JP2007298503A - Propagation delay time measuring device and radar device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately measure the propagation delay time by acquiring a frequency transfer function by reducing the degradation of SNR and then performing high-resolution processing. <P>SOLUTION: This propagation delay time measuring device comprises a cross correlation function calculation section for calculating a cross correlation function of a received diffusion signal and a reference signal, a self correlation function calculation section for calculating a self correlation function of the reference signal, first and second extraction sections for extracting the periphery of each peak value of the cross correlation function and self correlation function, first and second Fourier transforming sections for acquiring first and second frequency functions by Fourier transforming the periphery of each peak value of the cross correlation function and self correlation function, a division section for acquiring a frequency transfer function by dividing the first frequency function by the second frequency function in a frequency range for reducing the degradation of the SNR, and a high-resolution processing section that separates a multi-path wave included in the peak value of the cross correlation function based on the frequency transfer function and calculates an evaluation function indicating a true time delay of the received diffusion signal. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a propagation delay time measuring apparatus and a radar apparatus that measure the propagation delay time of radio waves, sound waves, or light whose carrier wave is modulated by a known spread signal.

  2. Description of the Related Art Radio wave propagation delay time measuring devices are used to specify the position of radio wave transmitters and receivers such as radars, GPS (Global Positioning System) receivers, and portable terminals. Conventionally, as such a radio wave delay time estimation method, a method in which a correlation peak is detected by a cross-correlation function between a transmission signal itself or a reference signal similar to the transmission signal and a reception signal, and a delay time is estimated from the peak position is generally used. Has been used.

  Furthermore, Non-Patent Document 1 and Non-Patent Document 2 describe methods for obtaining a higher time resolution than the correlation calculation using a high resolution algorithm such as MUSIC (Multiple Signal Classification) and ESPRIT (Estimation of Signal Parameters via Rotational Invariance Techniques). Are listed. In this delay time estimation method using a high-resolution algorithm, a radio wave modulated by a PN (Pseudo Noise) code is received by a receiver, and a signal obtained by performing a Fourier transform on the radio wave is converted into the same PN code as the modulated code (here, (Referred to as a reference signal) is divided by a Fourier-transformed signal, and the MUSIC algorithm is applied to the obtained frequency transfer function to estimate the delay time with high accuracy. The problem with this method is that the frequency spectrum of the received signal is divided by the frequency spectrum of the reference signal, and the signal-to-noise ratio (hereinafter referred to as SNR) may deteriorate. is there. In general, when the SNR deteriorates, signals that are close in time cannot be separated even if the high resolution algorithm is used, and as a result, the signal propagation delay time estimation accuracy deteriorates.

  In order to solve the above problem, the frequency transfer function generated by dividing the frequency spectrum of the received signal by the frequency spectrum of the reference signal is inverse FFTed and converted to the time domain, and then the noise is removed and removed in the time domain. For example, Patent Document 1 discloses a method for obtaining a transfer function obtained by performing FFT on a processed signal and removing a noise component.

Japanese Patent Laid-Open No. 11-261444 (FIG. 1) Nobuyoshi Kikuma "Adaptive signal processing by array antenna" Science and Technology Publishing (1998) Nakahara, Ogawa, Kikuma, Inagaki, B-10, 'Comparison study of multipath propagation delay time resolution between FFT-MUSIC method and FFT operation type correlation method', 1995 IEICE General Conference

  Also in the technique described in Patent Document 1, in order to obtain a frequency transfer function, the frequency spectrum of the received signal is divided by the frequency spectrum of the reference signal. When the SNR at the time of incidence is low, the SNR deteriorates due to the division. There is a problem.

  The present invention has been made to solve the above problems, and obtains a frequency transfer function by reducing the degradation of SNR, and then measures the propagation delay time with high accuracy by performing high resolution processing. An object of the present invention is to obtain a propagation delay time measuring apparatus capable of performing the above and a radar apparatus to which the apparatus is applied.

  The propagation delay time measuring apparatus according to the present invention receives a radio wave, a sound wave or a light wave whose carrier wave is modulated by a known spread signal, converts it into an electrical signal, extracts the received spread signal from the converted signal, In a propagation delay time measuring apparatus that generates a reception spread signal digitized by D conversion and measures a propagation delay time of the received radio wave, sound wave or light wave based on the digitized reception spread signal and a reference signal A reference signal generation unit that generates the same reference signal as the spread signal at the time of transmission, and compensates for the frequency shift amount generated in the propagation process of the digitized received spread signal, and the mutual relationship between the compensated received spread signal and the reference signal A correlation function is calculated and a cross-correlation function calculator that estimates the propagation delay time of the received spread signal and the periphery of the peak value of the cross-correlation function that represents the estimated propagation delay time are extracted Using the correlation peak position calculated by the first extraction unit, the first Fourier transform unit that obtains a frequency function by performing Fourier transform around the peak value of the extracted cross correlation function, and the cross correlation function calculation unit, An autocorrelation function calculation unit that calculates an autocorrelation function of the reference signal, a second extraction unit that extracts the periphery of the peak value of the calculated autocorrelation function, and a Fourier transform of the periphery of the peak value of the extracted autocorrelation function A frequency function obtained by the second Fourier transform unit and a frequency function obtained by the first Fourier transform unit in a frequency range in which the deterioration of the signal-to-noise ratio is reduced. Divide by function to obtain a frequency transfer function, and based on the frequency transfer function obtained by the divider, separate and receive multipath waves that approximate the received spread signal included in the peak value of the cross-correlation function Those with high-resolution processing unit for calculating an evaluation function that represents the true delay time of the scattered signal.

  According to the present invention, even when the SNR of the received signal is low, there is an effect that the delay signal that is close in time can be separated and the propagation delay time of the spread signal can be measured with high accuracy. In addition, since the periphery of the peak value of the cross-correlation function is extracted and only a part of the extracted part is subjected to high resolution processing, the amount of calculation can be reduced as compared with the conventional method.

Embodiment 1 FIG.
Although the present invention can be used for a radio terminal such as a radar device, a GPS positioning device, and a mobile phone, Embodiments 1 to 6 describe examples in which the present invention is applied to a GPS positioning device. In addition, the GPS positioning device has a configuration (assist type GPS receiver) that receives information such as the approximate position of the satellite and GPS time through a separate server (assist type GPS receiver). In the first to sixth embodiments, the assistance is required. A case where the present invention is applied to a self-supporting GPS receiver will be described.

  The spread signal handled in the present invention corresponds to a C / A code for generating a GPS signal when GPS is taken as an example. The configuration of this C / A code is shown in FIG. The horizontal axis represents time. A GPS signal that is a radio wave transmitted from a GPS satellite is a signal obtained by BPSK modulation (Binary Phase Shift Keying) of a signal having a carrier frequency L1 (LINK 1: 1575.42 MHz). The basis of the modulation is a signal called C / A code (Clear / Acquisition Code). The actual GPS signal is a signal obtained by modulating the carrier frequency L1 with a P code (Precision Code) and then further modulating with a C / A code. However, the description of the P code is omitted here. The duration (cycle) of the C / A code is 1 millisecond, and has 1,023 bits (1.023 Mbps) therebetween. Accordingly, one bit of the C / A code is about 1 μsec, and one bit of the C / A code is generally called one chip. The 20 repetitions of the C / A code are a unit. This corresponds to 1 bit of the navigation data, and the navigation data is represented by a signal which is 20 repetitions of the C / A code or a signal whose polarity is inverted (in the case of a modulated GPS signal, the phase differs by 180 °). Therefore, one bit of navigation data is 20 milliseconds.

1 is a block diagram showing a functional configuration of a propagation delay time estimation apparatus according to Embodiment 1 of the present invention.
The propagation delay time estimation apparatus includes a reception antenna 111, a spread signal acquisition unit 110 including a reception unit 101 and an A / D conversion unit 102, a reference signal generation unit 103, a cross-correlation function calculation unit 104, extraction units 1051, 1052, and Fourier transform. Sections 1061 and 1062, a division section 107, an autocorrelation function calculation section 108, and a high resolution processing section 109.

Next, the operation will be described.
The spread signal acquisition unit 110 receives a radio wave whose carrier wave is modulated by a known spread signal, converts the electric signal, extracts a received spread signal from the converted received signal, and then digitizes the signal by A / D conversion. Means for obtaining a received spread signal. This process has been performed conventionally, and specifically, is as follows.
The receiving antenna 111 receives a radio wave of a GPS signal transmitted from a GPS satellite and sends the received signal to the receiving unit 101. The receiving unit 101 performs amplification, frequency conversion, and the like on the received GPS signal. For example, a signal other than about 2 to 20 MHz before and after the carrier frequency L1 (1575.42 MHz) is removed by a band pass filter, a sine wave having the same frequency as the carrier frequency L1 is multiplied, and a harmonic component is removed by a low pass filter. Thus, the frequency is directly converted into a baseband signal, and the received spread signal is extracted. As another method, a configuration in which a received signal is once converted into an IF (intermediate frequency) band by multiplying by a frequency smaller than the carrier frequency and then detected to extract a baseband signal is considered. A description will be given of the former method of converting to a baseband signal.

  At this time, the receiving unit 101 extracts the real part and the imaginary part of the signal using two sine waves whose phases are different by 90 degrees. Since these two signals are different in phase by 90 degrees, these are regarded as a complex number having the real part and the imaginary part as absolute values and treated as one signal having phase and amplitude information, and this is called a complex signal. In the following description, it is assumed that the extracted signal is a complex number. A configuration is also possible in which only the real part is taken out and converted into digital data by the A / D converter 102, and then the real part and the imaginary part are taken out using two sine waves whose phases are different by 90 degrees. A configuration in which the real part and the imaginary part are extracted before A / D conversion and handled as a complex signal will be described.

  If the carrier frequency of the received signal matches the frequency of the sine wave multiplied by this, the phase of the received spread signal is constant. However, due to the relative speed between the GPS satellite and the GPS positioning device 100, the carrier wave frequency of the received GPS signal does not match the frequency (1575.42 MHz) transmitted by the GPS satellite. In addition, since there is an error in the oscillation frequency of the oscillator included in the receiving unit 101, the sine wave multiplied by the received signal also has a frequency different from 1575.42 MHz. Due to the difference in frequency, the phase of the signal converted by the receiving unit 101 changes. That is, the phase of the converted signal rotates with the frequency difference as a frequency. Considering this, the receiving unit 101 outputs a complex signal having phase and amplitude information so that the difference in frequency can be corrected later.

  The A / D conversion unit 102 samples the reception spread signal output from the reception unit 101 at a predetermined sampling frequency and converts it into a digital signal. In this case, the real part and the imaginary part of the signal output from the receiving unit 101 are respectively converted, and a complex number represented by a real / imaginary part pair is output. The C / A code is composed of 1,023 bits in one cycle. Since one cycle of the C / A code is 1 millisecond, it is necessary to perform sampling at a frequency of 2.046 MHz or higher.

  The reference signal generation unit 103 is means for generating the same reference signal as the spread signal at the time of transmission. Specifically, a spread signal that matches the C / A code used by the GPS satellite is generated as a reference signal from the number of the GPS satellite that transmitted the received GPS signal (binary value of 1 or −1). At this time, the number of data is adjusted so as to be synchronized with the sampling frequency of the digital data output from the receiving unit 101. For example, if the sampling frequency is 2.046 MHz, two pieces of data are generated for one bit of the C / A code.

The cross-correlation function calculation unit 104 compensates for the frequency shift amount generated in the propagation process of the digitized reception spread signal acquired by the spread signal acquisition unit 110, and calculates the cross-correlation function between the compensated reception spread signal and the reference signal. It is a means for calculating and estimating the propagation delay time of the received spread signal. Specifically, the interval for obtaining this correlation is one cycle of the C / A code (1 millisecond). The reception spread signal acquired by the spread signal acquisition unit 110 has been subjected to a frequency shift due to the error of the Doppler frequency accompanying the satellite motion and the oscillation frequency of the transmitter included in the reception unit 101 as described above. To compensate. Then, a cross-correlation function ccf between the compensated received spread signal and the C / A code that is the reference signal generated by the reference signal generation unit 103 is calculated by the processing of equation (1). Since the amount of frequency shift is unknown, ccf is a function of the starting point p and the frequency difference f _d of the correlation interval.

ここで、Ｎは１ｍｓ内に含まれるサンプル数であり、ｆ_dは参照信号であるＣ／Ａコードの周波数シフト、ΔＴはサンプリング周期である。＊は複素共役を計算することを意味する。ただし、ｃｃｆ_q（ｐ、ｆ_d）は相互相関関数（複素数）、ｑは受信信号を１ｍｓ毎に分割したときのＣ／Ａコード周期番号インデックス、ｐは相関区間の開始時点（サンプリングデータの数を単位とする整数。ｐ＝０，１，・・・，Ｎ−１）である。また、ｆ_dは周波数差、ＮはＣ／Ａコード一周期あたりのサンプリングデータ数（自然数。Ｃ／Ａコード一周期は１ミリ秒なので、サンプリング周波数は１０００Ｎ）、ｖ_q（ｘ）は拡散信号取得部１１０が取得した受信拡散信号のｑ番目Ｃ／Ａコード周期におけるｘ＋１番目のサンプリングデータ（複素数）である。さらに、Ｃ_ca（ｘ）は参照信号生成部１０３が生成するｘ＋１番目のＣ／Ａコードであり、Ｃ_ca（ｘ+Ｎ）＝Ｃ_ca（ｘ）の周期性が成り立つとする。ｅは自然対数の底、ｊは虚数単位、πは円周率である。
なお、相互相関関数ｃｃｆはＦＦＴ（Fast Fourier Transform：高速フーリエ変換）により計算してもよい。その場合、計算量を削減することができる。

Here, N is the number of samples included in 1 ms, f _d is the frequency shift of the C / A code that is the reference signal, and ΔT is the sampling period. * Means to calculate the complex conjugate. Where ccf _q (p, f _d ) is a cross-correlation function (complex number), q is a C / A code period number index when the received signal is divided every 1 ms, and p is the start time of the correlation interval (number of sampling data An integer with a unit of p = 0, 1,. F _d is the frequency difference, N is the number of sampling data per C / A code period (natural number. Since one C / A code period is 1 millisecond, the sampling frequency is 1000 N), and v _q (x) is the spread signal. This is the (x + 1) th sampling data (complex number) in the qth C / A code period of the received spread signal acquired by the acquisition unit 110. Further, C _ca (x) is the (x + 1) th C / A code generated by the reference signal generation unit 103, and it is assumed that the periodicity of C _ca (x + N) = C _ca (x) is established. e is the base of the natural logarithm, j is the imaginary unit, and π is the pi.
The cross-correlation function ccf may be calculated by FFT (Fast Fourier Transform). In that case, the amount of calculation can be reduced.

FIG. 3 shows an example of a time-frequency profile of the absolute value of the cross correlation function ccf calculated by the cross correlation function calculation unit 104. Cross-correlation function ccf is a function of the correlation start time p and the frequency difference f _d as a variable. When the frequency difference f _d matches the actual frequency difference f _T , the absolute value of the cross-correlation function ccf has a peak as shown in FIG. This peak appears when the start point p of the correlation interval coincides with the start point of the received C / A code. If p × ΔT, it can be estimated as the received C / A code delay time. However, when the frequency difference f _d does not match the actual frequency difference, such a peak does not appear. In the following description, it is assumed that the frequency difference f _d is equal to the actual frequency difference f _T, and the cross-correlation function ccf (p, f _T ) when f _d = f _T is assumed to be ccf (p). The correlation start time for each 1 ms period of the C / A code, that is, pΔT will be described as a delay time hereinafter. The calculated ccf (p) is passed to the extraction unit 1051, and the peak position p _peak where the absolute value of ccf (p) is maximum is passed to the autocorrelation function calculation unit 108.

  FIG. 4 shows an example of the cross-correlation function ccf (p) when a multipath wave is incident in close proximity to the received spread signal. It was assumed that there was no noise. The difference in delay time between the multipath wave and the direct wave (received spread signal) is about 0.2 μsec, and the frequency shift and amplitude value of both are the same. As shown in the figure, when the multipath wave is close in time, the shape of the cross-correlation function ccf (p) is distorted, and the signal delay time of the direct wave cannot be measured accurately.

The extraction unit (first extraction unit) 1051 is a means for extracting the periphery of the correlation peak value of the cross-correlation function representing the propagation delay time estimated by the cross-correlation function calculation unit 104. Specifically, extracts 2N _ab pieces near _{p peak -N ab ~p peak + N} ab -1 of the correlation peak p _peak of the cross-correlation function ccf calculated by the cross-correlation function calculation section 104, the extracted cross-correlation The function is newly set to ccfe (i) (i = 0, 1,..., 2N _ab −1). The extraction width is set wider to ensure the number of samples when the high resolution processing unit 109 performs high resolution processing. N _ab is set so that, for example, N _ab ΔT = 10 μsec.
A Fourier transform unit (first Fourier transform unit) 1061 is means for performing Fourier transform around the peak value of the cross-correlation function extracted by the extraction unit 1051. Specifically, the Fourier transform processing of the cross-correlation function ccfe extracted by the extraction unit 1051 is performed as shown in Equation (2) using FFT.

The autocorrelation function calculation unit 108 is a unit that calculates the autocorrelation function of the reference signal generated by the reference signal generation unit 103 using the correlation peak position calculated by the cross correlation function calculation unit 104. Specifically, using the correlation peak position p _peak calculated by the cross-correlation function calculation unit 104, the reference signal generation unit 103 generates a C / A code that is a reference signal by the processing of equation (3). An autocorrelation function acf (p) is calculated. It can be seen that the autocorrelation function acf (p) has a _{peak at} p = p _peak .

The extraction unit (second extraction unit) 1052 is means for extracting the periphery of the calculated peak value of the autocorrelation function. Specifically, the autocorrelation function acf (p) calculated by the autocorrelation function calculation unit 108 has the same vicinity as the width extracted by the extraction unit 1051, ie, p _peak −N _{ab to} p _peak. + N _ab −1 is extracted. The extracted autocorrelation function is assumed as “fefe (i)”.
The Fourier transform unit (second Fourier transform unit) 1062 is means for performing a Fourier transform around the peak value of the autocorrelation function extracted by the extraction unit 1052. Specifically, the Fourier transform of the autocorrelation function “acfe” extracted by the extraction unit 1052 is performed using FFT to obtain a frequency function as shown in Equation (4).

ここで、Ｋは除算を行う正の周波数範囲における周波数ビンの数であり、負の周波数も考慮すると除算を行う周波数ビンの総数は２Ｋ＋１となる。
また、ＦＦＴ後の周波数分解能Δｆは２Ｎ_abΔＴとなるので除算を行う周波数範囲は、−２ＫＮ_abΔＴ［Ｈｚ］〜２ＫＮ_abΔＴ［Ｈｚ］となる。除算する周波数範囲は、受信機のバンドパスフィルタの帯域幅程度が妥当である。例えば、仮に受信機バンドパスフィルタの帯域が８ＭＨｚであった場合、除算を行う範囲は、−４ＭＨｚ〜４ＭＨｚとする。また、ＰＮＲ（Peak to Noise floor level Ratio；相互相関関数のピークとノイズフロアの比）によって除算する周波数範囲を変化させる構成でもよい。すなわち、ＰＮＲが低い場合には、除算する周波数範囲を狭めＧＰＳ信号のメインローブの幅（±１ＭＨｚ）程度とするが、ＰＮＲがある程度高い場合には、除算する周波数範囲を受信機帯域幅（±４ＭＨｚ）程度とするような構成である。これは、ＰＮＲが低い場合、ＧＰＳ信号のメインローブ以外の周波数ではノイズが支配的になるので、（５）式の除算を行う場合においてノイズが増幅される可能性があるからである。 The division unit 107 is a unit that obtains a frequency transfer function by dividing the frequency function obtained by the Fourier transform unit 1061 by the frequency function obtained by the Fourier transform unit 1062 in a frequency range that reduces the deterioration of the signal-to-noise ratio. is there. Specifically, the frequency function of the cross-correlation function input from the Fourier transform unit 1061 shown in the equation (2) is the frequency function of the autocorrelation function input from the Fourier transform unit 1062 shown in the equation (4). Divide using the equation (5) in the frequency bin.

Here, K is the number of frequency bins in the positive frequency range to be divided, and the total number of frequency bins to be divided is 2K + 1 considering the negative frequency.
Further, since the frequency resolution Δf after FFT is 2N _ab ΔT, the frequency range to be divided is −2 KN _ab ΔT [Hz] to 2 KN _ab ΔT [Hz]. The frequency range to be divided is appropriate to the bandwidth of the band-pass filter of the receiver. For example, if the band of the receiver bandpass filter is 8 MHz, the range in which the division is performed is -4 MHz to 4 MHz. Moreover, the structure which changes the frequency range divided by PNR (Peak to Noise floor level Ratio; ratio of the peak of a cross-correlation function and a noise floor) may be sufficient. That is, when the PNR is low, the frequency range to be divided is narrowed to about the width of the main lobe of the GPS signal (± 1 MHz), but when the PNR is somewhat high, the frequency range to be divided is set to the receiver bandwidth (± 4 MHz). This is because when PNR is low, noise becomes dominant at frequencies other than the main lobe of the GPS signal, and therefore noise may be amplified when performing the division of equation (5).

ここで、η（ｋ）は内部雑音、Ｌは伝搬経路数、ｃ_lは各伝搬経路の複素振幅係数、２Ｎ_abは抽出部１０５１，１０５２で抽出された相互相関関数、自己相関関数のサンプル数である。また、τ_lはｌ番目伝搬経路における、相互相関関数のピーク位置ｐ_peakを時間に直した値ΔＴｐ_peakからの遅延時間シフト量である。 Based on the frequency transfer function obtained by the division unit 107, the high resolution processing unit 109 separates the multipath wave approximated to the received spread signal included in the peak value of the cross correlation function calculated by the cross correlation function calculation unit 104. And an evaluation function representing the true delay time of the received spread signal. Specifically, this is performed as follows.
The high resolution processing includes a MUSIC method, an ESPRIT method, a maximum likelihood estimation method, and the like. Here, a case where the MUSIC method is applied will be described. A detailed description of the MUSIC method is described in detail by Nobuyoshi Kikuma, “Adaptive signal processing using an array antenna”, Science and Technology Publishing (1998), pages 191-202 and pages 269-282. A brief explanation will be given.
It is assumed that the frequency transfer function obtained by the division process of the division unit 107 can be expressed by equation (6).

Here, eta (k) is the internal noise, L is the number of propagation paths, c _l is the number of samples of the complex amplitude coefficients, 2N _ab cross-correlation function extracted by the extraction unit 1051 and 1052, the autocorrelation function of the propagation path It is. Further, τ _l is a delay time shift amount from a value ΔTp _peak obtained by correcting the peak position p _peak of the cross correlation function in the l-th propagation path.

  In the MUSIC algorithm, the propagation delay time is estimated with high accuracy by calculating the correlation matrix of the signal of the above equation (6) and obtaining the eigenvalue and eigenvector of the correlation matrix. However, when a multipath wave having a high correlation with the direct wave is incident as in the case of equation (6), the rank of the correlation matrix becomes 1, and estimation by the MUSIC algorithm becomes impossible. Therefore, in order to suppress the cross-correlation between the direct wave and the multipath wave and restore the rank of the correlation matrix, it is necessary to pre-process the frequency averaging operation on the frequency function of equation (6).

ここで、Ｃは移動平均後のフルランクのＭ×Ｍの複素振幅相関行列であり、ＩはＭ×Ｍの単位行列である。またσ²は内部雑音電力である。 First, 2K−M + 2 sub-array data of M (M <2K + 1) components are extracted from the 2K + 1 component frequency array data while shifting one component at a time. The correlation matrix R _α (αth partial correlation matrix) of each subarray is calculated as shown in Equation (8). The 2K−M + 2 partial correlation matrices are averaged, and the covariance matrix R _xx after the moving average is calculated as shown in Equation (9).

Here, C is a full rank M × M complex amplitude correlation matrix after moving average, and I is an M × M unit matrix. Σ ² is the internal noise power.

Next, eigenvalues and eigenvectors are obtained for the covariance matrix R _xx . If the eigenvalue and eigenvector are λ _m and e _m (m = 1,..., M), equation (13) holds.

さらに、行列ＡとＣがフルランクであることから、（１５）式が成り立つ。

したがって、（１６）式が成り立つ。

The eigenvalue of the covariance matrix R _xx is obtained, and the arrival wave number L is estimated from the number of eigenvalues larger than the internal noise power σ ² . Further, the number of incoming waves L may be determined in advance. Equation (14) holds for the eigenvector corresponding to the eigenvalue equal to the internal noise power σ ² .

Furthermore, since the matrices A and C are full rank, equation (15) is established.

Therefore, equation (16) is established.

（１７）式の評価関数Ｐ_MU（τ）は、各パス波の遅延時間シフト量τ_lと一致した場合にピークを持つことが分かる。なぜなら、（１６）式において遅延時間ベクトルａ_lと雑音固有値に対応する雑音固有ベクトルｅ_m（ｍ＝Ｌ＋１，…，Ｍ）との内積が０となる、つまりＰ_MU（τ）の分母が０となるからである。 Therefore, the function expressed by the equation (17) using the following delay time vector a (τ) is used as the evaluation function of the high resolution processing unit 109.

It can be seen that the evaluation function P _MU (τ) in the equation (17) has a peak when it coincides with the delay time shift amount τ _{l of} each path wave. This is because the noise eigenvectors _{e m (m = L + 1} , ..., M) becomes the inner product is zero and, that is the denominator of the P _MU (tau) is 0 corresponding to the delay time vector a _l and noise eigenvalues in (16) Because it becomes.

Ｐ_MU（τ）において推定を行う遅延時間シフト量τの幅を−Ｎ_abΔＴ〜＋Ｎ_abΔＴとして、Ｐ_MU（τ）を計算する。Ｐ_MU（τ）のピークサーチを行い、１個のピークを検出する。各ピークに対応する遅延時間シフト量τ^(l) _peakと相互相関関数のピーク値ｐ_peakを用いて、各伝搬経路のＣ／Ａコード１ｍｓ内における伝搬遅延時間はτ^(l) _peak＋ΔＴｐ_peakとして推定することができる。 The delay time range τ _max that can be estimated in the evaluation function P _MU (τ) is expressed by Equation (20) due to the periodicity of the delay time mode vector a (τ).

The width of the delay time shift tau to estimate as -N _ab ΔT~ + N _ab ΔT in P _MU (τ), to calculate the P _MU (τ). A peak search of P _MU (τ) is performed to detect one peak. Using the delay time shift amount τ ^(l) _peak corresponding to each peak and the peak value p _peak of the cross-correlation function, the propagation delay time in the C / A code 1 ms of each propagation path is expressed as τ ^(l) _peak + ΔTp _peak Can be estimated.

  FIG. 5 shows an example of the evaluation result in the first embodiment. The horizontal axis represents the signal delay time in one cycle (within 1 ms) of the C / A code, and the vertical axis represents the normalized evaluation function value. While the multipath wave close to the direct wave cannot be separated by the delay profile based on the cross correlation function, the delay profile of this example separates the two waves and accurately estimates the delay time of both the direct wave and the multipath wave. You can see that it is made.

  As described above, according to the first embodiment, the cross-correlation function between the received spread signal compensated for the frequency shift amount generated in the propagation process and the same reference signal as the spread signal at the time of transmission by the cross-correlation function calculation unit 104. The first extraction unit 1051 extracts the periphery of the peak value of the cross-correlation function representing the estimated propagation delay time, and the first Fourier transform unit 1061 Fourier-transforms the periphery of the peak value of the cross-correlation function. The autocorrelation function calculation unit 108 calculates the autocorrelation function of the reference signal using the correlation peak position calculated by the crosscorrelation function calculation unit 104, and the peak value around the calculated autocorrelation function. Is extracted by the second extraction unit 1052, and the second Fourier transform unit 1062 performs Fourier transform around the peak value of the autocorrelation function to obtain a frequency function, and the division unit 107 Therefore, the frequency transfer function is obtained by dividing the frequency function obtained by the first Fourier transform unit 1061 by the frequency function obtained by the second Fourier transform unit 1062 in a frequency range that reduces the deterioration of the signal-to-noise ratio. In the high-resolution processing unit 109, an evaluation function representing the true delay time of the received spread signal by separating the multipath wave approximated to the received spread signal included in the peak value of the cross-correlation function based on the frequency transfer function Is calculated. Therefore, even when the SNR of the received signal is low, the delay signals that are close in time can be separated and the propagation delay time of the spread signal can be measured with high accuracy. In addition, since the periphery of the peak value of the cross-correlation function is extracted and only a part of the extracted part is subjected to high resolution processing, the amount of calculation can be reduced as compared with the conventional method.

一般に方形波による関数の抽出を行った場合、抽出点における不連続性により周波数スペクトル上においてサイドローブが発生し、高分解能処理において精度を劣化させる恐れがある。そこで、この実施の形態２では、相互相関関数と自己相関関数の抽出において時間窓関数を乗算することにより、周波数スペクトルのサイドローブを低減し、遅延時間分解能を向上させることを目的とする。 Embodiment 2. FIG.
In the extraction units 1051 and 1052 in the first embodiment, the extraction around the peak values of the cross-correlation function and the autocorrelation function simply extracts a signal within a certain time range. Will be multiplied.

In general, when a function is extracted by a square wave, side lobes are generated on the frequency spectrum due to discontinuity at the extraction point, and there is a risk of degrading accuracy in high resolution processing. Therefore, in the second embodiment, an object is to reduce the side lobe of the frequency spectrum and improve the delay time resolution by multiplying the time window function in the extraction of the cross-correlation function and the autocorrelation function.

FIG. 6 is a block diagram showing a functional configuration of a propagation delay time measuring apparatus according to Embodiment 2 of the present invention. In the figure, functional components corresponding to those in FIG. 1 are denoted by the same reference numerals, and description thereof will be omitted in principle. The second embodiment is different from the first embodiment in that time window function multiplication units 6011 and 6012 are provided in place of the extraction units 1051 and 1052.
The time window function multiplication unit (first time window function multiplication unit) 6011 is means for multiplying the peak value of the cross correlation function calculated by the cross correlation function calculation unit 104 by the time window function in the range to be extracted. . Also, the time window function multiplier (second time window function multiplier) 6012 multiplies the time window function within the same extraction range around the peak value of the autocorrelation function calculated by the autocorrelation function calculator 108. It is means to do.
As an example, description will be given of a configuration for extracting using Hamming window in a section _{p peak -N ab ~p peak + N} ab -1.

ｐ_peak−Ｎ_ab〜ｐ_peak＋Ｎ_ab−１の区間で抽出された相互相関関数ｃｃｆｅ（ｉ）および自己相関関数ａｃｆｅ（ｉ）に、上記ハミング窓関数を乗算し、乗算した相互相関関数および自己相関関数を新たにｃｃｆｅｗ（ｉ）、ａｃｆｅｗ（ｉ）とおくと、ｃｃｆｅｗ（ｉ）をフーリエ変換部１０６１に、ａｃｆｅｗ（ｉ）をフーリエ変換部１０６２に送る。また、ここではハミング窓を使用した場合について説明したが、これが三角窓、ハニング窓等であってもよい。 A time window function w (i) in the case of using a Hamming window is expressed by equation (22).

the _{p peak -N ab ~p peak + N} ab extracted in a section -1 cross-correlation function ccfe (i) and the autocorrelation function ACFE (i), multiplied by the Hamming window function, multiplied by cross-correlation function and self If the correlation function is newly set as ccfw (i) and acfw (i), ccfw (i) is sent to the Fourier transform unit 1061 and acfw (i) is sent to the Fourier transform unit 1062. Although the case where a Hamming window is used has been described here, this may be a triangular window, a Hanning window, or the like.

  FIG. 7 shows an example of the evaluation result in the second embodiment obtained by the high resolution processing unit 108. Although the evaluation conditions are the same as those in the first embodiment, the peaks of the direct wave and the multipath wave are sharper than in the case of the first embodiment (FIG. 5) by multiplying the Hamming window function. It can be seen that the delay time resolution is improved.

Embodiment 3.
FIG. 8 is a block diagram showing a functional configuration of a propagation delay time measuring apparatus according to Embodiment 3 of the present invention. In the figure, functional components corresponding to those in FIG. 1 are denoted by the same reference numerals, and description thereof will be omitted in principle. The third embodiment is different from the first embodiment in that an addition processing unit 801 is newly provided between the cross-correlation function calculation unit 104 and the extraction unit 1051.

ここで、ｑはＣ／Ａコード周期番号である。上記加算処理は，相互相関関数ｃｃｆ(ｐ)においてピーク位置ｐ_peakが得られているならば、ｐ_peak周辺のみを式（２３）に従い実行すればよい。しかし、ＳＮＲが低くｃｃｆ（ｐ）においてピークが得られない場合には、全てのｐすなわちｐ＝０，１，…，Ｎ−１において式（２３）に従い加算処理を行う必要がある。Ｎａｖ（ｑ）はＣ／Ａコード１周期毎の航法ビット変調系列であり、ｅｘｐ（０）あるいはｅｘｐ（ｊπ）のどちらかを取る。
上述したように、ＧＰＳ信号はＣ／Ａコード２０周期（２０ｍｓ）毎に航法データビット系列によりＢＰＳＫ変調されている。したがって、２０ｍｓ以上の加算処理を行う場合、航法ビット系列による位相変調を補償して積分する必要がある。このＮａｖ（ｑ）は事前の処理において求めておく。また式（２３）の加算処理を行う場合の注意点として、相互相関関数算出部１０４における周波数シフト量ｆｄは１／加算処理時間の精度で、真の周波数シフトｆｒと一致している必要がある。さもなければ、加算処理時間の増加に従い、ｆｄとｆｒの周波数差によって受信信号位相が回転し、積分によるＳＮＲ改善効果が少なくなってしまうからである。例えば、１秒の加算処理を行う場合、ｆｄとｆｒは１Ｈｚ単位で一致している必要がある。周期加算値ｓｃｃｆは抽出部１０５１に送られる。抽出部１０５１以降の処理は実施の形態１と同様である。 In the addition processing unit 801, for example, the cross-correlation function ccf (p) calculated by the cross-correlation function calculating unit 104 is added for each period according to the equation (23).

Here, q is a C / A code cycle number. If the peak position p _peak is obtained in the cross-correlation function ccf (p), only the vicinity of p _peak may be executed according to the equation (23). However, when the SNR is low and no peak is obtained at ccf (p), it is necessary to perform addition processing according to the equation (23) in all p, that is, p = 0, 1,. Nav (q) is a navigation bit modulation sequence for each cycle of the C / A code, and takes either exp (0) or exp (jπ).
As described above, the GPS signal is BPSK-modulated with the navigation data bit sequence every 20 C / A codes (20 ms). Therefore, when performing addition processing of 20 ms or more, it is necessary to compensate and integrate phase modulation by the navigation bit sequence. This Nav (q) is obtained in advance processing. As a precaution when performing the addition process of Expression (23), the frequency shift amount fd in the cross-correlation function calculation unit 104 needs to coincide with the true frequency shift fr with the accuracy of 1 / addition process time. . Otherwise, as the addition processing time increases, the received signal phase is rotated by the frequency difference between fd and fr, and the SNR improvement effect by integration is reduced. For example, when performing addition processing for 1 second, fd and fr need to correspond in 1 Hz units. The period addition value sccf is sent to the extraction unit 1051. The processes after the extraction unit 1051 are the same as those in the first embodiment.

  As described above, according to the third embodiment, since the addition process is performed on the cross correlation function, the SNR can be further improved. In addition, even when a signal peak cannot be obtained by the cross-correlation function, it is possible to detect the peak by performing addition processing, and then processing by the high-resolution processing unit makes it possible to increase the peak even when the signal cannot be detected by the cross-correlation function The signal delay time can be estimated with high accuracy.

Embodiment 4.
FIG. 9 is a block diagram showing a functional configuration of a propagation delay time measuring apparatus according to Embodiment 4 of the present invention. In the figure, functional parts corresponding to those in FIG. 6 of the second embodiment are given the same reference numerals, and description thereof will be omitted in principle. The fourth embodiment differs from the second embodiment in that an addition processing unit 801 is newly provided between the cross-correlation function calculation unit 104 and the time window function multiplication unit 6011.
In the fourth embodiment, as described in the third embodiment, after the addition is performed for each period of the cross-correlation function ccf (p), the time window function multiplication unit 6011 calculates the integration value of the cross-correlation function. After the range is multiplied by the window function, the Fourier transform is performed. The time window function multiplier 6012 multiplies the range of the autocorrelation function to be extracted by the window function in the same manner as in the second embodiment, and then performs Fourier transform.

  As described above, according to the fourth embodiment, after the SNR is recovered in the addition processing unit 801, the vicinity of the correlation peak is extracted by the window function multiplication, thereby improving the SNR in the vicinity of the correlation peak, The high resolution processing unit 109 can estimate the signal delay time with high accuracy.

Embodiment 5 FIG.
The extraction unit (first extraction unit) 1051 in the first embodiment extracts the vicinity of the start time point p _peak of the correlation section where the cross-correlation function ccf has a peak, so that the mutual around the true value of the signal delay time is obtained. The correlation function is extracted. Similarly, in the time window function multiplying unit 6011 in the second embodiment, the time window function is multiplied around the start time point p _peak of the correlation interval of the cross correlation function ccf, thereby obtaining a signal delay time around the true value. The cross correlation function is extracted. However, in a low SNR environment, the start point p _{peak of the} correlation interval in which the cross-correlation function ccf has a peak is not always around the true value of the signal delay time. That is, as the signal power decreases, the correlation peak (hereinafter referred to as the signal peak) generated around the true value of the signal delay time decreases, and the peak due to the correlation between the noise and the reference signal (hereinafter referred to as the noise peak). The value may be higher. In such a case, when a peak of the cross correlation function ccf is simply detected, a noise peak is detected. Therefore, an erroneous signal delay time is estimated by the high resolution processing unit 109. Therefore, in the fifth embodiment, as will be described below, means for enabling accurate estimation of the delay time is provided.

  FIG. 10 is a block diagram showing a functional configuration of the delay time measuring apparatus according to the fifth embodiment of the present invention. In the figure, the same reference numerals are given to the functional parts corresponding to those in FIG. 1 of the first embodiment, and the description thereof will be omitted in principle. The fifth embodiment is different from the first embodiment in that a detection unit 1102 is provided instead of the extraction unit 1051, and a determination unit (first determination unit) 1104 is newly added to the high resolution processing unit 109. The resolution processing unit 1101 is provided. Note that the high resolution processing unit 109 according to the first embodiment has a configuration in which the eigenanalysis unit 1103 and the delay time estimation unit 1105 are combined together. The high resolution processing unit 1101 is provided. Hereinafter, processing of parts different from the first embodiment will be described.

The detection unit 1102 is a means for extracting the periphery of all peak values for which the absolute value of the cross correlation function calculated by the cross correlation function calculation unit 104 exceeds a predetermined threshold. Specifically, the correlation start point p _peak ^(h) (h is an integer from 1 to ^H) when the absolute value of the cross-correlation function ccf (p) exceeds a set threshold value is calculated. That is, p _peak ^(h) is an index of a peak candidate value including a signal peak and a noise peak. The threshold value for detection is the average value of the absolute value of the cross-correlation function ccf (p) or, if the approximate received signal power is known in advance, the noise floor level of the cross-correlation function estimated from it. It may be used. Alternatively, H may be set in advance, which is the total number of peak candidate value indexes, and p _peak ^(h) may be detected in descending order of the absolute value of the cross-correlation function ccf (p) until H is satisfied. Next, a cross-correlation function ccf (p) around p _peak ^(h ) is extracted. Here, the number of samples to be extracted is the same as in the first embodiment, 2N _{ab of} p _peak ^(h) −N _{ab to} p _peak ^(h) + N _ab −1 are extracted, and the extracted ccf is ccfe ^(h). (I) (i = 0, 1,..., 2N _ab −1). The extracted peak candidate value index p _peak ^(h) is sent to the autocorrelation function calculation unit 108. Also, ccfe ^(h) (i) is sent to the Fourier transform unit 1061.
The autocorrelation function calculation unit 108 calculates the autocorrelation function acfe ^(h) (i) according to the equation (3) based on the peak candidate value index p _peak ^(h) sent from the detection unit 1051. The calculated cake ^(h) (i) is sent to the extraction unit 1052.

Next, the operation of the high resolution processing unit 1101 will be described.
The eigenanalysis unit 1103 first calculates the correlation matrix of the frequency transfer function x (k) ^(h) sent from the division unit 107, and calculates Rxx ^(h) subjected to the moving average process. Next, eigenvalue expansion of Rxx ^(h) is performed to obtain an eigenvector e _m ^(h) and an eigenvalue λ _m ^(h) . This corresponds to the processing of the expressions (7) to (13) in the first embodiment. The obtained eigenvector e _m ^(h) and eigenvalue λ _m ^(h) are sent to the determination unit 1104. The determination unit 1104 outputs a signal to the ccfe ^(h) (i) extracted by the detection unit 1102 from the distribution of the eigenvalue λ _m ^(h) of the frequency transfer function x (k) ^(h) sent from the division unit 107. It is determined whether a peak or a noise peak is included. If a noise peak is included, ccfe ^(h) (i) includes a noise component and a side lobe of an autocorrelation function. On the other hand, only the periphery of the peak of the autocorrelation function is included in cafe ^(h) (i), the correlation between the peak periphery of the autocorrelation function and the side lobe is small, and x (k) ^(h) obtained by the division unit 107 It is assumed that the noise component becomes dominant in. Therefore, it is expected that the eigenvalue of x (k) ^(h) is only the noise eigenvalue and does not include the signal eigenvalue. When only the noise eigenvalue is included, all eigenvalues have almost the same value from the equation (13), so it is easy to determine whether the signal peak is included from the eigenvalue distribution.

Here, as the determination method of the determination unit 1104, a real number times the average value of the eigenvalues λ _m ^(h) as a threshold value for determination, contains signal peaks when the eigenvalues lambda _m exceeds the threshold value ^(h) is present Judge that For the real number multiple coefficient, an appropriate value may be obtained by a prior simulation. Further, since the noise eigenvalue is equal to the noise power according to the equation (13), when the noise power is known in advance, the value may be used as a threshold for eigenvalue detection. All the eigenvalues ^{_{λ m (h) (h =}} 1,2, ···, H) performs the determination process with respect to, the h _sig which is determined to include the signal peak, and ^e _m ^(hsig) λ _m ^(hsig) is sent to the delay time estimation unit 1105.
The delay time estimation unit 1105 uses the eigenvector and the eigenvalues e _m ^(hsig) and λ _m ^(hsig) sent from the determination unit 1104 to calculate the evaluation function of equation (17), and the evaluation function reaches a peak. The delay time τ ^{(hsig, l)} _peak is obtained, and the propagation delay time within 1 ms of the C / A code of each propagation path can be estimated as τ ^(l) _peak + ΔTp _peak ^(hsig) . In addition, the delay time estimation unit 1105 uses the transfer function x (k) ^(hsig) obtained by extracting the periphery of p _peak ^(hsig) determined as the signal peak by the determination unit 1104, and uses the ESPRIT method or the maximum likelihood. The configuration may be such that the delay time is estimated using an estimation method.

As described above, in the fifth embodiment, a threshold is provided when detecting the peak of the cross-correlation function calculated by the cross-correlation function calculation unit 104, and all the peak candidates that exceed the threshold are surrounded by the peak value. A detection unit 1102 that performs extraction is provided, and the high resolution processing unit 1101 determines a signal peak from a distribution of eigenvalues for a correlation peak that exceeds a threshold by a determination unit (first determination unit) 1104, and Since the delay time is estimated only when the signal is detected, it is possible to detect the signal peak even under a low SNR environment and accurately estimate the delay time.
The detection unit 1102 extracts the periphery of all peak values when the absolute value of the cross-correlation function calculated by the cross-correlation function calculation unit 104 exceeds a predetermined threshold. The periphery for the peak value may be extracted after multiplying the periphery by the time window.
In addition, in the above example, the detection unit 1102 is configured to directly detect the peak value for the cross-correlation function calculated by the cross-correlation function calculation unit 104. However, the addition processing unit 801 in the third embodiment is used. It may be provided in place of the subsequent extraction unit 1051. In this case, since the addition processing unit 801 adds the cross-correlation function calculated by the cross-correlation function calculating unit 104 for each period, the detection unit 1102 determines that the calculated period addition value of the cross-correlation function is a predetermined threshold value. The periphery of all peak values exceeding the threshold value is extracted, or the periphery of the peak value is extracted after being multiplied by a time window function. Therefore, the SNR can be improved, and even when the signal peak cannot be obtained by the cross-correlation function, it is possible to detect the peak by performing the addition process, and the signal of the high-resolution processing unit 1101 can be detected with high accuracy. The delay time can be estimated.

Embodiment 6 FIG.
In the fifth embodiment, the signal peak and the noise peak are determined from the eigenvalue distribution of the correlation matrix obtained from the frequency transfer function. However, when the noise eigenvalue varies, the determination may be erroneous. In view of this, the sixth embodiment provides means for increasing the probability of detecting a signal peak as described below.
FIG. 11 is a block diagram showing a functional configuration of a delay time measuring apparatus according to Embodiment 6 of the present invention. In the figure, functional parts corresponding to those in FIG. 10 of the fifth embodiment are denoted by the same reference numerals, and description thereof will be omitted in principle. The sixth embodiment is different from the fifth embodiment in that a delay time evaluation function calculation unit 1201 and a determination unit (second determination unit) 1203 are newly provided in place of the delay time estimation unit 1105. Hereinafter, processing of parts different from the fifth embodiment will be described.

The delay time evaluation function calculation unit 1201 calculates a delay time evaluation function based on the noise eigenvector calculated by the eigenanalysis unit 1103 when the determination unit (first determination unit) 1104 determines that a signal peak is included. It is a means for calculating. Specifically, the MUSIC evaluation function (delay time evaluation function) of Expression (17) using the noise eigenvector determined from the frequency transfer function x (k) ^(h) determined by the determination unit 1104 to include a signal peak. Is calculated and sent to another determination unit (second determination unit) 1203.
The determination unit 1203 is a unit that outputs the delay time evaluation function value as the delay time estimation value only when the calculated delay time evaluation function value exceeds a predetermined threshold value. Specifically, using the MUSIC evaluation function sent from the delay time evaluation function calculation unit 1201, it is determined again whether the frequency transfer function x (k) ^(h) includes a signal peak or a noise peak. To do. When a signal peak is included in the frequency transfer function x (k) ^(h) , the inner product of the noise eigenvector and the delay time mode vector obtained from the covariance matrix of x (k) ^(h) approaches 0, and the delay time Near the true value, the MUSIC evaluation function has a sharp peak. Therefore, it is possible to estimate the signal delay time by detecting the peak of the evaluation function. However, when a noise peak is included in the frequency transfer function x (k) ^(h) , the orthogonality between the noise eigenvector obtained from the correlation matrix of x (k) ^(h) and the delay time mode vector is not guaranteed, and the inner product Can be greater than zero. Therefore, when the MUSIC evaluation function value exceeds the set threshold value, it is determined as a signal peak, and when it is equal to or less than the threshold value, it is determined as a noise peak. As the threshold of the MUSIC evaluation function, a value obtained in advance by computer simulation or experiment is used. When the threshold value is exceeded, the peak position of the MUSIC evaluation function determined as the signal peak is calculated and output as the signal delay time. Therefore, it is possible to reduce the determination that the noise peak is erroneously set as the signal peak.

  As described above, according to the sixth embodiment, when the determination unit (first determination unit) 1104 determines that a signal peak is included, the delay time evaluation function calculation unit 1201 performs the eigenanalysis unit 1103. A MUSIC evaluation function (delay time evaluation function) is calculated based on the calculated noise eigenvector, and a signal peak is calculated only when the MUSIC evaluation function value exceeds a set threshold by the determination unit (second determination unit) 1203. Since it is determined that the peak is included and the peak position of the MUSIC evaluation function is calculated and output as the signal delay time, the probability of detecting the signal peak can be increased.

Embodiment 7.
In the first embodiment, the propagation delay time measuring apparatus serving as the basis of the present invention has been described. In the seventh embodiment, a radar apparatus to which the propagation delay time measuring apparatus is applied will be described.
FIG. 11 is a block diagram showing a functional configuration of a radar apparatus according to Embodiment 7 of the present invention. In the figure, functional parts corresponding to those in FIG. 1 of the first embodiment are denoted by the same reference numerals, and description thereof will be omitted in principle. The radar apparatus includes a transmission / reception antenna 1005, a transmission / reception unit 1001, a spread signal generation unit 1003, an A / D conversion unit 1002, a cross-correlation function calculation unit 104, extraction units 1051 and 1052, Fourier transform units 1061 and 1062, a division unit 107, An autocorrelation function calculation unit 108 and a high resolution processing unit 109 are provided.

The spread signal generation unit 1003 generates a pulse signal or a pseudo noise code as a spread signal. The generated spread signal is distributed by the combiner 1004 and output to the transmission / reception unit 1001, the cross-correlation function calculation unit 104, and the autocorrelation function calculation unit 108. In the transmission / reception unit 1001, the internal transmitter modulates the carrier wave with the input spread signal to generate a transmission signal, and then amplifies the power and feeds power to the transmission / reception antenna 1005 through a transmission / reception switch (not shown). . The transmission / reception antenna 1005 converts the transmission signal into a radio wave, radiates it into the space, receives the radio wave reflected back from the target and converts it into an electrical signal, and receives the signal inside the transmission / reception unit 1001 via the transmission / reception switch. To give. The receiver of the transmission / reception unit 1001 performs a known process such as amplification and frequency conversion on the received signal to extract the received spread signal and outputs the signal to the A / D conversion unit 1002. A / D conversion section 1002 A / D converts the received spread signal to generate a digitized received spread signal, and outputs it to cross-correlation function calculation section 104. The subsequent operation of each functional part is the same as the operation of the propagation delay time measuring apparatus described in detail in the first embodiment, and will be omitted.
Note that the spread signal at the time of transmission given to the cross-correlation function calculation unit 104 and the autocorrelation function calculation unit 108 is obtained from the transmission signal by a demodulator having the same configuration including the receiver of the transmission / reception unit 1001 and the A / D conversion unit 1002. There is also a method using a demodulated signal.

As described above, in the seventh embodiment, since the propagation delay time is estimated using the same propagation delay time measurement method as in the first embodiment, even when the SNR is low, the high-resolution processing has high accuracy. The target distance can be estimated. In addition, it is possible to measure adjacent targets separately.
In the seventh embodiment, the radar apparatus to which the first embodiment is applied has been described. However, even if any one of the propagation delay time measuring methods of the second to sixth embodiments is applied, the respective implementations are performed. An effect similar to that of the embodiment can be obtained.
In each embodiment, the case of a radio wave whose spread signal is known has been described. However, it may be applied to light instead of the radio wave, and a light wave radar having the same effect can be obtained. Moreover, you may apply with respect to a sound wave and can obtain the sonar with the same effect.

It is a block diagram which shows the function structure of the propagation delay time measuring apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the structure of the C / A code for producing | generating a GPS signal as an example of a spreading | diffusion signal. It is explanatory drawing which shows an example of the cross correlation function which concerns on Embodiment 1 of this invention. It is explanatory drawing which shows an example of a cross-correlation function when the adjacent multipath wave concerning Embodiment 1 of this invention mixes. It is explanatory drawing which shows the example of an evaluation result which concerns on Embodiment 1 of this invention. It is a block diagram which shows the function structure of the propagation delay time measuring apparatus by Embodiment 2 of this invention. It is explanatory drawing which shows the example of an evaluation result which concerns on Embodiment 2 of this invention. It is a block diagram which shows the function structure of the propagation delay time measuring apparatus by Embodiment 3 of this invention. It is a block diagram which shows the function structure of the propagation delay time measuring apparatus by Embodiment 4 of this invention. It is a block diagram which shows the function structure of the propagation delay time measuring apparatus by Embodiment 5 of this invention. It is a block diagram which shows the function structure of the propagation delay time measuring apparatus by Embodiment 6 of this invention. It is a block diagram which shows the function structure of the radar apparatus by Embodiment 7 of this invention.

Explanation of symbols

  101 receiving unit, 102, 1002 A / D conversion unit, 103 reference signal generating unit, 104 cross correlation function calculating unit, 107 dividing unit, 108 autocorrelation function calculating unit, 109, 1101 high resolution processing unit, 111 receiving antenna, 801 Addition processing unit, 1001 transmission / reception unit, 1003 spread signal generation unit, 1004 combiner, 1005 transmission / reception antenna, 1051 extraction unit (first extraction unit), 1052 extraction unit (second extraction unit), 1061 Fourier transform unit (first 1 Fourier transform unit), 1062 Fourier transform unit (second Fourier transform unit), 1103 eigenvalue analysis unit, 1104 determination unit (first determination unit), 1105 delay time estimation unit, 1201 delay time evaluation function calculation unit, 1203 Determination unit (second determination unit), 6011, 6012 Time window function multiplication unit.

Claims (14)

Received radio wave, sound wave, or light wave whose carrier wave is modulated by a known spread signal, convert it to an electrical signal, extract the received spread signal from the converted signal, and then digitize the received spread signal by A / D conversion In the propagation delay time measuring device for measuring the propagation delay time of the received radio wave, sound wave or light wave based on the digitized reception spread signal and the reference signal,
A reference signal generator that generates the same reference signal as the spread signal at the time of transmission;
Compensates for the amount of frequency shift generated in the propagation process of the digitized received spread signal, calculates the cross-correlation function between the compensated received spread signal and the reference signal, and estimates the propagation delay time of the received spread signal A correlation function calculator;
A first extraction unit for extracting a periphery of a peak value of a cross-correlation function representing the estimated propagation delay time;
A first Fourier transform unit that obtains a frequency function by performing a Fourier transform around the peak value of the extracted cross-correlation function;
An autocorrelation function calculation unit that calculates an autocorrelation function of the reference signal using the correlation peak position calculated by the cross correlation function calculation unit;
A second extraction unit for extracting a periphery of a peak value of the calculated autocorrelation function;
A second Fourier transform unit for obtaining a frequency function by performing a Fourier transform around the peak value of the extracted autocorrelation function;
Division to obtain a frequency transfer function by dividing the frequency function obtained by the first Fourier transform unit by the frequency function obtained by the second Fourier transform unit in a frequency range in which deterioration of the signal-to-noise ratio is reduced. And
Evaluation representing the true delay time of the received spread signal by separating the multipath wave approximated to the received spread signal included in the peak value of the cross correlation function based on the frequency transfer function obtained by the division unit A propagation delay time measuring device comprising a high resolution processing unit for calculating a function. A first time window function multiplying unit that is provided in place of the first extraction means and multiplies the time window function of the range to be extracted around the peak value of the cross correlation function calculated by the cross correlation function calculating unit;
A second time window function multiplying unit provided in place of the second extracting means and multiplying the time window function of the range to be similarly extracted around the peak value of the autocorrelation function calculated by the autocorrelation function calculating unit. The propagation delay time measuring apparatus according to claim 1, further comprising: An addition processing unit that adds the cross-correlation function calculated by the cross-correlation function calculation unit for each period;
2. The propagation delay time measuring apparatus according to claim 1, wherein the first extraction unit extracts a period addition value of the calculated cross-correlation function. An addition processing unit that adds the cross-correlation function calculated by the cross-correlation function calculation unit for each period;
The first time window function multiplication unit multiplies the period addition value of the cross-correlation function obtained by the coherent integration unit by a time window function in a range to be extracted. Propagation delay time measurement device. Provided in place of the first extraction unit and extract the periphery of all peak values for which the absolute value of the cross-correlation function calculated by the cross-correlation function calculation unit exceeds a predetermined threshold, or the peak value A detection unit that extracts the periphery of the peak value after multiplying the periphery of the time window,
The high resolution processor
An eigenanalysis unit for calculating eigenvalues and eigenvectors from the correlation matrix of the frequency transfer function obtained by the division unit;
A first determination unit that determines whether a signal peak is included from the eigenvalue obtained by the eigenanalysis unit;
When the first determination unit determines that a signal peak is included, the multipath wave approximated to the received spread signal included in the peak value of the cross-correlation function is separated based on the frequency transfer function. The propagation delay time measuring apparatus according to claim 1, further comprising a delay time estimation unit that calculates an evaluation function representing a true delay time of the received spread signal. An addition processing unit that adds the cross-correlation function calculated by the cross-correlation function calculation unit for each period;
The detection unit extracts the periphery of all peak values for which the periodic addition value of the cross-correlation function calculated by the addition processing unit exceeds a predetermined threshold value, or adds a time window function around the peak value. 6. The propagation delay time measuring apparatus according to claim 5, wherein the periphery of the peak value is extracted after multiplication. Instead of the delay time estimation unit, the high resolution processing unit
A delay time evaluation function calculation unit that calculates a delay time evaluation function based on the noise eigenvector calculated by the eigenanalysis unit when it is determined by the first determination unit that a signal peak is included;
6. A second determination unit that outputs a delay time evaluation function value as a delay time estimation value only when the calculated delay time evaluation function value exceeds a predetermined threshold value. 6. The propagation delay time measuring device according to 6. A transmission signal whose carrier wave is modulated by a spread signal is transmitted as a radio wave, sound wave, or light wave. The radio wave, sound wave, or light wave that is reflected back from the target is received, converted into an electrical signal, and received from the converted signal. After extracting the spread signal, the reception spread signal digitized by A / D conversion is generated, and the propagation delay time of the received radio wave, sound wave or light wave based on the digitized reception spread signal and the transmission signal In a radar device that measures
A transmission signal generation unit for generating a predetermined spread signal;
Compensates for the amount of frequency shift generated in the propagation process of the digitized received spread signal, calculates the cross-correlation function between the compensated received spread signal and the predetermined spread signal at the time of transmission, and propagates the propagation delay of the received spread signal A cross-correlation function calculator for estimating time;
A first extraction unit for extracting a periphery of a peak value of a cross-correlation function representing the estimated propagation delay time;
A first Fourier transform unit that obtains a frequency function by performing a Fourier transform around the peak value of the extracted cross-correlation function;
An autocorrelation function calculation unit that calculates an autocorrelation function of the digitized transmission spread signal using the correlation peak position calculated by the cross correlation function calculation unit;
A second extraction unit for extracting a periphery of a peak value of the calculated autocorrelation function;
A second Fourier transform unit for obtaining a frequency function by performing a Fourier transform around the peak value of the extracted autocorrelation function;
Division to obtain a frequency transfer function by dividing the frequency function obtained by the first Fourier transform unit by the frequency function obtained by the second Fourier transform unit in a frequency range in which deterioration of the signal-to-noise ratio is reduced. And
Evaluation representing the true delay time of the received spread signal by separating the multipath wave approximated to the received spread signal included in the peak value of the cross correlation function based on the frequency transfer function obtained by the division unit A radar apparatus comprising a high resolution processing unit for calculating a function. A first time window function multiplying unit that is provided in place of the first extraction means and multiplies the time window function of the range to be extracted around the peak value of the cross correlation function calculated by the cross correlation function calculating unit;
A second time window function multiplying unit provided in place of the second extracting means and multiplying the time window function of the range to be similarly extracted around the peak value of the autocorrelation function calculated by the autocorrelation function calculating unit. The radar apparatus according to claim 8, further comprising: An addition processing unit that adds the cross-correlation function calculated by the cross-correlation function calculation unit for each period;
The radar apparatus according to claim 8, wherein the first extraction unit extracts a period addition value of the calculated cross-correlation function. An addition processing unit that adds the cross-correlation function calculated by the cross-correlation function calculation unit for each period;
The first time window function multiplication unit multiplies the period addition value of the cross-correlation function obtained by the addition processing unit by a time window function in a range to be extracted. Radar device. Provided in place of the first extraction unit and extract the periphery of all peak values for which the absolute value of the cross-correlation function calculated by the cross-correlation function calculation unit exceeds a predetermined threshold, or the peak value A detection unit that extracts the periphery of the peak value after multiplying the periphery of the time window,
The high resolution processor
An eigenanalysis unit for calculating eigenvalues and eigenvectors from the correlation matrix of the frequency transfer function obtained by the division unit;
A first determination unit that determines whether a signal peak is included from the eigenvalue obtained by the eigenanalysis unit;
When the first determination unit determines that a signal peak is included, the multipath wave approximated to the received spread signal included in the peak value of the cross-correlation function is separated based on the frequency transfer function. 9. The radar apparatus according to claim 8, further comprising a delay time estimation unit that calculates an evaluation function representing a true delay time of the received spread signal. An addition processing unit that adds the cross-correlation function calculated by the cross-correlation function calculation unit for each period;
The detection unit extracts the periphery of all peak values for which the periodic addition value of the cross-correlation function calculated by the addition processing unit exceeds a predetermined threshold value, or adds a time window function around the peak value. 13. The radar apparatus according to claim 12, wherein the periphery of the peak value is extracted after multiplication. Instead of the delay time estimation unit, the high resolution processing unit
A delay time evaluation function calculation unit that calculates a delay time evaluation function based on the noise eigenvector obtained by the eigenanalysis unit when it is determined by the first determination unit that a signal peak is included;
13. The apparatus according to claim 12, further comprising a second determination unit that outputs the delay time evaluation function value as the delay time estimation value only when the calculated delay time evaluation function value exceeds a predetermined threshold value. 13. The radar device according to 13.

Images