positioning apparatus (positioning apparatus carrying the arrival time estimation apparatus) by Embodiment 3 of this invention. It is a block diagram which shows the receiver position positioning apparatus (positioning apparatus carrying the arrival time estimation apparatus) by Embodiment 4 of this invention. It is a block diagram which shows the GPS receiver (positioning apparatus which mounts the arrival time estimation apparatus) by Embodiment 5 of this invention.

Explanation of symbols

1 reception signal sampling unit (signal reception unit), 2 reception signal covariance matrix calculation unit (covariance matrix generation unit), 3 mode vector calculation unit (mode vector generation unit), 4 MUSIC processing unit (arrival time estimation unit), 11 Eigenvalue / Eigenvector Calculation Unit, 12 Noise Eigenvalue Determination Unit, 13 MUSIC Spectrum Calculation Unit, 21 Transmission Signal Source, 22 Signal Combiner, 23 Circulator, 24 Transmit / Receive Antenna, 25 Receiver, 26 A / D Converter, 27 A / D converter, 28 arrival time estimation device, 31 radio wave transmitter, 32 pseudo noise code generation unit, 33 transmission signal generation unit, 34 modulation unit, 35 transmission antenna, 41 reception antenna, 42 receiver, 43 arrival time estimation device, 44 Pseudo-noise code generation unit, 45 Direct wave selection unit, 46 Positioning calculation unit, 51 Radio wave transmission station, 52 Signal receiver, 53 Receiving antenna, 54 receiver, 55 A / D converter, 56 transmitter discriminating unit, 57 arrival time estimating device,
58 pseudo-noise code generation unit, 59 direct wave selection unit, 60 positioning calculation unit, 71 GPS satellite, 72 GPS receiver, 73 reception antenna, 74 receiver, 75 A / D converter, 76 arrival time estimation device, 77 satellite Number selection unit, 78 pseudo noise code generation unit, 79 direct wave selection unit, 80 positioning calculation unit.

Claims (10)

  Signal receiving means for receiving a signal periodically transmitted to space, covariance matrix generating means for generating a covariance matrix from a time waveform of the signal received by the signal receiving means, and periodically transmitted to space Mode vector generating means for generating a mode vector that is a function of the arrival time and the frequency shift amount from the time waveform of the received signal, the covariance matrix generated by the covariance matrix generating means, and the mode vector generating means An arrival time estimation device comprising: arrival time estimation means for estimating an arrival time of a signal received by the signal reception means by performing super-resolution processing using a mode vector.   Signal receiving means for receiving a signal periodically transmitted to space, covariance matrix generating means for generating a covariance matrix from a time waveform of the signal received by the signal receiving means, and periodically transmitted to space Mode vector generating means for generating a mode vector that is a function of the arrival time and the frequency shift amount from the time waveform of the reference signal correlated with the received signal, the covariance matrix generated by the covariance matrix generating means, and the mode An arrival time estimation apparatus comprising: an arrival time estimation unit that estimates an arrival time of a signal received by the signal reception unit by performing super-resolution processing using a mode vector generated by a vector generation unit.   The arrival time estimation device according to claim 1 or 2, wherein the arrival time estimation means performs MUSIC processing as super-resolution processing.   The arrival time estimating means calculates eigenvalues and eigenvectors of the covariance matrix generated by the covariance matrix generating means, specifies eigenvalues equal to noise power among the eigenvalues of the covariance matrix, and corresponds to the eigenvalues A MUSIC spectrum relating to the arrival time and the frequency shift amount is calculated from the eigenvector and the mode vector generated by the mode vector generating means, and the arrival time of the signal received by the signal receiving means is estimated from the peak position of the MUSIC spectrum. The arrival time estimation apparatus according to claim 3.   5. The arrival time estimation apparatus according to claim 4, wherein the arrival time estimation means estimates the number of propagation paths by comparing eigenvalues of the covariance matrix and noise power.   5. The arrival time estimation apparatus according to claim 4, wherein the arrival time estimation means estimates the frequency shift amount of the signal received by the signal reception means from the peak position of the MUSIC spectrum.   The arrival time estimation device according to any one of claims 1 to 6, wherein the estimation result of the arrival time estimation means is output to a target detection device that detects a target existing in space.   The arrival time estimation device according to any one of claims 1 to 6, wherein the estimation result of the arrival time estimation means is output to a transmitter position positioning device that measures the position of the signal transmitter.   The arrival time estimation apparatus according to any one of claims 1 to 6, wherein the estimation result of the arrival time estimation means is output to a receiver position positioning apparatus that measures the position of the signal receiver. The arrival time estimation apparatus according to any one of claims 1 to 6, wherein the estimation result of the arrival time estimation means is output to a GPS receiver that measures the current position.

Images