JP2003318792A - Apparatus for estimating arrival wave direction, and receiver provided with the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To shorten a processing time by reducing an operation amount, without deteriorating the arrival wave direction estimation accuracy. <P>SOLUTION: A/D conversion sections 103-1 to 103-3 sample analog data with a first cycle to convert them into digital data. Down-sampling sections 104-1 to 104-3 perform sampling with a second cycle longer than the first cycle. A relative matrix operating section 105 determines relative matrices, by using plural data items which are sampled by the sections 104-1 to 104-3. A relative matrix memory 106 stores relative matrices, every time the relative matrices are calculated. A relative matrix averaging section 107 decides the number of relative matrices to be used for averaging, stored in the memory 106 and a 'forgetting' factor β to determine an average relative matrix. An arrival wave direction estimating section 109 estimates an arrival wave direction by using the average relative matrix. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an incoming wave direction estimating apparatus and a receiving apparatus including the incoming wave direction estimating apparatus.

[0002]

2. Description of the Related Art A conventional direction-of-arrival estimating apparatus samples a received signal at a predetermined cycle, obtains a correlation matrix using a plurality of sampled data, and estimates the direction of arrival using the correlation matrix. ing. In such a direction-of-arrival estimation apparatus, in order to improve the accuracy of the direction-of-arrival estimation, it is necessary to increase the number of samplings and increase the number of data when obtaining the correlation matrix.

[0003]

However, the conventional arriving wave direction estimating device has a problem that the amount of calculation increases and the processing time increases when the number of data used in obtaining the correlation matrix is increased.

The present invention has been made in view of the above point, and an arrival wave direction estimation device and an arrival wave direction estimation device capable of shortening the processing time by reducing the calculation amount without reducing the accuracy of the arrival wave estimation. An object of the present invention is to provide a receiving device including the device.

[0005]

An arrival wave direction estimating apparatus according to the present invention comprises an analog / digital conversion means for sampling a received arrival wave at a first cycle and converting the analog signal into a digital signal. Sampling means for sampling a digital signal in a second cycle longer than the first cycle; correlation matrix computing means for obtaining a correlation matrix using a plurality of data sampled by the sampling means; The correlation matrix calculated by the correlation matrix calculating means is stored at each calculation timing in the correlation matrix calculating means, and each correlation matrix stored in the correlation matrix storing means is weighted, and each weighted correlation matrix Correlation matrix averaging means for obtaining an averaged correlation matrix using, and estimating the arrival wave direction from the averaged correlation matrix A configuration that includes the incoming wave direction estimating means.

According to this configuration, by using the data sampled at a period longer than the sampling period at the time of analog / digital conversion, the amount of calculation required for one correlation matrix calculation is reduced and the processing time is reduced. By shortening and calculating the averaged correlation matrix by weighting a plurality of stored correlation matrices, it is possible to perform the DOA estimation without lowering the accuracy.

The arrival wave direction estimating apparatus of the present invention has an averaging correlation matrix storing means for storing the averaging correlation matrix obtained by the correlation matrix averaging means, and the correlation matrix averaging means has the averaging means. The subtraction of the correlation matrix initially stored in the correlation matrix storage means from the averaged correlation matrix stored in the averaged correlation matrix storage means, and the averaged correlation matrix stored in the averaged correlation matrix storage means The correlation matrix thus obtained is added to obtain a new averaging matrix.

According to this configuration, when obtaining the averaged correlation matrix from the second time onward, the correlation matrix stored first is subtracted from the averaged correlation matrix obtained first, and
Instead, the averaged correlation matrix is obtained only by adding a newly calculated correlation matrix, so that the amount of calculation can be reduced and the processing time can be shortened as compared with the case where the averaged correlation matrix is obtained from the beginning.

The arrival wave direction estimating device of the present invention has a time variation detecting means for detecting an angular velocity from the change in the direction estimated by the arrival wave direction estimating means, and the arithmetic processing control means comprises:
Depending on the angular velocity detected by the time variation detection means, the number of correlation matrices used by the correlation matrix averaging means is changed and weighting is performed according to the order of storage in the correlation matrix storage means, and the sampling means A configuration is adopted in which the second cycle is changed and sampling is performed in accordance with the angular velocity detected by the time variation detection means.

According to this configuration, for a communication partner with a large detected angular velocity, the sampling period in the sampling means is shortened, and the number of correlation matrices used for averaging is reduced to approximate the current correlation matrix. The greater the weight, the more the average correlation matrix is obtained, and for the communication partner whose detected angular velocity is small, the sampling period in the sampling means is increased, and
Since the average correlation matrix is obtained by increasing the number of correlation matrices used for averaging and increasing the weight as the correlation matrix becomes past, it is possible to accurately estimate the arrival direction according to the position of the communication partner.

An arrival wave direction estimating apparatus of the present invention has a received power value detecting means for detecting a received power value, and the arithmetic processing control means has a function according to the received power value detected by the received power value detecting means. The number of correlation matrices used in the correlation matrix averaging unit and the weighting of the correlation matrix are performed, and the sampling unit changes the period of the second period.

According to this structure, when the received power value is large, the sampling period in the sampling means is increased, and the number of correlation matrices used for averaging is reduced, so that the closer the correlation matrix is to the current one, the more weighting is performed. When the received power value is small, the averaged correlation matrix is increased to reduce the sampling period in the sampling means, and the number of correlation matrices used for averaging is increased to increase the weighting as the correlation matrix becomes past. Since the averaged correlation matrix can be obtained in this way, it is possible to accurately estimate the direction of the incoming wave according to the received power value.

The arrival wave direction estimating apparatus of the present invention has a communication type detecting means for detecting the communication type of the communication partner, and the arithmetic processing control means, in accordance with the communication type detected by the communication type detecting means, Changing the number of correlation matrices used in the correlation matrix averaging means and performing weighting according to the order of storing in the correlation matrix storage means, the sampling means, according to the communication type detected by the communication type detection means, A configuration is adopted in which the second cycle is changed and sampling is performed.

According to this structure, if the communication type used by the communication partner is high speed, the sampling cycle in the sampling means is shortened and the number of correlation matrices used for averaging is increased to obtain a past correlation matrix. Since the weighting is increased as is to obtain the averaged correlation matrix,
The direction of the incoming wave can be accurately estimated according to the communication type of the communication partner.

The receiving apparatus of the present invention has a configuration including any one of the above-mentioned incoming wave direction estimating apparatuses. Further, the transmitting apparatus of the present invention adopts a configuration including any one of the above-mentioned arrival wave direction estimating apparatuses.

According to these configurations, since the correlation matrix is obtained by sampling at a period longer than the sampling period at the time of analog / digital conversion, it is possible to reduce the calculation amount when obtaining the correlation matrix and to reduce the processing time. Can be shortened.

The arrival wave direction estimation method of the present invention includes an analog / digital conversion step of converting an analog signal into a digital signal by sampling the arrival wave at a first cycle, and the first signal for the digital signal. Sampling step in which the sampling is performed in a second cycle that is longer than the cycle, a correlation matrix calculation step that obtains a correlation matrix using a plurality of sampled data, and the obtained correlation matrix is stored at each correlation matrix calculation timing. A correlation matrix storing step,
Weighting each stored correlation matrix, a correlation matrix averaging step of obtaining an averaged correlation matrix using each weighted correlation matrix, and an arrival wave direction estimation step of estimating an arrival wave direction from the averaged correlation matrix, It is to have.

According to this method, by using the data sampled at a period longer than the sampling period at the time of analog / digital conversion, the amount of calculation required for one correlation matrix calculation is reduced and the processing time is reduced. By shortening and calculating the averaged correlation matrix by weighting a plurality of stored correlation matrices, it is possible to perform the DOA estimation without lowering the accuracy.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION The essence of the present invention is to store a correlation matrix in a memory at each operation timing of the correlation matrix and to use the stored correlation matrix for averaging. In addition to controlling the number of correlation matrices used for, the stored correlation matrices are weighted to obtain an averaged correlation matrix, and the direction of arrival is estimated using the obtained averaged correlation matrix.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(Embodiment 1) FIG. 1 is a diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 1 of the present invention. The arriving wave estimation apparatus is provided with antennas 101-1 to 101-
3, frequency converters 102-1 to 102-3, analog /
Digital (hereinafter referred to as "A / D") conversion unit 103-
1 to 103-3, down-sampling units 104-1 to 1
04-3, correlation matrix calculation unit 105, a plurality of memories 106
It mainly includes a correlation matrix memory unit 106 including -1 to 106-n, a correlation matrix averaging unit 107, a matrix calculation unit 108, and an arrival wave direction estimation unit 109.

Antennas 101-1 to 101-3 receive incoming waves from a plurality of paths and output the received signals to frequency converters 102-1 to 102-3. The frequency conversion units 102-1 to 102-3 include antennas 101-1 to 10-3.
The received signal received at 1-3 is converted into an intermediate frequency or a baseband frequency, and A / D conversion units 103-1 to 103-1
Output to 03-3.

The A / D converters 103-1 to 103-3 are
The received signals input from the frequency conversion units 102-1 to 102-3 are converted from analog signals to digital signals and output to the down sampling units 104-1 to 104-3. As shown in FIG. 2, the conversion from the analog signal to the digital signal is performed by a predetermined sampling time interval H which is the first cycle.
Conversion is performed by sampling every one. That is, sampling is performed at times t1, t2, t3, ... And the down-sampling unit 104-
1 to 104-3.

The down-sampling units 104-1 to 104-3, which are sampling means, have a predetermined sampling time interval which is a second cycle so that T predetermined sample data are output in one frame. H2
Is set and sampling is performed, and the data at the set sampling time is output to the correlation matrix calculation unit 105. That is, as shown in FIG.
In 04-1 to 104-3, the A / D conversion unit 103
−1 to 103-3 sampling times t1, t2, t
Sampling times t1, t so that T sampling data can be obtained from the data in 3, ...
Data of 3, t5, ... is thinned out. Then, only the data at the sampling times t2, t4, t6, ... Is selected, and the selected sampling times t2, t4, t
.. are output to the correlation matrix calculation unit 105. In addition, in FIG. 2, the length of each arrow represents the amplitude value, and the rotation angle of the arrow represents the phase.

Correlation matrix calculation section 105 obtains a correlation matrix using T pieces of data input from down-sampling sections 104-1 to 104-3, and outputs the obtained correlation matrix to correlation matrix memory section 106. In correlation matrix calculation section 105, the correlation matrix is obtained based on equation (1).

[0026]

[Equation 1] However, T: Number of sampling data X (t): Received signal H: Conjugate transpose Rxx: Correlation matrix The correlation matrix memory unit 106 which is a correlation matrix storage unit is L
Each of the memories 106-1 to 106-n is provided. The memories 106-1 to 106-n sequentially store the correlation matrix input from the correlation matrix calculation unit 105 at each calculation timing of the correlation matrix in the order of input, and the M memories 106-1 to 106-n.
When all 106-n are stored, the stored correlation matrix is output to the correlation matrix averaging unit 107 from the L (M ≧ L) memories 106-1 to 106-n. Correlation matrix averaging unit 107
Calculates the averaged correlation matrix from the correlation matrix input from the memories 106-1 to 106-n. The memory 1 used when the averaged correlation matrix is obtained by the correlation matrix averaging unit 107
The number of 06-1 to 106-n is arbitrary. The averaged correlation matrix is obtained from equation (2).

[0027]

[Equation 2] Here, β: forgetting coefficient (0 <β ≦ 1 and real number) Rxx (l): latest correlation matrix aveRxx: averaged correlation matrix L: correlation matrix number forgetting coefficient used when obtaining the averaged correlation matrix β is stored in the memories 106-1 to 106-n when obtaining the averaged correlation matrix in the correlation matrix averaging unit 107.
The weighting is applied to each correlation matrix stored in. If the forgetting factor β is made small, the past data will be averaged at a small ratio to the current data and the current data will be weighted, and if the forgetting factor β is made large, the current data and the past data will be Will be the same ratio.

The number of elements of Rxx and the value of L can be set arbitrarily. Also, the total number of sample data used in the averaged correlation matrix is TL. Correlation matrix averaging unit 1
The averaged correlation matrix obtained in 07 is output to the matrix calculation unit 108.

The matrix calculation unit 108 is the correlation matrix averaging unit 1
Matrix calculation processing is performed on the averaged correlation matrix input from 07 using any of various known methods, and the calculation result is output to the arrival wave direction estimation unit 109. For example, the calculation in the matrix calculation unit 108 is the inverse matrix calculation of the correlation matrix in the direction estimation by the Capon method, and
In the direction estimation by the SIC method or ESPRIT method, the eigenvalue / eigenvector decomposition of the correlation matrix is performed. The arrival wave direction estimation unit 109 estimates the arrival direction from the calculation result input from the matrix calculation unit 108. For example, in the Capon method and the MUSIC method, the evaluation function is calculated using the operation result of the matrix operation unit 108 and the arrival direction is estimated by the peak search. In the ESPRIT method, the ESPRIT algorithm is also calculated using the operation result. There is a calculation of.

As described above, according to the arrival wave direction estimation apparatus of the present embodiment, the downsampling units 104-1 to 104-1 are provided.
By reducing the number of samplings in 04-3, the number of data used when the correlation matrix calculation unit 105 obtains the correlation matrix decreases, so that the calculation amount can be reduced and the processing time can be shortened. Further, since the forgetting factor β is used to weight each correlation matrix used when obtaining the averaged correlation matrix to obtain the averaged correlation matrix, the correlation matrix calculation unit 105 performs the calculation. Even if the number of data used in is reduced, the accuracy of incoming wave estimation does not deteriorate.

In the present embodiment, the down-sampling units 104-1 to 104-3 have equal sampling intervals, but similar effects can be obtained even if the sampling intervals are unequal. At this time, within one frame,
By performing the downsampling so that the first half of the frame is small and the second half is large, it is possible to increase the number of latest data with respect to the entire data in the frame. Also,
When estimating the arrival directions of a plurality of users, a method of sequentially processing a plurality of users to perform serial processing in time, or a correlation matrix memory unit 106, a correlation matrix averaging unit 107, a matrix calculation unit 108, and Arrival wave direction estimation unit 10
9 may be provided for the number of users to process a plurality of users in parallel in time. In addition, a despreading unit may be provided and applied to a spread spectrum signal. Further, by having a plurality of despreading units, CDMA (Code
It may be applied to a Division Multiple Access) system. The same effect can be obtained by using a correlation vector instead of the correlation matrix.

(Embodiment 2) FIG. 3 is a diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 2 of the present invention. In this embodiment mode, the configuration in which the memory 301 is provided is different from the configuration in FIG. 1, and other configurations are the same as those in FIG.
Since the configuration is the same as the above, the same reference numerals are given and the description thereof is omitted.

A memory 30 which is an averaged correlation matrix storage means.
The correlation matrix averaging unit 107 stores the averaged correlation matrix obtained by the first calculation in the correlation matrix averaging unit 107.
At the time of the second and subsequent calculations in 7, the stored averaged correlation matrix is output to the correlation matrix averaging unit 107. Then, the calculation results of the second and subsequent times in the correlation matrix averaging unit 107 are sequentially updated and stored. The correlation matrix averaging unit 107, in the first calculation, stores in the memories 106-1 to 10-1.
The averaged correlation matrix is obtained using the correlation matrices up to 06-n, and the obtained averaged correlation matrix is output to the matrix calculation unit 108. During the second and subsequent calculations in the correlation matrix averaging unit 107, an averaged correlation matrix is obtained using the correlation matrix input from the memory 301, the correlation matrix calculating unit 105, and the memory 106-n to the correlation matrix averaging unit 107. . This operation is
The formula (3) is used.

[0034]

[Equation 3] Here, aveRxx (m) is an averaged correlation matrix, and aveRxx (m-1) is the averaged correlation matrix Rxx (1) stored in the memory 301, and the latest correlation matrix Rxx (L) is the memory 106-. The correlation matrix stored in n, that is, from the equation (3), the averaged correlation matrix aveRxx (m−1) calculated from the memory 301 and stored in the correlation matrix memory 106 is stored in the memory 106-n. The averaged correlation matrix aveRxx (m) is obtained by subtracting the correlation matrix Rxx (L) and further adding the latest correlation matrix Rxx (1) obtained by the correlation matrix calculation unit 105. In this case, the stored averaged correlation matrix aveRx
Correlation matrix Rxx (m-1) is set to 1 for x (m-1).
The averaged correlation matrix can be obtained by simply adding one and subtracting one from the correlation matrix Rxx (l). Here, as an example, the number of elements is 6
Consider the number of additions when calculating the averaged correlation matrix. In Embodiment 1, the correlation matrix is a matrix having 6 × 6 elements, and when L = 10, the number of additions is 6 × 6 × 10, except for multiplication by β and division by L. =
It will be 360 additions. On the other hand, in the present embodiment, when subtraction is regarded as equivalent to addition, 6 × 6 × 2 = 72
It will be an addition of times.

As described above, the arrival wave direction estimating apparatus according to the present embodiment has the advantages of
To obtain the averaged correlation matrix after the first time, simply subtract the first stored correlation matrix from the first obtained averaged correlation matrix and add the newly calculated correlation matrix instead. Since the averaged correlation matrix is obtained, the amount of calculation can be reduced and the processing time can be shortened as compared with the case where the averaged correlation matrix is obtained from the beginning.

The memory 301 is assumed to store sequentially from the averaged correlation matrix calculated in the correlation matrix averaging unit 107 for the first time. However, the memory 301 may be sequentially stored from the calculation result after an arbitrary number of times after the second time. May be.

(Embodiment 3) FIG. 4 is a diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 3 of the present invention. In this embodiment, the configuration in which the time variation detection unit 401 and the arithmetic processing control unit 402 are provided is different from the configuration in FIG. 1, and the other configurations are the same as the configurations in FIG. The description is omitted.

The time variation detecting section 401 stores the estimated direction input from the incoming wave direction estimating section 109, detects the temporal variation as an angular velocity from the trajectory of the angle of the estimated direction,
The detection result is output to the arithmetic processing control unit 402. The magnitude of the angular velocity mainly changes depending on the distance to the communication partner and the moving speed of the communication partner. The arithmetic processing control unit 402 has a reference table as shown in FIG. 5 in advance. Figure 5
In the reference table shown in (1), the frame length is 10 ms and the number of data in the frame before down-sampling is 150 bits at the maximum when the direction estimation is performed once for each frame. Angular velocity is 0.1 per frame
If it is less than or equal to degrees, the forgetting factor β is set to a value between 0.95 and 1.00, the number of data T is set to a value between 10 and 30, and the correlation matrix L is set to a value of 10 or more. When the angular velocity is between 0.1 degree and 1.0 degree per frame, the forgetting factor β is set to a value between 0.93 and 0.98 and the number of data T is set. When the control is performed by setting the value between 30 and 50 and the correlation matrix L between 5 and 10, and the angular velocity is 1.0 degree or more per frame,
Set the forgetting factor β to a value between 0.90 and 0.95,
Control is performed to set the number of data T to a value between 50 and 150 and to set the correlation matrix L to a value between 1 and 5. Note that the reference table in FIG. 5 is an example, and a reference table having a setting different from that in FIG. 5 may be used.

In general, these controls are performed so as to improve the followability to the fluctuation when the fluctuation value of the angle in the estimated direction detected by the time fluctuation detecting section 401 is large. That is, by reducing the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 and reducing the value of the forgetting factor β, the influence of past sample data in the arrival wave estimation is reduced. Further, in each of the down-sampling units 104-1 to 104-3, the sampling interval is narrowed to change the sampling period and the number of sample data T is increased, so that the influence of the current sample data in the arrival wave estimation is increased. To do. As a result, the weight of new data in the data used for direction estimation becomes large, and the followability to fluctuations becomes high.

On the other hand, when the time variation in the estimated direction is small, control is performed so as to reduce the calculation amount while ensuring the estimation accuracy. That is, by increasing the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 and increasing the value of the forgetting factor β, the influence of past sample data in the arrival wave estimation is increased. In each of the down-sampling units 104-1 to 104-3, the sampling interval is widened to reduce the sample data number T, thereby reducing the influence of the current sample data on the arrival wave estimation. As a result, it is possible to reduce the calculation amount required for the correlation matrix calculation while suppressing deterioration of the estimation accuracy by using a large amount of past data.

When estimation is performed for a plurality of users, control is performed so that all users have the same estimation accuracy within the limited computing capacity. That is, the communication partner when the time variation in the estimated direction is large reduces the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107,
Further, by reducing the value of the forgetting factor β, the influence of past sample data in the arrival wave estimation is reduced. In addition, each of the downsampling units 104-1 to 104
In -3, by increasing the number of sample data T by narrowing the sampling interval, the influence of the current sample data on the arrival wave estimation is increased. As a result, the followability with respect to the fluctuation is improved.

On the other hand, when the time variation in the estimated direction is small, the communication partner increases the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 and increases the value of the forgetting factor β. Increase the influence of past sample data on incoming wave estimation. In each of the down-sampling units 104-1 to 104-3, the sampling interval is widened to reduce the sample data number T, thereby reducing the influence of the current sample data on the arrival wave estimation. As a result, it is possible to reduce the calculation amount required for the correlation matrix calculation while suppressing deterioration of the estimation accuracy by using a large amount of past data. In particular, the sample data number T affects the calculation amount and the estimation accuracy in the correlation matrix calculation unit 105, so that all users can obtain the same estimation accuracy and maximize the calculation ability. , The sample data number T is controlled for each user.

Thus, in addition to the effects of the first embodiment, the arrival wave direction estimation apparatus according to the present embodiment detects time fluctuations in the estimated direction as angular velocities in time fluctuation detection section 401, and detects the fluctuations. Since the correlation matrix calculation is performed by controlling the number L of the correlation matrix to be averaged according to the size, the forgetting factor β, and the sample data number T, it is possible to suppress the deterioration of the estimation accuracy due to the deterioration of the followability with respect to time fluctuation.
In addition, for a communication partner whose detected angular velocity is small and whose fluctuations are small, the influence of past sample data in arriving wave estimation is reduced and the influence of the present sample data is increased, and the communication whose detected angular velocity is large and fluctuates greatly. For the other party, the influence of the past sampling data in the arrival wave estimation is increased and the influence of the current sample data is reduced. It can be estimated with estimation accuracy.

(Embodiment 4) FIG. 6 is a diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 4 of the present invention. In this embodiment, the configuration in which the power value detection unit 601 and the arithmetic processing control unit 602 are provided is different from the configuration in FIG. 1, and the other configurations are the same configurations as in FIG. The description will be omitted.

The estimation accuracy in the arrival wave direction estimation technique is
It depends on the signal-to-noise power ratio of the received signal, and the estimation accuracy improves when the signal-to-noise power ratio is high. For this reason, the correlation matrix calculation realizes a high signal-to-noise power ratio state by using a large amount of data.

The power value detecting section 601 includes the A / D converting sections 1
The received power value is detected from the digital signals output from 03-1 to 103-3, and the detected received power value is output to the arithmetic processing control unit 602. The arithmetic processing control unit 602 compares the received power value that is the output of the power value detection unit 601 with a preset threshold value, and the sampling interval and correlation matrix average in each of the downsampling units 104-1 to 104-3. The number L of correlation matrices and the forgetting factor β for performing averaging in the averaging unit 107 are controlled. The control method is, for example, as follows. Power value detection unit 601
From the received power value detected in
02, when the signal-to-noise power ratio is expected to increase, that is, when the received power value is equal to or greater than the threshold value, the estimation accuracy can be secured without using a large amount of data in the correlation matrix calculation. Therefore, control is performed so as to reduce the calculation amount required for the correlation matrix calculation. This control is
In each of the downsampling units 104-1 to 104-3, the number of sample data T is reduced by increasing the sampling interval. Thereby, the correlation matrix calculation unit 10
The calculation amount in 5 is reduced. Further, by reducing the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 and reducing the value of the forgetting factor β, the estimation accuracy can be improved by increasing the weight of new data having a large received power value. Improve.

On the other hand, when the signal-to-noise power ratio is expected to be small, that is, when the received power value is less than the threshold value, control is performed so as to improve the estimation accuracy. This control is performed by each of the downsampling units 104-1 to 104.
At -3, the number of sample data T is increased by narrowing the sampling interval. This increases the amount of calculation in the correlation matrix calculation unit 105, but improves the estimation accuracy by using a large amount of data. Further, the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 is increased,
Further, by increasing the value of the forgetting factor β, the estimation accuracy is improved by using a lot of past data.

Further, in the case of performing estimation for a plurality of users, control is performed so that all users have the same estimation accuracy within the limited computing capacity. In the case of a user determined to have a large received power value, the estimation accuracy can be secured without using a large amount of data in the correlation matrix calculation, so control is performed to reduce the calculation amount required for the correlation matrix calculation. That is, each downsampling unit 104-1
In 104-3, the sampling interval is widened to reduce the sample data number T. As a result, the calculation amount in the correlation matrix calculation unit 105 is reduced. Further, by reducing the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 and reducing the value of the forgetting factor β, the estimation accuracy can be improved by increasing the weight of new data having a large received power value. Improve.

On the other hand, the user determined to have a small received power value is the downsampling units 104-1 to 104-.
In No. 3, the number of sample data T is increased by narrowing the sampling interval. Thereby, the correlation matrix calculation unit 1
Although the calculation amount in 05 increases, the estimation accuracy improves by using a large amount of data. Further, by increasing the number L of correlation matrices to be averaged in the correlation matrix averaging unit 107 and increasing the value of the forgetting factor β, the estimation accuracy is improved by using a large amount of past data. In particular, the sample data number T affects the calculation amount and the estimation accuracy in the correlation matrix calculation unit 105, so that all users can obtain the same estimation accuracy and maximize the calculation ability. , The sample data number T is controlled for each user.

Thus, in addition to the effects of the first embodiment, the arrival direction estimating apparatus of the present embodiment detects the received signal power in power value detection section 601 and determines the magnitude of the received power value. Thus, by controlling the correlation matrix calculation, the calculation amount can be reduced while maintaining the estimation accuracy. Further, when there are a plurality of users, it is possible to make an estimation with the same estimation system for all users by making the most of the computing power.

Although the received power value is detected from the output digital signals of the A / D converters 103-1 to 103-3 in the present embodiment, each array antenna 10 is detected.
Outputs of 1-1 to 101-3 or each frequency conversion unit 10
The same effect can be obtained by using the analog signals output from 2-1 to 102-3.

(Embodiment 5) FIG. 7 is a diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 5 of the present invention. In the present embodiment, the user information detection unit 701
1 is different from the configuration of FIG. 1 in that the arithmetic processing control unit 702 is provided, and the other configurations are the same as those in FIG.

The user information detection unit 701 detects the communication type of the communication partner from the digital signals output from the A / D conversion units 103-1 to 103-3, and the arithmetic processing control unit 904.
Output to. The arithmetic processing control unit 702 determines the sampling interval in each of the downsampling units 104-1 to 104-3 and the averaging in the correlation matrix averaging unit 107 according to the communication type that is the detection result input from the user information detecting unit 701. The number L of correlation matrices to be performed and the forgetting factor β are controlled. The arithmetic processing control unit 702 has a reference table as shown in FIG. 8 in advance.

The lookup table shown in FIG. 8 has a frame length of 10 m when the direction is estimated once for each frame.
s, and the maximum number of data in the frame before down-sampling is 150 bits. When the communication service of the communication partner is voice communication, the forgetting factor β is set to a value between 0.93 and 1.00 and the number of data T is 10 and 50.
When the communication service of the communication partner is high-speed data communication, the forgetting factor β is set to 0.95 and 1.00. The control is performed by setting a value between 2 and 5, a data number T between 50 and 150, and a correlation matrix number L between 5 and 10. Note that the reference table in FIG. 8 is an example, and a reference table having a setting different from that in FIG. 8 may be used.

In general, these controls are performed by the arithmetic processing control unit 702 so as to increase the estimation accuracy as the communication speed of the user, which is the user information detected by the user information detection unit 701, increases. That is, for a user of high-speed communication, the number L of correlation matrices to be averaged by the correlation matrix averaging unit 107 is increased, and the value of the forgetting factor β is increased. In addition, each downsampling unit 1
In 04-1 to 104-3, the number of sample data T is increased by narrowing the sampling interval. This allows
Since the number of data used for the direction estimation increases, the estimation accuracy increases.

Further, when the estimation is performed for a plurality of users, the arithmetic processing control unit 702 has the user information detection unit 70.
As the communication speed of the user, which is the information for each user detected in 1, increases, the control is performed so as to increase the estimation accuracy. That is, for a user of high-speed communication, the number L of correlation matrices to be averaged by the correlation matrix averaging unit 107 is increased, and the value of the forgetting factor β is increased. In each of the downsampling units 104-1 to 104-3, the number of sample data T is increased by narrowing the sampling interval. As a result, the number of data used for the direction estimation increases, and the estimation accuracy increases.

As described above, according to the arrival wave estimating apparatus of the present embodiment, in addition to the effects of the first embodiment, the correlation matrix calculation is performed according to the communication service of the communication partner detected by the user information detecting unit 701. It is possible to secure the estimation accuracy according to the state of the user by controlling the. Further, when there are a plurality of users, it is possible to realize the estimation according to the communication states of all the users by making the most of the computing power.

In this embodiment, the forgetting factor β, the number of data T, and the number of correlation matrices L are controlled according to the communication type of the communication partner, but communication with the communication partner is performed instead of the communication type. The control may be performed using other information such as the communication speed in.

(Embodiment 6) FIG. 9 is a diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 6 of the present invention. In this embodiment, the configuration in which the power value detection unit 901, the user information detection unit 902, the time variation detection unit 903, and the arithmetic processing control unit 904 are provided is different from the configuration in FIG. 1, and other configurations are the same as those in FIG. Since it has the same configuration as the above, the same reference numerals are given and the description thereof is omitted.

The power value detecting section 901 is provided for each A / D converting section 1
The reception power value is detected from the digital signals output from 03-1 to 103-3, and the detected reception power value is output to the arithmetic processing control unit 904. The user information detection unit 902 determines each A
The contents of the communication service are detected from the digital signals output from the / D conversion units 103-1 to 103-3 and output to the arithmetic processing control unit 904. The time variation detection unit 903 stores the estimated direction input from the arrival direction estimation unit 109, detects an angular velocity as a temporal variation from the locus, and outputs the detection result to the arithmetic processing control unit 904. Arithmetic processing control unit 9
04 is the received power value input from the power value detection unit 901,
Using the content of the communication service input from the user information detection unit 902 and the angular velocity input from the time variation detection unit 903, the forgetting factor β and the number of data T in the correlation matrix averaging unit 107, and the downsampling units 104-1 to 104-104.
-3 controls the sampling period. In this case, the priority order of which input is to be adopted is set for the inputs from the arithmetic processing control unit 904, the power value detection unit 901, the user information detection unit 902, and the time variation detection unit 903, and the provided priority order. Control according to. The priority order can be arbitrarily changed according to the communication environment at that time.

Incidentally, the above-mentioned first to sixth embodiments
The arriving wave direction estimating device up to the above can be applied to the receiving device and the transmitting device.

[0062]

As described above, according to the present invention,
The processing time can be shortened by reducing the calculation amount without lowering the accuracy of arrival wave estimation.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an arrival wave direction estimation device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing sampling according to the first embodiment of the present invention.

FIG. 3 is a block diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 2 of the present invention.

FIG. 4 is a block diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 3 of the present invention.

FIG. 5 is a reference table according to the third embodiment of the present invention.

FIG. 6 is a block diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 4 of the present invention.

FIG. 7 is a block diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 5 of the present invention.

FIG. 8 is a reference table according to the fifth embodiment of the present invention.

FIG. 9 is a block diagram showing the configuration of an arrival wave direction estimation apparatus according to Embodiment 6 of the present invention.

[Explanation of symbols]

104-1 to 104-3 Down-sampling unit 105 correlation matrix calculator 106 correlation matrix memory unit 107 correlation matrix averaging unit 109 Direction of Arrival Estimation Unit 401, 903 Time fluctuation detector 402, 602, 702, 904 Arithmetic processing control unit 601, 901 Power value detection unit 701, 902 User information detection unit

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Takashi Fukagawa             1006 Kadoma, Kadoma-shi, Osaka Matsushita Electric             Sangyo Co., Ltd. (72) Inventor Masayuki Hoshino             3-1, Tsunashima-Higashi 4-chome, Kohoku-ku, Yokohama-shi, Kanagawa             Matsushita Communication Industry Co., Ltd. F term (reference) 5J021 AA05 AA06 CA06 DB02 DB03                       EA04 FA13 FA14 FA15 FA16                       FA17 FA20 FA26 FA29 FA30                       FA32 GA02 HA05 JA10                 5K059 CC04 DD02 DD10

Claims (8)

[Claims] 1. An analog / digital conversion means for sampling a received incoming wave at a first cycle and converting an analog signal into a digital signal, and a first signal longer than the first cycle for the digital signal. Sampling means for sampling in two cycles, correlation matrix calculating means for obtaining a correlation matrix using a plurality of data sampled by the sampling means, and correlation matrix obtained by the correlation matrix calculating means for the correlation matrix Correlation matrix storage means for storing each calculation timing in the calculation means, and correlation matrix averaging means for weighting each correlation matrix stored in the correlation matrix storage means and obtaining an averaged correlation matrix using each weighted correlation matrix And an arrival wave direction estimating means for estimating the arrival wave direction from the averaged correlation matrix. Arrival direction estimating device. 2. An averaged correlation matrix storage means for storing the averaged correlation matrix obtained by the correlation matrix averaging means,
The correlation matrix averaging means subtracts the correlation matrix initially stored in the correlation matrix storage means from the averaged correlation matrix stored in the averaged correlation matrix storage means, and also stores the average stored in the averaged correlation matrix storage means. The arrival wave direction estimating apparatus according to claim 1, wherein a new averaging matrix is obtained by adding the correlation matrix obtained by the correlation matrix calculating means to the generalized correlation matrix. 3. A time variation detecting means for detecting an angular velocity from the change in the direction estimated by the incoming wave direction estimating means, and the arithmetic processing control means, in accordance with the angular velocity detected by the time variation detecting means, The number of correlation matrices used by the correlation matrix averaging means is changed, and weighting is performed according to the order of storing in the correlation matrix storage means, and the sampling means, in accordance with the angular velocity detected by the time variation detection means, The arrival wave direction estimating apparatus according to claim 1 or 2, wherein the second period is changed to perform sampling. 4. A reception power value detecting means for detecting a reception power value is provided, and the arithmetic processing control means uses the correlation matrix averaging means in accordance with the reception power value detected by the reception power value detecting means. The arrival wave direction estimation apparatus according to claim 1 or 2, wherein the number of correlation matrices and the weighting of the correlation matrix are performed, and the sampling means changes the period of the second period. 5. A communication type detecting means for detecting a communication type of a communication partner, wherein the arithmetic processing control means uses the correlation matrix averaging means according to the communication type detected by the communication type detecting means. The number of correlation matrices is changed and weighting is performed according to the order of storage in the correlation matrix storage unit, and the sampling unit changes the second cycle according to the communication type detected by the communication type detection unit. The incoming wave direction estimation device according to any one of claims 1 to 4, wherein the sampling is performed by sampling. 6. A receiving apparatus comprising the arrival wave direction estimating apparatus according to claim 1. Description: 7. A transmission apparatus comprising the arrival wave direction estimation apparatus according to any one of claims 1 to 5. 8. An analog / digital conversion step of sampling an incoming wave at a first cycle to convert an analog signal into a digital signal, and a second step longer than the first cycle for the digital signal. A sampling step of sampling in a cycle, a correlation matrix calculation step of obtaining a correlation matrix using a plurality of sampled data,
A correlation matrix storing step of storing the obtained correlation matrix at each operation timing of the correlation matrix, and a correlation matrix averaging step of weighting each stored correlation matrix and obtaining an averaged correlation matrix using each weighted correlation matrix, And an arrival wave direction estimation step of estimating the arrival wave direction from the averaged correlation matrix.