JP6396073B2 - Measurement method of signal propagation characteristics - Google Patents

Description

  The purpose of the present invention is to measure minute fluctuations in a spatial propagation path to an observation point by secondarily using, for example, a digital terrestrial broadcast signal, and in particular, signal propagation for accurately measuring propagation time fluctuations. It relates to a method for measuring properties.

The subject of the present invention is, for example, minute fluctuations in the propagation time, and one of the causes of the minute fluctuations is delay time fluctuation due to the influence of temperature and humidity in the atmosphere on the propagation path. This is part of meteorological measurements such as rainfall, and it is expected to capture changes in the amount of water vapor that is the source of clouds such as guerrilla heavy rain.
The target is to detect the fluctuation and deflection of the transmission point, the detection of the fluctuation of a high-rise building using a reflected wave, the positioning of the reception point, etc., and the measurement resolution is set to several millimeters.

  As a conventional example for accurately estimating transmission path characteristics, for example, Patent Document 1 (Japanese Patent Laid-Open No. 2010-158003) discloses a multicarrier signal such as an OFDM (orthogonal frequency-division multiplexing) signal. A method and system for determining a channel frequency response corresponding to is disclosed. In this disclosure, the following procedure is described. (1) First, an OFDM signal having a pilot signal or a pilot is received. (2) Next, the channel frequency response of the pilot carrier frequency is determined from the pilot. (3) Bandpass filtering / masking noise in the time domain channel frequency to transform the filtered / masked channel frequency response into the frequency domain. (4) Frequency interpolate the filtered / masked channel frequency response to generate a channel response for the remaining carrier frequencies.

  Further, for example, Patent Document 2 (Japanese Patent Application Laid-Open No. 2004-266814) obtains a communication apparatus that can realize highly accurate transmission path estimation even in a communication environment where intersymbol interference and intercarrier interference exist. There is a disclosure aimed at that. This disclosure is, for example, a communication apparatus on the receiving side that demodulates a received signal in which a pilot signal is repeatedly inserted at a constant period using, for example, an OFDM signal, and (1) a frequency characteristic based on the pilot signal extracted after Fourier transform (2) delay profile generation means for generating a delay profile by performing inverse Fourier transform on the frequency characteristics, and (3) a predetermined threshold based on the delay profile. Interference component removing means for removing an interference signal component equal to or less than the value; and (4) performing a Fourier transform on the time-axis signal based on the delay profile after removing the interference component, A channel estimation value generating means for generating a channel estimation value).

  An example in which an OFDM signal is used for ranging is disclosed in Patent Document 3 (WO 200501020600A2). This disclosure relates to a method for estimating the distance between two communication stations, (1) receiving the OFDM signal from the second communication station at the first communication station, (2) performing received signal processing and demodulation, and (3) The received signal is classified for each transmission path component, (4) the component of the shortest path component is determined, and (5) the propagation time is evaluated to calculate the distance between the two communication stations.

  As is well known, transmission and reception in OFDM television broadcasting in terrestrial digital broadcasting are performed as follows. First, an OFDM signal is a transmission signal for transmitting and receiving a signal using a number of synchronized carriers having a certain frequency difference. FIG. 1 shows a procedure for transmitting a digital signal using this signal. First, in order to improve the data transmission error rate, the digital signal is replaced by a symbol mapper, the digital signal is assigned to each of the above-mentioned carrier waves, and each discrete carrier wave having an amplitude according to the assigned digital signal is obtained. By performing Fourier transform (IDFT) and superimposing the IDFT signals, parallel-serial conversion is performed to obtain an OFDM baseband signal. The OFDM baseband signal thus obtained is orthogonally modulated on the carrier wave, then amplified and output to the transmission line.

On the receiving side, as shown in FIG. 2, the reception signal obtained from the transmission path is amplified and removed from the noise at the front end unit, and after orthogonal demodulation, the demodulated signal is digitalized by sampling. In order to remove the influence of the transmission line characteristics after performing the serial-parallel conversion that performs the reverse operation of the parallel-serial conversion of the transmitter, performing the discrete Fourier transform (DFT) on the parallelized signal After performing equalization processing, parallel-to-serial conversion and superposition, determination for digitization is performed.
This equalization process is a process for removing the rotation caused by the transmission path of signal points in the signal space diagram.

In receiving such an OFDM signal, it is known that the following synchronization is required.
1. Symbol timing synchronization 2. Carrier frequency synchronization Sampling frequency synchronization It is known that these synchronizations can be obtained by a method using correlation of guard intervals, for example.

In order to further improve the synchronization performance, pilot symbols are used. For example, the following pilot symbols are known.
1. Continuous pilot symbol 2. Pilot carrier

The transmission path characteristics necessary for performing the equalization process are estimated by the following method.
A method in which data symbols are mixed and distributed (an arrangement called a scattered pilot symbol). In this case, transmission path characteristics for data symbols are estimated by interpolation using pilot symbols distributed in the time and frequency directions.
There is also a method of concentrating and arranging at the top of the frame, which is a method used in a wireless LAN or the like. In the case of a wireless LAN, processing is performed in units of packets, and therefore a header for equalization, synchronization, and estimation of transmission path characteristics is added to the beginning of the packet. The receiver estimates the transmission path using the pilot symbols in the header and uses it for equalization processing.

Although there is a method based on blind estimation, this method is a method for estimating transmission path characteristics by performing statistical processing of transmission contents, and therefore it is difficult to apply in the present invention, and detailed description thereof is omitted.
However, if the transmission content conforms to the pilot signal or pilot symbol, it should be considered that the pilot signal or pilot symbol is substantially used.

  Non-Patent Document 1 (Mizuno et al., “Study on Positioning System Using Digital Terrestrial Broadcasting”) has a report of obtaining a delay profile by synchronously demodulating a received signal, and measuring a distance variation from the phase variation. . In this method, an error of about 0.3 m was generated even when a high-accuracy frequency reference was used.

  In addition to white noise caused by thermal noise of the transmitter / receiver and urban noise in the propagation path, the error factors in the above report include (a) the effects of frequency stability and phase noise of the transmission system and the reference oscillator on the transmission side, (B) Influence of frequency stability and phase noise of the reference oscillator on the measurement side, and (c) Influence of temperature characteristics of the receiving cable. In addition, the effects of (d) bending or shaking of the transmitting antenna, (e) behavior of the receiving antenna, (f) ground reflection, multipath, fading, etc. are considered.

  As a countermeasure against the above-described phase noise, for example, it is effective to apply a phase noise correction method using a terrestrial digital broadcast retransmission signal (pass-through signal) of a CATV network which is a large-scale wired network. However, there are problems that the number of measurement signals increases and measurement is limited to the CATV service area. As a countermeasure against temperature fluctuation of the receiving cable, a temperature fluctuation compensation method using a loopback cable on the receiving side is effective.

JP 2010-158003 A JP 2004-266814 A International Publication No. 2005/1020600 Specification

Mizuno et al., “Study on positioning system using terrestrial digital broadcasting”, Eizo Technical Report, Vol.33, No.10, pp.17-20, BCT2009-37 (Feb.2009)

  This is to suppress phase noise that hinders measurement of propagation fluctuations and to capture propagation time fluctuations between a transmission point and a reception point.

  In summary, the present invention relates to phase noise that hinders measurement of propagation fluctuations: (1) precisely synchronizes multiple received digital broadcast channels, (2) obtains delay profile from transfer function, (3) A technique is used in which the phase noise on the transmission / reception side is canceled by directly obtaining the propagation time fluctuation from the phase difference. (4) By this method, the phase noise of the transmission / reception device is canceled, and it becomes possible to capture the propagation time fluctuation between the transmission point and the reception point.

Therefore, the signal propagation characteristic measurement method of the present invention transmits an OFDM signal including a pilot signal from a transmission point, receives the OFDM signal at the reception point, and changes the phase difference between the transmission point and the reception point. A signal propagation characteristic measurement method for measuring a change in a distance between a transmission point and a reception point or a change in a signal propagation time from a transmission point to a reception point,
For two pilot signal groups which are pilot signals of OFDM signals and belong to different frequency bands, a change in phase difference, which is a phase difference in the delay profile of each pilot signal group, is detected, and the above transmission point to reception point are detected. A procedure for extracting the change in the distance of the signal or the change in the signal propagation time from the transmission point to the reception point,
The antenna trajectory of the transmission point is tracked at a plurality of reception points located from the transmission point within a predetermined distance that can ignore the influence on the change in the phase difference due to the variation in the transmission path of the OFDM signal. ,
Using the reception points where the line segments extending from the transmission point to each of the plurality of reception points intersect with each other at a significant angle, the locus is plotted by synchronizing the measurement time of the measurement data at the plurality of reception points. Including procedures .

  The pilot signals are a relatively high frequency pilot signal group selected from one OFDM signal and a relatively low frequency pilot signal group.

The phase change of the pilot signal of the OFDM signal is a change of a rotation angle in a signal space diagram in orthogonal modulation.
The delay profile is normally used for equalization processing, but is processing for returning the rotation angle in the signal space diagram to a normal state.

In addition, when a common OFDM signal is received between the first and second reception points, the phase difference of the pilot signal between the reception points is changed between the transmission point and the first reception point. A difference between a change in phase difference (first phase difference change) and a change in phase difference between the transmission point and the second reception point (second phase difference change) is used.
Even when the transmission point and the first and second reception points are not in a straight line, it can be assumed that they are almost in a straight line, and the radio wave propagation state changes sufficiently slowly. As for the difference between the first phase difference change and the second phase difference change, the same result can be obtained depending on whether or not the above three members are on a straight line.

In addition, the antenna trajectory of the transmission point is obtained at a plurality of reception points located from the transmission point within a predetermined distance that can ignore the influence on the change in the phase difference due to the variation in the transmission path of the OFDM signal. To track,
Using the reception points where the line segments extending from the transmission point to each of the plurality of reception points intersect with each other at a significant angle, the locus is plotted by synchronizing the measurement time of the measurement data at the plurality of reception points. It is characterized by being.

The method further includes a procedure of subtracting the influence of the antenna locus from the change in phase difference detected from the delay profile of the OFDM signal.

  According to the present invention, phase noise and the like that are obstructive factors are suppressed in order to realize the measurement of minute fluctuations in propagation characteristics between transmission points and observation points (carrier wave propagation time fluctuations) by using digital broadcasting secondarily. I was able to do it. For example, it has become possible to suppress phase noise and the like using reception signals of a plurality of channels transmitted from the same transmission point. In addition, by receiving signals for one channel of digital broadcasting (single channel), it is possible to cancel the phase noise on the transmission / reception side even if there is only one measurement point by dividing the channel. .

It is a block diagram which shows the transmission procedure in OFDM television broadcasting in terrestrial digital broadcasting. It is a block diagram which shows the reception procedure in OFDM television broadcasting in terrestrial digital broadcasting. It is a figure which shows the systematic diagram of the measurement in this invention. It is a figure which shows the table | surface about the parameter in the example of computer simulation. It is a figure which shows the model assumed in order to demonstrate the actual evaluation of a phase noise for a computer simulation as an example. In the example of computer simulation, it is a figure which shows the result of calculating | requiring the difference (DELTA) (tau) of propagation time fluctuation | variation according to the process procedure of this invention, (a) is a phase noise, (b) is a fluctuation | variation of a transmission antenna, (c) is a water vapor | steam fluctuation | variation. The propagation time variation simulating the range, (d) shows the generated received signal. In the computer simulation example, it is shown that the minute fluctuations obtained by suppressing the phase noise of the oscillator and the like are extracted by the sum Δτ + τ _d + σ _n of the propagation time fluctuation components between A and B obtained by the processing procedure of the present invention. In the figure, (a) C / N = 25 dB, Span = 24 MHz, (b) C / N = 25 dB, Span = 6 MHz, (c) C / N = 15 dB, Span = 24 MHz, (d) Shows the analysis result of the sum Δτ + τ _d + σ _n when C / N = 15 dB and Span = 6 MHz. (E) is a 20-second moving average of (b), where C / N = 25 dB, Span = 6 MHz, (f ) Is the 20-second moving average of (d), where C / N = 15 dB and Span = 6 MHz. It is a figure which shows arrangement | positioning of the apparatus etc. in the analysis simulation of the movement behavior detection of a transmission point, (a) to the position which pinches | interposes a transmission antenna on a horizontal plane on the straight line with respect to the transmission point at the reception points A and B (B) shows an overview of transmission points and reception points. It is a figure which shows the example of a parameter in the analysis simulation of the movement behavior detection of a transmission point. (A), (b) is a figure which shows the fluctuation | variation of the transmission antenna which should be detected in an analysis simulation, (c) shows the setting value of fluctuation, (d) shows the detection value of fluctuation. It is a block diagram for demonstrating the procedure which observes the propagation time fluctuation | variation of space only with the signal of a single channel. It is a figure for demonstrating the parameter at the time of observing the propagation time fluctuation of space only with the signal of a single channel. It is a figure for demonstrating the example of the value of the parameter at the time of observing the propagation time fluctuation of space only with the signal of a single channel. It is a figure which shows the detection example of the analysis simulation which observes the propagation time fluctuation of space only with the signal of a single channel, (a) is a phase noise, (b) is a fluctuation | variation of a transmission antenna, (c) is a water vapor | steam fluctuation | variation. Simulated propagation time fluctuation, (d) shows the generated received signal. A view showing a detection example of analysis simulation performed only by observation of the propagation time variation of spatial signal of a single channel, with a propagation time variation component sum Δτ + τ _h + σ _n, given by suppressing the oscillator phase noise, etc. a diagram showing an example of extraction of a slight change, (a) the analysis result of the sum Δτ + τ _h + σ _n obtained is, (b) is an enlarged view of a portion of (a), 20 seconds of (c) is (a) Shows moving average. It is a figure which shows the arrangement | positioning condition used in the case of the analysis simulation of the behavior of the transmission point shaking. (A) is that it is arrangement | positioning that the line segment which connects the receiving point A and B point and a transmission point is orthogonally crossed horizontally. (B) shows an overview of a transmission point and a reception point. It is a figure which shows the example of a parameter of the analysis simulation of the behavior of the shaking of a transmission point. It is a figure which shows (a) the setting value which should be detected, and (b) the example of a detection by the analysis simulation of the behavior of the fluctuation | variation of a transmission point.

  Embodiments of the present invention will be described below in detail with reference to the drawings. In the following description, devices having the same function or similar functions are denoted by the same reference numerals unless there is a special reason.

First, the basic knowledge of the OFDM signal necessary for describing the present invention will be described below.
For an OFDM baseband signal, N is the number of subcarriers, f ₀ is the carrier frequency interval, τ _g is the guard interval length, τ _e is the effective symbol length, τ _s is the symbol length (τ _s = τ _g + τ _e ), When d (m, l ) is an OFDM symbol, l (el) is a carrier number, m is a symbol number, and g (t−mτ _s ) is a symbol interval gate, it is expressed by the following equation.

Further, the received signal, the k as a channel number, with respect to transmitted from the transmitting station at a carrier frequency f _k signal, there are variations in the change and the carrier frequency of the symbol length, also there is a white noise on the transmission line When there is a minute variation factor such as However, ΔT is a symbol length error, and τ ′ _s is a symbol length including the symbol length error (τ ′ _s = τ _s + ΔT). R is the amplitude, Δf _k is the carrier frequency error, Φ ′ _T (t) is the phase noise on the transmitting side, ζ ′ (t) is the distance variation between the transmitting and receiving points, C [m / s] is the speed of light, and T ′. (t) is the propagation time variation between transmission and reception, W (t) is white noise.

On the receiving side, the frequency of this signal is converted into a baseband signal. This frequency conversion can be performed by multiplying the received signal and the following signal. The baseband signal obtained by this operation is discretely sampled (sampled) at 1 / (Nf ₀ ) second intervals including the guard interval. For m + 1 th symbol, effective symbol part, and k from k = 0 to N-1, and k from k = -G of the guard interval portion to -1, for, k / (Nf ₀₎ discrete in time Sampling is performed. Here, G = Nτ _g / τ _e .

Next, an effective symbol is cut out from the symbol data (= continuation of guard interval data and effective symbol data) according to the result of symbol synchronization.
The effective symbol thus cut out can be subjected to Fourier transform to obtain the next received symbol for the l-th carrier.

  Here, h (m, l) is a term corresponding to a propagation profile, and w ′ (m, l) is a term corresponding to white noise.

  Here, when propagation is based on P multipaths, propagation profiles are superimposed on individual paths, so the propagation profile h (m, l) in the case of multipaths is as follows.

Here, τ _i is a delay time, but when P = 1, τ ₁ is τ. Also, Φ ^ _T is the phase noise on the transmission side, P is the number of multipath paths, the function T () is the time variation component of the propagation space, the function ζ () is the distance variation component between the transmission and reception points, and the function Φ ^ _R ( ) Is the phase noise on the receiving side, and θ ^ is the amount of phase rotation determined by the propagation path length and the initial phase difference between the carriers f _k and f ′ _k .

Further, the transfer function obtained from the received signal is as follows.

  The complex delay profile obtained by IDFT of the transfer function is as follows. In the following equation, γ is a delay time.

However, G (m, γ) is normalized by the number of carriers L.

  Since the time resolution of the delay profile is determined by the signal bandwidth, if the bandwidth is 5.6 MHz, the time resolution is about 0.17857 μsec, and the propagation path difference is about 53.57 m.

The above transfer function is affected by phase noise, carrier frequency error, whiteness noise, etc. even when reception is only a direct wave.
For example, for simplicity, the demodulated carrier f ′ _k is equal to the carrier frequency f _k , the carrier frequency error Δf _k is zero, frame synchronization, symbol synchronization, and clock recovery are ideal, and τ = 0 and ΔT = 0. Assuming that the wave is only a direct wave (P = 1), the complex delay profile is expressed by the following equation.

Here, w ^ (m, γ) is the IDFT of the noise component of the transfer function.

  When the displacement position Δγ of the delay time of the arriving wave to be observed is obtained from the delay profile and the temporal change of the phase at that position is observed, the following equation is obtained as the phase fluctuation of the arriving wave.

Here, angle (z) is a function for obtaining the phase angle from the complex number z, and is defined below.

Here, since the term including the carrier frequency interval f ₀ in Equation 7 is helically twisted on the frequency axis of the transfer function according to the variation of the propagation distance and propagation time, the phase of the delay profile rotates. It can be seen that the peak position moves with (that is, the phase is shifted).

In other sections

Are phase fluctuation components that do not affect the frequency characteristics, and if the distance (mτ _S ) between the transmission and reception points and the propagation time T (mτ _S ) change, the phase of the entire transfer function rotates accordingly (that is, phase shift) It is understood.

In the following, first, a phase noise correction method using a plurality of channels will be described.

  In broadcast radio waves, it is said that the propagation time fluctuation caused by the change in the amount of water vapor in the atmosphere is about 360 degrees in terms of phase. In the actual measurement, for example, there is a report that the result of actually measuring the phase fluctuation of the radio wave using five Rb (rubidium) oscillators as a signal source was about 180 degrees / hour on average. However, considering the phase change when the receiving cable (5D2V) is further exposed to the outside air, when the cable length is 50 m and the temperature change is 20 ° C., the phase change is about 222 degrees. From this report, it can be seen that in order to obtain precise observations, it is necessary to precisely correct phase noise with a large fluctuation range. It is obvious that phase noise and the like are independent for each broadcasting station and need to be individually corrected.

  As a measure for precise measurement, it is conceivable to share a clock between measurement points, but it is not easy to use a shared clock between measurement points separated by several kilometers or more.

  The use of GPS is also conceivable, but in this case, the instantaneous phase may change by about 10 times that of the Rb oscillator, and the phase value may sometimes jump due to time correction processing, which is practically solved. Issues remain.

  A method of using a CATV line that is a cable broadcasting network is also known. This is a method of using a CATV line pass-through signal as a common reference signal as a method of establishing synchronization between reception points in order to suppress phase noise. Although this method is effective for the above-mentioned purpose, it has a drawback that the measurable range is limited to the laying area of the CATV line.

  Therefore, a method for measuring the phase noise will be described below using reception signals of a plurality of channels of digital terrestrial broadcasting waves transmitted from the same transmission point without using a CATV line.

First, the measurement / analysis model is as follows.
In the case where the incoming wave is only a direct wave (P = 1) upon reception of the broadcast wave, if the demodulated carrier f ′ _k is ideally synchronized with the carrier frequency f _k , Δf _k = 0, ΔT = 0, γ = Δγ = 0, and Equation 7 is simplified as follows.
However, G (m, 0) is normalized by the number of carriers L.

Further, the phase fluctuation element of the received symbol can be expressed by the following phase term.

Where φ _{T, k} is the phase noise component of the transmitting-side oscillator, φ _X is the phase noise component of the receiving-side oscillator, ξ _{X, k} is a value obtained by converting the propagation distance variation into a phase, and ψ _{X, k} is the propagation A value obtained by converting the propagation time variation in the path into a phase, θ _{X, k} is a phase rotation amount determined by the propagation path length and the initial phase difference between the carriers f _k and f ′ _k . The subscript X indicates the observation point, and k indicates the channel number.

Each item arranged is as follows.

A system diagram of the measurement is shown in FIG. In this measurement, two measurement points on a straight line as viewed from the transmission point are set, and measurement is performed with sufficient antenna height and line-of-sight conditions secured. Here, the subscript is a transmission point T, the reception points A and B, k is an arbitrary channel number in the UHF band for broadcasting, and the symbol position is an arbitrary integer at a certain moment. For simplicity, T, A, and B are arranged on a straight line.
Φ _{T, k} , φ _A , φ _B are phase noise components of the respective oscillators, and ξ _{T, k} is a value obtained by converting the propagation distance variation of the transmission point due to the shaking of the antenna into a phase, ψ _{A, k} , ψ _{B and k} are values obtained by converting the propagation time fluctuation in the propagation path between the transmission point and the measurement point into a phase. On the measurement side, a plurality of channels are sampled in a wide band at once, and φ _A and φ _B are common to each receiving point.

  Through the analysis of the transfer function and complex delay profile from the received signal, the phase component of the incoming wave is expressed by the following equations 14 and 15.

From Equations 14 and 15, the difference in the phase component of the received signal between points A and B is given by the following equation.

The value to be suppressed is the phase noise φ _B −φ _A , and the values obtained as observation results are the fluctuation component ψ _{B, k} −ψ _{A, k} and the initial phase θ _{B, k} −θ _{A, k} .

Next, an analysis method using a plurality of channels will be described.
In this method, as a pre-processing, it is necessary to perform synchronization processing for each received channel and take the cross-correlation of the obtained delay profile to match the time position, and this processing accuracy affects the analysis result. , ΔT = 0, Δf _k = 0, γ _{A, k} = γ _{B, k} .

As a method for suppressing phase noise, reception signals of a plurality of channels transmitted from the same transmission point are simultaneously received at each measurement point, and a difference Δτ of propagation time fluctuations independent of the channel frequency is obtained between the measurement points.
In addition, since the phase drift due to the phase noise does not change within a short time, it is possible to effectively synchronize the observation time between the measurement points by adjusting the time with a GPS or a radio clock.

  From Equation 14, the phase difference between the α channel and the β channel is obtained for the received signal at point A. Here, the subscript (β−α) represents a difference between channels.

  Similarly, the phase difference between the α channel and the β channel is obtained for the received signal at point B from Equation 15.

Next, the difference between the two points is obtained from Equations 15 and 16.

However, it follows the abbreviation of Equation 20, and the last term is an error due to white noise.

Then, the fixed delay time difference τ _d due to the propagation path difference, the propagation time fluctuation difference Δτ, and the error σ _n due to the white noise are as follows when converted from the phase to the propagation time.

  From Equations 19 and 21, the value of the phase fluctuation at an arbitrary channel k between points A and B can be obtained by the following equation.

The final term on the right side of Equation 22 is an estimated value of white noise.

As described above, the phase noise is suppressed, and the difference in propagation fluctuation between the target points A and B is obtained.
When the white noise σ _n is larger than Δτ, the analysis result is affected, so that it is necessary to evaluate the effectiveness regarding the C / N (carrier power / noise power) ratio.

The C / N of the signal band is signal-to-noise power, and since the OFDM signal is stochastically constant power in the band, the C / N does not depend on the bandwidth. However, since the number of carriers that can be referred to in the band is limited, the number of delay profiles affects the Fourier transform. Since SP has 1872 subcarriers in the band, if the amplitude of the signal transfer function is normalized to 1, the signal amplitude of the delay profile is 43.3 and the average amplitude of noise is 1. Therefore, when the delay profile is obtained, the gain is improved by 32.8 dB.
Also, since the time / distance variation is obtained from the change difference using a plurality of channels, the value of the time / distance variation that does not depend on the frequency is proportional to the difference in the frequency of the channel to be compared to obtain ψ _Z increases but w _Z does not change. Therefore, the gain becomes more advantageous as the frequency interval of the channel to be compared is larger. Each time the channel interval to be compared is separated by 6 MHz, it is improved by 3 dB.

Next, the actual phase noise evaluation will be described by taking a computer simulation as an example.
In this simulation, in order to generate phase noise data, a model is assumed in which an Rb oscillator is connected to each of the modulator and the RF signal recording device as shown in FIG. In addition, phase noise φ _A , φ _B , φ _C , φ _D , φ _α , φ _β is recorded once by connecting the modulation signal to the RF signal recording device, and synchronous demodulation is performed to measure the time variation of the phase of the main wave. Use the extracted data. However, these are independent fluctuation data according to time.

Further, the measurement conditions are assumed to be reception at measurement points A and B located on one straight line from the transmission point shown in FIG. 3, and phase noise, spatial variation, initial phase, whiteness noise for the parameters shown in FIG. And the like, signals Ph _A and Ph _B are generated for channels α and β.

The results of obtaining the propagation time variation difference Δτ according to the above processing procedure are shown in FIGS.
6A shows the phase noise, FIG. 6B shows the fluctuation of the transmission antenna, FIG. 6C shows the propagation time fluctuation simulating the range of the water vapor fluctuation, and FIG. 6D shows the generated received signal.
FIG. 7 shows the sum Δτ + τ _d + σ _n of the propagation time fluctuation components between A and B obtained by the proposed method, and the minute fluctuation given by suppressing the phase noise of the oscillator is extracted.
FIG. 7A shows C / N = 25 dB, Span = 24 MHz, FIG. 7B shows C / N = 25 dB, Span = 6 MHz, and FIG. 7C shows C / N = 15 dB, Span = 24 MHz. 7 (d) is an analysis result of the sum Δτ + τ _d + σ _n when C / N = 15 dB and Span = 6 MHz.
The wider the frequency interval of the channels to be compared, the more advantageous for noise. However, if the average value of the short section is taken as shown in FIGS. 7 (e) and 7 (f), the noise is improved.
By the way, the reception environment of C / N = 15 dB is a reception level at which 64QAM digital broadcasting cannot be viewed.

[Analysis simulation of detection of shaking behavior of transmission point]
Since the present invention can also detect the movement behavior of the transmission point, its analysis simulation is shown below.

The condition in FIG. 8 is to measure behavior such as the deflection of the antenna at the transmission point, and points A and B are arranged at horizontal positions that sandwich the transmission antenna on a straight line passing through the transmission point. As the distance, the vicinity of the transmission point that is not easily affected by the propagation space is desirable.
In general, it is assumed that the transmitting antenna is rotating in an ellipse, but as shown in FIG. 8, if fluctuations are measured from two different directions of AB and CD passing through the transmission points, the elliptical shape is measured. The trajectory can be specified.
The principle is as follows. First, the amplitude of fluctuation in two directions connecting two pairs of two points is obtained by measurement between the two pairs of points. Although the constraint conditions on the outer periphery of the ellipse are determined by measurement in two directions, there can be infinite number of ellipse motions.
Therefore, one ellipse can be identified by plotting the locus by synchronizing the measurement times of the measurement data in two directions with respect to the period of amplitude fluctuation. At this time, the angles formed by AB and CD do not need to be orthogonal.
Here, for the sake of simplicity, the transmission point is assumed to oscillate at the resonance period of the transmission tower due to the influence of wind and the like, and the spiers are assumed to swing in a circle on a horizontal plane. Moreover, the measurement points C and D are arranged on a line orthogonal to the line AB on the same horizontal plane as the points A and B. That is, (θ _CD −θ _AB = π / 2).

For the parameters shown in FIG. 9, phase noise, spatial variation, initial phase, whiteness noise, and the like are given, and signals Ph _A , Ph _B , Ph _C , and Ph _D are generated for channels α and β. Then, the difference Δτ of the propagation time fluctuation is obtained by the above processing procedure.

  10 (a) and 10 (b) show the vibration of the transmitting antenna to be detected. However, as shown in FIG. 10 (c), the setting is such that a circle is drawn on the horizontal plane. Here, the fluctuation is shown as a phase variation of the carrier frequency.

In detecting the oscillation of the transmission antenna from the received signal, the phase fluctuation difference was extracted between A and B, and between C and D, and the sum Δτ + τ _d + σ _n was obtained.
First, the phase noise of the oscillator is cancelled because it is common regardless of the reception direction. If the shake of the antenna is also in the same direction, it is canceled out in the same phase, but because the reception is in the opposite direction across the antenna, the phase is reversed and a fluctuation value twice as large is obtained. Accordingly, FIG. 10D shows the result of halving the obtained value.

  For example, if the behavior of the transmission point is obtained in this way, for example, when the phase fluctuation is obtained from the measurement at the A and B points, the behavior can be used for correction in the measurement at the A and B points. Even when the observation points A and B and the transmission point are not located on the same line, the difference between the data at the points A and B can be obtained.

  Since the present embodiment detects a delay profile using a large number of SPs (scattered pilot signals) as described above, the delay profile is measured and evaluated even at a reception intensity level that is so low that viewing of digital broadcasting is severe. Is possible. For this reason, if sufficient reception C / N is obtained under the installation conditions of the antenna, the influence of noise is reduced, and measurement with high resolution is possible. In other words, there is a sufficient area where the present invention can be applied even if it is outside the broadcast receivable area. Also, in the case of measurement with a low antenna height or measurement with a moving object, there are cases where measurement is often performed at a low reception intensity level, but application to this case is also possible.

[Method of finding the delay profile by dividing the band and correcting the phase noise]
Next, a method for correcting the phase noise by obtaining a delay profile of the transfer function divided into bands will be described.

The above-mentioned analysis method using multiple channels requires processing to synchronize each received channel and take the cross-correlation of the obtained delay profile to match the time position. This processing accuracy affects the analysis results. To do. Furthermore, it is necessary to broadcast a plurality of channels from the same transmission point, and observation is not possible in a broadcasting area with only one channel such as a prefectural area station.
Moreover, in order to cancel out the phase noise on the transmission side, two observation points are required on a straight line seen from the transmission point, and the measurement time must be synchronized.
Even when two broadcasting channels are used, if symbol timing generation and frequency conversion to a carrier wave can be performed using a frequency reference common to each channel, the transmission side and the reception side can be determined at the stage of obtaining the phase difference of the delay profile. The phase noise is canceled at the same time to determine the amount of fluctuation in the propagation time. To realize this, it is necessary to change the broadcasting equipment to use a common frequency reference.

An example that can solve this problem is shown below. this is,
(A) A digital broadcast signal for one channel (single channel) is received and precisely synchronized and demodulated. (B) The transfer function is divided on the frequency axis to obtain respective delay profiles.
(C) By directly obtaining the propagation distance and propagation time fluctuation from the phase difference between the main waves,
This is a technique that can cancel the phase noise on the transmission / reception side even if there is only one measurement point (that is, reception point).

  If there is only one measurement point, even if the frequency reference is shared and the phase noise is canceled, the distance fluctuation and propagation time fluctuation are preserved and components such as the transmission antenna fluctuation remain. In order to obtain only the propagation time fluctuation, it is necessary to correct a distance fluctuation component such as a fluctuation of the transmitting antenna.

Such a shaking of the transmitting antenna is
(A) a method of subtracting from data separately observed in a relatively close vicinity of the antenna, which is a position where the propagation time fluctuation can be ignored,
(B) By taking a difference between the observation data of two points on a straight line from the transmission point and canceling the common transmission antenna fluctuation,
It is possible to eliminate it.

[Modeling of OFDM in-band division]
First, OFDM in-band division is modeled as follows.
FIG. 11 is a block diagram for explaining a procedure for observing a variation in spatial propagation time using only a single channel signal. Since an OFDM signal is a multicarrier, even if a single channel OFDM signal is divided into two on the frequency axis, it is equivalent to using two channels of signals.

Disadvantages of the method of dividing the channel into two are that the gain of the phase difference to be obtained is proportional to the frequency difference of the channels to be compared, the delay time resolution is about ½, and the gain of the complex delay profile is deteriorated by about 3 dB. Is a point.
On the other hand, the advantage is that the signal processing load is reduced because the demodulation timing is the same, and the delay time matching error does not occur because the compared delay profiles are the same. In addition, observation is possible even in broadcasting areas where only a single channel can be received, such as prefectural stations.

  For example, in the case of the current terrestrial digital broadcasting, the 6 MHz frequency band for a single channel is divided into two to make 3 MHz. In digital terrestrial broadcasting, the 6 MHz bandwidth is equally divided into 14 segments, of which 13 segments are assigned effective signal bandwidths, and the remaining one segment is assigned as a guard band with adjacent channels, half on both sides of the bandwidth. ing. Therefore, 6 MHz / (14 × 2) (≈214 kHz) is blanked on both sides of the band.

In the present embodiment, one band (for one segment) at the center of the effective signal band is excluded, and two bands of six segments on the upper and lower sides are formed. In this case, the centers of the bands are 3 MHz apart. The lower channel number of the received signal at point A is described as kL, and the upper channel number is described as kH.
When only a direct wave is received, Δf _k = 0, ΔT = 0, and γ = Δγ = 0 if synchronization is ideal.
In that case, the delay profile of the incoming wave of the kL channel observed at the point A is expressed as follows from Equation 7.

However, G _{A, kL} (m, 0) is normalized by the number of carriers L.
The phase fluctuation element of the received symbol can be expressed by the following phase term.

Where φ _{T, k} is the phase noise component of the transmitting-side oscillator, φ _X is the phase noise component of the receiving-side oscillator, and ξ _{T, k} is the value obtained by converting the propagation distance variation at the transmitting point due to the antenna fluctuation into a phase. , Ξ _{A, k} is a value obtained by converting the propagation distance fluctuation due to the movement of the position of the receiving point or the receiving antenna into a phase, ψ _{X, k} is a value obtained by converting the propagation time fluctuation in the propagation path into a phase, θ _{X, k} is a phase rotation amount determined by the propagation path length and the initial phase difference between the carrier waves f _k and f ′ _k . The subscript X indicates an observation point, and kH and kL indicate channel numbers.

[A technique for dividing the transfer function of a single channel into bands]
From Equation 24, the difference between kH and kL of the received signal band at point A is given by the following equation.

Here, the subscript (kH−kL) represents a difference between channels.

From Equation 25, it can be seen that the phase noise is canceled out, and the propagation distance fluctuation between the transmission point and the measurement point and the noise remaining in the propagation time fluctuation are obtained. If the transmitting antenna and the receiving antenna are not physically shaken or bent, the fluctuation between the transmitting and receiving points is obtained.
Then, the difference Δτ of the propagation time fluctuation, the difference τ _h of the propagation distance fluctuation, and the error σ _n due to the white noise are as follows when converted from the phase to the propagation time.

Since the sum (Δτ + τ _h + σ _n ) that does not depend on the frequency has been obtained, the value of the phase fluctuation at an arbitrary channel k between the points A can be obtained from the following equation from Equations 23 and 24.

Here, w ″ _{A, k} is an estimated value of white noise.

As described above, the phase noise is canceled, the difference in propagation fluctuation between the target points A and B is obtained, and the white noise σ _n is sufficient for Δτ when there is no fluctuation of the transmission point or movement of the reception point. If it is small, the fluctuation of the propagation time can be estimated.

[When analyzing using measurement results between two points]
In addition to the measurement at the point A, the difference between the received signal band kH and kL in the measurement at the point B is as follows.

Next, a difference is further calculated for Equation 25 and Equation 28.

ψ _{Z (kH−kL)} is a value obtained by converting a propagation time variation in space into a phase, and ξ _{Z (kH−kL)} is a value obtained by converting a propagation distance variation due to movement of the position of the reception point into a phase. w _{Z (kH−kL)} is an error due to noise.
Therefore, the difference Δτ of the propagation time fluctuation due to the path difference between the two points is as follows.

  It can be seen from Equation 29 and Equation 30 that the propagation time fluctuation and propagation distance fluctuation caused by the path difference between points A and B are as follows.

Therefore, the values of phase fluctuation and phase rotation due to the path difference are as follows.

Here, w ″ _{Z, k} is an estimated value of white noise. When the observation point is not on the straight line from the transmission point, the arrival angle of the horizontal plane from the transmission point is different. In such a case, it is necessary to sequentially obtain and correct the distance change amount in each observation point direction.

Thus, the phase noise and the variation of the transmission point are offset to obtain the difference of variation between the target points A and B. If there is no movement of the reception points A and B, the propagation time variation between A and B is obtained. . Moreover, if there is no propagation time fluctuation, a change difference in the distance between A and B with respect to the transmission point is obtained. When the white noise σ _n is larger than Δτ, the analysis result is affected, so that the effectiveness of C / N needs to be evaluated.

[Influence of noise]
The C / N of the signal band is signal-to-noise power, and since the OFDM signal is stochastically constant power in the band, the C / N does not depend on the bandwidth.
Regarding the delay profile, the number of carriers that can be referred to in the band is limited, and the number affects the Fourier transform. Since SP has 1872 subcarriers in the band, if the amplitude of the transfer function of the signal is normalized to 1, the signal amplitude of the delay profile is 43.3 and the average amplitude of noise is 1. Therefore, when the delay profile is obtained, the gain is improved by 32.8 dB.

  However, when dividing the band of one channel, it is necessary to divide the SP, and when the central segment is not used in the two divisions, 1872 × 6/13 = 864 is the number of SPs that can be used. In this case, the signal amplitude of the delay profile is 29.4, and the average noise amplitude is 1. Therefore, when the delay profile is obtained, the gain is improved by 29.4 dB.

[Evaluation by simulation]
The present embodiment will be described more specifically along with computer simulation. First, in order to generate phase noise data, a model is assumed in which an Rb oscillator is connected to each of the modulator and the RF signal recording device as shown in FIG. Here, the phase noises φ _A , φ _B , and φ _k use data obtained by connecting a modulation signal to an RF signal recording device, temporarily recording it, performing synchronous demodulation, and extracting the time variation of the phase of the main wave. However, these are independent fluctuation data according to time.

[Analysis simulation of variation between measurement points]
The measurement conditions are assumed to be received at the measurement point A shown in FIG. 11, and phase noise, spatial variation, initial phase, whiteness noise, etc. are given to the parameters shown in FIGS. 12 and 13, and the signals Ph _A and Ph _B are given. Are generated for channel frequencies kL and kH. Then, the difference Δτ of the propagation time fluctuation is obtained by the above processing procedure.

14A shows the phase noise, FIG. 14B shows the fluctuation of the transmitting antenna, FIG. 14C shows the propagation time fluctuation simulating the water vapor fluctuation, and FIG. 14D shows the generated received signal.
FIG. 15 shows a sum Δτ + τ _h + σ _n which is a propagation time fluctuation component at point A obtained by the method of the present invention, and a minute fluctuation given by suppressing the phase noise of the oscillator is extracted. FIG. 15A shows an analysis result of the obtained sum Δτ + τ _h + σ _n , and FIG. 15B is an enlarged view of a part of FIG. FIG. 15C is a 20-second moving average of FIG. 15A, which is improved against noise.

[Analysis simulation of shaking behavior of transmission point]
The condition in FIG. 16 is to measure behavior such as the deflection of the antenna at the transmission point, and the points A and B are arranged at positions as orthogonal as possible on the horizontal plane with the transmission point as the center. The distance is in the vicinity of the transmission point that is not easily affected by propagation space fluctuations.

In general, the transmission point may vibrate at a resonance period due to the influence of wind or the like. At this time, it is assumed that the transmission antenna is rotating in an ellipse. However, as shown in FIG. 16, the locus of the ellipse can be specified by measuring the above-mentioned fluctuations from two different directions connecting the transmission point and A or B.
The principle is similar to the above case and is as follows. First, the amplitude of the fluctuation in the two directions connecting the transmission point and the measurement point is obtained by measurement. This establishes the constraint condition of the outer periphery of the ellipse, but there can be infinite number of ellipse motions. Therefore, one ellipse can be identified by plotting the locus by synchronizing the measurement times of the measurement data in the two directions with respect to the period of amplitude fluctuation. At this time, the angle formed by A and B does not need to be orthogonal.
Here, for the sake of simplicity, a swing that draws a circle on a horizontal plane is assumed. Then, the measurement points A and B are arranged at an angle perpendicular to the horizontal plane. That is, (θ _B −θ _A = π / 2).

Signals Ph _A and Ph _B are generated for channel frequencies kL and kH by giving phase noise, spatial variation, initial phase, whiteness noise, and the like for the parameters shown in FIG. Then, the propagation time variation difference Δτ is obtained by the proposed processing procedure.

As described above, the set transmission antenna shake is a shake drawing a circle on a horizontal plane as shown in FIG. In this case, when the sum Δτ + τ _h + σ _n is obtained for A and B, the obtained result is the shaking component of the antenna shown in FIG. 18B, and the setting can be reproduced.

  If the behavior of the transmission point is obtained, it can be used as a correction value, so even if the observation point is not located on the same line when viewed from the transmission point, the difference between the data at the two points should be estimated and used for correction. Can do.

  The present invention has a wide range in which measurement is possible even at a reception level that is so low that viewing of digital broadcasting is severe. If sufficient reception C / N can be obtained under antenna installation conditions, the influence of noise is reduced and high. Measurement with resolution can be expected. In addition, it can be applied to measurement with a low antenna height and measurement with a moving object.

  By installing many observation points in a wide area that can be seen from the transmission point and measuring the propagation delay time between the observation points as described above, for example, the distribution of temperature changes and humidity changes in the propagation path, or raindrops It becomes possible to monitor the distribution of scatterers such as in real time.

d (m, l ) OFDM symbol f ′ _k demodulated carrier f ₀ carrier frequency interval f _k carrier frequency g (t−mτ _s ) symbol period gate k channel number kH, kL channel number L number of carriers l (el) carrier number m time variation component of the symbol number N number of routes of sub-carrier number P multipath r amplitude T propagation time T () propagation space T'(t) propagation time variation w'between transmission and reception (m, l) the white noise w _{"Z , k} Estimated value of whiteness noise X Observation point, eg A, B
γ delay time Δf _k carrier frequency error ΔT symbol length error Δτ propagation time variation difference θ ^ phase rotation amount θ _{X, k} propagation path length and carrier determined by propagation path length and initial phase difference between carriers f _k and f ′ _{k A} phase rotation amount ξ _{A, k} determined by the initial phase difference between f _k and f ′ _k and a value obtained by converting a propagation distance variation due to movement of the position of the receiving point and shaking of the receiving antenna into a phase ξ _{T, k} transmitting point due to shaking of _{the k} antenna Ξ _{X, k} Value converted from propagation distance variation into phase σ _n Error due to white noise τ ′ _s Symbol length τ _d Fixed delay time difference τ _e Effective symbol length due to propagation path difference τ _g Guard interval length τ _h Propagation distance fluctuation difference τ _s Symbol length ζ Distance between transmission and reception points ζ () Distance fluctuation component between transmission and reception points ζ ′ (t) Distance fluctuation between transmission and reception points Φ ^ _R () Reception side phase noise _Φ'T (t) phase noise φ _{T, k} the transmission side of the oscillator on the transmission side of Phase noise component [psi _X phase noise component phi _X receiving side of the _oscillator, the value [psi _{A, k} obtained by converting the propagation time variation of the phase of the _k propagation path, between [psi _{B, k} transmission point and the measurement point propagation path of the Value converted from propagation time variation to phase

Claims (5)

An OFDM signal including a pilot signal is transmitted from the transmission point, the OFDM signal is received at the reception point, a change in the phase difference between the transmission point and the reception point is detected, and the distance between the transmission point and the reception point is detected. A signal propagation characteristic measurement method for measuring a change or a change in signal propagation time from a transmission point to a reception point,
For two pilot signal groups which are pilot signals of OFDM signals and belong to different frequency bands, a change in phase difference, which is a phase difference in the delay profile of each pilot signal group, is detected, and the above transmission point to reception point are detected. A procedure for extracting the change in the distance of the signal or the change in the signal propagation time from the transmission point to the reception point ,
The antenna trajectory of the transmission point is tracked at a plurality of reception points located from the transmission point within a predetermined distance that can ignore the influence on the change in the phase difference due to the variation in the transmission path of the OFDM signal. ,
Using the reception points where the line segments extending from the transmission point to each of the plurality of reception points intersect with each other at a significant angle, the locus is plotted by synchronizing the measurement time of the measurement data at the plurality of reception points. A method for measuring signal propagation characteristics characterized by including a procedure .   2. The signal propagation characteristic according to claim 1, wherein the pilot signals are a relatively high frequency pilot signal group selected from one OFDM signal and a relatively low frequency pilot signal group. Measurement method.   3. The signal propagation characteristic measuring method according to claim 1, wherein the phase change of the pilot signal of the OFDM signal is a change of a rotation angle in a signal space diagram in quadrature modulation.   When a common OFDM signal is received between the first and second reception points, the phase difference between the transmission point and the first reception point is changed as a change in the phase difference of the pilot signal between the reception points. The difference between a change (first phase difference change) and a phase difference change (second phase difference change) between the transmission point and the second reception point is used. A method for measuring the signal propagation characteristics described in 1. From a change in the phase difference detected from the delay profile of the OFDM signal, measurement of the signal propagation characteristics according to claim 1, any one of 4, characterized in that it comprises a procedure of subtracting the influence of the trajectory of the antenna .