JP2013251684A - Diffusive article position estimation device, diffusive article position estimation method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for estimating the position of a diffusive article from only measurement data and performing clustering of a wavelet group and diffusive article position estimation in connection with each other.SOLUTION: The method includes: calculating, for each particle, a distance between propagation data and cluster-assumption data, where the former is obtained from measurement data of an electromagnetic wave received by a receiver, while the latter is an assumption value of propagation data of a cluster assumed by using a diffusive article estimation position; correlating, for each particle, a smallest-distance cluster with another cluster; calculating, as an evaluation value, for each particle, a ratio of the sum of electric power of propagation data whose distance is equal to or less than a threshold to the sum of electric power of propagation data of all measurement points and all wavelets; performing resampling in accordance with the evaluation value to calculate a particle; determining, as estimation-target data, for each particle calculated by the resampling, propagation data whose distance is the largest; estimating, for each particle, the position of the diffusive article anew, using the estimation-target data; and assuming cluster assumption data of a new cluster corresponding to the estimated diffusive article.

Description

  The present invention uses the propagation data obtained from the measurement data of the radio wave received by the receiver, and the position of the reflector and scatterer (hereinafter abbreviated as “scatterer”) in the propagation path of the wave arriving at the receiver. It is related with the technique which estimates.

  In cellular radio, it is possible to know the rough propagation path of an elementary wave group by estimating the position of the scatterer in the propagation path as well as the reception power and direction of arrival of the elementary wave group that has arrived at the receiver. . It is known that knowing the propagation path of an elementary wave group is advantageous for grasping area design, channel fluctuation characteristics, and the like.

In order to estimate the propagation path of the wave group at each point in the area, normally, measurement is continuously performed using a channel sounder or the like in the area, and propagation data is obtained using the measurement data. When the measurement is performed at the point I, the propagation data of the elementary wave n (1 ≦ n ≦ N, where N is the number of elementary waves) at the measurement point i (1 ≦ i ≦ I) is generally _expressed as x _n (i). = {TOA _n (i), AOA _n (i), AOD _n (i), P _n (i)}. Each element in the propagation data includes propagation delay (ToA: Time of Arrival), arrival azimuth angle (AoA), outgoing angle (AoD: Angle of Departure), (reception) power (P : Power). Patent Document 1 is known as a conventional technique for obtaining propagation data using measurement data.

  In order to estimate the position of the incoming scatterer from the propagation data, Non-Patent Document 1 proposes a method using measurement data and a building map in the measurement area. In Non-Patent Document 1, a straight line is drawn from the position of the receiver toward the arrival azimuth angle of the elementary wave, and a point that intersects an obstacle such as a wall in the building map is estimated as the position of the scatterer.

  Also, the propagation data of elementary waves usually contains many errors, and many estimation errors occur when the position of the scatterer is estimated individually from the propagation data of each elementary wave. In order to reduce the estimation error, clustering (grouping) of elementary wave groups is required in advance. Non-patent document 2 proposes a clustering method (K-Power-Means) according to the similarity of propagation data of the wave, and non-patent document 3 proposes a clustering method by visual observation of the wave data.

JP 2011-14980 A

Poutanen, J .; Haneda, K .; Salmi, J .; Kolmonen, V.-M .; Richter, A .; Almers, P .; Vainikainen, P., “Development of measurement-based ray tracer for multi-link double directional propagation parameters ”, EuCAP 2009. 3rd European Conference on Antennas and Propagation, 2009, pp 2622-2626 Czink N, Ruiyuan Tian, Wyne S, Tufvesson F, Nuutinen J.-P, Ylitalo J, Bonek E, Molisch A.F., "Tracking Time-Variant Cluster Parameters in MIMO Channel Measurements", 2007, IEEE CHINACOM '07, pp22-24 Oestges, C. Clerckx, B., “Modeling Outdoor Macrocellular Clusters Based on 1.9-GHz Experimental Data”, Trans. On IEEE VTC2007-Fall, 2007, vol.55, Issue 5, pp.2821-2830

  In the scatterer position estimation in Non-Patent Document 1, a building map is required in addition to the measurement data. A building map is not always obtained, and an estimation error occurs when there is a difference between the actual building layout and the building map.

  In the clustering method in Non-Patent Document 2, the wave group is clustered only according to the similarity of the propagation data. Therefore, there is a possibility that element waves having different positions of scatterers (different propagation paths) are classified into the same cluster. In that case, the position of the scatterer cannot be estimated accurately.

  The clustering method in Non-Patent Document 3 is a method in which a human visually observes a wave group from propagation data and a building map. Objectivity is not guaranteed, such as different standards depending on the worker.

  An object of the present invention is to provide a technique for estimating the position of a scatterer only from measurement data and performing clustering of the wave group and scatterer position estimation in conjunction with each other.

  In order to solve the above-described problems, according to the first aspect of the present invention, the scatterer position estimation device includes, for each particle, propagation data obtained from radio wave measurement data received by a receiver, and scatterer estimation. A distance calculation unit that calculates the distance between the assumed cluster propagation data that is assumed using the position and the cluster assumption data, and a clustering unit that associates another cluster with the cluster having the smallest distance for each particle. And, for each particle, an evaluation value calculation unit that calculates, as an evaluation value, a ratio of the sum of power of propagation data whose distance is equal to or less than a threshold value in the sum of power of propagation data of all measurement points and all the elementary waves, and evaluation Resampler that performs resampling according to the value, and estimation target data that determines the propagation data with the longest distance as estimation target data for each particle determined by resampling A scatterer position estimation unit that estimates a new scatterer position using estimation target data for each particle and assumes cluster assumption data of a new cluster corresponding to the estimated scatterer; Including.

  In order to solve the above problems, according to the second aspect of the present invention, the scatterer position estimation method includes, for each particle, propagation data obtained from radio wave measurement data received by a receiver, and scatterer estimation. A distance calculating step for calculating a distance between the assumed cluster propagation data assumed by using the position and a cluster assumed data, and a clustering step for associating another cluster with a cluster having the smallest distance for each particle. And an evaluation value calculation step for calculating, as an evaluation value, the ratio of the sum of the power of the propagation data whose distance is equal to or less than the threshold to the sum of the power of the propagation data of all the measurement points and all the elementary waves for each particle, and the evaluation Resampling is performed according to the value, the resampling step for obtaining particles, and the propagation data with the largest distance for each particle obtained by resampling The estimation target data determination step to be determined, and for each particle, the estimation target data is used to newly estimate the position of the scatterer, and the scatterer position assuming the cluster assumption data of a new cluster corresponding to the estimated scatterer An estimation step.

  According to the present invention, since the position of the scatterer is estimated only from the measurement data and the clustering of the wave group and the scatterer position estimation are performed in conjunction with each other, a building map is not required, and the objectivity is maintained while maintaining the objectivity. In addition, the position of the scatterer can be estimated.

The figure for demonstrating the basic concept of 1st embodiment. The functional block diagram of a scatterer estimation apparatus. The functional block diagram of a scatterer position estimation part. The figure which shows the processing flow of a scatterer estimation apparatus. The figure for demonstrating the detail of the process in a scatterer position estimation part. The figure which shows the processing flow of a scatterer position estimation part.

  Hereinafter, embodiments of the present invention will be described. In the drawings used for the following description, constituent parts having the same function and steps for performing the same process are denoted by the same reference numerals, and redundant description is omitted.

<First embodiment>
FIG. 1 shows the basic concept of this embodiment. Since the wave scatterer position is a latent variable that cannot be obtained directly from the measurement data, P kinds of scatterer positions (L ₁ , L ₂ ,..., L _P ) are estimated within the scatterer existence range. A scatterer estimated position and parameters obtained from the scatterer estimated position (for example, a propagation delay from a transmitter to a scatterer, an outgoing angle of an elementary wave, and the like) are called particles (state model). In each particle U _p (1 ≦ p ≦ P, P is the number of particles used for estimation), cluster c (1 ≦ c) at each measurement point i (1 ≦ i ≦ I, where I is the number of measurement points) ≦ C, where C is the number of clusters used for estimation, and is the number of scatterers to be estimated (assumed to be C = 0 in the initial estimation value). Transition) can be assumed using the estimated position of the scatterer. The estimated cluster data obtained in this way is compared with the propagation data obtained from the actual measurement data, and corresponds to the assumed cluster data (assumed value 2 in the figure) close to the propagation data obtained from the actual measurement data. Judge particles as a promising estimation result.

FIG. 2 shows a functional block diagram of the scatterer position estimation apparatus 100 in the present embodiment. The scatterer position estimation apparatus 100 receives N pieces of propagation data x _n (i) obtained from measurement data. However, n is an index of a wave and 1 ≦ n ≦ N. The estimated positions of C scatterers are output using N pieces of propagation data x _n (i). In addition, it is good also as a structure provided with the propagation data estimation apparatus of patent document 1 in the front | former stage of the scatterer position estimation apparatus 100 by making measurement data into an input.

The scatterer position estimation apparatus 100 includes P distance calculation units 110 _p , P clustering units 120 _p , P evaluation value calculation units 130 _p that calculate evaluation values of each particle based on the clustering results, It includes a resampling unit 140 that samples according to the evaluation value for particles, P estimation target data determination units 150 _p , P scatterer position estimation units 160 _p, and a determination unit 170. Moreover, the scatterer position estimating unit 160 _p includes respective scatterers existence range estimation unit 161, a scatterer disposed portion 163 and the cluster parameter calculating unit 165 (see FIG. 3). Propagation data x _{n (i)} and the resampling unit 140 is common to all of the particles, each data and each section other than it is present individually for each particle U _p.

FIG. 4 shows a processing flow of the scatterer position estimation apparatus 100. It is assumed that measurement data is continuously acquired at measurement points I along the measurement course while moving the receiver. The measurement data at each measurement point is composed of N pieces of wave data, and the propagation data of the wave n at the measurement point i is expressed as x _n (i) = {ToA _n (i), AoA _n (i), AoD _n (i ), P _n (i)}. Each element in the propagation data x _n (i) is the propagation delay, arrival azimuth angle, emission angle, and power of the elementary wave n at the measurement point i. Cluster assumed data cluster c in particles _{U p x p, c (i} ) = is denoted as _{{ToA p, c (i)} , AoA p, c (i), AoD p, c (i)}. Each element in the assumed cluster data x _{p, c} (i) is the propagation delay, arrival azimuth angle, and emission angle of the elementary wave n and the cluster c at the measurement point i. A method of calculating the cluster assumption data x _{p, c} (i) will be described later.

<Distance calculation unit 110 _p >
The distance calculation unit 110 _p receives N × I pieces of propagation data x _n (i) and C × I pieces of cluster assumed data x _{p, c} (i) corresponding to all measurement data, and each piece of propagation data x _n. N × C × I distances D _{n, p, c} (i) between (i) and each cluster assumption data x _{p , c} (i) are calculated (s1) and output to the clustering unit 120 _p . In this embodiment, MCD (Multipath Component Distance) is used as the distance D _{n, p, c} (i). MCD _{n, p, c} (i) is defined as follows.

However, _δToA , _δAoA, and _δToD are parameters for normalization (normalization) of propagation delay, arrival azimuth angle, and emission angle, respectively.

<Clustering unit 120 _p >
The clustering unit 120 _p receives N × C × I MCD _{n, p, c} (i), and most MCD _{n, p, c} (i) for each elementary wave n and measurement point i as shown in the following equation. ) Is associated with other clusters (s2), and N × I MCD _{n, p} (i) is output to the evaluation value calculation unit 130 _p .

Further, N × I MCD _{n, p} (i) and particle _Up are associated and stored in a storage unit (not shown). The smallest MCD indicates that the propagation data and the cluster assumed data are closest. Therefore, clustering is performed by associating other cluster assumption data with the closest cluster assumption data to reduce the estimation error.

<Evaluation Value Calculation Unit 130 _p >
The evaluation value calculation unit 130 _p receives N × I MCD _{n, p} (i) and N × I propagation data x _n (i) power _pn (i) as in the following equation: , the sum of the power occupying the sum of electric power of the propagation data _x n for all the measurement points i and all rays n _{(i), MCD n,} propagation p (i) is equal to or less than the threshold _{MCD th} data _x n (i) the proportion of, calculated as the evaluation value _{E p} of the particles _{U p} (s3), and outputs to the judging unit 170 and the resampling unit 140.

<Determining unit 170>
Determination unit 170 receives the P-number of evaluation values E _p, the largest value among the evaluation values determined whether the threshold E _th or (s4), in the case of more than the threshold value E _th is the particle U _The estimated scatterer position included in _p is output as a final estimated value U, and a control signal is output so as to end the processing in each unit. If it is less than the threshold E _th , the determination unit 170 outputs a control signal so as to repeat the processing to each unit.

<Resampling unit 140>
Resampler 140 receives the P-number of evaluation values E _p, performs (duplicate samples in this embodiment) resampling in accordance with the value, determined P number of particles U _p (s4), respectively estimation target data determination part Output to 150 _p . For example, P particles U _p are selected with a probability proportional to the evaluation value E _p . In the selection, it is permitted to select the same particle a plurality of times.

<Estimation target data determination unit 150 _p >
Estimation target data determination part 0.99 _p receives the particles _{U p} obtained by resampling, taken the particles _U N × associated with _p I pieces of _{MCD n,} from the storage unit not shown p (i). The estimation target data determination unit 150 _p selects the largest MCD _{n, p} (i) from N × I MCD _{n, p} (i) as shown in the following equation, and selects one corresponding to this. estimating the propagation data target data _{x n_tp (i_t} p) (where subscript tp represents a _{t p)} is determined as (s6), and outputs the scatterer position estimating unit 160 _p.

This means that the propagation data corresponding to the largest MCD is different from any cluster assumption data, so this propagation data is considered to be propagation data based on a scatterer that is not assumed. Estimated.

In the estimated initial value (C = 0), the following equation shall be the most selective power large measurement data as scatterer estimation target data _{x n_tp} (i_t _p).

The estimation target data determination unit 150 _p increments (increases by one) the number C of clusters as in the following equation after the above processing.
C = C + 1
The _{reason why} there are no P evaluation values E _p that exceed the threshold E _th is that the number of assumed scatterers differs from the actual number of scatterers. This is because the number is considered to be smaller than the actual number of scatterers. The scatterer position estimation unit 160 _p assumes cluster assumption data relating to a newly provided C-th cluster.

<Scatterer position estimation unit 160 _p >
Scatterer position estimating unit 160 _p receives the estimated target data _x n_tp _{(i_t p),} using this value, a new position of the scatterers and estimates, I new cluster C corresponding to the estimated scatterer Suppose cluster assumed data x _{p, C} (i) (s7), and output to the distance calculation unit 110 _p together with (C-1) × I assumed cluster data obtained up to the previous iteration process. To do. The distance calculation unit 110 _p stores (C−1) × I cluster assumption data obtained up to the previous repetitive processing, and the scatterer position estimation unit 160 _p stores the I of the new cluster C. Only one cluster assumed data x _{p, C} (i) may be output.

It should be noted that the processing is repeated in each unit until the determination unit 170 determines that the largest value among the evaluation values is equal to or greater than the threshold value _Eth .

Details of processing in the scatterer position estimation unit 160 _p will be described with reference to FIGS. 3, 5, and 6.

The particle U _p includes C cluster parameters μ _{p, c} and is expressed as U _p = {μ _{p, 1} , μ _{p, 2} ,..., Μ _{p, C} }. One cluster parameter μ _{p, c} corresponds to one scatterer and is expressed as μ _{p, c} = {(x _{p, c} , y _{p, c} ), Delay _{p, c} , AoD _{p, c} }, Each element is the position of the scatterer, the propagation delay from the transmitter to the scatterer, and the outgoing angle of the elementary wave. Let {x _r (i_t _p ), y _r (i_t _p )} be the relative coordinates of the measurement point i_t _p (receiver) with respect to the transmitter. It is assumed that the scatterer position estimation apparatus 100 can acquire relative coordinates in advance.

The scatterer position estimation unit 160 _p includes a scatterer existence range estimation unit 161, a scatterer arrangement unit 163, and a cluster parameter calculation unit 165 (see FIG. 3).

(Scatterer existence range estimation unit 161)
Scatterer existence range estimation unit 161 receives the estimated target data _x n_tp _{(i_t p),} is present which may be the range of the scattering bodies (hereinafter referred to as "scatterer existing range"), and focus the transmitter and receiver ellipse range, and is limited within the range defined by the angle of arrival of the propagation data for estimation target data _{x n_tp (i_t p) (s71} ).

First, the estimated target data _x N_tp propagation delay _ToA N_tp in (i_t _p) (i_t _p) and by using the high-speed Vc, corresponding to the propagation path length _L n_tp _{(i_t p)} of rays = ToA n_tp _{(i_t} p) × Vc
Calculate as The relative coordinates of the measurement points i_t _p for the transmitter is _{{x r (i_t p),} y r (i_t p)}, the distance between the transmitter and the receiver (measuring point),
L _{tx, rx} (i_t _p ) = √ {x _r (i_t _p ) ² + y _r (i_t _p ) ² }
_Therefore , the coordinates of the scatterer are limited to an ellipse in which the sum from the transmitter and the receiver (measurement point) is the propagation path length L _{n_tp of the} elementary wave. In other words, it is limited to an ellipse with the transmitter and receiver (measurement points) as the focal points.

In this embodiment then, the coordinates of the scatterer, using AoA _AoA N_tp estimation target data _{x n_tp (i_t p) (i_t} p), AoA n_tp (i_t p) ± f (0, ρ AoA) range Limited to within. However, f (0, ρ _AoA ) is a normal random number having an average of 0 and a standard deviation ρ _AoA . ρ _AoA is a parameter. Thus, by limiting the scatterer existence range, the scatterer position can be estimated more efficiently.

(Scatterer arrangement part 163)
The scatterer arrangement unit 163 newly arranges a scatterer with a uniform probability within the scatterer existence range (s72), determines the scatterer position {xp _{, C} , yp _{, C} }, and calculates cluster parameters. To the unit 165.

(Cluster parameter calculation unit 165)
Cluster parameter calculation unit 165, the scatterer locations _{{x p, C, y p} , C} receive scatterer position _{{x p, C, y p} , C} and the position _{x r (i_t _p receiver ), Y _r (i_t _p )}, the cluster parameters μ _{p, C} = {(x _{p, C} , y _{p, C} ), Delay _{p, C} , of the new cluster C as follows: Calculate AoD _{p, C} }.

However, L _{rx, s} (i_t _p ) is a distance between the receiver and the scatterer.

Next, the cluster parameter calculation unit 165 uses the new cluster parameters μ _{p, C} and uses the new cluster parameters μ _{p, C,} as shown in the following equation, I cluster assumed data x _{p, C} (i) = {ToA _{p , C} (i), AoA _{p, C} (i), AoD _{p, C} (i)} are calculated (s73).

<Other variations>
The present invention is not limited to the above-described embodiments and modifications. For example, the various processes described above are not only executed in time series according to the description, but may also be executed in parallel or individually as required by the processing capability of the apparatus that executes the processes. In addition, it can change suitably in the range which does not deviate from the meaning of this invention.

<Program and recording medium>
The scatterer estimation device described above can also be operated by a computer. In this case, each process of a program for causing a computer to function as a target device (a device having the functional configuration shown in the drawings in various embodiments) or a process procedure (shown in each embodiment) is processed by the computer. A program to be executed by the computer may be downloaded from a recording medium such as a CD-ROM, a magnetic disk, or a semiconductor storage device or via a communication line into the computer, and the program may be executed.

DESCRIPTION OF SYMBOLS 100 Scatterer position estimation apparatus 110 _p distance calculation part 120 _p clustering part 130 _p evaluation value calculation part 140 resampling part 150 _p estimation object data determination part 160 _p scatterer position estimation part 161 Scatterer existence range estimation part 163 Scatterer arrangement Unit 165 cluster parameter calculation unit 170 determination unit

Claims (7)

For each particle, calculate the distance between the propagation data obtained from the radio wave measurement data received by the receiver and the assumed cluster data that is the assumed propagation data of the cluster using the estimated position of the scatterer A distance calculator;
For each particle, a clustering unit that associates another cluster with a cluster having the shortest distance;
For each particle, an evaluation value calculation unit that calculates, as an evaluation value, the ratio of the sum of power of propagation data whose distance is equal to or less than a threshold value in the sum of power of propagation data of all measurement points and all the elementary waves;
Re-sampling according to the evaluation value to obtain particles; and
For each particle obtained by resampling, an estimation target data determination unit that determines propagation data with the longest distance as estimation target data;
A scatterer position estimation unit that estimates the position of the scatterer newly using the estimation target data for each particle, and assumes cluster assumption data of a new cluster corresponding to the estimated scatterer,
Scatterer position estimation device. The scatterer position estimation apparatus according to claim 1,
The scatterer position estimation unit is
Including a scatterer existence range estimation unit that limits a scatterer existence range within an ellipse that focuses on a transmitter and a receiver and within a range defined by an arrival angle of propagation data with respect to estimation target data;
Scatterer position estimation device. The scatterer position estimation device according to claim 1 or 2,
The scatterer position estimation unit is
A scatterer arrangement part for arranging a scatterer with a uniform probability within the scatterer existence range;
Based on the relationship between the scatterer position obtained by the scatterer placement unit and the position of the receiver, a cluster parameter of a new cluster is calculated, and cluster assumption data at other measurement points is calculated using the cluster parameter. A cluster parameter calculation unit to
Increase the number of clusters until the largest value among the evaluation values obtained in the evaluation value calculation unit is equal to or greater than a threshold value, and repeat the processing in each unit.
Scatterer position estimation device. For each particle, calculate the distance between the propagation data obtained from the radio wave measurement data received by the receiver and the assumed cluster data that is the assumed propagation data of the cluster using the estimated position of the scatterer A distance calculation step;
A clustering step of associating other clusters with the smallest distance cluster for each particle;
For each particle, an evaluation value calculation step for calculating, as an evaluation value, a ratio of the sum of power of propagation data in which the distance is equal to or less than a threshold value in the sum of power of propagation data of all measurement points and all elementary waves;
Resampling according to the evaluation value to obtain a particle;
For each particle obtained by resampling, an estimation target data determination step for determining propagation data with the longest distance as estimation target data;
For each particle, the estimation target data is used to newly estimate the position of the scatterer, and a scatterer position estimation step that assumes cluster assumption data of a new cluster corresponding to the estimated scatterer.
Scatterer position estimation method. The scatterer position estimation method according to claim 4,
The scatterer position estimating step includes:
A scatterer presence range estimation step for limiting the scatterer presence range within an ellipse focused on the transmitter and the receiver and within a range defined by an arrival angle of propagation data with respect to estimation target data,
Scatterer position estimation method. The scatterer position estimation method according to claim 4 or 5,
The scatterer position estimating step includes:
A scatterer placement step of placing a scatterer with a uniform probability within the scatterer presence range;
Based on the relationship between the position of the scatterer obtained in the scatterer placement step and the position of the receiver, a cluster parameter of a new cluster is calculated, and cluster assumption data at other measurement points is calculated using the cluster parameter. A cluster parameter calculation step to
Increase the number of clusters until the largest value among the evaluation values obtained in the evaluation value calculation step is equal to or greater than a threshold value, and repeat the processing in each step.
Scatterer position estimation method.   The program for functioning a computer as a scatterer position estimation apparatus in any one of Claims 1-3.

Images